What happens when molecules collide?

How HTC helps researchers to simulate chemical reactions

Chemical reactions are at the core of everything that happens in the Universe. From the thermonuclear fusion that powers the Sun, to how antibiotics help to fight pneumonia, everything depends on what happens when molecules collide and interact to form new compounds.

Chemist Ernesto García, based at the University of the Basque Country in Vitoria (Spain), has dedicated his academic career to understanding chemical reactions from a theoretical perspective. “My main scientific goal is to compute accurately the efficiency of molecular processes in which molecules collide to react, dissociate, exchange energy and deform,” says García.

García creates computational models to describe reactions that are important for studying natural phenomena or industrial processes. Having good theoretical models to predict molecular behaviour means that simulations will be realistic and useful to tackle research problems in the real world.

Accurate models of molecular collisions take into account many types of parameters (for example kinetic energies, the shape of the molecules, thermal properties). García uses a workflow called Grid Empowered Molecular Simulator (GEMS) to streamline the computational work of the calculations.

GEMS was developed by the team of Antonio Laganà at the University of Perugia in Italy and is powered by High-Throughput Compute resources made available by the CompChem Virtual Organisation.

In the last four years, García worked in projects ranging from astronomy, to applied chemistry and atmospheric science. He has submitted about 2.5 millions of jobs for a total of 31 millions of CPU hours and published eight papers in peer-reviewed journals and a lot of results awaiting publication.

GEMS in action

Chemical evolution of interstellar clouds

Interstellar clouds are amalgamations of gas, plasma and dust scattered across the Universe. In Rampino et al. 2016, Garcia and his team looked into how temperature influences their chemical evolution.

The team modelled the formation of C2+ (an ion with a chemical bond between to carbon atoms and therefore a precursor of longer hydrocarbon chains) from one atom of carbon and the methylidine radical, CH+ (ubiquitous throughout the interstellar space) and found something surprising: its rates of formation in the interstellar clouds are several orders of magnitude different from the values used in current astronomical models.

Modelling nitrogen plasma

In Esposito et al. 2017 the team modelled nitrogen plasmas, like those that surround spacecrafts when they enter the Earth’s or Titan’s atmosphere. Under these circumstances, the temperature can reach tens of thousands of degrees.

Thanks to the EGI grid, it was possible to calculate the collision induced dissociation rate of the nitrogen molecules in several vibrational excited states by collision with both nitrogen atoms and nitrogen molecules.

Ernesto García’s EGI usage

In the last 4 years, García submitted about 2.5 millions High-Throughput Compute jobs for a total consumption of 31 millions CPU hours. The compute resources were provided by the compchem virtual organisation.

About GEMS

Laganà, A. Costantini, O. Gervasi, N. Faginas Lago, C. Manuali, S. Rampino: COMPCHEM: progress towards GEMS a Grid Empowered Molecular Simulator and beyond, Journal of Grid Computing, 8(4), 571-586 (2010)

Selected papers

Esposito et al. 2017 Plasma Sources Science and Technology. doi:10.1088/1361-6595/aa5d27.
(abstract only)

Rampino et al. 2016 Monthly Notices of the Royal Astronomical Society. doi: 10.1093/mnras/stw1116.
(abstract only)

Garcia et al. 2016 Journal of Physical Chemistry. doi:10.1021/acs.jpcb.5b06423.
(abstract only)

Pacifici et al. 2016 Journal Physical Chemistry A. doi:10.1021/acs.jpca.6b00564.
(abstract only)

Garcia et al. 2015 Chemical Physics Letters. doi: 10.1016/j.cplett.2014.12.021.
(abstract only)

Kurnosov et al. 2014 Journal of Computational Chemistry. doi: 10.1002/jcc.23545.
(abstract only)