On the Completeness of Atomic Structure Representations

Sergey N. Pozdnyakov, Michael J. Willatt, Albert P. Bartók, Christoph Ortner, Gábor Csányi, Michele Ceriotti. arXiv:2001:11696v1
Contributed by Jesper Madsen

Here, I highlight an interesting recent preprint that tries to formalize and quantify something that I previously have posted here at Computational Chemistry Highlights (see the post on Atomistic Fingerprints here), namely how to best describe atomic environments in all their many-body glory. A widely held perception among practitioners of the "art" of molecular simulation is that while we usually restrict ourselves to 2-body effects for efficiency purposes, 3-body descriptions uniquely specify the atomic environment (up to a rotation and permutation of like atoms). Not the case (!) and the authors effectively debunk this belief with several concrete counter-examples. 

FIG. 1: "(a) Two structures with the same histogram of triangles; (angles 45, 45, 90, 135, 135, 180 degrees) (b) A manifold of degenerate pairs of environments: In addition to three points A,B,B′ a fourth point C+ or C− is added leading to two degenerate environments, X + and X − . (c) Degeneracies induce a transformation of feature space so that structures that should be far apart are brought close together."

Perhaps the most important implication of the work is that it in part helps us understand why modern machine-learning (ML) force fields appears to be so successful. At first sight the conclusion we face is daunting: for arbitrarily high accuracy, no n-point correlation cutoff may suffice to reconstruct the environment faithfully. Why, then, can recent ML force fields so accurately be used to calculate extensive properties such as the molecular energy? According to the results of Pozdnyakov, Willatt et al.'s work, low-correlation order representations often suffice in practice because, as they state, "the presence of many neighbors or of different species (that provide distinct “labels” to associate groups of distances and angles to specific atoms), and the possibility of using representations centred on nearby atoms to lift the degeneracy of environments reduces the detrimental effects of the lack of uniqueness of the power spectrum [the power spectrum is equivalent to the 3-body correlation, Madsen], when learning extensive properties such as the energy." However, the authors do suggest that introducing higher order invariants that lift the detrimental degeneracies might be a better approach in general. In any case, the preprint raises many technical and highly relevant issues; and it would be well worth going over if you don't mind getting in the weeds with Maths.   

Electron-Driven Proton Transfer Along H2O Wires Enables Photorelaxation of πσ* States in Chromophore−Water Clusters

More than just shifting state energies, polar solvents may actively participate in the photochemistry of excited molecules

Szabla, R.; Šponer, J.; Góra, R. W. J. Phys. Chem. Lett. 2015, 6, 1467-1471.

Highlighted by Mario Barbatti

For decades, it has been well known that solvents, especially polar ones, have a large impact on the photodynamics of chromophores. Phenomenological models, such as the Lim Proximity effect (1), for instance, have been developed to describe how the energy shift caused by the solvent molecules determines radiative and non-radiative rates.

These early approaches focused mostly on the description of the shape of the potential energy surfaces and on the relative shift between them. From that tradition, we learned, for example, the important heuristic rule telling that, in comparison to the gas phase, water stabilizes the ππ* state of the chromophore, while it destabilizes the nπ* state. Those models, however, did not consider that new electronic states arising from the chromophore-solvent interaction could play a major role in the fate of the excited system.

From the photochemical point of view, the solvent was understood as an important but passive factor. Many simulations relied on this important-but-passive hypothesis, to restrict, for instance, the quantum-mechanical region in QM/MM modelling to the chromophore only, saving precious computational time.

In the last years, however, the important-but-passive hypothesis has been challenged by a number case studies (2). Diverse computational simulations have shown that the solvent may indeed play an active role in photochemistry. (I have myself contributed to the field by showing a case where a water-to-chromophore electron transfer could create a conical intersection (3).)

The paper by Szabla, Šponer, and Góra (4) belongs to this new tradition, the important-and-active hypothesis.

Using surface hopping simulations based on ADC(2) excited states, they investigated the ultrafast dynamics of 2-aminooxazole (AMOX) microsolvated by water. They found out that an important radiationless pathway for the AMOX-(H₂O)₅ cluster involves an electron-driven proton transfer along water wires. This process occurs in an electronic state characterized by an electron transfer from an n orbital at the chromophore to a σ* orbital in one of the water molecules (Fig. 1).

Fig. 1 - Electron-driven proton transfer along water wires.

Szabla, Šponer, and Góra note that similar deactivation mechanism has been observed before in simulations of other heterocyclic chromophores. This means that it may be a common pattern in the photochemistry of these compounds.

Although this is is matter for speculation, all these examples imply that we cannot restrict ourselves to credit polar solvents a passive role only. Any new investigation, either experimental or theoretical, has now to take into account the possibility that the solvent may actively be contributing to the photochemistry.

In particular, for the next generation of excited-state QM/MM simulations, the message (and the cost) is clear: quantum-mechanical microsolvation is simply mandatory.

References
(1) Lim, E. C. Proximity effect in molecular photophysics: dynamical consequences of pseudo-Jahn-Teller interaction. J. Phys. Chem. 1986, 90, 6770-6777. doi: 10.1021/j100284a012

(2) Liu, X.; Sobolewski, A. L.; Borrelli, R.; Domcke, W. Computational investigation of the photoinduced homolytic dissociation of water in the pyridine-water complex. Phys. Chem. Chem. Phys. 2013, 15, 5957-5966. doi: 10.1039/C3CP44585B

(3) Barbatti, M. Photorelaxation Induced by Water–Chromophore Electron Transfer. J. Am. Chem. Soc. 2014, 136, 10246-10249. doi: 10.1021/ja505387c

(4) Szabla, R.; Šponer, J.; Góra, R. W. Electron-Driven Proton Transfer Along H₂O Wires Enables Photorelaxation of πσ* States in Chromophore–Water Clusters. J. Phys. Chem. Lett. 2015, 6, 1467-1471. doi: 10.1021/acs.jpclett.5b00261

Selected publications of Stefan Grimme, Leibniz prize winner 2015

Contributed by Martin Korth

Yesterday, Stefan Grimme received a Leibniz prize, which is awarded by the German Research Foundation DFG every year to ten outstanding German scientists across all(!) fields and including a research grant of 2.5 million Euro each (probably the reason why it is called the 'German Nobel prize'): http://www.dfg.de/en/funded_projects/prizewinners/leibniz_prize/2015/index.html

Only a handful of theoretical chemists can boast to have one dangling over their Victorian fireplace; S. Peyerimhoff (1989), H.-J. Werner (2000), J. Gauß (2005), F. Neese (2010) - probably a club few would mind to join. Other chemists who have received the prize include H. Michel, G. Ertl, H. Schwarz, F. Schüth, ... - again not the usual bunch. (Like Philosophy? J. Habermas got one in 1986, Historical Science? J. Osterhammel did it in 2010, ... just look up the list on Wikipedia)

In honor of Grimme winning the prize, this highlight is devoted to a selection of his papers, with a focus on method development. He is of course already a well-known figure in our community (being amongst the 200 most cited chemists now), but not everyone might be aware of the breadth also of his methodological work - though CCH did it's best with no less than 7 Highlights devoted to his work over the last 3 years!

Here's the list:

DFT/MRCI 1996 1998
Spin-scaled methods 2003a 2003b (2012 review)
DFT-D 2004 2006 2010 - see CompChemHighlight (2011 review)
Double Hybrid functionals 2006a 2006b 2007a 2007b (2014 review)
gCP 2012 - see CompChemHighlight
Supramolecular binding 2012 - see CompChemHighlight
HF-3c 2013 - see  CompChemHighlight
QCEIMS 2013 - see CompChemHighlight 2014
simplified TDA 2013 2014
QMDFF 2014 - see CompChemHighlight
Crystal structure prediction 2014a 2014b

And the bonus numbers are:

Do special pi-pi interactions exist? 2008
Why not to use B3LYP/6-31G* 2012
Dispersion effects 2013 (amongst many other papers on this topic) - see CompChemHighlight
GMTKN benchmark databases 2010 2011 - is anyone NOT using them?

Congratulations to Stefan Grimme, we're looking forward to extend the list!

Computational Chemistry: 2014 in numbers

In 2014 we learnt that two of the germinal DFT papers (by Becke and Lee, Yang and Parr) are amongst the top ten most cited scientific papers of all time, and of chemistry papers published in the last ten years, the fourth and fifth most cited again relate to computational research (Truhlar and Hess, respectively). In this vein I thought it would be interesting to perform a (pseudo)-scientific analysis of the usage of computation in chemistry research in, and in the years leading up to, 2014 as judged by bibliometric data.

Searching all 2014 chemistry papers in the Web of Science for mention of "computation" or "computational" in either the article title, abstract or keywords suggests that approximately 2.7% of chemistry research involved computation of some variety this year (9,101 of a staggering 331,699 papers). This is most likely an underestimate since searching for more specific phrases such as "DFT" will turn up more hits. The same analysis over previous years reveals a steady increase in the proportion of chemistry research using computation from 0.6% in 1994, to 1.1% in 2004 and 2.2% in 2010.

Around 20% of all the computational chemistry papers published in 2014 emanate from the USA, more than double the closest competitor, China. The top ten nations in terms of publications are USA 19.5%, China 9.3%, Germany 6.1%, India 4.3%, France 4.0%, Italy 3.8%, Spain 3.7%, England 3.6%, Japan 2.7% and Canada 2.4% - making nearly 60% of the total output. A decade ago in 2004 the ten most prolific countries accounted for around 87% of total output, which indicates that recent years have witnessed a greater global involvement  in computational chemistry. Noticeable trends are seen in individual nations share of the computational chemistry pie, with the USA and some European nations effectively halving their fraction of papers between 2004 and 2014, with China's output nearly doubling from 5.4% in 2004 to 9.3% in 2014 and the emergence of India from outside the top ten into fourth place in 2014. It should be borne in mind that the globalization of science will inevitably lead to some over counting of papers here, due to multiple addresses appearing on the same paper.