Search

The QM9 dataset [Scientific Data, Vol. 1, 140022 (2014)] became a standard dataset to benchmark machine learning methods, especially on molecular graphs. It contains geometries as well as multiple computed molecular properties of 133,885 compounds at B3LYP/6-31G(2df,p) level of theory, including frontier orbitals (HOMO and LUMO) energies. However, the accuracy of HOMO/LUMO predictions from density functional theory, including hybrid methods such as B3LYP, is limited for many applications. In contrast, the GW method significantly improves HOMO/LUMO prediction accuracy, with mean unsigned errors in the GW100 benchmark dataset of 100 meV. In this work, we present a new dataset of HOMO/LUMO energies for the QM9 compounds, computed using the GW method. This database may serve as a benchmark of HOMO/LUMO prediction, delta-learning, and transfer learning, particularly for larger molecules where GW is the most accurate but still numerically feasible method. We expect this dataset to enable the development of more accurate machine learning models for predicting molecular properties

Open access to research data is increasingly important for accelerating research. Grant authorities therefore request detailed plans for how data is managed in the projects they finance. We have recently developed such a plan for the EU-H2020 BIG-MAP project - a cross-disciplinary project targeting disruptive battery-material discoveries. Essential for reaching the goal is extensive sharing of research data across scales, disciplines and stakeholders, not limited to BIG-MAP and the European BATTERY 2030+ initiative but within the entire battery community. The key challenges faced in developing the data management plan for such a large and complex project were to generate an overview of the enormous amount of data that will be produced, to build an understanding of the data flow within the project and to agree on a roadmap for making all data FAIR. This paper describes the process we followed and how we structured the plan., Comment: 12 pages, 4 figures; the full DMP is published as Supporting Information

Organic semiconductors are indispensable for today's display technologies in form of organic light emitting diodes (OLEDs) and further optoelectronic applications. However, organic materials do not reach the same charge carrier mobility as inorganic semiconductors, limiting the efficiency of devices. To find or even design new organic semiconductors with higher charge carrier mobility, computational approaches, in particular multiscale models, are becoming increasingly important. However, such models are computationally very costly, especially when large systems and long time scales are required, which is the case to compute static and dynamic energy disorder, i.e. dominant factor to determine charge transport. Here we overcome this drawback by integrating machine learning models into multiscale simulations. This allows us to obtain unprecedented insight into relevant microscopic materials properties, in particular static and dynamic disorder contributions for a series of application-relevant molecules. We find that static disorder and thus the distribution of shallow traps is highly asymmetrical for many materials, impacting widely considered Gaussian disorder models. We furthermore analyse characteristic energy level fluctuation times and compare them to typical hopping rates to evaluate the importance of dynamic disorder for charge transport. We hope that our findings will significantly improve the accuracy of computational methods used to predict application relevant materials properties of organic semiconductors, and thus make these methods applicable for virtual materials design.

Since the discovery of the TTF-TCNQ charge transfer complex as the first metallic material composed of molecules, many other molecular metals were reported. It was however only recently that the first metal-free single-component organic metal was characterized by Kobayashi et al. Although the measured properties of a poorly crystalline sample clearly showed metallic behavior, the crystal structure itself could not be solved, so that the conduction mechanism in this material is still unknown. Here, we present the results of theoretical crystal structure prediction calculations for the TED molecule, accompanied by electronic DOS and band structure calculations which indicate band transport., Comment: Main text: 5 pages including 3 figures. Appendix: 2 tables and 1 figure

Amorphous organic semiconductors based on small molecules and polymers are used in many applications, most prominently organic light emitting diodes (OLEDs) and organic solar cells. Impurities and charge traps are omnipresent in most currently available organic semiconductors and limit charge transport and thus device efficiency. The microscopic cause as well as the chemical nature of these traps are presently not well understood. Using a multiscale model we characterize the influence of impurities on the density of states and charge transport in small-molecule amorphous organic semiconductors. We use the model to quantitatively describe the influence of water molecules and water-oxygen complexes on the electron and hole mobilities. These species are seen to impact the shape of the density of states and to act as explicit charge traps within the energy gap. Our results show that trap states introduced by molecular oxygen can be deep enough to limit the electron mobility in widely used materials., Comment: 13 pages + SI, 7 figures + TOC-graphic + 2 figures in SI

Catalytic hydrogenation of graphite has recently attracted renewed attention, as a route for nano-patterning of graphene and to produce graphene nano-ribbons. These reports show that metallic nanoparticles etch surface layers of graphite, or graphene anisotropically along the crystallographic zigzag <11-20> or armchair <1010> directions. On graphene the etching direction can be influenced by external magnetic fields or the substrate. Here we report the sub-surface etching of highly oriented pyrolytic graphite (HOPG) by Ni nanoparticles, to form a network of tunnels, as seen by SEM and STM. In this new nanoporous form of graphite, the top layers bend inward on top of the tunnels, while their local density of states remains fundamentally unchanged. Engineered nanoporous tunnel networks in graphite allow further chemical modification and may find applications in storage or sensing.

Theoretical approaches of electronic transport in correlated molecules usually consider an extended molecule, which includes, in addition to the molecule itself, parts of electrodes. In the case where electron correlations remain confined within the molecule, and the extended molecule is sufficiently large, the current can be expressed by means of Laudauer-type formulae. Electron correlations are embodied into the retarded Green function of a sufficiently large but isolated extended molecule, which represents the key quantity that can be accurately determined by means of ab initio quantum chemical calculations. To exemplify these ideas, we present and analyze numerical results obtained within full CI calculations for an extended molecule described by the interacting resonant level model. Based on them, we argue that for organic electrodes the transport properties can be reliably computed, because the extended molecule can be chosen sufficiently small to be tackled within accurate ab initio methods. For metallic electrodes, larger extended molecules have to be considered in general, but a (semi-)quantitative description of the transport should still be possible particularly in the typical cases where electron transport proceeds by off-resonant tunneling. Our numerical results also demonstrate that, contrary to the usual claim, the ratio between the characteristic Coulomb strength and the level width due to molecule-electrode coupling is not the only quantity needed to assess whether electron correlation effects are strong or weak.

We study the current and shot noise in a linear array of metallic nanoparticles taking explicitly into consideration their discrete electronic spectra. Phonon assisted tunneling and dissipative effects on single nanoparticles are incorporated as well. The capacitance matrix which determines the classical Coulomb interaction within the capacitance model is calculated numerically from a realistic geometry. A Monte Carlo algorithm which self-adapts to the size of the system allows us to simulate the single-electron transport properties within a semiclassical framework. We present several effects that are related to the geometry and the one-electron level spacing like e.g. a negative differential conductance (NDC) effect. Consequently these effects are designable by the choice of the size and arrangement of the nanoparticles., Comment: 13 pages, 12 figures

We investigate the momentum distribution function near the Mott-Hubbard transition in the one-dimensional t1-t2 Hubbard model (the zig-zag Hubbard chain), with the density-matrix renormalization-group technique. We show that for strong interactions the Mott-Hubbard transition occurs between the metallic-phase and an insulating dimerized phase with incommensurate spin excitations, suggesting a decoupling of magnetic and charge excitations not present in weak coupling. We illustrate the signatures for the Mott-Hubbard transition and the commensurate-incommensurate transition in the insulating spin-gapped state in their respective ground-state momentum distribution functions.

We investigate an antiferromagnetic s=1/2 quantum spin system with anisotropic spin exchange on a fractal lattice, the Sierpinski gasket. We introduce a novel approximative numerical method, the configuration selective diagonalization (CSD) and apply this method to the Sierpinski gasket with N=42. Using this and other methods we calculate ground state energies, spin gap, spin-spin correlations and specific heat data and conclude that the s=1/2 quantum antiferromagnet on the Sierpinski gasket shows a disordered magnetic ground state with a very short correlation length of about 1 and an, albeit very small, spin gap. This conclusion holds for Heisenberg as well a for XY exchange., Comment: LaTeX: 16 pages, 9 figures, 1 table

Using the next-nearest neighbor (zig-zag) Hubbard chain as an one dimemensional model, we investigate the influence of interactions on the position of the Fermi wavevectors with the density-matrix renormalization-group technique (DMRG). For suitable choices of the hopping parameters we observe that electron-electron correlations induce very different renormalizations for the two different Fermi wavevectors, which ultimately lead to a complete destruction of one section of the Fermi sea in a quantum critical point.

We present a simple model of electrical transport through a metal-molecule-metal nanojunction that includes charging effects as well as aspects of the electronic structure of the molecule. The interplay of a large charging energy and an asymmetry of the metal-molecule coupling can lead to various effects in non-linear electrical transport. In particular, strong negative differential conductance is observed under certain conditions., Comment: 7 pages, 5 figures, accepted by Europhys. Lett

Recent experimental evidence suggest the existence of three distinct V-valence states (V^{+4}, V^{+4.5} and V^{+5}) in the low-temperature phase of NaV_2O_5 in apparent discrepancy with the observed spin-gap. We investigate a novel spin cluster model, consisting of weakly coupled, frustrated four-spin clusters aligned along the crystallographic b-axis that was recently proposed to reconcile these experimental observations. We have studied the phase diagram and the magnon dispersion relation of this model using DMRG, exact diagonalization and a novel cluster-operator theory. We find a spin-gap for all parameter values and two distinct phases, a cluster phase and a Haldane phase. We evaluate the size of the gap and the magnon dispersion and find no parameter regime which would reproduce the experimental results. We conclude that this model is inappropriate for the low-temperature regime of NaV_2O_5.