Can Ice Be Described from First Principles? Meet Young CAS PI Sigbjørn Løland Bore

Ice is everywhere around us – from the poles and glaciers to the small cubes clinking in a glass of water. It plays a crucial role in shaping our planet, but even with today’s powerful computers, researchers still struggle to describe its behaviour precisely. That challenge is at the heart of Sigbjørn Løland Bore’s Young CAS project, Can Ice Be Described from First Principles?

Bore, a postdoctoral associate at the University of Oslo, aims to improve how scientists model ice at the atomic level. By developing new empirical corrections to what are known as first-principle calculations, his work could make computer simulations of ice more accurate and reliable.

“Ice is a constant companion for us humans on Earth,” Bore writes in his project proposal. “Of all the fresh water available, ice makes up 69 % and covers 10 % of the Earth’s surface. On the one hand, ice is fundamental for humans, keeping water levels manageable for current human habitation, reflecting sunlight to slow down global warming, preserving food, or facilitating leisure activities, such as skiing in Nordmarka during winter. But, on the other hand, ice can pose an existential threat, for example, making you miss out on Christmas dinner due to long queues while waiting for the deicing of the airplane.”

Although researchers have used first-principle calculations for decades to study materials like ice, these models still fail to predict correctly the temperatures and pressures at which ice melts. Bore’s project brings together experts on ice and computational chemistry to tackle this long-standing problem. “In practice, approximation leads to simulations failing at predicting when the ice melts, often by as much as 30 Celsius or even worse,” he explains. “My project gathers experts on ice and first-principles calculations to find a solution to this problem.”

If successful, the work will enable more trustworthy computer simulations of ice – or what Bore calls “computational microscopy” – making it possible to study its properties under conditions that are hard or impossible to reproduce in a lab.
 

From Stavanger to San Diego – and back again

Bore grew up in Stavanger and has followed an international path through physics and chemistry. He holds a master’s in applied physics from NTNU in Trondheim and another in Physics of Complex Systems from Université Pierre et Marie Curie in Paris (now Sorbonne).

“After physics, I luckily stumbled into the Department of Chemistry at the University of Oslo to the Centre for Theoretical and Computational Chemistry (CTCC), which was followed by a new centre of excellence called Hylleraas for a Ph.D. in theoretical chemistry,” he says. There, under the supervision of Professor Michele Cascella, he began working on approximate methods to simulate large systems of atoms – “typically millions.”

After his PhD, Bore spent two years as a postdoctoral researcher at the University of California, San Diego, under Professor Francesco Paesani. “Here, my research shifted away from very large biological systems to working simply on water, modelling only a few hundred molecules at the time,” he says. “It turns out that small systems can be just as interesting and, for me, really satisfying as one can apply more accurate methods that really get to the bottom of how nature, or in my case, how water, behaves.”

Now back at the Hylleraas Centre, Bore is continuing to develop computational tools – including machine learning – to make simulations faster and more efficient. “Essentially, the goal is to reduce the time a computer experiment takes by, instead of solving the Schrödinger equation numerically, taking minutes, predicting the solution by machine learning model in milliseconds,” he explains. “This allows us to do computer experiments on days that would otherwise take years.”

The Young CAS project will, in turn, strengthen this work: “For such machine learning models to accurately model ice, the machine learning models need to be trained on very accurate energies from first principle calculations. Therefore succeeding with the research of Young CAS will provide me with a powerful tool for my current research.”
 

Fresh ideas for an old problem

Bore first realised the limitations in the current models while working in California. “It was during this postdoc that I came to understand the limitations in current modelling from first principles, which can capture an incredible amount of chemistry, but really struggle to predict ice melts,” he says. “It was this realisation and the thought that there must be some way of fixing this problem just for water that made me apply for the Young CAS programme.”

The Young CAS Grant gives early-career researchers the opportunity to lead their own collaborative projects at the Centre for Advanced Study (CAS). For Bore, it offers the right mix of focus, independence, and collaboration. “I thought that this fellowship was perfect for this problem, as it would allow gathering a group of young scientists, all under forty, with fresh ideas and the overconfidence needed to crack this long-standing problem,” he says with a smile.
 

Looking forward

“What I am most looking forward to,” he adds, “is having the opportunity of focusing together with an excellent group of young scientists, whom I would not have the opportunity of working with without Young CAS, on a problem with a really specific goal. Then, I am also really looking forward to spending time on the beautiful premises of the National Academy, where I have previously enviously watched other CAS and Young CAS receivers from the Hylleraas Centre having the time of their life.”

Through this project, Bore and his collaborators aim to get one step closer to understanding ice – a material that affects everything from our climate to everyday life. And in doing so, the project captures what the Young CAS initiative is all about: giving young researchers the space to explore big scientific questions with curiosity, rigour, and creativity.Bore's project aims to enhance the accuracy of computational models used to study ice, by developing new empirical corrections to first principle calculations.