Accelerated end-to-end chemical synthesis development with large language models DOI Creative Commons
Yixiang Ruan,

Chenyin Lu,

Ning Xu

и другие.

Опубликована: Май 8, 2024

The rapid emergence of large language model (LLM) technology presents significant opportunities to facilitate the development synthetic reactions. In this work, we leveraged power GPT-4 build a multi-agent system handle fundamental tasks involved throughout chemical synthesis process. comprises six specialized LLM-based agents, including Literature Scouter, Experiment Designer, Hardware Executor, Spectrum Analyzer, Separation Instructor, and Result Interpreter, which are pre-prompted accomplish designated tasks. A web application was built with as backend allow chemist users interact experimental platforms analyze results via natural language, thus, requiring zero-coding skills easy access for all chemists. We demonstrated on recently developed copper/TEMPO catalyzed aerobic alcohol oxidation aldehyde reaction, LLM copiloted end-to-end reaction process includes: literature search information extraction, substrate scope condition screening, kinetics study, optimization, scale-up product purification. This work showcases trilogy among users, automated reform traditional expert-centric labor-intensive workflow.

Язык: Английский

Structural Design for EMI Shielding: From Underlying Mechanisms to Common Pitfalls DOI Creative Commons
Ali Akbar Isari,

Ahmadreza Ghaffarkhah,

Seyyed Alireza Hashemi

и другие.

Advanced Materials, Год журнала: 2024, Номер 36(24)

Опубликована: Март 12, 2024

Abstract Modern human civilization deeply relies on the rapid advancement of cutting‐edge electronic systems that have revolutionized communication, education, aviation, and entertainment. However, electromagnetic interference (EMI) generated by digital poses a significant threat to society, potentially leading future crisis. While numerous efforts are made develop nanotechnological shielding mitigate detrimental effects EMI, there is limited focus creating absorption‐dominant solutions. Achieving EMI shields requires careful structural design engineering, starting from smallest components considering most effective wave attenuating factors. This review offers comprehensive overview structures, emphasizing critical elements design, mechanisms, limitations both traditional shields, common misconceptions about foundational principles science. systematic serves as scientific guide for designing structures prioritize absorption, highlighting an often‐overlooked aspect

Язык: Английский

Процитировано

84
Progress toward the computational discovery of new metal–organic framework adsorbents for energy applications DOI
Peyman Z. Moghadam, Yongchul G. Chung, Randall Q. Snurr

и другие.

Nature Energy, Год журнала: 2024, Номер 9(2), С. 121 - 133

Опубликована: Янв. 9, 2024

Язык: Английский

Процитировано

67
Nanozymes for nanohealthcare DOI
Yihong Zhang, Gen Wei, W. Liu

и другие.

Nature Reviews Methods Primers, Год журнала: 2024, Номер 4(1)

Опубликована: Май 30, 2024

Язык: Английский

Процитировано

59
ChatGPT Research Group for Optimizing the Crystallinity of MOFs and COFs DOI Creative Commons
Zhiling Zheng, Oufan Zhang, Ha L. Nguyen

и другие.

ACS Central Science, Год журнала: 2023, Номер 9(11), С. 2161 - 2170

Опубликована: Ноя. 10, 2023

We leveraged the power of ChatGPT and Bayesian optimization in development a multi-AI-driven system, backed by seven large language model-based assistants equipped with machine learning algorithms, that seamlessly orchestrates multitude research aspects chemistry laboratory (termed Research Group). Our approach accelerated discovery optimal microwave synthesis conditions, enhancing crystallinity MOF-321, MOF-322, COF-323 achieving desired porosity water capacity. In this human researchers gained assistance from these diverse AI collaborators, each unique role within environment, spanning strategy planning, literature search, coding, robotic operation, labware design, safety inspection, data analysis. Such comprehensive enables single researcher working concert to achieve productivity levels analogous those an entire traditional scientific team. Furthermore, reducing biases screening experimental conditions deftly balancing exploration exploitation parameters, our search precisely zeroed on pool 6 million significantly shortened time scale. This work serves as compelling proof concept for AI-driven revolution laboratory, painting future where becomes efficient collaborator, liberating us routine tasks focus pushing boundaries innovation.

Язык: Английский

Процитировано

56
Shaping the Water-Harvesting Behavior of Metal–Organic Frameworks Aided by Fine-Tuned GPT Models DOI
Zhiling Zheng, Ali H. Alawadhi, Saumil Chheda

и другие.

Journal of the American Chemical Society, Год журнала: 2023, Номер 145(51), С. 28284 - 28295

Опубликована: Дек. 13, 2023

We construct a data set of metal-organic framework (MOF) linkers and employ fine-tuned GPT assistant to propose MOF linker designs by mutating modifying the existing structures. This strategy allows model learn intricate language chemistry in molecular representations, thereby achieving an enhanced accuracy generating structures compared with its base models. Aiming highlight significance design strategies advancing discovery water-harvesting MOFs, we conducted systematic variant expansion upon state-of-the-art MOF-303 utilizing multidimensional approach that integrates extension multivariate tuning strategies. synthesized series isoreticular aluminum termed Long-Arm MOFs (LAMOF-1 LAMOF-10), featuring bear various combinations heteroatoms their five-membered ring moiety, replacing pyrazole either thiophene, furan, or thiazole rings combination two. Beyond consistent robust architecture, as demonstrated permanent porosity thermal stability, LAMOF offers generalizable synthesis strategy. Importantly, these 10 LAMOFs establish new benchmarks for water uptake (up 0.64 g g-1) operational humidity ranges (between 13 53%), expanding diversity MOFs.

Язык: Английский

Процитировано

50
Machine Learning-Assisted Discovery of Propane-Selective Metal–Organic Frameworks DOI
Ying Wang, Zhijie Jiang,

Dong-Rong Wang

и другие.

Journal of the American Chemical Society, Год журнала: 2024, Номер 146(10), С. 6955 - 6961

Опубликована: Фев. 29, 2024

Machine learning is gaining momentum in the prediction and discovery of materials for specific applications. Given abundance metal–organic frameworks (MOFs), computational screening existing MOFs propane/propylene (C3H8/C3H6) separation could be equally important developing new MOFs. Herein, we report a machine learning-assisted strategy C3H8-selective from CoRE MOF database. Among four algorithms applied learning, random forest (RF) algorithm displays highest degree accuracy. We experimentally verified identified top-performing (JNU-90) with its benchmark selectivity performance directly producing C3H6. Considering excellent hydrolytic stability, JNU-90 shows great promise energy-efficient C3H8/C3H6. This work may accelerate development challenging separations.

Язык: Английский

Процитировано

27
Image and data mining in reticular chemistry powered by GPT-4V DOI Creative Commons
Zhiling Zheng,

Zhiguo He,

Omar Khattab

и другие.

Digital Discovery, Год журнала: 2024, Номер 3(3), С. 491 - 501

Опубликована: Янв. 1, 2024

The integration of artificial intelligence into scientific research opens new avenues with the advent GPT-4V, a large language model equipped vision capabilities.

Язык: Английский

Процитировано

18
Accelerating materials language processing with large language models DOI Creative Commons
Jaewoong Choi, Byungju Lee

Communications Materials, Год журнала: 2024, Номер 5(1)

Опубликована: Фев. 15, 2024

Abstract Materials language processing (MLP) can facilitate materials science research by automating the extraction of structured data from papers. Despite existence deep learning models for MLP tasks, there are ongoing practical issues associated with complex model architectures, extensive fine-tuning, and substantial human-labelled datasets. Here, we introduce use large models, such as generative pretrained transformer (GPT), to replace architectures prior strategic designs prompt engineering. We find that in-context GPT few or zero-shots provide high performance text classification, named entity recognition extractive question answering limited datasets, demonstrated various classes materials. These also help identify incorrect annotated data. Our GPT-based approach assist material scientists in solving knowledge-intensive even if they lack relevant expertise, offering guidelines applicable any domain. In addition, outcomes expected reduce workload researchers, manual labelling, producing an initial labelling set verifying human-annotations.

Язык: Английский

Процитировано

18
Large language models for reticular chemistry DOI
Zhiling Zheng, Nakul Rampal,

Theo Jaffrelot Inizan

и другие.

Nature Reviews Materials, Год журнала: 2025, Номер unknown

Опубликована: Янв. 31, 2025

Язык: Английский

Процитировано

3
Evaluation of Open-Source Large Language Models for Metal–Organic Frameworks Research DOI

Xuefeng Bai,

Ya-Bo Xie, Xin Zhang

и другие.

Journal of Chemical Information and Modeling, Год журнала: 2024, Номер 64(13), С. 4958 - 4965

Опубликована: Март 26, 2024

Along with the development of machine learning, deep and large language models (LLMs) such as GPT-4 (GPT: Generative Pre-Trained Transformer), artificial intelligence (AI) tools have been playing an increasingly important role in chemical material research to facilitate screening design. Despite exciting progress based AI assistance, open-source LLMs not gained much attention from scientific community. This work primarily focused on metal–organic frameworks (MOFs) a subdomain chemistry evaluated six top-rated comprehensive set tasks including MOFs knowledge, basic in-depth knowledge extraction, database reading, predicting property, experiment design, computational scripts generation, guiding experiment, data analysis, paper polishing, which covers units research. In general, these were capable most tasks. Especially, Llama2-7B ChatGLM2-6B found perform particularly well moderate resources. Additionally, performance different parameter versions same model was compared, revealed superior higher versions.

Язык: Английский

Процитировано

14