Welcome to the
Organic Materials Database
The organic materials database is an open access electronic structure database for 3-dimensional organic crystals, developed and hosted at the Nordic Institute for Theoretical Physics – Nordita.
It provides tools for search queries based on data-mining and machine learning techniques. The universal features provided on our web interface facilitate the design of novel functional organic materials with a wide-range of applications.
Shifting computational boundaries for complex organic materials
In this article we discuss the future potential of organic materials and the databases that are at the heart of it. Artificial intelligence is enabling simulation of highly complex but fascinating organic crystals that were previously computationally intractable.
Within this domain of complex organic materials, the Organic Materials Database is the largest database of its kind in the world, containing the electronic, magnetic and crystallographic properties of the materials.
R. Matthias Geilhufe, Bart Olsthoorn, Alexander V. Balatsky, Nature Physics (2021).
Identification of strongly interacting organic semimetals
Dirac and Weyl point- and line-node semimetals are characterized by a zero band gap with simultaneously vanishing density of states. Although several inorganic nodal semimetals are known, the space of organics remains fairly unexplored. We combined machine learning, density functional theory, and effective models to identify three organic semimetals.
The three materials highlighted in the publication can also be visited by clicking the buttons below:
R. Matthias Geilhufe and Bart Olsthoorn, Phys. Rev. B 102, 205134 (2020).
New dataset for excitations of magnetic metal organic materials
We have released a new dataset containing information about local magnetic moments, magnetic exchange coupling, expected magnetic ground state, as well as spin-wave excitation spectra and the dynamical structure factor S(Q,ω) for hundreds of previously synthesized organic molecular crystals and metal organic frameworks. The new data is embedded into the existing framework of the OMDB and can be explored using our interactive statistics and non-trivial search tools. The dataset represents the first high-throughput database of magnetic properties of organometallic materials, opening the path towards establishing general concepts of materials statistics and state-of-the-art machine learning applications.
Browse magnetic information
Band gap prediction for large organic crystal structures with machine learning
Large datasets of ab initio calculations enabled the pioneering studies of machine learning applied to quantum-chemical systems. For example, the popular QM9 dataset of small organic molecules and their properties can now be modelled by machine learning with chemical accuracy. However, a more challenging dataset of large organic crystal structures is not yet available. Therefore, we present a new consistent dataset of 12,500 crystal structures and their corresponding DFT band gap, freely available at https://omdb.mathub.io/dataset/.
In this paper, we also evaluate the performance of two recent machine learning models on this new dataset: Kernel Ridge Regression with the Smooth Overlap of Atomic Positions (SOAP) and the deep learning model SchNet.
Materials Informatics for Dark Matter Detection
Dark Matter particles are commonly assumed to be weakly interacting massive particles (WIMPs) with a mass in the GeV to TeV range. However, recent interest has shifted towards lighter WIMPs, which are more difficult to probe experimentally. WIMP detection would require the use of small gap materials in sensors. We propose the use of informatics tools to rapidly assay materials band structures to search for small gap semiconductors and semimetals. Using recent estimates of the WIMP mass, we identify the relevant target space towards small band gap materials (100-10 meV).
We use a large dataset search to identify candidate materials, rather than focusing on a few preselected compounds. Dirac Materials, a class of small- or zero-gap materials, emerge as natural candidates for sensors for Dark Matter detection. As a specific example of the proposed search strategy, we use the Organic Materials Database to identify few organic candidates for sensors: the narrow band gap semiconductors BNQ-TTF and DEBTTT with gaps of 40 and 38 meV, and the Dirac-line semimetal (BEDT-TTF)⋅B which exhibits a tiny gap of ≈ 50 meV when spin-orbit coupling is included. We outline a novel approach to search for sensor materials where a rapid assay of materials using informatics tools yields more candidates and thus provides a useful tool to identify sensor materials for Dark Matter detection.
Semimetallicity and topological nodes induced by chemical strain
Organic charge transfer salts, especially quasi two-dimensional (BEDT-TTF) based crystals, have generated great interest over the past few decades. Depending on the structural phase, the acceptor molecules, and other factors, these can be associated with a rich variety of physical phenomena such as charge ordering, metal-insulator transitions, unconventional superconductivity, antiferromagnetism and spin liquids.
Especially α-(BEDT-TTF)2I3 has been shown to exhibit a semiconductor-semimetal phase transition under applying high pressure, where the low-energy excitations in the semimetallic phase behave as massless Dirac fermions leading to unique transport properties. In a series of papers it has been shown that under a hypothetical substitution of iodine with the halogens bromine, chlorine and fluorine the unit cell size tremendously decreases introducing topolgical nodes and semimetallicity due to chemical strain. Currently the OMDB stores hundreds of electronic structures of charge transfer salts and therefore represents a powerful tool for identifying such novel phases of matter.
R. M. Geilhufe, B. Commeau, and G. W. Fernando, physica status solidi (RRL)–Rapid Research Letters (2018)
B. Commeau, R. M. Geilhufe, G. W. Fernando, A. V. Balatsky, Physical Review B 96.12 (2017): 125135.
MATDAT18
Our proposed project is about predicting novel stable crystal structures using the recent advances in generative modeling from the machine learning field, bypassing computationally demanding first-principles optimization calculations.
Particular focus is made on prediction of organic crystals with metallic and semiconducting properties.
Finding such materials in the organic space, which is mainly populated by wide-gap insulators, is a highly non-trivial task.
At the same time, these functional materials offer promising technological applications in electronics and offer a basis for environmentally friendly technologies. The electronic structure data from the OMDB will be available for training.
Topological organic materials
The study of topological materials has tremendously stimulated condensed matter research. Based on a data-mining exercise within the Organic Materials Database OMDB it was possible to identify the nature of topological band crossings based on the underlying crystalline symmetry.
In a collaboration of researchers from Nordita, the Royal Institute of Technology KTH and Uppsala University pattern matching was applied to identify examples of topological crossings in materials with space group 14 and 19 and to explain their stability based on topological invariants.
It could be shown that such topological invariants can be deduced from the irreducible representations associated to the electronic band structure.
R. M. Geilhufe, A. Bouhon, S. S. Borysov, A. V. Balatsky, Physical Review B 95:4 (2017): 041103.
R. M. Geilhufe, S. S. Borysov, A. Bouhon, A. V. Balatsky, Scientific Reports 7:1 (2017): 7298.
Towards high-Tc organic superconductors using similarity search tools
Even though many organic superconductors are known, the highest superconducting transition temperatures found so far are in the range of 10 Kelvin.
A novel density of states similarity search tool developed for the OMDB will help future research to identify potential materials with similar electronic features to a well-known prototype.
To stimulate the experimental verification of a recently reported organic high-temperature superconductor based on p-terphenyl, a similarity search for p-terphenyl has resulted in 15 novel candidates which potentially exhibit similar properties.
Identifying topological quasiparticles with pattern matching
A novel pattern matching tool for electronic band structures developed for the OMDB will stimulate future research identifying specific crossings in organic materials.
Via a graphical interface users are able to draw patterns of band crossings which are used as a search query to find similar features within the electronic band structures stored in the OMDB.
Excitations within materials hosting topological crossings can mimic condensed matter realizations of quasiparticles like Dirac or Weyl fermions (linear crossings), nearly free electrons (parabolic band touching), and more exotic particles such as hourglass fermions.
The search tool facilitates finding real material realizations for any kind of point- and line-node material with various bands involved.
Be part of the OMDB community fostering data-driven organic materials design.
Register for free today
Why?
The OMDB provides data-driven online services specifically for organic materials, a class of high technological relevance.The elasticity of these compounds connected to the endless configuration space and tuning opportunities open the path towards various basic research and industrial applications such as flexible electronics, carbon-sequestration materials and photovoltaics.
How?
We bring together scientists, publishers, and industrial partners of various background interacting via the OMDB.The OMDB community provides additional resources and services like computational support, experimental verification and synthesis.
The social component of the OMDB enables the connection and stimulates the formation of virtual research groups fostering cross-institutional interactions world wide.
