Materials shaped human civilization, impacting on the evolution of technologies and the benefits of the lifestyle. Typically, every new material was discovered by chance – the so called serendipity! – or through painstaking search. Just think of Edison, who tried more than 1600 materials for over 14 months to find the best one for the light bulb. At the current stage of scientific and technological advances, there is an ever-increasing request for more efficient, more resistant, smarter and greener materials.
In fact, advanced materials play a crucial role in many societal challenges: from energy harvesting and storage to healthcare and transportation. At the same time, experiments can be difficult, time-consuming and expensive, and not at the pace of industry demands.
Research on new materials, traditionally based on experimentation, is increasingly carried out through simulations that, by exploiting the knowledge of quantum physics, nanotechnologies and big data algorithms, allow one to design materials with specific properties or optimal performance, and at the same time to save enormously time and costs. Through simulations, scientists can either design novel and not existing materials from scratch or study existing ones from a different perspective. They can even envisage how to manipulate them to meet the technological request.
Simulating new materials requires scientific codes (software) to study their physical, chemical, and electronic properties, and also powerful supercomputers to run such complex and energy-demanding codes.
The next further leap to exascale computing (i.e. machines capable of 10 18, one billion of billions of floating point operations per second) will transform the study and design of new materials and their functions in a disruptive way. It is expected that “materials design” through simulations will have an enormous impact in the industry, of the same magnitude of, or even larger than the one induced by fluid dynamics simulations in the automotive sector or by molecular dynamics simulations in drug design a decade ago.
To respond to these significant challenges, Cnr Nano is at the forefront in taking up and developing High-Performance Computing (HPC) applications to the field of materials research, while fostering open science through international collaborations to enable the advancement of materials discovery and design.
The next further leap to exascale computing will transform the study and design of new materials and their functions in a disruptive way.
The High-Performance Computing activity at Cnr Nano includes the first-principles prediction of structural, electronic, magnetic, and optical properties of materials, from crystals to molecules, low dimensional systems and devices. Theoretical developments and their implementation and optimization in HPC software are key aspects of the research activity, together with the development of automated workflows for computational high-throughput screening and production of curated data.
Cnr Nano is also committed to enhancing the data and HPC ecosystem to take advantage of the new computational power while widening the access to codes and engaging academic and industrial communities, by providing workflows and turn-key solutions and also training for a broader and different pool of users and developers.
We are a hub of expertise in High-Performance Computing applications at the national and European level, and our cutting-edge capabilities in materials design and discovery enable us to impact science on a global scale.
Below is a short overview of our participation in international collaborations that highlights our capabilities as well as our open-data, open-access and open-source culture.
Cnr Nano coordinates MaX – materials design at the exascale – a European Centre of Excellence for HPC applications dedicated to materials design. MaX is funded by the EU within the EuroHPC strategy, a plan to strengthen European leadership in the applications of supercomputing, and is a coordinated effort of an outstanding team of European leaders in the materials domain. MaX develops open-source codes, workflows, and data platforms and makes them available for a large and growing base of researchers in the materials domain.
Cnr Nano also coordinates the INTERSECT project (Interoperable Material to Device simulation box for disruptive electronics) dedicated to the development of an interoperable software platform (IM2D) for the modelling of synaptic electronics and neuromorphic computing. IM2D is conceived as an advanced solution for industry-driven research: a platform that will enable the material-to-device and the device-to-material simulation workflows for the characterization and design of materials at the device level. All this will enhance the fast exploration of materials’ properties (e.g. switching memory effects for analog memristive computing), the ability to link materials properties to performance in unexplored device architectures, and to assess their business potential.
Cnr Nano researchers works together with a team of 34 European partners on the design of materials for the development of cheaper, better and more sustainable batteries. The ambitious BIG-MAP project (Battery Interface Genome – Materials Acceleration Platform), set within the Battery 2030+ EU research initiative, aims at speeding-up the battery discovery and development processes by changing the way batteries are designed and optimized. The BIG-MAP ‘engine’ is a new materials acceleration platform that combines physical models at different scales with machine learning, artificial intelligence and large-scale computer simulations, and integrates them with autonomous synthesis robotics and large-scale, high-throughput experimental characterization. Cnr Nano team will focus on atomistic simulations of materials and interfaces and automated predictions of spectroscopies, exploiting the potential of extreme (pre)exascale HPC in battery materials research.
Cnr Nano participates in LESGO, a European Union’s Horizon 2020 FET-PROACT project, aimed at engineering an integrated system for extracting hydrogen from water by photochemical means, storing and using it in fuel cells to produce electricity within a clean energy production-and-use cycle. The role of Cnr Nano researchers is specifically to optimize the structure and composition of reduced graphene oxide and to simulate the different phases of the processes involved in the cycle, by means of a multi-scale modeling approach, high throughput computing and machine learning based analyses.