Woody biomass and wheat straw are all sources of the natural polymer lignin. More than 50 megatons of lignin are produced annually on a commercial scale. Most of this is burned to produce energy, while it could also be used to make useful chemicals. However, what is a major problem with producing chemicals from lignin is that the properties of lignin vary from source to source and from season to season. This variability can affect the yield and quality of the chemicals produced from lignin. In a study led by TU/e, researchers have therefore developed and tested a new and efficient model. With this model it is possible to predict the yield of lignin, including the specific chemical properties that are important for the production of biobased chemicals, materials or fuels.
To date, most lignin, from sources such as agricultural waste or woody biomass, is burned to produce energy. Since lignin is a renewable resource, this can be seen as waste. Researchers are looking for ways to use organic lignin as a reliable raw material for the chemical industry, for the production of resins, foams and biofuels.
As a material, woody biomass can be grown relatively quickly, making it attractive to use lignin for long-term chemical production.
However, this is an idealized picture of the situation. “The big problem is that the properties of lignin are both unpredictable and variable. This affects usability,” says Mark Vis, assistant professor at the Department of Chemical Engineering and Chemistry and research leader of the new research published in Green chemistry.
Why is the unpredictability of lignin properties a problem? Vis explains: “Suppose we want to make a certain chemical with lignin, but we need lignin with a specific chemical composition to make the chemical. There can be a million different types of lignin in any sample of lignin, and isolating the right type of lignin to make the chemical is the heart of the problem. Lignin does not have a clear chemical structure, unlike the raw materials used to make conventional chemicals.”
The needle in the haystack
This sounds like looking for the proverbial needle in the haystack.
A treatment process known as solvent fractionation can help find the desired types of lignin in the haystack. In this process, the types of lignin with the desired chemical properties are dissolved using a solvent. The product can later be further purified by removing the lignin from the solvent.
“Fractionation can reduce the number of lignin species, but with millions of lignin species in a sample, it is difficult to be sure that a particular solvent will isolate a particular lignin species,” says Vis. “Theoretical calculations can help predict the outcome of fractionation, but current theories are too complex to apply to lignin. And this is the problem we solved.”
The solution of Vis and his employees from TU/e (including first author Stijn van Leuken and postdoc Dannie van Osch), Maastricht University and the spin-off Vertoro is a new model that accurately and quickly predicts the fractionation of lignin in a mixture of solvents with methanol and ethyl acetate. It turns out that a mixture works better to isolate exactly the lignin fraction that is needed.
Validation
The researchers' model is based on the Flory-Huggins solution theory, a famous mathematical method to quantify the solubility of polymers. Typically, this model is applied to study how a polymer reacts with a solvent, but the researchers went a few steps further.
“Our model is very suitable when you have hundreds of different types of polymer at the same time. “With the solvent we can model the interactions of numerous types of lignin polymer with different chemical properties, such as polymer chain length and composition,” says Vis. “Understanding these interactions is crucial because they influence whether or not a particular type of lignin dissolves in a particular solvent.”
To validate the new model, the researchers calculated the fractionation of lignin from wheat straw. They then compared the model data with experiments based on the same materials. Vis adds: “We tested our model against existing data from a commonly used industrial lignin in a different mixture of solvents (methanol and dichloromethane). Our model was applied with minimal effort and described the data very well,” adds Vis.
So far, the use of numerical tools to predict lignin yield is all rather speculative. “There are not many people who use the theory to predict yield,” says Remco Tuinier, professor at the Department of Chemical Engineering and Chemistry and co-author of the article. “Our model makes it possible to easily predict which lignin can be isolated with a certain mixture of solvents. It is an important development for the field.”
Next steps
Now that the model has proven to be so successful in predicting lignin yields, the question arises: how can this model have an impact in the industry? Panos Kouris, Chief Technology Officer and co-founder of Vertoro and co-author of the paper: “This model provides a springboard for all lignin valorization activities; both in academia and industry.”
Vertoro is a spin-off company from a public-private partnership with, among others, TU/e and wants to offer viable and affordable biobased alternatives to fossil raw materials. Kouris and his colleagues are therefore well aware of the impact the model can have on both academia and industry.
“In academia, the model could spark new lines of research into new solvents and types of lignin, in addition to finding ways to use certain chemical properties of lignin for certain applications,” Kouris notes. “And in the biomass biorefinery industry, the model can provide many insights and contribute to the design of new products based on lignin.”
To prepare
In theory, biorefineries can already use the model to investigate the valorization of certain lignin species, but much still needs to be done before the model is ready for large-scale commercial use.
First, the model must be validated for the most common types of lignin processed by the industry. Subsequently, the solvent fractionation technology itself must have reached the level where the technology is ready for commercial use. Finally, clear applications of the end products on the market are also needed, such as bio-based packaging or biofuels.
Meeting these demands will take time, but Kouris and his colleagues at Vertoro are optimistic that the model will have an impact on biorefineries sooner rather than later. “At Vertoro, we expect the model to be expanded to test various commercially available sources of lignin in the first half of 2024. This mainly concerns biorefineries of second generation cellulosic ethanol, which are actively looking for technologies and solutions to valorize lignin.”
Full article details
'Quantitative Prediction of the Solvent Fractionation of Lignin', Stijn van Leuken et al, Green Chemistry, (2023).
