A one-shot automated framework based on large language model and AutoML: Accelerating the design of porous carbon materials and carbon capture optimization
{"title":"A one-shot automated framework based on large language model and AutoML: Accelerating the design of porous carbon materials and carbon capture optimization","authors":"Lin Hu, Zhaorong Zhou, Guozhu Jia","doi":"10.1016/j.seppur.2025.133487","DOIUrl":null,"url":null,"abstract":"<div><div>With the escalating challenges of global warming and increasing demand for carbon capture, machine learning has become a popular approach. Traditional machine learning methods face challenges such as a lack of portability and complex operations, which prevent effective global optimization. To address this limitation, we present a one-shot automated framework, designed to accelerate the development of carbon capture materials and optimize carbon capture processes. It integrates experimental design, data collection, model training, and result analysis. By leveraging the large language model (LLM) for text mining and active learning, we expanded our dataset to over 10,000 entries. H<sub>2</sub>O AutoML enabled optimal model identification for material design and synthesis, achieving R<sup>2</sup> <span><math><mo>></mo></math></span> 0.95 and F1 Score <span><math><mo>></mo></math></span> 0.8. More than 100 optimized material designs with over 90<span><math><mi>%</mi></math></span> accuracy were generated within hours, representing a tenfold expansion of the experimental design space. Comparative analysis revealed a 60-100 fold increase in efficiency over conventional methods. It achieved exceptional carbon capture efficiency (7.5 mmol/g–9.5 mmol/g) and porous carbon synthesis yield (0.95), significantly exceeding current benchmarks. These results demonstrate the One-Shot automated framework’s potential for guiding intelligent industrial carbon capture processes.</div></div>","PeriodicalId":427,"journal":{"name":"Separation and Purification Technology","volume":"376 ","pages":"Article 133487"},"PeriodicalIF":9.0000,"publicationDate":"2025-06-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Separation and Purification Technology","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S1383586625020842","RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, CHEMICAL","Score":null,"Total":0}
引用次数: 0
Abstract
With the escalating challenges of global warming and increasing demand for carbon capture, machine learning has become a popular approach. Traditional machine learning methods face challenges such as a lack of portability and complex operations, which prevent effective global optimization. To address this limitation, we present a one-shot automated framework, designed to accelerate the development of carbon capture materials and optimize carbon capture processes. It integrates experimental design, data collection, model training, and result analysis. By leveraging the large language model (LLM) for text mining and active learning, we expanded our dataset to over 10,000 entries. H2O AutoML enabled optimal model identification for material design and synthesis, achieving R2 0.95 and F1 Score 0.8. More than 100 optimized material designs with over 90 accuracy were generated within hours, representing a tenfold expansion of the experimental design space. Comparative analysis revealed a 60-100 fold increase in efficiency over conventional methods. It achieved exceptional carbon capture efficiency (7.5 mmol/g–9.5 mmol/g) and porous carbon synthesis yield (0.95), significantly exceeding current benchmarks. These results demonstrate the One-Shot automated framework’s potential for guiding intelligent industrial carbon capture processes.
期刊介绍:
Separation and Purification Technology is a premier journal committed to sharing innovative methods for separation and purification in chemical and environmental engineering, encompassing both homogeneous solutions and heterogeneous mixtures. Our scope includes the separation and/or purification of liquids, vapors, and gases, as well as carbon capture and separation techniques. However, it's important to note that methods solely intended for analytical purposes are not within the scope of the journal. Additionally, disciplines such as soil science, polymer science, and metallurgy fall outside the purview of Separation and Purification Technology. Join us in advancing the field of separation and purification methods for sustainable solutions in chemical and environmental engineering.