The sp, sp2, and sp3 carbon hybridizations allow an almost infinite number of different structures with tunable mechanical and electronic properties. These structures can exhibit different topologies with different electronic dimensions (0-fullerenes,1-nanotubes, 2-graphene, 3-diamond). These topologies can be exploited to create a large class of different materials, such as bucky papers [1], carbon nanotube-based artificial muscles [2,3], foams [4], auxetic crystals [5], etc. These materials present extremely complex morphologies, which results in a difficult challenge to realistically model their mechanical and structural properties. In this talk, we will present and discuss multi-scale (from fully atomistic to macroscale) approaches to model these materials, including the use of artificial intelligence methods (such as the bioinspired ANT algorithms). Of particular interest are the new molecular dynamics simulation techniques based on reactive potentials that allow handling multi-million atom systems. These techniques can be also used for non-carbon materials, such as chalcogenides [6] and other non-van der Waals solids [7].

[1] L. J. Hall, V. R. Coluci, D. S. Galvao, M. E. Kozlov, M. Zhang, S.
O.Dantas, and R. H. Baughman, Science v320, 5875 (2008).
[2] M. D. Lima et al., Science v338, 6109 (2012).
[3] Z. F. Liu et al., Science v349, 400 (2015),
[4] R. Wang et al., Science v366, 216 (2019).
[5] R. H. Baughman and D. S. Galvao, Nature v375, 735 (1993).
[6] S. Lei et al., Nature Nanotech. v11, 465 (2016)
[7] A. P. Balan et al., Nature Nanotech. v13, 602 (2018).