O(n) Models: XY and Heisenberg

Overview

The O($n$) model describes a classical spin $\mathbf{S}_i \in \mathbb{R}^n$ with $|\mathbf{S}_i| = 1$ on each lattice site, coupled by the Hamiltonian

\[H = -J \sum_{\langle i,j \rangle} \mathbf{S}_i \cdot \mathbf{S}_j\]

The global symmetry is the orthogonal group O($n$). Special cases:

$n$NameSymmetryPhysical realisations
1Ising$\mathbb{Z}_2$Uniaxial magnets, lattice gas
2XYO(2) $\cong$ U(1)Superfluid $^4$He, thin-film magnets
3HeisenbergO(3) $\cong$ SU(2)Isotropic magnets (EuO, EuS)

XY Model ($n = 2$)

$d = 2$ –- Berezinskii-Kosterlitz-Thouless (BKT) Transition

The 2D XY model has no spontaneous symmetry breaking at any $T > 0$ (Mermin-Wagner theorem). Instead, it undergoes a BKT transition at $T_{\mathrm{BKT}}$: below $T_{\mathrm{BKT}}$ correlations decay algebraically (quasi-long-range order), while above they decay exponentially.

The BKT transition is not described by standard power-law critical exponents. Its hallmarks are:

PropertyValue / behaviourNote
$\eta(T_{\mathrm{BKT}})$$1/4$Universal jump at the transition
Correlation length above $T_{\mathrm{BKT}}$$\xi \sim \exp(b/\sqrt{T - T_{\mathrm{BKT}}})$Essential singularity, not power law
Superfluid stiffnessUniversal jump $\rho_s(T_{\mathrm{BKT}}^-) = 2T_{\mathrm{BKT}}/\pi$Nelson-Kosterlitz criterion
Vortex-antivortex pairsBound for $T < T_{\mathrm{BKT}}$, unbound aboveTopological defect mechanism
No standard critical exponents

The BKT transition has $\nu = \infty$ in the usual sense (essential singularity). Standard exponents $\beta, \gamma, \delta$ are not defined for this transition because there is no true order parameter. QAtlas stores the universal value $\eta = 1/4$ at $T_c$ and flags the BKT nature.

$d = 3$ –- Conformal Bootstrap

In $d = 3$, the XY model has a conventional second-order transition. The most precise exponents come from the O(2) conformal bootstrap.

ExponentValueUncertaintyReference
$\alpha$$-0.0146(8)$Chester, Landry, Liu, Poland, Simmons-Duffin, Su, Vichi (2020)
$\beta$$0.3485(2)$JHEP 2020, 142
$\gamma$$1.3177(5)$"
$\nu$$0.6717(1)$"
$\eta$$0.0381(2)$"

Heisenberg Model ($n = 3$)

$d \leq 2$ –- Mermin-Wagner Theorem

The Mermin-Wagner theorem forbids spontaneous breaking of a continuous symmetry in $d \leq 2$ at $T > 0$ for short-range interactions. Consequently:

  • $d = 1$: The 1D Heisenberg chain is disordered at all $T > 0$. At $T = 0$ the quantum spin-1/2 chain is critical with $c = 1$ (Luttinger liquid), described by the SU(2)$_1$ WZW model.
  • $d = 2$: The 2D classical Heisenberg model has no finite-$T$ phase transition. (In contrast, the 2D XY model has the BKT transition, which does not break the continuous symmetry.)

$d = 3$ –- Conformal Bootstrap

In $d = 3$, the Heisenberg model has a conventional second-order transition. Best estimates from the O(3) conformal bootstrap:

ExponentValueUncertaintyReference
$\alpha$$-0.1336(15)$Chester, Landry, Liu, Poland, Simmons-Duffin, Su, Vichi (2020)
$\beta$$0.3689(3)$JHEP 2020, 142
$\gamma$$1.3960(9)$"
$\nu$$0.7112(5)$"
$\eta$$0.0378(3)$"

$d \geq 4$ –- Mean-Field

The upper critical dimension for O($n$) models (all $n \geq 1$) is $d_c = 4$. For $d \geq 4$ the exponents take mean-field values: $\beta = 1/2$, $\nu = 1/2$, $\gamma = 1$, $\eta = 0$.


QAtlas API

using QAtlas

# XY d = 3: numerical (Float64 + _err)
e_xy = QAtlas.fetch(Universality(:XY), CriticalExponents(); d=3)
# (β = 0.3485, β_err = 2e-4, ν = 0.6717, ν_err = 1e-4, ...)

# Heisenberg d = 3
e_heis = QAtlas.fetch(Universality(:Heisenberg), CriticalExponents(); d=3)
# (β = 0.3689, β_err = 3e-4, ν = 0.7112, ν_err = 5e-4, ...)

# d ≥ 4: mean-field
e_mf = QAtlas.fetch(Universality(:XY), CriticalExponents(); d=4)
# → same as fetch(MeanField(), CriticalExponents())

References

  • V. L. Berezinskii, Sov. Phys. JETP 32, 493 (1971) –- BKT transition (part I).
  • J. M. Kosterlitz, D. J. Thouless, J. Phys. C 6, 1181 (1973) –- BKT transition.
  • N. D. Mermin, H. Wagner, Phys. Rev. Lett. 17, 1133 (1966) –- absence of long-range order in $d \leq 2$.
  • S. M. Chester, W. Landry, J. Liu, D. Poland, D. Simmons-Duffin, N. Su, A. Vichi, "Carving out OPE space and precise O(2) model critical exponents", JHEP 2020, 142 –- O(2) and O(3) bootstrap.
  • D. R. Nelson, J. M. Kosterlitz, Phys. Rev. Lett. 39, 1201 (1977) –- universal superfluid-stiffness jump.

Connections

  • Ising: the $n = 1$ case; see Ising.
  • Heisenberg chain ($T = 0$): quantum $S = 1/2$ chain with $c = 1$; see Heisenberg model.
  • Mean-Field: baseline for $d \geq 4$; see mean-field.