Polarons: Electrons Dressed in Their Own Distortion

The core idea: a self-consistent hole an electron digs for itself

Put a charge −e at the origin of a polar crystal. Its Coulomb field polarizes the medium: positive ions creep toward it, negative ions away. That rearrangement is a real, physical distortion of the lattice, and it costs elastic energy but lowers the electrostatic energy of the electron. The net gain binds the electron to its own distortion — the two move together as a single quasiparticle, the polaron.

The crucial subtlety is self-consistency. The distortion is created by the electron, and the electron's wavefunction is in turn shaped by the potential well of that distortion. Solving one requires the other, exactly like a marble sitting in the dent it makes in a stretched rubber sheet. Because the lattice can only respond as fast as its ions can move (the phonon frequency ω_LO, typically 10¹³–10¹⁴ rad/s), a slow-moving electron drags a fully-formed cloud, while a fast one outruns it and stays nearly bare.

• Only the ionic (lattice) polarization matters. The electronic polarizability responds instantly and is already folded into the high-frequency permittivity ε_∞. It is the sluggish, lattice part — the difference 1/ε_∞ − 1/ε₀ — that traps the carrier.
• This is why polarons live in polar and deformable solids — ionic crystals, oxides, polar semiconductors, molecular and organic crystals — but are absent in a non-polar, rigid lattice like pure silicon, where the effect is negligibly small.

The Fröhlich Hamiltonian and the coupling constant α

The standard model is the Fröhlich Hamiltonian (1954), which couples a band electron of mass m* to the longitudinal-optical (LO) phonon mode of frequency ω_LO:

• H = p²/2m* + Σ_q ℏω_LO a†_q a_q + Σ_q (V_q e^{iq·r} a_q + h.c.)
• where p is the electron momentum, a†_q creates an LO phonon of wavevector q, and the coupling amplitude scales as |V_q| ∝ 1/q — the hallmark long-range signature of Coulomb coupling to a polar mode.

All the material physics collapses into one dimensionless number, the Fröhlich coupling constant:

• α = (e²/4πε₀ℏ)·√(m*/2ℏω_LO)·(1/ε_∞ − 1/ε₀)
• Here ε_∞ is the high-frequency (optical) relative permittivity, ε₀ (in this context) the static one, and e, ε₀-vacuum, ℏ carry their usual meaning. Large α means strong dressing.

Plugging in real numbers: GaAs (m* = 0.067mₑ, ℏω_LO = 36.2 meV, ε_∞ = 10.9, ε_static = 12.9) gives α ≈ 0.07 — barely dressed. NaCl gives α ≈ 3.7, and SrTiO₃ lands near α ≈ 2. The full range across solids runs from ~0.02 to ~5, spanning the weak-, intermediate- and strong-coupling regimes.

How heavy and how big: mass enhancement and polaron radius

In the weak-coupling limit (α ≪ 1), second-order perturbation theory gives clean, memorable results:

• Self-energy: the ground state is lowered by ΔE = −α ℏω_LO. For GaAs this is only −0.07 × 36.2 meV ≈ −2.6 meV, a tiny binding.
• Effective mass: m*_polaron = m*/(1 − α/6). GaAs is enhanced by ~1.2%; an α ≈ 3 crystal already sits well outside this formula and can more than double the mass.

Feynman's 1955 all-coupling path-integral treatment interpolates smoothly across the whole range and remains the benchmark. In the strong-coupling limit (α ≳ 6) the mass enhancement grows dramatically, roughly m*_pol/m* ∝ α⁴, and the polaron localizes.

The characteristic size of a large (Fröhlich) polaron is set by the phonon-limited length scale:

• r_p = √(ℏ / 2m*ω_LO)
• For GaAs this gives r_p ≈ 4 nm — many lattice constants (a ≈ 0.565 nm), which is exactly why the continuum Fröhlich model is valid there. When r_p shrinks to about one unit cell, the continuum picture breaks and you enter small-polaron territory.

Two species: the large Fröhlich polaron and the small Holstein polaron

Polarons come in two physically distinct flavors, and confusing them is the most common error.

The large (Fröhlich) polaron spreads over many unit cells. Its distortion is gentle and continuous, the lattice is treated as a polarizable dielectric, and the carrier stays in an itinerant Bloch-like state with a modestly heavier mass. Transport is band-like: mobility decreases with rising temperature because more phonons scatter the carrier. GaAs, CdTe, SrTiO₃ and the halide perovskites host large polarons.

The small (Holstein) polaron, from Theodore Holstein's 1959 molecular-crystal model, is a short-range coupling to a local deformation. The carrier digs a well deep enough to self-trap in a single site, becoming enormously heavy — the mass grows exponentially, m* ∝ exp(g²) with dimensionless coupling g. Motion is then thermally-activated hopping from site to site:

• μ ∝ (1/T) · exp(−E_a / k_BT), with activation energy E_a typically 0.1–0.5 eV
• so mobility increases with temperature — the opposite temperature dependence, and a smoking-gun experimental signature. Rutile TiO₂, NiO and many transition-metal oxides, plus most organic/polymer semiconductors, form small polarons.

How we measure them, and where they matter

Polarons are not abstractions — they leave clear fingerprints:

• Cyclotron resonance directly measures the enhanced mass; in the 1970s it confirmed Fröhlich-polaron mass enhancement in AgBr and AgCl to within a few percent, including the resonant coupling near ω_c ≈ ω_LO.
• Optical absorption shows a mid-infrared polaron band from transitions that shake off phonons; small-polaron oxides display a characteristic broad, asymmetric peak.
• ARPES (angle-resolved photoemission) resolves phonon replica bands and the mass renormalization directly, spectacularly in oxide surfaces like SrTiO₃ and anatase TiO₂.

Why care? Polarons set the ceiling on carrier mobility in the technologically hottest materials of the last decade:

• Halide perovskites (e.g. CH₃NH₃PbI₃): large-polaron formation is now thought to protect photo-carriers from scattering, helping explain the remarkably long carrier lifetimes (~microseconds) and diffusion lengths (>1 μm) behind >25% efficient solar cells.
• Oxide electronics: SrTiO₃'s polaronic carriers, its dilute superconductivity, and its high permittivity all trace to strong electron-phonon coupling.
• Organic semiconductors and OLEDs: charge transport is polaron hopping; device mobilities of 0.1–10 cm²/V·s are governed by polaron binding energies.

Bipolarons, and the connection to superconductivity

If two electrons each dress themselves in lattice distortion, can they share one well and bind? Yes — under strong enough coupling and weak enough Coulomb repulsion, two polarons pair into a bipolaron, a boson with charge −2e. This is more than a curiosity: it is one candidate route to superconductivity. In conventional BCS superconductors the electron-phonon interaction glues Cooper pairs in the weak-coupling limit; bipolaron theories push into the strong-coupling regime where real-space pairs could, in principle, Bose-condense.

The debate over whether bipolarons drive high-Tc cuprate superconductivity ran for decades and remains partly open, but the physics is real and measurable. Bipolarons appear in materials like Ti₄O₇ and are actively studied in the doped perovskites. The key competition is always the same tug-of-war:

• Phonon-mediated attraction (∝ the polaron binding energy) tries to bind the pair,
• against the bare Coulomb repulsion (∝ e²/4πε₀r) pushing the like charges apart.

The winner depends on α, the ratio of static to optical permittivity, and the carrier density.

Subtleties and common misconceptions

A polaron is not a bound electron-ion state. Nothing is chemically captured; the electron remains free to move through the whole crystal, just heavier and slower because it must drag its phonon cloud. Remove the electron and the distortion relaxes away.

It is not the same as an exciton or a defect. An exciton is a bound electron-hole pair; a polaron is a single charge plus its lattice cloud, and forms in a perfect, defect-free crystal. Trapping at a defect is a different mechanism, though real materials often blend the two.

The two ε's must be different for a Fröhlich polaron to exist. If ε_∞ = ε₀ the coupling term (1/ε_∞ − 1/ε₀) vanishes: no ionic polarization, no polaron. This is precisely why covalent, non-polar crystals show essentially none of the effect.

• Temperature dependence is diagnostic: mobility falling with T ⇒ band-like large polaron; mobility rising with T (activated hopping) ⇒ small polaron. Watch the sign of dμ/dT.
• Holes dress too. A positive carrier in a polar lattice forms an equally valid polaron; hole-polaron masses matter as much as electron ones in p-type oxides.
• α is a material constant, not a universal number. Always quote it with the material — an α of 0.07 (GaAs) and 3.7 (NaCl) describe qualitatively different physics.

Two limits of the same object: the extended Fröhlich (large) polaron versus the self-trapped Holstein (small) polaron.

Property	Large (Fröhlich) polaron	Small (Holstein) polaron
Spatial extent	Many unit cells (~1–10 nm)	One unit cell (self-trapped)
Dominant coupling	Long-range Coulomb to LO phonons	Short-range deformation potential
Mass enhancement	Modest, m*/(1−α/6); α ≲ 6	Exponentially large, ∝ exp(g²)
Transport	Band-like, mobility falls with T	Thermally-activated hopping, rises with T
Example materials	GaAs, CdTe, SrTiO₃, halide perovskites	TiO₂ (rutile), NiO, many oxides & polymers
Discoverer/model	Fröhlich Hamiltonian (1954)	Holstein molecular-crystal model (1959)