Period 4

Period 4 is where the periodic table gets wide. The two rows before it hold eight elements each; this one holds eighteen, and the extra ten arrive in a block wedged between calcium and gallium that has no counterpart above it. Everything distinctive about the row comes from that insertion — where it comes from, what it does to the trends running across the row, and what it does to the row below.

Why the row jumps from eight to eighteen

The 3d orbitals came into existence one row earlier, when n reached 3, and stayed empty. They fill here because of the order in which orbital energies actually run in a many-electron atom, which is not the order of the principal quantum number.

That ordering has a name and a full statement, along with the tie-break it needs and the elements it gets wrong, on the electron configuration page. The part of it that matters here is one comparison. Add the two quantum numbers together and 4s scores 4 while 3d scores 5, which puts 4s first — so potassium and calcium spend their new electrons on 4s even though a 3d orbital is nominally closer in, and scandium does not open the 3d series until 4s is already occupied.

The underlying reason is penetration. A 4s orbital has a small but real probability density close in to the nucleus, inside the argon core, where it experiences almost the full nuclear charge. A 3d orbital does not penetrate that far; it is held out beyond the core and is screened more effectively. Being more diffuse but better positioned beats being compact and shielded, by a margin that is narrow — which is why the ordering is fragile enough to flip.

It fills first and it empties first

And flip it does. Once the 3d subshell has electrons in it, the added nuclear charge stabilises those d electrons more than it stabilises the 4s electrons, and 3d drops below 4s in energy. When a transition metal atom is ionised, the electrons removed are the 4s ones.

The iron(II) ion is therefore [Ar]3d⁶, not [Ar]3d⁴4s². The vanadium(III) ion is 3d², not 3d⁰4s² minus something. This is the most reliably failed question in the whole of electron configuration, and it is failed because "fills first, empties first" sounds like it ought to be a rule.

It is worth being precise about why there is no contradiction. Orbital energies are not fixed properties of an element; they depend on the whole configuration and on the charge. A neutral potassium atom and an Fe²⁺ ion are different systems, and there is no reason the same ordering should hold for both. The apparently paradoxical statement is only paradoxical if orbitals are imagined as fixed shelves that atoms load in sequence.

Two of the ten 3d elements do not follow the plain filling order at all, chromium and copper each taking a single 4s electron rather than two. Two exceptions in ten is the best behaviour any d series manages — the 4d row in period 5 departs from the pattern far more often, for the same reason this row's ordering is fragile.

Ten elements that barely change

Crossing period 2 or period 3 changes an element's character completely, from an alkali metal to a noble gas in eight steps. Crossing the d-block changes almost nothing.

Atomic radii from titanium through copper vary by only a handful of picometres. Melting points stay high throughout. All ten are dense, hard, lustrous metals, and they alloy with one another in almost any proportion. The reason is that each added electron goes into an inner subshell that partly shields the added proton, so the effective nuclear charge experienced by the outermost 4s electrons creeps up instead of climbing.

That flatness is why the row's chemistry is dominated by a different variable: not position, but oxidation state. What distinguishes manganese from iron in practice is less their size than the range of charges they can hold, which is treated on the transition metals page rather than here.

Biologically, the block is remarkably productive. Chromium through zinc — seven consecutive elements — all have established roles in human biochemistry, a hit rate no other stretch of the table approaches. They were abundant and soluble in the early ocean, and their capacity to sit in more than one oxidation state is exactly what an enzyme active site needs.

Gallium is not bigger than aluminum

The d-block leaves a mark on the elements that come after it. Gallium sits directly below aluminum and should, by the normal group trend, be appreciably larger. It is not: their covalent radii are almost identical, gallium's first ionisation energy is fractionally the higher of the two, and gallium is distinctly the more electronegative.

Ten additional protons have been added between calcium and gallium, and the ten 3d electrons that came with them shield those protons poorly. The result — sometimes called the d-block or scandide contraction — is that the whole post-transition part of period 4 is pulled in tighter than the trend from period 3 would predict. Germanium, arsenic, selenium and bromine all inherit it.

Once past the block the row resumes ordinary behaviour: gallium and germanium are followed by arsenic and selenium, then bromine, then krypton closing the period at 4s²4p⁶. The metal-to-nonmetal walk that period 3 completes in eight steps happens here too, but it is interrupted in the middle by ten elements that are all metals and all much the same.

The row where fusion stops paying

Period 4 contains the point at which stars run out of profitable reactions, and it is worth stating carefully because the version usually repeated is slightly wrong.

Fusing light nuclei releases energy because the product is more tightly bound per nucleon than the reactants. That binding energy per nucleon rises steeply through the light elements and then flattens into a broad maximum in this row — the iron peak. Beyond it, fusion costs energy rather than yielding it, so a massive star that has built an iron core has no further fuel and collapses.

The nucleus with the single highest binding energy per nucleon is nickel-62, at about 8.79 MeV, with iron-58 and iron-56 immediately behind it and the differences in the third decimal place. Iron-56 is nonetheless the most abundant product, partly because the alpha-particle chain that builds heavy nuclei in a star arrives naturally at nickel-56, which then decays to iron-56, and partly because the photodisintegration equilibrium at those temperatures favours it. So "iron is the most tightly bound nucleus" is a good approximation to a more interesting truth, and the elements immediately around it in this row are all part of the same peak.

One consequence is that every iron atom in the crust, in haemoglobin and in a steel beam was made either in the last moments of a massive star or in the merger of two neutron stars, and none of it was made by ordinary stellar burning after the fact. Another is that the row opens with a long-lived radioactive isotope: potassium-40, about one part in eight thousand of natural potassium, and slow enough to halve only once in something over a billion years. It is the largest single source of radioactivity inside an ordinary human body, which is a consequence of potassium being the dominant cation inside cells rather than of anything unusual about the isotope.