Why can geckos walk up walls? Why does nitrogen become liquid at –196 °C? Many everyday phenomena can be explained by van der Waals forces – weak bonds between molecules that are notoriously difficult to calculate. For years, scientists have struggled with the fact that different computational methods produced conflicting results.
Now, researchers at TU Wien have resolved this discrepancy and found a solution. Ironically, it was the very method long considered the “gold standard” of quantum chemistry that turned out to be the source of the error: it systematically overestimates the energy contained in certain molecular bonds. With an improved variant, the TU Wien team can now correctly predict the behavior of large molecules – an essential step for understanding biological systems and for advancing renewable energy technologies.
A Mystery of Chemistry
“To describe the bonds between large molecules, scientists use different computational approaches,” explain Tobias Schäfer and Andreas Irmler, first authors of the new study. Together with Alejandro Gallo and research group leader Prof. Andreas Grüneis, they compared the most widely used methods.
“One option is to use quantum Monte Carlo simulations,” says Schäfer. “Here, the computer explores countless possible arrangements of electrons – keeping energetically favorable ones and discarding unfavorable ones. Another option is the so-called coupled-cluster approach,” adds Irmler. “In that case, the molecules are treated in their low-energy states, and higher-energy configurations are added later as a kind of correction.”
“This coupled-cluster method has long been regarded as the gold standard,” says Schäfer. “But the closer we looked, the clearer it became that there were small yet persistent deviations compared to the Monte Carlo results – and for years, nobody knew why.”
Now the TU Wien team has found the answer: “We discovered that the coupled-cluster method systematically overestimates binding energies in large, highly polarizable molecules,” explains Irmler. “Our improved variant corrects this deviation without significantly increasing the computational cost.” With this correction, the results now align much more closely with quantum Monte Carlo data.
Large Molecules – Large Importance
This advance is particularly crucial for large molecular systems. “If you want to describe molecules containing up to a hundred atoms, the computational effort becomes enormous,” says Alejandro Gallo. “Even the world’s largest supercomputers reach their limits. To achieve reliable predictions, we need highly sophisticated approximation methods.”
And large molecules are becoming ever more important – in fields ranging from materials research to pharmaceutical development. “If we want to understand how a drug crystallizes inside a tablet, or how strongly a material binds hydrogen for energy storage, we need to model van der Waals forces accurately,” says Schäfer.
From Fundamental Theory to Practical Applications
The new method enables more reliable reference data – not only for traditional simulations but also as training data for AI models. Such models are already being used to design new materials and pharmaceuticals in virtual environments.
“We are building a bridge between ultimate accuracy and practical usability,” says Prof. Andreas Grüneis from the Institute of Theoretical Physics at TU Wien. “This opens up new possibilities for materials science. Our results show that even well-established methods must be continuously re-examined to keep pace with the growing demands of modern research.”
Original publication
Contact
Prof. Andreas Grüneis
Institute for Theoretical Physics
TU Wien
+43 158801 13662
andreas.grueneis@tuwien.ac.at
https://cqc.itp.tuwien.ac.at, opens an external URL in a new window