Students develop model to better understand the chemical process of renewable fuels

Editors' notes

This article has been reviewed according to Science X's editorial process and policies. Editors have highlighted the following attributes while ensuring the content's credibility:

fact-checked

peer-reviewed publication

trusted source

proofread

A group of TU Delft bachelor students has developed, as part of the students' minor in Computational Science and Engineering, a new model that accurately predicts the molecular properties of alkanes. This knowledge is crucial for the development of renewable fuels. The model is now regularly used by scientists and has also led to a publication in The Journal of Physical Chemistry B.

Fuels and chemicals from renewable sources will become the standard in the future. But there are still many technological barriers to overcome in the production process.

"For sustainable aviation fuel, shorter alkanes, which are hydrocarbon molecules, are highly suitable because of their favorable molecular properties," explains Thijs Vlugt, professor of Engineering Thermodynamics. "The splitting of long alkanes into short alkanes in zeolites, a substance with nanopores, is essential, but we still don't fully understand what happens during this process."

Four bachelor students in the Computational Science and Engineering minor dedicated six months to this challenge, in addition to their regular courses. "Our goal was to develop a model that could more accurately predict the properties of longer alkanes, as little is known about them," says Josh Sleijfer, one of the students.

"Existing models focus mainly on the carbon atoms in the molecule and their immediate neighbors, but we went a step further and also looked at the 'neighbors of the neighbors.' This broader approach allows us to make much more accurate predictions."

The model is proving invaluable to Vlugt's research. "There are a huge number of possible structures for alkanes. If the molecule contains 20 carbon atoms, this means that almost a quarter of a million variations of the molecule are possible. That's simply too many to measure." A computer model that can predict the properties of all these possible alkanes is therefore very welcome.

"By better understanding the conversion process in zeolites, we can optimize the production process and thus develop better renewable fuels," says Vlugt.

Fresh perspective

Sleijfer, who has a bachelor's degree in both Applied Mathematics and Applied Physics, initially knew little about chemical reactions or the context of the research. "It took some time to fully grasp what it was all about, but once the penny dropped, everything made sense," he says.

Vlugt appreciates the students' diverse backgrounds and their fresh, open-minded approach to the subject. "It challenges you, as a supervisor, to explain the task thoroughly and clearly, which only benefits the research."

A useful contribution to research

The efforts of the group, which also included Jeroen op de Beek, Stach van der Zeeuw and Daniil Zorzos, eventually led to a publication. Sleijfer is very proud of this. "Although the project required some extra time, I was extremely motivated by the fact that our results were actually useful. In a bachelor project, you often work for your own learning experience, but these projects were designed to actually make a useful contribution to ongoing research."

To have a publication to your name before you start your master's degree is something few students can say. Sleijfer is now doing a joint master's in Computer Simulations for Science and Engineering, while his group's model is regularly used in Vlugt's research group. "We're now considering extending the model to other quantities. Maybe that will be another interesting challenge for a new team of students," says Sleijfer.

More information: Shrinjay Sharma et al, Prediction of Thermochemical Properties of Long-Chain Alkanes Using Linear Regression: Application to Hydroisomerization, The Journal of Physical Chemistry B (2024). DOI: 10.1021/acs.jpcb.4c05355

Journal information: Journal of Physical Chemistry B

Provided by Delft University of Technology