What is the difference between ferroelectric and piezoelectric?

[NAP2626]

Hey All,

This is a question I have faced many times, but I never fully grasp the difference.

All ferroelectrics are piezoelectric, but not all piezoelectrics are ferroelectric...

I understand that ferroeletrics are due to their noncentrosymetric crystal structure and they can therefore have a spontaneous polarization.

And Piezoelectrics can generated voltage and reverse their polarization about the application of a stress.

But I don't understand the relationship mentioned above. How can a piezoelectric not be ferroelectric as well. Quartz is an example of this. I believe this is because of the crystal structure and ion charge balance present in the crystal structure. Therefore quartz will NOT have spontanous polarization, but if a stress is applied to its crystal structure a charge potentail can be created.

Any insight how to clearly or better understand this and explain the difference would be greatly apperciated!

 

[adrianpoll]

With NAP2626's comments I think we are getting close. This is helping me to remember now from acoustics school:

Piezoelectricity gives a linear relationship between stress and voltage.
Whereas Ferroelectricity gives fundamentally a quadratic relationship.
That is why ferroelectric ceramics have to be "poled in": they end up operating along a short, therefore approximately linear stretch on a much bigger parabolic curve.

So I like what NAP2626 says about piezoelectric crystals not actually having an electric polarization in the unstressed state. They don't have polarization; it's just that their lattice lacks a center of symmetry. This is the reason underlying their piezoelectric behavior and the criterion for a material to show piezoelectricity. But this effect as we all know is very weak compared to ferroelectricity.

This makes me wonder, is it a grayscale transition from the one to the other or is there a fundamental difference ? What if you had a lattice with just the tiniest tendency to a dipole offset, would that make it ferroelectric ? If I understand it right, the critical difference is when that tiny bit of eccentricity in the single cell gets to be swept up in a long range interaction or long range order, so that a whole chunk of lattice has the tiny electric dipole moments of all the lattice cells all aligned instead of haphazard. That is the essence of what makes it ferroelectric. This is how ferroelectricity is analogous to ferromagnetism: in ferromagnetism there is a long range order such that the spins (magnetic fields) of certain of the electrons in the individual atoms get to be aligned everywhere within the boundaries of the ferromagnetic domain. So indeed, we can speak of ferroelectric domains and we can discuss the polarization process in terms of the movements of ferroelectric domain boundaries, just as in ferromagnetism.