Eur. Phys. J. D 34, 59–61 (2005) DOI: 10.1140/epjd/e2005-00119-4 T HE EUROPEAN P HYSICAL JOURNAL D Structural and electronic properties of C 6 cluster D. Zhang 1 , F. Jin 1 , and J. Yuan 1, a Department of Applied Physics, National University of Defense Technology, Changsha, 410073, P.R. China Received 6 September 2004 Published online 13 July 2005 – c  EDP Sciences, Societ`a Italiana di Fisica, Springer-Verlag 2005 Abstract. First-principles total-energy calculations of structural properties of the carbon cluster C6 have been made using the full-potential augmented plane-waves plus local orbital (APW+LO) method with the generalized gradient approximation (GGA). Initiated from a hexagonal configuration, we performed geometry optimization with damped Newton dynamics. The computed ground state atomic configuration for C6 belongs to a monocyclic D 3h structure. The average bond length is 1.52 a.u. and the bond angle is 90.2 ◦ , respectively, which are in agreement with the reported results. PACS. 31.15.Ar Ab initio calculations – 31.15.Qg Molecule dynamics and other numerical method – 36.40.-c Atomic and molecular clusters Owing to their unusual electronic and structural proper- ties, small clusters have received considerable attention recently [1,2]. Intense interest in the small carbon clus- ters C 2-5 is one of the good examples of the flurry of studies [3–5]. A remarkable characteristic of clusters is the changes in their physical and chemical properties with the size. In early studies, most of the larger carbon clusters were believed to be linear species. However, recent work has pointed out that the configurations of small carbon clusters may be cyclic [6]. So it is a fundamental problem to understand the evolution of configurations and proper- ties of carbon clusters by increasing the size from an atom to a bulk crystal. For small carbon clusters, some of the changes in properties with size is striking. These changes are strongly related with their atomic structures, which differ substantially from that of the bulk crystal. Clearly, the role of the surface induced properties become increas- ingly important with decreasing size of clusters. However, most of structural calculations have been performed either using semi-empirical techniques or have been restricted to small basis set Hartree-Fork studies [7]. For bare carbon clusters, unsaturated carbon bonds can exit on the sur- face. These dangling carbon bonds can result in unusual structures and properties. This phenomenon can be seen on the free surface of a bulk crystal. The impressing ex- ample is the (7 × 7) reconstruction for the (111) cleavage plane of the Si crystal [8]. In small carbon clusters, where almost all the carbon atoms reside on the surface, this phenomenon is reflected in structures that deviate sub- stantially from that of their crystalline material. a e-mail: jmyuan@nudt.edu.cn While the properties of clusters offer new and novel states of matter, the synthesis and characterization of small clusters encounter numerous difficulties. Since clus- ters of atoms are, by definition, only stable in isolation, experimentalists face the challenge of maintaining an iso- lated system while probing its properties. Typical experi- mental techniques include cluster deposition on the inert substrate [9], and the use of cluster beam for photoemis- sion measurements [10]. Theoretical studies of clusters are also challenging due to the many degrees of freedom, and the lack of good structural information. Structural deter- minations are not easy considering that the atomic ar- rangements and coordination can differ considerably from that of the equivalent crystal. Some of the different con- figurations of a moderately sized cluster have almost the same local minima in the potential-energy surface, which further complicated this problem. For clusters with more than ten or so atoms, it is impossible to make a inventory and decide by direct calculations which geometry is low- est in energy. At present there are few reliable methods for predicting the structure of a given cluster. In this paper, we will present first-principles calculations for the ground state configuration of carbon cluster C 6 . The key problem in determining the structure of a cluster is to develop a method which can accurately ac- count for energies and interatomic forces. One approach applies empirical interatomic potentials, which are usu- ally obtained by fitting to the known crystalline forms. The method is not necessarily a good approximation, for it requires a careful construction and accurate knowledge of the chemical bonding. Since clusters often have bonding configurations that differ substantially from the crystalline phase, the database used to fit empirical potentials may often prove inadequate when determining the parameters.