The surface science of graphene: Metal interfaces, CVD synthesis, nanoribbons, chemical modifications, and defects
Introduction
Research on graphene has experienced explosive growth in the last few years. Initial excitement for graphene came from its unusual linear dispersion of the -band at the Fermi level, which gives rise to new physical properties. The isolation of single layer free-standing graphene from highly oriented pyrolytic graphite (HOPG) and the measurements of many exotic electronic properties of graphene have earned Kostya S. Novoselov and Andre K. Geim the Physics Nobel Prize in 2010. In addition to the exciting physics of the electrons in this material, graphene also possesses other interesting physical properties, such as being the ‘strongest’ material (by weight) [1] and exhibiting a negative thermal expansion coefficient [2]. Such superlatives and unique properties of graphene motivated much of the fundamental research on this special material. However, this alone does not explain the enormous interest in graphene. Soon after its synthesis, the applied research community became interested in graphene. In particular, many from the carbon-nanotube community changed their focus toward graphene. One practical aspect for the interest in graphene was the prospect of large area graphene wafers and thus the use of lithographical methods for patterning and device fabrication. Such processes are more compatible with existing technologies than, for example, the selection and assembly of carbon-nanotube architectures. Graphene wafers were already available by graphitization of SiC, even before the groundbreaking experiments by Novoselov and Geim, and soon afterward large-area graphene was synthesized by chemical vapor deposition (CVD) growth on metal substrates. Development of transfer procedures from the metal substrate on which graphene is grown on, allowed these wafers to be placed on any arbitrary substrate. The synthesis of graphene on SiC and on metal substrates also brought many surface scientists into the arena of graphene research. The graphene formation on metals had been well known in the surface science community for decades. The increasing interest in graphene meant that these graphene synthesis processes were now re-investigated in more detail and with modern surface science techniques such as low energy electron microscopy (LEEM), scanning tunneling microscopy (STM) and angle resolved photoemission spectroscopy (ARPES) to obtain structural and electronic information. In general the 2D nature of graphene with all its atoms situated at the surface makes this material an obvious object for surface science studies.
Graphene research may be divided in three sub-areas: (i) the characterization of the special physical properties that are originating from the 2D nature of the material and its special electronic band structure, (ii) device applications, and (iii) the materials science of graphene, i.e. making and processing of graphene wafers, doping of graphene by impurities, and interface formation between graphene and dissimilar materials. This review is primarily concerned with the latter. Furthermore, we exclude areas that are already thoroughly reviewed. In particular single and multilayer graphene on SiC single crystal wafers is not included in this review. Information on surface studies of graphene on SiC may be found in other recent reviews [3], [4], [5]. Raman spectroscopy has played an important role in the characterization of graphene, Dresselhaus and co-workers have written excellent reviews on the application of this technique for graphene characterization [6], [7] and therefore we do not include any Raman spectroscopy studies in this article. Also, formation of graphene by reduction of graphene oxide is not included here.
Although graphene has just recently become a ‘hot topic’, it is not a new material and therefore we start this review by describing historical highlights in the materials science of graphene.
Graphene, a single layer of sp2 bonded carbon atoms, is the basic building block of carbon nanotubes, fullerenes and, of course, graphite and HOPG. Therefore, it is somewhat ironic that the most basic ‘building block’ achieves ‘maturity’ last among these carbon materials. However, the concept of graphene is not new. As early as 1947 Wallace [8] showed that a single sheet of sp2 hybridized carbon would have a linear energy dispersion as function of electron-momentum vector () at the -point of the Brillouin zone. The first synthesis of single and multilayer graphene was probably achieved by Boehm in 1962 [9]. Boehm used a method based on the reduction of graphene oxide, which has now been re-discovered as chemical synthesis method of graphene. For many other experimental works on sp2 carbon the synthesis and commercialization of HOPG crystals in the 1960s was essential. HOPG would also lay the foundation for the work by Novoselov and Geim four decades later. But first, in the 1970’s, most interests in carbon materials were on intercalation compounds [10]. Intercalation compounds are single or multilayers of graphene sandwiched in between intercalant layers of guest atoms or molecules.
Before the interest in graphene as a ‘nanomaterial’, the discovery of fullerenes by Kroto et al. [11] in 1985 and then the identification of carbon nanotubes by Iijima [12] in 1991 were the center of attention. The existence of free-standing graphene was dismissed for some time as not being thermodynamically stable; first by Landau [13] and later by Mermin [14]. On the other hand, formation of supported mono- and multilayer graphene on transition metal substrates was observed by surface scientists in vacuum, either by segregation of carbon containing samples or by exposure of hot samples to hydrocarbons. Graphene was first suggested to form on transition metals by hydrocarbon dissociation from low energy electron diffraction (LEED) observations on Pt(100) in 1968 [15], [16]. First scanning tunneling microscopy investigations of graphene on Pt(111) surfaces was performed by the Comsa group [17] in 1991. Blakely and co-workers, studied the segregation behavior of carbon from Ni-crystals by Auger-electron spectroscopy (AES) in 1974 and found the formation of single layer graphene as a thermodynamic stable surface termination while multilayer graphene formed by precipitation of carbon if the samples are cooled to lower temperatures [18]. Single layer graphene on Ni was also obtained by exposure of pure Ni-crystals to hydrocarbons in ultrahigh vacuum (UHV) [19] and electronic decoupling of graphene from the Ni-substrate was investigated by metal intercalation [20]. These early surface science studies of graphene on metals had, however, little motivation in the direction of fundamental properties or applications of graphene in electronic devices. This was different for the studies of graphene formation on SiC. The formation of graphene on SiC was first observed in 1975 by van Bommel et al. [21]. In 2001 the de Heer group developed the process of forming planar graphene layers on SiC substrates further by heating SiC wafers to above 1300 °C. These studies were motivated by the prospect of 2D electronics and in 2004 they published a paper highlighting the 2D electron gas properties of the graphene charge carriers in an electric field [22]. In terms of the characterization of basic physical properties of graphene the breakthrough came, however, with the development of a mechanical exfoliation method of single and multilayer graphene from HOPG by Novoselov and Geim [23], [24], [25], [26]. Also at the same time the group around Kim [27], [28] developed their own exfoliation method for few-layer graphene. This preparation of high quality graphene allowed the verification of many predicted exotic behaviors of the charge carriers in graphene for which the Physics Nobel Prize was awarded in 2010. From an application perspective the exfoliation method is, however, unlikely to produce the graphene-materials needed for scalable production and device fabrication. Therefore, the above mentioned processes of graphene oxide reduction [29], [30] and chemical vapor deposition (CVD) growth processes on metal supports [31], [32] have been refined in recent years.
The literature on graphene is growing almost exponentially and therefore new developments are reported at an extraordinary pace. Nevertheless, for many aspects a thorough understanding is emerging which will be the basis for future studies and therefore we believe now is a good time to provide a summary of the experimental materials research of graphene. This review summarizes published results up to summer 2011. This review focuses on experimental advances. Theoretical studies are considered if they provide support of experiments in this review.
We start by investigating metal/graphene interfaces. This is a field with a long tradition in the surface science community and has reached maturity. There exist a large number of studies on single crystal substrates and the structure of graphene on these metal substrates is thoroughly characterized for a number of materials. The initial fundamental interest in these metal/graphene interfaces has become of applied interest due to its importance in graphene synthesis by CVD processes. In Section 3 we will investigate processes to form graphene with limited lateral extension, such as graphene nanoribbons. Graphene nanostructures are an important field of graphene research and lateral electron confinement in these nanostructures enables opening of a band gap in graphene. In Section 4 we focus on atomic-scale defects in graphene and at its edges. High resolution transmission electron microscopy and scanning tunneling microscopy have given us insight into defect formation in graphene at the atomic level. Defects can be utilized for modifying graphene properties and may play an important role for chemical functionalization of graphene. In this section we also consider substitutional impurity doping, namely nitrogen and boron doping as extrinsic defects in the graphene sheet. Furthermore, 1D defects, such as grain boundaries of graphene are included in Section 4. Section 5 discusses briefly the formation of chemical derivatives of graphene such as hydrogenated or fluorinated graphene. Section 6 studies interfaces between graphene and molecules. Especially organic interfaces have attracted interest for doping graphene by charge transfer from the organic molecules to graphene to create p- and n-type graphene. Section 7 concludes this review.
Section snippets
Graphene/metal interfaces
The observation of graphene-formation on transition metal surfaces dates back to the beginning of surface science studies on single crystal metals. On Pt(111) and Ru(0001) characteristic LEED patterns were observed after annealing to high temperatures. These LEED patterns were identified as originating from carbon segregation from the bulk and formation of graphitic layers. Later, graphene was grown on different metals intentionally by either saturating crystals with carbon outside of the UHV
Making ribbons and other graphene structures: cutting, etching, and template-growth of graphene
One attractive prospect of graphene, in contrast for example to carbon nanotubes, is that it can be handled as a wafer and patterned and cut by lithography methods much like in today’s silicon technology. A potentially major application of graphene is in high frequency field effect transistors (FETs). To use graphene in FETs a sizable band gap is required, however [133]. One way to introduce a band gap in graphene is to laterally confine the charge carriers, which may be achieved by making
Atomic-scale imperfections in the graphene
Defects can be characterized by their extent as zero, one, or two dimensional defects. A point defect (zero dimensional) may be vacancies or interstitials; in graphene a structural defect described by a bond rotation (Stone–Wales defects) is also possible that does not require any change in the carbon density. The restructuring of the carbon lattice surrounding these point defects causes them to exhibit a lateral extent of ∼1 nm. Extrinsic defects such as substitutional dopants on the other
Chemically modified graphene: graphane, fluorographene, and related materials
Local chemical modifications of graphene may be another approach to pattern extended graphene sheets. The chemical surface modifications and the formation of graphene with hydrogen or fluorine have been of particular interest. Graphane is a theoretically predicted compound with one hydrogen atom attached to each carbon atom on alternating sides of the graphene sheet [230]. According to computational results graphane is a wide-band gap semiconductor, whereas a half-hydrogenated graphene would
Molecular adsorption on graphene
For many electronic applications it is a prerequisite to control the doping of graphene, i.e. to make graphene p- or n-type by shifting the Fermi-level up or down from the Dirac point. Substitutional dopants such as B or N as discussed in Section 4.2 strongly affect charge carrier mobility in graphene and as of yet have to be better controlled. An alternative mechanism for shifting the Fermi-level, which is particular attractive for 2D materials like graphene, is to bring graphene in contact
Conclusion and outlook
The isolation of monolayer graphene as a freestanding material by mechanical exfoliation has enabled the investigation of the special physical properties arising from its linear band dispersion at the Dirac point. In addition, properties such as high charge carrier mobility, good optical transparency, and mechanical toughness makes it a promising material for microelectronic devices, transparent electrodes in optoelectronics, and a huge array of other potential applications that exploit the
Acknowledgment
The financial support from the Office of Naval Research through award # N00014-10-1-0668 is acknowledged.
References (266)
- et al.
Prog. Surf. Sci.
(2009) - et al.
Surf. Sci.
(1968) Surf. Sci.
(1969)- et al.
Surf. Sci.
(1992) - et al.
Surf. Sci.
(1974) - et al.
Surf. Sci.
(1997) - et al.
Surf. Sci.
(1975) - et al.
Surf. Sci.
(2009) - et al.
Surf. Sci.
(1997) - et al.
Surf. Sci.
(1993)