Iterated Hamiltonian Intertwining in Quantum Mechanics

• Iterated Hamiltonian intertwining is a method that repeatedly applies intertwining operators to relate different Hamiltonians, enabling systematic spectral construction in quantum systems.
• It leverages techniques such as shape invariance, higher-order matrix operators, and spectral intertwining to transfer eigenvalues and eigenfunctions across hierarchies.
• Applications span supersymmetric quantum mechanics, relativistic Dirac systems, and integrable many-body models, illustrating its versatility in spectral design and quantum integrability.

Iterated Hamiltonian intertwining denotes the repeated use of operators satisfying intertwining relations between Hamiltonians in order to generate chains, hierarchies, or grids of related spectral problems. In its most familiar form, one seeks operators $A$ such that $H_1A^\dagger=A^\dagger H_0$, and then iterates the construction to transfer eigenvalues, eigenfunctions, or integrals of motion across a family of Hamiltonians. In the literature this pattern appears in supersymmetric quantum mechanics, relativistic Dirac systems, matrix Hamiltonians, non-self-adjoint settings, and integrable many-body models. This suggests that “iterated Hamiltonian intertwining” is best understood as an umbrella description for a recurrent operator-theoretic mechanism rather than a single universally standardized formalism (Contreras-Astorga et al., 2012, Houri et al., 2017, Sokolov, 2013, Correa et al., 1 Apr 2025).

1. Foundational operator scheme

In the nonrelativistic context, the intertwining operator technique is synonymous with supersymmetric quantum mechanics. For two scalar Hamiltonians $H_0$ and $H_1$, a first-order differential intertwining operator $A_1^\dagger$ satisfies

$H_1A_1^\dagger=A_1^\dagger H_0,$

with the associated superpotential determined by Riccati-type equations. The basic algebraic purpose of the construction is to relate different Hamiltonians whose eigenvalue problems can be solved algebraically (Contreras-Astorga et al., 2012).

Iteration enters when one replaces a single pair $(H_0,H_1)$ by a hierarchy $\{H_n\}$ linked by intertwiners $\{A_{n+1}^\dagger\}$. In that setting, each step removes a ground-state-like solution from the spectrum, and the action of the intertwining operators on ground states of higher Hamiltonians produces the excited states of the initial Hamiltonian. The chain enables systematic construction of an entire system’s spectrum and eigenfunctions, including all bound states, not just the ground state. This iterative reading is the core meaning of Hamiltonian intertwining in the exact-solvability literature (Contreras-Astorga et al., 2012).

A central sufficient condition for complete algebraic solvability is shape invariance. Two potentials $V_1(x;a_1)$ and $H_1A^\dagger=A^\dagger H_0$0 are shape invariant when

$H_1A^\dagger=A^\dagger H_0$1

for a parameter transformation $H_1A^\dagger=A^\dagger H_0$2 and a remainder $H_1A^\dagger=A^\dagger H_0$3 independent of $H_1A^\dagger=A^\dagger H_0$4. When an intertwining hierarchy possesses shape invariance, spectral data at one step relate simply to those at the next, and the full spectrum can be reconstructed from ground-state information plus repeated application of the intertwiners (Contreras-Astorga et al., 2012).

2. Relativistic realization: Dirac Hamiltonians in a magnetic field

A concrete relativistic implementation is provided by the Dirac equation for a spin-$H_1A^\dagger=A^\dagger H_0$5 particle in a magnetic field with cylindrical symmetry, whose intensity decreases as the distance to the symmetry axis grows and whose field lines are parallel to the $H_1A^\dagger=A^\dagger H_0$6-$H_1A^\dagger=A^\dagger H_0$7 plane. After separation into radial components and a unitary transformation, the problem is recast in terms of an effective radial Dirac Hamiltonian with block structure

$H_1A^\dagger=A^\dagger H_0$8

where $H_1A^\dagger=A^\dagger H_0$9 is a $H_0$0 matrix differential operator (Contreras-Astorga et al., 2012).

The iterative step is the introduction of a hierarchy of radial Dirac Hamiltonians $H_0$1 and block-diagonal intertwiners

$H_0$2

Here

$H_0$3

with $H_0$4 a $H_0$5 matrix function. The parameters satisfy

$H_0$6

and the explicit $H_0$7 obtained in the paper generalizes one-dimensional scalar intertwiners to matrix-valued Dirac operators (Contreras-Astorga et al., 2012).

The hierarchy is shape invariant, and this is the decisive structural property. It permits analytical determination of the full discrete spectrum,

$H_0$8

together with four-component eigenspinors obtained from repeated application of the intertwining operators. The paper formulates the resulting wavefunctions through expressions such as

$H_0$9

This is an exact relativistic example in which iteration, matrix structure, and shape invariance coincide (Contreras-Astorga et al., 2012).

3. Matrix Hamiltonians, higher-order operators, and polynomial supersymmetry

For one-dimensional $H_1$0 matrix Schrödinger Hamiltonians,

$H_1$1

the intertwining relation

$H_1$2

is imposed with an $H_1$3 matrix differential operator

$H_1$4

whose highest coefficient is constant and invertible. The kernel of $H_1$5 is invariant under $H_1$6, and explicit construction can be carried out from transformation vector-functions with nonvanishing Wronskian. For first-order operators this leads to

$H_1$7

These formulas make the iterative, spectral-design aspect transparent: the choice of transformation vector-functions fixes which bound states or Jordan structures are inserted into the partner Hamiltonian (Sokolov, 2014).

The matrix setting supports both chains of first-order transformations and genuinely higher-order constructions. A key result is that for any minimal-order intertwining operator $H_1$8, there exists another matrix differential operator $H_1$9, generally of different order, such that

$A_1^\dagger$0

and

$A_1^\dagger$1

with $A_1^\dagger$2 a polynomial in the համապատասխան Hamiltonian. This yields a polynomial supersymmetry algebra and extends ordinary first-order supersymmetry to higher-order matrix contexts (Sokolov, 2013).

Iteration in the matrix case is more subtle than in the scalar case. The literature develops criteria for minimizability and reducibility, but also shows that there are absolutely irreducible matrix intertwining operators, in contrast to the scalar case. Hence repeated intertwining need not reduce to a product of elementary Darboux-like steps. At the same time, explicit first-order spectral design remains possible: for a $A_1^\dagger$3 matrix Hamiltonian intertwined with the zero-potential Hamiltonian, one can add either up to two bound states for different energy values, or up to two bound states described by vector-eigenfunctions for the same energy value, or up to two bound states described by a vector-eigenfunction and an associated vector-function for the same energy value (Sokolov, 2013, Sokolov, 2014).

4. Beyond ordinary intertwining: spectral, non-isospectral, and antilinear variants

Standard intertwining does not exhaust the operator relations available in exactly solvable systems. A broader notion is the spectral intertwining relation for a parametrized Hamiltonian $A_1^\dagger$4,

$A_1^\dagger$5

This unifies ladder and intertwining relations in an $A_1^\dagger$6-dependent way and allows operators to connect eigenfunctions of different energy eigenvalues belonging to two different Hamiltonians. The paper introducing this framework emphasizes that such relations can connect two different Hamiltonians in ways that cannot be obtained by shape invariance, and it gives explicit examples for the hydrogen atom and the Rosen–Morse potential (Houri et al., 2017).

A different generalization drops isospectrality. Starting from $A_1^\dagger$7 and an operator $A_1^\dagger$8 with $A_1^\dagger$9, the traditional isospectral construction uses

$H_1A_1^\dagger=A_1^\dagger H_0,$0

when $H_1A_1^\dagger=A_1^\dagger H_0,$1 is invertible. The non-isospectral extension instead sets

$H_1A_1^\dagger=A_1^\dagger H_0,$2

without multiplying by $H_1A_1^\dagger=A_1^\dagger H_0,$3. Then eigenvectors of $H_1A_1^\dagger=A_1^\dagger H_0,$4 are still related to those of $H_1A_1^\dagger=A_1^\dagger H_0,$5 by $H_1A_1^\dagger=A_1^\dagger H_0,$6, but the eigenvalues are generally different. The procedure can be iterated, in principle, to produce chains of Hamiltonians with related but differing spectra, and it extends to crypto-hermitian Hamiltonians (Bagarello, 2011).

In non-self-adjoint finite-dimensional systems with complex eigenvalues, ordinary linear intertwining relations can fail because $H_1A_1^\dagger=A_1^\dagger H_0,$7 and $H_1A_1^\dagger=A_1^\dagger H_0,$8 are not isospectral. The remedy developed in this setting uses antilinear operators

$H_1A_1^\dagger=A_1^\dagger H_0,$9

leading to the antilinear partner $(H_0,H_1)$0 and the linear iterated operator

$(H_0,H_1)$1

The corresponding intertwining relations

$(H_0,H_1)$2

restore a version of the intertwining mechanism compatible with complex spectra. This enlarges the meaning of Hamiltonian intertwining beyond the Hermitian case and shows that iteration may involve linear and antilinear steps in alternation (Bagarello, 2016).

5. Integrable many-body systems and multidirectional iteration

In quantum integrable systems, intertwining can relate different couplings, different root multiplicities, or even different particle numbers. For the generalized Calogero–Moser–Sutherland Hamiltonian associated with the vector configuration $(H_0,H_1)$3, there exists a third-order differential operator $(H_0,H_1)$4 such that

$(H_0,H_1)$5

where $(H_0,H_1)$6 is the $(H_0,H_1)$7 Calogero–Moser–Sutherland Hamiltonian. The formal adjoint satisfies $(H_0,H_1)$8, and therefore

$(H_0,H_1)$9

This yields an order-6 quantum integral for $\{H_n\}$0. More generally, if $\{H_n\}$1 commutes with $\{H_n\}$2, then $\{H_n\}$3 commutes with $\{H_n\}$4, so higher-order integrals arise by conjugation through the intertwiner (Feigin et al., 2019).

A more explicitly multidirectional pattern appears in the quantum Calogero model. For fixed particle number, there are intertwining operators that increase or decrease the coupling constant by an integer amount; these are named horizontal. The 2025 work also constructs vertical intertwiners, which change the number of interacting particles for a fixed but integer value of the coupling constant. The resulting grid of intertwiners exists only in the algebraically integrable situation, and it allows one to obtain each Liouville charge from the free power sum in the particle momenta by iterated intertwining either horizontally or vertically. The paper further presents recursion formulae for the intertwiners as a factorization problem for partial differential operators and proves their existence for small values of particle number and coupling; as a byproduct, a new basis of non-symmetric Liouville integrals appears, algebraically related to the standard symmetric one (Correa et al., 1 Apr 2025).

These integrable examples show that iteration need not be one-dimensional. In some systems it is naturally organized as a linear hierarchy; in others it becomes a two-parameter grid. A plausible implication is that the geometry of the parameter space—coupling, multiplicity, particle number, or representation data—determines whether the intertwining structure is best viewed as a ladder, a chain, or a lattice.

6. Scope, limitations, and common misconceptions

A frequent misconception is to identify iterated intertwining with shape invariance. Shape invariance is a powerful sufficient mechanism for exact solvability, but it is not necessary: spectral intertwining relations were introduced precisely to connect eigenfunctions between different Hamiltonians in ways that cannot be obtained by previously known structures such as shape invariance (Houri et al., 2017).

A second misconception is that intertwining necessarily preserves the spectrum. The non-isospectral construction $\{H_n\}$5 shows that eigenvectors may remain explicitly related while eigenvalues change, and the procedure can still be iterated. Thus “intertwining” in current usage includes both isospectral and non-isospectral spectral engineering (Bagarello, 2011).

A third misconception is that any higher-order intertwiner is merely a product of first-order ones. Matrix theory invalidates that scalar intuition: reducibility criteria exist, but absolutely irreducible matrix intertwining operators also exist. This makes the matrix case structurally richer and places limits on naive factorization pictures of iterative schemes (Sokolov, 2013).

A fourth misconception concerns Hermiticity. In finite-dimensional non-self-adjoint systems with complex eigenvalues, standard linear intertwining between $\{H_n\}$6 and $\{H_n\}$7 can break down, and suitable antilinear operators are required to recover corresponding relations. Iteration in this setting remains possible, but its algebraic form differs from the Hermitian case (Bagarello, 2016).

Finally, the phrase should be distinguished from unrelated uses of “Hamiltonian” and “iterated” elsewhere in the literature. It is not the graph-theoretic theory of Hamiltonian paths in iterated line graphs, where the central objects are $\{H_n\}$8, traceability, and the hamiltonian path index (Nou et al., 2020). Nor is it the Floer-theoretic study of the behavior under iterations of the filtered and local Floer homology of a Hamiltonian, with cyclic group actions, supertraces, Lefschetz indices, and Smith-type inequalities (Cineli et al., 2019). Within quantum operator theory, by contrast, iterated Hamiltonian intertwining refers specifically to repeated intertwining constructions that organize solvability, spectral transport, and integrability.