Editor-in-Chief Harry Levinson examines how the use of curvy features exemplifies that advanced lithography continues to be a dynamic field.

Guest Editors Patrick Naulleau and Gregg Gallatin introduce the Special Section on Metrology for EUV.

Mirror-based and zone plate-based imaging systems are being used in actinic extreme ultraviolet (EUV) reticle review tools. With regard to zone plates, a short working distance is advantageous in terms of the required spectral bandwidth, manufacturability, and potential throughput and imaging performance. Zone plates therefore typically have a short working distance. The industry has adopted the use of an EUV pellicle to protect the photomask. Imaging photomask through-pellicle requires a working distance larger than 2.5 mm. A zone-plate-based EUV mask microscope with a 3-mm working distance has been commissioned at beamline 11.3.2 of the Advanced Light Source. Through-pellicle imaging at an exposure time of 2 s is demonstrated. The instrument achieves an image contrast of 95% on large features on a photomask with a tantalum-based absorber. Imaging down to 45-nm half pitch (mask scale) is demonstrated. A NILS of 2.55 is achieved on 60-nm half-pitch (mask scale) lines and spaces. These results demonstrate that zone-plate-based imaging systems can meet the requirements of an actinic EUV mask review tool in terms of imaging performance and throughput in an instrument compatible with EUV pellicles.

The interaction of extreme ultraviolet (EUV) light with matter is a critical step in EUV lithographic processes, and optimization of the optical material properties of all elements in the lithographic chain (from optical coatings and pellicles to photoresists) is crucial to harnessing the full power of EUV lithography. To optimize these materials, accurate measurements of EUV absorption and reflection are needed to extract the corresponding actinic optical properties and structural parameters. Here, we report on actinic EUV metrology-based absorption and reflection measurements enabled by coherent table-top EUV sources based on high-harmonic generation. We demonstrate the capabilities and flexibility of our setup with measurements on crystalline films, photoresist systems, and carbon nanotube membranes and provide extracted optical parameters, absorption kinetics, and 2D transmission maps, respectively. These results showcase the power of lab-based actinic inspection methods based on compact, coherent EUV sources for providing crucial data for material optimization and lithographic simulation.

EUV lithography is rapidly being pushed to its resolution limit, where tradeoffs are heightened among resolution, throughput, and stochastics. Mitigation strategies include attenuated phase shift masks (aPSMs) and thinner high-k absorbers. Furthermore, multilayer bandwidth and phase shift may need to be reassessed. All these improvements relate to mask 3D effects (M3Ds), arising from several causes: first, phase shift versus pitch, which sets the aPSM target phase shift around 1.2π instead of π; second, absorber thickness effects, which directly relate to the promise of high-k absorbers; and third, multilayer effects such as bandwidth and phase versus angle. We quantify these effects in simulation for different EUV scanner generations (0.33 and 0.55 NA). Moreover, we demonstrate the measurement of these effects with the EUV Tech ENK (EUV N&K and Phase Measurement Tool) using actinic scatterometry. The complexity of M3Ds suggests that new metrics of multilayer and absorber performance beyond reflectivity will need to be considered. Actinic scatterometry provides a promising route to measuring M3Ds due to its sensitivity to wavelength, angle, and feature size.

Guest Editors Linyong (Leo) Pang and Danping Peng summarize the Special Section on Curvilinear Masks.

The photomask industry is experiencing a fundamental shift from Manhattan masks to curvilinear masks. In the recent lithography and mask technology conferences, there were many papers and talks on curvilinear masks, curvilinear optical proximity correction, curvilinear inverse lithography technology (ILT), curvilinear mask process correction, and curvilinear mask formats. Step by step, the photomask industry has started a transition from Manhattan to curvilinear, enabled by the adoption of the new multi-beam mask writers and the advent of practical full-chip curvilinear ILT. The benefits of curvilinear masks go much deeper than is immediately obvious. We will share our insights on why the mask world is moving toward curvilinear mask shapes. We will demonstrate that curvilinear masks are more reliably manufacturable. We will evaluate the benefits of curvilinear in terms of process window, mask rules, mask error enhancement factor, and mask variation.

We demonstrate through comprehensive analysis and empirical evidence the inherent advantages of pixel-based computing for curvilinear photomasks when using a graphics processing unit (GPU)-based platform. The Single Instruction, Multiple Data nature of GPUs enables fast pixel-based computations. We study the advantages of using GPU acceleration for pixel-based computing in various mask processing and verification steps. We highlight the natural runtime predictability of pixel-based computing, which is in the order of number of pixels, or O(p), irrespective of the complexity of the mask shapes. We also emphasize that pixel dose equivalence and information theory provide a mathematical basis for the practical accuracy of a pixel-based approach toward mask data preparation. The Nyquist limit guides the mask rules for masks written using multi-beam writers as the mask is essentially sampled into pixels. This sampling acts as a low-pass filter. We also perform a case study of smooth shapes to show how bandlimited shapes enabled by pixel-based multi-beam writers are more manufacturable. We conclude that the O(p) approach for GPU acceleration enables accurate and practical data processing for curvy masks governed by information theory as we leap into an increasingly complex curvy world.

Single-exposure extreme ultraviolet (EUV) lithography is quickly advancing as a replacement for argon fluoride immersion (ArFi)-based multiple patterning approaches for printing the most critical features in semiconductor devices. However, the dimensional scaling of EUV lithography patterns is hampered by stochastic effects, resulting in rough patterns and increased defectivity. A promising solution to mitigate these stochastic pattern variations is complementing top–down EUV lithography with bottom–up directed self-assembly (DSA) of block copolymers (BCPs). We investigated an EUV + DSA complementary process for the rectification of pitch 28-nm line/space (L/S) patterns on high-volume manufacturing compatible processing tools. We found that several DSA material and process parameters contribute to minimize the roughness of the rectified patterns. In particular, the BCP size and film thickness are the most critical parameters. In terms of defectivity, a combination of optical inspection and e-beam review pointed out that dislocations are not a major concern for EUV + DSA patterning due to the fast assembly kinetics. Instead, bridge and cluster defects are the main defect modes and minimum defectivity can be achieved by controlling the geometry of the guide pattern. Finally, the impact of pattern density multiplication by DSA was assessed by comparing the performance of the current EUV + DSA rectification process to the ArFi + DSA technology, both for generating a pitch 28 nm L/S pattern.

We have developed an imaging spectroscopic reflectometry (ISR) method based on hyperspectral imaging and deep learning to detect defects in the bottom region of high-aspect-ratio nanostructures. ISR enables fast and non-destructive imaging of the bottom critical dimension (BCD) of channel holes (CHH) on a chip die of vertical NAND (V-NAND). A supervised learning model is built to predict the BCD by associating a pre-measured hyperspectral cube with scanning electron microscopy images after decapsulation of the top of the sample. The BCD predicted by ISR shows a high correlation of R2=0.72 with the actual BCD, and the distribution of CHH not open (NOP) defects on the chip die identified by bright field inspection after decapsulation is consistent with the BCD image obtained by ISR. In addition, ISR can detect defects that occur at arbitrary positions relative to the optical critical dimension (OCD) of the die. On fully integrated V-NAND chips, the ISR result showed a high correlation (R2=0.82) with the failure rate caused by CHH NOP, while the conventional spot OCD showed only R2=0.41. Thus, ISR can be used to optimize the etch process for weak wafer edge regions and to detect defective etch equipment.

Metrology plays a crucial role in semiconductor manufacturing by providing accurate and precise measurement and characterization of critical parameters. With the development of high-resolution extreme ultraviolet lithography (EUVL) processes, critical dimensions are shrinking to sub-10 nm. Resist materials encounter the challenge of providing heightened sensitivity and a handle on exacerbating stochastic variations. A comprehensive understanding of the chemical profile of the latent image is pivotal for mitigating stochastic effects and optimizing pattern quality. However, the subtle differences in chemistry between the exposed and unexposed regions of the resists make it extremely challenging to characterize the latent images with sub-nanometer precision. Here, we develop the metrology with critical-dimension resonant soft X-ray scattering (CD-RSoXS) to probe the chemical profiles of latent images stored in resist after exposure. The combination of absorption spectroscopy and enhanced scattering contrast makes it possible to characterize the subtle structural and chemical variations in the latent image. Moreover, the results of the measurements are compared with the simulations with a finite element method–based Maxwell solver to extract a detailed profile of the latent and developed images. We demonstrate that the CD-RSoXS technique can provide valuable insights into the high spatial resolution and local chemical sensitivity simultaneously, which is crucial to understanding the resolution limits and stochastic effects in EUVL processes.

Advanced semiconductor devices are moving toward three-dimensional (3D) geometries due to scaling demands and performance requirements. Non-destructive metrology necessary for process control of 3D structures must be advanced to facilitate their transition from technology development to high-volume manufacturing. Thin film metrology using Mueller matrix spectroscopic ellipsometry (MMSE) and X-ray diffraction (XRD) film metrology, as well as patterned structure metrology using optical critical dimension (OCD) and X-ray fluorescence (XRF) techniques, have proved capable of measuring the Si/Si1−xGex superlattices and gate-all-around transistor test structures. Because these techniques are indirect, their limitations associated with superlattice device structures need to be further understood. To understand these limitations, a four-superlattice layer Si/Si1−xGex structure was measured at four process steps: as an unpatterned film stack, after anisotropic column etch, and at low and high levels of cavity etch. Thin film samples were analyzed with XRD and MMSE, and patterned samples were analyzed using OCD, as well as XRF. A model was developed describing the primary and secondary process effects on the structure. This was evaluated for consistency on datasets collected at different measurement azimuth angles. Square error–based methods were evaluated to quantify OCD model detectability of fit variable step deviations, as well as sensitivity relative to the model to measurement error. OCD and XRF results were compared with reference scanning transmission electron microscopy (STEM) images of nanowire test structure lamellae. Dual-azimuth fit OCD models were found to be within 0.3 nm of the STEM reference mean.

Spectroscopic ellipsometry has been widely used as one of the metrology methods of choice in various industries: microelectronics, photovoltaic, optoelectronics, flat panel display, etc. We present an example of the characterization of dielectric multilayer structures on the substrates with unintended surface modifications involving macroscopic roughness. We assume that under our inspection conditions, the effect of macrorough surfaces can be treated as the presence of a specially designed overlayer on top of the ordinary substrate. A systematic procedure was then proposed to simulate the dielectric response of the overlayer. This approach is quite useful in a practical sense and provides more accurate process monitoring and control in a production environment.

We focus on detecting defects during hotspot monitoring using scanning electron microscope (SEM) images from the after-development inspection (ADI) phase in semiconductor manufacturing. A primary challenge lies in the significant imbalance between defective and defect-free samples, complicating binary classification. To address this, data augmentation and the synthetic minority over-sampling technique (SMOTE) are applied to generate additional samples for the minority class. A two-stage learning approach is then proposed: first, deep convolutional neural networks are trained on SMOTE-synthesized data to develop a pre-trained defect classification model. Subsequently, this model is fine-tuned with raw, augmented data to enhance its performance on targeted data. Extensive experiments demonstrate that this method improves efficiency, reduces computational costs, and outperforms traditional techniques. The results show that integrating SMOTE-based augmentation with a two-stage learning process mitigates data imbalance and improves defect detection accuracy. This proposed methodology offers a robust solution for automating SEM image inspections during hotspot monitoring, addressing the scalability challenges posed by increasing design complexity and emerging defect types in advanced semiconductor processes.

In the case of low-k1 imaging, critical dimension (CD) variation on extreme ultraviolet (EUV) mask enhances on wafer, which is called mask error enhancement factor (MEEF). CD variation for tighter line and space (L/S) features on the photomask has taken up a large fraction of wafer CD variation. Understanding the mask contribution to EUV lithography (EUVL) is one of the important initiatives to further optimize 0.33-NA EUVL employability. The optical properties of the EUV mask absorber have the potential to improve Normalized Image Log-Slope (NILS). Low-n attenuated phase-shift mask (attPSM) is expected to mitigate mask three-dimensional effects and to enhance the image contrast compared with a traditional Ta-based mask. We experimentally evaluated the MEEF on the drawn feature size, illumination, and mask absorber properties to characterize the impact of optical factors on MEEF using the NXE:3400B EUV scanner. This paper covers two types of mask absorber materials which are Ta-based and low-n attPSM. We experimentally demonstrated that a low-n mask is expected to improve the balance of MEEF at vertical L/S through pitch due to suppression of best focus variation. We also tried to identify the suitable absorber optical properties in terms of MEEF and NILS on both 0.33/0.55-NA EUVL through lithographic simulations.

Erratum corrects errors in STEM measurements in Table 1.