• dft
• graphene corrugation

Due to its high chemical energy content and to the fact that its oxidation byproduct
is water, hydrogen is currently considered one of the most promising green fuels.
Despite its good perspective as an energy vector, it presents some problems in
concerning the storage and transport. The most obvious storage system, namely the
compression within tanks, presents some safety risk, related to the fact that it is
highly explosive under pressure. In addition this kind of storage requires heavy
tanks that decrease the overall gravimetric and volumetric capacity of the system.
One possibility to overcome safety and capacity problems is to considered new
specifically designed materials. One currently considered is graphene, the two
dimensional honeycomb lattice of sp2 hybridized carbon. Specifically, hydrogen
can be either physisorbed in molecular form onto graphene sheets (or within them),
or chemisorbed, forming covalent C-H bonds and generating the graphene
hydrogenated counterpart, called graphane. While physisorption is a very weak
linkage based on Van der Waals interactions, which leads to reasonable storage
capacity only at low temperatures and high pressures, chemisorption creates a very
stable linkage and relatively high storage capacities (up to 7-8% of hydrogen mass
stored with respect to the total mass of the mean), and gives the possibility of safe
transportation. However this is obtained at the expenses of high barriers for the
adsorption/desorption, leading to slow kinetic for the loading/discharge, or
imposing the use of high temperatures.These problems are common to all the others proposed materials that use
chemisorption as storage process (e.g. N or B hydrides, hydrocarbons). However,
at variance with those, and because of its unique mechanical and structural
properties, graphene offers alternative ways to overcome these problems. Graphene
is extremely resistant and flexible, thus it can be rippled by compression or by
mean of interactions with specific substrates over which it is deposited. Moreover,
it was shown that convexities have enhanced reactivity towards hydrogen. This
considered, corrugation and its control was proposed as a possible mean to control
the adsorption/desorption of hydrogen. It is to be observed that this possibility is
related not only to the exceptional properties of graphene, but also to the specific
fact that it is a 2D crystal, and therefore its flexibility can be exploited.
This thesis work focus on the structural, electronic and dynamical properties of
rippled graphene, with the specific focus of its interaction with hydrogen. The
methods used for the study are numerical calculations and simulations based on the
Density Functional Theory.
First, different rippled structures are created by lateral compression of a supercell
including ~54 C atoms. The symmetry of the graphene supercell is chosen so to
match the structure of the natural rippling observed in graphene grown by Si
evaporation from SiC. This allows a direct comparison with available experimental
determination of the structure and electronic structure by STM. In fact, the effect
of rippling on the band structure and electronic properties of graphene is evaluated
by calculating them on relaxed structures at different levels of rippling, and
compared with experiment.
Subsequently, rippled structures were hydrogenated by adding hydrogen on
specific sites. At each stage, hydrogen is added on those sites whose binding
energy is larger, so to mimic the natural hydrogenation process. Structures with
different hydrogen decoration were relaxed and their electronic structures evaluated. Even in this case, comparison with corresponding experimental
determinations are possible.
Finally, a structure with maximum hydrogen loading was created, and its stability
is studied by means of an extensive molecular dynamics simulation. The starting
structure was relaxed and then subjected to simulated annealing. During this phase
hydrogen hopping is observed, until a stabler configuration/decoration is reached.
The electronic properties of the maximally loaded structure are then evaluated.
All the calculations and simulations are performed with the QuantumEspresso
code, implemented on the FERMI Blue Gene Q parallel system at CINECA.
This study gives indication on the hydrogenation process of a model system
mimicking a real one, i.e. the naturally rippled graphene on SiC, and hints on
possible strategies to efficiently the loading/discharge kinetics.