LearnChem · Part II — Why do atoms stick together?
Line up a column of the table and you get a family photograph: the same face, five times over, growing more extreme down the rows. Two dials — size and pull — explain the resemblance. And the one member who does not fit.
“Why is sodium something that explodes in water, when we eat it every day?”
Last time, each glue left its fingerprint — the sea's shine, the chessboard's cleave, the network's hold. But one thing was left as brute fact: which glue an element ends up held by. Why does sodium shed an electron into a sea for almost nothing, while chlorine's electrons sit at full price? Read down the table's columns and the same outcomes repeat, row after row — families of elements, each with a personality. This module reads the columns — and meets the member who breaks the family rules.
Take the table's left-hand column, below hydrogen: lithium, sodium, potassium, rubidium, caesium. Five metals, and one description covers them all. Soft enough to cut with a knife — a fresh cut gleams, then dulls in front of you as the air gets to work. Too quick to combine to be found alone in the ground, ever. And every one of them does the same thing to water: tears it apart, sets hydrogen free, and leaves the water alkaline — which is where the family name comes from, the alkali metals. (Hydrogen, printed above them, is not a member: one electron, yes, but a gas of shared pairs, not a metal — a family of one.)
The reaction is one sentence of bookkeeping — 2 Na(s) + 2 H2O(l) → 2 NaOH(aq) + H2(g) — and the family performs it in order. Lithium fizzes and skates. Sodium dashes about as a molten bead and may flare. Potassium ignites the instant it lands, burning lilac — the flame carries potassium's own line-spectrum, module 04's fingerprint. Rubidium and caesium are not demonstrated on an open bench at all. Same family, same one sentence of chemistry — performed each row down with more violence.
Lithium fizzes in water. Sodium, one row down, dashes and may flame. Potassium ignites at once. Same family, same reaction. Commit — why does the violence grow down the column?
Commit before you look. First answers are counted anonymously, never named.
What makes them a family is one number: each holds exactly one outer electron, alone on its atom's top floor. Module 05 priced that electron — the farthest out, the best screened, the cheapest thing the atom owns. And module 06 showed what one cheap electron buys: hand it to a strong enough puller and the ledger closes well in the black. All five metals are that same transaction walking about in different sizes.
The violence, though, is the size speaking. Each row down adds a floor, so the one electron starts farther from the nucleus, with a thicker cushion of inner floors screening the pull. Farther and better screened means cheaper to take — and the first step of the water reaction is exactly that removal. Caesium's electron is the family's cheapest; caesium is the family's most violent. The family resemblance is one electron; the family gradient is the price of letting it go. One honesty note, kept short: the growing violence is a matter of speed, not of total payout — lithium's tiny ion actually collects the biggest hydration payoff in the family, §3's trick in miniature, and modules 12 and 21 settle that ledger properly. What falls down the column is the price of the first, slowest step. And the melting points read the same way. Lithium holds solid to 180.5 °C; sodium lets go at 97.8, potassium at 63.5, rubidium at 39.3. Caesium, at 28.5, would melt in a hand — if anyone were unwise enough to try. The sea gets shallower as the atoms get bigger: module 09's ladder, read straight down a family.
And here the question at the door gets its answer. The sodium you ate today is not this metal. Table salt holds the ion — sodium after the transaction, its one cheaply-taken electron already handed over, the violent step already paid, ages ago, wherever that salt's sodium last met a strong puller. The metal is an atom still carrying its lit fuse; the ion is the same atom after the fuse has burned. And the ion has nothing cheap left to offer: module 05's staircase leaps exactly here — the next electron priced nine times dearer — so water finds no affordable transaction at all. Chlorine plays the identical trick on the other side: the gas that was used as a weapon becomes, once it has taken its electron, the placid chloride half of the same white crystal. Nobody keeps sodium metal at home — and every kitchen keeps its ion by the kilogram.
Quick check — sodium metal explodes in water, and you ate sodium today. How is table salt safe?
Alkali metals in contact with water release hydrogen with enough heat that it can ignite — the UK schools' safety sheets (CLEAPSS SSS 80) put lithium, sodium and potassium under teacher-only demonstration, behind screens, with rubidium and caesium not allowed in schools at all. This course never asks you to handle any of them. Watch the transaction done properly instead: the University of Nottingham's sodium and water in slow motion (Periodic Videos) — the molten bead, the hydrogen flame, the whole §1 argument in half a minute of footage.
One family down. But the claim on the door was bigger: any family's behaviour comes from the same two dials. Dial one is size — how far out the working electrons start. Dial two is pull — how much nucleus those electrons actually feel through the screening, module 05's half-felt charge. Everything a family does steadily, row by row, is one of these two dials turning. And down a column the two turn together, because size decides how much of the nucleus's pull survives the distance and the screening. (Across a row they part company — module 05 walked that direction.)
The dials are not opinions. The sizes and the prices are measured; the pull χ is a scale computed from measured bond energies — module 05's dial, carried forward. Before you open the bench, make two predictions and hold yourself to them. Rubidium sits one row below potassium: bigger or smaller? cheaper or dearer to ionise? Now check yourself:
Read the two columns of the “both dials” view against each other. Down the alkali metals, the atom swells from lithium's 128 picometres to caesium's 244 — nearly doubling — while the ionisation price falls from 520 to 376 kJ/mol. The dials move together, in opposite directions, because they are one mechanism seen twice: a new floor each row pushes the working electron out, distance and screening dilute the pull it feels, and a weaker hold is a cheaper eviction. That is the whole engine of §1's gradient — no new idea, just module 05's staircase read downwards.
Now switch families. The halogens — Greek for salt-formers — run the same two dials from the other side: one vacancy instead of one spare, so the family trade is taking an electron, not shedding one — and the taking weakens down the column exactly as the shedding cheapens. Fluorine's pull, χ = 3.98, is the strongest on the whole table; iodine's 2.66 is merely firm. A halogen's fierceness fades with size broadly as an alkali metal's grows with it, and both facts are the same fact: the working floor gets farther from the nucleus. The symbol χ, remember, is that same dial from module 05 — the pull an atom exerts on bonding electrons.
Quick check — read Bench 1 with both dials up. Which pair moves together down Group 1?
Two dials, and the table becomes a prediction machine. Given only where an element sits, you can now say how big it runs and how dearly it holds its electrons — before ever meeting it. That is what the periodic table is for. The rest of this module stress-tests the machine: first on the element that seems to break it, then on a resemblance the columns cannot explain.
Every family photograph has one member standing slightly wrong. In the halogens it is fluorine — the top of the column, the strongest puller, the one you would expect to be simply the most halogen of the halogens. Is it? Start with the bond fluorine makes to its own twin. Climbing the family from the bottom, the X–X bond has been strengthening: iodine 152.5, bromine 193.9, chlorine 242.6 kJ/mol — shorter bond, tighter share. One more row up, to the smallest and hardest-pulling of them all:
Chlorine's Cl–Cl bond holds 242.6 kJ/mol, and the family's bonds have been climbing: 152.5, 193.9, 242.6. Fluorine is smaller still. Commit — the F–F bond holds about…
Measured: 156.9 kJ/mol — far below bromine's, all the way back down at the level of iodine, the supposed runt of the family. The climb does not just stall at the top; it collapses. And the culprit is the very smallness that makes fluorine fierce. Squeeze two fluorine atoms close enough to share a pair, and each atom's other electrons — three untouched pairs apiece, crammed into the tiniest shell in the family — arrive almost on top of the neighbour's. Module 06's soaring push — the steep cost of overlapping filled clouds — starts charging inside the bond itself. The shared pair pays the family's usual wage; the crowding claws most of it back. Chlorine's roomier shell keeps its spectators out of each other's laps, so chlorine keeps the wage.
The same crowding taxes fluorine's other famous number. An atom's electron catch — the energy paid out when it takes one electron in — ought to peak at the strongest puller. Measured, per mole: fluorine 328.2, chlorine 348.6, bromine 324.5, iodine 295.2 kJ. Chlorine out-earns fluorine, because the incoming electron must move into fluorine's already-crowded quarters, and the repulsion bill comes off the payout. Small is strong; very small is strong and cramped, and cramped costs. One cause, two broken trends.
Quick check — which halogen's electron-catch pays out the most energy?
So fluorine makes a weak bond with itself and a second-best catch. It should be a middling oxidiser. It is the fiercest in the whole table — the one gas that attacks glass, platinum, and nearly every element it meets. The paradox dissolves the moment you stop reading single numbers and add up the whole transaction. Watch the ledger:
The ledger settles it. To turn half a fluorine molecule into a dissolved fluoride ion: pay 78.5 to break the feeble bond. Collect 328.2 for the catch. Then collect 505 — the largest hydration payoff of any halide, because water's crowds grip a tiny ion hardest. Total: near −755 kJ/mol. Chlorine's same three lines — pay 121.3, collect 348.6, collect 363 — come to −590. Fluorine finishes about 164 kJ/mol ahead: one line lost (the catch, by 20), two lines won (the cheap bond, by 43; the hydration, by 142). All three lines have one cause. Small means a crowded shell — a cheap bond to break and a discounted catch. Small also means a dense little ion that water rewards extravagantly. The oxidising panel keeps the score in volts: fluorine 2.87, and nobody else close. One dial — size — set every line of the ledger. The member who breaks the family's rules is obeying them harder than anyone.
A chemist's honesty note: this three-line ledger is a back-of-envelope — the full account books in a few more terms and a second currency called entropy, and modules 12 and 21 audit it properly. The back of the envelope already calls the winner correctly, which is why chemists keep one in every pocket.
The same catch that drives the ledger makes the light halogens dangerous to breathe. Chlorine gas corrodes eyes, skin and airway on contact — high exposures flood the lungs and can kill, which is exactly why it was used as a weapon (ATSDR ToxFAQs). Fluorine is harsher still, attacking skin and airway on contact, and is handled only in specialist facilities (ATSDR; RSC). The halogens in this module stay on the page and in cited footage — never in a home experiment. And the one real halogen risk in an ordinary home: never mix bleach with an acid cleaner — the mixture releases chlorine gas into the room.
The columns explain so much that it is worth showing where they stop. Lithium is a well-behaved alkali metal — one electron, water, hydrogen, alkali, the family sentence. But it has habits no one else in its column shares. Heat any other Group 1 carbonate and it sits there, stable past the reach of a Bunsen flame; heat lithium's and it gives up — Li2CO3(s) → Li2O(s) + CO2(g). Leave lithium in plain nitrogen — the gas that famously does nothing — and it reacts, alone in its family: 6 Li(s) + N2(g) → 2 Li3N(s). Habits are caused, and causes have addresses on the table.
Lithium, alone in its family, bonds straight to nitrogen, and its carbonate gives up on heating — tricks no other Group 1 metal manages. Commit — whose chemistry does lithium's most resemble?
Magnesium — one row down, one column right — decomposes its carbonate at Bunsen temperatures, MgCO3(s) → MgO(s) + CO2(g), and burns in nitrogen to a nitride. Lithium's oddities are magnesium's ordinary habits, and chemists call the pattern the diagonal relationship. The mechanism, honestly stated: stepping down the table swells the ion and dilutes its pull; stepping right raises the charge and shrinks it back. For lithium and magnesium it is the size that does the pairing — the bench will show you two ions of nearly one radius, double the charge on one of them. For beryllium and aluminium it is the packing of charge per size that nearly matches. No single number carries the whole diagonal; what repeats is the trade — one step's dilution answered by the other step's concentration.
Put Li⁺ against Mg²⁺ on the bench and the headline is the radii themselves: 76 against 72 picometres — near twins — while lithium's own sibling Na⁺ has swelled to 102 and left it behind. A crystal or a molecule meeting Li⁺ meets, size-wise, a magnesium ion with a gentler charge; the nitride and carbonate habits follow. Follow one down: a small, concentrated ion pressed against a big, floppy carbonate drags the carbonate's electron cloud toward itself, and the distorted ion falls apart on gentle heating into the tidier oxide. Sodium's big, dilute ion distorts nothing, and its carbonate shrugs off the flame. Put Be²⁺ against Al³⁺ and it is the ratio's turn: 4.4 against 5.6 in charge per hundred picometres, with beryllium's true column-mate magnesium back at 2.8 — the down-step dilutes, the right-step restores. The habits run in parallel here too: both oxides answer acids and bases, both chlorides are covalent, bridged molecules rather than chessboards — though this pair's habit list reached only a weaker reference when the page was built, and the bench flags it. The ratio is a guide, not a law — the cited habits are the evidence, and the guide is why the habits pair diagonally. The family photograph tells the truth; the diagonal is the table's fine print.
Now stand back from the two columns you know and read the whole table the way this module has taught you: as a map of the two dials. Left edge: big, loosely held electrons, shed cheaply — metals, seas, violence rising downwards. Right edge, one short of the end: small, hard-pulling vacancies, fierceness rising upwards. The far-right column has no vacancy and no spare — noble, in the old chemists' word. The far-right column almost never buys or sells. And every column in between is a family whose personality you can now price from its position before you have met a single member.
Try it on a family you have never been introduced to — and watch the machine fail, instructively. The coinage metals — copper, silver, gold, one column of the table's middle block — should, by the size dial, shed electrons ever more cheaply down the column. The measured order runs the other way: gold, the biggest of the three, grips its electrons hardest and reacts with almost nothing, which is why civilisations kept their accounts in it. Part of the explanation is the middle block's own crowd — its d electrons screen the nucleus poorly, so the pull climbs faster down this column than the size can dilute it — and gold's full answer needs the same borrowed physics as the box below. The dials are the rule; this family marks the rule's edge. The everyday test sleeps in a drawer: a silver bangle dulls beside a steel spoon that stays bright, because even this near-noble family has a weak spot — a trace of sulphur in ordinary air is enough for silver. (What the sulphur builds, and how to undo it, is module 21's electricity-meets-chemistry territory.)
Module 09's ladder gives the honest first half: mercury lets go at −38.8 °C, so at room temperature its sea has already let go — a shallow sea, not a different glue. Liquid mercury still conducts and still shines; the metal personality is intact, only weakly held. But why is the sea of a dense, heavy metal so shallow, when its neighbours gold and lead hold theirs firmly? Here this course owes you a confession: the full answer sits next door, in physics. Mercury's electrons move so fast near its massive nucleus that Einstein's relativity changes their behaviour — the very effect the LearnPhysics course builds up to — and calculations published in 2013 showed that without relativity, mercury would melt around 80 °C: solid on your desk. (Norrby 1991 — a paper titled, in full, “Why is mercury liquid?” — and Calvo et al. 2013.) The same physics, it turns out, is behind another question waiting in the next module: why gold — and, less loudly, copper — carries colour when nearly every other metal is grey. Two mysteries, one borrowed answer.
Quick check — mercury is a metal that pours at room temperature. On module 09's melting ladder, what does that low letting-go point say?
Children have been drawn to the pourable metal for centuries, and it has poisoned some of them. Mercury is toxic, and the main danger is the one you cannot see: its vapour — a spilled bead keeps evaporating, and about 80% of inhaled mercury vapour passes straight into the bloodstream (ATSDR). It is why mercury thermometers have been retired from homes and clinics. If you ever meet spilled mercury, leave it, ventilate, and tell an adult who can arrange proper cleanup — never a vacuum cleaner, which turns the bead into breathable mist.
Who measured all this? The radii come from X-ray crystallography — thousands of measured crystals, averaged (Shannon's 1976 tables, on Bench 3, are exactly such an average, and module 30 does the measuring). The ionisation prices come from spectroscopy: shine light on a vapour, find the exact energy at which the electron leaves. The bond strengths come from calorimetry and spectroscopy agreeing with each other. Almost none of the table's personality arrives as theory: the theory is the two dials, and the numbers they must explain were measured first.
And that is the family photograph, finished: the resemblance caused, the gradient priced, the misfits explained — or honestly referred next door.
These check themselves, and “New numbers” deals a fresh set — there is nothing to memorise. Each stem says how exact to be.
1. Read Bench 1: the first ionisation price of , in kJ/mol. (To the nearest whole number, ±1.)
2. Bench 3's arithmetic, by hand: . Charge ÷ radius, in charge per hundred picometres. (To 0.1.)
3. Module 05's staircase, revisited: sodium's ionisation prices start with one modest step, then hit a cliff. After how many removals does the cliff come?
4. Bench 2's ledger for : half the bond price, minus the catch, minus the hydration payoff. Give the total, in kJ/mol, sign included. (To ±5.)
Do fluorine's whole ledger yourself, from Bench 2's cited prices: breaking half an F–F bond costs 156.9 ÷ 2 kJ/mol; the catch pays back 328.2; the hydration of the tiny F⁻ ion pays back 505. Add the three lines, sign by sign, and give fluorine's total in kJ/mol. Then check yourself below — and notice which single line did the winning.
Worked through. Pay 78.5 (half of 156.9). Collect 328.2. Collect 505. Total: 78.5 − 328.2 − 505 = −754.7 kJ/mol. Chlorine's same three lines — pay 121.3, collect 348.6, collect 363 — come to −590.3. Fluorine wins by about 164 kJ/mol: the catch loses it 20, the cheap bond wins it 43, and the hydration line wins it 142 — water's grip on the smallest ion does most of the work. (Bench 2, carrying the unrounded 78.45, prints −754.8: same ledger.) (Back-of-envelope, remember: the full audit, with entropy in the books, is modules 12 and 21.)
Group 2 — beryllium, magnesium, calcium, strontium, barium — sits one column right of the alkali metals. You know the dials; you have never been introduced to barium. Predict three things about barium against magnesium: bigger or smaller? ionisation cheaper or dearer? calmer or fiercer in water? Then a harder question: §4's diagonal pattern was shown for the table's top rows — would you expect it to matter as much this far down? Write your four verdicts before looking.
The discussion. Barium sits three rows below magnesium: bigger by far, its working pair starting floors farther out; cheaper to ionise, by the same distance-and-screening argument; and accordingly fiercer in water — barium reacts the way the mid-table alkali metals do, where magnesium barely troubles cold water. And the diagonal: it fades. The relationship earns its name at the table's top, where one step is a large share of an ion's whole size — lithium to magnesium nearly doubles the charge on almost the same radius. Six rows down, one step changes the proportions far less, and barium's nearest kin in behaviour is simply its row-mate caesium — both huge, both cheap, both violent; caesium sheds one electron, barium two. The prediction machine runs on families you have never met; it just pays to know which of its rules are top-of-the-table rules.
One outer-electron count makes a family; size and screening set the gradient.