Peierls Distortion

The Peierls distortion (also known as the Peierls transition or Peierls instability) is a fundamental quantum-mechanical theorem of solid-state physics and polymer chemistry formulated in 1955 by the German-British physicist Sir Rudolf Peierls (and discovered independently in the context of conjugated polyenes by H. Christopher Longuet-Higgins and L. Salem in 1959). Peierls proved that a hypothetical one-dimensional (1D) metallic crystal with a partially filled conduction band and equally spaced atoms is thermodynamically unstable at absolute zero (T = 0 K) with respect to a periodic static lattice distortion of wavevector q = 2k_F (where k_F is the Fermi wavevector).

In physical and chemical terms, the Peierls distortion is the macroscopic, infinite-chain solid-state analogue of the molecular Jahn–Teller effect and pseudo Jahn–Teller effect. In the simplest case of a half-filled 1D band (one conduction electron per atomic site, k_F = π / 2a), the uniform chain of lattice constant a spontaneously dimerizes into alternating short and long bonds (a − u, a + u), doubling the real-space unit cell to 2a and halving the first Brillouin zone to ±π / (2a) = ±k_F. This symmetry-breaking electron–phonon vibronic coupling opens a Peierls energy band gap (2Δ) right at the Fermi level, lowering the energies of all occupied electronic states below E_F, raising the unoccupied states above E_F, and transforming the 1D metal into a Peierls insulator or semiconductor accompanied by a periodic Charge Density Wave (CDW).

Interactive VSEPR Lab: Visualize square-planar Pt(II) coordination units and linear chain geometries in 3D inside the VSEPR Lab!

Half-Filled 1D Tight-Binding Chain and Zone-Folding at k_F = π / (2a)

Consider a uniform one-dimensional chain of N identical atoms (such as a hypothetical chain of hydrogen atoms, sodium atoms, or equidistant carbon π-radicals in polyacetylene) spaced by a uniform lattice constant a, with one valence electron per atom. In the Hückel or tight-binding approximation with nearest-neighbor resonance (hopping) integral t₀ (or β₀), the single-particle electronic dispersion relation is a continuous cosine band spanning bandwidth 4t₀:

E(k) = − 2 t₀ cos(k a)     for     −π / a < k ≤ +π / a

Because each spatial Bloch orbital k holds two electrons of opposite spin (↑ and ↓), one valence electron per atom fills the band exactly halfway, occupying all states between the Fermi points ±k_F = ±π / (2a) up to the Fermi energy E_F = E(±k_F) = 0. Notice that at k = +k_F and k = −k_F, the two states are exactly degenerate (E(+k_F) = E(−k_F) = 0) and separated in momentum space by the nesting wavevector q = 2k_F = π / a.

Now introduce a periodic dimerization in which every odd atom displaces by +u and every even atom displaces by −u, alternating the bond lengths between r_short = a − 2u and r_long = a + 2u. This periodic modulation of period λ_CDW = 2π / (2k_F) = 2a alternates the nearest-neighbor hopping integral between t₁ = t₀ + αu (across short bonds) and t₂ = t₀ − αu (across long bonds), where α = −(∂t/∂r)₀ is the electron–phonon (vibronic) coupling constant. Doubling the real-space unit cell from a to 2a folds the Brillouin zone in half (to −π/2a ≤ k ≤ +π/2a) and mixes every state |k⟩ with |k − 2k_F⟩ via the half-gap parameter Δ = 2 α u, splitting the single metallic band into a filled valence band (E₋) and an empty conduction band (E₊):

E_±(k) = ± √[ 4 t₀² cos²(k a) + Δ² sin²(k a) ]     where     Δ = 2 α u

At the new Brillouin zone boundaries k = ±k_F = ±π / (2a), cos(k_Fa) = 0 and sin(k_Fa) = ±1, so a direct Peierls band gap of E_g = 2Δ = 4 α u opens between the top of the valence band (−Δ) and the bottom of the conduction band (+Δ).

Why the Electronic Energy Lowering Always Beats the Elastic Cost in 1D

Why does this dimerization occur spontaneously for any nonzero electron–phonon coupling α, no matter how stiff the lattice spring constant K? Integrating the electronic energy change over all occupied states (−k_F ≤ k ≤ +k_F) for small dimerization amplitude u (Δ ≪ 2t₀) and adding the harmonic elastic restoring penalty ½K (2u)² = 2K u² per dimer yields the total energy per unit cell E_tot(u):

E_tot(u) = E_el(u) + E_elastic(u) ≈ − [ Δ² / (2 π t₀) ] ln( 4 t₀ / Δ ) + ½ K_eff u²

Look closely at the electronic stabilization term −(Δ² / 2πt₀) ln(4t₀ / Δ) ∝ −u² ln(1 / |u|):

• Because of the logarithmic factor ln(4t₀ / Δ)—which stems from the perfect 1D Fermi surface nesting where two parallel flat Fermi sheets at ±k_F are connected by a single wavevector q = 2k_F—for sufficiently small displacements u → 0, the negative logarithmic electronic energy gain −u² ln(1/|u|) always outweighs the positive quadratic elastic cost +½K_effu², regardless of how large K_eff is!
• Minimizing E_tot(Δ) with respect to Δ (setting dE_tot/dΔ = 0) yields the equilibrium Peierls gap Δ₀:

Δ₀ = 4 t₀ exp( − 1 / λ_ep )     where     λ_ep = 4 α² / (π K_eff t₀)

where λ_ep is the dimensionless electron–phonon coupling constant. Simultaneously, the acoustic phonon frequency ω(q) at wavevector q = 2k_F undergoes a giant logarithmic dip known as a Kohn anomaly, softening all the way to ω(2k_F) → 0 at the mean-field Peierls transition temperature k_BT_P ≈ 0.57 Δ₀, below which the phonon freezes into a static lattice distortion.

Bond-Length Alternation in trans-Polyacetylene [–(CH=CH)n–] and Su–Schrieffer–Heeger (SSH) Solitons

In organic chemistry and molecular biophysics, the simplest infinite conjugated polyene is trans-polyacetylene, [–CH=CH–]_n. If all C–C bonds along the carbon backbone were equal in length (~1.40 Å, as in benzene), each sp² carbon atom would contribute one unpaired 2p_z electron to a half-filled 1D π-band, making undoped polyacetylene a metallic conductor with zero band gap.

In reality, as modeled quantitatively by the Su–Schrieffer–Heeger (SSH) Hamiltonian (1979), trans-polyacetylene undergoes a robust Peierls dimerization (k_F = π / 2a):

• The carbon chain spontaneously alternates between short C=C double bonds (1.36 Å) and long C–C single bonds (1.44 Å) (a dimerization displacement of u ≈ ±0.04 Å), localizing π electron density onto the short bonds.
• With resonance integral t₀ ≈ 2.5 eV and electron–phonon coupling α ≈ 4.1 eV Å⁻¹, this bond-length alternation opens a Peierls semiconductor optical band gap of 2Δ ≈ 1.4 to 1.8 eV, explaining why pristine trans-polyacetylene is an insulator rather than a metal. (In biological polyenes such as β-carotene and the retinal chromophore of vision, this Peierls-like bond-length alternation governs the visible absorption wavelength and reverses sign upon photoexcitation toward a conical intersection.)
• Because trans-polyacetylene possesses two degenerate dimerization phases—Phase A (–C=C–C=C–) and Phase B (=C–C=C–C=)—a topological boundary between Phase A and Phase B forms a mobile, non-bonding mid-gap radical defect known as a topological soliton, whose discovery and chemical doping into a metallic state by Alan Heeger, Alan MacDiarmid, and Hideki Shirakawa earned the 2000 Nobel Prize in Chemistry.

Charge Density Waves (CDWs) in Krogmann's Salt (KCP) and NbSe₃

In quasi-1D inorganic crystals—where transition-metal chains are packed parallel to one another with strong intrachain d-orbital overlap along the chain axis and weak interchain coupling—the Peierls instability creates a periodic modulation of the conduction electron density ρ(x) = ρ₀ + ρ₁ cos(2k_Fx + φ) called a Charge Density Wave (CDW):

1. Krogmann's Salt, K₂Pt(CN)₄Br_0.30·3H₂O (KCP): Square-planar [Pt(CN)₄]²⁻ units stack face-to-face along the c-axis with a short Pt–Pt distance of 2.89 Å (nearly matching metallic platinum at 2.77 Å), forming a 1D conduction band out of overlapping Pt 5d_z² orbitals. Partial oxidation by 0.30 equivalents of bromine removes 0.30 electrons per Pt atom, leaving a 5/6-filled 1D band with Fermi wavevector k_F = (1 − 0.30/2)(π/c) = 0.85(π/c). Diffuse X-ray and inelastic neutron scattering reveal a giant Kohn anomaly and an incommensurate Peierls distortion of wavevector 2k_F = 0.30(2π/c) (period λ_CDW = 6.67 c), opening a pseudogap that turns KCP from a lustrous bronze-colored 1D conductor along the needle axis into a low-temperature insulator.
2. Transition-Metal Trichalcogenides (NbSe₃, TaS₃) and Blue Bronzes (K_0.30MoO₃): Quasi-1D trigonal-prismatic chains in NbSe₃ undergo two successive Peierls CDW transitions at T_P1 = 145 K and T_P2 = 59 K, while blue bronze (K_0.30MoO₃) undergoes a sharp metal-to-semiconductor Peierls transition at T_P = 180 K. When a small DC electric field exceeds a threshold value E_T, the incommensurate CDW depins from crystal impurities and slides collectively through the lattice (Fröhlich sliding CDW conductivity).
3. Spin-Peierls Transition (CuGeO₃, TTF–CuS₄C₄(CF₃)₄): Even in a 1D Mott insulator where strong on-site Coulomb repulsion U prevents metallic conduction (leaving one localized S = ½ spin per site coupled antiferromagnetically by exchange J), coupling between the 1D Heisenberg spin chain and lattice phonons drives a spin-Peierls dimerization below T_SP (e.g., 14 K in inorganic CuGeO₃), pairing adjacent spins into non-magnetic singlets (S = 0) separated by a spin gap.

How is the Peierls distortion related to the molecular Jahn–Teller effect?

As Roald Hoffmann and Isaac Bersuker demonstrated, the Peierls distortion and the molecular Jahn–Teller effect are the exact same vibronic phenomenon viewed from opposite ends of the size scale (solid-state band theory vs. molecular orbital theory). Consider building a 1D chain from cyclic rings of N carbon atoms (annulenes C_NH_N): in cyclobutadiene (C₄H₄, N = 4) or planar cyclooctatetraene (C₈H₈, N = 8), the Hückel frontier molecular orbitals at E = 0 are doubly degenerate, so the ring undergoes a molecular Jahn–Teller (or pseudo Jahn–Teller) distortion from a regular polygon into alternating single and double bonds. As N → ∞, the discrete Hückel molecular orbitals merge into a continuous 1D tight-binding band E(k) = −2t₀ cos(ka), the degenerate frontier orbitals at E = 0 become the Fermi points k = ±k_F = ±π/(2a), and the bond-alternating Jahn–Teller vibration becomes the zone-boundary phonon at q = 2k_F = π/a.

Why don't ordinary 3D metals like copper, silver, or sodium undergo a Peierls transition at low temperature?

The instability of a 1D metal relies on perfect Fermi surface nesting: in one dimension, the "Fermi surface" consists of just two flat points at +k_F and −k_F, so a single phonon wavevector q = 2k_F simultaneously connects 100% of the states at the Fermi level, producing a divergent logarithmic electronic susceptibility χ(2k_F) ∝ ln(E_F / k_BT) that guarantees instability at T → 0 K. In a two-dimensional or three-dimensional metal (like Cu, Ag, Au, or Na), the Fermi surface is a curved circle or sphere. A single wavevector q can connect only a tiny geometric fraction of states on opposite sides of a sphere, eliminating the logarithmic divergence in χ(q). Consequently, the small electronic energy gain from opening a partial gap on a tiny patch of the Fermi sphere is far too weak to overcome the positive elastic strain energy ½K u² of distorting the 3D crystal lattice.

What is the difference between a commensurate and an incommensurate Peierls Charge Density Wave?

In a commensurate Peierls distortion, the Fermi wavevector k_F is a simple rational fraction (M/P) of the reciprocal lattice vector (for example, a half-filled band with P = 2 in trans-polyacetylene, where λ_CDW = 2a, or a quarter-filled band with P = 4, where λ_CDW = 4a). The resulting Charge Density Wave (CDW) locks into register with the underlying atomic lattice at specific preferred positions. In an incommensurate Peierls distortion—such as Krogmann's salt K₂Pt(CN)₄Br_0.30·3H₂O, where non-stoichiometric bromine oxidation gives λ_CDW = (2/0.30)c = 6.67 c—the period of the CDW is not an integer multiple of the lattice spacing c. Because the energy of an ideal incommensurate CDW is invariant under a rigid phase shift φ along the chain, its phase excitations form a gapless collective mode (phason), allowing the CDW to slide and carry collective electrical current above a small depinning threshold electric field.

1. Peierls, R. E. (1955). Quantum Theory of Solids (pp. 108–112). Oxford University Press. https://doi.org/10.1093/acprof:oso/9780198507819.001.0001
2. Su, W. P., Schrieffer, J. R., & Heeger, A. J. (1979). Solitons in polyacetylene. Physical Review Letters, 42(25), 1698–1701. https://doi.org/10.1103/PhysRevLett.42.1698
3. Hoffmann, R. (1987). How chemistry and physics meet in the solid state. Angewandte Chemie International Edition in English, 26(9), 846–878. https://doi.org/10.1002/anie.198708461
4. Grüner, G. (1988). The dynamics of charge-density waves. Reviews of Modern Physics, 60(4), 1129–1181. https://doi.org/10.1103/RevModPhys.60.1129