This PDF file contains the front matter associated with SPIE Proceedings Volume 12472 including the Title Page, Copyright information, Table of Contents, and Conference Committee Page.

As part of the National Agenda for Quantum Technology, QuTech (TU Delft and TNO) has agreed to make quantum technology accessible to society and industry via its full-stack prototype: Quantum Inspire. This system includes two different types of programmable quantum chips: circuits made from superconducting materials (transmons), and circuits made from silicon-based materials that localize and control single-electron spins (spin qubits). Silicon-based spin qubits are a natural match to the semiconductor manufacturing community, and several industrial fabrication facilities are already producing spin-qubit chips. Here, we discuss our latest results in spin-qubit technology and highlight where the semiconducting community has opportunities to drive the field forward. Specifically, developments in the following areas would enable fabrication of more powerful spin-qubit based quantum computing devices: circuit design rules implementing cryogenic device physics models, high-fidelity gate patterning of low resistance or superconducting metals, gate-oxide defect mitigation in relevant materials, silicon-germanium heterostructure optimization, and accurate magnetic field generation from on-chip micromagnets.

Background: In practice, manufactured lithography masks come with a certain number of unintended defects. Therefore, mask fabrication is accompanied by a subsequent repair, performed via etching or material deposition by a gas-assisted focused electron beam. Aim: The goal of this work is to assess the lithographic impact of mask defects and corresponding repair by simulations. Approach: For this purpose, a novel analytical method was developed to retrieve exact repair shapes f rom scanning electron microscope (SEM) images of the mask patterns. A developed method, based on computer vision and image processing, is combined with a dedicated artificial intelligence (AI) network trained to detect defective contact and line/space patterns from mask SEM images. Lithography simulations were done for 3D masks derived from the real SEM images. Results: 3D masks with the 13 nm lines and 18 nm contact holes are simulated, and corresponding aerial images are computed. Different typical defects are investigated and demonstrate the robustness and effectiveness of the developed software. Conclusions: The developed analytical algorithm demonstrates a stable and accurate extraction of repair shapes from given mask SEM images. Using our simulation procedure, the impact of each defect from a variety of SEM images was assessed, and lithographic performance after a repair was predicted. In the simulations, the determination of the optimum repair shape is implemented as a two-step procedure providing a large overlap of process windows of defect-free and repaired features, hence high-quality lithography output.

The imaging performance of a mask in EUV lithography is governed by the optical properties of the absorber material, namely the refractive index n and extinction coefficient k, and by its thickness. The imaging metrics viz. Normalized Intensity Log Slope (NILS), Telecentricity Error (TCE) and Best Focus Variation (BFV) through pitch, exhibit a tradeoff. In addition, the choice of illumination has a significant influence on these imaging metrics. Most of the previous studies have focused on either reflectivity or phase shift induced by the absorber to determine the optimum absorber thickness. The limitation of this approach is that the structure of the patterns on the mask is ignored. This simulation study is intended to facilitate the selection of the optimum absorber thickness with an emphasis on diffraction order analysis and the impact of illumination source shape using a case study of TaCo alloy. The behavior of imaging metrics is investigated as a function of absorber thickness in combination with illumination source shapes recommended in the literature. Maximal NILS, TCE within specified limits, balancing of diffraction order amplitudes with a minimum phase difference, and throughput criterion, are the important parameters that are considered when selecting the optimum absorber thickness. We evaluate and compare the through pitch imaging performance of TaCo alloy with recommended thicknesses, to that of the reference TaBN 60 nm absorber using Leaf shape Dipole (LDP), Inner Half Leaf shape Dipole (IHLDP) and Outer Half Leaf shape Dipole (OHLDP) for Line and Space (LnS) pattern with trench width of 10nm and the smallest pitch of 20 nm. The study confirms that TaCo alloy exhibits improved NILS and lower BFV compared to the reference TaBN 60 nm absorber.

This paper presents contour-based methods to assess mask variability. Mask certification depends on the measurement reliability and on criteria relevance. By now, ST and its maskshop partners rely mostly on CDSEM measurements for mask certification. However, this kind of metrology has limitations and, looking at the future, we think it would be timely to search for metrology which bypass those limitations. That is why we are looking at 2D metrology ^[1], especially to area and contour measurements ^[2] on SEM images using extracted contours. Thanks to the added value of 2D metrology, we expect to assess mask variability, mask uniformity and pattern fidelity. We also take the opportunity to compare the results on two FOVs (field of view) from the images provided by mask shops. Finally, we also intend to automate the whole measurement process to make it easier to use.

The shift of semiconductor industry applications into demanding markets as spatial and automotive led to high quality requirements to guaranty good performances and reliability in harsh environments. As reliability is directly related to a well-controlled process, characterizing the local overlay and its variations inside the chip itself becomes a real asset. While most available in-chip overlay metrologies require dedicated target or dedicated tools, we developed a new method that aims to augment the current SEM tool park into measuring the local overlay directly on the product. In a previous proceeding, this on-device and target-free overlay measurement based on CD-SEM contours has been assessed on SRAM patterns and showed promising results. The work presented here pushes forward this assessment using SEM synthetic images generated from the open-source Nebula simulator of electron-matter interaction. From a layout, a 3D geometry of the measured pattern can be generated, with materials and interfaces carefully defined. Then, a GPU-accelerated Monte-Carlo model simulates in tens of seconds the SEM image. This fast generation of images enables the use of synthetic SEM images in a digital twin system: they can be used to characterize and to challenge the overlay metrology, before applying it to real products. Indeed, a known overlay can be programmed in these images. This way the performances of the measurement algorithm can be assessed with a ground truth reference. Firstly, imaging parameters such as pixel size and noise have been varied in a wide range. This demonstrated a good accuracy and precision inside a defined measurement window with a coefficient of correlation above 0.996 and an offset lower than 0.2nm. In a second part, the influence of the pattern measured has been investigated and experimental results on SRAM could be reproduced using synthetic images. The origin of the loss of sensitivity has been identified and improvements in the contour extractions and used template led to a correlation with a slope of 1.03, an offset of 0.1nm and a Root Mean Square Deviation of 1.36 nm. Finally, the developed digital twin already showed behaviors in the measurement that were hidden in the on-wafer experiments, that helped assessing the method and which will be used in the future to define guidelines for template-based SEM-OVL measurements.

In this work, we focus on the application of the "three-state lithography model" developed for the production of 3D-topographies in photoresist through grayscale lithography. We demonstrate in detail how the variables of the model are determined and optimized in a parameter definition procedure. The principle work ow for a automated mask generation is shown on a pyramid sample structure. Additionally, we tested a top and bottom anti-reflective coating for the use of surface smoothening. Experiments reveal bottom anti-reflective coating as method of choice to smoothen the surfaces on manufactured 3D-topographies.

Controlling the Local CD Uniformity is important for the implementation of EUV lithography in high-volume production. Spatial frequency breakdown of stochastic effects and identification of stochastic noise contributors may help us to understand the current performance and suggest possibilities and pathways for future improvement. In this work, we look for potentially hidden sources of systematic local variability by collecting and analyzing CD metrology data over lengths greater than a single SEM field of view (FOV). Fourier analysis of the CD data is used to identify any systematic variability. This work will enable a more accurate breakdown of local variability. Additionally, using the length scale of any observed systematic signal we can attempt to trace back the origin and reduce or eliminate its source.

Background: Stochastic effects in DUV lithography are manifested by variabilities in critical dimension (CD), in placement or in shape. A combination of these very local variabilities can lead to yield killer open contacts. Traditionally, opens are massively measured with Voltage Contrast (VC) tools, returning the defects density after etching and metal filling. Aim: A set of contour-based metrics for the quantification of stochastic effects in DUV has already been presented. In this paper, we correlate these metrics and open count to predict failure risk. Approach: With an in-depth analysis of post-lithography CD-SEM images, we investigate if variabilities inside the metrology target are forerunners of open risk inside the product. It is challenging because of the difference between the surface inspected with defectivity tools and the one measured with CD-SEM. Results: We applied the methodology on contacts of a 28 nm node technology, on a Focus Exposure Matrix (FEM) wafer, to obtain post-lithography contour-based metrics mappings. A new metric has been computed: the classification of shapes inside the image. After post-processing, the correlations between contour-based metrics and the log value of open count are presented. A threshold value of size variability emerges above which open risk is too high, enabling process monitoring. Conclusion: As contour-based metrology offers complementary metrics not only related to CD metrology, we can now predict open probability with new indicators coming from traditional CD-SEM images. This early detection of an atypical situation allows the process assessment.

In this paper, we describe a lithographic technique of exposing complex patterns with an advanced resist processing that connects the high resolution of electron beam lithography and the fast exposure of optical i-line stepper lithography via an Intra Level Mix and Match (ILM&M) approach. The key element of our approach is that we use two successive exposures on one single resist layer directly followed by a single resist development. Process and resist characterization of negative tone resist ma-N 1402 as well as a resolution study for each lithographic tools involved. Lithographic performance of negative tone resist ma-N 1402 has shown structures with dimensions of 55 nm with 300 nm pitch for ebeam lithography (VISTEC SB254, shaped beam) and 350 nm structures for i-line stepper (Nikon NSR 2205i11D). Resist footing problem in structures exposed by i-line stepper is solved by introducing a 200 nm thick bottom antireflective coating AZ BARLI II in ILM&M resist processing sequence. A general processing recipe for electron beam/i-line stepper ILM&M with negative tone resist ma-N 1402 is successfully developed and patterns with different dimensions ranging from sub 100 nm to μm scale were reproducibly fabricated on the same resist layer.

EUV lithography is currently setting the pace for the semiconductor industry’s expectations on future progress towards the 3nm node and beyond. This technology also defines the upcoming challenges for equipment providers upstream and downstream of the production line among which wafer-level overlay and CD error requirements stand out most prominent. Registration errors on the mask, both local (mid-range) and global (long-range), contribute to overlay errors on the wafer. Here, we will present novel calibration strategies for the IMS Multi-Beam Mask Writer (MBMW) by ZEISS PROVE measurements to meet the mask registration requirements: First, we showcase how we can efficiently leverage the high precision, resolution and fast capture time of the PROVE tool to allow for extensive control and tuning of MBMW properties that affect local registration (LREG) such as systematic residual errors originating from the electron beam optics. Second, we provide insights into the MBMW Registration Improvement Correction (RIC) calibrated with PROVE technology. This feature allows removing remaining systematic local registration errors in the MBMW electron beam array field (82μm x 82μm) resulting in LREG improvement by 30% from 1.2nm to 0.8nm three-sigma. Third, we show how the PROVE technology can be applied efficiently for the calibration of the MBMW’s Thermal Expansion Correction (TEC) that allows compensating systematic global registration errors originating from thermal-mechanical deformations of the mask during the writing process.

Any modeling of an interaction between photons and matter is based on the optical parameters. The determination of these parameters, also called optical constants or refractive indices, is an indispensable component for the development of new optical elements such as mirrors, gratings, or lithography photomasks. Especially in the extreme ultraviolet (EUV) spectral region, existing databases for the refractive indices of many materials and compositions are inadequate or are a mixture of experimentally measured and calculated values from atomic scattering factors. Synchrotron radiation is of course ideally suited to verify such material parameters due to the tuneability of photon energy. However, due to the large number of possible compounds and alloys, the development of EUV laboratory reflectometers is essential to keep pace with the development of materials science and allow for inline or on-site quality control. Additionally, optical constants are also essential for EUV metrology techniques that aim to achieve dimensional reconstruction of nanopatterned structures with sub-nm resolution. For this purpose, we studied a TaTeN grating created on an EUV Mo/Si multilayer mirror, to mimic a novel absorber EUV photomask. We present here a first reconstruction comparison of these structures, measured by EUV scatterometry at the electron storage ring BESSYII and with a laboratory setup of a spectrally-resolved EUV reflectometer developed at RWTH Aachen University. Both approaches differ in several aspects reaching from setup size to spectral quality (brilliance, bandwidth and coherence) as well as the measured and simulated data.

Due to the increasing availability of Multi Beam Mask Writers (MBMW) and the associated wafer lithography benefits, efficient handling of curvilinear mask shapes in the post tape-out flow becomes increasingly important. The presence of complex curvilinear mask shapes generated by Inverse Lithography Technology (ILT), or similar techniques requires new features and algorithms, especially in the context of applying Mask Process Correction (MPC). This paper discusses the status of Curvilinear Mask Process Correction (CLMPC) and gives a brief outlook into the near future of handling curvilinear mask shapes in the post tape-out flow. A summary of MRC requirements for incoming data and implications for mask inspection are included as well to put CLMPC into the right context. Further, approaches for CLMPC file size compaction along with new methods of data representation based on native curve formats are also discussed.

Some specific applications, such as optical devices, require non-conventional layouts. In this context, the known OPC solutions developed during decades and optimized for CMOS planar applications are facing significant challenges. Standard design files format as well as OPC algorithms are indeed suitable for 0-45-90° edges (also called Manhattan layouts) and other angle edges can lead to bad OPC results, huge run time, large file size, and even run crashes. While innovative developments are on going from OPC suppliers’ side, we have to use smartly the conventional OPC platforms to achieve accurate, fast and cost-effective solutions. Taking the example of optical diffusers application, we will discuss the implementation of such an OPC flow, including rule-based correction, SRAF insertion, model-based correction, and mask sign-off strategy.