Logo: to the web site of Uppsala University

Biological structure determination has had new avenues of investigation opened due to the introduction of X-ray free-electron lasers (XFELs). These X-ray lasers provide an extreme amount of photons on ultrafast timescales used to probe matter, and in particular biomolecules. The high intensity of the X-rays destroys the sample, though not before structural information has been acquired. The unique properties of the probe provide the unprecedented opportunity to study the un-crystallized form of biological macromolecules, small crystals of biomolecules and their dynamics. 

In this work, we study processes in XFEL imaging experiments that could affect the achievable resolution of the protein structure in a diffraction experiment. Elastic scattering is the process which provides structural information and leaves the sample unperturbed. This interaction occurs far less often compared to damage inducing processes, such as photoabsorption, which leads to rapid ionization of the studied sample. By using density functional theory, we study the effect of ultrahigh charge states in small systems, such as amino acids and peptides, on the subsequent bond breaking and charge dynamics. Reproducible fragmentation patterns are studied in order to find features that could be understood in larger systems, such as proteins. 

Biomolecules are dynamical systems, and the currently used pulse duration is not short enough to outrun the movement of the atoms. The diffraction patterns acquired in an experiment are therefore an incoherent sum of slightly different conformations of the same system. Water can help to reduce these structural variations, but the water molecules themselves will then be a source of noise. Using classical molecular dynamics, we study the optimal amount of water that should be used to achieve the highest resolution. 

To simulate ultrafast molecular dynamics of larger systems such as proteins, we develop a hybrid Monte Carlo/molecular dynamics model. We utilize it to simulate the fragmentation dynamics of small proteins and investigate the possibility to extract structural information from the fragmentation patterns. For larger systems exposed to X-ray lasers, such as viruses and crystals, we develop a hybrid collisional-radiative and classical molecular dynamics approach. The method is used in several projects, both in theoretical studies and to support experiments conducted at XFEL facilities. In particular, we simulate the interaction of hexagonal ice with an X-ray laser, and show the structure makes a phase transition from the native crystal state to a plasma, while still partly retaining structural order. Furthermore, we note that the structural changes occur in an anisotropic manner, where different local structural configurations in ice decay on different time-scales. 

Preliminary experimental results show this anisotropic dynamics in an X-ray pump-probe serial femtosecond X-ray crystallography experiment performed on  I3C crystals. The real space dynamics as a function of probe delay given by our theoretical model and the experiment both show good agreement, where the iodine atoms exhibit correlated motion. The model is also used to calculate the expected atomic displacement and ionization in a hemoglobin crystal, revealing the time and length scales of the dynamics in the protein during the experiment.