LearnChem · Part II — Why do atoms stick together?
A spoon contains no molecules. Neither does salt — and a diamond is one single molecule the size of the stone. Three new glues hold them, and each leaves its own fingerprint: shine, shatter, hardness.
“If atoms are mostly empty space, why can't my hand pass through the table?”
Last time, the crowd held hands: intact water molecules gripping each other at a twentieth of a bond's price — a grip so cheap that a sunny morning undoes it, and the ice is a puddle by ten. But now look at your spoon. There is no crowd in it, and no hands: a steel spoon and a salt crystal contain no molecules at all, and a diamond is a single molecule the size of the stone. Whatever holds them together is a different kind of glue entirely — and a sunny morning means nothing to it. This module sorts the glues.
The question at the door deserves a straight answer, because module 04 sharpened it: nearly everything thrown at a gold foil sailed straight through, and the nucleus turned out to be a marble in a stadium — the mass all in the marble, the space nearly all empty. Your hand is a cloud of such stadiums. So is the table. Two collections of almost-nothing approach each other… and stop dead, every single time.
An atom is nearly all empty space — module 04's marble in a stadium. Your hand is mostly nothing; the table is mostly nothing. Commit — why does the hand stop?
Commit before you look. First answers are counted anonymously, never named.
The atom is empty of stuff, not of charge. Its outer electrons fill the stadium the way a hum fills a hall — thinly spread, but everywhere. Push two atoms together and those outer clouds arrive first, and module 06's valley already showed you what happens next: remember its left wall, the cliff the energy curve climbed when the two atoms came too close. Squeezing electron clouds into each other costs energy at a ferocious and rising rate. Your hand pressing on the table is that cliff, met at trillions upon trillions of atoms at once. Nothing needs to touch, in the marble sense, for the price of one more nanometre to exceed anything your arm can pay. Solidity is not fullness. Solidity is a wall of cost.
So any collection of atoms resists being squeezed — that part is universal. What makes a spoon a spoon, salt salt and diamond diamond is the other side of the ledger: what holds the atoms together, against being pulled apart or slid about. Ice answered with hands — weak grips between intact molecules. This module meets the three glues that work without molecules: the shared sea, the charge checkerboard, and the bonded network. Three glues, and you already know the fourth. Four personalities. Everything on your table is one of them.
Begin with the spoon. A metal atom holds its outermost electrons cheaply — module 05 priced them: the far floor, well screened, the least costly electrons the atom owns. Pack metal atoms together and those outermost electrons stop belonging to anyone. They pour off into a shared pool that runs through the whole stack — chemists call it the electron sea — and what remains is a lattice of positive cores sitting in it, every core held by the sea around it, the sea held by every core. That mutual grip is the metallic glue.
Notice what this glue lacks entirely: direction. A core in the sea is held the same whichever neighbours it has — no arms pointing anywhere, no fixed partners. The atoms are, as far as the glue is concerned, identical spheres to be packed — and identical spheres, left to settle, pack the way oranges pack on a vendor's cart. Watch a stack find this out for itself:
Pour identical discs, and the close-packed rows appear on their own — inner discs ringed by six neighbours, the rows running level — because settling under gravity is a search for lowest energy, module 06's downhill run played with a hundred and fifty balls at once. (Gravity here stands in for the sea's all-round pull; any steady pull toward togetherness runs the same search.) Nobody drew the rows; jiggle the pile and they re-form. That is the deep reason metals crystallise so readily: to a direction-blind glue the atoms are interchangeable spheres, and interchangeable spheres have one best answer.
Now pour the mismatched discs. Same gravity, same search — and the rows never come, because two sizes thrown together at random cannot sort themselves into a shared pattern by jiggling alone. (Ordered two-size stacks do exist — salt, two sections on, is exactly one — but they need a rule to dictate who sits where, and these discs have none.) The search is frustrated: the pile locks into a dense, orderless jam. Hold that jammed pour in mind — it has a famous name, and §5 pays it off.
The sea explains the spoon's whole character. Bendable: push a row of cores one step sideways and the sea simply keeps holding — no bond pointed anywhere in particular, so the slid layer is gripped exactly as the old one was; the metal dents and flows rather than cracking. That is why a steel bangle bends where a glass one shatters (pure metals, at least — a cast-iron kadai, its sea interrupted by lumps of carbon, cracks like stone if dropped), and why a goldsmith can hammer metal into leaf: the silver vark on mithai is a sea rolled thinner than paper. Conducting: the sea is loose charge; tilt it electrically and it flows — a current is the sea on the move. And shiny: light is a rippling electric field, and the sea's charges are free to ripple in answer, taking the arriving light and throwing it straight back out. A polished insulator can only ever gleam; the silver mirror of a spoon is the sea itself, answering.
Quick check — polish a block of wood to a mirror finish and it still shows you almost nothing. Polish steel and it shows you your face. What makes a metal shine?
The settled stack repeats. One small box of it — a few atoms in a fixed arrangement — is stamped over and over in every direction, the way one tile repeats across a floor, and chemists call the box a unit cell. The whole crystal is the cell, copied without end. Which means the whole crystal can be weighed from the cell — if you can count what a cell really contains, and that counting has a famous trap in it.
Quick check — picture a 2 × 2 × 2 stack of little boxes, eight dice glued into one cube, and an atom sitting at the very centre point where they meet. How many boxes share that atom — so what fraction of it belongs to each box?
Do the count for real, and then collect the payoff — a number you could check with a kitchen scale and a measuring jug:
Run copper. Eight corners at an eighth each is one atom; six faces at a half each is three more: four atoms to a cell. The cell's edge — measured by X-rays to better than a thousandth of an ångström, 3.6150 — gives its volume; four coppers' worth of mass divided by that volume predicts 8.93 grams per cubic centimetre. The measured density of copper, weighed the ordinary way, is 8.96. The invisible box, counted in eighths and halves, reproduces the weighed number to a third of a percent. This is why chemists trust the cell: it earns its keep in weighable currency. And notice the filled fractions as you switch structures — copper's tidy stack fills 74% of space, the best identical spheres can ever do, while diamond's cell, for a reason §4 makes plain, stands about two-thirds empty.
Nobody measures a cell with a ruler. The 3.6150 comes from X-rays again: waves short enough to feel the rows, bounced off the stack, agreeing brightly only at angles set by the row spacing. From the angles, the spacing; from the spacing, the cell; from the cell, the density you just checked — module 30 does the bouncing itself. For now, note that the prediction agreeing with the bathroom scale is the evidence the X-rays were read right.
Now bring in the second glue. Module 06 built it and priced it: sodium hands over an electron, chlorine takes it, and the ions stack + − + − in exactly the chessboard you can turn over on Bench 2 (pick rock salt) — every positive ringed by six negatives, every negative by six positives, the whole ledger closing at −787 kJ/mol. Strong glue: salt shrugs off any kitchen flame and holds solid to 801 °C. So here is a puzzle. The glue is strong — and the crystal is fragile. Drop a salt crystal on a stone floor and it does not dent like the spoon. It splits, along flat faces so clean they look machine-cut.
Tap a copper wire with a hammer: it dents and bends. Tap a salt crystal exactly as hard: it splits clean along a flat face. Both are tidy stacks of spheres. Commit — why does the ionic stack crack where the metal stack bends?
Brittleness is not weakness. It is strength that depends on position. The metal's sea holds whichever cores sit wherever — slide a layer and the glue travels with it unchanged. The chessboard's glue is the alternation itself: every attraction in it exists because + sits beside −. Shear one layer a single step and the pattern that held becomes its own opposite — plus over plus, minus over minus, attraction flipped to repulsion along the whole plane at once — and the crystal does not so much break as fling itself apart along that plane. That is why rock salt — halite, once it is a mineral in the ground — splits along those uncanny flat faces: they are the planes where one step of slide is catastrophic. And notice: the charge alternation is also the rule that sorts two ion sizes into one perfect pattern — the rule Bench 1's mismatched discs lacked. Strong and rigid, therefore brittle; bendable because direction-blind. The personality is the glue's geometry, not its strength. The flat cleaved face is the chessboard's fingerprint.
The third glue you have known since module 06: the full-price covalent bond — and its personality shows best in the one element that plays it two ways. In diamond, every carbon bonds to four neighbours at module 07's tetrahedral angles, and the four arms repeat outward until the whole crystal is one molecule the size of the stone — a network in which going anywhere means breaking full-price bonds. (Four arms at fixed corners also hold the atoms well apart — which is why Bench 2 found diamond's cell only a third full.) That is hardness, the network's fingerprint: scratching diamond means snapping short, braced covalent bonds by the row, in every direction at once, and almost nothing arriving at its surface can pay.
In graphite — same atoms, same purity — each carbon bonds to only three neighbours, in strong flat sheets, atoms closer within a sheet (1.42 Å) than diamond's are to each other (1.54); but between the sheets, 3.35 Å apart, run no bonds at all — only the faint universal cling of the bonding world's loose change, the cling module 08 met and left nameless. So the sheets shed. A pencil writes by leaving sheets of graphite behind on the paper; a diamond writes on glass by breaking it. Same element. The wiring is the personality.
And the fourth glue is last week's news: sugar is a molecular solid — intact molecules, module 08's crowd, held by hands and loose change. Which settles a kitchen puzzle: sugar and salt, two white crystals a cook can barely tell apart, part company the moment the tawa warms. Sugar's grips are cheap: it softens, browns and burns before 200 °C. Salt's chessboard holds to 801 — six hundred degrees past it — without a tremor. And drop each into water: both vanish, but only the salt water conducts, because the chessboard split into free charges while the sugar is still whole molecules, hands held, with no charge to carry. Identical to the eye; different glue; different world.
Quick check — pencil lead and diamond are both pure carbon. What single difference makes one write and the other cut glass?
You cannot see a glue. But you can test for it — put heat in and watch what the solid does. Each answer below is one glue's fingerprint, read under heat.
A copper wire's electron sea is already free — that is what makes it a conductor. Commit — heat the wire, and does it conduct better, worse, or just the same?
The sea's problem is never freedom; it is the obstacle course. Current is the sea flowing through the stack, and a cold stack stands in still, tidy rows. Heat the stack and the rows jiggle in place — and a jiggling row scatters the flow, the way carts lurching about a narrow lane scatter a cyclist threading it. The measurements say so with satisfying steadiness:
From 100 K to 700 K, copper's resistivity climbs thirteen-fold — steady, measured proof that the obstacle course, not the carrier count, is what temperature changes in a metal. But hold on to the wrong answer above, because it is exactly right for a different family. In pure silicon — intrinsic, chemists say, meaning nothing added — almost no charge is free to begin with: at room temperature fewer than one atom in a million million contributes a carrier, and the material resists about 10¹¹ times harder than copper. There, heat's main job is setting carriers free: every ten-degree rise roughly doubles their number, and silicon conducts better as it warms — the opposite slope. One measurement, run on two solids, sorts the metals from the semiconductors, and the slope names the glue.
Heat reads the other fingerprints just as plainly. Why does a steel spoon feel colder than the wooden one that shared its drawer all night? Not temperature — one drawer, one night, one temperature. The sea carries heat as well as charge: steel conducts heat about a hundred times better than wood (steel about 15 W/m·K to hardwood's 0.16; copper, all sea, reaches 401), so it pulls warmth out of your fingertip several times faster, and your skin reads the drain rate and calls it cold. Dip both spoons in hot tea and the verdict reverses: the steel handle warms first — same sea, same drain, other direction.
The melting ladder. Line the solids up by the heat their glue withstands, and each family signs in its own range. Molecular solids let go first: ice at 0 °C, its hands undone by a warm morning. The chessboard holds to 801. Quartz's network reaches about 1710 — and carbon, kept from air, never melts at ordinary pressure at all: near 3,800 °C it sublimes straight to vapour (in air, a diamond burns long before any such heat). The metals' seas, meanwhile, span almost the whole ladder — sodium's shallow sea lets go at 97.8 °C, leaving it soft enough to cut with a knife (watched on the demonstrator's bench, never handled), while copper's deeper sea holds to 1085 — so a low melting point never convicts a whole family; it may only mean a shallow sea.
Two answers stack. First: why does light pass through glass at all? Glass is a network solid — silicon and oxygen, cousin to quartz — and its bonding electrons are held so deeply that one packet of visible light (a photon, carrying at most about 3 eV on module 01's energy scale) cannot promote a single electron: the cheapest jump in the material costs around 9 eV. Unabsorbed, the light passes through. (Metals are the mirror image, literally: their sea answers and returns nearly everything — the part copper's sea keeps is exactly why copper is not silver-white.) Then why isn't sand transparent? It is — grain by grain. Look at beach sand with a magnifier and the quartz grains are tiny windows; a heap is opaque because light crossing it meets thousands of grain surfaces, bending and bouncing at each until it staggers back out the way it came — white, like mist, which is transparent water droplets playing the same trick. Melt the grains into one continuous piece, no internal surfaces left, and the scattering stops: glass.
And here Bench 1's mismatched pour collects its payoff. Silica can order — every sand grain is a quartz crystal — but a silica melt is a tangled network, so sluggish that ordinary cooling locks it before the rows can form: a solid wearing a liquid's disorder, held rigid. (The disorder is not what makes glass clear — a quartz crystal is just as clear; clarity is the 9 eV gap's doing.) And a glass bangle, all strong bonds and no forgiving sea, shatters like salt rather than bending like steel.
Quick check — a steel spoon and a wooden spoon sat in the same drawer all night. The steel one feels colder. Why?
These check themselves, and “New numbers” deals a fresh set — there is nothing to memorise. Each stem says how exact to be.
1. Cell bookkeeping: . Using corners ÷ 8, faces ÷ 2, body × 1 — how many atoms does the cell really contain? (A whole number.)
2. Read it off Bench 2: . What fraction of space does it fill, in percent? (To 0.2.)
3. Module 06's dial, revisited: the glue between . Look both pulls up on the electronegativity (χ) table (module 05's Bench 3, or module 06's dial) and give the mismatch, Δχ, to two decimals — the number that sorted sharing from handover.
4. The density payoff: . Predicted density = count × atomic mass ÷ (cell volume × NA), with NA = 6.022 × 10²³. Give it in g/cm³. (To 0.05.)
Diamond's unit cell: edge a = 3.5667 Å (that is 3.5667 × 10⁻⁸ cm), and the cell's own bookkeeping — corners, faces and four carbons wholly inside — totals 8 atoms. Carbon's atomic mass is 12.011 g/mol. Convert the edge to centimetres, cube it for the volume, and work out the density diamond's invisible box predicts, in g/cm³, to one decimal place. Then check yourself below.
Worked through. Volume first: (3.5667 × 10⁻⁸ cm)³ = 4.537 × 10⁻²³ cm³. Mass next: 8 atoms × 12.011 g/mol ÷ 6.022 × 10²³ /mol = 1.596 × 10⁻²² g. Divide: 1.596 × 10⁻²² ÷ 4.537 × 10⁻²³ = 3.52 g/cm³. The NBS calculated value from the same cell is 3.515; the measured density of gem diamond is 3.513. You have just weighed a crystal by geometry — a box no eye has ever seen, counted with eighths and halves, agreeing with a balance to a tenth of a percent. This is the same trick Bench 2 played on copper, done with your own hands.
Five solids from your own kitchen: a steel spoon, table salt, sugar, a glass tumbler, pencil lead. For each one predict three behaviours — bend or shatter? conduct or not? gives in easily to the stove or shrugs it off? — and then name the glue. Write your five verdicts before looking; the discussion sorts them.
The discussion. The spoon: electron sea — bends, conducts (charge and heat both, hence its false coldness), shrugs off any stove. Salt: chessboard — shatters along flat planes, conducts nothing as a solid (every charge locked in place; melt it and the freed ions conduct — electricity and chemistry meeting, module 21's territory), and holds to 801 °C. Sugar: molecular crowd — crumbles softly, conducts nothing, and gives up on the tawa before the ghee is even hot, browning as its cheap grips let go. The tumbler: a covalent network locked mid-search, Bench 1's mismatched pour made real — no sea to forgive a blow, so it shatters; no free charge, so it conducts nothing; strong bonds, so it takes serious heat, though its frozen disorder means it softens gradually rather than melting at one sharp temperature. Pencil lead: the two-personality element — it smears rather than shatters (the sheets shed under any blow), it rides out a stove (the clay binder chars before the carbon gives), and, unusually for a non-metal, it conducts, because each carbon bonded three ways instead of four leaves electrons loose in the sheets — a thin, flat cousin of the sea. Five solids, four glues, one drawer.
The glue decides everything: seas bend and shine, chessboards shatter, networks hold.