Electrons and Holes: Defects Without a Missing Atom
In this rung so far, every defect has been about atoms: an ion gone missing (a vacancy), an ion shoved into a gap (an interstitial), a foreign ion sitting in for the host (a dopant). But a crystal can also be defective in its charge alone, with no atom out of place. Warm any solid and now and then an electron gets kicked up across the energy gap into the empty band above, where it can roam — and the empty seat it leaves behind in the full band below also moves, behaving for all the world like a positive particle. That freed electron and that empty seat, a hole, are the two electronic defects, and they are the missing half of the point-defect story you have been building.
ENERGY (up) | | ============================ CONDUCTION BAND (empty when cold) | e- e- <- electrons : n-type carriers | ---------------------------- Ec | (D) donor level ~0.1 eV O-vacancy / reduced cation | | (F) colour centre mid-gap electron trapped in a vacancy | --> absorbs a visible photon, so the crystal looks coloured | | (A) acceptor level ~0.1 eV cation vacancy / oxidised cation | ---------------------------- Ev | h+ h+ <- holes : p-type carriers | ============================ VALENCE BAND (full when cold) | band gap Eg : Al2O3 ~9 eV (insulator) TiO2 ~3 eV NiO / Fe-oxides small
How many of these appear is set by one number: the energy gap they must jump, the band gap Eg. Making an electron-hole pair costs Eg, so their numbers follow the same Boltzmann law as the ionic defects — the count of electrons n equals the count of holes p, and each falls off as exp(-Eg/(2 times k times T)), the identical form you met for Schottky pairs, obeying the same law of mass action with n times p held to a constant. The gap is the whole story: in alumina Eg is near 9 eV, so at any ordinary temperature the pair count is unimaginably tiny and Al2O3 is a superb insulator; in rutile TiO2 the gap is only about 3 eV, and in the small-gap transition-metal oxides it is smaller still, which is the first reason some ceramics can behave as a semiconductor.
The Atmosphere Sets the Carriers
In real oxides, though, the biggest source of electrons and holes is rarely that clean band-to-band jump. It is nonstoichiometry — the drift in composition with atmosphere you met in guide 4 — because the atoms that leave or arrive drop off charge as they go. Fire TiO2 in a low-oxygen furnace and it gives up a little oxygen to the gas, written in Kroger-Vink as O(O)x -> V(O)** + 2 e' + 1/2 O2(g): each departing oxygen leaves behind an oxygen vacancy and, crucially, two electrons that settle onto neighbouring titanium ions, turning Ti4+ into Ti3+. The crystal is now oxygen-deficient TiO2-x, stuffed with electrons, an n-type conductor — and it has visibly darkened to blue-black.
- Look at the atmosphere. A low-oxygen (reducing) furnace tends to pull oxygen OUT of the oxide; a high-oxygen (oxidizing) one tends to push oxygen IN. This one fact decides the sign of the carriers.
- Reducing: write O(O)x -> V(O)** + 2 e' + 1/2 O2(g). Oxygen leaves, electrons stay behind on the cation, so the oxide turns n-type; the electron count climbs as the oxygen pressure falls, as n proportional to pO2^(-1/6).
- Oxidizing: write 1/2 O2(g) -> O(O)x + V(M)'' + 2 h*. Oxygen enters and builds new lattice, leaving cation vacancies and holes, so the oxide turns p-type; the hole count climbs as the oxygen pressure rises, as p proportional to pO2^(+1/6).
- Check where the charge lands. If the host cation has a spare valence state — Ti4+/Ti3+, Ni2+/Ni3+, Fe2+/Fe3+ — the electron or hole sits on it as a changed oxidation state, and it will conduct by hopping: the small polaron of the next section.
Plot the conductivity of such an oxide against oxygen pressure and you get a lopsided U: an n-type branch sloping down as pO2 rises on the left, a flat near-intrinsic plateau in the middle where ionic or fixed-dopant defects rule, and a p-type branch climbing on the right. That U is simply the electronic reading of the Brouwer diagram from guide 1, its slopes handed to you by the mass-action exponents. NiO is the textbook case: its conductivity really does rise as pO2^(+1/6) as you oxidize it, holes hopping on Ni3+. Be honest about the exponent, though — whether it is 1/6 or 1/4 depends on how many times the vacancy is ionized, so the slope is a diagnostic of the dominant defect, not a universal constant.
Small Polarons: An Electron That Drags Its Cage
Why do those electrons hop rather than fly? In a silicon crystal an extra electron races through wide, gently-bonded bands almost like a free particle. In an ionic oxide it cannot: drop an electron onto a titanium ion and its negative charge shoves the surrounding O2- ions outward, digging a small bowl of lattice distortion — and the electron then sits trapped in the very dent it made. That package of one electron plus the distortion it carries is a small polaron (see polaron). Picture a heavy ball on a trampoline, forever resting in the dip it presses into the fabric; to go anywhere it must climb out and press a fresh dip next door.
That climb-and-press is the key to how these oxides conduct. The polaron moves by hopping to the next cation, dragging its distortion along, and each hop needs a thermal kick to get over the barrier — so the conductivity rises with temperature, sigma proportional to (1/T) times exp(-Eh/(k times T)) with a hopping energy Eh of maybe 0.1 to 0.5 eV. That is exactly backwards from a metal, which conducts worse when it gets hot. The hopping highway is a mixed-valence pair on neighbouring sites: an electron trading between Fe2+ and Fe3+ makes magnetite one of the more conductive oxides, and the same Ti3+/Ti4+ or Ni2+/Ni3+ shuffles carry current in reduced titania and oxidized nickel oxide.
This thermally-activated hopping is not a curiosity — it is a product you can buy. An NTC thermistor is a little bead of transition-metal-oxide spinel (manganese, nickel, and cobalt oxides) whose resistance drops steeply and predictably as it warms, and it is the temperature sensor tucked into nearly every battery pack, engine, and thermostat. Be honest about the trade-off, though: polaron mobilities are tiny next to silicon's, so these are mediocre conductors in absolute terms. What makes them useful is not high conductivity but the steep, reliable way that conductivity changes with temperature.
Color Centers: Trapped Charge You Can See
This is the most vivid electronic defect of all — one you can see with your own eyes. Trap an electron not on a cation but inside an empty anion seat — an oxygen or halide vacancy — and it behaves like an electron in a tiny box, its allowed energy levels spaced just right to absorb a photon of visible light. Absorb the green and the crystal transmits its complement and glows purple; absorb the blue and it looks amber. This trapped-charge defect is a color center, and the classic version — an electron in an anion vacancy — is called an F-center, from the German Farbe, colour. A water-clear crystal turns a deep colour with not one atom of its composition changed; only where the charge sits has moved.
Color centers are born two ways. Heat NaCl in sodium vapour and the extra metal forces electrons into anion vacancies — the crystal comes out amber; this is additive coloration. Or bombard a crystal with radiation, which knocks electrons loose to be caught in traps: the brown of smoky quartz is a hole trapped at an aluminium impurity, created by natural radioactivity over geological time, and amethyst's purple is an iron-based color center made the same way. Because the charge is only trapped, not chemically bonded, the process reverses — a little heat or bright light can bleach a color center back to clear, which is exactly why heating amethyst turns it to yellow citrine. This makes color centers the working principle of radiation dosimeters, and a real headache for optical windows and lenses in reactors and in space, which steadily darken under the dose.
From Defect to Device, and On to Diffusion
Pull the camera back and the payoff of this whole rung comes into view: by planting electronic defects you turn an inert ionic ceramic into a working n-type or p-type material. The polaron-hopping thermistor measures temperature; a zinc-oxide varistor built from n-type ZnO grains switches on to swallow a lightning surge and protect the circuit behind it; a phosphor doped with the right activator ion turns an electron-hole recombination into a photon of coloured light; and an oxygen sensor reads the atmosphere straight off the nonstoichiometry we traced above. Each device is just a chosen electronic-defect population, set by dopant and atmosphere, doing useful work.
And a bridge forward. The electrons and holes of this guide ride on top of the ionic defects, but it is those ionic vacancies underneath that let whole atoms change seats. The next rung follows exactly that: diffusion, the slow migration of atoms through the lattice by way of the vacancies you now understand, which is the engine of every high-temperature reaction and, above all, of sintering — the welding-shut of pores that turns loose powder into a dense part without ever melting it. Everything electrical you just met, and everything that will densify in a furnace, traces straight back to the point defects of this rung.