Electron Microscopy-based Automatic Defect Inspection for Semiconductor Manufacturing: A Systematic Review
0
Sign in to get full access
Overview
- Systematic review of using electron microscopy for automatic defect inspection in semiconductor manufacturing
- Covers key aspects of computer vision and image processing techniques applied to this problem
- Examines the state-of-the-art research and emerging trends in this field
Plain English Explanation
Semiconductor chips power many of the electronic devices we use every day, from smartphones to computers. However, during the manufacturing process, tiny defects can sometimes occur on the surface of these chips. These defects can cause the chips to malfunction, so it's important to detect and fix them before the chips are sold.
One way to inspect for these defects is to use a powerful microscope called an electron microscope. Electron microscopes can produce very detailed images that can reveal even the smallest imperfections on the chip's surface. Researchers have been developing computer vision and image processing techniques to automatically analyze these electron microscope images and detect any defects. This can help speed up the inspection process and improve quality control in semiconductor manufacturing.
The paper provides a comprehensive review of the latest research in this area, covering the key techniques and approaches that have been used. It also discusses some of the challenges and limitations of this technology, and highlights promising directions for future development, such as advancing SEM-based nano-scale defect analysis and using portable smartphone microscopes for on-site inspection.
Technical Explanation
The paper presents a systematic review of the use of electron microscopy-based techniques for automatic defect inspection in semiconductor manufacturing. The authors cover the key computer vision and image processing methods that have been applied to this problem, including techniques for process monitoring, image segmentation, and defect classification.
The review examines the various stages of the defect inspection pipeline, from image acquisition using electron microscopes to the final detection and categorization of defects. The authors also discuss the challenges associated with this task, such as dealing with the large size and complexity of electron microscope images, as well as the need to handle different types of defects and manufacturing processes.
The paper also covers emerging trends in this field, including the use of portable smartphone microscopes for on-site inspection and the development of more advanced nano-scale defect analysis techniques. The authors highlight the potential of these new approaches to improve the efficiency and accuracy of semiconductor defect inspection.
Critical Analysis
The paper provides a comprehensive and well-structured review of the state-of-the-art in electron microscopy-based automatic defect inspection for semiconductor manufacturing. The authors have done an excellent job of covering the key technical aspects of this field, including the various computer vision and image processing techniques that have been applied.
One potential limitation of the review is that it may not have captured the most recent developments in this rapidly evolving field. The paper was likely written some time ago, and there may have been significant advancements since then. Additionally, the review focuses primarily on the technical aspects of the problem, without much discussion of the practical challenges and considerations that companies might face when implementing these technologies in a real-world manufacturing environment.
That said, the paper does a good job of highlighting some of the emerging trends and promising directions for future research, such as the use of portable smartphone microscopes and advanced nano-scale defect analysis. These areas could be worth further exploration to address the limitations of current inspection methods.
Overall, this review serves as a valuable resource for researchers and practitioners in the field of semiconductor manufacturing, providing a solid foundation for understanding the current state of the art and identifying potential areas for future development.
Conclusion
This systematic review of electron microscopy-based automatic defect inspection for semiconductor manufacturing highlights the significant advancements that have been made in this field. By leveraging computer vision and image processing techniques, researchers have developed powerful tools for detecting and classifying defects on semiconductor chips with a high degree of accuracy.
The review covers the key technical aspects of this problem, from image acquisition to defect detection and classification. It also examines emerging trends, such as the use of portable smartphone microscopes and advanced nano-scale defect analysis, which hold promise for improving the efficiency and accessibility of semiconductor inspection.
While the review may not capture the most recent developments in this rapidly evolving field, it provides a solid foundation for understanding the current state of the art and identifying areas for future research and innovation. As the semiconductor industry continues to push the boundaries of technology, the ability to quickly and accurately detect defects will become increasingly critical for maintaining high product quality and yield.
This summary was produced with help from an AI and may contain inaccuracies - check out the links to read the original source documents!
Related Papers
0
Electron Microscopy-based Automatic Defect Inspection for Semiconductor Manufacturing: A Systematic Review
Enrique Dehaerne, Bappaditya Dey, Victor Blanco, Jesse Davis
In this review, automatic defect inspection algorithms that analyze Electron Microscope (EM) images of Semiconductor Manufacturing (SM) products are identified, categorized, and discussed. This is a topic of critical importance for the SM industry as the continuous shrinking of device patterns has led to increasing defectivity and a greater prevalence of higher-resolution imaging tools such as EM. These aspects among others threaten to increase costs as a result of increased inspection time-to-solution and decreased yield, respectively. Relevant research papers were systematically identified in four popular publication databases in January 2024. A total of 103 papers were selected after screening for novel contributions relating to automatic EM image analysis algorithms for semiconductor defect inspection. These papers were then categorized based on the inspection tasks they addressed, their evaluation metrics, and the type of algorithms used. A notable finding from this categorization is that reference-based defect detection algorithms were the most popular algorithm type until 2020 when deep learning-based inspection algorithms became more popular, especially for defect classification. Furthermore, four broader research questions were discussed to come to the following conclusions: (i) the key components of inspection algorithms are set up, pre-processing, feature extraction, and final prediction; (ii) the maturity of the manufacturing process affects the data availability and required sensitivity of inspection algorithms; (iii) key challenges for these algorithms relate to the desiderata of minimizing time-to-solution which pushes for high imaging throughput, reducing manual input during algorithm setup, and higher processing throughput; and (iv) three promising directions for future work are suggested based on gaps in the reviewed literature that address key remaining limitations.
Read more9/12/2024
0
An Evaluation of Continual Learning for Advanced Node Semiconductor Defect Inspection
Amit Prasad, Bappaditya Dey, Victor Blanco, Sandip Halder
Deep learning-based semiconductor defect inspection has gained traction in recent years, offering a powerful and versatile approach that provides high accuracy, adaptability, and efficiency in detecting and classifying nano-scale defects. However, semiconductor manufacturing processes are continually evolving, leading to the emergence of new types of defects over time. This presents a significant challenge for conventional supervised defect detectors, as they may suffer from catastrophic forgetting when trained on new defect datasets, potentially compromising performance on previously learned tasks. An alternative approach involves the constant storage of previously trained datasets alongside pre-trained model versions, which can be utilized for (re-)training from scratch or fine-tuning whenever encountering a new defect dataset. However, adhering to such a storage template is impractical in terms of size, particularly when considering High-Volume Manufacturing (HVM). Additionally, semiconductor defect datasets, especially those encompassing stochastic defects, are often limited and expensive to obtain, thus lacking sufficient representation of the entire universal set of defectivity. This work introduces a task-agnostic, meta-learning approach aimed at addressing this challenge, which enables the incremental addition of new defect classes and scales to create a more robust and generalized model for semiconductor defect inspection. We have benchmarked our approach using real resist-wafer SEM (Scanning Electron Microscopy) datasets for two process steps, ADI and AEI, demonstrating its superior performance compared to conventional supervised training methods.
Read more7/18/2024
0
Addressing Class Imbalance and Data Limitations in Advanced Node Semiconductor Defect Inspection: A Generative Approach for SEM Images
Bappaditya Dey, Vic De Ridder, Victor Blanco, Sandip Halder, Bartel Van Waeyenberge
Precision in identifying nanometer-scale device-killer defects is crucial in both semiconductor research and development as well as in production processes. The effectiveness of existing ML-based approaches in this context is largely limited by the scarcity of data, as the production of real semiconductor wafer data for training these models involves high financial and time costs. Moreover, the existing simulation methods fall short of replicating images with identical noise characteristics, surface roughness and stochastic variations at advanced nodes. We propose a method for generating synthetic semiconductor SEM images using a diffusion model within a limited data regime. In contrast to images generated through conventional simulation methods, SEM images generated through our proposed DL method closely resemble real SEM images, replicating their noise characteristics and surface roughness adaptively. Our main contributions, which are validated on three different real semiconductor datasets, are: i) proposing a patch-based generative framework utilizing DDPM to create SEM images with intended defect classes, addressing challenges related to class-imbalance and data insufficiency, ii) demonstrating generated synthetic images closely resemble real SEM images acquired from the tool, preserving all imaging conditions and metrology characteristics without any metadata supervision, iii) demonstrating a defect detector trained on generated defect dataset, either independently or combined with a limited real dataset, can achieve similar or improved performance on real wafer SEM images during validation/testing compared to exclusive training on a real defect dataset, iv) demonstrating the ability of the proposed approach to transfer defect types, critical dimensions, and imaging conditions from one specified CD/Pitch and metrology specifications to another, thereby highlighting its versatility.
Read more7/16/2024
0
Advancing SEM Based Nano-Scale Defect Analysis in Semiconductor Manufacturing for Advanced IC Nodes
Bappaditya Dey, Matthias Monden, Victor Blanco, Sandip Halder, Stefan De Gendt
In this research, we introduce a unified end-to-end Automated Defect Classification-Detection-Segmentation (ADCDS) framework for classifying, detecting, and segmenting multiple instances of semiconductor defects for advanced nodes. This framework consists of two modules: (a) a defect detection module, followed by (b) a defect segmentation module. The defect detection module employs Deformable DETR to aid in the classification and detection of nano-scale defects, while the segmentation module utilizes BoxSnake. BoxSnake facilitates box-supervised instance segmentation of nano-scale defects, supported by the former module. This simplifies the process by eliminating the laborious requirement for ground-truth pixel-wise mask annotation by human experts, which is typically associated with training conventional segmentation models. We have evaluated the performance of our ADCDS framework using two distinct process datasets from real wafers, as ADI and AEI, specifically focusing on Line-space patterns. We have demonstrated the applicability and significance of our proposed methodology, particularly in the nano-scale segmentation and generation of binary defect masks, using the challenging ADI SEM dataset where ground-truth pixelwise segmentation annotations were unavailable. Furthermore, we have presented a comparative analysis of our proposed framework against previous approaches to demonstrate its effectiveness. Our proposed framework achieved an overall [email protected] of 72.19 for detection and 78.86 for segmentation on the ADI dataset. Similarly, for the AEI dataset, these metrics were 90.38 for detection and 95.48 for segmentation. Thus, our proposed framework effectively fulfils the requirements of advanced defect analysis while addressing significant constraints.
Read more9/9/2024