From a Bond Length to a Radius
In the last guide you met the ionic model: picture a crystal as a packing of charged spheres, positive cations and negative anions held together by their opposite charges. If ions really were spheres, then the cation-to-anion distance an X-ray machine measures should just be the sum of two radii, one for each ion. The wonderful surprise is that bond lengths behave almost exactly that way — the magnesium-to-oxygen distance stays near 2.1 Å in MgO, in magnesium silicate, and in Mg(OH)2 alike. That additivity is what lets us hand every ion its own effective ionic radius.
There is a catch worth stating plainly. Experiment only ever gives you the sum — the bond length. How to split 2.1 Å into 'so much cation, so much anion' is a matter of convention, not measurement. Pauling, Goldschmidt, and Shannon each anchored their scale a little differently, most often by fixing the oxygen ion near 1.40 Å and letting the rest follow. So an ionic radius is a bookkeeping figure, not a hard physical edge, and different tables can differ by a few pm. Throughout we use Shannon's widely adopted set. (Units reminder: 1 Å, one ångström, equals 100 pm equals 0.1 nm equals 10^-10 m.)
Why Cations Shrink and Anions Swell
Turn a neutral atom into a cation by stripping away electrons — often the whole outer shell simply disappears. Now the same nuclear charge pulls on fewer electrons, so each one is held more tightly and the cloud tightens. A cation is therefore always smaller than its parent atom, sometimes dramatically: a sodium atom is about 1.9 Å in radius, but Na+ is only about 1.0 Å. An anion does the opposite. Adding electrons piles on electron-electron repulsion without adding any nuclear charge, so the cloud puffs outward — a bare oxygen atom is roughly 0.7 Å, while O2- swells to about 1.4 Å, twice as large.
The cleanest way to see the trend is an isoelectronic series — a set of ions that all carry the exact same number of electrons. Take five ions with ten electrons each, the configuration of neon: O2-, F-, Na+, Mg2+, and Al3+. The electron cloud is, in principle, the same size, but the number of protons climbs from 8 up to 13. More protons pulling on the same cloud squeeze it down step by step.
Isoelectronic series -- every ion has 10 electrons (neon core) ion protons effective ionic radius, CN 6 ----- ------- ---------------------------- O2- 8 140 pm <-- anion, largest F- 9 133 pm Na+ 11 102 pm <-- big drop crossing to a cation Mg2+ 12 72 pm Al3+ 13 53.5 pm <-- smallest same-size electron cloud + more protons = smaller ion
Two things jump out. First, the big fall from F- at 133 pm to Na+ at 102 pm: the moment an ion goes from net-negative to net-positive, its grip on the electrons tightens and it shrinks sharply. Second, Al3+ at just 53.5 pm is barely more than a third the size of O2-. This is the deep reason most oxide ceramics are built as a close-packed array of big oxygen anions with the small cations tucked into the gaps between them — a picture the next rung will lean on heavily.
More Charge, Same Element, Smaller Ion
Fix the element and change only its charge, and the rule holds: more positive charge means a smaller ion. Pull off one more electron and the survivors are held tighter still. Iron is the classic case — Fe2+ measures about 0.78 Å, but Fe3+ shrinks to about 0.645 Å, roughly 17 percent smaller. This is the mirror image of a rule from the previous guide: a higher-charge ion pulls shorter, stronger bonds — and here we see that the same higher charge also makes the ion itself count as smaller.
The other knob is which row of the periodic table you are on. Going down a group adds a whole new electron shell, so the ion grows: Mg2+ is 0.72 Å, Ca2+ is 1.00 Å, and Ba2+ is a hefty 1.35 Å — each step down the column wraps on another shell. So far, then, two things fix an ion's size: how many electrons it has kept (its charge and its group), and — coming next — how many neighbours crowd around it.
Neighbours Change the Size, Too
Here is the twist that trips up newcomers. The very same ion, at the very same charge, has a different radius depending on its coordination number — the count of nearest neighbours it holds. Crowd more anions around a cation and those anions repel one another, prying every bond a little longer; to keep the radii additive, the cation's tabulated radius has to grow as its coordination rises. Size is not a fixed property of the ion alone — it depends on the company it keeps.
The numbers make it concrete. Na+ is about 0.99 Å with 4 neighbours, 1.02 Å with 6, and 1.18 Å with 8. Zr4+ swells from 0.72 Å in six-fold coordination to 0.84 Å in eight-fold. Nothing about the ion changed except the size of the crowd. This is exactly why a good table like Shannon's lists a separate radius for each coordination number, and why quoting an ionic radius without its coordination is meaningless.
Far from being a nuisance, this coupling is the whole plot. Because the ratio of the cation radius to the anion radius decides how many anions can physically pack around a cation, ion size and coordination lock into a feedback loop. Turning that ratio into a prediction of coordination is precisely the radius-ratio rule of the next guide — for a quick taste, Mg2+ over O2- is 0.72 / 1.40, about 0.51, which lands each Mg2+ in a six-neighbour octahedral hole and gives MgO the rock-salt structure.
Ionic Radius, Covalent Radius, and Why Size Runs the Show
When a bond is shared rather than transferred — the covalent end of the ionic-covalent spectrum from the last guide — we switch to a covalent radius, defined as half the bond length between two identical atoms. The carbon-carbon bond in diamond is 1.54 Å, so carbon's covalent radius is 0.77 Å; silicon's works out to about 1.11 Å. Add the two and you predict the silicon-carbon bond in SiC at about 1.88 Å — which is exactly what it measures.
The same atom can wear wildly different 'radii' depending on how it bonds. Silicon as a covalent atom is 1.11 Å, but as the ion Si4+ in four-fold coordination it is a tiny 0.26 Å. Which do you use? It depends where the bond sits on the spectrum. Here is the honest part: the Si-O bond in SiO2 is only about half ionic, yet Si4+ (0.26) plus O2- (1.40) gives 1.66 Å, remarkably close to the real 1.62 Å — the ionic bookkeeping works even for a half-covalent bond. Just never mistake the radius for a fixed fact of nature. And a big, soft anion sitting beside a small, high-charge cation gets its cloud pulled out of shape, nudging the bond toward covalent — the same polarisation idea you met a guide ago.
- Start from the neutral atom and adjust for charge: losing electrons (a cation) makes it smaller; gaining electrons (an anion) makes it larger.
- Within one element, more positive charge means a smaller ion; moving down a group adds electron shells and makes it larger.
- For an isoelectronic set (same electron count), more protons means a smaller ion.
- Choose the radius listed for the right coordination number — a larger coordination number means a larger tabulated radius.
- Stay on one consistent scale, such as Shannon's, and remember that only the bond-length sum is truly measured.
Why does all this careful sizing matter? Two of the biggest facts about any ceramic fall straight out of it. First, the cation-to-anion radius ratio sets the coordination and therefore the crystal structure — the entire subject of the next guide. Second, small, highly charged ions form short, strong bonds, and short strong bonds mean high melting points and great hardness: MgO, with its small Mg2+, melts near 2850 degrees C, and alumina near 2054 degrees C. Get the ion sizes right and the rest of ceramic crystal chemistry clicks into place — which is exactly where this rung is heading.