Final Report Summary - QUASINANO (Quantum-mechanical simulations for the nanoscale)
A leading scientific software company and a renowned university research group have teamed up to develop new, quantum-theory based methods and concepts to allow quantum-mechanical computer simulations of systems in the 100,000 atoms range, and apply them to study challenging applications. As a result, a new chapter of atomistic computer simulation is being opened, as quantum-mechanical simulations become feasible to study systems and processes in nanotechnology, biochemistry, supramolecular and inorganic chemistry, and in particular interfaces between these domains.
Scientific Computing & Modelling (SCM) N.V. is an SME founded in 1998 as a spin-off of the Vrije University in Amsterdam, and has developed since then into one of the market leaders for quantum chemical software. The Computational Materials Science group of Prof. Heine at Jacobs University in Bremen has a long experience in nanomolecular sciences, including organic and inorganic nanomaterials, spectroscopy, dynamic effects etc. The project is based on recent developments of the Heine group, aiming at developing, implementing, and applying new methods and algorithms that enable quantum-mechanical simulations of nanoscale systems of up to 100,000 atoms.
This has been achieved by extending the DFTB (Density Functional-based Tight Binding) method - an approximate Density Functional Theory (DFT) approach which is applicable throughout the periodic table - to overcome certain historical limitations. In particular, the extended method covers a larger part of the periodic table and reaches a higher accuracy by improving the quality of the initial electron density, which is constructed by superposing electron densities of fragments, and by other improvements. Earlier alternative methods available for the target systems include empirical force fields, which are fast but limited in scope and transferability of their predictive power. What is more, typical force field methods do not allow bond-breaking and give no information on electronic structure and spectroscopic properties.
A universal set of DFTB parameters for the calculation of electronic band structures has been successfully obtained and already used for production work. Thanks to QUASINANO those universal parameters already make the method available for most of the periodic table of elements, and there is ongoing work to include the rest. A following step has been to calculate spectroscopic properties with the new method and also include dynamics (computer simulations of atoms moving in time). A newly-introduced description of the repulsion energy term has been proven to work, and the basic algorithms for the optimization of parameters and the calculation of the repulsion energy terms have been implemented. Thus, the project has brought us much closer to obtaining a full set of parameters for the repulsion energy part, a critical success factor for the broad applicability of the method. Furthermore, we have provided an alternative DFT-DFTB hybrid approach via the Frozen Density Embedding (FDE) method currently implemented in ADF. With the availability of parameters throughout the periodic table and the possibility to compute properties directly from DFTB (UV-Vis and excitons, topological analysis of the electron density, electron transport, vibrational spectroscopy and phonons) the DFTB implementation is offering more than originally expected at the start of the project. These developments have already led to a new Marie Curie project (PROPAGATE), which is taking on some of the remaining challenges. QUASINANO is therefore finding continuity in the new project.
The method extensions have been brought through secondments to SCM where they have been implemented into its software product ADF and brought into the market. All software developments have been done in a professional manner to produce commercial-grade implementations, including Graphical User Interfaces, documentation, tutorials, etc. SCM is maintaining and commercializing the resulting software, and will continue to do so. Special emphasis has been placed on efficient parallel implementations of the new method, with thorough quality control. The software development has been done in a modular fashion that enables reuse of methods. The new developments also build upon the functionality already available in the ADF (molecular DFT) and BAND (periodic DFT) codes from SCM, and their libraries.
The new methods and software have been applied to challenging applications in the field of environmental nanoscience. The first group involve the structure and spectroscopic properties of large finite systems, i.e. molecular containers and polyoxometalates. The second group are extended systems, and involve nanoporous materials (metal organic frameworks), which are investigated for their ability to conduct protons, an important property for fuel cell membrane materials. Finally, the extended method has been used to study inorganic nanostructures, such as inorganic nanotubes and nanowires, which are interesting for nanoelectronic devices and materials with high adsorption rates of environmentally harmful mining disposals.
Thanks to this initiative the ADF Modelling Suite is becoming significantly more competitive and SCM is strengthening its position in the market of quantum-theory based computer software. Knowledge has been exchanged between the academic and private-sector partners through extensive secondments of newly recruited Experienced Researchers, Early Stage Researchers and also staff members . Two international workshops took place in Leiden and Bremen to disseminate results from the project, which have also been published in almost 50 papers in peer-reviewed journals.
More information can be found at the project website, as well as contacting SCM and the Heine group:
• www.scm.com/quasinano/
• http://www.jacobs-university.de/ses/theine
Scientific Computing & Modelling (SCM) N.V. is an SME founded in 1998 as a spin-off of the Vrije University in Amsterdam, and has developed since then into one of the market leaders for quantum chemical software. The Computational Materials Science group of Prof. Heine at Jacobs University in Bremen has a long experience in nanomolecular sciences, including organic and inorganic nanomaterials, spectroscopy, dynamic effects etc. The project is based on recent developments of the Heine group, aiming at developing, implementing, and applying new methods and algorithms that enable quantum-mechanical simulations of nanoscale systems of up to 100,000 atoms.
This has been achieved by extending the DFTB (Density Functional-based Tight Binding) method - an approximate Density Functional Theory (DFT) approach which is applicable throughout the periodic table - to overcome certain historical limitations. In particular, the extended method covers a larger part of the periodic table and reaches a higher accuracy by improving the quality of the initial electron density, which is constructed by superposing electron densities of fragments, and by other improvements. Earlier alternative methods available for the target systems include empirical force fields, which are fast but limited in scope and transferability of their predictive power. What is more, typical force field methods do not allow bond-breaking and give no information on electronic structure and spectroscopic properties.
A universal set of DFTB parameters for the calculation of electronic band structures has been successfully obtained and already used for production work. Thanks to QUASINANO those universal parameters already make the method available for most of the periodic table of elements, and there is ongoing work to include the rest. A following step has been to calculate spectroscopic properties with the new method and also include dynamics (computer simulations of atoms moving in time). A newly-introduced description of the repulsion energy term has been proven to work, and the basic algorithms for the optimization of parameters and the calculation of the repulsion energy terms have been implemented. Thus, the project has brought us much closer to obtaining a full set of parameters for the repulsion energy part, a critical success factor for the broad applicability of the method. Furthermore, we have provided an alternative DFT-DFTB hybrid approach via the Frozen Density Embedding (FDE) method currently implemented in ADF. With the availability of parameters throughout the periodic table and the possibility to compute properties directly from DFTB (UV-Vis and excitons, topological analysis of the electron density, electron transport, vibrational spectroscopy and phonons) the DFTB implementation is offering more than originally expected at the start of the project. These developments have already led to a new Marie Curie project (PROPAGATE), which is taking on some of the remaining challenges. QUASINANO is therefore finding continuity in the new project.
The method extensions have been brought through secondments to SCM where they have been implemented into its software product ADF and brought into the market. All software developments have been done in a professional manner to produce commercial-grade implementations, including Graphical User Interfaces, documentation, tutorials, etc. SCM is maintaining and commercializing the resulting software, and will continue to do so. Special emphasis has been placed on efficient parallel implementations of the new method, with thorough quality control. The software development has been done in a modular fashion that enables reuse of methods. The new developments also build upon the functionality already available in the ADF (molecular DFT) and BAND (periodic DFT) codes from SCM, and their libraries.
The new methods and software have been applied to challenging applications in the field of environmental nanoscience. The first group involve the structure and spectroscopic properties of large finite systems, i.e. molecular containers and polyoxometalates. The second group are extended systems, and involve nanoporous materials (metal organic frameworks), which are investigated for their ability to conduct protons, an important property for fuel cell membrane materials. Finally, the extended method has been used to study inorganic nanostructures, such as inorganic nanotubes and nanowires, which are interesting for nanoelectronic devices and materials with high adsorption rates of environmentally harmful mining disposals.
Thanks to this initiative the ADF Modelling Suite is becoming significantly more competitive and SCM is strengthening its position in the market of quantum-theory based computer software. Knowledge has been exchanged between the academic and private-sector partners through extensive secondments of newly recruited Experienced Researchers, Early Stage Researchers and also staff members . Two international workshops took place in Leiden and Bremen to disseminate results from the project, which have also been published in almost 50 papers in peer-reviewed journals.
More information can be found at the project website, as well as contacting SCM and the Heine group:
• www.scm.com/quasinano/
• http://www.jacobs-university.de/ses/theine