---
title: Generalized Laughlin Wave Functions
url: https://www.emergentmind.com/topics/generalized-laughlin-wave-functions
type: topic
---

# Generalized Laughlin Wave Functions

Generalized Laughlin wave functions are extensions of Laughlin’s correlated lowest-Landau-level ansatz that preserve some subset of its defining structures—most notably Jastrow zeros, analyticity, incompressibility, topological degeneracy, or parent-Hamiltonian annihilation conditions—while changing geometry, Hilbert space, internal degrees of freedom, or excitation content. In the contemporary literature, the term does not refer to a single canonical family. It encompasses analytic deformations of the form \(\Psi_F=F\Psi_{\mathrm{Lau}}\), torus and cylinder adaptations, exact lattice and matrix-product realizations, multicomponent Halperin states, second-quantized zero-mode modules, coupled-condensate hierarchy constructions, and spin or lattice analogues derived from conformal field theory [1609.03818, 1206.3071, 1405.0742, 2604.21434, 1201.3096].

## 1. Foundational definition and preserved structure

The standard Laughlin state at filling \(1/\ell\) is written, in planar complex coordinates \(z_j\), as
\[
\Psi_{\mathrm{Lau}}(z_1,\dots,z_N)=c_{\mathrm{Lau}}\prod_{1\le i<j\le N}(z_i-z_j)^\ell \exp\!\left(-\frac12\sum_{i=1}^N |z_i|^2\right),
\]
with \(\ell\) odd for fermions and even for bosons in the conventions used in the rigidity literature [1609.03818]. The factor \(\prod_{i<j}(z_i-z_j)^\ell\) enforces the characteristic Laughlin correlation hole: every pair still vanishes at least as \((z_i-z_j)^\ell\) when particles coincide. A central modern definition of generalized Laughlin states therefore takes the form
\[
\Psi_F(z_1,\dots,z_N)=F(z_1,\dots,z_N)\,\Psi_{\mathrm{Lau}}(z_1,\dots,z_N),
\]
where \(F\) is analytic and symmetric [1707.05059].

This class is often called the class of **fully correlated states**. It is broad enough to include quasi-holes, disorder-induced deformations, trap-induced shape changes, and more general analytic many-body modifications, but it is not the full lowest-Landau-level Hilbert space. The distinction is essential: arbitrary lowest-Landau-level states may alter the coincidence zeros, whereas \(\Psi_F=F\Psi_{\mathrm{Lau}}\) preserves the Laughlin vanishing pattern and thus the short-range mechanism suppressing repulsive interactions [1701.09064]. A common restricted subclass is
\[
\Psi_f(z_1,\dots,z_N)=\Bigl(\prod_{j=1}^N f(z_j)\Bigr)\Psi_{\mathrm{Lau}}(z_1,\dots,z_N),
\]
with \(f\) analytic. In this factorized case the zeros of \(f\) are interpreted as uncorrelated quasi-holes [1906.11656].

The same foundational idea reappears in less direct guises. In second quantization, generalized Laughlin states become zero modes obtained by acting on the densest Laughlin zero mode with an algebra representing symmetric functions [1409.3577]. In multicomponent settings, the role of the Laughlin Jastrow factor is played by coupled intra- and inter-component Jastrow products, as in Halperin \((m,m,n)\) states [2604.21434]. In lattice and spin formulations, fixed-site coordinates replace mobile continuum coordinates, but the same logic of analyticity and pairwise zeros survives [1201.3096].

## 2. Analytic deformations, quasi-holes, and rigidity

For the fully correlated class \(\Psi_F=F\Psi_{\mathrm{Lau}}\), the principal rigorous result is an incompressibility bound. If \(\rho_F\) denotes the one-particle density of a normalized \(\Psi_F\), then for any \(\alpha>1/4\) and any disk \(D\) of radius \(N^\alpha\),
\[
\int_D \rho_F \le \frac{1}{\pi\ell}\,|D|\,(1+o(1)),
\]
uniformly in the analytic symmetric prefactor \(F\), as \(N\to\infty\) [1609.03818]. This is a mesoscopic averaged statement rather than a pointwise finite-\(N\) bound. The same theme is developed in a sharper local-incompressibility framework for generalized Coulomb gases generated by Laughlin’s plasma analogy [1701.09064].

The plasma formulation is central. After rescaling, \(|\Psi_F|^2\) becomes a Gibbs measure with effective Hamiltonian
\[
H_N(Z_N) = \sum_{j=1}^N |z_j|^2 - \frac{2\ell}{N}\sum_{1\le i<j\le N}\log|z_i-z_j| + W_N(Z_N),
\qquad
W_N(Z_N)= -\frac{2}{N}\log|F(\sqrt N\,Z_N)|.
\]
Because \(F\) is analytic, \(W_N\) is superharmonic in each coordinate. That superharmonicity is the structural reason the extra factor can redistribute or lower density but cannot create local overcompression [1707.05059]. In the corresponding classical exclusion problem, screening regions derived from an incompressible Thomas–Fermi model enforce asymptotic density at most the neutralizing-background density, which maps back to the Laughlin value \(1/(\pi\ell)\) [2203.06952].

These bounds lead directly to variational consequences. For confining potentials \(V\), the energy minimization problem over all fully correlated states is asymptotically governed by the bathtub functional with hard cap \(1/(\pi\ell)\). More strongly, for a suitable polynomial \(f\), the restricted quasi-hole class already attains the same large-\(N\) energy:
\[
\lim_{N\to\infty}\frac{E_V(N,\ell)}{e_V(N,\ell)}=
\lim_{N\to\infty}\frac{E_V(N,\ell)}{E_V^{\rm bt}(N,\ell)}=1.
\]
Thus, within the stated scaling regime, independent quasi-holes suffice to asymptotically optimize the response to trapping and disorder [1707.05059]. A common misconception is that this proves every generalized Laughlin state reduces to a product quasi-hole state. It does not. The rigorous statement is variational and asymptotic: the full class is larger, but the smaller factorized subclass is asymptotically sufficient for the energy problems considered [1906.11656].

## 3. Geometry-adapted continuum generalizations

On the torus, generalized Laughlin wave functions replace the planar polynomial factor by Jacobi theta functions. For filling \(\nu=1/m\), the many-body holomorphic factor takes the form
\[
f[z]=F(Z)\prod_{i<j}\vartheta_{\frac12,\frac12}(z_i-z_j|\tau)^m,
\qquad
Z=\sum_i z_i,
\]
where the center-of-mass factor \(F(Z)\) carries the residual \(m\) zeros required by torus boundary conditions [1405.0742]. This yields the expected \(m\)-fold topological degeneracy on genus one. The same formalism produces torus quasi-hole states straightforwardly and, more nontrivially, a quasi-electron state involving derivative operators and an additional Gaussian operator needed to satisfy magnetic-translation boundary conditions [1405.0742].

On the cylinder, the Laughlin state becomes
\[
\Psi_N(z_1,\dots,z_N)=\kappa_N \prod_{1\le j<k\le N}\left(e^{z_k/R}-e^{z_j/R}\right)^p
\exp\!\left(-\frac{1}{2\ell^2}\sum_{k=1}^N x_k^2\right),
\]
with \(p\) odd for fermions and even for bosons [1109.4022]. In the thermodynamic limit at fixed radius, all correlation functions have a unique bulk limit, and the limit state exhibits a nontrivial axial period \(p\ell^2/R\). The result holds for both bosonic and fermionic statistics and persists from thin to thick cylinders [1109.4022]. This geometry therefore reveals an emergent periodic ordering that is absent from the planar homogeneous-liquid description.

A distinct but related geometric generalization is the thin-torus-orbital reduction to one-dimensional lattice models. For \(\nu=1/q\), one can construct finite-range one-dimensional Hamiltonians whose exact frustration-free ground states are squeezed descendants of the Tao–Thouless root patterns and admit exact matrix-product representations [1206.3071]. For fermions with odd \(q\),
\[
\mathcal H_{1/q}=\sum_{l=1}^{[q/2]}\sum_i\bigl[Q_i^{(l)\dagger}Q_i^{(l)}+P_i^{(l)\dagger}P_i^{(l)}\bigr],
\]
and the exact zero mode is built by local squeezing operators acting on the root configuration \(10^{q-1}10^{q-1}\cdots\) [1206.3071]. These states are not claimed to be identical to the full continuum torus Laughlin wave function at generic aspect ratio; rather, they share the same pseudopotential-like algebraic structure near the thin-torus limit and show high overlap with the torus Laughlin state over a substantial range of circumference.

## 4. Lattice, spin, and crystalline analogues

One line of generalization retains Laughlin-type topological data while replacing continuum coordinates by lattice sites. In the Moore–Read/CFT-inspired lattice construction, the amplitudes
\[
\psi_q^\eta(n_1,\ldots,n_N)\propto \delta_n \xi_{\eta}
\prod_{i<k}(z_i-z_k)^{q n_i n_k}\prod_{j\neq l}(z_l-z_j)^{-\eta n_l}
\]
define lattice analogues of bosonic and fermionic Laughlin states with \(\nu=1/q\), but with site filling \(\mu=\eta/q\) allowed to differ from \(\nu\) [1609.02435]. This disentangles continuum filling from lattice filling. The same family contains particle-hole-symmetric half-filled states at \(\mu=1/2\), supports edge-state constructions by varying particle number at fixed analytic form, and admits exact parent Hamiltonians for \(\mu\le 1/q\), \(\mu\ge 1-1/q\), and a special half-filled \(q=4\) case [1609.02435].

The lattice phase structure is geometry dependent. On the square lattice, half-filled states with \(q\le 4\) retain the topological properties of continuum Laughlin states, whereas for larger \(q\) there is a transition toward long-range-ordered antiferromagnets. This effect is absent on triangular and Kagome lattices, where topological properties are recovered, indicating that lattice frustration stabilizes the generalized Laughlin phase [1609.02435]. A different lattice-spin route starts from free-boson conformal blocks:
\[
\psi_{s_1,\ldots,s_N}(z_1,\ldots,z_N)=\delta_{\mathbf s}\,\prod_{p=1}^N \chi_{p,s_p}\prod_{n<m}(z_n-z_m)^{\alpha s_n s_m},
\]
with \(s_n=\pm1\). After reinterpreting \(s_n=2q_n-1\), these become hard-core-boson lattice Laughlin states with effective filling \(\nu=(4\alpha)^{-1}\) [1201.3096]. For \(\alpha=1/2\), the Kalmeyer–Laughlin state is recovered, and an exact parent Hamiltonian is obtained from \(SU(2)_1\) null-field decoupling [1201.3096].

A further analogue maps on-site singlet Mott states of spin-\(S\) bosons to bosonic Laughlin states on the sphere. For \(n=S/l+1\), the on-site state
\[
|\varphi\rangle_i=\prod_{1=\alpha<\beta}^{n}
\left(a^\dagger_{i\alpha}b^\dagger_{i\beta}-b^\dagger_{i\alpha}a^\dagger_{i\beta}\right)^{2l}|0\rangle_i
\]
becomes, in spin-coherent coordinates, the bosonic Laughlin wave function at filling \(\nu=1/(2l)\), and the construction generalizes to a \(q\)-family associated with \(\nu=q/(2l)\) [1310.6105]. By contrast, in periodic torus geometry one may also hybridize Laughlin correlations with crystalline Hartree–Fock structure. The resulting many-body states preserve the same zeros as periodic Laughlin states but include a determinantal crystalline factor, and Monte Carlo data up to \(16\) particles, with preliminary results at \(36\), show lower energy than the periodic Laughlin state even at \(\nu=1/3\) [1006.0838]. This suggests a possible coexistence or competition between Laughlin-type short-range correlations and broken translation symmetry, although the thermodynamic-limit conclusion remains explicitly preliminary.

## 5. Operator-algebraic, multicomponent, and hierarchy extensions

In second quantization, generalized Laughlin states arise as the full zero-mode module of pseudopotential Hamiltonians. Writing the parent Hamiltonian in lattice/orbital form,
\[
H^M=\sum_{\substack{0\le m<M\\ (-1)^m=(-1)^M}} V_m H_m,
\qquad
H_m=\sum_R (T_R^m)^\dagger T_R^m,
\]
one finds that the entire zero-mode subspace satisfies
\[
Z=\mathcal A |\psi_{1/M}\rangle,
\]
where \(\mathcal A\) is the commutative algebra generated by the operators \(\mathcal O_d=\sum_r c_{r+d}^\dagger c_r\), equivalently by the elementary-symmetric-function operators \(e_d\) obeying the operator Newton–Girard relations [1409.3577]. In first-quantized language this is the statement that generalized Laughlin descendants are precisely \(\Psi_{1/M}\) multiplied by symmetric polynomials. The same framework yields admissible partitions, dominance patterns, a second-quantized quasi-hole operator
\[
\hat U(\zeta)=\sum_{d=0}^N (-\zeta)^{N-d}e_d,
\]
and Read’s nonlocal order parameter [1409.3577].

A multicomponent generalization replaces a single Jastrow factor by coupled intra- and inter-component factors. The canonical example is the Halperin state
\[
\Psi_{\rm H}^{(mmn)}=
\prod_{i<j}(u_i-u_j)^m
\prod_{i<j}(v_i-v_j)^m
\prod_{i,j}(u_i-v_j)^n.
\]
Using operator contraction multiplication, one can exactly decompose ordinary Laughlin states and then extend the method to Halperin states by expressing the mixed Jastrow factor as a resultant of two polynomials and rewriting it through elementary symmetric polynomials [2604.21434]. For the Halperin \((2,2,1)\) state, the construction yields an explicit basic expansion, identifies the root configurations \({\bf U}_{[30]}{\bf V}_{[41]}\) and \({\bf U}_{[41]}{\bf V}_{[30]}\), and distinguishes intra-color from inter-color squeezing, thereby exposing the underlying generalized Pauli principle [2604.21434].

Another hierarchy extension recasts Jain-type composite-fermion states as generalized Laughlin condensates. Starting from a modified quasiparticle operator acting on the \(\nu=1/m\) Laughlin state, one constructs families at
\[
\nu_{n,m}^{\pm}=\frac{n}{n(m\mp 1)\pm 1},
\]
interpreted as coherent superpositions of \(n\) coupled Laughlin condensates and their conjugates in different projected Hilbert subspaces [1507.02161]. The stated result is that the corresponding wave functions are identical to the noninteracting composite-fermion wave functions, including the \(\nu_{n,m}^{-}\) branch, which is presented as especially surprising in this language [1507.02161]. A different modern extension uses operator-contraction rules to decompose Halperin and Laughlin states at very large Hilbert-space dimensions, but the common structural theme is that generalized Laughlin physics can be reformulated as an algebraic problem of root configurations, squeezing, and operator-generated descendants rather than only as explicit first-quantized polynomials [2604.21434].

## 6. Correlations, entanglement, and complexity diagnostics

Generalized Laughlin wave functions are often assessed not only by overlap or energy, but also by the extent to which they reproduce the correlation, entanglement, and edge structures associated with the Laughlin phase. In the one-dimensional exact lattice realization of the Laughlin series, the matrix-product form allows analytic access to densities, correlation lengths, and entanglement spectra [1206.3071]. The transfer matrix has only two nonzero eigenvalues, giving an exact correlation length
\[
\xi^{-1}=-\frac{1}{q}\ln\left|\frac{\lambda_-}{\lambda_+}\right|.
\]
On the cylinder, the entanglement spectrum forms an approximately linear low-lying branch interpreted as the chiral Tomonaga–Luttinger-liquid edge mode; on the torus, the spectrum exhibits the additive “diamond” structure expected from independent left- and right-moving edges [1206.3071]. These signatures support the view that the exact lattice states encode nontrivial Laughlin-phase data even though they are not translationally invariant homogeneous liquids at finite circumference.

Other diagnostics probe how far Laughlin-type states lie from single-determinant or Fermi-sea descriptions. For the fermionic Laughlin state, the geometric entanglement
\[
E_G=-\ln\!\left(\max_S |\langle S|\Psi_L\rangle|^2\right)
\]
is numerically close to a linear function of particle number \(N\), especially for \(\nu=1/3\), and the fitted constant term does not agree with the expected topological entropy \(\ln\sqrt m\) [1610.03904]. This directly cautions against identifying every subleading constant in a non-bipartite entanglement measure with topological entanglement entropy. A complementary diagnostic uses Dyson orbitals. For the Laughlin state \(\Psi_{(m,N)}\), the Dyson orbital for the \(N\to N+1\) transition is exactly the Landau orbital \(\varphi_{mN}\), yet the optimal particle-addition overlap \(\mathcal O_{\max}\) decays with \(N\); numerically, for fermions \(m=3,5,7\) and bosons \(m=2,4,6\), the data are consistent with
\[
\mathcal O_{\max}\propto N^{-(m-1)/2}.
\]
This is presented as quantitative evidence that Laughlin states are strongly correlated and non-Fermi-liquid-like [2606.05502].

The same care is needed when interpreting improved variational families. The Girvin–Jach modified Laughlin states introduce a real parameter \(d\) through the factor \(\prod_{i<j}|z_i-z_j|^{2d}\), followed by LLL projection. On the sphere,
\[
\psi^{(q,d)}=\prod_{i<j}(u_iv_j-u_jv_i)^q|u_iv_j-u_jv_i|^{2d},
\]
and on the torus a corresponding \(\vartheta_1\)-based form is used [1601.06736]. These states are not proposed as a different topological phase; rather, the stated conclusion is that they probably represent the same universality class as Laughlin while yielding significant improvements in variational energy, overlap with the exact Coulomb ground state, and aspects of the entanglement spectrum on both sphere and torus [1601.06736]. More generally, a recurring limitation across the literature is that exact topological equivalence is seldom established directly. Depending on the construction, support comes from incompressibility theorems, entanglement-spectrum counting, high overlaps, parent-Hamiltonian annihilation, or adiabatic-continuity arguments, and these criteria are related but not identical.

Source: https://www.emergentmind.com/topics/generalized-laughlin-wave-functions