Molecules with largest dipole moments aren’t due to electronegativity differences

Computational analysis has revealed that diatomic molecules with two metal atoms of similar electronegativities can have a much greater dipole moment than those conventionally thought of as polar. This challenges chemists’ understanding of what polarity is, with these compounds potentially finding use in ultra-cold experiments that need highly polar gases.

‘If you ask any chemist “What is the diatomic molecule with the largest dipole moment?”, they’re [likely] going to pick fluorine, because it’s the halogen with the largest electronegativity, and then they will pick something like caesium [with a very small electronegativity value],’ says Jesús Pérez Ríos at Stony Brook University in the US.

However, rationalising a diatomic molecule’s dipole moment goes beyond Linus Pauling’s idea of electronegativity differences, with bond length and sizes of atoms also influencing bond polarity, says Pérez Ríos. Chemists recently calculated that there are some molecules with higher dipole moments than caesium fluoride (7.88 Debye), even when they have similar electronegativities.¹ This includes copper or silver atoms bonded to group 1 or 2 metals, where these diatomic molecules can have a dipole moment up to 13 Debye.

Pérez Ríos and his team have now developed a machine learning model that analyses all 7021 possible combinations of atoms in diatomic molecules – 6903 heteronuclear and 118 homonuclear pairings – to see if there are any other unusually large dipole moments.²

The team trained the model using both experimental data of dipole moments of existing diatomic molecules, as well as dipole moments of theorised compounds.

Analysis of modelled molecules that the system hadn’t been trained on revealed that heavy halogen atoms – such as iodine or astatine – bonded to large alkali atoms like caesium or francium had some of the greatest dipole moments. Swapping the halogen atom out for gold also gave unexpectedly high dipole moments, with the caesium–gold bond reaching over 11 Debye. Pérez Ríos explains this is due to gold’s partially filled d-orbital ‘hole’ that can accept electron density – much like a halogen atom.

‘[Researchers] are crazy about polar gases with a huge dipole moment so that [they] can study long-range interactions [in cold environments],’ says Pérez Ríos. He thinks that caesium–gold could be an alternative to sodium–caesium, which is the current limit for polar gases with a dipole moment around 5 Debye.

While the model is not ‘super accurate’ at predicting the exact dipole moment of a given molecule, Pérez Ríos says that it is able to correctly predict the trends as to which types of molecules will give large dipole moments.

Laura McKemmish, a computational chemist at the University of New South Wales, Australia, explains that ‘some of the trends seen in diatomics may not translate directly to more complex chemistry’. ‘Studies like this are particularly useful for highlighting unusual cases that prompt chemists to look more closely at the underlying physics.’