First principles investigation of hydrogen physical adsorption on graphynes' layers

Abstract

Graphynes are 2D porous structures deriving from graphene featuring triangular and regularly distributed subnanometer pores, which may be exploited to host small gaseous species. First principles adsorption energies of molecular hydrogen (H₂) on graphene, graphdiyne and graphtriyne molecular prototypes are obtained at the MP2C level of theory. First, a single layer is investigated and it is found that graphynes are more suited than graphene for H₂ physical adsorption since they provide larger binding energies at equilibrium distances much closer to the 2D plane. In particular, for graphtriyne a flat minimum located right in the geometric center of the pore is identified. A novel graphite composed of graphtriyne stacked sheets is then proposed and an estimation of its 3D arrangement is obtained at the DFT level of theory. In contrast to pristine graphite this new carbon material allow both H₂ intercalation and out-of-plane diffusion by exploiting the larger volume provided by its nanopores. Related H₂ binding energies for intercalation and in-pore adsorption are around 0.1 eV and they could lead to high storage capacities. The proposed carbon-based layered material may represent a safer and potentially cheaper alternative for hydrogen on-board storage than conventional solutions based on cryogenic liquefaction and/or high compression.

Introduction

Hydrogen is a clean-burning energy carrier which is supposed to provide the most promising alternative to fossil fuels to be employed in propulsion systems for automotive applications. However, some drawbacks such as the safe and convenient hydrogen storage [1] and release in reversible cycles still represent major challenges to the scientific community. The use of advanced solid state adsorbents has recently received wide attention since they could provide solutions to achieve the 2017 targets for on-board vehicle hydrogen storage established by the U.S. Department of Energy (DOE): they are actually 5.5 wt% and 0.040 kg L⁻¹ for gravimetric and volumetric capacities, respectively [2], and they have not been simultaneously reached yet in practice at moderate temperature and pressure.

As a matter of fact, the physisorption of molecular hydrogen (H₂) in porous materials represents a safer, simpler and potentially cheaper option for gas storage than conventional solutions, based on liquefaction at low temperature (and/or high compression at room temperature). Materials traditionally used in these applications include activated carbons and zeolites and in recent years novel and promising adsorbents, namely metal-organic frameworks (MOF) and covalent organic frameworks (COF) [3], are also available. MOF and COF possess advantages with respect to traditional porous materials due to their crystalline structure: in fact, their regular and ordered porosity, together with tunable size and shape of the openings, has led to very high performances in gas uptake capacities [4]. However, they present some shortcomings mainly related to their thermal and chemical stability. Moreover, the presence of heavy metals in MOFs, which in general enhances the binding strength of a given adsorbate, could also induce its dissociation which would alter the desirable gas storage and delivery reversible cycle. Therefore, investigations on alternative porous materials are highly advisable and, specifically, we here focus on carbon-based layers, which in general provide a large specific surface area coupled with a lower weight.

In the last years, by following “bottom-up” assembly processes two-dimensional (2D) materials similar to ubiquitous graphene but with regular and uniformly distributed subnanometer pores have been synthesized in large area films [5], [6], and among them graphdiyne is of particularly interest for our purposes. Graphdiyne is actually a member of γ-graphynes which are new 2D carbon allotropes formed by sp–sp² hybridized carbon atoms. They can be thought as deriving from graphene by replacing one-third of its C–C bonds with mono(poly)-acetylenic units. The number n of acetylenic linkages which connect the benzene rings defines the different graphyne-n species and the first three members of the family are known as graphyne, graphdiyne and graphtriyne, respectively [7], and they feature nanopores of increasing size. The successful synthesis of graphynes has led to important theoretical studies devoted to their application as effective single-layer membranes for gas separation and water filtration technologies [8], [9], [10], [11].

In this work we want to assess the capability of graphynes as efficient materials for the reversible storage of molecular hydrogen and a fundamental point is to address the possibility of gas adsorption by exploiting its intercalation. As it is well known intercalation of gases between pristine graphene layers it is not feasible since large values of the interlayer distance would be needed for an effective storing [12], [13] (roughly the double with respect to graphite equilibrium distance). This requirement is difficult to fulfill in practice since it depends on an “a priori” forced separation of the graphene layers. On the contrary novel graphites composed of multi-layer graphynes could allow intercalation without significantly altering the interlayer separation since a larger available volume is in principle provided by the naturally occurring nano-pores.

The interaction of nonpolar H₂ molecules with carbon-based substrates is mainly the London dispersion and expected values for its binding energy with a single (multi)-layer are around 50 meV [14], [15], [16]. The H₂–graphdiyne single-layer interaction was previously theoretically investigated [17], [8] by means of dispersion-corrected density functional theory (DFT) approaches but the focus was mostly on the penetration barrier in order to propose graphdiyne as an optimal platform for hydrogen purification. The storage capability of Ca and Li decorated graphynes [18], [19] was recently theoretically studied by means of DFT calculations and it was found that the metal decoration can lead to more favorable H₂ binding energies (∼200–300 meV) with respect to the non-doped materials. The three-dimensional (3D) diffusion of H₂ in bulk graphdiyne was also recently investigated [20] but the obtained dispersion-corrected DFT interaction energies were unreasonably too high with binding energies larger than 400 meV. More accurate estimations for the noncovalent H₂–graphynes interaction are therefore desirable and their calculations would require reliable computational approaches. Our choice is to use the “coupled” supermolecular second-order Møller-Plesset perturbation theory (MP2C) which has been reported to provide reliable estimations for weakly bound systems such as rare gas–fullerene [21] and -coronene [22] as well as water (rare gas)–graphynes’ pores [10], [11].

The work reported in this paper is based on the following scheme. First, we carry out accurate estimations of the interaction energy of H₂ adsorbed on graphene, graphdiyne and graphtriyne molecular prototypes. Then, after assessing the most appropriate graphyne for H₂ physisorption, we obtain the equilibrium 3D structure of the related graphyne multi-layer. Finally, the H₂ interaction with the proposed novel graphite composed of graphyne layers is also investigated.