Hydrogen atoms cannot hide from x-rays anymore but can instead be detected very reliably in routine measurements.

High-resolution low-temperature synchrotron X-ray diffraction data of the salt L-phenylalaninium hydrogen maleate are used to test the new automated iterative Hirshfeld atom refinement (HAR) procedure for the modelling of strong hydrogen bonds. The HAR models used present the first examples of Z' > 1 treatments in the framework of wavefunction-based refinement methods. L-Phenylalaninium hydrogen maleate exhibits several hydrogen bonds in its crystal structure, of which the shortest and the most challenging to model is the O-H...O intramolecular hydrogen bond present in the hydrogen maleate anion (O...O distance is about 2.41 Å). In particular, the reconstruction of the electron density in the hydrogen maleate moiety and the determination of hydrogen-atom properties [positions, bond distances and anisotropic displacement parameters (ADPs)] are the focus of the study. For comparison to the HAR results, different spherical (independent atom model, IAM) and aspherical (free multipole model, MM; transferable aspherical atom model, TAAM) X-ray refinement techniques as well as results from a low-temperature neutron-diffraction experiment are employed. Hydrogen-atom ADPs are furthermore compared to those derived from a TLS/rigid-body (SHADE) treatment of the X-ray structures. The reference neutron-diffraction experiment reveals a truly symmetric hydrogen bond in the hydrogen maleate anion. Only with HAR is it possible to freely refine hydrogen-atom positions and ADPs from the X-ray data, which leads to the best electron-density model and the closest agreement with the structural parameters derived from the neutron-diffraction experiment, e.g. the symmetric hydrogen position can be reproduced. The multipole-based refinement techniques (MM and TAAM) yield slightly asymmetric positions, whereas the IAM yields a significantly asymmetric position.

In this work, the quality of the electron density in crystals reconstructed by the multipolar model (MM) and by X-ray wavefunction refinement (XWR) is tested on a set of high-resolution X-ray diffraction data sets of four amino acids and six tripeptides. It results in the first thorough validation of XWR. Agreement statistics, figures of merit, residual- and deformation-density maps, as well as atomic displacement parameters are used to measure the quality of the reconstruction relative to the measured structure factors. Topological analysis of the reconstructed density is carried out to obtain atomic and bond-topological properties, which are subsequently compared to the values derived from benchmarking periodic DFT geometry optimizations. XWR is simultaneously in better agreement than the MM with both benchmarking theory and the measured diffraction pattern. In particular, the obvious problems with the description of polar bonds in the MM are significantly reduced by using XWR. Similarly, modeling of electron density in the vicinity of hydrogen atoms with XWR is visibly improved.

Hirshfeld atom refinement is one of the most successful methods for the accurate determination of structural parameters for hydrogen atoms from X-ray diffraction data. This work introduces a generalization of the method [generalized atom refinement (GAR)], consisting of the application of various methods of partitioning electron density into atomic contributions. These were tested on three organic structures using the following partitions: Hirshfeld, iterative Hirshfeld, iterative stockholder, minimal basis iterative stockholder and Becke. The effects of partition choice were also compared with those caused by other factors such as quantum chemical methodology, basis set, representation of the crystal field and a combination of these factors. The differences between the partitions were small in terms of R factor (e.g. much smaller than for refinements with different quantum chemistry methods, i.e. Hartree–Fock and coupled cluster) and therefore no single partition was clearly the best in terms of experimental data reconstruction. In the case of structural parameters the differences between the partitions are comparable to those related to the choice of other factors. We have observed the systematic effects of the partition choice on bond lengths and ADP values of polar hydrogen atoms. The bond lengths were also systematically influenced by the choice of electron density calculation methodology. This suggests that GAR-derived structural parameters could be systematically improved by selecting an optimal combination of the partition and quantum chemistry method. The results of the refinements were compared with those of neutron diffraction experiments. This allowed a selection of the most promising partition methods for further optimization of GAR settings, namely the Hirshfeld, iterative stockholder and minimal basis iterative stockholder.

No abstract

Hydrogen atom positions can be obtained accurately from X-ray diffraction data of hydrogen maleate salts via Hirshfeld atom refinement.

The position of the hydrogen atom inside the strong and short intramolecular hydrogen bond of the hydrogen maleate anion strongly varies depending on the crystalline environment. Therefore, it has not been possible in the past to accurately determine it using X-ray diffraction data although there are 292 hydrogen maleate crystal structures with different cations in the literature. In this study, a geometric correlation for the accurate prediction of the hydrogen position in the short intramolecular hydrogen bond is presented. The results used to derive the correlation are obtained from low-temperature neutron-diffraction studies on nine different hydrogen maleate salts that span the whole range from perfectly symmetric to highly asymmetric intramolecular hydrogen bonds. Since the only variable in the correlation as derived from the neutron data is the O•••O distance, the hydrogen atom position in question can subsequently be predicted using information that is accurately available from routine X-ray data. The procedure is tested using high-resolution low-temperature synchrotron X-ray diffraction structures of the same compounds, before it is applied to X-ray data sets found in the literature in which the hydrogen atom position was not determined accurately or not determined at all, e.g., using a riding model.

We focused on elastic properties of B-and N-doped graphene in wide range of concentrations up to 20%. The Young's, bulk and shear moduli and Poisson's ratio have been calculated by means of the density functional theory for a representative set of supercells with disordered impurity patterns including aggregates. In contrast to earlier work, it is demonstrated that doping with nitrogen even strengthens the graphene layers, whereas incorporation of boron induces large structural and morphological changes seen in simulated STM images. Young's and shear moduli increase or decrease with the doping strength for nitrogen or boron, respectively, while bulk modulus and Poisson's ratio exhibit opposite trends. Elastic properties of samples for both types of impurities are strongly related to the electronic structures, especially for heavy doping (>12%). Local arrangements of dopants and an agregation or separation of impurities play crucial role in the determination of stiffness in the investigated systems. Interestingly, these findings are opossed for B-and N-contained samples.