All-boron aromatic clusters as potential new inorganic ligands and building blocks in chemistry

Abstract

Small boron clusters as individual species in the gas phase are reviewed. While the family of known boron compounds is rich and diverse, a large body of hitherto unknown chemistry of boron has been recently identified. Free boron clusters have been recently characterized using photoelectron spectroscopy and ab initio calculations, which have established the planar or quasi-planar shapes of small boron clusters for the first time. This has surprised the scientific community, as the chemistry of boron has been diversely featured by three-dimensional structures. The planarity of the species has been further elucidated on the basis of multiple aromaticity, multiple antiaromaticity, and conflicting aromaticity.

Although mostly observed in the gas phase, pure boron clusters are promising molecules for coordination chemistry as potential new ligands and for materials science as new building blocks. The use of pure boron species as novel ligands has commenced, suggesting many new chemistries are ahead of us.

Introduction

Boron compounds have been known to humankind since ancient times, when it was used to prepare hard glasses and glazes [1]. Nowadays the use of boron compounds ranges from hard materials and semiconductors to antitumor medicines, and its importance cannot be overestimated. Neighboring carbon in the periodic table, boron has one electron less than valence orbitals, and that makes a huge difference in determining the chemistry of boron. Although boron has a rich and diverse chemistry, it differs substantially from that of carbon [1], [2].

Boron compounds played essential roles in advancing chemical bonding models. A few milestones in the history of contemporary boron chemistry should be pointed out. In 1912 Stock reported his pioneering work on boranes [3], which led to the identification of the neutral boron hydrides with formulas B₂H₆, B₄H₁₀, B₅H₉, B₅H₁₁, and B₆H₁₀. These compounds were characterized as toxic, air-, and water-sensitive gases, or volatile liquids. A larger compound, B₁₀H₁₄, was isolated as a volatile solid. Although the structures of the boranes were established then, the chemical bonding within them remained unclear, as the stoichiometry of the species contradicted the postulates of valence theory. Even the reason for the rapid dimerization of BH₃ into B₂H₆ was a puzzle. The structure of B₂H₆ with bridging H-atoms was proposed in 1921 by Dilthey [4]. However, it was not considered seriously until the 1940s, when infrared spectroscopy data [5], [6], [7] supported the structure. Later, electron diffraction [8] and low-temperature X-ray diffraction [9] also confirmed the bridged structure for the diborane. The chemical bonding in boranes was first considered by Pitzer, who proposed the concept of a “protonated double bond” [10].

Further, Lipscomb and co-workers [11] put forward the concept of three-center two-electron (3c-2e) bonding, which, in the case of the B₂H₆ diborane, consisted of two 3c-2e B–H–B bonds involving the bridging hydrogen atoms. Lipscomb also explained the structure of all known boron hydrides, in which the bridging B–H–B bond appeared to be the key structural unit [9]. In the 3c-2e bonding three atoms supply three orbitals, one on each atom. These atomic orbitals interact to form one bonding and two antibonding orbitals. The two available electrons may thus fill the bonding orbital to form a 3c-2e bond. In the n-atomic species, there are n atomic orbitals, and only n/3 bonding molecular orbitals, which can be occupied by 2n/3 electrons. Thus, the reason for certain boranes to exhibit special stability was elucidated. In principle, Lipscomb's concept of the 3c-2e bond, along with aromaticity, is one of the ways of describing electron deficient bonding, even though aromaticity is more common in chemistry and, in a way, more clear. The work of Lipscomb on the chemical bonding of the boranes eventually led to his winning of the Nobel Prize and opened the gateway to understanding the chemistry of boron.

Allard in 1932 [12], and Pauling and Weinbaum in 1934 [13] showed the existence of regular octahedra of boron atoms in several metal hexaborides of the general formula MB₆. These early works represented the first experimental demonstration of closed boron polyhedra in a chemical structure. A subsequent related work by Longuet-Higgins and Roberts [14] used molecular orbital theory to show that the [B₆]²⁻ unit has a closed-shell electronic arrangement of high stability. Longuet-Higgins and Roberts [15] also used a similar approach to study the B₁₂ icosahedron, a dominant structural unit of various allotropes of boron [16]. Their work indicated that the B₁₂ icosahedron has 13 skeletal bonding orbitals and 12 outward pointing external orbitals. They concluded that a borane, B₁₂H₁₂, would be stable only as a dianion, B₁₂H₁₂²⁻, as well as a series of stable cage-like borane dianions of the general formula, B_nH_n²⁻. The conclusion was supported by experimental data provided by Hawthorne and Pitochelli, who synthesized salts of the borane anion, B₁₂H₁₂²⁻ [17], [18]. From X-ray diffraction experiments, the structure of the B₁₂H₁₂²⁻ anion was indeed shown to be icosahedral [19] and another B₁₀H₁₀²⁻ anion was shown to be a bicapped square antiprism [20]. Further experiments [21], [22], [23] indicated the existence of other deltahedral boranes, B₁₁H₁₁²⁻, B₉H₉²⁻, B₈H₈²⁻, B₇H₇²⁻, and B₆H₆²⁻.

Different rules were developed for the number of atoms, bonds, electrons, and orbitals in stable boranes [24], [25], [26], [27], [28]. Dixon et al. made the first attempt to describe boranes in terms of resonance of Kekule-type structures with alternating 2c-2e and 3c-2e bonds [29]. In 1971 Williams [30] recognized the closo-, nido-, and arachno-structural motifs in the chemistry of deltahedral boranes. The most spherical deltahedra have a formula B_nH_n+2 (or B_nH_n²⁻). The loss of a vertex from this closo-form results in the B_nH_n+4 (B_nH_n+2²⁻) nido-structure. The arachno-structure, B_nH_n+6 (B_nH_n+4²⁻), can be formed by the removal of yet another vertex from the deltahedron. Wade [31] recognized that this structural relationship could be associated with the number of skeletal electrons in a borane. Namely, he understood that closo-, and the nido- and arachno-shapes, have the same number of skeletal electrons: 2n + 2. These species have the same number of molecular orbitals belonging to the boron skeleton: closo-boranes have to have n + 1 MOs, nido-boranes have to have n + 2 MOs, and arachno-boranes have to have n + 3 MOs. This set of rules is known as Wade's rules for boranes and carboranes. Gellespie et al. also proposed another scheme, in which skeletal electron pairs (2n electrons) remained localized on each vertex, whereas two electrons participated in the delocalized bonding over the spherically symmetric polyhedron [32]. Furthermore, the relative stabilities of deltahedral boranes form a series [33]:

B₁₂H₁₂²⁻ > B₁₀H₁₀²⁻ > B₁₁H₁₁²⁻ > B₉H₉²⁻ ∼ B₈H₈²⁻ ∼ B₆H₆²⁻ > B₇H₇²⁻

In 1959 Lipscomb et al. [34] proposed the term “superaromaticity” to explain the 3D aromaticity of B₁₂H₁₂²⁻. Chen and King recently reviewed [35] the introduction of 3D aromaticity in chemistry. Explicitly, the idea of aromaticity of deltahedral boranes was put forward by Aihara [36], and by King and Rouvray [37] in 1978. Carboranes of the formula C₂B_n−2H_n also exhibit deltahedral topology, and the chemical bonding within them can be described in the same manner [38]. The only difference in the borane-carborane-electronic relationship is that C₂B₃H₅, being an electronic analog of B₅H₅²⁻, is a known compound, while B₅H₅²⁻ itself has not been synthesized. The three-dimensional aromaticity in boranes, B_nH_n²⁻, closo-monocarborane anions, CB_n−1H_n⁻, and closo-dicarboranes, C₂B_n−2H_n was also computationally studied by Schleyer and Najafian [39], who used the NICS index values as criteria of aromaticity. The authors showed the general trend in such systems: the stability increases with increasing verteces from 5 to 12. The comparison of the chemistry of boron and carbon on the example of boron and carbon hydrides was also discussed by Jemmis and Jayasree [40].

In 2003, Boldyrev and co-workers theoretically predicted the new family of planar aromatic highly charged boranes, such as B₆H₆⁶⁻ stabilized by six Li⁺ cations surrounding the species (structure A, Fig. 1) [41].

Importantly, the planar B₆-fragment has been previously known to be the building block of the MgB₂ solid, which is a recently discovered high-temperature superconductor [42]. An extended analysis of the chemical bonding in the lowest energy planar and nonplanar isomers involving B₆H₆⁶⁻ has been performed [41b]. Salts like Li₆B₆H₆ with the B₆H₆⁶⁻ benzene-like analog are still a theoretical prediction, but Fehlner and co-workers [43] recently reported synthesis and crystal structures of remarkable triple-decker (Cp^*ReH₂)B₅Cl₅ and (Cp^*)₂B₆H₄Cl₂ compounds containing planar B₅Cl₅ and B₆H₄Cl₂ structural fragments, respectively (Fig. 2).

It is believed that the planar B₅Cl₅ and B₆H₄Cl₂ structural fragments in Felhner's compounds acquire six electrons from the Re atoms formally and thus become six π-electron aromatic compounds similar to the predicted B₅H₅⁶⁻ and B₆H₆⁶⁻ building blocks in the Li₆B₅H₅ and Li₆B₆H₆ salt molecules.

Beyond boranes and carboranes a very rich family of metallocarboranes has been synthesized and characterized [1], [2], [38], [43], [44], [45], [46], [47]. One of the exciting new applications of these species was a demonstration of rotary motion of a carborane cage ligand (7,8-dicarbollide) around a nickel axle controlled by electrical or light energy, thus moving us closer to inorganic nanomachines [45]. Boron atoms can also be incorporated into transition metal clusters [44], [45], [46], [47] further extending the rich boron chemistry.

Experimental studies of bare boron clusters have been surprisingly limited despite its proximity to carbon in the periodic table and the extensive research effort on carbon clusters and the fullerenes ([48], [49], [50], [51], [52] and references therein). Until very recently the only experimental studies on boron clusters were carried out by Anderson and co-workers in the late 1980s [53], [54], [55]. These authors produced boron cluster cations using laser vaporization and studied their chemical reactivity and fragmentation properties [53], [54], [55], [56], [57], [58], [59]. Their observation of the prominent B₁₃⁺ peak in mass spectra stimulated further computational efforts [60], [61], [62], [63], [64]. In 1992, La Placa et al. reported a mass spectrum of bare boron clusters consisting of 2–52 atoms, generated by laser vaporization of a boron nitride target in an effort to produce BN clusters [65]. Little information regarding the structural and electronic properties of boron clusters can be drawn from the mass-spectrometry-based studies. And there have been no spectroscopic studies until very recently.

Starting from 2001, we have conducted extensive photoelectron spectroscopy (PES) studies on a series of boron cluster anions, B_n⁻ (n = 3–20), which were combined with state-of-the-art computational studies to elucidate the structural and electronic properties and chemical bonding in the species [66], [67], [68], [69], [70], [71], [72]. These works established for the first time experimentally planarity or quasi-planarity in small boron clusters for as large as 20 atoms. Among our findings are the observation and characterization of the hepta- and octa-coordinated pure boron molecular wheels in B₈ and B₉ [70], the observation of an unusually large HOMO–LUMO gap in B₁₂ [71], and the discovery that the planarity or quasi-planarity of boron clusters can be explained on the basis of multiple σ- and π-aromaticity, multiple σ- and π-antiaromaticity, and conflicting aromaticity (simultaneous presence of σ-aromaticity and π-antiaromaticity or σ-antiaromaticity and π-aromaticity) [66], [67], [68], [69], [70], [71], [72]. We also found that the π-aromaticity in small boron clusters up to n = 15 seems to follow the Hückel's rules, analogous to hydrocarbons [71]. We showed that the B₂₀ neutral cluster appears to possess a ring-like 3D ground state geometry although both planar and 3D structures are nearly isoenergetic for the B₂₀⁻ anion [72]. The B₂₀ neutral therefore represents the planar-to-tubular transition in small boron clusters and may be viewed as the embryo of the thinnest single-walled boron nanotubes with a diameter as small as 5.2 Å [72]. Very recently, several other experimental efforts have also appeared on boron clusters [73], [74], [75], [76], [77].

In spite of the enormous variety of boron chemical compounds and their great influence on developing modern chemical bonding theory, bare boron clusters as ligands in chemical compounds are still absent. Given the fact that our recent spectroscopic and theoretical studies have shown that isolated boron clusters have planar geometries and exhibit aromatic and antiaromatic electronic properties analogous to hydrocarbons, it is believed that bare boron clusters could potentially be new ligands or building blocks of new solids. It is worth mentioning that a planar aromatic cyclopentadien anion C₅H₅⁻ is one of the most common ligands in coordination chemistry. In this review we summarize theoretical and spectroscopic studies of isolated all-boron clusters with the hope that the latest experimental and theoretical understanding may stimulate further investigations leading to novel coordination compounds containing the planar all-boron clusters as ligands or new building blocks.