Search NASA⌕ Search

SEARCH · Search NASA

Results for “Design, synthesis and processing”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 19 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

A technoeconomic analysis of poly- and single-crystalline NMCxyz from material synthesis to battery pack design

A process model was developed for estimating the cost of manufacturing lithium nickel manganese cobalt oxide (LiNi x Mn y Co z O 2 , NMCxyz), the main cathode active material in lithium-ion batteries used for electric vehicles in the United States. The model was used to estimate the prices of NMC622, NMC811, and NMC955 with poly- and single-crystalline morphologies. The Battery Performance and Cost Model (BatPaC) was used to translate the NMC prices into battery pack prices. A decrease in cobalt content from NMC622 (20% Co) to NMC955 (5% Ni) decreases the material price by $\$$0.85/kg (−3%) due to a decrease in the cost of battery materials, which account for >60% of the total price. NMC622 only produce cheaper packs if the nickel sulfate cost is > 4 × its baseline, indicating higher nickel materials will lower pack cost under normal market conditions. Single-crystalline materials are $\$$2/kg (+8%) more expensive than their polycrystalline counterparts due to higher manufacturing costs from longer, hotter calcinations in less densely packed saggars. This increases the pack price by ∼$\$$3/kWh, assuming identical electrochemical properties. In conclusion, the single-crystalline materials could yield cheaper packs if cycled to higher upper cutoff voltages (i.e., 4.35 to 4.43 V for the single-crystalline material vs. 4.25 V for their polycrystalline counterparts).

Cost modeling↗
DOE OSTI2025

Accurate and uncertainty-aware multi-task prediction of HEA properties using prior-guided deep Gaussian processes

Surrogate modeling techniques have become indispensable in accelerating the discovery and optimization of high-entropy alloys (HEAs), especially when integrating computational predictions with sparse experimental observations. This study systematically evaluates the training and testing performance of four prominent surrogate models—conventional Gaussian processes (cGP), Deep Gaussian processes (DGP), encoder-decoder neural networks for multi-output regression and eXtreme Gradient Boosting (XGBoost)—applied to a hybrid dataset of experimental and computational properties of the 8-component HEA system Al-Co-Cr-Cu-Fe-Mn-Ni-V. We specifically assess their capabilities in predicting correlated material properties, including yield strength, hardness, modulus, ultimate tensile strength, elongation, and average hardness under dynamic/quasi-static conditions, alongside auxiliary computational properties. The comparison highlights the strengths of hierarchical deep modeling approaches in handling heteroscedastic, heterotopic, and incomplete data commonly encountered in materials science. Our findings illustrate that combined surrogate models such as DGPs infused with machine-learned priors outperform other surrogates by effectively capturing inter-property correlations and by assimilating prior knowledge. This enhanced predictive accuracy positions the combined surrogate models as powerful tools for robust and data-efficient materials design.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Grayscale projection two-photon lithography using sub-diffraction motifs for ultrafast and precise nanoscale 3D printing

Rapid and high-fidelity nanoscale 3D printing is highly desirable, but it is difficult due to the tradeoff between speed and accuracy. Although optical projection techniques can massively scale up printing, fidelity is compromised due to the difficulty in precisely controlling the light dosage over the entire field. This challenge is typically addressed by using multiple projections, but it slows down printing. Here, we present grayscale projection two-photon lithography to overcome this tradeoff. Despite using a binary mask, it enables projecting more than 15,000 focal spots, each with independently tunable intensity. It advantageously leverages constraints imposed by optical diffraction to achieve grayscale tuning over the entire field at once. By directly tuning the focal spot intensities, we demonstrate suppression of proximity effects, compensation of non-uniform illumination, compensation of stitching artefacts, and rapid 3D printing with a single femtosecond pulse per layer. We demonstrate printing of nanowires as thin as 55 nm and achieve rates of 1.7 billion voxels/s and 215 mm 3 /hr.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Optimized Photoemission from Organic Molecules in 2D Layered Halide Perovskites

In recent years, hybrid organic−inorganic metal halides have been at the forefront of materials research. Typically, the functional (e.g., optoelectronic) properties of hybrid halides are derived from the inorganic structural part, whereas the organic structural units can add extra advantages in terms of stability, rigidity, and processability. Here, we report the design, synthesis, and characterization of two new hybrid materials in which the outstanding photophysical properties originate from the organic structural part. The new compounds, (C 15 H 16 N) 2 CdCl 4 and ((Br)C 15 H 15 N) 2 CdCl 4 , have 2D layered Ruddlesden−Poppertype perovskite structures. These hybrids are blue-white light emitters just like their corresponding pure organic salts, but with much improved emission efficiencies. Optical spectroscopy and density functional theory (DFT) studies confirm that photoemission comes from the trans-stilbene organic cations. The photoluminescence quantum yield (PLQY) values of these new materials are among the highest known, 50.83% and 26.60% for (C 15 H 16 N) 2 CdCl 4 and ((Br)C 15 H 15 N) 2 CdCl 4 , respectively. This is up to a 5-fold increase as compared to the light emission efficiency of the precursor salt C 15 H 16 NCl (PLQY of 10.33%). Alongside their outstanding optical properties, their environmental and thermal stability allow their consideration for potential practical applications such as radiation detection. This work shows that hybrid metal halides can be compositionally and structurally engineered to have highly efficient photoemission originating from the organic components for fast scintillation applications.

Halogens↗
DOE OSTI2025

Extending High-Level Synthesis with AI/ML Methods

Artificial Intelligence (AI) and Machine Learning (ML) methods provide significant opportunities of improving quality of results when performing high-level synthesis (HLS). For example, they can be used to model and predict metrics of the final design (e.g., area, considering aspects such as interconnect overhead for different device technologies), facilitating exploration when searching for the best design trade-offs. They can also enable identifying hidden correlations across the various phases of the synthesis and the various optimizations performed, identifying the most effective pipelines. Finally, in more general terms, bio-inspired heuristic algorithms can improve the design space exploration for the synthesis process in terms of time and quality of the result. This paper discusses opportunities and challenges to augment HLS with AI/ML using as example flow the SODA Synthesizer, an open-source hardware generation toolchain which includes SODA-OPT, a hardware/software partitioning and pre-optimization tool developed with the MLIR framework, and PandA-Bambu, a state-of-the art HLS tool. SODA interfaces with OpenROAD to provide a complete end-to-end toolchain.

artificial intelligence↗
DOE OSTI2025

Using Electrochemistry to Benchmark, Understand, and Develop Noble Metal Nanoparticle Syntheses

The complex chemical nature of metal nanoparticle synthesis presents obstacles for the mechanistic understanding of nanoparticle growth and predictive synthesis design, despite significant progress in this area. Real-time characterization of the chemical processes that take place throughout nanoparticle growth will enable progress toward addressing outstanding challenges in metal nanoparticle synthesis, such as mitigating synthetic reproducibility issues, defining chemical mechanisms that direct nanoparticle growth, and designing synthetic conditions for previously unachievable combinations of nanoparticle shape and composition. In this Perspective, we present open-circuit potential (OCP) measurements as an in situ, real-time method for characterizing chemical changes during nanoparticle growth and discuss the method’s strengths in comparison to and in combination with other characterization techniques. We propose the use of OCP measurements as benchmarks for troubleshooting irreproducibility and streamlining synthetic optimization. Finally, we explore possibilities for using the increased parameter space accessible by electrodeposition to accelerate the development of shape-selective nanoparticle syntheses.

benchmarking↗
DOE OSTI2025

Enabling Partnership between South Carolina and NREL for Advancing Opportunities in Plastics Recycling Research (EPSCOR for Plastics Recycling)

Proposal Objectives: 1. Utilize depolymerization/fractionation techniques to recover highly processable and reactive feedstocks for polymer synthesis from lignin. 2. Synthesize lignin‐derived non‐isocyanate polyurethane, epoxy, and polyamide using non‐ toxic, biobased route designed for chemical recycling. Characterize resulting materials. 3. Design a high‐yielding chemical recycling process for as‐synthesized materials yielding usable building blocks for many generations of polymer synthesis. 4. Optimize chemical recycling of PET waste for the synthesis of lignin‐based polymers. Compare properties to commercial materials. 5. Optimize reaction conditions and recycling steps to facilitate enhanced sustainability of the synthetic steps and final properties of materials. 6. Complete a lifecycle assessment of lignin utilization and chemical recycling to compare their environmental performance to that of materials produced from virgin material. Identify hot spots and benefits using the chemical recycling process.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Autonomous Synthesis and Inverse Design of Electrochromic Polymers with High Efficiency and Accuracy

Here, the design and synthesis of functional polymers, aimed at targeted properties through specific structures, have long been challenged by their complex and often nonlinear structure–property relationships. Key processes, including knowledge accumulation for predictive design and experimental refinement and validation, are traditionally labor-insensitive and time-consuming, making it difficult to balance accuracy and efficiency. Here, we introduce an accelerated, autonomous system for the on-demand synthesis of electronic polymers that achieves the desired electrochromic functionality with high accuracy and efficiency. Our approach leverages large language model-assisted data mining, a physics-informed copolymer machine learning model, and an AI-driven autonomous robotic workflow in the Polybot lab. Within 72 h, Polybot autonomously synthesized electrochromic polymers (ECPs) with targeted, previously-unreported color values, including green polymers with specific absorption profiles, precisely fine-tuning copolymer structures with a 5% step size in comonomer composition within a three-monomer system. A publicly accessible ECP informatics database has also been created to foster knowledge exchange.

AI-driven Robotic Lab↗
DOE OSTI2026

wa-hls4ml: A Benchmark and Surrogate Models for hls4ml Resource and Latency Estimation

As machine learning (ML) is increasingly implemented in hardware to address real-time challenges in scientific applications, the development of advanced toolchains has significantly reduced the time required to iterate on various designs. These advancements have solved major obstacles, but also exposed new challenges. For example, processes that were not previously considered bottlenecks, such as hardware synthesis, are becoming limiting factors in the rapid iteration of designs. To mitigate these emerging constraints, multiple efforts have been undertaken to develop an ML-based surrogate model that estimates resource usage of ML accelerator architectures. We introduce wa-hls4ml, a benchmark for ML accelerator resource and latency estimation, and its corresponding initial dataset of over 680,000 fully connected and convolutional neural networks, all synthesized using hls4ml and targeting Xilinx FPGAs. The benchmark evaluates the performance of resource and latency predictors against several common ML model architectures, primarily originating from scientific domains, as exemplar models, and the average performance across a subset of the dataset. Additionally, we introduce GNN- and transformer-based surrogate models that predict latency and resources for ML accelerators. We present the architecture and performance of the models and find that the models generally predict latency and resources for the 75% percentile within several percent of the synthesized resources on the synthetic test dataset.

Hawks, Benjamin [Fermilab] (ORCID:0000000157000288↗
DOE OSTI2025

wa-hls4ml: A Benchmark and Surrogate Models for hls4ml Resource and Latency Estimation

As machine learning (ML) is increasingly implemented in hardware to address real-time challenges in scientific applications, the development of advanced toolchains has significantly reduced the time required to iterate on various designs. These advancements have solved major obstacles, but also exposed new challenges. For example, processes that were not previously considered bottlenecks, such as hardware synthesis, are becoming limiting factors in the rapid iteration of designs. To mitigate these emerging constraints, multipleefforts have been undertaken to develop an ML-based surrogate model that estimates resource usage of synthesized ML accelerator architectures. We introduce wa-hls4ml, a benchmark for ML accelerator resource and latency estimation, and its corresponding initial dataset of over 680 000 fully connected and convolutional neural networks, all synthesized using hls4ml and targeting Xilinx FPGAs. The benchmark evaluates the performance of resource and latency predictors against several common ML model architectures, primarily originating from scientific domains, as exemplar models, and the average performance across a subset of the dataset. Additionally, we introduce GNN- and transformer-based surrogate models that predict latency and resources for ML accelerators. We present the architecture and performance of the models and find that the models generally predict latency and resources for the 75% percentile within several percent of the synthesized resources on the synthetic test dataset.

Hawks, Benjamin G. [Fermilab]↗
DOE OSTI2025

wa-hls4ml and lui-gnn: A benchmark and GNN-based surrogate model for hls4ml resource and latency estimation

As machine learning (ML) increasingly serves as a tool for addressing real-time challenges in scientific applications, the development of advanced tooling has significantly reduced the time required to iterate on various designs. These advancements have solved major obstacles, but also exposed new challenges. For example, processes that were not previously considered bottlenecks, such as model synthesis, are now becoming limiting factors in the rapid iteration of designs. To reduce these emerging constraints, multiple efforts are being launched toward designing an ML-based surrogate model that estimates resource usage of synthesized accelerator architectures. This model would reduce the design iteration time, especially when designing within a set of given hardware constraints. This approach shows considerable potential, but as it stands, the effort is early and would benefit from coordination and standardization to assist future work as it emerges. We introduce wa-hls4ml, a benchmark for ML accelerator resource and latency estimation, and its corresponding initial dataset of more than 100,000 fully connected neural networks, all synthesized using hls4ml and targeting Xilinx FPGAs. In addition to the resource utilization and latency data provided, the dataset includes generated artifacts and log files for many of the synthesized neural networks, in order to support future research in ML-based code generation. The benchmark evaluates the performance of resource and latency predictors against several common ML model architectures, primarily originating from scientific domains, as exemplar models, as well as the average performance across a subset of the dataset. We measure the performance of a given predictor model through multiple metrics, including $R^2$ score and SMAPE on regression tasks, as well as inference time to further characterize the estimator under test. Additionally, we introduce the latency/utilization inference graph neural network (lui-gnn), a surrogate model that uses a graph neural network to represent input architectures in the form of a directed graph. This graph representation allows for a diverse set of model architectures to all be effectively handled by a surrogate model. We present the architecture and performance of the model, as evaluated by the new proposed benchmark, including SMAPE, $R^2$ score, and inference times, and find that lui-gnn generally predicts latency and utilization for the 75\% quantile within several percent of the synthesized resources on the synthetic test dataset, indicating that this approach of estimating resource and latency via a surrogate models has promise and warrants further research.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2025

High-throughput synthesis of high-entropy alloys via parallelized electric field assisted sintering

Materials discovery and design is an expensive and time-consuming process, though necessary to advance many engineering fields. In this work, a novel tooling design is utilized in conjunction with electric field assisted sintering (EFAS) to effectively create a new high-throughput synthesis technique: parallelized EFAS. Through this technique, a wide range of material compositions and geometries can be synthesized in parallel as isolated samples or as part of contiguous arrays. Multiple tooling designs are explored to examine both the flexibility and limitations of the technique. A series of increasing complex alloys is produced simultaneously using in situ alloying, beginning with pure Ni and adding equimolar constituents up to the septenary high-entropy alloy AlCoCrCuFeMnNi. Microstructural characterization reveals each sample is effectively fully dense and chemically homogenous while exhibiting phases in agreement with CALPHAD predictions. Scalability of parallelized EFAS is then experimentally demonstrated and the implications for materials discovery and automation are discussed.

36 - MATERIALS SCIENCE↗
DOE OSTI2026

Understanding Solvent-Induced Glass Transition in Polymer Thin Films Using Absorption–Desorption Isotherms

The fundamental thermodynamic and mechanical underpinnings of polymer thin films exposed to solvent vapor are critical for the development of advanced nanolithography and high-performance coatings. This work investigates the solvent− polymer interactions of glassy thin films by using the solvent absorption−desorption isotherms. An analogous relationship to the Flory−Fox equation was observed between solvent−induced glass transition, swelling, Flory−Huggins interaction parameter, and molecular weight. Isothermal swelling measurements revealed that the glass transition trends are more robust in the absorption curve compared to desorption, contrary to previous reports. Excess osmotic pressure analysis of the isotherm provides a measure of the degree of physical aging in thin films annealed below the glass transition. This is further validated in the ordering of block copolymer (BCP) films annealed at low solvent activity. In agreement with the thermal analysis, free-surface plasticization effects become the most prominent below 100 nm. However, solvent annealing is largely dependent on solvent mass transport, as made evident by the strong dependence on solvent viscosity. From these observations, four general types of isotherms are identified that graphically capture distinct solvent−polymer interaction regimes. More broadly, these results inform solvent vapor annealing-induced self-assembly, sequential infiltration synthesis, membrane-based separations, adsorptive processes, and swelling-based responsive materials design.

Hendeniya, Nayanathara [Iowa State Univ., Ames, IA↗
DOE OSTI2026

Facile synthesis of Co(OH)2 nanoneedle arrays grown on stainless steel for industrial electrochemical oxygen evolution reaction

The electrochemical oxygen evolution reaction (OER) is a critical half-reaction in a variety of energy conversion and storage applications, however, OER suffers from sluggish kinetics due to the four proton-electron transfer processes. To remedy this, interfacial engineering proves an effective strategy to design high active OER catalysts. Herein, we report a facile synthesis of Co(OH)2 nanoneedle (Co-NN) arrays grown on stainless steel (SS) via a one-step hydrothermal reaction, and the resultant impressive OER performance. Particularly, the Co-NN/SS needs 267, 307, and 576 mV overpotentials to reach 10, 100, and 1000 mA cm−2 in 1 M KOH & room temperature, and the overpotentials drop to 199, 257, and 283 mV to achieve the same current densities in 30 wt% KOH & 80 °C. Additionally, an alkaline water electrolysis (AWE) cell coupled with Co-NN/SS anode and PtRu/NiMo cathode only requires 1.65 and 1.73 V iR-free cell voltage to reach 1.0 and 2.0 A cm−2, respectively. The sterling OER performance could be attributed to four possible reasons, the Fe in SS surface tailoring the electronic properties of Co(OH)2, the mixed metal oxides on SS surface accelerating surface reconstruction of Co(OH)2 nanoneedles into CoOOH active species, a Ni-rich surface layer formation cooperatively boosting the OER activity, and a local electric field effect concentrating reactants on the tip surface and accelerating the mass transfer. This work provides a facile approach for synthesizing highly efficient OER catalysts for industrial applications by interfacial engineering.

Lyu, Xiang [ORNL] (ORCID:0000000208673248)↗
DOE OSTI2026

Intelligent Experiments through Real-Time AI: Fast Data Processing and Autonomous Detector Control for High-Energy Nuclear Experiments

The aim of this project is to develop software and hardware for fast real-time data processing and autonomous detector control and calibration for the sPHENIX and the future EIC experiments. Below summarizes Georgia Tech team efforts in the past year: 1. We developed a real-time clustering algorithm and FPGA-based pipeline architecture for processing fired pixel data from ALPIDE sensors in sPHENIX experiments. Our Columnar Clustering Co-Design introduces a hardware-aware, stream-friendly approach that segments pixel data by column pairs using a Column Pair Clustering (CPC) strategy, followed by Cluster Stitching to merge adjacent subclusters. Implemented in Vitis HLS, the pipeline comprises five stages—read-in, subclustering, stitching, analysis, and write-out—connected by tagged HLS streams with custom end-of-event signaling for robust synchronization. We designed a pipelined dataflow model optimized for throughput, low latency, and minimal buffering, enabling scalable clustering across events of arbitrary size. Our system maintains spatial precision via center-of-mass and shape key extraction and efficiently handles edge cases such as fragmented or nested clusters. Compared against DBSCAN in both software and hardware, our approach demonstrates competitive performance under FPGA constraints. 2. We also conducted a comprehensive algorithm-to-hardware co-design of connected component analysis tailored for sPHENIX experiments, focusing on real-time, low-latency processing using FPGAs and High-Level Synthesis (HLS). Starting from a Python-based particle tracking pipeline, the team translated the core logic—graph traversal via DFS and Union-Find—into an HLS-compatible C++ model, replacing dynamic memory and recursion with static arrays and pipelined control flow. The final design includes a fully streamed and dataflow-compatible Union-Find kernel optimized across five iterations, incorporating loop pipelining, array partitioning, AXI/FIFO interface tuning, and function flattening. Experimental results show up to 14.8× speedup over the CPU baseline, reducing per-graph latency to 1.58 μs and demonstrating strong resource efficiency with only ~7k LUTs and zero BRAM usage. The design maintains functional correctness against the Python reference using a Python-based C-simulation framework and Mean Squared Error metrics. This work validates the potential of HLS-driven FPGA designs for edge-level HEP data acquisition, laying a scalable foundation for future integration with real-time detector pipelines and multi-graph processing systems.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2026

Design Principles for the Synthesis of Self-Pillared ZSM-5 Zeolite Nanosheets

The design of next-generation materials for emerging energy and environmental applications heavily relies on empirical approaches to direct nonclassical nucleation and crystal growth pathways, polymorphism, intercrystalline transformations, and seed-assisted growth processes. A long-standing obstacle to nanoporous materials design is the complexity of their crystallization, which hinders the development of predictive models and/or physical descriptors that can guide their synthesis. In this study, we use a combination of state-of-the-art synthesis, characterization, and computational design to prepare hierarchical MFI-type zeolites, which we couple with benchmark catalytic testing to assess structure-property-performance relationships. These hierarchical materials are intergrowths of two commercially relevant zeolite frameworks, MFI and MEL, prepared as self-pillared pentasil (SPP) zeolites through seed-assisted, organic-free syntheses for which little theoretical guidance existed. Here, by comparing a large library of zeolite seeds with different pore sizes, dimensions, and structural composite building units, we determined the relative impact of seed and silica source selection, among other synthesis variables. Combined experimental and computational studies are used to test several hypotheses in literature to rationalize the choice of seed structure and establish a more robust selection criteria for seed-assisted synthesis of zeolites. Specifically, we show that a data-driven approach to develop structural descriptors correlates to new, facile routes to rationally design SPP zeolites, addressing knowledge gaps in the fundamental understanding of (non)classical crystal growth mechanisms that are characteristic of nanoporous aluminosilicates.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Selectivity mechanisms of ion intercalation in Prussian blue analogs

Prussian blue analogs (PBAs) are a family of materials with facile, reversible, and selective ion transport capability for various ions via electrochemical intercalation, owing to their vacancy structure. The large tunable compositional space of PBAs allows for manipulation of intercalation behavior and selectivity by controlling structural vacancy level through choice of transition metal centers and modifications to the synthesis process. However, a lack of understanding of the mechanisms of ion selectivity hinders the material’s design process. Here, for this work, we investigated the origins of ion selectivity using a model PBA, copper hexacyanoferrate, and focused on eight technologically and biologically prominent ions, for which we determined a sequence of selectivity: Rb + > K + > Na + > Ba 2+ > Sr 2+ ≈ Ca 2+ > Mg 2+ > Li + . We provide electrochemical, structural, and redox evidence of strong correlation between the ion identity, the dominant charge-compensating redox, and preferred occupancy site. Specifically, using synchrotron anomalous X-ray diffraction (AXRD), we reveal that monovalent ions exhibit significant association with the corner sites of the unit cell and iron redox, whereas divalent ions display affinity toward the center site with higher ratios of copper redox. Informed by selectivity results, we applied CuHCFe to Li purification and achieved 99.9% purity. Our findings demonstrate an approach to elucidating ion intercalation behavior in order to distinguish and manipulate material properties to optimize separation performance.

Prussian blue analog↗
DOE OSTI2025

Computer-aided design of stability enhanced nicotinamide cofactor biomimetics for cell-free biocatalysis

Cell-free biocatalysis (CFB) is an efficient and environmentally friendly method to synthesize molecules such as pharmaceuticals, biochemicals, and biofuels through the in vitro use of enzyme cascades. These enzymes often require redox cofactors to drive chemical reactions. Natural redox cofactors (NAD(P)H) are expensive to isolate, motivating synthetic nicotinamide cofactor biomimetics (NCBs) as a cost-effective solution. A select handful of NCBs have been identified as potential NAD(P)H alternatives with comparable or improved redox capabilities, however, they display a tendency to degrade in common buffers. In this study, a library of 132 NCB candidates is systematically generated, over 85% of which have not been characterized in the literature, to expand the diversity of currently explored NCBs. The decomposition mechanism of NCBs in phosphate is evaluated using density functional theory (DFT), revealing protonation at the nicotinamide C5 position as a reporter of cofactor stability. Based on this result, we trained a linear regression model on DFT calculated descriptors to predict NCB stability in phosphate buffer, achieving mean absolute error (MAE) and root mean squared error (RMSE) values within computational accuracy. Analysis of key atomic descriptors and qualitative trends in our dataset informed the design of novel NCB candidates we propose with optimized stability. This work enables researchers to predict the relative stability of NCBs before synthesis, thereby streamlining the process to make CFB more affordable and viable at industry scales.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗