System and method for generating predictive images for wafer inspection using machine learning
A system and method for generating predictive images for wafer inspection using machine learning are provided. Some embodiments of the system and method include acquiring the wafer after a photoresist applied to the wafer has been developed; imaging a portion of a segment of the developed wafer; acquiring the wafer after the wafer has been etched; imaging the segment of the etched wafer; training a machine learning model using the imaged portion of the developed wafer and the imaged segment of the etched wafer; and applying the trained machine learning model using the imaged segment of the etched wafer to generate predictive images of a developed wafer. Some embodiments include imaging a segment of the developed wafer; imaging a portion of the segment of the etched wafer; training a machine learning model; and applying the trained machine learning model to generate predictive after-etch images of the developed wafer.
1 . A method for using a generated SEM image for wafer inspection, comprising:
obtaining a first image of a feature of a first wafer after the first wafer has been etched;
analyzing the first image with a trained machine learning model configured to;
receive images of the first wafer after the first wafer has been etched; and
generate images of the first wafer based on a photoresist application to the first wafer; and
generating, using the machine learning model and without using an image of the feature after a photoresist applied to the first wafer has been developed, an after development image of the feature of the first wafer corresponding to the feature of the first image.
2 . The method of claim 1 , wherein:
the machine learning model is trained based on a plurality of images of features of the first wafer; and
the plurality of images includes images of features after a photoresist is applied to the first wafer and images of corresponding features after the first wafer has been etched.
3 . The method of claim 2 , further comprising training the machine learning model by:
transforming the images of features of the etched first wafer into one or more feature maps by applying a convolutional neural network (CNN), wherein the one or more feature maps contain visual features of the images of the etched first wafer.
4 . The method of claim 3 , further comprising:
transforming the corresponding images of features after the photoresist is applied to the first wafer into one or more feature maps by applying the CNN, wherein the one or more feature maps contain visual features of the images of the features after the photoresist is applied to the first wafer; and
mapping the visual features associated with the images of the etched first wafer to the visual features associated with the images of the features after the photoresist is applied to the first wafer.
5 . The method of claim 4 , further comprising:
training a CNN encoder, using the mapping, to construct the images of the features after the photoresist is applied to the first wafer from the images of the etched first wafer;
extracting trained features from the construction,
inputting the trained features into a CNN decoder; and
adjusting weights of the CNN decoder by training the CNN decoder over the trained features.
6 . The method of claim 2 , wherein the first image comprises a portion of the first wafer that is different from portions of the first wafer used to train the machine learning model.
7 . The method of claim 1 , wherein the machine learning model is further configured to:
transform the first image into one or more feature maps by applying a convolutional neural network (CNN), wherein the one or more feature maps contain visual features of the first image;
extract the visual features associated with the first image using a CNN encoder, wherein the extraction comprises using one or more convolutional pooling layers to learn hierarchical features from the first image;
identify trained features using the extracted visual features; and
input the trained features into a CNN decoder.
8 . The method of claim 1 , wherein the first image comprises a plurality of images.
9 . The method of claim 1 , wherein the machine learning model comprises a plurality of machine learning models.
10 . The method of claim 9 , wherein each machine learning model corresponds to a different portion of the first wafer.
11 . The method of claim 1 , wherein the machine learning model is performed by one machine learning model, the method further comprising inputting etching parameters that characterize etching of the first wafer.
12 . The method of claim 1 , further comprising:
obtaining a second image of a feature of a third wafer after the third wafer has been etched;
analyzing the second image with the trained machine learning model; and
generating, using the machine learning model, an image of a feature that is based on a photoresist application to the third wafer corresponding to the feature of the second image.
13 . The method of claim 12 , wherein the machine learning model of the second image is configured to:
transform the second image into one or more feature maps by applying a convolutional neural network (CNN), wherein the one or more feature maps contain visual features of the second image;
extract the visual features associated with the second image using a CNN encoder, wherein the extraction comprises using one or more convolutional pooling layers to learn hierarchical features from the second image;
identify trained features using the extracted visual features; and
input the trained features into a CNN decoder.
14 . The method of claim 1 , further comprising:
performing metrology on the generated image of the first wafer to enable reworking a second wafer using the metrology results prior to etching the second wafer.
15 . A charged particle multi-beam system for using a generated image for wafer inspection, the system comprising:
a controller including circuitry to:
image a portion of a segment of a first wafer after a photoresist applied to the first wafer has been developed;
image the segment of the first wafer after the first wafer has been etched;
train a machine learning model using the imaged portion of the developed first wafer and the imaged segment of the etched first wafer; and
wherein the trained machine learning model is used to generate, without using an image of a portion of the first wafer after a photoresist applied to the wafer has been developed, an after development image of the portion of the first wafer.
16 . The system of claim 15 , wherein training the machine learning model comprises:
transforming the imaged segment of the etched first wafer into one or more feature maps by applying a convolutional neural network (CNN), wherein the one or more feature maps contain visual features of the imaged segment of the etched first wafer.
17 . The system of claim 16 , wherein training of the machine learning model further comprises:
transformation of the imaged portion of the developed first wafer into one or more feature maps by applying the CNN, wherein the one or more feature maps contain visual features of the imaged portion of the developed first wafer; and
mapping of the visual features associated with the imaged segment of the etched first wafer to the visual features associated with the imaged portion of the developed first wafer.
18 . The system of claim 17 , wherein training the machine learning model further comprises:
training of a CNN encoder, using the mapping, to construct the imaged portion of the developed first wafer from the imaged segment of the etched first wafer;
extraction of trained features from the construction,
inputting of the trained features into a CNN decoder; and
adjustment of weights of the CNN decoder by training the CNN decoder over the trained features.
19 . The system of claim 15 , wherein;
the imaged segment of the etched first wafer comprises a plurality of images; and
the imaged portion of the developed first wafer comprises a plurality of images.