Piezoelectricity and PZT Devices

A Crystal That Talks Both Ways

In the last guide you watched barium titanate become ferroelectric: below its Curie point the little Ti4+ ion slips off-centre inside its oxygen cage, giving the unit cell a permanent electric dipole, and a field can flip which way that dipole points — the switchable spontaneous polarization behind the hysteresis loop. Piezoelectricity is the close cousin of that same off-centre trick. If the crystal already carries a built-in dipole because its ions sit lopsided, then anything that moves those ions — a squeeze, a stretch, an applied field — must change the dipole too. Piezoelectricity is exactly that link: a two-way street between mechanical push and electrical charge.

Run the street one way and you have the direct effect: squeeze the crystal, shove its ions a little closer, and the net dipole shifts — so charge appears on the faces and a voltage builds across them. This is what fires the clicker on a gas lighter or a barbecue: a spring-loaded hammer whacks a small poled disc of PZT, the sudden stress produces a pulse of many kilovolts — tens of thousands of volts across a millimetre-thin disc — and that is more than enough to arc across an air gap and light the gas. No battery, no wire, just a mechanical blow turned straight into a spark.

Run the street the other way and you have the converse effect: apply a field, push the off-centre ions a hair further off-centre, and the whole crystal changes shape — it strains. The motion is tiny, a strain of a fraction of a thousandth, but it is exquisitely controllable: the shape follows the voltage almost perfectly and almost instantly. The beautiful economy of piezoelectricity is that a single material constant, the piezoelectric coefficient d, governs both directions at once — the charge you get per unit of force pressing in (measured in pC/N) is numerically the very same d as the displacement you get per unit of voltage applied (in pm/V). One crystal, one number, both a sensor and a motor.

It Takes a Lopsided Crystal

Why can only some crystals do this? The deciding property is symmetry — specifically, whether the crystal has a centre of symmetry. Picture a unit cell that is perfectly balanced: for every ion at some position there is an identical ion at the exact mirror-image position through the centre. Now squeeze it uniformly. Every ion shifts, yes — but its mirror partner shifts by the equal and opposite amount, so all the little dipole contributions cancel in pairs and the net polarization stays stubbornly zero. A centrosymmetric crystal simply cannot turn a squeeze into charge. Rock salt and plain cubic MgO are like this; squeeze them all day and nothing electrical happens.

Break that symmetry and everything changes. If the crystal is non-centrosymmetric — if there is no such mirror centre — then a squeeze shifts the positive and negative ions by different amounts, the cancellation fails, and a net dipole appears. This is precisely the situation in the distorted perovskite of the last guide: once the Ti4+ has slid off-centre, the BaTiO3 cell has lost its centre of symmetry, so it is guaranteed to be piezoelectric. The lopsidedness that made barium titanate ferroelectric is the very same lopsidedness that makes it (and PZT) piezoelectric. They come as a package.

  THE NESTED PIEZOELECTRIC FAMILY (32 crystal classes)

  [ 32 point groups : every crystal ]
     |
     +-- [ 21 NON-CENTROSYMMETRIC : no centre of symmetry ]
            |
            +-- [ 20 PIEZOELECTRIC : stress makes charge ]      <- quartz stops here
                   |
                   +-- [ 10 PYROELECTRIC : a built-in polar axis ]
                          |
                          +-- [ FERROELECTRIC : polarity can be SWITCHED ]  <- BaTiO3, PZT

  quartz       = piezoelectric only  (not pyro-, not ferro-)
  BaTiO3, PZT  = all the way in, so automatically piezoelectric too

That nested picture also settles a common confusion. Quartz is piezoelectric but not ferroelectric: it makes charge under stress, yet it has no switchable spontaneous polarization and no Curie point to lose it at — which is exactly why quartz is prized for the ultra-stable oscillators that keep time in watches and radios. Barium titanate and PZT, sitting in the innermost box, are ferroelectric, pyroelectric, and piezoelectric all at once. The middle box, pyroelectricity — a polarization that shifts when the crystal merely warms or cools — is the star of the next guide's infrared detectors, so keep it in view.

Poling: Waking the Ceramic Up

Here is a snag that single-crystal quartz never faces but every PZT device must. A fired ceramic is a mob of millions of tiny grains, each grain sliced into ferroelectric domains whose dipoles point whichever way the crystal happened to freeze. Add up all those dipoles pointing in every direction and they cancel to almost nothing: as it comes out of the furnace, a PZT disc has essentially zero net polarization and shows no useful piezoelectric effect at all. Each grain is willing; the crowd is just a disorganised jumble. To make a device we have to give the whole mob a common direction.

1. Heat the ceramic up, often to somewhere below but near its Curie point, where the domain walls are at their most mobile and easiest to move.
2. Apply a strong DC field across it — a few kV per millimetre — pointing along the axis you want the finished device to work on. This is the poling field.
3. Let the domains switch. Wherever a domain can reorient to line up better with the field, it does; the crowd swings toward one common direction.
4. Cool the ceramic while the field stays on, then remove the field. The alignment freezes in as a permanent remanent polarization — the mob now points, on average, one way.

Poling transforms the material. That frozen-in remanent polarization defines a unique axis — engineers call it the 3-direction — and along it the ceramic now behaves like one big polar crystal: squeeze it and charge appears, drive it and it strains. The ceramic has been woken up into a piezoelectric. But the same wake-up can be undone. Heat the part back above its Curie temperature and the polarization vanishes and the alignment is lost; hit it with a huge reverse field or a crushing stress and the domains scramble again. This depoling is a real service limit — it is why a PZT device carries a maximum operating temperature and a maximum drive, and why the higher Curie point of PZT (often 300 to 400 degrees C) is one of its practical advantages over plain barium titanate, whose polarization fades away above only about 130 degrees C.

PZT and the Morphotropic Sweet Spot

So why PZT rather than the barium titanate we started with? PZT — lead zirconate titanate, Pb(Zr,Ti)O3 — is a perovskite solid solution, a continuous blend between two end members, lead zirconate (PbZrO3) and lead titanate (PbTiO3). You can dial the Zr-to-Ti ratio anywhere from all-zirconium to all-titanium, and as you do, the crystal's preferred distortion changes: titanium-rich PZT freezes into a tetragonal form, its dipole pointing along a cube edge, while zirconium-rich PZT freezes into a rhombohedral form, its dipole pointing along a body diagonal. Somewhere between them the two structures must meet.

That meeting place, at roughly 52 percent zirconium to 48 percent titanium, is the morphotropic phase boundary (MPB), and it is where PZT becomes magical. Right at the boundary the tetragonal and rhombohedral phases coexist, so the polarization has an unusually rich menu of directions it is allowed to point — many nearly-equal choices at once. That makes the crystal extraordinarily easy to pole and easy to nudge, so the piezoelectric coefficient and the permittivity both spike to a sharp peak. The numbers tell the story: quartz has a d of about 2 pC/N, poled barium titanate about 190 pC/N, but PZT tuned to its MPB reaches 400 to 600 pC/N. Engineers then fine-tune with dopants — soft PZT (donor-doped) gives the highest d and sensitivity for sensors and small actuators, while hard PZT (acceptor-doped) trades some d for low loss and stability to survive high-power sonar and ultrasonic cleaning.

One honest caveat sits at the heart of PZT: it is roughly three-fifths lead oxide (PbO) by weight. Lead is toxic, its mining, firing, and disposal all carry an environmental cost, and modern regulations such as RoHS restrict lead in electronics. PZT survives today only under a specific exemption, granted because — bluntly — no lead-free ceramic yet matches it. A whole research field chases replacements: sodium-potassium niobate (KNN), bismuth sodium titanate, and other lead-free perovskites have closed much of the gap but still trail PZT in performance and stability. It is a genuine open frontier, and a reminder that the best-performing material is not always the one we would choose if we were starting clean.

From Igniters to Ultrasound

With a poled PZT element in hand, the two directions of the street become two whole families of devices. Run it in the direct effect — stress in, charge out — and you have sensors. A PZT accelerometer turns the tiny force of vibration into a readable voltage; a hydrophone in a sonar array listens for the faint pressure of a distant echo; a piezo microphone, a force gauge, a knock sensor in an engine, and the spark igniter we started with all work the same way. There is even a growing use in energy harvesting — a chip of PZT under a floor tile or on a vibrating machine skims usable electricity from motion that would otherwise be wasted.

Run the same element in the converse effect — voltage in, motion out — and you have actuators, prized for motion so fine and so fast that no gear or motor can match it. Here is the catch, and its cure. Take a good soft PZT with d33 = 500 pm/V. The thickness change is simply displacement = d33 x V, so at 100 V a single plate moves 500 pm/V x 100 V = 50,000 pm = 50 nm — real, but a rounding error for most machines. The fix is the exact trick from the capacitor guide: stack many thin poled layers and co-fire them into one block, a multilayer actuator, so their motions add. A hundred layers gives a hundred times the stroke — about 5 microns at 100 V — and that is enough to drive a diesel fuel injector, jet an ink droplet, nudge a telescope mirror, or position an atomic-force-microscope tip to the width of a single atom.

The richest devices use both directions of the street in the same instant, and these are the transducers behind sonar and medical ultrasound. A single poled PZT disc is pulsed with a burst of voltage; by the converse effect it snaps into motion and rings, launching a pulse of ultrasound into the water or the body. Then it falls silent and listens: the echo bouncing back off a submarine, a shoal of fish, or a beating heart presses on the very same disc, and by the direct effect that pressure becomes a faint voltage the machine reads out. Transmit by converse, receive by direct, all with one piece of ceramic — that elegant double duty is how a sonogram is made and how a ship sees in the dark of the deep sea.