Home > Seminars > The high-throughput highway to accelerated materials design (fasten your seat belts: highway without speed limits)

The high-throughput highway to accelerated materials design (fasten your seat belts: highway without speed limits) By Prof. Stefano Curtarolo (Duke University, USA)

  • Class schedule:  Tuesday, Mar. 25th, 2014 from 04:00p m to 05:00 pm 
  • Location: Building 9, level 2, Lecture Hall 2, Room 2325
  • Refreshments:  Available @ 03:45 pm


High-throughput computational materials design is a rapidly emerging area of materials science [1]. By combining advanced thermodynamic and electronic-structure methods with intelligent data mining and database construction, and exploiting the power of current supercomputer architectures, scientists generate, manage and analyze enormous data repositories for the discovery of novel materials [2].
The key for discovering new materials is the availability of descriptors. These are physically sound empirical quantities, not necessarily observables, connecting the calculated microscopic parameters to macroscopic properties of the materials. In other words, the descriptor is the language with which the researcher speaks to the database, the heart of any effective HT implementation.
In this talk (i) we provide a current snapshot of this rapidly evolving field and illustrate our recent efforts (e.g, thermodynamics [3], catalytic materials [4], thermoelectricity [5,6], topological insulators [7], nuclear detection [8], and so on), (ii) we highlight the challenges and opportunities that lie ahead, and (iii) we illustrate the needs and goals of the communities involved: open and free repositories with widely accepted standards [9] and interfaces [10].
[1] Nature Mater, 12, 191 (2013).  [2] e.g., http://aflowlib.org [3] Comp. Mat. Sci. 58, 227 (2012). [4] Phys. Rev. X 3, 041035 (2013). [5] Phys. Rev. X 1, 021012 (2011). [6] Phys. Rev. X 4, 011019 (2014) [7] Nature Materials 11(7), 614 (2012). [8] ACS Comb. Sci. 13, 382 (2011). [9] “Fuelling discovery by sharing”, Nature Mater. 12, 173 (2013). [10] Curtarolo et al. “A RESTful API for exchanging Materials Data in the AFLOWLIB.org consortium”, submitted to NPG - Scientific Data (2014).


After studying Electrical Engineering and Physics in Padova, Italy, Stefano Curtarolo received his PhD in Materials Science from MIT in 2003. Since then, he was faculty of Materials Science and Physics at Duke University. During his time at Duke, SC received the ONR-Young-Investigator, the CALPHAD best paper award, the NSF-Career, the Presidential PECASE Awards, the International Union of Pure and Applied Physics - Young Scientist Prize in Computational Physics, the 2013 Duke Stansell Research Award, the 2013 MURI Award for strategies in element replacement and the 2014 APS Fellowship for the developing of scientific tools for Materials Genomics. SC was promoted to Associate in Oct.2008, to Full Professor in Feb. 2012 and to Center Director in 2012. Currently he has more than 95 refereed publications and more than 140 invited departmental seminars and talks in national and international conferences, including keynote and plenary contributions.
At Duke University, the SC's group started and maintains the ``on-line ab-initio aflowlib.org'' consortium containing free energy information and electronic characterization of more than 600,000 entries/compounds equipped with online tools for accelerated materials development. He is the leading author of the Nature Materials review on High-throughput computational materials science [doi:10.1038/nmat3568]. He likes (and drives) old air-cooled Volkswagens.