top of page

AI drastically accelerates the optimization of biorefineries

Biorefineries, which transform biomass such as wood and agricultural residues into valuable products and energy, are on the brink of a groundbreaking transformation thanks to artificial intelligence (AI). Finnish and German research teams are leveraging AI to enhance the efficiency and output of biorefineries, aiming for a more comprehensive utilization of natural resources. Traditionally, developing these processes has been slow due to the need to optimize numerous factors, but AI offers a game-changing solution.



A multidisciplinary consortium of scientists from Aalto University and Åbo Akademi in Finland, along with the Technical University of Munich in Germany, is advancing a flexible biorefinery concept that exploits lignin-carbohydrate complexes (LCCs) - inherent components of wood that contribute to its rigidity and structure. Unlike conventional pulping processes, which break down LCCs to separate cellulose, hemicellulose, and lignin, this project aims to harness LCCs to create bio-based antioxidants, effective surfactants, and polymer additives.


Isolated lignin-carbohydrate complexes to be tested for new applications. Photo: Aalto University
Isolated lignin-carbohydrate complexes to be tested for new applications. Photo: Aalto University

The teams of late Professor of Practice Mikhail Balakshin of Aalto University, Professor Patrick Rinke from both Aalto University and Technical University of Munich, and Professor Chunlin Xu from Åbo Akademi have developped process named AquaSolv Omni (AqSO) that employs modified hydrothermal treatment and solvent extraction to produce high-yield LCCs with customizable properties.


Professor Rinke's group has integrated Bayesian Optimization to iteratively gather data and ascertain the effects of various process conditions (P-factor, temperature, and liquid-to-solid ratio) on yield and carbohydrate content. By using Pareto front analysis, the team has successfully maximized both yield and carbohydrate content. To assess the potential of LCC's for high-value application potential of LCCs, their properties such as antioxidant activity, surface tension, and glass transition temperature were investigated in collaboration with experts from Professor Xu’s team of biomass chemists.


Doctoral candidate Daryna Diment elaborates, “High carbohydrate content in LCCs significantly benefits reducing the glass transition temperature and surface tension of aqueous solutions, indicating their potential in thermoplastic formulations and as bio-based surfactants. LCCs processed under intense conditions (high temperature and extended time) also exhibited notable antioxidant activity.”


The novel process represents a significant leap over traditional biorefinery methods for isolating LCCs, since it utilizes only hydrothermal treatment followed by acetone extraction, making it environmentally friendly. The flexibility afforded by Bayesian Optimization allows for the production of LCCs with specific, high-yield properties. This research not only points to LCCs as a promising new biorefinery product but also highlights AI's role in speeding up technological advancements for better valuation of local and natural resources.


This research was published by Diment et al. in ChemSusChem in November 2024.


This work was a part of consortium project “AI-4-LCC: Exploiting Lignin-Carbohydrate Complex through Artificial Intelligence” funded by the Research Council of Finland. The work has been also supported by Flagship programmes FinnCERES (Competence Center for Materials Bioeconomy) and FCAI (the Finnish Center for Artificial Intelligence).


bottom of page