Unveiling the role of lignin in biomass-derived hard carbon anodes via machine learning

Aug 4, 2026

Upload by Riley

Biomass-derived hard carbon provides a sustainable, high-potential anode for sodium-ion batteries, yet precursor-dependent variations strongly affect attainable specific capacity. This study compiles 149 literature data points from the past decade and applies machine learning to quantify how lignin content and lignin structure in biomass precursors govern hard-carbon performance. Tree-based ensemble models, especially XGB and GBDT, achieve strong predictive accuracy (R² up to 0.99 training, 0.60 testing) and reveal via interpretable analysis that higher lignin content and well-defined structures enhance capacity, while pyrolysis conditions also critically control outcomes.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages48

Reading Time2 h 1 min

AccessFree to download

Abstract

Introduction

Motivation for sodium-ion batteries

Role of hard carbon anodes

Biomass precursors and lignin influence

UC Berkeley UC Berkeley Previously Published Works Title Unveiling the role of lignin in biomass-derived hard carbon anodes via machine learning Permalink [https://escholarship.org/uc/item/5fh9m6t1](https://escholarship.org/uc/item/5fh9m6t1) Authors Li, Junxiao Jin, Yanghao Sun, Kang et al. Publication Date 2025-03-01 DOI 10.1016/j.jpowsour.2025.236323 Peer reviewed [eScholarship.org](eScholarship.org) Powered by the California Digital Library University of California 1 Unveiling the Role of Lignin in Biomass-Derived Hard 2 Carbon Anodes via Machine Learning 3 Junxiao Lia, Yanghao Jinb, Kang Suna,c*,, Ao Wanga,c,d, Gaoyue Zhanga,c, Limin Zhou a,c, 4 Weihong Yangb, Mengmeng Fana,c, Jianchun Jianga,c, Yuming Wene, Shule Wangf,a,c* 5 6 a Jiangsu Province Key Laboratory of Biomass Energy and Materials, National 7 Engineering Laboratory for Biomass Chemical Utilization, Institute of Chemical 8 Industry of Forest Products, Chinese Academy of Forestry (CAF), Nanjing, 210042, 9 China. 10 b Department of Materials Science and Engineering, KTH Royal Institute of 11 Technology, SE-100 44, Stockholm, Sweden. 12 c Jiangsu Co-Innovation Center for Efficient Processing and Utilization of Forest 13 Resources, College of Chemical Engineering, Nanjing Forestry University, Longpan 14 Road 159, Nanjing 210037, China. 15 d School of Chemistry, Chemical Engineering and Biotechnology, Nanyang 16 Technological University, 62 Nanyang Drive, Singapore, 637459 Singapore. 17 e Department of Chemical and Biomolecular Engineering, National University of 18 Singapore, 4 Engineering Drive 4, E5 \#02-09, Singapore 117585. 19 f Department of Environmental Science. Policy and Management, University of 20 California, Berkeley, CA 94720, USA 21 22 *[Email: Kang Sun: W19852858766@163.com](Email: Kang Sun: W19852858766@163.com) ; Shule Wang: [shule@berkeley.edu](shule@berkeley.edu). 23 Abstract 24 Biomass-derived hard carbon is a sustainable and promising anode material for 25 sodium-ion batteries. Variations in biomass precursors lead to substantial differences in 26 capacity, necessitating a deeper understanding of the underlying mechanisms. This 27 study collected data from 149 relevant literature in the past decade. We used machine 28 learning models to analyze the impact of lignin content and its structure in biomass 29 precursors on the specific capacity of the derived hard carbon. The tree-based ensemble 30 algorithms, particularly XGB and GBDT, showed superior performance, with the 31 optimal model having a R²value of up to 0.99 for training and 0.60 for testing. 32 Interpretable machine learning models identified lignin content and its structure as 33 crucial factors, Shapley value analysis highlighted that higher lignin content and well- 34 defined lignin structures positively influence capacity. Also, it is found that optimal 35 pyrolysis temperatures (1000-1400℃) and appropriate retention times are critical for 36 enhancing performance. This work provides insights into optimizing biomass precursor 37 selection and processing for high-performance hard carbon anodes. 38 Key words: biomass hard carbon, sodium-ion battery, machine learning, XGB, 39 precursor selection, lignin 40 1.Introduction 41 Lithium-ion batteries (LIBs) have been widely commercialized due to their high 42 specific capacity, excellent cycle life, and rate performance [1] . However, the limited 43 availability of lithium resources and their high costs render LIBs less suitable for large- 44 scale energy storage applications [2] . This has driven the exploration of alternative 45 battery technologies that offer similar performance at a lower cost and with more 46 abundant raw materials [3, 4] . Sodium-ion batteries (SIBs) have emerged as one of the 47 most promising alternatives to LIBs due to the low cost and abundant availability of 48 sodium [5] . Despite these

FAQs

Why is biomass-derived hard carbon important for sodium-ion batteries?

It is considered a sustainable and promising anode material, and its performance depends strongly on the biomass precursor composition.

How does the study use machine learning in analyzing hard carbon anodes?

It collects 149 data points from literature and trains tree-based ensemble models to analyze how lignin content and structure affect specific capacity.

Which factors identified by interpretable analysis are most influential for capacity?

Higher lignin content and well-defined lignin structures are highlighted as crucial factors that positively influence capacity.

Aug 4, 2026

Upload by Riley