It takes a lot of brain power to be a theoretical physicist. It also
takes far more than brain power to be a theoretical physicist.
The calculating minds of University of Delaware physicist Krzysztof
Szalewicz and his collaborators, for example, use more than 26 million
hours annually on Department of Defense computers. They routinely use
UD’s High Performance Computing clusters as well.
And that’s what it takes now to produce increasingly precise information to support new science and advanced applications.
Such muscular machines weren’t available 30 years ago, when an active
debate was going on about the likelihood of ammonia dimers — two joined
ammonia molecules — forming hydrogen bonds.
The debate was an important one. Ammonia is a molecule of
significance on many fronts, including those on our planet and far
beyond it. Understanding the properties of ammonia molecules and how
they interact with other molecules has critical value for industry,
pharmaceuticals, biology and production of environmentally sustainable
fuels, for example.
Szalewicz, an expert in the study and calculation of intermolecular
forces, and his collaborators found a reliable, highly accurate answer
to the question. Their findings were published recently in Nature Communications.
Aling Jing, a graduate student on Szalewicz’s team, was the lead
author. Ad van der Avoird, a theoretical chemist from the Netherlands,
was a third collaborator.
Their work resolves the debate and gives chemists and biologists and
other scientists new confidence as they develop new experiments,
materials and processes.
A bit of background on hydrogen bonds may be helpful to understand how Szalewicz’s new calculations shed light on this issue.
When two hydrogen atoms connect with one oxygen atom the bonds are
strong and are called “covalent” bonds. These strong bonds form water
molecules — H2O.
When two water molecules are near each other, a hydrogen atom from
one molecule will form a bond with the oxygen atom of the other
molecule. This is a hydrogen bond, which is not as strong as the
covalent bond intrinsic to the water molecule, but is still a powerful
part of intermolecular dynamics.
The covalent bond is what holds the water molecule together. The
hydrogen bond is what holds multiple water molecules together, making it
possible to pour yourself a big glass of water.
The hydrogen bonds between water molecules are settled science.
Until 1985, the ammonia-hydrogen bond question was considered
settled, too. An ammonia molecule (NH3) is made of one nitrogen atom
connected to three hydrogen atoms by covalent bonds.
The debate about whether ammonia molecules could form hydrogen bonds
with other ammonia molecules was reopened in 1985, when new experiments
suggested that the ammonia dimers — pairs of ammonia molecules — are not
hydrogen bound, in contrast to the predictions of previous theories.
More calculations, experiments and debate followed.
“Finally, people said ‘It is too hard. We cannot do anything more,’” Szalewicz said.
But as computing muscle became increasingly available, more accurate
calculations were possible, providing increasingly precise pictures of
the mechanisms in play.
Szalewicz and his collaborators now have produced a calculation of
the potential energy surface of the ammonia dimer, which shows how the
interaction energy of the molecules is related to their geometric
shapes.
“What we have found now is that, yes, it was a hard problem,”
Szalewicz said. “The answer is not completely ‘yes, period.’ We cannot
say that.”
What they have shown, with highest confidence, is that ammonia dimers
are quite flexible, not rigid, as the 1985 experiment concluded. This
means that a broad range of intermolecular separations and orientations
is covered during the intermolecular motions.
The published experimental configuration turned out to be an average
between two hydrogen-bonded configurations. This is like a snapshot of
intermolecular motion, which was assumed by the experimental group to be
the most likely configuration, but actually is fairly rare.
By factoring in many more data points, Szalewicz and collaborators
went far beyond single configurations to show that the hydrogen bonds
were far more likely than not. That kind of precision makes a huge
difference in how you incorporate ammonia molecules in various
applications and has many other implications for chemistry.
Szalewicz compares it to taking an extended hike through a mountain range.
“You go up, up, up from a valley to a pass,” he said. “Then you go
down to another valley. If the pass is high above the valley, it is a
hard hike. The valley corresponds to the hydrogen-bonded configurations
and with a high pass, getting from one valley to another is difficult.
Thus, molecules stay mostly in the valleys and finding the dimer at the
top of the pass is a very rare event.
“The ammonia-dimer valley surface is different from those of typical
hydrogen-bonded molecules. Instead of two well-separated valleys, there
is one very narrow one containing both hydrogen-bonded configurations,
with almost no pass between them. Finding the dimer at the top of the
pass is a fairly likely event. Therefore, it could be observed in
experiments.”
This is why experimental physicists need theoretical physicists and
also why theoretical physicists need experimental physicists.
“There is an old joke that is actually very true,” Szalewicz said.
“When an experimentalist publishes a result, everyone believes it —
except the experimentalist, who always knows they might have overlooked
something. When a theorist publishes a result, nobody believes it —
except the theorists.”
When they work together, as they must, great insights are likely.
Experiments also measure excitations of intermolecular motions.
Szalewica and collaborators performed quantum-mechanical calculations of
such excitations, obtaining excellent agreement with the experiment.
This is a strong validation of the correctness of the surface developed
in the calculations.
Using similar calculations with water, Szalewicz has previously
published potential energy surfaces that help to explain properties of
water that have not been previously explained. They now are used by
industrial chemists who work on steam engines and need to know those
properties at various temperatures.
The National Institute of Standards and Technology now recommends
using these theoretical calculations, which have shown greater accuracy
than experimental measurements.
The research was supported by a grant from the National Science Foundation.