LAMP

Motivation

‍Extreme Ultraviolet (EUV) lithography is the current technology for semiconductor device manufacturing, and it allows to uphold Moore’s law in upcoming years. The use of EUV wavelength (13.5 nm) allows a tremendous increase in the resolution, but at the same time, it also increases the cost of the infrastructure for the lithography process. PSI is pioneering a lensless imaging approach - ptychography or coherent diffraction imaging (CDI) - for EUV mask inspection. A dedicated imaging tool, RESCAN, has been built and developed in recent years. With RESCAN, the main challenge is that a very large computational infrastructure is needed to achieve the desired reconstruction speed, which is highly impractical. The objective of the project was to explore alternative data processing approaches to achieve the required reconstruction speed with an affordable hardware infrastructure. The development of a cost-effective framework for mask inspection would have an enormous impact on future technology development in semiconductor device manufacturing.

Proposed Approach / Solution

‍The proposed work introduced Ptycho-LDM, a novel deep learning-based framework for efficient and accurate phase retrieval of extreme ultraviolet (EUV) photomasks from diffraction patterns. It combined a physics-informed ptychographic algorithm (Fig. 1a) with a conditional Latent Diffusion Model (LDM) (Fig. 1b), a more computationally efficient variant of Denoising Diffusion Probabilistic Models (DDPMs). Unlike traditional approaches, Ptycho-LDM does not require additional posterior sampling steps, enabling faster inference.

To overcome the inefficiencies of standard ptychographic algorithms like Difference Map (DifMap)—which require many iterations and probe positions—Ptycho-LDM uses a hybrid strategy. First, a coarse spatial reconstruction is obtained with a reduced-iteration ptychographic method. Then, this is refined using the generative power of LDMs in a low-dimensional latent space, significantly reducing computational cost and reconstruction time.

Experiments show that Ptycho-LDM achieves a 10x speedup, reducing phase retrieval time from 174 seconds to just 0.5 seconds, while maintaining high fidelity. This approach leverages both the physical understanding of diffraction and the learning capabilities of deep generative models, setting a new benchmark in EUV mask imaging. It highlights the potential of integrating likelihood-based generative models with classical reconstruction techniques for real-world scientific applications.

A visual comparison of ptychographic phase retrieval is shown in Fig. 2 and  visual results of EUV Photomask Inspection is shown in Fig. 3.

Impact

‍The impact of this work is significant in advancing semiconductor manufacturing technologies. Firstly, by implementing deep generative models for phase retrieval in ptychography, this approach has the potential to enhance the accuracy and resolution of defect detection on EUV photomasks. This improvement directly supports the industry's ability to enable the production of smaller, more powerful semiconductor devices. Secondly, the use of such advanced models reduces the computational load and associated costs compared to traditional methods, making the process more economically viable and potentially lowering the barrier to adoption for manufacturers. Finally, the application of machine learning and computer vision techniques, such as deep generative modeling and image-to-image translation, provides a novel pathway to overcome the challenges of high data volumes and dynamic range in lithography, thus improving the efficiency and speed of the imaging process.