Groups and Periods

Students are taught that the periodic table is organised into columns called groups and rows called periods, as though the two axes were equivalent ways of slicing the same thing. They are not, and treating them as equivalent is the source of a lot of wrong predictions.

Knowing an element's group tells you a great deal about how it behaves. Knowing its period tells you almost nothing.

Sodium and chlorine sit in the same row, three squares apart. One is a soft metal that reacts vigorously with water; the other is a pale green gas that forms salts with metals. Nothing about being in period 3 makes them alike. Sodium and potassium sit in the same column, and everything you know about one transfers to the other.

Why a group works

The reason is the valence electrons. Every element in group 1 has one electron in its outermost shell; every element in group 17 has seven. Since chemical behaviour is almost entirely the behaviour of those outer electrons, elements with the same number of them do the same kinds of thing.

Crossing a period, that number changes at every step — one, two, three, and so on up to eight. The outer shell is the same shell throughout, which is why atomic size and nuclear pull vary smoothly, but the chemistry is transformed, because the number of electrons available for bonding is different at every square.

A group is therefore a family in a meaningful sense: its members react with the same partners, in the same ratios, forming compounds with matching formulas. A period is a survey of what happens as you progressively fill one shell, and its members are related by mechanism rather than by resemblance.

The names, and where they came from

Only some groups have names in general use, and the ones that do carry their chemistry in their etymology.

• Group 1, the alkali metals. From the Arabic al-qalī, the ashes of certain plants, which yield potassium carbonate. Their compounds were known long before the metals were isolated.
• Group 2, the alkaline earth metals. "Earth" was an old term for an oxide that resisted heating; these oxides are alkaline as well as refractory.
• Group 15, the pnictogens. From a Greek root meaning to choke — nitrogen suffocates, being breathable only when diluted with oxygen. The name was coined in the twentieth century and adopted by IUPAC comparatively recently.
• Group 16, the chalcogens. "Ore formers", because oxygen and sulfur account for the great majority of the mineral ores from which metals are extracted.
• Group 17, the halogens. "Salt formers". Every one of them combines with a metal to give a salt, and common salt is the archetype.
• Group 18, the noble gases. Noble in the sense used of the noble metals: aloof, disinclined to react with anything.

Groups 13 and 14 have proposed names — triels and tetrels — that almost nobody uses, and the d-block is called by its position rather than by a family name.

Reactivity goes down group 17 and up group 1

Here is a pattern that looks contradictory and is not, and it catches out anyone who has memorised trends without their reasons.

Group 1 gets more reactive as you go down. Lithium reacts with water steadily; sodium vigorously; cesium violently.

Group 17 gets less reactive as you go down. Fluorine is the most aggressive of all; chlorine is milder; iodine is comparatively sluggish.

Both follow from the same fact — the outermost electrons lie further from the nucleus in the heavier members — because the two groups want opposite things. An alkali metal reacts by losing an electron, so holding it more loosely makes the element more reactive. A halogen reacts by gaining one, so a more distant outer shell attracts the incoming electron less strongly and makes the element less reactive.

The general form of the rule is worth carrying: descending a group makes an element more metallic, and whether that counts as more or less reactive depends entirely on which direction its chemistry runs.

The first element of every group is the odd one

If a group were a perfect family, its topmost member would be a typical example. In practice the first element of each main group is the least typical, and the gap between it and the second member is larger than any other gap in the column.

• Lithium forms a nitride directly with nitrogen gas, which no other alkali metal does. Its carbonate decomposes on heating, unlike the rest of group 1, and lithium fluoride is poorly soluble where the other alkali fluorides dissolve freely.
• Beryllium forms a covalent chloride and an amphoteric oxide, where magnesium and calcium are firmly ionic and basic.
• Nitrogen exists as N₂, two atoms held by a triple bond among the strongest known. Phosphorus, directly below, forms P₄ tetrahedra with single bonds instead, because a third-period atom is too large for effective sideways orbital overlap.
• Fluorine has an anomalously weak bond in F₂ — about 158 kJ/mol, weaker than either Cl₂ at 242 or Br₂ at 193 — because its two atoms are so small that their lone pairs are forced close enough to repel each other significantly.

The common thread is size. Second-period atoms are much smaller than their neighbours below, so charge densities are higher, orbital overlap works differently, and the maximum number of atoms that can be crowded around them is lower. Where a question asks why an element does not fit its group, "it is the first member and it is very small" is the beginning of the answer more often than not.

The pattern is strong enough to produce sideways resemblances instead: lithium behaves more like magnesium than like sodium, in the diagonal relationships covered under periodic trends.

What a period does share

Rows are not meaningless. What they share is the shell being filled, which makes the pattern of change across them consistent from one period to the next.

The clearest expression is stoichiometric. The oxides across period 3 run Na₂O, MgO, Al₂O₃, SiO₂, P₄O₁₀, SO₃ and Cl₂O₇ — and the ratio of oxygen to the element climbs steadily, because the number of valence electrons available to bond with oxygen climbs from one to seven. Period 2 does the same thing with the same shape, one shell in.

That is what a period offers: not similarity, but a repeatable progression. Two elements in the same period are related the way two frames of the same film are related, and the periodic table works precisely because the film runs again in the next row.

In the d-block, the rows are more alike than the columns

The generalisation reverses for the transition metals, and this is worth knowing because it makes nonsense of trying to apply main-group habits there.

Iron, cobalt and nickel are horizontal neighbours and are strikingly similar — all three ferromagnetic, all three forming +2 and +3 ions, all three used as catalysts. Iron, ruthenium and osmium are vertical neighbours in group 8, and they are not remotely alike: one is the commonest structural metal on Earth, the other two are rare platinum-group metals whose highest oxides are volatile.

The reason is that d-block elements bond largely with their outer s electrons, which do not change in number across the row, while the d electrons that do change are buried one shell in. Horizontal neighbours therefore end up with similar valence behaviour.

The f-block takes this to its limit. The lanthanides differ only in how many electrons occupy a 4f subshell that is shielded from the outside world almost completely, so all fourteen behave nearly identically and nearly all form 3+ ions. That similarity is why they were discovered so late and, once found, took over a century to separate cleanly.