Substrate processing apparatus, substrate processing method, and storage medium
A substrate processing apparatus includes: an imaging portion configured to acquire a surface image of a film formed on a surface of a substrate; an optical property estimation portion configured to estimate an optical property of the film based on process information acquired during formation of the film; and a film thickness estimation portion configured to estimate a film thickness of the film based on the surface image and an estimation result of the optical property.
1. A substrate processing apparatus comprising:
an imaging portion configured to acquire a surface image of a film formed on a surface of a substrate, including:
a housing accommodating the substrate; and
a camera imaging the surface image of the film formed on the surface of the substrate; and
a controller including a processor and a memory storing instructions thereon, the instructions when executed by the processor cause the processor to:
estimate an optical property of the film based on process information acquired during formation of the film by inputting the process information into an optical property model that represents a relationship between the process information and the optical property, and the optical property model outputting the estimated optical property of the film in response to the process information, wherein the optical property includes at least one of a refractive index and an extinction coefficient and the process information includes surrounding environment information of the substrate when the film is formed on the surface of the substrate;
estimate a film thickness of the film by inputting the surface image into a film thickness model that represents a relationship between the surface image and the film thickness, the film thickness model outputting an estimated thickness of the film in response to the surface image; and
correct the estimated thickness of the film based on the estimated optical property output by the optical property model.
2. The substrate processing apparatus of claim 1 , wherein the instructions when executed by the processor further cause the processor to estimate the film thickness by estimating an error by inputting the optical property to an error model that is generated in advance to represent a relationship between the optical property and the error of the film thickness and the estimated optical property, and the error model outputting the error in response to the optical property.
3. The substrate processing apparatus of claim 1 , wherein the film thickness model is generated in advance to represent a relationship between the surface image and film thicknesses at a plurality of locations on the surface of the substrate, and
wherein the instructions when executed by the processor further cause the processor to estimate the film thicknesses at the plurality of locations based on the film thickness model and the surface image.
4. The substrate processing apparatus of claim 3 , wherein the optical property model is generated in advance to represent a relationship between the process information and the optical properties at the plurality of locations, and
wherein the instructions when executed by the processor further cause the processor to:
estimate the optical properties at the plurality of locations based on the optical property model and the process information, and
correct each of the estimation results of the film thicknesses at the plurality of locations based on an estimation result of the optical properties at the plurality of locations.
5. The substrate processing apparatus of claim 1 , wherein the instructions when executed by the processor further cause the processor to:
generate the optical property model by machine learning based on process database accumulated by associating the process information with an actual measurement result of the optical property; and
generate the film thickness model by machine learning based on image database accumulated by associating the surface image with an actual measurement result of the film thickness.
6. The substrate processing apparatus of claim 5 , wherein the instructions when executed by the processor further cause the processor to correct the film thickness model based on a comparison between an estimation result of the film thickness and the actual measurement result of the film thickness.
7. The substrate processing apparatus of claim 1 , wherein the process information includes surrounding environment information of the substrate when the film formed on the surface of the substrate is partially removed.
8. A substrate processing method, comprising:
acquiring a surface image of a film formed on a surface of a substrate by a camera;
estimating an optical property of the film based on process information acquired during formation of the film by inputting the process information into an optical property model that represents a relationship between the process information and the optical property, and the optical property model outputting the estimated optical property of the film in response to the process information, wherein the optical property includes at least one of a refractive index and an extinction coefficient and the process information includes surrounding environment information of the substrate when the film is formed on the surface of the substrate;
estimating a film thickness of the film by inputting the surface image into a film thickness model that represents a relationship between the surface image and the film thickness, the film thickness model outputting an estimated thickness of the film in response to the surface image; and
correcting the estimated thickness of the film based on the estimated optical property output by the optical property model.
9. A non-transitory computer-readable storage medium storing instructions thereon, the instructions when executed by a processor cause the processor to:
acquire a surface image of a film formed on a surface of a substrate by a camera;
estimate an optical property of the film by inputting process information into an optical property model that represents a relationship between the process information and the optical property, and the optical property model outputting the estimated optical property of the film in response to the process information, wherein the optical property includes at least one of a refractive index and an extinction coefficient and the process information includes surrounding environment information of the substrate when the film is formed on the surface of the substrate;
estimate a film thickness of the film by inputting the surface image into a film thickness model that represents a relationship between the surface image and the film thickness, the film thickness model outputting an estimated thickness of the film in response to the surface image; and
correcting the estimated thickness of the film based on the estimated optical property output by the optical property model.