# Can an atom bond with more than 8 other atoms?

Is it possible for an atom to bond with 8 other elements (same or other type)? If yes, then please give some examples. If no, then what could be the possible reason for it?

My question is not about valence shell electrons. My question is about whether an element could combine with more than 8 other same or different elements? It could be rather covalent bonds or any bond.

• I was thinking in 3D; most 6 coordinate atoms have an octahedral shape. But bon's answer made me think of ferrocene.
– LDC3
May 24 '15 at 18:24
• May 24 '15 at 18:50

Yes, there are coordination complexes of large elements which have coordination numbers greater than eight. Some examples are:

• $\ce{[ReH9]^2-}$ with a tricapped trigonal prismatic structure. The nine hydride ligands are small enough to fit around the relatively large rhenium atom fairly easily. This ion can be isolated as a potassium salt $\ce{K2ReH9}$.

• $\ce{[U(NO3)6]^2-}$ with an impressive coordination number of 12, consisting of six nitrate ligands each forming two coordinate bonds to the central uranium ion.

• Counting only monohaptic ligands, coordination number 9 seems to be the highest currently known, in compounds such $\ce{[ReH9]^{2-}}$, though a coordination number as high as 15 has been predicted for $\ce{[PbHe_15]^{2+}}$. Allowing for multihaptic ligands, a coordination number of 20 is found in $\ce{Th}\mathrm{(\eta ^5-}\ce{C5H5}\mathrm{)_4}$. May 24 '15 at 18:59
• @Alizter Orbitals from all 20 carbon atoms are involved in the bond with the central thorium ion, so yes. May 24 '15 at 19:07
• @NicolauSakerNeto I've seen a number of papers argue that 20-coordination in $\ce{M(C5H5)4}$ is not an accurate view. The argument goes that each double bond should count as one 2-electron contact. Hence there would be 3 contacts for each $\ce{C5H5}$ for a total (Werner) coordination number of 12.
– ron
May 24 '15 at 20:50
• @ron The x-ray data for the cyclopentadienyl ligands have equal bond distances and the entity is flat, which means that the bonding of cyclopentadienyl is distributed equally on the ring. Since there are 5 orbitals available, how could the bonding be characterized as 3?
– LDC3
May 24 '15 at 21:16
• @LDC3 Read the second paragraph in ref_2 in my answer. Also, Wikipedia says, "There are various ways of assigning the contribution made to the coordination number of the central iron atom by each cyclopentadienide ligand. The contribution could be assigned as one since there is one ligand, or as five since there are five neighbouring atoms, or as three since there are three electron pairs involved. Normally the count of electron pairs is taken" (emphasis mine).
– ron
May 25 '15 at 16:23

14 coordination is claimed in $\ce{U(BH4)4}$ (ref_1, p. 268). The molecule exists as a polymer in the solid state. Six hydrogens from two of the $\ce{BH4}$ groups bond between the boron and uranium (a bridge bond). Two hydrogens from each of the two remaining $\ce{BH4}$ groups also bridge bond to uranium; the other two hydrogens bond to an adjacent uranium atom to build the polymer. So far this takes us to 10 hydrogens (6+4) bonding to the central uranium. Two more hydrogens in two $\ce{BH4}$ groups in two adjacent uraniums also bond to the initial uranium, just as described above, as part of the polymer extension. These 4 additional bonds to hydrogen take the coordination number to 14.

Coordination number 15 has been found in $\ce{Th(H3BNMe2BH3)4}$, and it is suggested that the molecule would have a coordination number of 16 in the gas phase (ref_2)

(image source: see ref_2 above)

There are 7 short $\ce{B-Th}$ contacts and one long $\ce{B-Th}$ contact. Each short contact provides 2 hydrogens that bond to thorium, while the long contact provides 1 hydrogen for bonding to thorium, making for a total of 15 $\ce{Th-H}$ bonds.

There has been quite some interesting work by my former co-worker and my supervisor on metal-rich molecule with the co-ordination higher than eight.

For more see Timo Bollermann, Thomas Cadenbach, Christian Gemel, Moritz von Hopffgarten, Gernot Frenking, Roland A. Fischer, Chem. Eur. J., 2010, 16 (45), 13372-13384. Also available through researchgate.com. For the purely theoretical bonding analysis see Moritz von Hopffgarten, Gernot Frenking, J. Phys. Chem. A, 2011, 115 (45), 12758–12768.

In october 2015, 5 months after all the current answers were given, a new cluster compound witn C.N. 16, $\ce{CoB16-}$, has been reported [1]:

Here we report the observation of a large metal-doped boron cluster of $\ce{CoB16-}$, which is produced using a laser vaporization cluster source and characterized by photoelectron spectroscopy (PES). Extensive computational searches reveal that there are two nearly degenerate structures for $\ce{CoB16-}$, which are indistinguishable at the highest level of theory employed. They both possess tubular double-ring framework and give similar photoelectron spectral patterns. The structures can be viewed as two $\ce{B8}$ rings sandwiching a $\ce{Co}$ atom, reminiscent of a drum and giving rise to the highest coordination number known in chemistry thus far. [...]

The previous highest coordination number known experimentally was 15 for $\ce{[Th(H3BNMe2BH3)4]}$ (ref. 33), though theoretical studies have suggested the highest coordination numbers of 15 in $\ce{PbHe15^2+}$ (ref. 34) and 16 in the Friauf–Laves phases in $\ce{MgZn2}$ or $\ce{MgNi2}$ (ref. 35). Endohedral fullerenes ($\ce{M@C60}$) have been observed 36,37, but the metal atom in those cases interacts with the $\ce{C60}$ shell primarily ionically and it does not stay in the centre of $\ce{C60}$.

It needs to be pointed out, that no crystal structure has yet been determined (as of 2017Q3).

### References

1. Popov, I. A.; Jian, T.; Lopez, G. V.; Boldyrev, A. I.; Wang, L.-S. Nature Communications 2015, 6, ncomms9654 DOI: 10.1038/ncomms9654. (Open Access)
• "no crystal structure has yet been determined (as of 2017Q3)." do you have any updates on this now that Q1 2018 is about to end? Thanks! Apr 16 '18 at 15:50

Maybe it's a "cheat", but coordination to ten atoms has been known since ferrocene. The iron is bonded symmetrically to two five-carbon rings. Variations with larger rings and thus more atoms bonded to the metal are also known, as in uranocene.

• With π-donor ligands it's not that trivial. Actual C.N. has little to do with the number of atoms in each ring, instead one has to determine the number of donated π-electron pairs. So, actual maximum estimated C.N. of ferrocene and uranocene are 6 and 10, respectively. Jan 4 '19 at 3:07

In metals with hexagonal or face-centered cubic packing, every atom has 12 closest neighbors. Libretext illustration