LearnChem · Part I — What is stuff made of?
That odd castle with the two towers and the long middle is not a design. It is a measurement.
“Why eight? Why not ten?”
Last time, foil that bullets bounce off forced the nucleus into the open — and hydrogen's glow showed its one electron sitting only on fixed energy rungs. But hydrogen is the easiest instrument there is: one electron, one nucleus. Gold has seventy-nine electrons. What do the rungs do with a crowd? And why, walking the elements in order, does the same personality keep returning every eighth step? This module reads both answers off the measurements — and the answer is the table itself.
Line the elements up in order — lightest atom to heaviest is nearly right; counting protons, which module 04 handed you, is exact — and walk. A soft silvery metal that fizzes in water: lithium. Eight steps later, another soft silvery metal that fizzes in water: sodium. Eight more: potassium, softer and fiercer still — chemists call this family the alkali metals. The same act keeps returning, like a chorus. Fluorine, chlorine, bromine: three corrosive, choking elements, one under the other — the halogens. And one step past each halogen, a gas that reacts with nothing at all — the noble gases, so called because they mix with nobody. (The Victorians never heard that part of the chorus: those gases hid until the 1890s, precisely because they do nothing.)
Nineteenth-century chemists heard the rest of it and could not explain a bar. John Newlands pointed out the every-eighth-step return in 1865, and was mocked for it when he read it to London's Chemical Society. Dmitri Mendeleev, in 1869, did the brave thing: he took the pattern as law, arranged the elements so the family lines came out true — and where the pattern demanded an element nobody had found, he left a hole. By 1871 he had published the missing occupants' properties in advance. Weight, density, the chemistry they would have: predictions for metals nobody had yet seen.
In 1875 a French chemist found a new metal, gallium, and it slotted into a hole — matching the hole's predictions closely enough to make Europe's chemists sit up. When gallium's first measured density disagreed with the prediction, Mendeleev published a note suggesting the sample was impure. Re-purified and re-measured, the density landed where the hole had said. The pattern had predictive teeth. Nobody on Earth knew why.
“Why eight?” is therefore the right question, asked for fifty years by the best chemists alive. The answer had to wait for the atom to be opened — for module 04's nucleus and floors. You now hold the pieces they lacked. This module assembles them.
Quick check — Mendeleev left holes in his table and published predictions for the missing occupants. What made that bold rather than careless?
To see what the atom's crowd of electrons is doing, ask a blunt question: what does it cost to pull the electrons out, one at a time? The price of removing an electron is called the ionisation energy, and it is paid in the atom-sized unit from the course's first page, the eV. Strip an atom electron by electron, writing down each price, and the bill tells you more about the atom's inside than any picture could.
Take sodium, the fizzing metal. Its first electron is famously cheap: 5.1 eV, among the cheapest in the table — that looseness is half of sodium's personality.
Stripping sodium's first electron costs 5.1 eV. Commit to a prediction for the second.
Commit before you look. First answers are counted anonymously, never named.
47.3 eV — nine times the first price. Call that multiple the jump factor: the new price over the old, and nine is enormous. Not a gentle rise: a cliff. And then, stranger, the cliff stops. The third electron costs 71.6, the fourth 98.9 — a staircase climbing steadily again, as if nothing had happened. Eight steady steps. Then, at electron ten, the bill goes mad: 1,465 eV. A second cliff, five times its neighbour.
Why should the prices climb at all? Each electron removed leaves the ion one more unit of net positive charge behind, so every electron that stays is held a little harder: steady steps, jump factors near one. But one extra unit of charge cannot turn 5 eV into 47. Something else changed at that step — the second electron must live somewhere else entirely: far deeper, far closer to the nucleus. A jump factor above about four, then, is not a step but a cliff. Walk the whole bill yourself, and watch where the cliffs fall:
Read the bill as a floor plan. Small steady steps: electrons living at about the same distance from the nucleus, in the same neighbourhood. A cliff: the next electron is not in that neighbourhood — it lives far deeper, held many times harder. Sodium's bill says: one loosely held electron on the outside; then a tight neighbourhood of eight; then, deepest of all, a pair costing over a thousand eV each. One, eight, two. The atom has floors — and switching elements moves the first cliff exactly one step along: magnesium's comes after two electrons, aluminium's after three. The floors are real, and the outermost floor's headcount is precisely what changes from element to element.
Chemists call the floors shells. Hydrogen's ladder of rungs, from module 04, was the one-electron version of the same fact; the staircase is what the floors look like when a crowd is actually living on them.
Nobody pulls electrons off one by one with tweezers. The prices are read from light: each ionisation stage of an element has its own ladder of spectral rungs, and the rungs crowd together towards a top — the energy at which the electron comes off entirely. That top, the series limit, is the ionisation energy, which is why the values are known so sharply. (The deepest few prices are calculated rather than read off a spectrum, and cross-checked against it.) The staircase above is bookkeeping on light.
Quick check — an element's ionisation energies run 7.6, 15.0, then 80.1 eV. How many electrons sit on its top floor?
Now count floors across many elements — the bench walked you through three; chemists hold the full bill for all ninety-odd — and the numbers come out with total regularity. The ground floor holds at most 2 electrons. The next floor holds at most 8. The next: 8 again, though with a surprise hiding in it.
Module 04 left the electron as a standing wave — a cloud of probability with fixed energy floors. The same wave mathematics that fixes the floors also divides each floor into rooms. The ground floor has one room: a round one (call it an s-room). The second floor has an s-room and three rooms of a dumbbell kind (p-rooms) — three because a dumbbell can point along each direction of space: left–right, front–back, up–down. And every room, of either kind, takes two electrons and no more — a measured rule of the electron itself, never once broken; its name and its reason arrive in a later module. So the second floor holds 2 + 6 = 8. There is the eight. The recurring chorus of §1 is the top floor refilling by the same plan, element after element. Lithium, sodium and potassium each stand with exactly one electron on a fresh top floor; that lone loose electron is the family personality. The noble gases stand where a floor's s- and p-rooms have just filled — a closed eight: nothing cheap to give, no place free to take.
But the wave mathematics gives floor 3 more than four rooms: an s-room, three p-rooms — and five further rooms of a third kind (d-rooms), places for ten more. Eighteen places in all. So the honest question is the reverse of the famous one: not “why eight?” but “why does period 3 — the table's third row — stop at eight, when its floor has room for eighteen?”
Floor 3 has an s-room, three p-rooms and five d-rooms — places for eighteen. Yet period 3 of the table ends after eight, at argon. Why?
First, the building's move-in rule, which the staircase never needed: an arriving electron always takes the lowest empty place — filling is move-in, not eviction. Now the building's broken promise: “height” here is energy, not metres, and at this point in the table, floor 3's d-rooms sit higher than floor 4's bottom room. So after argon fills floor 3's s- and p-rooms, the next two electrons — potassium's and calcium's — move in upstairs, skipping the d-rooms entirely. Only then, from scandium onward, do the five d-rooms of floor 3 fill. The table obligingly grows a ten-column middle block to hold them: the transition metals, the long low wing of the castle. Period 4 runs 2 + 6 + 10 = 18 wide — count it on any table, potassium to krypton. (And once tenants actually occupy the d-rooms, the price list shifts again; the transition-metal module tells that story.)
So the table's whole silhouette — two tall towers, a long middle, the deep basement rows at the bottom (the f-rooms, seven per floor, fourteen places) — is nothing but the floor plan, drawn left to right in move-in order. Why eight? Because one floor's s- and p-rooms hold eight. Why not ten? Because rooms come in sets of 1, 3, 5 and 7, holding 2, 6, 10 and 14 — and a row's width is always a sum of complete sets: 2, then 2 + 6 = 8, then 2 + 6 + 10 = 18, then 32. No sum makes ten. Ten does exist in the shape — it is the width of the long middle block, the d-set filling — but it never stands alone as a row. The castle is not a design. It is the electron's wave mathematics, printed on paper.
Floors explain the table's rows. To explain the trends — how size and grip change along a row, and down a family — one more idea is needed, and it is bookkeeping again.
An outer electron in sodium does not deal with the full nucleus. Eleven protons pull it in; but ten inner electrons sit between, and their repulsion cancels most of that pull. Chemists call the cancelling screening: the outer electron feels the full charge, minus the crowd standing in the stairwell. What is left over — the charge the electron actually does business with — is the effective nuclear charge. In 1930 the physicist John Slater published a set of counting rules for estimating it: same-floor neighbours cancel a little (0.35 each), the floor below cancels a lot (0.85 each), and anything deeper cancels completely (1.00 each). Crude, but good enough to carry every trend in this module:
Slide across a row and watch the felt charge climb: sodium's outer electron feels about 2.2, chlorine's about 6.1. Sodium, worked in one line: 11 protons, minus 8 × 0.85 for the floor below, minus 2 × 1.00 for the deep pair — 11 − 6.8 − 2 = 2.2. Same floor all the way to chlorine, but each step rightward adds a full proton to the nucleus and only a same-floor neighbour to the screen, and same-floor neighbours cancel worst of all. The grip on the outer electrons therefore tightens across every row. Then drop from any element to the one below it. Potassium, same sum one floor later: 19 − (8 × 0.85) − (10 × 1.00) = 2.2 — the felt charge comes out almost unchanged, but the electron holding it starts a whole floor further out, and distance wins: the grip loosens. So two dials set every outer electron's life: dial one, the felt charge; dial two, which floor it lives on. The next section cashes them in.
Quick check — potassium's nucleus holds 8 more protons than sodium's, yet its outer electron costs less to remove. What pays for that?
Time to cash the two dials in — felt charge, and floor. Start with size, and commit first:
A sodium atom and a chlorine atom, both from row 3. Commit: which is wider?
Sodium, and it is not close — half as wide again as chlorine. The six extra electrons all join the same floor, while six extra protons join the nucleus; the felt charge nearly triples across the row, and the whole floor is reeled inward. More electrons, smaller atom: one of chemistry's best jokes, and the §4 arithmetic delivers the punchline. Now walk all three trends at once, in measured data:
Three saw-tooth waves, all cut by the same two dials. Size drifts smaller across each row and leaps at each new floor — the leaps at sodium and potassium are the biggest moves on the whole chart. (The points wobble near each row's end and through the d-block middle: this radius set is one consistent fit to real and computed bond lengths, and reality is under no obligation to be tidy.) Ionisation energy runs the mirror image: climbing across each row as the grip tightens, crashing at each fresh floor. Every crash lands on an alkali metal, every crest on a noble gas — and every crash in this curve is the first step of a fresh staircase: the cheap lone electron on a new floor. (The climb has small stumbles — beryllium to boron, nitrogen to oxygen — fine structure inside a floor, worth noticing and not yet worth explaining.) Mendeleev's chorus, drawn in eV.
The third row needs one new word. Put two different atoms in a tug-of-war over a shared pair of electrons — bonds do exactly this, and module 06 opens that box — and the winner is decided by the two dials again: felt charge, and distance. The pulling strength has a name, electronegativity, written χ (chi), and its league table broadly follows the dials: rising across rows, easing down families, peaking hard at fluorine — small, barely screened, the hardest puller in the table. (The d-block middle wobbles; module 11 enjoys that.) Helium, neon and argon sit as blanks in this row: electronegativity is measured from compounds, and those three offer almost nothing to measure. A cousin of the idea, met going the other way — the energy change when an atom accepts an extra electron, the electron affinity — runs on the same dials, with more fine print than this course needs yet; χ is the number that will work for a living from module 06 on.
The dials reach everyday life. “Low-sodium” table salt is the family pattern sold in shops: part of the sodium chloride is swapped for potassium chloride, one alkali-metal column-mate for another — cousins close enough that your tongue mostly accepts the swap. Helium lifts party balloons because a closed floor cannot burn; hydrogen, one proton away and cheaper, is the filling that makes the news when a street seller's cylinder goes wrong. And the argon in every breath you take — about 1 particle in 100 of dry air — has passed through countless lungs and reacted with none of them: a closed eight, since before the dinosaurs.
The table maps every kind of ordinary matter there is — so where were its entries manufactured? Almost all the hydrogen and helium are original stock from the universe's first minutes. Nearly everything else was assembled inside stars: nuclear work at nuclear prices — module 01's MeV, not eV — run at millions of degrees for millions of years, and scattered when the stars died. The iron in your blood and the sodium in your salt are star output; the goldsmith's furnace from module 01 fails at gold because it is a candle next to the machines that actually make elements. The full story is physics' to tell — the sister course takes it up — but nearly every element this module just sorted into floors was sorted first by a star.
One last walk, straight down the first column. Below potassium: rubidium, then caesium, then, at the very bottom, francium — by the two dials the loosest metal the table can offer. Yet nobody has ever seen a lump of it, and nobody ever will: francium's nuclei fall apart within minutes of being made. The electrons write an element's whole personality — but, module 01's line read from the other side, the nucleus decides whether the element gets to exist at all.
Quick check — fluorine tops the electronegativity table. Reading the two trends that build it, fluorine sits…
These check themselves, and “New numbers” deals a fresh set — there is nothing to memorise. Within a few per cent counts as right unless the check says otherwise.
1. An element's first ionisation energies, in eV: . How many electrons on its top floor?
2. The counting bridge from module 03, revisited: a strip of aluminium foil weighs . Aluminium is 26.98 g/mol — how many moles of atoms is that?
3. Slater's bookkeeping: protons in the nucleus, and the inner crowd screens of them away. The outer electron's felt charge is…
4. Reading a cliff: . The jump factor at that cliff is…
An element from period 3. Its successive ionisation energies, in eV: 5.99 — 18.83 — 28.45 — 120.0 — 153.8. Find the cliff, count the top floor, name the element. Type its name or symbol.
Worked through. Hunt the cliff: 5.99 → 18.83 is ×3.1; 18.83 → 28.45 is ×1.5; 28.45 → 120.0 is ×4.2 — the cliff. Three electrons come off before the wall, so the top floor holds three. A period-3 element with three outer electrons stands in the third column of the towers: aluminium. (The bill continues in the same NIST ledger as Bench 1: the tenth step ends near 400 eV, and the eleventh — into the deepest pair — passes 2,000. One, two, three floors: 3, 8, 2, reading down.)
Rubidium sits directly below potassium — one row past the edge of Bench 3. Play Mendeleev: predict, with a one-line reason each, whether rubidium is wider or narrower than potassium; whether its first ionisation energy sits above or below potassium's 4.3 eV; and which way its electronegativity should move from potassium's 0.82. Then check your machinery on the sting §5 ended with: why exactly does francium's missing lump not contradict a single one of your predictions? There is no marking here; write your reasoning down before looking.
The discussion. Down one step: a new floor of screeners moves in, so rubidium is wider; its outer electron sits further out behind the same felt charge, so its ionisation energy falls below 4.3 eV. The dials point χ downward too — and here the measurements teach one more lesson: on the printed Pauling scale, potassium and rubidium both read 0.82. The trend is real but smaller than the scale's two decimal places — a league table is a coarse instrument, and knowing how coarse is part of reading it. (Rubidium's measured ionisation energy does come out below potassium's: 4.18 against 4.34 eV.) As for francium: your predictions describe what its electrons would do — and its nuclei never stay assembled long enough for a lump to exist. The predictions aren't wrong; they are unredeemable. Electrons propose; the nucleus disposes. That is the game Mendeleev played, and the whole reason it worked: the dials predict, no lookup needed.
The table's shape is the shape of the electron shells, and nothing else.