# How many bonds does oxygen really make in minerals?

I'm studying some metallurgy things and interested in iron oxides.

The simple 2-D representation of a mineral like hematite shows $$\ce{O=Fe-O-Fe=O}$$.

But the crystal structure seems to show each oxygen making 4 bonds. Crystalline structure from Virtual museum of molecules and minerals[1]

So whats the deal here?

Reference:

[1]: Hematite – virtual museum of molecules and minerals
https://virtual-museum.soils.wisc.edu/display/hematite/
(accessed Jun 16, 2021).

This is a human.

Well, sort of.

All right, call it a simplified representation. But at least it shows the arms/legs ratio accurately, you can't deny that.

Now that was about as good as the molecular formula of $$\ce{Fe2O3}$$ (a compound which has no molecules in the first place):

As for the crystal structure image, that's what $$\ce{Fe2O3}$$ actually is. The sticks are drawn where the two atoms touch. Call them bonds or otherwise, the Nature does not care.

When you learn about the coordination compounds, the idea of atoms making far more bonds than their formal valence allows won't look that weird.

So it goes.

• thanks for the answer. As im interested in the redox reactions taking place in these crystals with CO and H2 reduction, I'm really curious about the electrons being shared. Commented Jun 17, 2021 at 9:12
• Reactions don't take place in the crystals. On the surface, maybe. Commented Jun 17, 2021 at 9:31
• @NilayGhosh Still not as cursed as this $\ce{O=Fe-O-Fe=O}$ abomination. Commented Jun 17, 2021 at 9:35

There is no relation between the formula of a substance and the number of neighbors any given atom has in its vicinity. For example, each sodium cation $$\ce{Na+}$$ is surrounded by $$6$$ $$\ce{Cl-}$$ ions in the crystal of salt, although its formula is $$\ce{NaCl}$$. Each sodium atom has $$6$$ neighbors, but it stays at oxidation number $$+1$$.

• So the crystal structure is not showing bonds, what is it showing? Commented Jun 16, 2021 at 19:06
• It is showing the local nearest neighbours for each atom. Don't think of ionic bonding as occurring pairwise. As coulomb interactions are very long ranged an ion sits in the electric potential due to the whole of the lattice, and so is interacting with every other ion in the lattice. A bonds to B in such a situation is just not a good model. Commented Jun 16, 2021 at 19:18