Research
Method Development - X-ray Wavefunction Refinement
Conventional methods for structure determination using single-crystal X-ray diffraction data neglect the deformation of the valence electron density, but only model atoms with spherical electron densities. However, it is exactly those valence deformations into bonding and lone-pair regions that are the heart of chemistry. Therefore methods were developed to determine the total electron density experimentally (multipole model, maximum entropy methods), which are unfortunately only accessible to experts. Our new method X-ray wavefunction refinement (XWR) employs quantum chemistry in order to interpret the diffraction experiment in a simple fashion. Its first step Hirshfeld Atom Refinement (HAR) makes localisation of hydrogen atoms from the X-ray data as precise and accurate as from neutron-diffraction data. The second step X-ray constrained wavefunction (XCW) fitting allows to extract crystal field effects, electron correlation and relativistic effects from the experimental data. We continuously improve XWR and work on a corresponding software ( HARt, implemented into Olex2).
Relativistic Effects in the Electron Density
In order to extract relativistic effects with our new method XWR from single-crystal X-ray diffraction data, we synthesise and crystallise organo-metallic molecular compounds bearing heavy elements of the 6th period (e.g., Pt, Au, Hg, Tl, Pb, Bi). Crystal quality must be exceedingly good, so that ultra-high resolution data sets can be measured at the synchrotron SPring-8 in Japan at very low temperatures (< 20K). Subsequently, the data are treated with the method IOTC (infinit order two component) during the crystallographic refinement. Moreover, we carry out many theoretical calculations on test molecules in order to separate effects such as electron correlation, polarisation, core deformation and relativistics from each other.
Electron-Density – Property Relationships in Inorganic Chemistry
We synthesise systematic arrays of compounds that only vary in a single substituent in order to correlate geometric with electron-density parameters across the array. This way, reactions or other chemical processes can be simluated through static crystallographic snapshots along a pseudo-coordinate. Each of these snapshots exhibits a complete experimental electron-density study so that deep insights into the electronic nature of the processes can be gained. Currently we work on penta-coordinated silyl naphtalene compounds peri-substituted with amines that represent an attacking group in a nucleophilic substitution reaction with the varying substituent at the silicon atom representing the leaving group. Other compounds of recent interest are siloxanes where we investigate the change of basicity relative to the Si-O-Si bond angle. The concept can be extended to many other systems and chemical processes.
Experimental Electron Density of Protease Inhibitors
We study to which extend the molecular electron density of a biologically active compound - obtained from the crystal of the pure compound - can be used to simulate the electron density of the same molecule in the biological environment, such as bonded to the enzyme pocket, and to which extend this helps to make predictions about the compound's activity. The current field of application is protease inhibition, where molecules with electrophilic centres (epoxides, vinyl sulfones) block thiol functions in cysteine proteases. The concept can be extended to other families of active ingredients.