JP7623174B2 - Film thickness estimation method, film thickness estimation device, and etching method - Google Patents

Description

This invention relates to a film thickness estimation technique for estimating the film thickness of the top layer of a substrate having a multilayer structure in which at least one thin film layer is laminated on a base layer after processing the substrate, and an etching method using the film thickness estimation technique. Here, the substrates include semiconductor wafers, glass substrates for liquid crystal display devices, glass substrates for plasma displays, substrates for optical disks, substrates for magnetic disks, substrates for magneto-optical disks, glass substrates for photomasks, substrates for solar cells, etc. (hereinafter simply referred to as "substrates").

Known techniques for measuring the thickness of the top layer formed on a substrate such as a semiconductor wafer include a stylus-type surface profile measurement method, a method using terahertz waves (see Patent Document 1), and a method using a spectroscopic ellipsometer or an interference film thickness meter (see Patent Document 2).

Patent No. 6589239 JP 2001-41713 A

When attempting to apply such conventional film thickness measurement technology to a substrate processing apparatus that processes substrates, such as an etching apparatus, the following problems arise. Stylus-type surface shape measurement causes inconveniences such as contamination and damage to the substrate. In addition, methods that use terahertz waves require the incorporation of an expensive film thickness measurement device into the etching apparatus, which increases the cost of the etching apparatus. Furthermore, when using a spectroscopic ellipsometer or interference film thickness gauge as a film thickness measurement device, it is necessary to equip it with spectroscopic optical components such as an acousto-optical spectroscopic filter. This not only results in an expensive device, but also makes it difficult to measure film thickness in real time, as the wavelength of the observation light must be scanned within a certain band.

This invention has been made in consideration of the above problems, and aims to provide a film thickness estimation technology that can determine the film thickness of the top layer of a substrate having a multi-layer structure at low cost and in real time, and an etching method that uses the film thickness estimation technology.

The first aspect of the present invention is a film thickness estimation method for estimating the film thickness of the top layer of a substrate having a multilayer structure in which at least one thin film layer is laminated on a base layer after the substrate has been treated, and is characterized by comprising the steps of: (a) irradiating the treated substrate with observation light of wavelength λ to obtain an image of the top layer of the substrate with an imaging unit; (b) storing the refractive index and extinction coefficient of each layer constituting the multilayer structure at wavelength λ and the film thickness of each layer of the multilayer structure except for the top layer; and (c) estimating the film thickness of the top layer from the luminance value of the pixel constituting the image of the top layer based on the refractive index, extinction coefficient, and film thickness stored in step (b) and the inverse function of a function (R(λ)=f(d1)) showing the relationship between the reflectance R(λ) of the substrate at wavelength λ and the film thickness d1 of the top layer.

The second aspect of the present invention is a film thickness estimation device that estimates the film thickness of the top layer of a substrate having a multilayer structure in which at least one thin film layer is laminated on a base layer after processing the substrate, and is characterized by comprising: an illumination unit that irradiates the processed substrate with observation light of wavelength λ; an imaging unit that images the substrate irradiated with the observation light to obtain an image of the top layer of the substrate; a memory unit that stores the refractive index and extinction coefficient of each layer that constitutes the multilayer structure at wavelength λ and the film thickness of each layer of the multilayer structure except for the top layer; and a film thickness estimation unit that estimates the film thickness of the top layer from the luminance value of the pixel that constitutes the image of the top layer based on the refractive index, extinction coefficient, and film thickness stored in the memory unit and the inverse function of a function (R(λ)=f(d1)) that indicates the relationship between the reflectance R(λ) of the substrate and the film thickness d1 of the top layer.

The third aspect of the present invention is an etching method for supplying a processing liquid to a substrate having a multilayer structure in which at least one thin film layer is laminated on a base layer, and etching the top layer of the substrate, characterized in that the above-mentioned film thickness estimation method is used to determine the film thickness of the top layer of the substrate being treated with the processing liquid, and the supply of the processing liquid is controlled in accordance with the estimated value.

In the invention configured in this manner, the refractive index and extinction coefficient of each layer constituting the multilayer structure at wavelength λ, and the film thickness of each layer of the multilayer structure except for the top layer are stored in advance. When estimating the film thickness of the top layer of a processed substrate, the substrate is irradiated with observation light of wavelength λ, and an image of the top layer of the substrate is acquired by the imaging unit. Then, based on the previously stored refractive index, extinction coefficient, and film thickness, and the inverse function of a function (R(λ)=f(d1)) showing the relationship between the reflectance R(λ) of the substrate at wavelength λ and the film thickness d1 of the top layer, the film thickness of the top layer is obtained from the luminance value of the pixels constituting the image of the top layer.

According to the invention configured in this way, the film thickness of the topmost layer of a substrate can be obtained from an image of the topmost layer obtained by imaging the substrate illuminated with observation light of a specific wavelength λ. In this way, it is possible to estimate the film thickness of the topmost layer using observation light of a specific wavelength λ without the need for wavelength scanning of the observation light as in conventional technology. As a result, the film thickness of the topmost layer of a substrate having a multi-layer structure can be obtained at low cost and in real time.

1 is a diagram showing an example of a substrate processing apparatus equipped with a film thickness estimation mechanism that is an embodiment of a film thickness estimation device according to the present invention; 2 is a plan view of the substrate processing apparatus shown in FIG. 1 as viewed from above. 3 is a diagram showing a configuration of a film thickness estimation mechanism provided in the substrate processing apparatus shown in FIGS. 1 and 2; FIG. 3 is a flowchart showing an example of a bevel etching operation performed by the substrate processing apparatus shown in FIGS. 1 and 2 . 10 is a flowchart showing an example of a calibration value acquisition operation executed before a bell etching process. 13 is a flowchart showing an operation for estimating a film thickness of an uppermost layer. 11A and 11B are diagrams for explaining the film thickness estimation operation of the uppermost layer; 11A and 11B are diagrams for explaining the film thickness estimation operation of the uppermost layer; 13 is a flowchart showing a process of determining a film thickness candidate value.

Figure 1 shows an example of a substrate processing apparatus equipped with a film thickness estimation mechanism, which is one embodiment of a film thickness estimation device according to the present invention. Figure 2 is a plan view of the substrate processing apparatus shown in Figure 1 from above. Figure 3 shows the configuration of the film thickness estimation mechanism equipped in the substrate processing apparatus shown in Figures 1 and 2. Note that in the figures referred to below, the dimensions and numbers of each part may be exaggerated or simplified for ease of understanding. The up-down direction is the vertical direction, and the substrate side is upper with respect to the spin chuck.

The substrate processing apparatus 1 includes a rotating and holding mechanism 2, a scattering prevention unit 3, a surface protection unit 4, a processing unit 5, a nozzle movement mechanism 6, a heating mechanism 7, a film thickness estimation mechanism 8 corresponding to one embodiment of the film thickness estimation apparatus according to the present invention, and a control unit 10. Each of these units 2 to 8 is electrically connected to the control unit 10 and operates according to instructions from the control unit 10. The control unit 10 may be, for example, the same as a general computer. That is, the control unit 10 includes, for example, a CPU that performs various arithmetic processing, a ROM that is a read-only memory that stores a basic program, a RAM that is a readable and writable memory that stores various information, a magnetic disk that stores control software, data, and the like. In the control unit 10, the CPU as the main control unit performs arithmetic processing according to a procedure described in the program, thereby controlling each unit of the substrate processing apparatus 1.

This substrate processing apparatus 1 is a so-called bevel etching apparatus that performs an etching process on the peripheral portion of a substrate W having a q-layer structure in which at least one thin film layer is stacked on a base layer Lq, as shown in FIG. 3. In the substrate processing apparatus 1, a film thickness estimation mechanism 8 determines the film thickness d1 of the top layer L1 on the peripheral portion of the substrate W at a predetermined sampling period. Then, during the etching process, the control unit 10 detects the point in time when the film thickness d1 determined by the film thickness estimation mechanism 8 becomes zero as an endpoint, and stops the etching process at that detection point. The film thickness sampling and bevel etching operation in the substrate processing apparatus 1 will be described in detail later.

The rotating and holding mechanism 2 is a mechanism capable of rotating the substrate W while holding it in a substantially horizontal position with its surface facing upward. The rotating and holding mechanism 2 rotates the substrate W around a vertical rotation axis a1 passing through the center c1 of the main surface. The rotating and holding mechanism 2 is equipped with a spin chuck ("substrate holding part") 21, which is a disk-shaped member smaller than the substrate W. The spin chuck 21 is provided so that its upper surface is substantially horizontal and its central axis coincides with the rotation axis a1. A cylindrical rotating shaft part 22 is connected to the lower surface of the spin chuck 21. The rotating shaft part 22 extends in the vertical direction with its axis coincident with the rotation axis a1. A rotating drive part (e.g., a motor) 23 is also connected to the rotating shaft part 22. The rotating drive part 23 rotates the rotating shaft part 22 around its axis in response to a rotation command from the control part 10. Therefore, the spin chuck 21 can rotate around the rotation axis a1 together with the rotating shaft portion 22. The rotation drive portion 23 and the rotating shaft portion 22 form a rotation mechanism 231 that rotates the spin chuck 21 around the rotation axis a1. The rotating shaft portion 22 and the rotation drive portion 23 are housed in a cylindrical casing 24.

A through hole (not shown) is provided in the center of the spin chuck 21, and communicates with the internal space of the rotating shaft portion 22. A pump (not shown) is connected to the internal space via piping and an on-off valve (not shown). The pump and on-off valve are electrically connected to the control unit 10. The control unit 10 controls the operation of the pump and the on-off valve. The pump can selectively supply negative pressure or positive pressure under the control of the control unit 10. When the pump supplies negative pressure with the substrate W placed in a substantially horizontal position on the upper surface of the spin chuck 21, the spin chuck 21 adsorbs and holds the substrate W from below. When the pump supplies positive pressure, the substrate W can be removed from the upper surface of the spin chuck 21.

In this configuration, when the spin chuck 21 holds the substrate W by suction, the rotation drive unit 23 rotates the rotation shaft unit 22, causing the spin chuck 21 to rotate about an axis along the vertical direction. As a result, the substrate W held on the spin chuck 21 is rotated in the direction of arrow AR1 around the vertical rotation axis a1 passing through the center c1 in its plane.

The method of holding the substrate W is not limited to this, but may be a so-called mechanical chuck method in which the substrate W is held by multiple chuck pins (e.g., six).

The splash prevention unit 3 catches processing liquid and the like that splashes from the substrate W that is rotated together with the spin chuck 21. The splash prevention unit 3 includes a splash guard 31. The splash guard 31 is a cylindrical member with an open upper end, and is provided to surround the rotating holding mechanism 2. A guard drive mechanism (not shown) that moves the splash guard 31 up and down is connected to the splash guard 31, and is driven in response to a lift command from the control unit 10.

The surface protection unit 4 is provided with a gas discharge mechanism that discharges a flow of inert gas so as to strike the peripheral portion of the upper surface (symbol Wu in FIG. 4) of the substrate W held and rotated on the spin chuck 21. The "inert gas" is a gas that is poorly reactive with the material of the substrate W and the thin film formed on its surface, such as nitrogen (N2) gas, argon gas, helium gas, etc. In this embodiment, gas discharge mechanisms 41 and 42 are provided. The gas discharge mechanisms 41 and 42 discharge the inert gas, for example, as a gas column-shaped gas flow. The gas discharge mechanism 42 discharges the inert gas flow so as to strike a position upstream in the direction of rotation of the substrate W from the position where the gas flow discharged by the gas discharge mechanism 41 strikes the peripheral portion of the substrate W.

The surface protection unit 4 further includes a gas discharge mechanism 43 that discharges a flow of inert gas toward the center of the upper surface of the substrate W that is held and rotated on the spin chuck 21. By discharging a flow of inert gas from the gas discharge mechanisms 41 to 43 toward the upper surface of the substrate W, the surface protection unit 4 protects a non-processing area of the upper surface of the substrate W from processing liquid, etc., that is discharged so as to hit an annular processing area defined by the peripheral portion of the upper surface of the substrate W.

The gas discharge mechanisms 41 and 42 include a nozzle head 44. The gas discharge mechanism 43 includes a nozzle head 45. The nozzle heads 44 and 45 are attached to the tips of arms 61 and 62, respectively, of the nozzle movement mechanism 6, which will be described later. The arms 61 and 62 extend along a horizontal plane. The nozzle movement mechanism 6 moves the arms 61 and 62 to move the nozzle heads 44 and 45 between their respective processing positions (solid line positions in FIG. 2) and retracted positions (double-dashed line positions in FIG. 2).

The nozzle head 44 has two nozzles 46 and 47, and is attached to the tip of the arm 61. The nozzles 46 and 47 have their tips (lower ends) protruding downward from the lower surface of the nozzle head 44, and their upper ends protruding upward from the upper surface. One end of a pipe 411 is connected to the upper end of one of the nozzles 46. The other end of the pipe 411 is connected to a gas supply source 412. In addition, a flow rate controller 413 and an on-off valve 414 are provided in this order from the gas supply source 412 side in the middle of the path of the pipe 411. One end of a pipe 421 is also connected to the other nozzle 47. The other end of the pipe 421 is connected to a gas supply source 422. In addition, a flow rate controller 423 and an on-off valve 424 are provided in this order from the gas supply source 422 side in the middle of the path of the pipe 421.

Here, when the nozzle movement mechanism 6 positions the nozzle head 44 at its processing position, the outlet of the nozzle 46 faces a part of the rotational trajectory of the peripheral edge of the substrate W rotated by the rotating and holding mechanism 2, and the outlet of the nozzle 47 faces another part of the rotational trajectory.

With the nozzle head 44 positioned at the processing position, the nozzles 46, 47 are supplied with an inert gas (nitrogen (N2) gas in the illustrated example) from the gas supply sources 412, 422. The nozzle 46 discharges the supplied inert gas flow from above so that it hits a position defined by the rotation trajectory of the peripheral edge of the substrate W. The nozzle 46 discharges the gas flow from the discharge port in a specified direction so that the discharged gas flow, after reaching the position, flows from the position toward the peripheral edge of the substrate W. The nozzle 47 discharges the supplied inert gas flow from above so that it hits a position defined on the rotation trajectory. The nozzle 47 discharges the gas flow from the discharge port in a specified direction so that the discharged gas flow, after reaching the position, flows from the position toward the peripheral edge of the substrate W.

The nozzle head 45 of the gas discharge mechanism 43 includes a cylindrical member 93 attached to the underside of the tip of the arm 62, a disk-shaped blocking plate 90 attached to the underside of the cylindrical member 93, and a cylindrical nozzle 48. The axis of the cylindrical member 93 and the axis of the blocking plate 90 are aligned vertically. The underside of the blocking plate 90 is aligned horizontally. The nozzle 48 penetrates the cylindrical member 93 and the blocking plate 90 vertically so that its axis is aligned with the axes of the blocking plate 90 and the cylindrical member 93. The upper end of the nozzle 48 also penetrates the tip of the arm 62 and opens on the upper surface of the arm 62. One end of a pipe 431 is connected to the upper opening of the nozzle 48. The other end of the pipe 431 is connected to a gas supply source 432. A flow rate controller 433 and an opening/closing valve 434 are provided in the middle of the pipe 431, in this order from the gas supply source 432 side. The lower end of the nozzle 48 opens into the lower surface of the blocking plate 90. This opening is the discharge port of the nozzle 48.

When the nozzle moving mechanism 6 places the nozzle head 45 at its processing position, the nozzle 48 outlet faces near the center of the upper surface of the substrate W. In this state, the nozzle 48 is supplied with an inert gas (nitrogen (N2) gas in the illustrated example) from a gas supply source 432 via a pipe 431. The nozzle 48 discharges the supplied inert gas as a gas flow of the inert gas toward near the center of the upper surface of the substrate W. The gas flow spreads radially from above the central portion of the substrate W toward the periphery of the substrate W. In other words, the gas discharge mechanism 43 discharges the inert gas from above the central portion of the upper surface of the substrate W to generate a gas flow that spreads from above the central portion toward the periphery of the substrate W.

The processing section 5 processes a processing region on the peripheral portion of the upper surface of the substrate W held on the spin chuck 21. Specifically, the processing section 5 supplies a processing liquid to the processing region of the substrate W held on the spin chuck 21. The processing section 5 includes a processing liquid discharge mechanism 51A. The processing liquid discharge mechanism 51A discharges a liquid flow of the processing liquid so as to hit a part of the peripheral portion of the upper surface (processing surface) of the substrate W held and rotated on the spin chuck 21. The liquid flow is in the form of a liquid column. The processing liquid discharge mechanism 51A includes a nozzle head 50. The nozzle head 50 is attached to the tip of a long arm 63 included in the nozzle movement mechanism 6. The arm 63 extends along a horizontal plane. The nozzle movement mechanism 6 moves the arm 63 to move the nozzle head 50 between its processing position (solid line position in FIG. 2) and a retracted position (two-dot chain line position in FIG. 2).

The nozzle head 50 has four nozzles 5a to 5d and is attached to the tip of the arm 63. The nozzles 5a to 5d are arranged in a row along the extension direction of the arm 63. In the nozzle head 50, the tips (lower ends) of the nozzles 5a to 5d protrude downward, and their base ends (upper ends) protrude upward. The nozzles 5a to 5d are connected to a processing liquid supply unit 51, which is a piping system that supplies processing liquid to the nozzles. Specifically, one end of the piping 53a to 53d of the processing liquid supply unit 51 is connected to the upper ends of the nozzles 5a to 5d. The nozzles 5a to 5d are each supplied with processing liquid from the processing liquid supply unit 51, and each of the nozzles 5a to 5d ejects the supplied processing liquid from an ejection port at the tip. The processing liquid ejection mechanism 51A ejects a flow of processing liquid from one of the nozzles 5a to 5d, which is determined by control information set in the control unit 10, according to the control of the control unit 10.

Specifically, the processing liquid supply unit 51 is configured by combining an SC-1 supply source 52a, a DHF supply source 52b, an SC-2 supply source 52c, a rinsing liquid supply source 52d, a number of pipes 53a, 53b, 53c, 53d, and a number of on-off valves 54a to 54d. SC-1, DHF, and SC-2 are chemical liquids. Therefore, the processing liquid discharge mechanism 51A is a chemical liquid discharge unit that discharges chemical liquid onto the peripheral edge of the substrate W.

The SC-1 supply source 52a is a supply source that supplies SC-1. The SC-1 supply source 52a is connected to the nozzle 5a via a pipe 53a in which an on-off valve 54a is inserted. Therefore, when the on-off valve 54a is opened, the SC-1 supplied from the SC-1 supply source 52a is discharged from the nozzle 5a.

The DHF supply source 52b is a supply source that supplies DHF. The DHF supply source 52b is connected to the nozzle 5b via a pipe 53b in which an on-off valve 54b is inserted. Therefore, when the on-off valve 54b is opened, the DHF supplied from the DHF supply source 52b is discharged from the nozzle 5b.

The SC-2 supply source 52c is a supply source that supplies SC-2. The SC-2 supply source 52c is connected to the nozzle 5c via a pipe 53c in which an on-off valve 54c is inserted. Therefore, when the on-off valve 54c is opened, the SC-2 supplied from the SC-2 supply source 52c is discharged from the nozzle 5c.

The rinse liquid supply source 52d is a supply source that supplies the rinse liquid. Here, the rinse liquid supply source 52d supplies, for example, pure water or DIW (deionized water) as the rinse liquid. The rinse liquid supply source 52d is connected to the nozzle 56 via a pipe 53d in which an on-off valve 54d is inserted. Therefore, when the on-off valve 54d is opened, the rinse liquid supplied from the rinse liquid supply source 52d is discharged from the nozzle 56. Note that pure water, warm water, ozone water, magnetic water, reduced water (hydrogen water), various organic solvents (ionized water, IPA (isopropyl alcohol), functional water (CO2 water, etc.), etc. may also be used as the rinse liquid.

The processing liquid supply unit 51 selectively supplies SC-1, DHF, SC-2, and a rinse liquid. When the processing liquid (SC-1, DHF, SC-2, or rinse liquid) is supplied from the processing liquid supply unit 51 to the corresponding nozzle among the nozzles 5a to 5d, the nozzle discharges a liquid flow of the processing liquid so as to hit the processing area of the upper surface peripheral portion of the rotating substrate W. However, each of the opening and closing valves 54a to 54d provided in the processing liquid supply unit 51 is opened and closed under the control of the control unit 10 by a valve opening and closing mechanism (not shown) electrically connected to the control unit 10. In other words, the discharge mode of the processing liquid from the nozzle of the nozzle head 50 (specifically, the type of processing liquid discharged, the discharge start timing, the discharge end timing, the discharge flow rate, etc.) is controlled by the control unit 10. In other words, the processing liquid discharge mechanism 51A discharges a liquid flow of the processing liquid so as to hit a position in the rotation locus of the upper surface peripheral portion of the substrate W rotating around the rotation axis a1 under the control of the control unit 10.

The nozzle movement mechanism 6 is a mechanism that moves the nozzle heads 44, 45, and 50 independently between their respective processing positions and retracted positions. The nozzle movement mechanism 6 comprises horizontally extending arms 61-63, nozzle bases 64-66, and drive units 67-69. The nozzle heads 44, 45, and 50 are attached to the tips of the arms 61-63, respectively.

The base ends of the arms 61 to 63 are connected to the upper end portions of the nozzle bases 64 to 66, respectively. The nozzle bases 64 to 66 are arranged around the casing 24 in a distributed manner with their axes aligned vertically. The nozzle bases 64 to 66 each have a rotating shaft that extends vertically along its axis and can rotate around the axis. The axis of the nozzle bases 64 to 66 coincides with the axis of each rotating shaft. The upper end portions of the nozzle bases 64 to 66 are attached to the upper ends of each rotating shaft. When each rotating shaft rotates, the upper end portions of the nozzle bases 64 to 66 rotate around the axis of each rotating shaft, i.e., the axis of the nozzle bases 64 to 66. The nozzle bases 64 to 66 are provided with drive units 67 to 69 that rotate the respective rotating shafts around their axes. The drive units 67 to 69 are each configured to include, for example, a stepping motor or the like.

The drive units 67-69 rotate the upper end portions of the nozzle bases 64-66 via the rotation shafts of the nozzle bases 64-66, respectively. As each upper end portion rotates, the nozzle heads 44, 45, 50 also rotate around the axis of the nozzle bases 64-66. As a result, the drive units 67-69 move the nozzle heads 44, 45, 50 horizontally between their respective processing positions and the retracted position.

When the nozzle head 44 is positioned at the processing position, the outlet of the nozzle 46 faces a part of the rotational trajectory of the peripheral edge of the substrate W rotated by the rotating and holding mechanism 2, and the outlet of the nozzle 47 faces another part of the rotational trajectory.

When the nozzle head 45 is placed in the processing position, the nozzle 48 is located above the center c1 of the substrate W, and the axis of the nozzle 48 coincides with the rotation axis a1 of the spin chuck 21. The outlet (lower opening) of the nozzle 48 faces the center of the substrate W. In addition, the lower surface of the blocking plate 90 faces and is parallel to the upper surface of the substrate W. The blocking plate 90 is close to the upper surface of the substrate W in a non-contact state.

When the nozzle head 50 is placed in the processing position, the nozzles 5a to 5d are placed in the processing position. The retracted positions of the nozzle heads 44, 45, and 50 are positions where they do not interfere with the transport path of the substrate W and do not interfere with each other. Each retracted position is, for example, a position outside and above the splash guard 31.

A heating mechanism 7 is provided below the peripheral edge of the lower surface of the substrate W. The heating mechanism 7 includes an annular heater (not shown) that extends in the circumferential direction of the substrate W along the peripheral edge of the lower surface of the substrate W, and an electric circuit (not shown) that supplies power to the heater under the control of the control unit 10. When power is supplied to the heater while the heater is positioned directly below the peripheral edge of the lower surface of the substrate W, the heater generates heat and heats the peripheral edge of the substrate W.

The film thickness estimation mechanism 8 includes an imaging head 81. As shown in FIG. 3, the imaging head 81 includes an illumination unit 82 that irradiates the substrate W with observation light L (λ) having a wavelength λ at a predetermined incident angle θ (e.g., 30°) to illuminate the substrate W, and an imaging unit 83 that receives reflected light RL reflected by the substrate W to image the top layer L1 of the substrate W and obtain an image of the top layer L1 (symbol IM in FIG. 7). In this embodiment, the imaging unit 83 is configured as a monochrome two-dimensional camera. As shown in FIG. 2, the imaging head 81 is attached to the tip of an arm 84. The arm 84 extends along a horizontal plane. The base end of the arm 84 is connected to the upper end portion of a head base 85. The head base 85 is arranged around the casing 24 with its axis aligned vertically. The head base 85 extends vertically along its axis and includes a rotation shaft that can rotate around the axis. The axis of the head base 85 and the axis of the rotation shaft coincide with each other. The upper end of each rotating shaft is attached to the upper end portion of a head base 85. When each rotating shaft rotates, the upper end portion of the head base 85 rotates around the axis of the rotating shaft, i.e., the axis of the head base 85. The head base 85 is provided with a drive unit 86 that rotates each rotating shaft around its axis. The drive unit 86 is configured with, for example, a stepping motor.

The film thickness estimation mechanism 8 has a film thickness estimation control unit 87 that controls the illumination unit 82, the imaging unit 83, and the driving unit 86 in response to a film thickness estimation command from the control unit 10 to obtain the film thickness d1 of the top layer L1 of the substrate W and outputs information about the film thickness d1 to the control unit 10. The film thickness estimation control unit 87, like the control unit 10, has a CPU 871 that performs various calculation processes and a memory unit 872 that stores a film thickness estimation program and various information. The CPU 871 cooperates with the control unit 10 in accordance with the film thickness estimation program to obtain related information necessary for film thickness estimation before the bevel etching process, a calibration value, and the film thickness immediately before the top layer L1, and estimates the film thickness d1 of the top layer L1 during the bevel etching process. In other words, the CPU 871 and the memory unit 872 function as the "film thickness estimation unit" and "memory unit" of the present invention, respectively. In this embodiment, the film thickness estimation control unit 87 is provided independent of the control unit 10, but the function of the film thickness estimation control unit 87 may be incorporated into the control unit 10, and the CPU and magnetic disk of the control unit 10 may function as the "film thickness estimation unit" and "storage unit" of the present invention, respectively.

Next, the bevel etching operation by the substrate processing apparatus 1 configured as described above will be described with reference to Figures 4 to 9. Figure 4 is a flow chart showing an example of the bevel etching operation by the substrate processing apparatus shown in Figures 1 and 2.

In the substrate processing apparatus 1, before performing the bevel etching process, relevant information of the substrate W to be processed is acquired and stored in the memory unit 872 (step S1). Here, the relevant information refers to the refractive index n(λ,1), n(λ,2), ..., n(λ,q) and extinction coefficient k(λ,1), k(λ,2), ..., k(λ,q) of each layer L1 to Lq of the substrate W at the wavelength λ, the film thickness d2 to dq of each layer of the multilayer structure except for the top layer L1, the angle of incidence θ, and the wavelength λ. The main purpose of acquiring such relevant information in advance and storing it in the memory unit 872 is closely related to the film thickness estimation principle in this embodiment.

The film thickness estimation in this embodiment is based on the method of calculating the reflectance described in "Spectral Characteristics Simulation of Optical Thin Films," Kuriyama, Keiji, Surface Technology, vol. 48, No. 9, 1997 (hereinafter referred to as "Non-Patent Document"). That is, in the Non-Patent Document, the reflectance R(λ) of an arbitrary multilayer film at a certain wavelength λ is determined by the number of layers q in the multilayer structure, the refractive indices n(λ,1), n(λ,2), ..., n(λ,q) of each layer L1 to Lq, the extinction coefficients k(λ,1), k(λ,2), ..., k(λ,q) of each layer L1 to Lq, the film thicknesses d1, d2, ..., dq, and the angle of incidence θ of the observation light L to the substrate W, as shown in FIG. In this case, if the relationship between the reflectance R(λ) and the thickness d1 of the top layer L1 is defined as a function R(λ)=f(d1), the thickness d1 of the top layer L1 can be calculated from the luminance value I(λ) of the pixel constituting the image of the top layer L1 acquired by the imaging unit 83 by the inverse function of the above function, more specifically, the following equation: d1=f ⁻¹ (I(λ)−Ioff)/A) … (1) where A is the gain of the imaging unit 83,
Ioff is the luminance value offset,
Here, the brightness value offset Ioff corresponds to the brightness value of the pixels constituting the image acquired by the imaging unit 83 in an environment where no illumination is provided by the illumination unit 82, i.e., in a dark environment. For example, for a substrate W having a two-layer structure in which a silicon oxide film is formed on a silicon wafer, the gain A can be calculated as follows. A bare silicon substrate is used as a reference sample (corresponding to the "calibration substrate" of the present invention), and while illuminating this with observation light L(λ), the following equation is calculated from the brightness value I(λ,Bare-Si) of the surface image of the bare silicon substrate acquired by the imaging unit 83 and the reflectance R(λ,Bare-Si) of the bare silicon substrate:
A can be calculated using the following equation: A=(I(λ,Bare-Si)-Ioff)/R(λ,Bare-Si) (2).

In this embodiment, therefore, in step S2, it is determined whether or not a calibration value has already been acquired. If a calibration value has already been acquired for the substrate W to be processed that is to be bevel etched, the process proceeds to step S4 without executing the calibration value acquisition process (step S3). On the other hand, if a calibration value has not been acquired ("NO" in step S2), the process proceeds to step S4 after the calibration value acquisition process is executed.

Figure 5 is a flow chart showing an example of the calibration value acquisition operation performed before the bell etching process. In order to facilitate understanding of the calibration value acquisition operation, the processing target substrate W is described here as a two-layer structure substrate in which a silicon oxide film is laminated on a silicon wafer, as described above. In the calibration value acquisition operation (step S3), the imaging unit 83 captures an image with the illumination unit 82 turned off. This allows acquisition of an image in a dark environment (step S31). Then, the CPU 871 of the film thickness estimation control unit 87 calculates the average value of the luminance values of the pixels constituting the image, and stores this in the memory unit 872 as the luminance value offset Ioff (step S32).

Subsequently, a reference sample, which is a substrate W on which only the top layer L1 is missing, is placed on the spin chuck 21. Here, a two-layer substrate on which the top layer (SiO2) L1 is missing, i.e., a bare silicon substrate, is used as the reference sample. The reference sample may be placed by a transfer robot (not shown) accessed from outside the substrate processing apparatus 1, or may be placed manually by an operator. Once the reference sample is placed on the spin chuck 21, a pump supplies negative pressure onto the spin chuck 21. This causes the reference sample to be adsorbed and held on the spin chuck 21. Thus, the loading of the reference sample is completed (step S33).

Next, the illumination unit 82 is turned on to irradiate the observation light L(λ) of wavelength λ onto the surface of the reference sample, and the imaging unit 83 receives the light reflected by the surface of the reference sample to obtain an image of the reference sample (for example, the symbol IM in the lower left drawing of FIG. 7, which will be described later) (step S34). Then, the CPU 871 obtains the average value of the luminance values of the pixels constituting the image of the reference sample as the luminance value I(λ, Bare-Si) (step S35). In parallel with this, the CPU 871 reads out from the memory unit 872 the refractive index n(λ, 2) of silicon at wavelength λ, the extinction coefficient k(λ, 2) at wavelength λ, the number of layers of the reference sample = 1, and the thickness d2 of the second layer L2 (here, the thickness of the bare silicon substrate). The CPU 871 also calculates the reflectance R(λ, Bare-Si) of the reference sample based on these values and the angle of incidence θ (e.g., 30°) of the observation light L(λ) with respect to the reference sample (step S36). The CPU 871 then substitutes these values into the above equation (1) and calculates the gain A (step S37).

The brightness value offset Ioff and gain A thus obtained are stored in the memory unit 872 as calibration values, and then the suction and holding by the spin chuck 21 is released. Then, the reference sample (bare silicon substrate) is unloaded from the spin chuck 21 by a transport robot, an operator, or the like (step S38). This completes the process of acquiring the calibration values.

Returning to Figure 4, the explanation continues. When the calibration values to be used in estimating the film thickness of the top layer L1 have been prepared, the spin chuck 21 is empty, and the substrate W to be processed is carried in and placed on it by a transport means such as a transport robot. Then, a pump supplies negative pressure onto the spin chuck 21, which causes the reference sample to be attracted and held on the spin chuck 21. In this way, the loading of the reference sample is completed (step S4). Meanwhile, the transport robot retreats from the substrate processing apparatus 1.

The substrate W held by the spin chuck 21 is basically manufactured with each layer having a value designed in advance, and it is highly likely that the thickness of the top layer L1 of the substrate W immediately after loading is the design value d1d or a value close to it. However, the substrate processing apparatus 1 does not have information on the substrate history, such as the time elapsed since the top layer L1 was formed or the processing steps. Therefore, it is desirable to obtain the thickness d1 of the top layer L1 at this point by the "thickness estimation of the top film", which is a characteristic part of the present invention. Therefore, in this embodiment, the thickness d1 of the top layer L1 immediately after loading is obtained in the same sequence as the thickness estimation of the top layer L1 performed at a certain time interval (sampling period) during the bevel etching process, as will be described later. More specifically, the design value d1d is set as the thickness of the top layer L1 immediately before the thickness estimation (hereinafter referred to as "previous thickness d1_pre") (step S5), and the CPU 871 uses this to execute the thickness estimation process shown in FIG. 6 to estimate the thickness d1 of the top layer L1 (step S6). Then, the film thickness d1 obtained in step S6 is set as the previous film thickness d1_pre immediately before the start of the bevel etching process (step S7).

Figure 6 is a flowchart showing the film thickness estimation operation of the top layer. Figures 7 and 8 are diagrams for explaining the film thickness estimation operation of the top layer. The substrate processing apparatus 1 is an apparatus for performing a bevel etching process, and the peripheral portion of the top layer L1 is selectively etched away. Therefore, the portion where the film thickness d1 of the top layer L1 should be accurately obtained is also the peripheral portion and its surroundings. Therefore, during the film thickness estimation operation of the top layer L1, the imaging head 81 is placed above the peripheral portion of the substrate W held by the spin chuck 21 as shown in Figure 2. After that, the illumination unit 82 is turned on to emit observation light L (λ) of wavelength λ. The observation light L (λ) is irradiated to the peripheral portion of the substrate W and its surroundings at an incident angle θ (Figure 3) as shown in Figure 7, thereby illuminating a part of the top layer L1.

The imaging unit 83 receives the light reflected multiple times by the top layer L1 to the substrate layer Lq. As a result, a two-dimensional image IM is obtained, for example, as shown in the enlarged view at the bottom left of FIG. 7 (step S61). The enlarged view at the bottom right of the figure shows an example of a two-dimensional image IM obtained during a bevel etching process, which will be described later. In the figure, hatched areas indicate areas that are not etched, and dotted areas indicate areas that are etched. In addition, in the figure, x and y coordinates are added to identify each pixel PX that constitutes the two-dimensional image. In other words, the two-dimensional image is composed of a group of pixels from pixel PX(0,0) to pixel PX(m,n). Then, steps S62 to S66 are performed for each pixel to obtain the estimated film thickness value d1(x,y).

In step S62, the CPU 871 derives the film thickness d1(x,y) of the top layer L1 at pixel PX(x,y) based on the above formula (1). However, the inverse function in formula (1) can have multiple solutions. That is, the reflectance R(λ) and the film thickness d1 have the relationship shown by the curve in FIG. 8, and the intersections of the curve and the measured reflectance R(λ) correspond to the multiple solutions described above and are obtained as film thickness candidates (hereinafter referred to as "film thickness candidate values").

Here, in order to obtain a stable and reliable film thickness d1, it is necessary to efficiently narrow down the solution from multiple film thickness candidate values. For example, in an ellipsometer, the solution is narrowed down by measuring the luminance value at multiple wavelengths and predicting and fitting the reflectance. However, this embodiment does not have a mechanism for changing the wavelength λ of the observation light L. Therefore, this embodiment cannot apply such an algorithm.

の範囲に含まれる膜厚候補値を抽出し、これを解とすることができる（ステップＳ６３）。ただし、上記範囲に膜厚候補値が１個のみ含まれる場合と、２個含まれる場合とが存在する。そこで、ＣＰＵ８７１は、上記範囲に含まれる膜厚候補値が１個であるか、２個であるかを判定する。そして、膜厚候補値が１個である場合（ステップＳ６４で「１個（ｄ１ａ）」）、ＣＰＵ８７１は当該膜厚候補値を膜厚推定値ｄ１（ｘ，ｙ）と認定する（ステップＳ６５）。一方、膜厚候補値が２個である場合（ステップＳ６４で「２個（ｄ１ａ、ｄ１ｂ）」）、ＣＰＵ８７１は膜厚候補値の決定処理を実行する（ステップＳ６６）。 Therefore, in this embodiment, the CPU 871 narrows down the solution by using the previous film thickness d1_pre. More specifically, when the phase difference φ (=4·π·n(λ,1)·d1·cos θ/λ) of the interference fringes in the uppermost layer L1 has a half period π, d1 = λ/(4·n(λ,1)·cos θ) is satisfied, as shown in FIG.

A thickness candidate value included in the range can be extracted and used as a solution (step S63). However, there are cases where the range includes only one thickness candidate value and cases where the range includes two thickness candidate values. Therefore, the CPU 871 judges whether the range includes one or two thickness candidate values. If there is one thickness candidate value (step S64: "one (d1a)"), the CPU 871 recognizes the thickness candidate value as the thickness estimated value d1(x, y) (step S65). On the other hand, if there are two thickness candidate values (step S64: "two (d1a, d1b)"), the CPU 871 executes a process of determining the thickness candidate value (step S66).

Figure 9 is a flowchart showing the process of determining the film thickness candidate value. In this determination process (step S66), one of the two film thickness candidate values is set as the film thickness candidate value d1a, and the other is set as the film thickness candidate value d1b. The film thickness estimate value d1 from one loop before in the estimation loop is set as the previous estimate value d1_pre(x,y). The design value d1d is set as the initial value of the previous estimate value d1_pre(x,y) in the estimation loop.

In step S661, the following equation is used:
d1a'=d1a-d1_pre(x,y)
d1b'=d1b-d1_pre(x,y)
The differential amounts d1a' and d1b' of the film thickness candidate value are calculated based on the above (step S661). Then, the changes Ca and Cb of the differential amounts are expressed by the following formulas, respectively:
Ca=d1a'-d1'(x,y)
Cb=d1b'-d1'(x,y)
(Step S662) where d1'(x, y) is the differential amount of the film thickness estimated value d1 in the previous loop in the estimation loop (hereinafter referred to as the "previous differential amount"), and zero is set as the initial value of the previous differential amount d1'(x, y) in the estimation loop.

In the next step S663, the amounts of change Ca and Cb are compared. If the amount of change Ca is smaller ("NO" in step S663), the estimated film thickness value d1(x, y) is determined to be the film thickness candidate value d1a (step S664). For the next loop, the previous estimated value d1_pre(x, y) and the previous differential amount d1'(x, y) are respectively set as follows: d1_pre(x, y)=d1a
d1'(x,y)=d1a'
and stored in the storage unit 872 (step S665).

On the other hand, if the change amount Cb is smaller ("YES" in step S663), the film thickness estimate value d1(x, y) is determined to be the film thickness candidate value d1b (step S666). Also, for the next loop, the previous estimate value d1_pre(x, y) and the previous differential amount d1'(x, y) are respectively expressed as follows: d1_pre(x, y)=d1b
d1'(x,y)=d1b'
and stored in storage unit 872 (step S667).

By executing the process for determining the film thickness candidate value (step S66) described above, a stable and reliable film thickness estimate value d1 can be obtained even when there are two film thickness candidate values.

6, the film thickness estimation loop is executed for all pixels PX. As a result, the following data is obtained and stored in the storage unit 872.
――――――――――――――――――――――――――
Pixel PX(0,0) Brightness value I(0,0) Estimated film thickness d1(0,0)
Pixel PX(1,0) Brightness value I(1,0) Estimated film thickness d1(1,0)
…
Pixel PX(m,p) Brightness value I(m,p) Estimated film thickness d1(m,p)
…
Pixel PX(m,n) Brightness value I(m,n) Estimated film thickness d1(m,n)
――――――――――――――――――――――――――

The CPU 871 extracts the film thickness estimates d1(0,0) to d1(m,p) for pixels PX(0,0) to PX(m,p) corresponding to the bevel etching region from the data. The CPU 871 then determines the film thickness d1 from the film thickness estimates d1(0,0) to d1(m,p) (step S67). For example, the average value of the film thickness estimates d1(0,0) to d1(m,p) may be set as the film thickness d1, or a weighted average value of the film thickness estimates d1(0,0) to d1(m,p) may be set as the film thickness d1 after applying a weighting coefficient to each pixel. In addition, since the bevel etching process has not yet started at this point, the film thickness d1 may be determined from the film thickness estimates d1(0,0) to d1(m,n) for all pixels.

Furthermore, returning to FIG. 4, the explanation will be continued. When the thickness d1 of the top layer L1 is obtained immediately before the bevel etching process as described above, the CPU 871 sets the thickness estimate d1 as the previous thickness d1_pre of the top layer L1 and stores it in the memory unit 872 (step S7).

Following the setting of the previous film thickness d1_pre, the bevel etching process is started (step S8). That is, the nozzle heads 44, 45, and 50 are placed at the processing position by the nozzle moving mechanism 6, and the splash guard 31 is placed at the upper position by the guard driving mechanism. Thus, when preparation for the substrate processing is completed, the rotation mechanism 231 of the substrate processing apparatus 1 starts rotating the spin chuck 21 that holds the substrate W. Also, power is supplied to the heater of the heating mechanism 7 at an appropriate timing, and heating of the peripheral portion of the substrate W is started. Also, the gas discharge mechanisms 41 and 42 start discharging a gas flow of the inert gas from the nozzles 46 and 47 of the nozzle head 44, and the gas discharge mechanism 43 starts discharging a gas flow of the inert gas from the nozzle 48 of the nozzle head 45. Then, after the temperature of the peripheral portion of the substrate W increases over time, the processing liquid discharge mechanism 51A starts processing the peripheral portion of the upper surface by discharging a liquid flow of the processing liquid (chemical liquid) so as to hit the peripheral portion of the upper surface of the substrate W.

After the start of the bevel etching process, each time it is confirmed in step S9 that a certain time has elapsed, the film thickness d1 of the top layer L1 is obtained by a process similar to that in step S6 (step S10). That is, as the bevel etching process progresses, the thickness of the top layer L1 decreases from moment to moment, and the film thickness d1 of the top layer L1 is sampled by the film thickness estimation mechanism 8 every certain time. Therefore, in this embodiment, it is determined whether the sampled film thickness d1 has reached zero (step S11). This determination is made by the control unit 10, and while the film thickness d1 has not reached zero, that is, the bevel etching process has not been completed ("NO" in step S11), the previous film thickness d1_pre is rewritten to the film thickness d1 obtained in step S10 (step S12), and then the process returns to step S9. In this way, the bevel etching process continues while the film thickness d1 of the top layer L1 is obtained at regular intervals.

On the other hand, when it is confirmed in step S11 that the film thickness d1 has reached zero and all of the peripheral portion of the top layer L1 has been etched away, the control unit 10 controls each part of the apparatus to terminate the bevel etching process (step S13). That is, the heating of the substrate W is stopped by stopping the power supply to the heater, and the discharge of the gas flow is also stopped. Furthermore, the rotation of the spin chuck 21 is stopped. After that, the hand of the transfer robot (not shown) enters the substrate processing apparatus 1 and receives the processed substrate W, and the substrate W is released from the spin chuck 21. The transfer robot then transports the received substrate W to the next substrate processing apparatus (step S14: unloading).

As described above, in this embodiment, the thickness d1 of the top layer L1 of the substrate W is estimated from an image (e.g., symbol IM in FIG. 7) of the top layer L1 of the substrate W obtained by imaging the substrate W illuminated with observation light L(λ) of a specific wavelength λ. In this way, the thickness d1 of the top layer L1 is obtained using only the observation light L(λ) of a specific wavelength λ, without requiring wavelength scanning of the observation light as in the conventional technology. In this way, according to this embodiment, the thickness d1 of the top layer L1 of a substrate W having a multi-layer structure can be obtained at low cost and in real time.

In addition, in the above embodiment, a thin film estimation process (step S6) is performed on the top layer L1 immediately before the start of the bevel etching process to set the previous film thickness d1_pre (step S7). Therefore, the film thickness d1 of the top layer L1 immediately before the bevel etching process can be determined with high accuracy, and the film thickness d1 of the top layer L1 can be determined stably even during the bevel etching process. As a result, the end timing of the bevel etching process can be accurately detected, and the occurrence of excessive or insufficient bevel etching can be effectively prevented.

The present invention is not limited to the above-mentioned embodiment, and various modifications can be made without departing from the spirit of the present invention. For example, in the embodiment, the film thickness estimation method and film thickness estimation device according to the present invention are applied to a substrate processing apparatus 1 that performs bevel etching processing, but the scope of application of the present invention is not limited to this, and the present invention can also be applied to an etching apparatus that etches the entire or part of the top layer L1 of the substrate W. In addition to detecting the etching end, the etching rate of the top layer L1 may be calculated by dividing the amount of change in film thickness d1 by the sampling period. The etching rate can also be used to maintain the etching rate and predict or estimate the etching end point. Furthermore, the present invention can also be applied to a substrate processing apparatus that grows the top layer L1 to increase the film thickness d1.

In the above embodiment, a bare silicon substrate is used as a reference sample (corresponding to the "calibration substrate" of the present invention) in response to the example of a two-layer substrate W in which a silicon oxide film is formed on a silicon wafer. However, the multi-layer structure is not limited to a (SiO2/Si) structure and is arbitrary. For example, the present invention can also be applied to a substrate processing apparatus that processes a substrate having an oxide film, nitride film, metal film, organic film, or the like other than silicon as the top layer L1.

A calibration substrate according to the multilayer structure can be used. For example, a reference sample composed only of the material constituting the second layer L2, as in the above embodiment, can be used as the "calibration substrate" of the present invention. A reference sample obtained by removing only the top layer from a substrate having a multilayer structure can be used as the "calibration substrate" of the present invention.

In addition, in the above embodiment, the thin film estimation process (step S6) of the top layer L1 is performed immediately before the start of the bevel etching process to set the previous film thickness d1_pre (step S7), but these steps may be omitted. In this case, the design value d1d of the top layer L1 becomes the previous film thickness d1_pre immediately before the bevel etching process.

In the above embodiment, the imaging unit 83 is configured with a monochrome two-dimensional camera, but other cameras can be used. For example, a monochrome line camera can be used to capture an image of the rotating substrate W to obtain a two-dimensional image. It can also be configured with a color two-dimensional camera. In this case, if the illumination unit 82 is configured with a light source that emits observation light having light components of multiple wavelengths, such as a white light source, the imaging unit 83 can capture an image for each wavelength. Then, by performing the above process for each wavelength, it is possible to obtain film thicknesses at different wavelengths. Therefore, simple spectroscopic measurement is possible by comparing and correcting these multiple film thickness estimation results, and measurement accuracy and reliability can be improved.

This invention can be applied to all film thickness estimation techniques for estimating the film thickness of the top layer of a substrate having a multilayer structure in which at least one thin film layer is laminated on a base layer after processing the substrate, and to all etching methods that use the film thickness estimation techniques.

1... Substrate processing apparatus 8... Film thickness estimation mechanism (film thickness estimation apparatus)
82: Illumination unit 83: Imaging unit 87: Film thickness estimation control unit 871: CPU
872...Memory unit A...Gain d1...Thickness (of the top layer) d1a, d1b...Thickness candidate value d1d...Designed thickness IM...Image (of the top layer) k...Extinction coefficient L...Observation light L1...Top layer (of the substrate) L2...Second layer (of the substrate) Lq...Base layer n...Refractive index PX...Pixel q...Number of layers R...Reflectance RL...Reflected light W...Substrate λ...Wavelength θ...Angle of incidence

Claims (9)

A film thickness estimation method for estimating a film thickness of a top layer of a substrate having a multilayer structure in which at least one thin film is laminated on a base layer after the substrate has been subjected to a process, comprising: (a) a step of irradiating the substrate having the process with observation light of a wavelength λ to obtain an image of the top layer of the substrate with an imaging unit; (b) a step of storing a refractive index and an extinction coefficient of each layer constituting the multilayer structure at the wavelength λ, and a film thickness of each layer of the multilayer structure except for the top layer; and (c) a step of estimating a film thickness of the top layer from a luminance value of a pixel constituting the image of the top layer, based on the refractive index, the extinction coefficient, and the film thickness stored in the step (b), and an inverse function of a function (R(λ)=f(d1)) showing a relationship between a reflectance R(λ) of the substrate at the wavelength λ and a film thickness d1 of the top layer;
A film thickness estimation method comprising: 2. The method of claim 1, further comprising: (d) determining a calibration value for performing the step (c),
The step (d) comprises: (d-1) a step of capturing an image by the imaging unit in a dark environment in which the observation light is not emitted; (d-2) a step of storing luminance values of pixels constituting the image acquired by the step (d-1) as a luminance value offset Ioff; (d-3) a step of irradiating the observation light onto a calibration substrate, the uppermost layer of which is made of a material constituting the second layer that is the second layer from the top in the multilayer structure, to acquire an image of the calibration substrate by the imaging unit; and (d-4) a step of calculating a gain A based on the luminance values of pixels constituting the image acquired by the step (d-3) and the reflectance of the calibration substrate,
The inverse function is d1=f ⁻¹ ((I(λ)−Ioff)/A),
where I(λ) is the luminance value of the pixel constituting the image acquired in step (a),
This is a method for estimating film thickness. The film thickness estimation method according to claim 2,
The gain A is expressed by the following formula: A=(I(λ, d-3)-Ioff)/R(λ, d-3))
where I(λ, d-3) is the luminance value of the pixel constituting the image acquired in the step (d-3),
R(λ, d−3) is the reflectance of the calibration substrate at wavelength λ;
A method for estimating film thickness. 4. The method of claim 1, further comprising: (e) acquiring a thickness of the uppermost layer before processing as a previous thickness;
The step (c) includes the steps of: (c-1) obtaining a candidate value for the film thickness of the uppermost layer from a luminance value of the pixel based on the refractive index, the extinction coefficient, and the film thickness stored in the step (b) and the inverse function, for each pixel constituting the image of the uppermost layer, and then determining a film thickness from the candidate value for the film thickness based on the previous film thickness;
(c-2) determining a thickness of the top layer from the thickness estimates for each pixel; and
The film thickness estimation method includes the steps of: The film thickness estimation method according to claim 4,
The step (e) includes: (e-1) acquiring a design value of the uppermost layer of the substrate before processing; (e-2) acquiring an image of the uppermost layer of the substrate by irradiating the substrate before processing with observation light; and (e-3) determining the previous film thickness based on the refractive index, the extinction coefficient, and the film thickness stored in the step (b) and the inverse function.
The film thickness estimation method includes the steps of: 2. The film thickness estimation method according to claim 1, wherein the step (a) is a step of irradiating light having light components of different wavelengths as the observation light and acquiring an image for each of the wavelengths by the imaging unit,
The film thickness estimating method, wherein the step (b) and the step (c) are performed for each of the wavelengths. 1. A film thickness estimation device for estimating a film thickness of a top layer of a substrate having a multi-layer structure in which at least one thin film layer is laminated on a base layer, after the substrate is subjected to a process, the device comprising:
an illumination unit that irradiates the substrate on which the treatment has been performed with observation light having a wavelength λ;
an imaging unit that captures an image of the substrate irradiated with the observation light to obtain an image of a top layer of the substrate;
a storage unit that stores the refractive index and extinction coefficient of each layer constituting the multilayer structure at the wavelength λ, and the film thickness of each layer of the multilayer structure except for the uppermost layer;
a film thickness estimation unit that estimates a film thickness of the uppermost layer from a luminance value of a pixel constituting an image of the uppermost layer based on the refractive index, the extinction coefficient, and the film thickness stored in the memory unit and an inverse function of a function (R(λ)=f(d1)) that indicates a relationship between a reflectance R(λ) of the substrate and a film thickness d1 of the uppermost layer;
A film thickness estimation device comprising: 1. An etching method for etching a top layer of a substrate having a multilayer structure in which at least one thin film layer is laminated on a base layer, comprising:
An etching method, comprising: determining a film thickness of a top layer of the substrate being treated with the treatment liquid using the film thickness estimation method according to any one of claims 1 to 6; and controlling the supply of the treatment liquid in accordance with the film thickness. 9. The etching method according to claim 8,
selectively etching the peripheral portion of the top layer by supplying the processing liquid to the peripheral portion of the substrate;
an etching method comprising: stopping supply of the processing liquid to a peripheral portion of the substrate when a film thickness of the uppermost layer estimated by the film thickness estimation method becomes zero;

Images