Organisers

• Nicola Marzari (Oxford University, United Kingdom)
• Angel Rubio (University of the Basque Country, Spain)

Supports

   CECAM

Description

Many of the topics and properties related to the workshop theme areat the frontier of what is currently possible with electronic-structure modeling.

1. Accurate exchange correlation functionals, covering topics such as dispersion forces, or self-interactions: several groups have made great progress in modeling dispersion forces, from the pioneering Lundqvist/Langreth effort, to more recent efforts in Davis (Galli), Trieste (de Gironcoli), and Wien (Kresse), to cite only a few, using RPA and/or the adiabatic connection. Progress in self-interaction free functionals has been slower, with studies by Spaldin (UCSB) and Tsemekhmann (PNNL) on the Perdew-Zunger functional, together with recent developments on the calculations of U and V parameters from first-principles (e.g. Coccioni, U. of Minnesota) or on more broadly-applicable self-interaction free functionals (Yang (Duke) and Dabo (Cermics, Paris)). Very relevant also for strongly-correlated systems (Gross (Berlin)).

2. charge-transfer reactions: these are challenging to the non-local nature of the excitations; in the context of DFT they have been studied e.g. in the group of Rubio (UPV) and Maitra (Hunter college); notable is also the intrudction of constrained functionals (especially Van Voorhis (MIT), but also Marzari (MIT) and Scheffler/Behler (FHI)).

3. reactions in the presence of an electrochemical potential: pioneering work by Norskov (DTU) and Sprik (U Cambridge).

4. electrical and thermal transport: work on Landauer and NEGF is very well established. Much less on inelastic effects in electronic transport (Mauri, Paris VI) and on thermal transport (Mingo (CEA), Broido (Boston College), Galli (UC Davis)).

5. optical absorption and exciton formation: a very active area, with pioneering efforts in Europe (Godby (York), Reining (Paris), Del Sole (Roma), Molinari (Modena), Rincke/Scheffler (FHI), Gonze (Louvain), and many others) and the US (Louie (Berkeley), Hybertsen (BNL), van Schilfgaarde (ASU)). Interesting recent developments in large-scale techniques (Umari (Trieste), and very lively interaction between issues in GW and TDDFT approaches.

6. ion dynamics in an excited state: a very difficult problem, with e.g. pioneering efforts by Martinez (Stanford), or in combined electronic and ionic dynamics by Horsfield (Imperial College)

7. catalytic activity of open-shell systems: abundandt DFT failures, even using hybrid functionals. Closely studied by Truhlar (Minnesota) and the chemistry community (e.g. Siegbahn, Stockholm). In DFT, recent efforts by Marzari (MIT), Oppeneer (Uppsala), using generalized Hubbard techniques

Scientific Objectives

In order to apply the power and gain the benefits of electronic-structure modeling to the characterization and engineering of novel materials, we need to accurately describe and explore extensively the ground-state and excited-state potential energy surfaces for complex reactions in complex environments – thus efficiency and accuracy are of paramount importance. For this reason, novel approaches based on DFT and TDDFT, multi-reference quantum-chemistry calculation, GW/Bethe-Salpeter, and quantum Monte Carlo on optimized nodal surfaces are required. Several key ingredients are needed to achieve these goals - algorithmic, conceptual, or simply computational:

1) Charge-transfer states and oxidation-reduction reactions. Self-interaction in most exchange-correlation functionals over-delocalizes electrons. Examples of this failure include the inability to describe the same ion in different oxidation states (e.g. a ferrous and a ferric ion in the same unit cell will split the extra electron in two, giving rise to two ions with a 2.5 oxidation state, irrespective of the distance or the electrostatic screening between the two ions), or to describe charge transfer between conducting polymers, where holes or electrons will incorrectly delocalize along one chain.

2) Electron-transfer and proton-coupled-electron-transfer processes, A wide variety of processes and reactions in electrochemistry, molecular electronics, and biochemistry involve a non-adiabatic electron transfer process from a “donor” to an “acceptor” via a “bridge”; the excited state (the Marcus energy gap) of an electron transferring from one complex to the other needs to be calculated correctly for thousands of configurations, in order to obtain the free-energy diabatic surfaces for the reactions at hand.

3) Potential energy surfaces of different spin multiplicities. Common exchange-correlation functionals, from PBE to B3LYP, often fail to represent accurately the multiplicity of the ground state or the multiplet splittings. These inaccuracies affect most transition-metal ions: one example is the inability to reproduce the high-spin and low-spin ground state for five-fold and six-fold coordinated iron in the porphyrin macrocycle.

4) Excited-state potential energy surfaces and dynamics. In any photo-induced or photocatalytic process, optical excitations force the system to evolve in a higher potential energy surface, with a competition between fast non-radiative decays (usually leading to the reaction desired) and the radiative re-emission of photons.

5) Realistic descriptions of large-scale systems in complex electrostatic and solvation environments. The electrostatic profile at the active site can greatly influence reaction rates – e.g. electrostatic fields in the heme pocket change the binding energies for CO. Thus it is essential to embed quantum calculations in the correct electrostatic environment 

6) Validation: novel development should be tested on paradigmatic, difficult cases: electronic-structure, vibrational frequencies, and potentially energy surfaces in ground and excited states of small transition-metal compounds, electron affinities of charge-transfer complexes, and optical excitations that include electron-hole exciton binding.

The approaches that are being pursues include the development and application of hybrid-functionals in large-scale DFT and TDDFT calculations; of constrained-, penalty-, and self-interaction corrected functionals that address correctly oxidation-reduction reactions; of non-empirical, Hubbard U approaches to describe faithfully different multiplet surfaces; of efficient numerical techniques to evaluate orbital-dependent exchange-correlation functionals, GW and Bethe-Salpeter approaches; of extensive validation with highly-accurate multi-reference states and quantum Monte Carlo approaches. Equally important will be the integration of complex electrostatic boundary conditions in order to take into account solvation or embedding media, or the large-scale electrostatic profile of the material, device, or molecular complex. 

The areas of application will include photocatalysis and Graetzel cells, energy transfer in nanoparticles, biomimetic catalysts such as oxidoreductases, electron-transfer reactions, polaron conductivity, and photosynthetic complexes. These material systems cover a wide range of chemical processes that are relevant to energy production, from oxygen reduction reactions, to methane monooxydation, to photoinduced hydrogen production, to efficient electron-hole separation in light-harvesting complexes and devices.