arXiv subjects

Thomas Stielow

Publications and source records attributed to Thomas Stielow.

5 recordsLinked to original sources

arXiv2021

Reconstruction of nanoscale particles from single-shot wide-angle FEL diffractions patterns with physics-informed neural networks

Single-shot wide-angle diffraction imaging is a widely used method to investigate the structure of non-crystallizing objects such as nanoclusters, large proteins or even viruses. Its main advantage is that information about the three-dimensional structure of the object is already contained in a single image. This makes it useful for the reconstruction of fragile and non-reproducible particles without the need for tomographic measurements. However, currently there is no efficient numerical inversion algorithm available that is capable of determining the object's structure in real time. Neural networks, on the other hand, excel in image processing tasks suited for such purpose. Here we show how a physics-informed deep neural network can be used to reconstruct complete three-dimensional object models of uniform, convex particles on a voxel grid from single two-dimensional wide-angle scattering patterns. We demonstrate its universal reconstruction capabilities for silver nanoclusters, where the network uncovers novel geometric structures that reproduce the experimental scattering data with very high precision.

physics.data-an↗

arXiv2020

Fast reconstruction of single-shot wide-angle diffraction images through deep learning

Single-shot X-ray imaging of short-lived nanostructures such as clusters and nanoparticles near a phase transition or non-crystalizing objects such as large proteins and viruses is currently the most elegant method for characterizing their structure. Using hard X-ray radiation provides scattering images that encode two-dimensional projections, which can be combined to identify the full three-dimensional object structure from multiple identical samples. Wide-angle scattering using XUV or soft X-rays, despite yielding lower resolution, provides three-dimensional structural information in a single shot and has opened routes towards the characterization of non-reproducible objects in the gas phase. The retrieval of the structural information contained in wide-angle scattering images is highly non-trivial, and currently no efficient rigorous algorithm is known. Here we show that deep learning networks, trained with simulated scattering data, allow for fast and accurate reconstruction of shape and orientation of nanoparticles from experimental images. The gain in speed compared to conventional retrieval techniques opens the route for automated structure reconstruction algorithms capable of real-time discrimination and pre-identification of nanostructures in scattering experiments with high repetition rate -- thus representing the enabling technology for fast femtosecond nanocrystallography.

physics.atm-clus↗

arXiv2017

A Green's function approach to giant-dipole systems

In this work we perform a Green's function analysis of giant-dipole systems. First we derive the Green's functions of different magnetically field-dressed systems, in particular of electronically highly excited atomic species in crossed electric and magnetic fields, so-called giant-dipole states. We determine the dynamical polarizability of atomic giant-dipole states as well as the adiabatic potential energy surfaces of giant-dipole molecules in the framework of the Green's function approach. Furthermore, we perform an comparative analysis of the latter to and exact diagonalization scheme and show the general divergence behavior of the widely applied Fermi-pseudopotential approach. Finally, we derive the giant-dipole's regularized Green's function representation.

physics.atom-ph↗

arXiv2017

Angular-Momentum Couplings in Ultra-Long-Range Giant Dipole Molecules

In this article we extend the theory of ultra-long-range giant dipole molecules, formed by an atom in a giant dipole state and a ground-state alkali atom, by angular-momentum couplings known from recent works on Rydberg molecules. In addition to $s$-wave scattering, the next higher order of $p$-wave scattering in the Fermi-pseudopotential describing the binding mechanism is considered. Furthermore, the singlet and triplet channels of the scattering interaction as well as angular-momentum couplings such as hyperfine interaction and Zeeman interactions are included. Within the framework of Born--Oppenheimer theory, potential energy surfaces are calculated in both first-order perturbation theory and exact diagonalization. Besides the known pure triplet states, mixed-spin character states are obtained, opening up a whole new landscape of molecular potentials. We determine exact binding energies and wave functions of the nuclear rotational and vibrational motion numerically from the various potential energy surfaces.

physics.atom-ph↗

arXiv2015

Casimir-Polder interaction from exact diagonalization and state mixing near surfaces

Dispersion forces have a sizeable effect on the energy levels of highly excited Rydberg atoms when brought close to material surfaces. Rydberg atoms experience energy shifts in the GHz range at micrometer distances, suggestive of considerable state admixture. We show that despite the non-applicability of perturbation theory for Rydberg atoms near a surface, the energy shift due to the dispersion interaction can be obtained from an exact diagonalization of the interaction Hamiltonian by finding the zeros of the Pick function. Moreover, we show that contrary to intuition from single-mode approaches, surface-induced state mixing is generally suppressed even for large interaction energies. We give a tailored example where mixing is observable despite this effect.

quant-ph↗