Being relatively small and electronegative, halogen atoms feature in a great number of high-valent species, such as NaCl3, HgF4 and IF7. The latter features heptacoordinate I(vii) — the highest coordination number for a main-group atom in a neutral compound. The chemical space occupied by iodine fluorides is rich, especially when one considers what can form at high pressures. A team led by Guochun Yang, Yan-ming Ma and Martin Rahm have used computational methods to survey this vast space, and they describe in Chemical Science how they came across IF8 as an energetically viable octacoordinate compound.

Credit: David Schilter/Springer Nature Limited

“Atomistic structure prediction of materials is extremely difficult because it involves classifying a huge number of energy minima on a multidimensional lattice energy surface,” reflects Ma. A particularly efficient way to explore such a surface is particle swarm optimization, whereby several potential solutions are each guided towards energy minima according to both the viability of their present positions and the viability of other solution positions. Beginning with I:F ratios, temperature (T→0) and pressure as inputs, one can search a surface for minimum energy structures, which can subsequently be optimized using density functional theory.

At ambient pressure, IF3, IF5 and IF7 exist as distorted T-shaped, square-pyramidal and pentagonal-biypyramidal molecules, respectively. When the pressure is ramped up in silico, Yang, Ma, Rahm and colleagues predict, for example, that IF3 polymerizes (23 GPa) before decomposing into IF5 and I2 (140 GPa). The team also discovered new species that are thermodynamically stable at 300 GPa. These include IF8 as well as the higher fluorides IF10, IF11 and IF12 that also feature octacoordinate I on account of having ‘free’ F2 in their lattices.

In atomic I, the 5d orbitals lie so high in energy above the 5p set that hybridization is impossible. The situation is very different when the same I atom is at the centre of a distorted cube defined by eight F atoms that come close when the system is subjected to 300 GPa. Here, the 5d orbitals, split by the cubic ligand field into lower-lying eg and higher-lying t2g sets, can overlap with filled F-centred orbitals in an interaction familiar to inorganic chemists. The team computed the relative occupancies of the I-centred valence orbitals to be s1p1.7d1, confirming the relevance of the 5d set to the frontier molecular orbitals of IF8, a truly hypervalent molecule. At ambient pressure, the anion [IF8]− has a square antiprismatic I(vii) centre and is also hypercoordinated. However, the I 5d orbitals are too high in energy to participate in covalent bonding and it would be contentious to call [IF8]− hypervalent.

Yang and co-workers acquired independent evidence for the involvement of 5d orbitals in the bonding of IF8 by calculating its projected density of (electronic) states, which indicates mixing of I 5d orbitals with other I and F valence orbitals. Moreover, the Fermi level intersects the F 2p band, suggesting that crystalline IF8 is metallic and has F-centred vacancies. Thus, IF8 is not an I(viii) complex but is best considered an I(vii) centre surrounded by a set of eight F− ligands in which there is one (delocalized) hole.

there is a growing need to expand our chemical design intuition to encompass the high-pressure regime

Although predicted to be stable at high pressures and ultralow temperatures, IF8 is dynamically unstable towards dissociation under ambient pressure — a likely reason why this compound, which would be challenging to prepare, has escaped the attention of many. Rahm points out “there is a growing need to expand our chemical design intuition to encompass the high-pressure regime.” Only then will we see how far we can push hypercoordination and hypervalence in main-group compounds.