Imperial College London

Henry Rzepa

This second report highlights two "themes", or common ideas that seem to
emerge spontaneously from diversely different talks. Most conferences do
have them.

The first is "*embedding*", which in this context means treating
different parts of a probably complex molecular system at different
levels of theory. Thus Emily Carter in her plenary described how a
periodic crystal treated by density functional theory, or DFT could have
an embedded component in which the electronic structures are described
instead by multi-reference correlated wave functions (CAS-PT2). She
illustrated this by discussing what happens when a triplet state oxygen
molecule approaches the surface of an aluminium crystal, and (mostly)
dissociates into surface bound oxygen atoms with Al-O bonds. The spin
state of the oxygen changes smoothly to an overall singlet, with a rapid
transfer of charge at the saddle point in the potential energy surface.
The numbered of embedded Al atoms had to be at least a cluster of 14 to
reproduce the observed reaction barriers (DFT on its own gets a zero
barrier!). This sort of study is important in understanding the details
of what is happening in metal surface catalysis.

Arieh Warshel then addressed the same theme with his own talk entitled
*Multiscale Modeling of Complex Biological Systems and Processes*. Here
you got quantum embedding in a mechanical force field description of
some very large molecules. This was a broad brush talk, but what I did
get out of it was the concept of asymmetry in molecular systems. Whereas
an organic chemist thinks of asymmetry as often relating to just a
single chiral carbon centre in a molecule, nature operates on vaster
scales. Thus the enzyme ATPase has a molecular axle or spindle, which
rotates to assemble the phosphate groups one at a time. This spindle
rotates asymmetrically, i.e. always in a specific direction, and Warshel
attempts to describe the origins of this rotational asymmetry at a
molecular level. Well, this is Nobel prize winning stuff! He followed
this up with filaments that "walk" along surfaces in one (asymmetric)
direction, first lifting up one point of attachment, and then
re-attaching at a different point such that the filament develops a
clear sense of direction in its walk. This of course is all done with
molecular dynamics, and (I think) has its origins in subtle
electrostatics.

Stefan Grimme in his plenary also described dynamic processes, this time
those that happen in a mass spectrometer when a molecule is ionised by
electron impact. Removal of an electron produces a complex set of
ionised states, in which many different single bonds may be weakened due
to this ionisation. He developed simplified  DFT (sDFT) methods that can
be applied to molecular dynamics, and assembled a "black box" which
predicts the expected fragmentations over a time scale of a ps or so. By
sampling the trajectories, he estimated the *intensities* of the various
positively charged species and overlaid this on the observed EI-MS. The
agreement was often spectacular. A particularly interesting example was
the fragmentation of taxol. Here, no molecular ion is found, only much
lighter ions. The molecular dynamics shows that rather than consecutive
single-bond fragmentations, you instead get multiple bonds more or less
all fragmenting at the same time. Tougher was to reproduce
rearrangements, such as the McLafferty. Here, the semi-empirical method
OM2 was more successful. His work means you can just
"*dial-a-mass-spectrum*" and he speculates whether getting a good fit
with the observed spectrum could tell you subtle aspects of the
gas-phase molecular species, what its tautomeric state might be or
perhaps even its conformation. He also described large-scale (800+) atom
simulations of electronic circular dichroism (ECD) spectra of
organometallic systems. Octahedral complexes can be prepared in chiral
form, and this theoretical ECD treatment allows determination of
absolute configuration of these often non-crystalline systems. Here you
often need to compute 1000 or more electronic states, and if you have
ever tried such ECD simulations, you will know that this is a lot of
states!

We had been expecting Stefan to talk about dispersion effects in
molecules, another emerging theme. Instead lots of other people
mentioned them. In [my
talk](http://dx.doi.org/10042/a3v14){target="_blank"} I showed how
including a D3-dispersion correction could dramatically change the
predicted enantioselectivity of a chiral aldol
condensation.\[cite\]http://dx.doi.org/10042/a3v14\[/cite\]

The above observations of course cannot be in the least representative;
typical of a modern conference there are five parallel sessions and 400+
posters, and so it represents a highly personal and selective snapshot.

::: {.wp_orcid_field}
[https://orcid.org/0000-0002-8635-8390](https://orcid.org/0000-0002-8635-8390){target="_blank"
rel="author"}
:::


This second report highlights two “themes”, or common ideas that seem to emerge spontaneously from diversely different talks. Most conferences do have them. The first is “
<em>
 embedding
</em>
”, which in this context means treating different parts of a probably complex molecular system at different levels of theory.

Messaggi di Rogue Scholar

WATOC2014 Conference report. Emergent themes.