Many groups of bound atoms complexively bearing a (negative) charge are called complex anions. Examples are $\ce{NO3-}$, $\ce{CO3^2-}$, $\ce{BO3^2-}$, $\ce{CrO4^2-}$, $\ce{Cr2O7^2-}$, $\ce{MoO4^2-}$, $\ce{WO4^2-}$, $\ce{PO4^3-}$, $\ce{AsO4^3-}$, $\ce{VO4^3-}$ and the several, sometimes polymeric, complex anions occurring in silicate and borate minerals.

I apologise if I am asking something trivial, but I cannot find a source, except for silicates, for which I read that the $\ce{Si-O}$ bonds are basically covalent, explaining whether the bonds within complex anions are covalent in general.

Are they always covalent? If they are not covalent in general, are they in the examples I have written?

  • 3
    $\begingroup$ All these anions have polarised covalent bonds. $\endgroup$ – Mithoron Jan 18 '15 at 3:15
  • 1
    $\begingroup$ The notion of the mentioned anions being "complex anions is that they don't dissociate if dissolved in water. Hence they have covalent bonds not ionic ones. $\endgroup$ – MaxW Sep 22 '17 at 14:57

All complex anions you noted that include only non-metal atoms are always considered covalent. Except for a few extreme examples including fluorine and hydrogen, non-metals only form predominantly covalent bonds between themselves.

It gets a little more tricky when considering those anions whose central atom is a transition metal. In this case at first approximation, we can consider the anion a coordination compound and the coordinate bonds within coordination compounds are typically considered more covalent. Moving on to a second approximation clarifies the picture: the formal oxido-ligands are in fact very strongly attached to the metal and do not undergo ligand substitution. This means that they are even more covalent than other coordination complexes.

Therefore, it is typically safe to consider the bonds within complex anions to be predominantly covalent.


"Electronegativity" is a measure of the power of an atom in a molecule to attract shared electrons to itself. If you look up the electronegativities for the atoms on either end of a bond, there's a quick and dirty way to decide for yourself how much covalent character the bonds have.

Calculate the electronegativity difference between the central atom and oxygen. The smaller the difference, the more covalent character the bond has. For example, for Si-O, Si and O have Pauling electronegativities of 1.8 and 3.5, respectively, so the electronegativity difference is 1.7.

Some textbooks set (rather arbitrary) thresholds for electronegativity differences to help general chemistry students classify bonds as non-polar covalent, polar covalent, or ionic. Pauling's original text says that if the electronegativity difference is 1.7, the bond has about 50% ionic character, so you can say an Si-O bond is about 50% ionic, 50% covalent. If the electronegativity difference is greater than 1.7, you can call it "predominantly ionic"; if it's less than 1.7, you can call it "predominantly covalent".

  • $\begingroup$ Therefore, if i correctly understand, the only ionicly bound complex anions would be those with Cr-O and V-O, having chromium and vanadium respectively electronegativity 1.66 and 1.63 versus the 1.8 of oxygen. Then we observe that $\text{CrO}_4$, $\text{Cr}_2\text{O}_7$ and $\text{VO}_4$ "get some electrons more" as a whole because they are anions. Thank you very much!!! $\endgroup$ – Self-teaching worker Jan 17 '15 at 23:54
  • 1
    $\begingroup$ You're welcome! But oxygen's Pauling electronegativity is 3.5 (a typo in your note?) And it isn't black and white covalent or ionic, if you have an electronegativity difference of 1.8, that means about 55% ionic, 45% covalent. $\endgroup$ – Fred Senese Jan 18 '15 at 0:02
  • $\begingroup$ Yes, a typo in my comment, I should have said "1.8 of silicon" and, if an element is less electronegative than Si, its bonds with O are more ionic than covalent. Thank you again!!! $\endgroup$ – Self-teaching worker Jan 18 '15 at 0:06
  • 1
    $\begingroup$ @Fred Senese I'm afraid that your answer is too long or too short, but definitely misleading - atoms in different oxidation states exhibit different "effective" electronegativities. Also Si has "standard" Pauling electronegativity 1.90 and oxygen 3.44. $\endgroup$ – Mithoron Jan 18 '15 at 3:13
  • 1
    $\begingroup$ @Mithoron, well, there are many different electronegativity scales, refinements, and corrections that could be used. I thought it best to answer the question at the genchem level, and I used Pauling's (original) electronegativity values in my answer, and I followed his approach as well. $\endgroup$ – Fred Senese Jan 18 '15 at 6:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.