As the semiconductor manufacturing process is moving towards the 3 nm node, there is a crucial need to reduce the edge placement error (EPE) to ensure proper functioning of the integrated circuit (IC) devices. EPE is the most important metric that quantify the fidelity of fabricated patterns in multi-patterning processes, and it is the combination of overlay errors and critical dimension (CD) errors. Recent advances in machine learning have enabled many new possibilities to improve the performance and efficiency of EPE optimization techniques. In this paper, we conducted a survey of recent research work that applied machine learning/ deep learning techniques for the purposes of enhancing virtual overlay metrology, reducing overlay error, and improving mask optimization methods for EPE reduction. Thorough discussions about the objectives, datasets, input features, models, key findings, and limitations are provided. In general, the results of the review work show a great potential of machine learning techniques in aiding the improvement of EPE in the field of semiconductor manufacturing. 

With the progression of deep learning algorithms in computer vision, a lot of research is taking place in the semiconductor industry towards improving real-time defect detection and classification analysis. An Automated Defect Classification and Detection (ADCD) framework not only enables rapid measurement of dimensions and classification of defects, but also helps minimize production costs, engineering time as well as tool cycle time associated with the defect inspection process. As we continue to shrink the pitch (below 36nm), defect characterization at wafer scale becomes a key issue as it demands rapid measurement but without losing accuracy and repeatability. Also, in the context of high NA lithography (thin resist), accurate metrology becomes difficult with very noisy as well as low contrast images (No BKM exists till now). Human eyes generally demonstrate close to the Bayesian Error limit in detecting smaller objects (for example, extracting contextual information instantaneously from nanoscale defects in SEM images). However, for most One-stage and Two-stage object detectors, this is still a very challenging task due to variable image resolution and SEM (scanning electron microscope) image quality (low SNR). In this research work, we have experimented with different modified YOLOv5 object detectors to improve challenging stochastic defect detection precision. In this work, we have proposed an ensemble strategy by empirically combining multiple custom-trained models (YOLOv5n, YOLOv5s, YOLOv5m, YOLOv5l, and YOLOv5x) together at the test and inference time. We have noticed four YOLOv5 architecture variants are outperforming against our previous Ensemble ResNets model with improvements of the average precision metric (AP) of the most difficult defect classes as p gap and microbridges as well as overall mAP accuracy. With Ensemble YOLOv5, the p gap AP and microbridge AP metrices have been improved by 35% and 25.33%, respectively, whereas the overall mAP has been improved by 6.25%. The proposed Automated Defect Classification and Detection (ADCD) framework can also be used for high resolution and high-speed metrology, providing rapid identification of defects with improved certainty and further root cause investigation. 

Defect inspection in semiconductor processes has become a challenging task due to continuous shrink of device patterns (pitches less than 32 nm) as we move from node to node. Current state-of-the-art defect detection tools (optical/e-beam) have certain limitations as these tools are driven by some rule-based techniques for defect classification and detection. These limitations often lead to misclassification of defects, which leads to increased engineering time to correctly classify different defect patterns. In this paper, we propose a novel ensemble deep learning-based model to accurately classify, detect and localize different defect categories for aggressive pitches and thin resists (High NA applications).In particular, we train RetinaNet models using different ResNet, VGGNet architectures as backbone and present the comparison between the accuracies of these models and their performance analysis on SEM images with different types of defect patterns such as bridge, break and line collapses. Finally, we propose a preference-based ensemble strategy to combine the output predictions from different models in order to achieve better performance on classification and detection of defects. As CDSEM images inherently contain a significant level of noise, detailed feature information is often shadowed by noise. For certain resist profiles, the challenge is also to differentiate between a microbridge, footing, break, and zones of probable breaks. Therefore, we have applied an unsupervised machine learning model to denoise the SEM images to remove the False-Positive defects and optimize the effect of stochastic noise on structured pixels for better metrology and enhanced defect inspection. We repeated the defect inspection step with the same trained model and performed a comparative analysis for “robustness” and “accuracy” metric with conventional approach for both noisy/denoised image pair. The proposed ensemble method demonstrates improvement of the average precision metric (mAP) of the most difficult defect classes. In this work we have developed a novel robust supervised deep learning training scheme to accurately classify as well as localize different defect types in SEM images with high degree of accuracy. Our proposed approach demonstrates its effectiveness both quantitatively and qualitatively. 

CD-SEM images inherently contain a significant level of noise. This is because a limited number of frames are used for averaging, which is critical to ensure throughput and minimize resist shrinkage. This noise level of SEM images may lead to false defect detections and erroneous metrology. Therefore, reducing noise in SEM images is of utmost importance. Both conventional noise filtering techniques and recent most discriminative deep-learning based denoising algorithms are restricted with certain limitations. The first enables the risk of loss of information content and the later mostly requires clean ground-truth or synthetic images to train with. In this paper, we have proposed an U-Net architecture based unsupervised machine learning approach for denoising CD-SEM images without the requirement of any such ground-truth or synthetic images in true sense. Also, we have analysed and validated our result using MetroLER, v2.2.5.0. library. We have compared the power spectral density (PSD) of both the original noisy and denoised images. The high frequency component related to noise is clearly affected, as expected, while the low frequency component, related to the actual morphology of the feature, is unaltered. This indicate that the information content of the denoised images was not degraded by the proposed denoising approach in comparison to other existing approaches.

In this post, we have presented a novel, robust, supervised deep learning training scheme based on RetinaNet architecture to accurately detect different defects in SEM images in aggressive pitches. This scheme includes: (i) Classification of defect types: bridge, line-collapse, line-breaks, (ii) Classification in more challenging scenario: micro-bridges, micro-gaps, and (iii) Detection/Localization of each distinct defect of interest in the SEM image. We have also investigated and demonstrated how the condition influences defect detection scenarios if the image is noisy or denoised and how denoised SEM images are aiding for better metrology and enhanced defect inspection. Future research direction can be extended to use data to model defect transfer from litho to etch as well as to other SEM applications (Logic/CH structures) as well as use TEM/AFM images.