Method for computational metrology and inspection for patterns to be manufactured on a substrate
Methods include generating a scanner aerial image using a neural network, where the scanner aerial image is generated using a mask inspection image that has been generated by a mask inspection machine. Embodiments also include training the neural network with a set of images, such as with a simulated scanner aerial image and another image selected from a simulated mask inspection image, a simulated Critical Dimension Scanning Electron Microscope (CD-SEM) image, a simulated scanner emulator image and a simulated actinic mask inspection image.
1. A method for determining a scanner aerial image from a mask inspection image, the method comprising:
inputting the mask inspection image, wherein the mask inspection image has been generated by a mask inspection machine; and
generating the scanner aerial image from the mask inspection image using a neural network.
2. The method of claim 1 , further comprising:
inputting a set of mask defect locations, wherein the set of mask defect locations has been generated by the mask inspection machine; and
identifying which mask defect locations of the set of mask defect locations result in defects on the generated scanner aerial image.
3. The method of claim 1 , further comprising training the neural network, the method comprising:
inputting a plurality of mask patterns;
simulating each pattern in the plurality of mask patterns using a detailed model of the mask inspection machine to create a simulated mask inspection image;
simulating each pattern in the plurality of mask patterns using a detailed model of a scanner to create a simulated scanner aerial image; and
training the neural network using corresponding pairs of the simulated mask inspection images and the simulated scanner aerial images.
4. A method for determining a scanner aerial image, the method comprising:
inputting a set of images, wherein a first image in the set of images is selected from the group consisting of a simulated mask inspection image, a simulated Critical Dimension Scanning Electron Microscope (CD-SEM) image, a simulated scanner emulator image and a simulated actinic mask inspection image; and wherein a second image in the set of images is a simulated scanner aerial image; and
training a neural network with the first image and the second image to generate the scanner aerial image.
5. The method of claim 4 , further comprising inputting a scanner illumination value, wherein the training comprises using the scanner illumination value.
6. The method of claim 4 , wherein the neural network comprises a convolutional neural network.
7. The method of claim 4 , wherein the neural network comprises a U-Net.
8. The method of claim 4 , wherein the neural network comprises a generative adversarial network (GAN).
9. A method for determining a scanner aerial image from a mask inspection image, the method comprising:
inputting the mask inspection image, wherein the mask inspection image is generated by a mask inspection machine;
generating a mask image from the mask inspection image using a first neural network; and
generating the scanner aerial image from the mask image using a second neural network.
10. The method of claim 9 , further comprising determining an optimized mask image from the mask inspection image, wherein the optimized mask image is used to train the first neural network.
11. The method of claim 10 , wherein the determining of the optimized mask image uses inversion.