Conductivity is a rate, not a reservoir
The last-but-one guide gave you specific heat — how much vibrational energy a body soaks up per degree, its size as a thermal sponge. This guide asks the sibling question, and it is a genuinely different one: hold one end of a bar in a flame, and how fast does the far end get hot? That speed is thermal conductivity, written k and measured in watts per metre per kelvin (W/m-K). It is the constant in a simple law of flow: heat streams from hot to cold down a temperature gradient, and k says how much heat crosses each square metre per second for a given steepness of that gradient. High k, and heat races through — grab a copper rod over a burner and it bites your other hand within seconds. Low k, and heat crawls — a wooden spoon can sit in boiling soup with a cool handle.
Keep capacity and conductivity firmly apart — they are the two independent numbers this rung keeps circling. Capacity is how much heat a body can hold; conductivity is how fast heat flows through it, and a material can be high in one and low in the other. Water is the classic split personality: a huge specific heat (about 4180 J/kg-K) makes it a greedy reservoir, yet its conductivity is a feeble 0.6 W/m-K, so it stores heat happily but passes it along slowly. Reservoir versus pipe: this guide is entirely about the pipe.
Two couriers: free electrons and phonons
In a metal, cast your mind back to the metallic bond: the atoms surrender their outer electrons into a shared sea that drifts freely through the whole crystal. That same sea that carries an electric current also carries heat. Warm one end and the electrons there speed up; being light and fast and unattached, they dart across to the cold end and dump their extra energy into the atoms they meet. This is electronic heat conduction, and it is fast — which is exactly why metals crowd the top of the conductivity ladder, with silver, copper, and aluminium leading the pack.
In ceramics, glass, and polymers there is no free electron sea — the electrons are locked tight in their bonds, which is precisely why these materials are electrical insulators. So the heat has no fast courier and must ride the lattice vibrations directly: the phonons from the heat-capacity guide, those indivisible packets of jiggle, ripple from the hot end toward the cold, each atom nudging its neighbour along the springs. This is phonon heat conduction, and it works — but it is generally slower and messier than electrons streaming through open space, so non-metals sit far lower on the ladder. One honest caveat: metals actually have both couriers, phonons and electrons, but the electrons out-carry the phonons by roughly ten to thirty times, so we fairly say a metal conducts 'by electrons' and quietly pocket the small phonon share.
Thermal conductivity k spans ~5 orders of magnitude (room temperature, approximate values) material k (W/m-K) main heat carrier ----------------- ---------- ------------------------ diamond ~2000 phonons (stiff, ordered) silver ~429 free electrons copper ~400 free electrons aluminium ~237 free electrons iron ~80 free electrons alumina (Al2O3) ~30 phonons stainless steel ~15 electrons (alloy-scattered) silica glass ~1.4 phonons (disordered) water ~0.6 molecular collisions polyethylene ~0.4 phonons (along chains) most plastics ~0.2 phonons (disordered) wood ~0.15 phonons + pores polymer foam ~0.03 trapped air (mostly) still air ~0.026 gas molecules Note: the electrical INSULATOR diamond out-conducts every metal.
Metals: one electron sea carries charge and heat together
Because the very same free electrons carry both the electric current and the heat, a good electrical conductor is almost automatically a good thermal conductor — silver and copper top both lists, and that is no coincidence. This tight link has a name: the Wiedemann-Franz law. It says the ratio of thermal conductivity k to electrical conductivity sigma is not random from metal to metal but settles onto a near-universal value: k / sigma is approximately L times T, where T is the temperature and L is the Lorenz number, about 2.44 x 10^-8 in units of W-ohm/K^2. The same electron sea sets both — so measure how well a metal conducts electricity and you can predict, to a good first approximation, how well it conducts heat.
Put a number through it. Copper's electrical conductivity is about 5.96 x 10^7 siemens per metre. At room temperature (T around 300 K), the law predicts k = L times sigma times T = 2.44 x 10^-8 times 5.96 x 10^7 times 300, which comes out to roughly 430 W/m-K — sitting right next to copper's measured value near 400. Be honest about the fit: Wiedemann-Franz is an approximation that works well for pure metals at and above room temperature, and it sags at intermediate temperatures where electron scattering changes character. But the deep point survives: a single feature of metallic bonding, the mobile electron sea, sets both conductivities at once.
This also explains a fact you can feel in a workshop: alloys conduct heat worse than the pure metals they are built from. Recall from the strengthening rung how solute atoms pin gliding dislocations by cluttering the lattice — those same foreign atoms clutter the path of the electrons, scattering them and slowing the heat. Pure iron sits near 80 W/m-K; stainless steel, iron loaded with chromium and nickel, drops all the way to about 15. It is the very same clutter that also raises electrical resistivity, so Wiedemann-Franz still holds — both conductivities fall together. The practical rule: if your job is to move heat, reach for pure copper or aluminium, not a clever alloy.
Non-metals: phonons, and why order and air matter
For phonons, the enemy is scattering. A vibration packet carries heat far and fast only if the lattice is stiff (so the waves travel quickly), made of light atoms (which vibrate faster), and orderly (so there is nothing for the phonon to bounce off). Meet all three and you get the great surprise of the whole table: diamond, at about 2000 W/m-K, out-conducts copper — yet diamond is an electrical insulator with no free electrons at all. Its bonds are ferociously stiff covalent bonds, its carbon atoms are light, and its lattice is nearly perfect, so phonons scream through almost unimpeded. The headline is worth stapling down: being a superb thermal conductor does not require free electrons, and 'metals conduct best' is a rule of thumb, not a law.
Now the opposite extreme: disorder. Glass is an amorphous solid — no long-range order, its atoms frozen in a jumble — so a phonon scatters every few atomic spacings and can hardly get anywhere; silica glass limps in near 1.4 W/m-K, far below the same silica in ordered crystalline quartz. Impurities, point defects, and grain boundaries scatter phonons too, so a dirty or fine-grained ceramic conducts less than a clean single crystal — the exact mirror image of how those same features strengthen a metal. Polymers are the same story in soft form: tangled, mostly amorphous chains, where heat can crawl along a covalent backbone but hops poorly from one chain to the next, leaving most plastics down around 0.2 W/m-K — natural insulators.
Which brings us to the champion insulators, and the twist is that they are barely solid at all. A polymer foam is a thin amorphous web of plastic enclosing millions of tiny pockets of trapped air. Gas is a dreadful conductor (still air is about 0.026 W/m-K) because its molecules are far apart and rarely collide to pass energy along; the plastic web's real job is simply to hold that air still so it cannot circulate and carry heat by convection. The result, near 0.03 W/m-K, is barely above air itself. It is the same trick everywhere you find warmth: wool, goose down, aerogel, the gap in double glazing — all of them are ways to trap still gas. Be honest about the mechanism: it is the trapped air doing the insulating, and the solid is mostly just the cage that keeps it from moving.
Why metals feel cold, and where this leads
Here is the everyday payoff. A metal doorknob and the wooden door around it are at exactly the same temperature — room temperature — yet the metal feels cold and the wood does not. Why? 'Feeling cold' is not your skin measuring temperature; it is your skin measuring how fast it is losing heat. The metal's high conductivity whisks warmth out of your finger fast, your skin cools, and the nerves report 'cold'. The wood, a poor conductor, barely draws any heat, your skin stays warm, and it reports 'warm'. You are feeling a flow rate, not a temperature. The same reasoning is a safety warning in reverse: a 60 degrees C metal bench burns you on contact while 60 degrees C wood is merely warm, because the metal delivers its heat to your skin so much faster. (The precise quantity for that first touch is effusivity, the square root of k times density times specific heat, blending conductivity with capacity — but conductivity is the star of it.)
Engineers work both ends of the ladder on purpose, and the next guides make a trade of it. When the job is to move heat away, you reach for high-k metals: the aluminium fins of a heat sink, the copper spreader under a computer chip. When the job is to keep heat out, you reach for low-k materials: foam in a wall, and the ceramic thermal-barrier coating sprayed onto a turbine blade — a thin layer of low-conductivity zirconia that lets the blade run in gas hotter than the bare metal could ever survive. And low conductivity has a sting in its tail: it is exactly what makes ceramics vulnerable to the next guide's problem, thermal shock. When heat cannot spread fast enough to even out the temperature across a part, one region expands while another does not, and the resulting stress can crack a brittle ceramic outright. Conductivity, expansion, and fracture are about to collide.