Constraint polynomial approach -- an alternative to the functional Bethe Ansatz method?

TL;DR

The paper introduces a constraint polynomial approach as an alternative to the functional Bethe Ansatz, simplifying the solution process for certain quantum potentials and enabling straightforward eigenvalue and wave function determination.

Contribution

It presents a new constraint polynomial method that replaces algebraic equations in the Bethe Ansatz, demonstrated on various quasi-exactly solvable potentials with advantages over existing models.

Findings

Constraint polynomials replace sets of algebraic equations in Bethe Ansatz.

The approach yields real, simple zeros indicating orthogonality.

Differences between the method and Rabi model generalizations are clarified.

Abstract

Recently developed general constraint polynomial approach is shown to replace a set of algebraic equations of the functional Bethe Ansatz method by a single polynomial constraint. As the proof of principle, the usefulness of the method is demonstrated for a number of quasi-exactly solvable (QES) potentials of the Schr\"odinger equation, such as two different sets of modified Manning potentials with three parameters, an electron in Coulomb and magnetic fields and relative motion of two electrons in an external oscillator potential, the hyperbolic Razavy potential, and a (perturbed) double sinh-Gordon system. The approach enables one to straightforwardly determine eigenvalues and wave functions. Odd parity solutions for the modified Manning potentials are also determined. For the QES examples considered here, constraint polynomials terminate a finite chain of orthogonal polynomials in an independent variable that need not to be necessarily energy. In the majority of cases the finite chain of orthogonal polynomials is characterized by a positive-definite moment functional ${\cal L}$, implying that a corresponding constraint polynomial has only real and simple zeros. Constraint polynomials are shown to be different from the weak orthogonal Bender-Dunne polynomials. At the same time the QES examples considered elucidate essential difference with various generalizations of the Rabi model. Whereas in the former case there are $n+1$ polynomial solutions at each point of a $n$th baseline, in the latter case there are at most $n+1$ polynomial solutions on entire $n$th baseline.