What Are Relaxor Ferroelectrics? Polar Nanoregions, Ergodic and Non-Ergodic States Explained

Introduction — What Makes a Ferroelectric a "Relaxor"?

Relaxor ferroelectrics are a special class of ferroelectric materials that show anomalous dielectric and polarization behaviour compared with conventional ferroelectrics. In a normal ferroelectric, there is a well-defined phase transition from the ferroelectric to the paraelectric state at a single temperature — the Curie temperature (T_C) — which appears as a sharp peak in the dielectric constant versus temperature plot (Figure 1). In relaxor ferroelectrics, this transition is not sharp but spread over a range of temperatures, a phenomenon known as a diffuse phase transition (DPT).

Why does this diffuseness arise? Because the crystal structure of a relaxor is chemically disordered, producing regions with different local environments. Instead of switching uniformly at one temperature, different regions respond at different temperatures, giving rise to a broad dielectric peak. A second defining feature is the frequency-dependent dielectric response (Figure 2): the temperature of maximum permittivity (T_m) shifts to higher values as the measurement frequency increases — something never observed in conventional ferroelectrics.

Conventional vs. Relaxor Ferroelectrics

In conventional ferroelectrics such as BaTiO₃ or PbTiO₃, the ferroelectric-to-paraelectric transition occurs sharply at the Curie temperature. The dielectric constant (ε′) shows a narrow, frequency-independent peak at T_C, indicating a well-defined structural change from a polar to a non-polar phase. (If the perovskite ABO₃ structure of these materials is new to you, revise it first — the same lattice hosts most relaxors.)

In contrast, relaxor ferroelectrics such as Pb(Mg_1/3Nb_2/3)O₃ (PMN) or Bi_0.5Na_0.5TiO₃ (BNT) show a diffuse transition spread over a wide temperature range. Their permittivity exhibits a broad maximum at T_m, and T_m itself shifts with measurement frequency. This tells us that polarization dynamics in relaxors are not uniform — different microscopic regions respond to electric fields on different timescales.

Why Are They Called "Relaxors"? Two Experimental Signatures

The name comes from their relaxation-type dielectric response, which resembles dipolar relaxation in glassy systems. Two experimental signatures justify the name.

Signature 1 — The Diffuse Phase Transition

Instead of a sharp ferroelectric–paraelectric transition, relaxors show a broad dielectric peak spread across a temperature range.

Signature 2 — Frequency-Dependent Permittivity

The dielectric constant changes with measurement frequency, indicating relaxation dynamics of dipoles rather than the switching of long-range-ordered ferroelectric domains.

How to Read the ε′–T Plot

Three observations summarise relaxor behaviour on the ε′–T plot. First, the normal ferroelectric (the sharp, tall curve) has one frequency-independent peak at T_C. Second, the relaxor curves show a broad maximum at T_m that shifts to higher temperature as frequency increases — the experimental hallmark of a relaxor. Third, polar nanoregions form below the Burns temperature (T_B) and progressively freeze below the freezing temperature (T_f), which is the microscopic origin of the relaxation-type response.

Polar Nanoregions: The Microscopic Origin of Relaxor Behaviour

The frequency-dependent dielectric behaviour of relaxors is attributed to polar nanoregions (PNRs) — nanometre-sized clusters of correlated dipoles that form below the Burns temperature and govern the material's relaxation dynamics.

Where Do PNRs Come From?

PNRs emerge because of chemical disorder and random electric fields within the perovskite lattice (ABO₃). In Pb(Mg_1/3Nb_2/3)O₃, for example, the unequal charges of Mg²⁺ and Nb⁵⁺ ions sharing the B-site create local fields that favour short-range polar ordering. Although each PNR is polar, the material as a whole remains macroscopically non-polar because the PNRs are randomly oriented and distributed throughout the crystal. Crucially, the system never develops long-range ferroelectric order on its own — it is the size, dynamics, and interaction of the PNRs that evolve with temperature.

Temperature Evolution of PNRs

At T_B, PNRs first appear inside the otherwise non-polar crystal. On cooling, these nano-polar clusters grow and start interacting, and near the freezing temperature T_f they become locked in place — producing the non-ergodic relaxor state described in the next section.

Ergodic and Non-Ergodic States of Relaxor Ferroelectrics

Up to this point, we have seen that the unusual dielectric response of relaxors arises from the dynamic interactions of PNRs within the crystal lattice. As temperature changes, the PNRs evolve from a dynamic to a frozen condition, giving rise to two distinct physical states — the ergodic and non-ergodic states. These states define how local polarization responds to external electric fields and to time, and they ultimately control the macroscopic dielectric and ferroelectric properties of the material.

Ergodic Relaxor State — The Dynamic Phase

The ergodic state is observed between the Burns temperature and the freezing temperature (T_f < T < T_B). Here the PNRs are highly dynamic: nanosized regions of polarization continuously form, grow, and dissolve within the non-polar matrix. Their dipoles reorient easily under an external electric field, so the system can explore all its configurations and reach thermodynamic equilibrium — which is precisely what "ergodic" means in statistical mechanics.

Because of this dynamic behaviour, the dielectric response in the ergodic state is strongly frequency-dependent: the higher the measurement frequency, the higher the temperature at which ε′ peaks. This is why relaxors display a broad dielectric maximum at T_m rather than the sharp peak of a conventional ferroelectric. The frequency dispersion of T_m follows the empirical Vogel–Fulcher relation:

The divergence of the relaxation time as T_m approaches T_f is direct evidence that PNR dynamics slow down and eventually freeze — behaviour closely analogous to structural glasses and spin glasses.

Non-Ergodic Relaxor State — The Frozen Phase

As the temperature drops below T_f, the PNRs lose mobility and eventually become frozen in random orientations. In this non-ergodic state, the system can no longer reach thermodynamic equilibrium within experimental timescales — it is trapped in one region of its configuration space. The dipoles are locked and cannot fully respond to external stimuli.

The result is behaviour resembling a dipolar glass: polarization becomes sluggish, and hysteresis appears in the polarization–electric field (P–E) loops. If a sufficiently strong electric field is applied, the material can be poled into a long-range-ordered, ferroelectric-like state with measurable remanent polarization. In many non-ergodic relaxors this field-induced state persists after the field is removed, and the disordered state is recovered only on heating above T_f — a signature of broken ergodicity.

Physically, this crossover from dynamic to frozen PNRs explains why relaxor ferroelectrics combine the field-responsiveness of a paraelectric phase with the memory effects of a ferroelectric phase — the very combination that makes them ideal for high-performance actuators, sensors, and capacitors.

Why It Matters: Scientific and Technological Significance

Understanding ergodic and non-ergodic behaviour is not just of academic interest — it is central to designing high-performance functional materials. By controlling the degree of chemical disorder and the interaction among PNRs, researchers can tune dielectric, ferroelectric, and piezoelectric properties for specific applications. Two examples illustrate this:

• PMN–PT and PZN–PT single crystals near their morphotropic phase boundary (MPB) exhibit giant piezoelectric coefficients, owing to the delicate balance between relaxor (short-range) and ferroelectric (long-range) polar order.
• Bi-based relaxors such as BNT–BT are among the most promising lead-free alternatives for environmentally friendly actuators, sensors, and energy-storage capacitors, where a field-induced ergodic-to-ferroelectric transition delivers large recoverable strain. Doping strongly modulates this behaviour: donor and acceptor substitution in BNT ceramics shifts the frequency-dependent ferro–antiferro transition and introduces internal bias fields that reshape the piezoelectric response [11].

Do Relaxor Ferroelectrics Occur Only in Perovskite Lattices?

No — relaxor behaviour is not restricted to perovskites, although the best-known relaxors (PMN, PZN, and their solid solutions) are perovskite-type oxides with the general formula ABO₃. The perovskite structure is an ideal host because it permits B-site cation disorder (for example Mg²⁺/Nb⁵⁺ in PMN), supports local charge imbalance and random electric fields, and thereby facilitates the formation of PNRs below the Burns temperature. (Dimensional and size effects add a further layer to this story in thin films and nanostructures — see the discussion of dimensional effects in ferroelectricity in [12].)

However, relaxor-like phenomena have also been reported in several non-perovskite systems:

• Aurivillius-type layered oxides, e.g. Bi₄Ti₃O₁₂-based compounds and La-modified SrBi₂Nb₂O₉;
• Tungsten–bronze structured ferroelectrics, e.g. Sr_xBa_1−xNb₂O₆ (SBN);
• Pyrochlore- and fluorite-type oxides; and
• Polymeric and organic ferroelectrics, e.g. irradiated or terpolymer PVDF-based systems, which show dipolar disorder and frequency-dependent dielectric relaxation.

Relaxor behaviour is therefore a physical phenomenon — arising from dipolar frustration, local disorder, and nanoscale polarization dynamics — not a property confined to one lattice type.

Future Research Directions

Current research focuses on resolving the atomic-scale origin and dynamics of PNRs using neutron and X-ray diffuse scattering, piezoresponse force microscopy (PFM), pair distribution function (PDF) analysis, and first-principles simulations. Field-induced phase transitions, temperature–frequency scaling, and lead-free relaxor systems remain highly active areas. Increasingly, machine-learning-assisted materials design is accelerating the discovery of relaxor compositions optimised for dielectric energy storage, electrocaloric solid-state cooling, and next-generation actuator technologies.

Practice Questions

Test your understanding with the following questions. Every correct answer is grounded in the content above — useful revision for viva, GATE, and CSIR-NET preparation.

Key Takeaways

Frequently Asked Questions (FAQs) on Relaxor Ferroelectrics

References

All references are in IEEE citation style. All sources are peer-reviewed journals or authoritative preprint archives.

1. A. A. Bokov and Z.-G. Ye, "Recent progress in relaxor ferroelectrics with perovskite structure," J. Mater. Sci., vol. 41, no. 1, pp. 31–52, 2006, doi: 10.1007/s10853-005-5915-7. [doi.org] — The definitive review of relaxor phenomenology: diffuse transitions, PNRs, ergodic/non-ergodic states.
2. M. E. Manley, J. W. Lynn, D. L. Abernathy, E. D. Specht, O. Delaire, A. R. Bishop, R. Sahul, and J. D. Budai, "Phonon localization drives polar nanoregions in a relaxor ferroelectric," Nat. Commun., vol. 5, Art. no. 3683, 2014, doi: 10.1038/ncomms4683. [doi.org] — Neutron-scattering evidence for the lattice-dynamical origin of PNRs.
3. H. Liu, X. Shi, Y. Yao, et al., "Emergence of high piezoelectricity from competing local polar order–disorder in relaxor ferroelectrics," Nat. Commun., vol. 14, Art. no. 1007, 2023, doi: 10.1038/s41467-023-36749-w. [doi.org] — Modern picture of how local order–disorder competition produces giant piezoelectricity.
4. L. Cai, R. Pattnaik, J. Lundeen, and J. Toulouse, "Piezoelectric polar nanoregions and relaxation-coupled resonances in relaxor ferroelectrics," arXiv, 2018, arXiv:1801.02655. [arxiv.org] — PNR-mediated electromechanical response.
5. H. Takenaka, I. Grinberg, and A. M. Rappe, "Anisotropic local correlations and dynamics in a relaxor ferroelectric," Phys. Rev. Lett., vol. 110, Art. no. 147602, 2013. [arxiv.org] — First-principles molecular dynamics of PNR correlations.
6. L. Cai, J. Toulouse, L. Harriger, R. G. Downing, and L. A. Boatner, "Origin of the crossover between a freezing and a structural transition at low concentration in the relaxor ferroelectric KTa₁₋ₓNbₓO₃," Phys. Rev. B, vol. 91, no. 13, Art. no. 134106, 2015, doi: 10.1103/PhysRevB.91.134106. [doi.org] — Freezing versus structural transition in a model relaxor.
7. V. V. Shvartsman and D. C. Lupascu, "Lead-free relaxor ferroelectrics," J. Am. Ceram. Soc., vol. 95, no. 1, pp. 1–26, 2012, doi: 10.1111/j.1551-2916.2011.04952.x. [doi.org] — Comprehensive review of BNT-based and other lead-free relaxor systems.
8. S. Huang, L. Sun, C. Feng, and L. Chen, "Relaxor behavior of layer-structured SrBi₁.₆₅La₀.₃₅Nb₂O₉," J. Appl. Phys., vol. 99, no. 7, Art. no. 076104, 2006, doi: 10.1063/1.2186975. [doi.org] — Relaxor behaviour in a non-perovskite (Aurivillius) lattice, cited in Section 7.
9. R. Verma and S. K. Rout, "Frequency-dependent ferro–antiferro phase transition and internal bias field influenced piezoelectric response of donor and acceptor doped bismuth sodium titanate ceramics," J. Appl. Phys., vol. 126, no. 9, Art. no. 094103, Sep. 2019, doi: 10.1063/1.5111505. [doi.org] — Author's research on doped BNT relaxor ceramics, cited in Section 6.
10. R. Verma and S. K. Rout, "The Mystery of Dimensional Effects in Ferroelectricity," in Recent Advances in Multifunctional Perovskite Materials. London, UK: IntechOpen, 2022, doi: 10.5772/intechopen.104435. [doi.org] — Author's chapter on size and dimensional effects in ferroelectric/perovskite systems, cited in Section 7.

Series Navigation — Ferroelectrics & Perovskite Hub