A thermal treatment of different carbon forms can lead to the formation of onion-)ike species as well. As for diamond particles, their surface structure plays an important role for the actual outcome of the process. If it is covered with functional groups, the bonding sites are saturated which renders a graphitization more difficult. From dangling bonds, on the other hand, graphitized domains will arise that can serve as nucleation center to the formation of carbon onions. In this process, a suitable orientation of lattice planes as well as a small particle size that entails a larger portion of partly graphitized structures favor the generation of onion-°ike material. Bulk diamond is hard to be converted into carbon onions, whereas nanoscale diamond particles already change into - initially unordered- sp2-structures at about 1170 °C. The conversion expectedly starts on the surface. The heating serves, among others, to the removal of functional groups from the surface and to the concomitant partial graphitization.
At higher temperatures then a faster generation of sp2-structures sets in and a particle with an onion-like shell on a diamond core is formed as an intermediate (Figure 4.25). For the close proximity of sp2-shell and sp3-core (the distance is never more than 0.35 nm), one can well assume a chemical bonding between them. Just like in the process induced by electron radiation, the transformation preferably starts from the diamond's (111)-planes as due to a reconstructive mechanism, they are prone to form graphitic surface structures (Section 6.2.2) . Again the (111)-planes of diamond are converted into the (001)-planes of the curved graphitic material (Section 126.96.36.199) . Depending on the structure of the starting material, either an exfoliation of graphene layers from the diamond surface or a transformation of (111)-planes confined on the edges is observed (Figure 4.26). Sometimes onions with a three-dimensionally spiral structure are found.
The formation of spiral intermediates is observed for the conversion of sp2 -carbons (induced both by electron bombardment or thermally) and, in parts, of nanodiamond as well. The resulting objects represent the three-dimensional equivalent to the rolled up graphene layers that have been discussed as a hypothetical structure of multiwalled nanotubes (Section 3.2.3) . This growth pattern has been termed "spiral" or "snow accreting mechanism." It is based on the assumption of the onion growing outward from the center, which is supported by electron microscopic observation. Bowl-shaped aromatic compounds constitute the initial nuclei in this process. They are available especially in soot-like starting materials. The continued addition of further carbon atoms results in an ever larger and increasingly spherical object which ideally is closed to a cage. In most cases, however, the open edges do not meet due to an unsuitable curvature, so an overlap is formed instead. The basic unit of a three- dimensional spiral has come into being. From this moment on, the van der Waals forces constantly affect a sufficiently low distance between the outermost edge and the layer below (Figure 4.27).
Further growth then always takes place on this outer edge with the mechanism resembling the accretion of a snowball. Due to this interaction with graphene layers situated farther inside, the addition may be considered some kind of "epitaxial" process, which ensures a constant interlayer distance. In macroscopic spiroids like the nautilus (Figure 4.28), on the other hand, this distance normally increases toward the outside.
The spiroids formed are always in equilibrium with partly concentric and really onion-like structures. Due to the strong curvature and the high concentration of dangling bonds, the inner edge of a carbon spiroid is rather reactive. Hence, it is able to attack on the neighboring outer layer (Figure 2.29). In doing so, a closed fullerene cage is completed in the center and a new edge of dangling bonds is formed on the adjacent shell. Likewise, the transformation proceeds toward the outside. Provided the electron radiation or heating are kept up, a completely concentric carbon onion is formed in the end. The processes and species involved here include both radical reactions and aryne structures (Figure 4.29 ).
It is clearly to be seen from electron micrographs that the conversion of carbon spiroids into nano- onions really proceeds from the core to the periphery. They show entirely spiral objects at first. These are transformed into completely concentric onions via the hybrid form of an onion core with a spiral shell (Figure 4.30). It is presumably even just a part of the spiroid structures actually present in the sample that are detected in these examinations as their projections appear onion-like at unfavorable orientation.
At very high temperatures like in the plasma zone of an arc discharge apparatus, a vaporization of small carbon clusters occurs, and even atomic carbon may be generated. In this case, the carbon onions do not grow by a displacement of atoms inside an already existing structure, but they rather assemble stepwise, starting from a nucleation center. Among the species observed in the gas phase of such a process, it is especially C2-clusters that are found in high concentration. At first,
under an electron microscope with the rearrangement proceeding from the outside to the inside (b), ©ACS 2002).
Figure 4.31 The accretion of curved graphitic structures and of annealed aromatic compounds also leads to a formation of carbon onions. The inner shells are not always completely closed, which explains for defects of the onion structure (© ACS
polycyclic aromatic structures emerge. Due to the tendency toward bond saturation, they will soon form the five-membered rings that also induce curvature.
This structure then keeps growing by the addition of further carbon units. The size and number of defects generated in doing so and the proportion between them give rise to different degrees of curvature which, in their turn, result in closed cages (Section 2.2.3) or in the aforementioned spiroids (Figure 4.27). The conversion into carbon onions then takes place as described above for the spiral mechanism. Still the growth of onion -like structures could also be explained by already closed cages adding further atoms and clusters. Especially in a plasma zone that is strongly confined by cooling, existing cages serve as nucleation centers indeed. Onion-growth occurs shell-wise according to this model, and inner layers do not necessarily have to be completed before another cluster can add to the surface (Figure 4.31). This hypothesis would also account for defects that are often observed inside of carbon onions.
Altogether it is reasonable to assume that different mechanisms, and predominantly among them a spiral and a shell-wise growth, proceed in parallel, regardless of starting material and method of preparation.
Was this article helpful?