Ionic, Covalent & Metallic Bonding

Why atoms bond at all, and three ways to do it

You climbed the first rung knowing how a single atom is built and why the periodic table arranges the elements the way it does. Now the obvious next question: once two atoms meet, what makes them stick? The short answer is energy. A pair of bonded atoms sits at lower total energy than the two atoms apart — the electrons get to feel the pull of two nuclei instead of one, or to leave a half-empty shell and reach a stable closed one. Everything in this guide is just three different physical answers to the same question of how atoms lower their energy together.

There are three classic, idealized pictures. In an ionic bond, one atom hands an electron over completely; the two atoms become oppositely charged ions and the bond is the electrostatic pull between them — think NaCl. In a covalent bond, neither atom will let go, so they compromise and share a pair of electrons sitting between the two nuclei — think H2 or N2. In a metallic bond, a whole crowd of atoms pool their outer electrons into a shared sea of mobile electrons that washes over fixed positive cores — think a block of copper. The crucial honesty up front: these three are limiting cases, the clean corners of a map. Almost every real bond lives somewhere inside that map, not exactly at a corner.

A recurring theme drives most of this: main-group atoms tend to bond until they are surrounded by eight valence electrons, the stable count of the noble gases — the octet rule. Ionic bonding reaches it by transfer (Na empties a shell, Cl fills one); covalent bonding reaches it by sharing (each shared pair counts for both atoms). It is a powerful default, not a law — boron settles for six and sulfur in SF6 holds more — but it is the bookkeeping behind why these bonds form at all.

Electronegativity: the dial that decides ionic versus covalent

What decides whether two atoms transfer an electron or share one? Mostly a single property you already met on the first rung: electronegativity, an atom's pull on the electrons in a bond. If both atoms pull equally hard — two identical atoms, as in Cl2 — the pair is shared dead evenly: a pure covalent bond. If one atom pulls much harder than the other, it drags the shared pair so far over that the bond is, for practical purposes, a transfer: ionic. In between, the sharing is just lopsided, and we call that a polar covalent bond, the shared pair pulled toward the greedier atom.

So the single most useful number to look at is the electronegativity difference, delta-EN, between the two atoms. Roughly: a very small delta-EN (say below about 0.5 on the Pauling scale) means an essentially nonpolar covalent bond; a moderate delta-EN (about 0.5 to 1.7) means a polar covalent bond; a large delta-EN (above about 1.7) means the bond is mostly ionic. Treat these cutoffs as soft guideposts, not hard walls — the boundaries are blurry on purpose, because the truth is a continuum. The same delta-EN that tells you HF is strongly polar (about 1.9) and yet still molecular, not a salt, is what makes the simple guideposts only approximately right.

the bonding triangle (van Arkel-Ketelaar)

            IONIC
            /    \
           /      \   (large delta-EN, one atom
          /        \   far more electronegative)
         /          \
  METALLIC ---------- COVALENT
  (both low EN,        (both high EN, similar:
   electrons pooled)    delta-EN ~ 0)

  average EN  ->  rises left-to-right along the base
  delta-EN    ->  rises toward the top apex

That triangle is worth keeping in your head: the base runs from metallic on the left to covalent on the right as the average electronegativity rises (low-EN elements pool electrons; high-EN elements share or grab them), while climbing toward the top apex means a growing difference and so growing ionic character. Two cesium atoms (low average EN, zero difference) sit at the metallic corner; two fluorines (high average EN, zero difference) sit at the covalent corner; cesium plus fluorine (one very low, one very high) climbs to the ionic apex. One picture, all three bond types, no sharp lines.

Blurring the line: Fajans' rules and ion polarization

Electronegativity gets you most of the way, but it leaves a puzzle. Both NaCl and AgCl are 1+/1- chlorides with similar delta-EN, so the simple rule says both should be cleanly ionic. Yet AgCl is far less soluble, lower-melting, and behaves much more covalently. To explain that, we look at the bond from the other end of the continuum: start from an ideal ionic lattice and ask how much covalent character creeps back in. The mechanism is ion polarization — a small, strongly positive cation tugs on the soft electron cloud of its anion neighbor, pulling some of that cloud into the space between them. Electrons sitting between two nuclei is exactly what a covalent bond is, so the more the cloud is distorted, the more covalent the nominally ionic bond becomes.

Fajans' rules are the handy summary, from the 1920s, of when this polarization is large. Covalent character rises when the cation is small and highly charged (a concentrated, intense pull — high polarizing power), when the anion is large and highly charged (its loose outer electrons are easily distorted — high polarizability), and when the cation lacks a noble-gas core. That last point cracks the AgCl puzzle: Ag+ is a d10 ion, and a d10 shell shields the nucleus poorly, so Ag+ polarizes far more strongly than a noble-gas-core cation like Na+ of similar size and charge. Hence Fajans' rules correctly predict AgCl to be much more covalent than NaCl, with everything that follows from it.

These rules quietly rationalize a lot of descriptive chemistry. Across AlF3, AlCl3, AlBr3, AlI3 the cation Al3+ is fixed but the anion grows softer and more polarizable, so covalent character climbs and melting points fall — AlF3 is a high-melting ionic solid while AlI3 is a low-melting covalent one. And the small, charge-dense Li+ and Be2+ are anomalously covalent for the s-block, which is one face of the diagonal relationship you saw on the first rung — lithium resembling magnesium, beryllium resembling aluminum. Just remember these are qualitative directions, not numbers: Fajans' rules tell you which way covalent character moves, never exactly how much.

The third corner: a sea of delocalized electrons

Ionic and covalent both keep electrons localized — pinned to an ion or parked between two nuclei. The third corner does something completely different. In a metal, the atoms have low electronegativity and few valence electrons that they hold loosely, so when millions of them pack together, no atom can win a tug-of-war for any particular electron. Instead they all give up. Each atom releases its valence electrons into a shared pool, and what is left behind is a regular array of positive ion cores bathed in a sea of delocalized electrons belonging to the whole piece at once, not to any single bond. The attraction between the positive cores and that negative sea is the metallic bond.

Almost every familiar property of metals falls straight out of this one picture. The electrons are not tied to any bond, so the ion cores can slide past one another without snapping anything — metals are malleable and ductile, where an ionic crystal would crack. The electrons drift freely under a voltage, so metals conduct electricity; they carry energy fast, so metals conduct heat; and they reflect light, so polished metals are lustrous. Bond strength varies enormously with how many electrons each atom donates and how tightly the cores grip them: mercury is liquid at room temperature, while tungsten melts above 3400 degrees Celsius.

Predicting the dominant bond type from the elements

Here is the payoff: given a formula, you can place its bonding on the triangle just by looking at where the elements sit in the periodic table. The fastest crude cut is metal-versus-nonmetal. Metal plus nonmetal (low-EN plus high-EN, large delta-EN) leans ionic — NaCl, MgO, CaF2. Nonmetal plus nonmetal (both high-EN, small delta-EN) leans covalent — CO2, NH3, SiO2. Metal plus metal (both low-EN) is metallic — bronze, brass, steel. The same low-EN-versus-high-EN logic that built the triangle is doing all the work here.

1. Identify the elements and roughly where they sit: metals are low-EN (left and bottom of the table), nonmetals are high-EN (top right). A metalloid in between will hedge its bets.
2. Estimate delta-EN. Near zero with both atoms high-EN means covalent; near zero with both low-EN means metallic; large means lean ionic. The cutoffs are soft, so do not over-trust a single number near a boundary.
3. If you called it ionic, apply Fajans' rules as a correction: a small, highly charged, or non-noble-gas-core cation, plus a large soft anion, drags real covalent character back in.
4. Sanity-check against properties: hard, brittle, high-melting, conducts only when molten = ionic; discrete molecules or low-melting solids = covalent; lustrous, malleable, conducts as a solid = metallic.