# Is calcium carbide an ionic or covalent compound?

We know that the bond which is formed by exchanging electrons between a metal and non-metal atom is called ionic bond. Calcium is a alkali earth metal and carbon is a non-metal. So, calcium carbide ($\ce{CaC2}$) should be an ionic bond. But it is not an ionic bond, why?

• strongly related: chemistry.stackexchange.com/questions/5227/… – Nilay Ghosh Nov 12 '17 at 5:55
• Because you said so. There is no other reason. – Ivan Neretin Nov 12 '17 at 6:04
• What's to understand? It is not ionic because you said so. Before that, it was ionic all right. – Ivan Neretin Nov 12 '17 at 7:07
• How am I to know the reason behind your words? As to why is it ionic, you just explained it yourself in your question. – Ivan Neretin Nov 12 '17 at 7:22
• @matt_black There is an issue with every single established fact in physics and chemistry. Every rule has either exceptions, or some subtle effects that are only visible at 9-digit precision, or some non-obvious aspect that would make it look entirely different, or something like that. Really, open a schoolbook and read a few rules at random; I bet you'll easily point out some exceptions or corrections to each of them. Some time ago I decided against delving into the subtleties too early. Simple things first. It is ionic, period. If someone wants to know more, there's a link by Nilay. – Ivan Neretin Nov 12 '17 at 12:03

Myth debunking time!

First, there is no such thing as a purely ionic bond and no such thing as a purely covalent bond. Each bond (yes, even the bond in $\ce{H2}$) has a partially covalent and a partially ionic character. However, it is true that we can classify bonds and interactions as predominantly covalent or predominantly ionic in many cases especially those that occur at a beginner’s level.

While it is a nice explanation for the creation of e.g. sodium chloride to think of a full electron transfer — not an exchange because only one electron per ion pair is transferred and only in one direction — this cannot explain the structures of e.g. sodium sulphate well. Instead, you should think of ionic compounds as those composed of differently charged ions between which the interactions are predominantly electrostatic (non-directional). This is not true for all ionic compounds (ammonium fluoride being an exception) but gets you very far.

Now that I have established that (those that can be classified as predominantly) ionic compounds are composed of charged particles that interact electrostatically, you may be able to understand that not each ionic compound consists of exactly one atom type as the anion and exactly one atom type as the cation. In fact, I want to postulate that the vast majority of ionic compounds do not follow this rule. These compounds contain polyatomic ions meaning that many atoms connected together by covalent bonds form one charged molecule (or molecular ion). One typical example is the sulphate anion $\ce{SO4^2-}$ that I mentioned earlier but also the ammonium cation $\ce{NH4+}$. Thus, a compound can be held together by both covalent and ionic bonds, if for example it is composed of molecular ions.

This is the case in calcium carbide. We do indeed have a calcium cation $\ce{Ca^2+}$. The counterion is a diatomic anion, the ethynediide anion $\ce{C2^2-}$. It consists of two carbon atoms connected by a triple bond with each carbon atom carrying an additional lone pair and thus formal negative charge for an overall charge of $2-$. The bond between the two carbon atoms is covalent. However, the bond between calcium and the $\ce{C2^2-}$ fragment is ionic.

Therefore, calcium carbide can be classified as an ionic compound.

If you dig deeper, you may come across the crystal structure of calcium carbide; see for example the following taken from Wikipedia:

(This structure actually depicts barium peroxide but calcium carbide crystallises in the same structure.)

The interatomic distances given by Atoji and Medrud[1] are: $$\begin{array}{lccc}\hline \text{atoms} & \ce{C#C} & \ce{C-Ca}\text{ (lin)} & \ce{C-Ca}\text{ (orth)}\\ \hline d/\mathrm{pm} & 120 & 259 & 282 \\ \hline\end{array}$$

The linear $\ce{(C#)C-Ca}$ distances are actually short enough to also consider a covalent interaction. Other acetylides can also be characterised with an intermediate or predominantly covalent bond. The picture becomes much more fuzzy once you take a closer look. The interaction of calcium with the $\ce{C2}$ fragment’s π system should also be considered as a potentially bonding interaction; think in the direction of transition metal complexes.

However, the similarity of both $\ce{C-Ca}$ distances (which becomes even more similar if you take the distance from calcium to the centre of the $\ce{C2}$ fragment) can also be seen as support for the argument of non-directional ionic bonding.

Considering all this experimental data, it is clear that the compound should have significant covalent bonding. Significant ionic bonding is not out of the question. The bonds are likely somewhere in the middle between covalent and ionic. Whatever the ‘final result’ e.g. of calculations is, it will not tip the scales strongly.

Reference:

[1]: M. Atoji, C. Medrud, J. Chem. Phys. 1959, 31, 332. DOI: 10.1063/1.1730352.

• Well, this looks rather like repeating high-school illusions then "Myth debunking". – Mithoron Nov 12 '17 at 20:22
• Starts like debunking, but "we can classify bonds and interactions as predominantly covalent or predominantly ionic in many cases — practically all cases that occur at a beginner’s level." is well... Most compounds of transition metals and even ones like very CaC2 aren't obvious at all, or are predominantly covalent. – Mithoron Nov 13 '17 at 1:41
• @Mithoron Not sure if I classify coordination compounds as truly beginner’s level, though. – Jan Nov 13 '17 at 1:44
• HgS, Au2S - very covalent and rather "beginner". – Mithoron Nov 13 '17 at 1:48