A through-focus scanning optical microscopy dimensional measurement method based on deep-learning classification model.

Through-focus scanning optical microscopy (TSOM) is an economical, non-contact and nondestructive method for rapid measurement of three-dimensional nanostructures. There are two methods using TSOM image to measure the dimensions of one sample, including the library-matching method and the machine-learning regression method. The first has the defects of small measurement range and strict environmental requirements; the other has the disadvantages of feature extraction method greatly influenced by human subjectivity and low measurement accuracy. To solve the problems above, a TSOM dimensional measurement method based on deep-learning classification model is proposed. TSOM images are used to train the ResNet50 and DenseNet121 classification model respectively in this paper, and the test images are used to test the model, the classification result of which is taken as the measurement value. The test results showed that with the number of training linewidths increasing, the mean square error (MSE) of the test images is 21.05 nm² for DenseNet121 model and 31.84 nm² for ResNet50 model, both far lower than machine-learning regression method, and the measurement accuracy is significantly improved. The feasibility of using deep-learning classification model, instead of machine-learning regression model, for dimensional measurement is verified, providing a theoretical basis for further improvement on the accuracy of dimensional measurement.

Imaging error compensation method for through-focus scanning optical microscopy images based on deep learning.

Through-focus scanning optical microscopy (TSOM) is a model-based nanoscale metrology technique which combines conventional bright-field microscopy and the relevant numerical simulations. A TSOM image is generated after through-focus scanning and data processing. However, the mechanical vibration and optical noise introduced into the TSOM image during image generation can affect the measurement accuracy. To reduce this effect, this paper proposes a imaging error compensation method for the TSOM image based on deep learning with U-Net. Here, the simulated TSOM image is regarded as the ground truth, and the U-Net is trained using the experimental TSOM images by means of a supervised learning strategy. The experimental TSOM image is first encoded and then decoded with the U-shaped structure of the U-Net. The difference between the experimental and simulated TSOM images is minimised by iteratively updating the weights and bias factors of the network, to obtain the compensated TSOM image. The proposed method is applied for optimising the TSOM images for nanoscale linewidth estimation. The results demonstrate that the proposed method performs as expected and provides a significant enhancement in accuracy.

Nondestructive shape process monitoring of three-dimensional high aspect ratio targets using through-focus scanning optical microscopy.

Low-cost, high-throughput and nondestructive metrology of truly three-dimensional (3-D) targets for process control/monitoring is a critically needed enabling technology for high-volume manufacturing (HVM) of nano/micro technologies in multi-disciplinary areas. In particular, a survey of the typically used metrology tools indicates the lack of a tool that truly satisfies the HVM metrology needs of 3-D targets, such as high aspect ratio (HAR) targets. Using HAR targets here we demonstrate that through-focus scanning optical microscopy (TSOM) is a strong contender to fill the gap for 3-D shape metrology. Differential TSOM (D-TSOM) images are extremely sensitive to small and/or dissimilar types of 3-D shape variations. Based on this here we propose a TSOM method that involves creating a database of cross-sectional profiles of the HAR targets along with their respective D-TSOM signals. Using the database, we present a simple-to-use, low-cost, high-throughput and nondestructive process-monitoring method suitable for HVM of truly 3-D targets, which also does not require optical simulations, making its use straightforward and automatable. Even though HAR targets are used for this demonstration, the similar process can be applied to any truly 3-D targets with dimensions ranging from micro-scale to nano-scale. The TSOM method couples the advantage of analyzing truly isolated targets with the ability to simultaneously analyze many targets present in the large field-of-view of a conventional optical microscope.

Optimizing noise for defect analysis with through-focus scanning optical microscopy.

Through-focus scanning optical microscopy (TSOM) shows promise for patterned defect analysis, but it is important to minimize total system noise. TSOM is a three-dimensional shape metrology method that can achieve sub-nanometer measurement sensitivity by analyzing sets of images acquired through-focus using a conventional optical microscope. Here we present a systematic noise-analysis study for optimizing data collection and data processing parameters for TSOM and then demonstrate how the optimized parameters affect defect analysis. We show that the best balance between signal-to-noise performance and acquisition time can be achieved by judicious spatial averaging. Correct background-signal subtraction of the imaging-system inhomogeneities is also critical, as well as careful alignment of the constituent images used in differential TSOM analysis.