Polynomial Eigenfunctions in Spectral Analysis

Updated 22 January 2026

Polynomial eigenfunctions are sequences of polynomials that serve as simultaneous eigenfunctions of linear differential operators with a triangular structure ensuring existence and uniqueness.
They establish a bispectral framework by satisfying both differential equations and finite-term recurrence relations, which aids in resolving inverse problems in spectral analysis.
Applications span classical orthogonal families (e.g., Hermite, Laguerre, Jacobi), representation theory, integrable systems, and stochastic models.

Polynomial eigenfunctions are sequences of polynomials that serve as simultaneous eigenfunctions of a differential (or, more generally, a linear) operator, typically parameterized by their degree. Their study forms a cornerstone of spectral theory, special functions, integrable systems, orthogonal polynomial theory, and modern mathematical physics. This subject encompasses both the classical families classified by Bochner and their generalizations via the bispectral problem, sl(2) and sl(N) representation theory, Fock space constructions, multidimensional extensions, and applications in matrix, elliptic, and stochastic settings.

1. Fundamental Theory: Differential Operators and Polynomial Eigenfunctions

Polynomial eigenfunctions arise from ordinary differential operators of the form $L = \sum_{i=0}^N a_i(x)\, \partial_x^i,$ where each $a_i(x)$ is a polynomial of degree at most $i$ and $a_0(x)=0$ (convention) (Rolanía, 2021, Anguas et al., 2023, Azad et al., 2010). The central spectral problem seeks a sequence of monic polynomials $\{P_n(x)\}$ , $\deg P_n = n$ , and eigenvalues $\{\lambda_n\}$ such that $L\,P_n(x) = \lambda_n\,P_n(x),$ for $n=0,1,2,\dots$ , with $\lambda_n$ generically determined by the leading (highest degree) coefficients:

$\lambda_n = \sum_{i=1}^{\min(n,N)} n(n-1)\cdots(n-i+1)\, \ell_{i,i},$

where $a_i(x) = \sum_{j=0}^i \ell_{i,j}\, x^j$ .

The triangular structure induced by such operators ensures the existence and uniqueness (up to normalization) of the polynomial eigenfunctions under a non-resonant spectrum (distinct $\lambda_n$ ). The coefficients of lower-degree terms are determined recursively or by closed-form multiple sum formulas (Rolanía, 2021, Anguas et al., 2023). Classical examples include Hermite, Laguerre, and Jacobi polynomials, which represent the only sequences admitting three-term recurrences and orthogonality for second-order $L$ (Bochner’s theorem).

2. The Bispectral Problem, Recurrences, and Inverse Construction

Given $L$ of suitable order, the sequence $\{P_n(x)\}$ may also satisfy a finite-term recurrence:

$x\, P_n(x) = \sum_{k=n-p}^{n+1} \gamma_{n,k}\, P_k(x),$

defining a bispectral setting where polynomials are eigenfunctions of both the differential operator $L$ (spectral in $x$ ) and a difference operator $D$ in the degree index $n$ (spectral in $n$ ) (Rolanía, 2021, Anguas et al., 2023). The bandwidth $p+2$ is precisely characterized by the rank of a Hankel-type matrix formed from coefficients of $\{P_n(x)\}$ .

The inverse problem involves reconstructing $L$ given a sequence $\{P_n\}$ and eigenvalues $\{\lambda_n\}$ , provided the recurrence and nondegeneracy conditions are met. Explicit expressions allow one to recover the operator coefficients via triangular systems and minor formulas, with vanishing support conditions defining the order (Rolanía, 2021).

3. Representation Theory: sl(2), Fock Spaces, and Algebraic Models

Polynomial eigenfunctions are naturally classified by finite-dimensional representations of Lie algebras, notably sl(2) in one-variable settings and sl(N) in multivariable or matrix settings (Turbiner et al., 8 Aug 2025, Turbiner et al., 2013, Turbiner, 2014, Sokolov et al., 2014).

Fock Space Construction: For the Heisenberg algebra $\mathcal{H}_3 = \mathrm{span}\{a, b, 1\}$ , the space of polynomials in the creation operator $b$ acting on the vacuum space provides a setting for polynomial eigenfunctions, where the Euler–Cartan (number) operator emerges as fundamental. The sl(2) subalgebra classifies exactly solvable (full infinite flag preserved) and quasi-exactly solvable (finite-dimensional invariant subspaces) operators; the latter admit only $n+1$ polynomial eigenfunctions for some discrete integer $n$ (Turbiner et al., 8 Aug 2025).
Algebraic Differential Realizations: Via gauge transformations and change of variables (often to invariants), elliptic or rational quantum models (e.g., $A_2/G_2$ Calogero–Moser, $BC_1$ models) yield polynomial eigenfunctions spanning finite-dimensional sl(N) or higher algebra modules. Hidden algebraic structures are exposed by expressing the transformed Hamiltonians as quadratic or higher elements in the enveloping algebra (e.g., $sl(3)$ , $g^{(2)}$ ), with invariance of the polynomial module determined by discrete coupling constants (Sokolov et al., 2014, Turbiner, 2014).

4. Multivariate, Matrix, and Nonclassical Extensions

Multivariate Orthogonal Polynomials

Several multivariate polynomial families serve as eigenfunctions for diffusion processes and birth–death generators. The multivariate Poisson–Charlier, Meixner, and Hermite–Chebycheff polynomials, constructed from Krawtchouk polynomials and classical one-dimensional bases, diagonalize transition kernels and infinitesimal generators of Markov chains, with spectral decompositions inheriting polynomial eigenbases and three-term recurrences (Griffiths, 2014).

Matrix-Valued Polynomial Eigenfunctions

Contrary to scalar orthogonal polynomials, which require operators of even order for orthogonal polynomial eigenbases, explicit constructions demonstrate the existence of weight matrices (non-reducible to scalars) such that the sequence of matrix orthogonal polynomials are eigenfunctions of odd-order (e.g., third, fifth) differential operators. The symmetry (matrix Pearson) equations allow odd order in the matrix setting (Durán et al., 25 Jan 2025).

Higher-Order and 2-Orthogonal Cases

Generalizations to degree-preserving operators admit polynomial eigenfunctions in more elaborate orthogonality structures, such as 2-orthogonal sequences with four-term recurrences and third-order operators. Algebraic criteria and explicit recursive relations produce unique eigenbases for specified operator classes (Mesquita, 2021).

Newtonian/Abstract Hypergeometric Polynomials

In abstract settings, polynomials constructed in a Newtonian basis (interpolation nodes) and defined as eigenfunctions of two-diagonal operators $L\varphi_n=\lambda_n\varphi_n+\tau_n\varphi_{n-1}$ unify the Askey–Wilson tableau via appropriate choices of spectra and recurrence parameters. Orthogonality and three-term recurrences follow from compatibility equations among generalized moments (Vinet et al., 2016).

5. Integrable Systems, Baker–Akhiezer Functions, and the DIM Algebra

The theory extends deeply into integrable systems and modern algebraic frameworks.

Baker–Akhiezer Functions (BAF): At non-generic parameter values (e.g., $t=q^{-m}$ for Macdonald polynomials), symmetrized combinations of complementary nonsymmetric polynomials yield explicit Baker–Akhiezer functions, which are polynomial eigenfunctions of first-order difference operators much simpler than the cut-and-join (Ruijsenaars) Hamiltonians. Extensions include twisted BAFs providing eigenfunctions for the commutative integer-ray subalgebras of the Ding–Iohara–Miki (DIM) algebra (Mironov et al., 2024, Mironov et al., 2024). Factorization properties and analytic continuation to non-integer parameters facilitate both representation-theory and integrable-model applications.

6. Applications in Physical, Matrix, and Stochastic Contexts

Polynomial eigenfunctions underpin the spectral analysis of elliptic, quantum, and stochastic operators.

Elliptic Operators and Special Curves: In two-dimensional contexts, polynomial eigenfunctions of Laplace–Beltrami operators with flat metrics give rise to families of algebraic curves with optimal singularity configurations (maximizing/free curves), whose bifurcation structures correspond to parameter-dependent Hamiltonian dynamics (Escudero, 23 Aug 2025).
Energy Balance in Stochastic Systems: Classical orthogonal polynomial families (Hermite, Zernike, spherical harmonics) emerge as eigenfunctions for dissipative differential operators governing the energy balance of stochastic dynamical systems. Projection of infinite-dimensional master integral operators yields matrix Lyapunov equations where the polynomial structure is preserved under uniform dissipation (Moriya, 18 Jun 2025).

7. Classical and Nonclassical Examples

Classical Families:

Hermite: $L = y'' - 2x\,y'$ , Eigenvalues $\lambda_n = -2n$ , $P_n(x) = H_n(x)$ (Rolanía, 2021, Azad et al., 2010).
Laguerre: $L = x\,y'' + (\alpha+1-x)y'$ , Eigenvalues $\lambda_n = -n$ , $P_n(x) = L_n^{(\alpha)}(x)$ .
Jacobi: $L = (1-x^2)y'' + [\beta-\alpha - (\alpha+\beta+2)x]y'$ , Eigenvalues $\lambda_n = -n(n+\alpha+\beta+1)$ .

Nonclassical and QES Examples:

Romanovski, identity-type, and higher-order, bispectral or quasi-exactly-solvable cases admit polynomial eigenfunctions with various orthogonality measures or finite orthogonality domains (Azad et al., 2010, Turbiner et al., 2013, Turbiner, 2014, Sokolov et al., 2014).