TYC Lunchtime Get Together

TYC@UCL Lunchtime Get-Togethers consist of two short presentations from either students, visitors or academics talking place over lunch and drinks. They are a great opportunity to get to meet other students, catch up with your friends and colleagues, and listen to members and visitors talk about their research.

Food and drink will be provided so please just bring yourselves!

This month's talks are from:

Proton ordering, nucleation and surface reactivity of cubic ice

Zamaan Raza

Hexagonal ice (Ih) is the condensed phase of water most commonly encountered on Earth. Cubic ice (Ic) is another ambient pressure phase that differs only in terms of the stacking order of hexagonal bilayers (ABAB for hexagonal, ABCABC for cubic). Cubic ice gradually anneals to hexagonal ice as it is heated from 180K; there is no distinct phase transition, and the associated enthalpy change is minuscule, estimated to be between -160 and -13J/mol. Ice nucleates homogeneously as cubic ice in water droplets with radii of 0.005 to 5nm, at temperatures between 160K and 243K, conditions which are typical of the Earth's middle to upper atmosphere. It is therefore expeted to have a significant impact on processes such as cloud dehydration and ozone depletion.

Using high precision density functional theory calculations, we have established the ground state proton ordered configuration of cubic ice, dubbed ice XIc (analogous to proton ordered hexagonal ice XI), and estimated the energy difference between proton disordered Ic and Ih by comparing the proton ordered ground states (Raza et al. Phys. Chem. Chem. Phys. 2011, 13, 19788). The variation in energy between bulk proton ordered configurations is less than 1kJ/mol, but proton ordering has been shown to affect the surface energy of Ih by an order of magnitude more and is a function of the pattern of dangling OH bonds. We find that the same relationship holds for cubic ice, albeit with a (surprisingly) higher surface energy.

We are currently performing brute force nucleation of ice using a variety of common empirical forcefield models in order to determine whether they are intrinsically biased towards nucleation of Ih or Ic. This could be particularly important with respect to biological simulations, where the TIP3P model is commonly employed. We are also examining the catalytic role of ice in reactions occuring in the Earth's atmosphere and interstellar clouds, such as the reaction of formaldehyde with monatomic hydrogen.


Application of the CASSCF approach to f-element complexes

Andy Kerridge

Complexes of the 4f (lanthanide) and 5f (actinide) elements provide a significant challenge to computational chemistry: strong electron correlation and relativistic effects combine with weak ligand fields to produce states which cannot be well described by methods that rely on single-configurational descriptions of the electronic structure. In this talk I will give an overview of the complete-active-space self-consistent-field (CASSCF) approach. CASSCF presents a method by which multi-configurational states can be described via a partitioning of the orbital space into chemically 'active' and 'inactive' subspaces: full configuration interaction with orbital relaxation is performed within the active space in order to i) give a multiconfigurational representation of the wavefunction, and ii) include the contribution to the total energy from static correlation.
Dynamical correlation both within the inactive space and between the active and inactive spaces is incorporated via perturbational methods.

We have used the CASSCF approach in order understand the electron structure of lanthanide and actinide metallocenes M(C8H8)2, M=(Ce, Th, U, Pu, Cm) and shed more light on the problem of the apparent oxidation state ambiguity of cerocene. We have also applied the methodology to studies or the uranyl dication, UO22+, which is a ubiquitous aqueous phase uranium complex. Uranyl, along with its neptunium and plutonium analogues can be complexed by the penta- and hexapyrrolic expanded porphyrins, and we are currently working on methods to allow us to describe such complexes within the CASSCF framework.