Search for: All records

We report what we believe to be a novel and unique approach for achieving high-performance and broadband THz phase shifting based on spatially-resolved photoconductivity modulation (SRPM). By changing the illumination area on a hybrid Au-Ge mesa-array (AGMA) structure in front of an indium tin oxide (ITO) layer for local photoconductivity modulation, the phase difference between the incident- and reflected-waves can be tuned nearly continuously with extremely low reflection loss. For a prototype demonstration, a photonically-driven THz phase shifting device based on the WR-5.1 (140-220 GHz) waveguide configuration was designed, modeled and simulated. To achieve phase tuning in the range of 0° to -180° at 180 GHz (band center frequency), a mesa-array consisting of 12 × 6 unit cells (each 105 μm × 105 μm) was designed, and a distancedof 250 μm between the AGMA and ITO was used. The SRPM is accomplished using computer-generated light patterns from a closely-coupled micro-LED array for through-ITO illumination, without the need for any biasing circuitry. Full wave simulation results have shown that pseudo-continuous and broadband phase shifting can be achieved in the entire WR-5.1 band, and a shifting range of 0° to -180° at 180 GHz can be realized as designed. In addition, by using light patterns of different combinations of vertical strips, a fine phase tuning step as small as ∼0.05° can be demonstrated. For all phase tuning states, the simulated reflection loss is generally less than 1 dB with low loss variation. The proposed technology for high-performance THz phase modulation is promising and powerful, while offering far more design flexibility and frequency scalability than the current state-of-the-art since it requires no biasing wires thus eliminating parasitic-related performance degradation. Therefore, this technology is suitable for the development of large-scale THz phased-arrays, reconfigurable reflectarrays, and tunable metasurfaces for dynamic beam steering/forming required in next generation (6G or beyond) wireless communications. 

Ensuring high-quality prints in additive manufacturing is a critical challenge due to the variability in materials, process parameters, and equipment. Machine learning models are increasingly being employed for real-time quality monitoring, enabling the detection and classification of defects such as under-extrusion and over-extrusion. Vision Transformers (ViTs), with their global self-attention mechanisms, offer a promising alternative to traditional convolutional neural networks (CNNs). This paper presents a transformer-based approach for print quality recognition in additive manufacturing technologies, with a focus on fused filament fabrication (FFF), leveraging advanced self-supervised representation learning techniques to enhance the robustness and generalizability of ViTs. We show that the ViT model effectively classifies printing quality into different levels of extrusion, achieving exceptional performance across varying dataset scales and noise levels. Training evaluations show a steady decrease in cross-entropy loss, with prediction accuracy, precision, recall, and the harmonic mean of precision and recall (F1) scores reaching close to 1 within 40 epochs, demonstrating excellent performance across all classes. The macro and micro F1 scores further emphasize the ability of ViT to handle both class imbalance and instance-level accuracy effectively. Our results also demonstrate that ViT outperforms CNN in all scenarios, particularly in noisy conditions and with small datasets. Comparative analysis reveals ViT advantages, particularly in leveraging global self-attention and robust feature extraction methods, enhancing its ability to generalize effectively and remain resilient with limited data. These findings underline the potential of the transformer-based approach as a scalable, interpretable, and reliable solution to real-time quality monitoring in FFF, addressing key challenges in additive manufacturing defect detection and ensuring process efficiency. 

Machine learning interatomic potential (MLIP) is an emerging technique that has helped achieve molecular dynamics simulations with unprecedented balance between efficiency and accuracy. Recently, the body of MLIP literature has been growing rapidly, which propels the need to automatically process relevant information for researchers to understand and utilize. Named entity recognition (NER), a natural language processing technique that identifies and categorizes information from texts, may help summarize key approaches and findings of relevant papers. In this work, we develop an NER model for MLIP literature by fine‐tuning a pre‐trained language model. To streamline text annotation, we build a user‐friendly web application for annotation and proofreading, which is seamlessly integrated into the training procedure. Our model can identify technical entities with an F1 score of 0.8 for new MLIP paper abstracts using only 60 training paper abstracts and up to 0.75 for scientific texts on different topics. Notably, some “errors” in predictions are actually reasonable decisions, showcasing the model's ability beyond what the performance metrics indicate. This work demonstrates the linguistic capabilities of the NER approach in processing textual information of a specific scientific domain and has the potential to accelerate materials research using language models and contribute to a user‐centric workflow. 

Abstract Automatic differentiation (AD) enables powerful metasurface inverse design but requires extensive theoretical and programming expertise. We present a Model Context Protocol (MCP) assisted framework that allows researchers to conduct inverse design with differentiable solvers through large language models (LLMs). Since LLMs inherently lack knowledge of specialized solvers, our proposed solution provides dynamic access to verified code templates and comprehensive documentation through dedicated servers. The LLM autonomously accesses these resources to generate complete inverse design codes without prescribed coordination rules. Evaluation on the Huygens meta-atom design task with the differentiable TorchRDIT solver shows that while both natural language and structured prompting strategies achieve high success rates, structured prompting significantly outperforms in design quality, workflow efficiency, computational cost, and error reduction. The minimalist server design, using only 5 APIs, demonstrates how MCP makes sophisticated computational tools accessible to researchers without programming expertise, offering a generalizable integration solution for other scientific tasks. 

The inverse design of meta-optics has received much attention in recent years. In this paper, we propose a GPU-friendly inverse design framework based on improved eigendecomposition-free rigorous diffraction interface theory, which offers up to 16.2 × speedup over the traditional inverse design based on rigorous coupled-wave analysis. We further improve the framework’s flexibility by introducing a hybrid parameterization combining neural-implicit and traditional shape optimization. We demonstrate the effectiveness of our framework through intricate tasks, including the inverse design of reconfigurable free-form meta-atoms. 

The flexibility of metal–organic frameworks (MOFs) affects their gas adsorption and diffusion properties. However, reliable force fields for simulating flexible MOFs are lacking. As a result, most atomistic simulations so far have been carried out assuming rigid MOFs, which inevitably overestimates the gas adsorption energy. Here, we show that this issue can be addressed by applying a machine-learning potential, trained on quantum chemistry data, to atomistic simulations. We find that inclusion of flexibility is particularly important for simulating CO2 chemisorption in MOFs with coordinatively unsaturated metal sites. Specifically, we demonstrate that the diffusion of CO2 in a flexible Mg-MOF-74 structure is about one order of magnitude faster than in a rigid one, challenging the rigid-MOF assumption in previous simulations. 

Graphene aerogel (GA), a 3D carbon-based nanostructure built on 2D graphene sheets, is well known for being the lightest solid material ever synthesized. It also possesses many other exceptional properties, such as high specific surface area and large liquid absorption capacity, thanks to its ultra-high porosity. Computationally, the mechanical properties of GA have been studied by molecular dynamics (MD) simulations, which uncover nanoscale mechanisms beyond experimental observations. However, studies on how GA structures and properties evolve in response to simulation parameter changes, which provide valuable insights to experimentalists, have been lacking. In addition, the differences between the calculated properties via simulations and experimental measurements have rarely been discussed. To address the shortcomings mentioned above, in this study, we systematically study various mechanical properties and the structural integrity of GA as a function of a wide range of simulation parameters. Results show that during the in silico GA preparation, smaller and less spherical inclusions (mimicking the effect of water clusters in experiments) are conducive to strength and stiffness but may lead to brittleness. Additionally, it is revealed that a structurally valid GA in the MD simulation requires the number of bonds per atom to be at least 1.40, otherwise the GA building blocks are not fully interconnected. Finally, our calculation results are compared with experiments to showcase both the power and the limitations of the simulation technique. This work may shed light on the improvement of computational approaches for GA as well as other novel nanomaterials.