After the discovery of fullerene-C60, it took almost two decades for the possibility of boron-based fullerene structures to be considered. So far, there has been no experimental evidence for these nanostructures, in spite of the progress made in theoretical investigations of their structure and bonding. Here we report the observation, by photoelectron spectroscopy, of an all-boron fullerene-like cage cluster at B40(-) with an extremely low electron-binding energy. Theoretical calculations show that this arises from a cage structure with a large energy gap, but that a quasi-planar isomer of B40(-) with two adjacent hexagonal holes is slightly more stable than the fullerene structure. In contrast, for neutral B40 the fullerene-like cage is calculated to be the most stable structure. The surface of the all-boron fullerene, bonded uniformly via delocalized σ and π bonds, is not perfectly smooth and exhibits unusual heptagonal faces, in contrast to C60 fullerene.

Boron is an interesting element with unusual polymorphism. While three-dimensional (3D) structural motifs are prevalent in bulk boron, atomic boron clusters are found to have planar or quasi-planar structures, stabilized by localized two-center-two-electron (2c-2e) σ bonds on the periphery and delocalized multicenter-two-electron (nc-2e) bonds in both σ and π frameworks. Electron delocalization is a result of boron's electron deficiency and leads to fluxional behavior, which has been observed in B13(+) and B19(-). A unique capability of the in-plane rotation of the inner atoms against the periphery of the cluster in a chosen direction by employing circularly polarized infrared radiation has been suggested. Such fluxional behaviors in boron clusters are interesting and have been proposed as molecular Wankel motors. The concepts of aromaticity and antiaromaticity have been extended beyond organic chemistry to planar boron clusters. The validity of these concepts in understanding the electronic structures of boron clusters is evident in the striking similarities of the π-systems of planar boron clusters to those of polycyclic aromatic hydrocarbons, such as benzene, naphthalene, coronene, anthracene, or phenanthrene. Chemical bonding models developed for boron clusters not only allowed the rationalization of the stability of boron clusters but also lead to the design of novel metal-centered boron wheels with a record-setting planar coordination number of 10. The unprecedented highly coordinated borometallic molecular wheels provide insights into the interactions between transition metals and boron and expand the frontier of boron chemistry. Another interesting feature discovered through cluster studies is boron transmutation. Even though it is well-known that B(-), formed by adding one electron to boron, is isoelectronic to carbon, cluster studies have considerably expanded the possibilities of new structures and new materials using the B(-)/C analogy. It is believed that the electronic transmutation concept will be effective and valuable in aiding the design of new boride materials with predictable properties. The study of boron clusters with intermediate properties between those of individual atoms and bulk solids has given rise to a unique opportunity to broaden the frontier of boron chemistry. Understanding boron clusters has spurred experimentalists and theoreticians to find new boron-based nanomaterials, such as boron fullerenes, nanotubes, two-dimensional boron, and new compounds containing boron clusters as building blocks. Here, a brief and timely overview is presented addressing the recent progress made on boron clusters and the approaches used in the authors' laboratories to determine the structure, stability, and chemical bonding of size-selected boron clusters by joint photoelectron spectroscopy and theoretical studies. Specifically, key findings on all-boron hydrocarbon analogues, metal-centered boron wheels, and electronic transmutation in boron clusters are summarized.

Boron is carbon's neighbour in the periodic table and has similar valence orbitals. However, boron cannot form graphene-like structures with a honeycomb hexagonal framework because of its electron deficiency. Computational studies suggest that extended boron sheets with partially filled hexagonal holes are stable; however, there has been no experimental evidence for such atom-thin boron nanostructures. Here, we show experimentally and theoretically that B36 is a highly stable quasiplanar boron cluster with a central hexagonal hole, providing the first experimental evidence that single-atom layer boron sheets with hexagonal vacancies are potentially viable. Photoelectron spectroscopy of B36(-) reveals a relatively simple spectrum, suggesting a symmetric cluster. Global minimum searches for B36(-) lead to a quasiplanar structure with a central hexagonal hole. Neutral B36 is the smallest boron cluster to have sixfold symmetry and a perfect hexagonal vacancy, and it can be viewed as a potential basis for extended two-dimensional boron sheets.