the Curie point
/ KYOOR-ee point /
Heat a ferroelectric high enough and its magic simply switches off: the built-in polarization vanishes, the giant permittivity collapses toward ordinary values, and it becomes a plain insulator. Cool it back down and the ferroelectricity returns. The temperature marking that on/off boundary is the Curie point, named after Pierre Curie who first studied the analogous effect in magnets. It is the single most important temperature on any ferroelectric or piezoelectric datasheet.
At the Curie point the crystal undergoes a change of symmetry. Below it the structure is distorted and polar — in barium titanate, tetragonal with the titanium ion off-centre, carrying a spontaneous polarization. At the Curie point thermal agitation wins and the structure snaps to a higher-symmetry, non-polar form (cubic, paraelectric) in which the ion sits centred and no spontaneous polarization can exist. Right at the transition the dielectric constant peaks sharply — the lattice is on a knife-edge and enormously easy to polarize — and just above it the permittivity falls off following the Curie-Weiss law, epsilon_r = C / (T minus T0), where C is the Curie constant and T0 a temperature close to the Curie point. Typical values: barium titanate about 130 degrees C, PZT around 350 degrees C, lead titanate about 490 degrees C.
The Curie point sets the usable temperature window of a device. A poled piezoelectric or a ferroelectric memory must operate well below its Curie point; approach it and the polarization weakens and the part can spontaneously depole, losing its function — which is why a PZT sonar transducer with a 350-degree Curie point is safe in warm water but a soldering iron will ruin it. Composition is chosen to place the Curie point where it is needed: high for a high-temperature actuator, or deliberately shifted and broadened for a temperature-stable capacitor. An honest caveat: exactly at the Curie point the permittivity is spectacular but wildly unstable with temperature, so capacitor designers formulate ceramics to sit beside the peak, never on it.
To wipe the slow capacitance drift out of an aged Class II ceramic capacitor, you simply heat it above the barium-titanate Curie point of about 130 degrees C and let it cool: the domains reset and the full room-temperature capacitance returns, restarting the aging clock.
Crossing the Curie point resets a ferroelectric's history — polarization, poling, and accumulated aging all vanish above it and re-form on cooling.
The Curie point is a ceiling, not a target. A device is designed to run below it with margin; sitting a capacitor right on the permittivity peak would give a huge but hopelessly temperature-unstable value.