Natural Language Processing for data extraction on CO2 reduction reactions

• Grant number: 24/13697-4
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: October 01, 2024
• End date: September 30, 2025
• Field of knowledge: Physical Sciences and Mathematics - Physics - Condensed Matter Physics
• Principal Investigator: James Moraes de Almeida
• Grantee: Paulo Henrique Alves dos Santos
• Host Institution: Centro Nacional de Pesquisa em Energia e Materiais (CNPEM). Ministério da Ciência, Tecnologia e Inovação (Brasil). Campinas , SP, Brazil
• Associated research grant: 23/09820-2 - Materials by design: from quantum materials to energy applications, AP.TEM

Abstract

Carbon dioxide (CO2) is one of the main greenhouse gases responsible for climate change, being excessively emitted into the atmosphere daily due to human activity. The electrochemical reduction of CO2 (CO2RR) emerges as a promising solution to mitigate these effects by converting CO2 into useful compounds. However, this process is energetically costly, and the efficiency varies depending on the catalyst used. Copper, in particular, stands out for its catalytic properties that allow for the efficient conversion of CO2 into commercially interesting products.This research project focuses on the use of Natural Language Processing (NLP) tools and Large Language Models (LLMs) to analyze the vast amount of scientific data published on CO2RR. Manual analysis of this data is impractical due to the growing volume of publications. NLP techniques, especially LLMs, are employed to automate the extraction and interpretation of information from articles, aiming to understand how variations in the composition of copper-based materials influence the Faradaic Efficiency (FE) and the reaction mechanisms.The specific objectives of the project include applying LLMs to study the variation of FE and the reaction mechanisms of copper in CO2RR. The student's work will be a continuation of another work carried out by the supervisor, in which a general survey of the catalysts used for CO2RR has already been performed, with a dataset of over 7000 articles already obtained.