Disorder-induced metallicity in amorphous graphene

Abstract

We predict a transition to metallicity when a sufficient amount of disorder is induced in graphene. Calculations were performed by means of a first principles stochastic quench method. The resulting amorphous graphene can be seen as nanopatches of graphene that are connected by a network of disordered small and large carbon rings. The buckling is minimal and we believe that it is a result of averaging of counteracting random in-plane stress forces. The linear response conductance is obtained by a model theory as function of lattice distortions. Such metallic behaviour is a much desired property for functionalisation of graphene to realize a transparent conductor, e.g. suitable for touch-screen devices.Since the discovery of graphene, and all its unique physical properties [1][2][3][4][5], attention is now turned to functionalization of this material to suit specific applications. One example is the suggestion of graphane, where H atoms are adsorbed on graphene. Graphane, which is an insulator, was predicted from first principles theory [6], and was subsequently realized experimentally [7]. Further theoretical works [8,9] addressed e.g. the value of the band gap, which was found to be of order 5.7 eV. In a similar fashion, fluorine can also be found to adsorb, and in theoretical works a band gap of 7.4 eV is found [10,11], which is larger than the measured value of 3.4 eV [12]. The reason that a large band gap opens up when hydrogen or fluorine is adsorbed on graphene is that sp 2 bonded C atoms become sp 3 bonded. Hence it seems possible to decrease the conductivity by chemical functionalization, and turn the semi-metallic graphene to a semi-conductor or even an insulator. Smaller values of the gap can be reached by a functionalization by organic molecules [13].The ways to increase the conductivity by chemical means has also been discussed, although here significantly less success can be identified. Adsorption by single impurities [14] have been explored, as well as the replacement of C atoms for B or N atoms [15], or even the creation of structural defects in the C matrix [16][17][18]. The electronic structure of non-crystalline graphene, e.g. around grain boundaries have also been under consideration, both from tight-binding analysis [19] and within the framework of first principles theory [20] in combination with calculations of transport properties. In the works mentioned above, the conducting properties are still those of a doped semiconductor, with relatively few charge carriers. A fully metallic behavior was however suggested in an amorphous structure of graphene [21], where Stone-Wales defects were introduced into graphene and geometry optimization was done according to a Keating-like potential. After geometry optimization, the electronic structure was calculated from a tight-binding Hamiltonian, and a non-zero value of the density of states (DOS) at the Fermi level (E F ) was obtained, suggesting that metallic conductivity is possible. The possibility of a transparent, mono-atomic thin material, has a g...

Charge and Spin Transport in Disordered Graphene-Based Materials

“…4.30(c) shows the density of states (DOS) of the two disordered samples, together with the pristine case (dashed line) for comparison. Sample 1, which keeps 52% of hexagonal rings, displays several noticeable features, similar to those found in previous studies [74,76]. First, the DOS at the charge neutrality point is found to be increased by a large amount.…”

“…From the theoretical side, models of the amorphous network have been proposed using stochastic quenching methods [74], and molecular dynamics [76,75,73]. Electronic structure calculations show that the amorphization yields a large increase of the density of states at and in the environment of the charge neutrality point [74,76,75]. Despite the expected reduction of the conduction properties due to strong localization effects, Holmström et al [76] suggest that disorder could enhance metallicity in amorphized samples, in contrast with the experimental evidence.…”

“…However, using a stochastic quenching method, Ref. [76] claimed that "we predict a transition to metallicity when a sufficient amount of disorder is induced in graphene...". In Chapter 4, by using Kubo-Greenwood calculation, we show that this conclusion is misleading and similar results have also been obtained recently in Ref.…”

“…Localization lengths were estimated to be of the range 0.1 to 10 nm in the amorphous samples, depending on the degree of amorphization. From the theoretical side, models of the amorphous network have been proposed using stochastic quenching methods [74], and molecular dynamics [76,75,73]. Electronic structure calculations show that the amorphization yields a large increase of the density of states at and in the environment of the charge neutrality point [74,76,75].…”

Insulating behavior of an amorphous graphene membrane

“…Localization lengths were estimated to be of the range 0.1 to 10 nm in the amorphous samples, depending on the degree of amorphization. From the theoretical side, models of the amorphous network have been proposed using stochastic quenching methods 15 , and molecular dynamics [16][17][18] . Electronic structure calculations show that the amorphization yields a large increase of the density of states in the close vicinity of the charge neutrality point [15][16][17] .…”