apply nanofabrication techniques

This work addresses the time-consuming nature of parameter extraction and inspection algorithm development caused by complex 3D structures and EUV stochastic defects in advanced semiconductor nodes. We propose a generalizable deep learning framework based on high-speed scanning probe microscopy images. By leveraging machine learning to replace conventional customized image processing algorithms, this framework enables automated feature recognition, critical dimension parameter extraction, and nanoscale defect detection across diverse layers, structures, and devices. The core innovation lies in establishing a unified intelligent analysis paradigm that significantly shortens algorithm development cycles while enhancing inspection accuracy. Consequently, the proposed approach effectively satisfies the stringent requirements of high-throughput in-line process control in semiconductor wafer fabs.

In nanoscale modular electronics (ME/NE) design automation, a fundamental trade-off exists between manufacturing time and bus length during component placement and routing. This work presents the first systematic exploration of this time–performance Pareto space, introducing an adaptive hierarchical algorithmic framework that integrates partitioning, floorplanning, stochasticity-aware component placement, and precise wire-printing–guided routing. The method explicitly accounts for the inherent randomness in micro/nanoscale deposition processes and supports dynamic reweighting of optimization objectives. Experimental results demonstrate that, with only a 21% increase in total wirelength, end-to-end fabrication time improves by up to 108×; alternatively, circuit performance can be prioritized without compromising manufacturability. This approach significantly enhances flexibility, scalability, and process–design co-optimization for heterogeneous nanocircuit synthesis.

In plasma etching of semiconductors, significant spatial variation in etch depth and process reliability degradation—driven jointly by aleatory and epistemic uncertainties—pose critical challenges. To address this, we propose an integrated uncertainty quantification and optimization framework combining heteroscedastic Gaussian process (hetGP) surrogate modeling with reliability-based design optimization (RBDO). Our approach innovatively decouples spatial variability, parametric perturbations, and model-form epistemic uncertainty, enabling explicit decomposition and synergistic modeling of multi-source uncertainties. Leveraging limited experimental data, the method accurately identifies optimal process parameter sets, reducing the spatial standard deviation of etch depth by over 35% on average while ensuring process robustness at ≥95% confidence. The framework demonstrates strong transferability and is readily extensible to other micro/nano-manufacturing processes, such as photolithography.

This work addresses the tight coupling between design intent and printer-specific representations in heterogeneous manufacturing, which hinders cross-platform reuse. The authors propose a novel compiler architecture that models fabrication-aware design as a staged, type-directed lowering process, decoupling source design, attribute translation, and backend compilation to enable manufacturing-agnostic expression. Introducing compiler paradigms to heterogeneous manufacturing for the first time, the approach unifies volumetric information—such as material composition, hardness, and color—through implicit geometry and typed spatial attribute fields, automatically generating voxel stacks, G-code, or slicer projects. Experiments demonstrate successful fabrication of complex objects embedding CT data, Shore hardness fields, and full-color fields on both material jetting and extrusion platforms, validating cross-process reusability. The accompanying Python toolkit is publicly released.

High-quality, physically plausible annotations for lithographic defect detection are scarce, hindering robust AI-based quality inspection. Method: We propose a design-driven, physics-constrained defect image generation framework: defects are synthetically placed on layout patterns using mathematical morphology; realistic images are acquired via DMD-based lithography and optical microscopy; and corresponding pixel-level contour masks are generated synchronously. Contribution/Results: We introduce LithoDefect-1K—the first open-source lithographic defect dataset comprising 13,365 instances across four defect types, spanning 3,530 images. Our paradigm enables controllable, interpretable, and high-fidelity annotation generation. Evaluated on Mask R-CNN, it achieves mean AP@0.5 of 78.6%–84.2%, outperforming Faster R-CNN by 34%–42%, thereby demonstrating substantial performance gains from physics-guided data synthesis for AI-powered inspection.

This study addresses the absence of evaluation benchmarks for language model agents in two-dimensional material flake field-effect transistor (FET) layout assessment by constructing the first benchmark comprising 128 tasks. Agents are required to generate GDSII-formatted layouts from microscopic image contours, with a deterministic verifier introduced to ensure geometric compliance and contour integrity. Methodologically, models such as GPT5.6-Luna are integrated with ReAct-3 and Plan-and-Execute strategies, utilizing Python scripts to generate polygon paths rendered into GDSII format, thereby supporting multi-flake and hole-containing complex tasks for the first time. Experimental results demonstrate that the optimal configuration achieves an 80.5% solution rate, 43.8% consistency, and approximately 60% expert audit acceptance, significantly outperforming single-pass planning baselines. These findings validate both the solvability of this task and the effectiveness of the proposed benchmark.

"This study addresses the challenges in data comparison and reutilization within the field of materials science, specifically due to the inconsistent reporting formats of atomic layer deposition (ALD) and atomic layer etching (ALE) experiments and simulations. The work proposes and expertly reviews four JSON schemas based on the QUDT standard, designed to standardize the description of materials, process conditions, configurations, and results in ALD/ALE processes, ensuring data consistency and interoperability. The innovation lies in the development and application of these specialized JSON schemas, which, through the use of a schema-miner toolset, enable effective extraction and structured representation of literature content, thereby promoting knowledge sharing. Additionally, the study provides a comparative analysis of the schemas and publishes related structured records on the ORKG platform, demonstrating their potential for guiding information extraction tasks."

This study addresses the challenges of complex rule interactions and fragmented expert knowledge in lithographic optical proximity correction (OPC) by proposing a large language model (LLM)-based virtual process engineer framework. Integrating a skill library with layout analysis, the method constructs evidence chains from measurement feedback under fixed model weights, leveraging an LLM reflection mechanism to guide iterative recipe editing for self-evolving corrections. The framework is further integrated with commercial tools to enable closed-loop evaluation. Experimental results demonstrate that the proposed approach significantly reduces the maximum edge placement error (EPE) on both Poly and Metal1 layers, with all final recipes satisfying quality constraints. This work provides an effective paradigm for intelligent lithography optimization.

This study addresses the challenges in nanomedicine development posed by the high sensitivity of nanoparticle size and polydispersity index to process parameters, which renders traditional trial-and-error approaches costly and time-consuming. To overcome this, the work integrates microfluidic experimentation with expert knowledge and, for the first time, incorporates shape constraints into a machine learning model. By leveraging a small amount of low-cost surrogate data, the approach accurately predicts critical physicochemical properties of lipid-based nanoparticles. The method achieves high-fidelity modeling of both liposome and lipid nanoparticle size and dispersity with minimal experimental samples, substantially reducing the need for extensive screening experiments and enabling rational, efficient design of nanomedicine manufacturing processes in continuous-flow systems.

This study addresses the challenges of manual operation, low throughput, and difficult statistical analysis in 4D-STEM data acquisition by proposing an interactive automated workflow with machine-driven decision-making. Integrating machine learning, programmatic instrument control, and automated ptychography screening, this approach enables intelligent acquisition and efficient interpretation of nanobeam diffraction and ptychographic data. The workflow successfully collected hundreds of datasets, effectively translating high-throughput instrumentation into statistically meaningful atomic-scale insights. Specifically, it reveals orientation and phase distribution patterns within platinum nanoparticle ensembles while precisely quantifying atomic-resolution phase and lattice strain at the single-grain level, thereby bridging the gap between raw data volume and rigorous quantitative materials characterization.