Determining pattern ranking based on measurement feedback from printed substrate
Methods for training a process model and determining ranking of simulated patterns (e.g., corresponding to hot spots). A method involves obtaining a training data set including: (i) a simulated pattern associated with a mask pattern to be printed on a substrate, (ii) inspection data of a printed pattern imaged on the substrate using the mask pattern, and (iii) measured values of a parameter of the patterning process applied during imaging of the mask pattern on the substrate; and training a machine learning model for the patterning process based on the training data set to predict a difference in a characteristic of the simulated pattern and the printed pattern. The trained machine learning model can be used for determining a ranking of hot spots. In another method a model is trained based on measurement data to predict ranking of the hot spots.
1. A method comprising:
obtaining a training data set comprising: (i) a simulated pattern associated with a mask pattern to be printed on a substrate, (ii) inspection data of a printed pattern imaged on the substrate using the mask pattern, and (iii) measured values of a parameter of the patterning process applied during imaging of the mask pattern on the substrate; and
training, by a hardware processor system, a machine learning model for the patterning process based on the training data set to predict a difference in a characteristic of the simulated pattern and the printed pattern.
2. The method of claim 1 , wherein the training is an iterative process comprising:
receiving the inspection data comprising a printed pattern image, and the measured values of the parameter of the patterning process;
obtaining, via simulating using a patterning process model and based on the measured values of the parameter of the patterning process, the simulated pattern;
predicting, via the machine learning model, an image corresponding to the simulated pattern;
determining the difference between the predicted image and the printed pattern image; and
adjusting one or more model parameter values of the machine learning model such that the difference between the predicted image and the printed pattern image is reduced.
3. The method of claim 2 , wherein the adjusting of the one or more model parameter values of the machine learning model comprises:
determining a gradient map of the difference between the predicted image and the printed pattern image as a function of a model parameter;
identifying the one or more model parameter values based on the gradient map such that the difference between the predicted image and the printed pattern image is reduced; and
adjusting the one or more model parameter values based on the identified values.
4. The method of claim 2 , wherein the machine learning model is a convolutional neural network.
5. The method of claim 4 , wherein the adjusting of the one or more model parameter values comprises adjusting one or more selected from:
one or more weights of a layer of the convolutional neural network,
one or more biases of a layer of the convolutional neural network,
one or more hyperparameters related to the convolutional neural network, and/or
a number of layers of the convolutional neural network.
6. The method of claim 1 , wherein the simulated pattern is an aerial image, a mask image, a resist image, and/or an etch image.
7. The method of claim 1 , wherein obtaining the simulated pattern comprises performing a simulation, using a patterning process model and using the mask pattern as an input, to produce the simulated pattern.
8. The method of claim 1 , wherein the simulated pattern comprises a hot spot that is likely to fail when imaged on the substrate.
9. The method of claim 1 , wherein the simulated pattern comprises a pattern that does not satisfy a lithography manufacturability check.
10. The method of claim 1 , wherein the characteristic of the simulated pattern or the printed pattern is a critical dimension of a feature of the simulated pattern, or a contour of a feature of the simulated pattern.
11. The method of claim 1 , wherein the inspection data comprises defect data of the imaged substrate obtained via an inspection apparatus.
12. The method of claim 1 , wherein the inspection data is an image of the printed substrate, wherein the image is a scanning electron beam image and/or an optical metrology image.
13. The method of claim 1 , wherein the parameter of the patterning process is a dose and/or a focus.
14. The method of claim 1 , wherein the measured values of the parameter of the patterning process are obtained via a metrology tool.
15. A computer program product comprising a non-transitory computer readable medium having instructions therein, the instructions, when executed by a computer system, configured to cause the computer system to at least:
obtain a training data set comprising: (i) a simulated pattern associated with a mask pattern to be printed on a substrate, (ii) inspection data of a printed pattern imaged on the substrate using the mask pattern, and (iii) measured values of a parameter of a patterning process applied during imaging of the mask pattern on the substrate; and
train a machine learning model of the patterning process based on the training data set to predict a difference in a characteristic of the simulated pattern and the printed pattern.
16. The computer program product of claim 15 , wherein the instructions configured to train a machine learning model are configured to do so iteratively and to cause the computer system to:
receive the inspection data comprising a printed pattern image, and the measured values of the parameter of the patterning process;
obtain, via simulation using a patterning process model and based on the measured values of the parameter of the patterning process, the simulated pattern;
predict, via the machine learning model, an image corresponding to the simulated pattern;
determine the difference between the predicted image and the printed pattern image; and
adjust one or more model parameter values of the machine learning model such that the difference between the predicted image and the printed pattern image is reduced.
17. The computer program product of claim 16 , wherein the instructions configured to adjust one or more model parameter values of the machine learning model are configured to cause the computer system to:
determine a gradient map of the difference between the predicted image and the printed pattern image as a function of a model parameter;
identify the one or more model parameter values based on the gradient map such that the difference between the predicted image and the printed pattern image is reduced; and
adjust the one or more model parameter values based on the identified values.
18. The computer program product of claim 16 , wherein the machine learning model is a convolutional neural network.
19. The computer program product of claim 18 , wherein adjustment of the one or more model parameter values comprises adjustment of one or more selected from:
one or more weights of a layer of the convolutional neural network,
one or more biases of a layer of the convolutional neural network,
one or more hyperparameters related to the convolutional neural network, and/or
a number of layers of the convolutional neural network.
20. The computer program product of claim 15 , wherein the simulated pattern comprises a hot spot that is likely to fail when imaged on the substrate and/or comprises a pattern that does not satisfy a lithography manufacturability check.