Editor-in-Chief Harry Levinson reflects on the evolution of the Photomask Japan meeting on the occasion of its 30th anniversary.

Elementary analysis shows that maximizing the aerial-image contrast in printing dense lines and spaces leads to a monotonically increasing relationship between (1−n) and κ, where n and κ are real and imaginary parts of the refractive index of the absorber material of the photomask, respectively, and this relationship is affected by the space-to-pitch ratio in the mask absorber. For the tantalum boron nitride mask absorber presently in use, improvement in image contrast can be made by displacing the illuminating dipoles away from their nominal positions in conjunction with a focus offset.

Tatiana Kovalevich of imec discusses her recent paper, Evaluation of Lines and Spaces printing and general understanding of imaging with dark field low-n mask, with JM³ Editor-in-Chief Harry Levinson.

In recent years, curvilinear mask technology has emerged as a next-generation resolution enhancement method for photomasks, providing optimal margins by maximizing the degree of freedom in pattern design. However, this technology presents challenges in defining the layout design rule limits based solely on geometric information, such as width, space, and corner-to-corner. With the introduction of multi-beam mask writers for curvilinear pattern production, a distinct set of defects that are difficult to pre-detect by conventional mask rule check have occurred, as these cannot be explained by geometry terms alone. In this study, we propose a deep learning-based mask check method, named mask deep check (MDC) for pre-detect defects in inspection. The proposed vector graphics transformer (VGT) uses the state-of-the-art transformer architecture, drawing an analogy between the vertices of curvilinear patterns and words in natural language. We demonstrate improved performance of VGT-based MDC compared to a traditional rule-based approach and a convolutional neural network-based MDC method. Importantly, VGT exhibits robustness in recall, ensuring that defective patterns are not misclassified as normal, thereby preventing missed defects. Moreover, by employing attention maps to visualize VGT results, we gain explainability and reveal that mask defects may arise from issues related to the fabrication of specific designs, rather than being solely attributable to geometric features. VGT-based MDC contributes to a better understanding of the challenges associated with curvilinear mask technology and offers an effective solution for detecting mask defects.

One of the most significant changes happening now in the photomask industry is the move from Manhattan masks to curvilinear masks. This change is enabled by multi-beam mask writers and is motivated by benefits to both mask and wafer by using curvilinear mask patterns. Following the breakthrough in full-chip curvilinear inverse lithography technology using GPU acceleration, the adaptation of curvilinear optical process correction (OPC) from edge-segment OPC, and the MULTGON format, wafer fabs are expanding the layers using curvilinear masks at a rapid pace. In the mask-making flow, mask writing, mask critical dimension (CD) scanning electron microscopy (SEM), mask inspection, mask aerial image review, and mask repair are all done in the image or pixel domain, so these processes can handle curvilinear masks naturally. However, for mask metrology there are some unresolved issues, especially the traditional mask CD spec, which is a shared standard between mask shop and wafer fabs and becomes unknown for curvilinear mask patterns. In this work, we propose an equivalent CD spec for curvilinear masks, and we will also use this proposed spec to demonstrate curvilinear masks have smaller mask variation than Manhattan masks.

A critical factor in the fabrication of complex nano- and microstructures with high quality and reproducibility is the determination of a suitable working point. This applies particularly to lithography, which is the basis for transferring the desired patterns onto the substrate. For this reason, we present a generic process optimization methodology that has been successfully applied to four chemically amplified positive and negative tone electron beam lithography photoresists with different sensitivities. The method is iterative and designed for the best possible results with a minimum use of resources. This is accomplished by identifying the critical key factors in photoresist processing using contrast curves and determining their impact. Starting with the most influential bake parameter, the maximum effect is achieved. The method used is similar to the Bossung plot procedure and aims for a maximum process window. After the bake parameters, the fundamentals of development kinetics are discussed, and a method for determining an appropriate development time is presented. A mask making approach is then used to investigate the ideal exposure conditions. This includes the determination of an appropriate base dose in conjunction with proximity effect correction and sizing. The evaluation of this method is demonstrated by critical dimension linearity plots and scanning electron microscope cross sectional analysis of resist profiles. The results presented demonstrate the universality of the optimization approach.

A silicon containing antireflection coating (SiARC) can effectively suppress standing waves when the photoresist thickness is limited, but the relationship between polymer structures and optical parameters is not clear. High-silicon-content polysiloxanes were synthesized by sol–gel processing, and SiARC was prepared by spin-coating method. The influence of Si/Ph on the optical parameters (n and k values) of SiARC was investigated. Finally, the effect of silicon content on the etching rate of fluorine ion and oxygen ion was studied.