The Luber group at University of Zurich develops forefront computational methods for the study and design of functional materials, e.g. as needed for sustainable hydrogen production.
Modelling is now an essential tool and has become a vital ingredient in today’s research. Our work deals with theoretical and computational methods and their applications at the interface of chemistry, biology, physics, and materials science. Our interdisciplinary group is especially interested in accurate approaches derived from quantum mechanics and which allow exhaustive insight into materials and processes at the atomistic level. This is the prerequisite to understand complex phenomena in great detail and lays the foundation for the informed design of novel materials with desired properties and functions. In this article we will showcase selected examples of our research, including on water oxidation.
The Sun as a clean energy source
Currently, humankind is facing a continuously increasing worldwide energy demand. This is especially serious in view of expected shortage of fossil fuels and global climate change. It is therefore crucial to develop technologies to harvest energy from renewable sources and to convert it into electricity or fuels. The biggest supplier of renewable energy is the Sun, but collecting, and in particular storing, the Sun’s energy is still a challenge.
Artificial water splitting is a promising solution where sun light is used to split water into molecular oxygen and hydrogen, the latter being desirable with respect to energy storage and conversion. The oxidation of water to generate oxygen is called water oxidation. This reaction is still one of the limiting steps for an overall efficient artificial water splitting device. We have therefore especially focused on this part of the water splitting process.
In nature, water oxidation is efficiently carried out in the photosystem II, a big protein complex found in plants, algae, and cyanobacteria. The catalyst itself has a cube-like structure (a so-called ‘cubane’) containing several manganese centers and one redox-inert calcium cation connected via oxo bridges.
Inspired by nature, a promising strategy for the development of novel efficient catalysts is thus to mimic the water oxidation in photosystem II. In collaboration with experimental groups, we have presented one of the rare stable and active artificial cubane catalysts for water oxidation, namely cubanes with cobalt metal centres, and investigated them extensively.
We have used forefront first-principles methods in order to obtain unprecedented access to the structure and behavior of these catalysts. Particular emphasis has been placed on an accurate description of the involved electronic structure of these catalysts as well as their nuclear dynamics. For the latter, we have applied high-performance ab initio molecular dynamics, possibly combined with enhanced sampling methods, which provides a more realistic picture than standard static computational approaches mostly used in this field. Sophisticated inclusion of finite temperature and environmental effects has been found to be key in order to capture the complex behaviour of the catalysts. This has revealed various factors determining catalytic activity.
Flexibility has turned out to be crucial for efficient catalysis. For instance, innovative computational approaches allowed us to observe great nuclear flexibility of the cubane core reminiscent of behaviour found in Nature, which has been observed for artificial water oxidation catalysts for the first time. Moreover, the calculations have demonstrated that subtle changes such as electron transfer between transition metal centers (redox isomerism), facile change of protonation states and spin states, and intramolecular base functionalities can significantly improve the overall water oxidation process.
This understanding has paved the way for important structure-activity relationships and informed design of more efficient catalysts. We have shown how, for example, the introduction of certain ligands and metal-metal co-operativity can improve the thermodynamics and kinetics of water oxidation, and how complex reaction networks including side and decomposition reactions influence catalytic activity. We have investigated Ru-based systems as well, for which a large variety of structurally simpler systems has been available, being perfect model systems for systematic in silico design of novel catalysts.
Linking experiment and theory: spectroscopy
An indispensable component in the development of materials is their spectroscopic characterisation. Depending on the question of interest, a variety of methods are available. Often, the analysis of spectroscopic data is hampered by missing knowledge from their interpretation. Computational exploration provides the necessary link and connects the data to atomistic structure and dynamics of the compounds under study.
On the one side, this requires high level calculations for a precise description of involved systems, which we have worked on using various approaches, lately especially for challenging dynamic ab initio methods. However, a large number of calculations can be necessary if many materials are investigated such as in commonly-used screening approaches. This makes purpose-driven, computationally efficient, yet sufficiently accurate methods necessary, and this has constituted another direction in our research.
X-ray absorption/emission spectroscopy is very valuable, especially for transition metal-containing compounds. We have used various calculations to elucidate the structure of catalysts/enzymes and developed refinement procedures leading to the improved agreement of experimental and computational spectra. Other approaches have concerned resonance with electronically excited states in the UV/Vis range, e.g. based on cutting-edge real time propagation of electronic densities.
Vibrational spectroscopy is another important tool, for which we have developed diverse approaches. Highly accurate simulations, in particular for condensed phase systems, can be achieved using ab initio molecular dynamics, which surpasses standard static approaches in several aspects: The former considers the dynamics of the system under study, finite temperature effects and more realistic thermodynamic conditions than non-dynamic, i.e. static approaches. Moreover, band shapes can be obtained in a more reliable way.
We have presented dynamic approaches for distinct systems, be it Raman spectroscopy for liquids, sum frequency generation for elucidation of interfaces/surfaces, or Raman optical activity for the elucidation of chiral compounds such as biomolecules. Purpose-driven approaches have included subspace iteration methods or embedding methods (using for example, density functional theory) for targeted access to certain properties and selected spectroscopic signatures, as well as in-depth analysis of composition of spectra and underlying mechanisms.
This short summary demonstrates that computational exploration and design is an essential part in modern research, providing highly-valuable additional information complementary to experimental work. Ideally, calculations are carried out at the first place before time- and cost-intensive, potentially toxic, experiments are conducted. Our developments have dealt with various novel directions for first-principles methods. This has led to highly sought-after insights into, for instance, solar light-driven water splitting for sustainable energy conversion and storage, the accurate spectroscopic characterisation of systems, and the innovative design of new materials for catalysis.
About the principal investigator
Sandra Luber is professor in theoretical and computational chemistry at University of Zurich, Switzerland. Luber received an MSc degree in chemistry from ETH Zurich in 2007 and a PhD degree in (relativistic) quantum chemistry from the same university in 2009. For postdoctoral studies, she joined Biozentrum at the University of Basel, Switzerland, (2010) and Yale University, USA (2010–2011). She later worked for BASF SE (2012) before becoming project group leader at the University of Zurich where she finished her habilitation and was promoted to SNSF professor in 2017.
Selected research awards include the IBM Research Prize for Computer Modelling and Simulations in Chemistry, Biology, and Materials Science and the ETH medal for an outstanding PhD thesis. Moreover, she has been the first theoretician to receive the renowned Clara Immerwahr Award and the first female scientist to obtain both the Hans G. A. Hellmann Award (awarded annually since 1999) and the Robin Hochstrasser Young Investigator Award for chemical physics. In 2018, she was awarded the prestigious Werner prize of the Swiss chemical society.
Professor Dr Sandra Luber
Department of Chemistry
University of Zurich
+41 44 635 4464