When oxygen atoms bind to a graphite surface, they fall into line and make bridges across carbon atoms. This is the spearhead of a chemical attack in which the atomic arrangement of solid carbon is torn apart.
What happens when we burn carbon? Combustion seems such a simple reaction, but at an atomic scale it is the result of several steps. When graphite burns, for example, an intermediate stage is the formation of a graphite oxide1. This process destabilizes the ordered arrangement of carbon atoms, so that cracks begin to form and the structure breaks up. Controlled oxidation reactions are useful for the preparation of very thin graphite flakes, or for chopping carbon nanotubes into shorter lengths, but the details of how oxygen attacks carbon bonds to break up the atomic structure of graphite have never been understood. Writing in Physical Review Letters, Je-Luen Li et al.2 provide an explanation for this fundamental process.
The carbon atoms in graphite are arranged in a hexagonal lattice like a honeycomb, with each carbon bonded to three others, forming flat sheets held together by weak van der Waals forces. The bonds between the atoms are chemically inert, so unless there are flaws in the lattice, such as missing carbon atoms, the interaction of oxygen gas with a graphite surface is weak. During oxidation, the graphite oxide that forms is a complex structure1 in which oxygen-containing groups are randomly attached to the honeycomb lattice. These groups, which can take part in varied surface chemistry, mainly attach at defects or edge-atom sites in the lattice, where the carbon atoms are not fully bonded to other atoms. So far so good, but how do the surface-bound oxygen-containing groups trigger the break-up of the carbon lattice, ultimately destroying the entire graphitic structure?
One kind of chemical group that forms on the graphite surface is known as an epoxy bridge, where a single oxygen atom bonds to two adjacent carbon atoms, forming a triangle. In their paper, Li et al.2 describe how the stress generated by these epoxy bridges leads to unravelling of the graphite lattice. Each epoxy bridge is severely strained, because the geo-metry of the incorporated carbon atoms has changed. Where the bridged atoms were once only bound to other carbon atoms in a planar, hexagonal honeycomb arrangement, they are now bound to an oxygen atom sitting above the lattice surface, in place of a lattice carbon atom, and adopt an almost three-dimensional, distorted form. This new geometry doesn't fit well in the remaining lattice — it's like trying to fit a square peg into a round hole.
Mechanistically, the oxygen atom acts as a minuscule wedge, pushing apart the bridge's carbon atoms and stretching the carbon–carbon bond. The epoxy groups do not act individually, but cooperate. The authors' extensive calculations2, based on density functional theory, show that side-by-side parallel positioning of the epoxy bridges is energetically favoured, so that they tend to line up on the graphite surface (Fig. 1). As a result, they collectively induce enough tension in the underlying lattice to break the native carbon bonds.
The feedback guiding such organized attachment of oxygen to graphite resembles generic brittle fracture, where the site of the next bond failure is determined by the existing crack as it propagates through the material. Having broken up carbon bonds, the bridging oxygen atoms go on to hold the fractured graphite sheet together, forming a seam between the separated lattice fragments. These fragments are held together at an angle to each other, to comply with the chemically preferred obtuse angle at the oxygen joints. The emerging network of such oxygen-zipped ridges results in the crumpling of a single graphite layer, known as a graphene sheet. This sheet flakes away from the stack of solid graphite, breaking the weak van der Waals forces that once bound it (Fig. 1).
The emerging picture of isolated epoxy groups falling into ranks, to take part in a well-orchestrated serial bond-breaking process, is rather appealing. But the kinetics require further explanation. Randomly bound epoxy groups are unaware of the energetic benefits of forming lines. The epoxy groups can only find such stable alignments as a result of rapid hopping around, but this is not easy to reconcile with the high energy barriers to such hopping. Further study is also required to determine exactly how visible faults or crack discontinuities emerge. Whatever the details may be, the lines of epoxy groups predetermine the tear pattern, like a perforation line directs the tear in a sheet of postage stamps. The resulting fault lines become clearly visible in optical microscope images of oxidized graphite2.
A similar sequence of events can also be expected in the oxidation of carbon nanotubes. These share the same bonding as in graphite, except at the end caps where defects are present. Oxidation treatment in strong acids has led to the selective removal of end caps3 and the breaking of long nanotubes into small pieces4. One would expect epoxy groups to line up circumferentially in a nanotube, leading to its transverse fracture into shorter segments, in a useful cutting process. On the other hand, uncontrolled oxidation can severely compromise the strength of nanotubes5.
The paper by Li et al.2 provides insight into the atomic-level mechanisms of oxidation in carbon. Graphite and its artefacts, such as carbon nanotubes, are materials with a wide range of uses, from lubrication to electronics. Controlled oxidative scission to extract nanoscale graphitic structures (for example, cut-to-size nanotubes6 or nanosize graphene sheets7) from larger domains of these materials would be an extremely powerful technique for all sorts of applications. Understanding how oxygen breaks up the atomic structure of graphite could lead to a whole new area of nanotechnology based on nanoscale graphite origami8.
About this article
Journal of Materials Science (2014)