Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2610.16613

Closed-form construction of exceptional points in a two-parametric PT-symmetric large-matrix model

Miloslav Znojil

quant-phmath-phSubmitted Oct 12, 2026• Announced Oct 13, 2026
arXivHTMLPDF

Abstract

Stable bound states living on a discrete one-dimensional N-point lattice are assumed generated by a certain two-parametric PT symmetric potential $V=V(A,B)$. At any $N$, the coordinates $A=A^{(EP)}$ and $B=B^{(EP)}$ of the Kato's exceptional point (marking the phase-transition instant of the loss of stability) are defined, exactly, via roots of a certain elementary quartic polynomial.

AI Summary

The paper studies exact localization of exceptional points (EPs) in a finite $\mathcal{PT}$-symmetric tight-binding chain. The $N\times N$ Hamiltonian has nearest-neighbour hopping $-1$, imaginary on-site potentials $-iA,-iB$ at one end and $iB,iA$ at the other, and zero potential elsewhere. The objective is to locate higher-order degeneracies at zero energy for arbitrary chain length without relying on numerical root searches. The main new construction concerns odd $N=2K+1$, extending earlier symbolic calculations restricted to $N\leq13$. Introducing $X=1+AB$ simplifies the two lowest nontrivial coefficients of the characteristic polynomial. Setting both coefficients to zero produces a zero eigenvalue of multiplicity at least five and reduces parameter localization to a quartic equation in $X$, with coefficients explicit in $K$. The original parameters follow from $A^2=[2+(K-1)X^2]/K$ and $B=(X-1)/A$. At large $K$, two real solution branches approach $X=\pm\sqrt{2}$; the remaining two quartic roots become nonreal. The paper also explains the one-parameter parity dependence: at $B=0$, the critical coupling is $|A|=1$ for even $N$ and $|A|=\sqrt{(K+1)/K}$ for odd $N$. Even-length results are largely recalled from earlier work, with an explicit $N=4$ construction and Jordan-chain verification. The approach is restricted to two boundary parameters; extensions to more parameters remain unresolved. General large-$N$ Jordan-form verification is discussed as requiring numerical checks rather than supplied explicitly.

No sign of AI-generated textIncrementalOverclaimsUnclear writing

Error found: Section 3.2 gives incorrect B coordinates for its negative-A EP4 roots: the corresponding values are approximately +1.691739510 and −0.4060952086, respectively. Section 4.3's second-order large-K root coefficients should be 3∓189√2/128 rather than 1∓189√2/128, and consequently both coefficients are positive rather than having opposite signs.

Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.

Comments
24 pp., 2 figures

Related papers

PT-symmetric Quantum Chain Models

Miloslav Znojil

A review is given of certain tridiagonal N-dimensional non-Hermitian J-parametric real-matrix quantum Hamiltonians H(N). The domains Ɗ(N) of reality of their spectra of energies are studied, with particular attention paid to their exceptional-point boundaries ∂Ɗ(N). The strongest admissible couplings are specified in closed form for all N.

Acta Polytechnica•Jan 2, 2007•5 citations

Asymptotic non-Hermitian degeneracy phenomenon and its exactly solvable simulation

Miloslav Znojil

Up to these days, the popular PT-symmetric imaginary cubic oscillator did not find any consistent probabilistic quantum-mechanical interpretation because its Hamiltonian has been shown, by mathematicians, intrinsic-exceptional-point (IEP) singular. In the paper we explain why there is even no reasonable small-perturbation-based regularization of the similar unacceptable (i.e., IEP-singular) quantum models. The explanation is based on a partial formal analogy of the IEP singularity with the conventional exceptional point (EP). What is important is that we are able to construct a simplified $N$ by $N$-matrix (and exactly solvable) toy-model Hamiltonian admitting the asymptotic (i.e., high-excitation) EP-related wave-function degeneracy which, in some sense (i.e., in the limit of large $N$) mimics several aspects of its IEP analogue. In this comparison, the difference is that the regularization of the EP singularities is possible (using an ad hoc perturbation of size ${\cal O}(1/N)$) while an analogous regularization of the IEP singularity is not (we have to consider $N \to \infty$).

quant-pharXiv:2603.13141•Symmetry•Mar 16, 2026•0 citations

Exceptional points and domains of unitarity for a class of strongly non-Hermitian real-matrix Hamiltonians

Miloslav Znojil

A family of non-Hermitian real and tridiagonal-matrix candidates H(N)(λ)=H0(N)+λW(N)(λ) for a hiddenly Hermitian (a.k.a. quasi-Hermitian) quantum Hamiltonian is proposed and studied. Fairly weak assumptions are imposed upon the unperturbed matrix [the square-well-simulating spectrum of H0(N) is not assumed equidistant)] and upon its maximally non-Hermitian N-parametric antisymmetric-matrix perturbations [matrix W(N)(λ) is not even required to be PT-symmetric]. Despite that, the “physical” parametric domain D[N] is (constructively) shown to exist, guaranteeing that in its interior, the spectrum remains real and non-degenerate, rendering the quantum evolution unitary. Among the non-Hermitian degeneracies occurring at the boundary ∂D[N] of the domain of stability, our main attention is paid to their extreme version corresponding to Kato’s exceptional point of order N (EPN). The localization of the EPNs and, in their vicinity, of the quantum-phase-transition boundaries ∂D[N] is found feasible, at the not too large N, using computer-assisted symbolic manipulations, including, in particular, the Gröbner-basis elimination and the high-precision arithmetics.

arXiv:2104.11016•Journal of Mathematical Physics•May 1, 2021•9 citations

Two-parametric PT-symmetric quartic family

Alexandre Erëmenko, Andrei M. Gabrielov

We describe a parametrization of the real spectral locus of the two-parametric family of PT-symmetric quartic oscillators. For this family, we find a parameter region where all eigenvalues are real, extending the results of Dorey et al (2007 J. Phys. A: Math Theor. 40 R205–83) and Shin (2005 J. Phys. A: Math. Gen. 38 6147–66; 2002 Commun. Math. Phys. 229 543–64).

arXiv:1105.2742•Journal of Physics A Mathematical and Theoretical•Apr 12, 2012•1 citation

The Intrinsic Exceptional Point: A Challenge in Quantum Theory

Miloslav Znojil

In spite of its unbroken PT symmetry, the popular imaginary cubic oscillator Hamiltonian H(IC)=p2+ix3 does not satisfy all of the necessary postulates of quantum mechanics. This failure is due to the “intrinsic exceptional point” (IEP) features of H(IC) and, in particular, to the phenomenon of a high-energy asymptotic parallelization of its bound-state-mimicking eigenvectors. In this paper, it is argued that the operator H(IC) (and the like) can only be interpreted as a manifestly unphysical, singular IEP limit of a hypothetical one-parametric family of certain standard quantum Hamiltonians. For explanation, ample use is made of perturbation theory and of multiple analogies between IEPs and conventional Kato’s exceptional points.

arXiv:2411.12501•Foundations•Mar 1, 2025•4 citations

Discrete PT-symmetric square-well oscillators

Miloslav Znojil

Exact solvability of the discretized N-point version of the PT-symmetric square-well model is pointed out. Its wave functions are found proportional to the classical Tshebyshev polynomials of a complex argument. At all N a compact secular equation is derived giving the real spectrum of energies at any non-Hermiticity strength Z below its finite and weakly N-dependent critical value. In the limit of vanishing Z the model degenerates to a Hermitian Hueckel Hamiltonian.

arXiv:quant-ph/0511110•Nov 11, 2005•0 citations
Stable bound states living on a discrete one-dimensional N-point lattice are assumed generated by a certain two-parametric PT symmetric potential V=V(A,B)V=V(A,B)V=V(A,B). At any NNN, the coordinates A=A(EP)A=A^{(EP)}A=A(EP) and B=B(EP)B=B^{(EP)}B=B(EP) of the Kato's exceptional point (marking the phase-transition instant of the loss of stability) are defined, exactly, via roots of a certain elementary quartic polynomial.
The paper studies exact localization of exceptional points (EPs) in a finite PT\mathcal{PT}PT-symmetric tight-binding chain. The N×NN\times NN×N Hamiltonian has nearest-neighbour hopping −1-1−1, imaginary on-site potentials −iA,−iB-iA,-iB−iA,−iB at one end and iB,iAiB,iAiB,iA at the other, and zero potential elsewhere. The objective is to locate higher-order degeneracies at zero energy for arbitrary chain length without relying on numerical root searches.
The main new construction concerns odd N=2K+1N=2K+1N=2K+1, extending earlier symbolic calculations restricted to N≤13N\leq13N≤13. Introducing X=1+ABX=1+ABX=1+AB simplifies the two lowest nontrivial coefficients of the characteristic polynomial. Setting both coefficients to zero produces a zero eigenvalue of multiplicity at least five and reduces parameter localization to a quartic equation in XXX, with coefficients explicit in KKK. The original parameters follow from A2=[2+(K−1)X2]/KA^2=[2+(K-1)X^2]/KA2=[2+(K−1)X2]/K and B=(X−1)/AB=(X-1)/AB=(X−1)/A. At large KKK, two real solution branches approach X=±2X=\pm\sqrt{2}X=±2​; the remaining two quartic roots become nonreal. The paper also explains the one-parameter parity dependence: at B=0B=0B=0, the critical coupling is ∣A∣=1|A|=1∣A∣=1 for even NNN and ∣A∣=(K+1)/K|A|=\sqrt{(K+1)/K}∣A∣=(K+1)/K​ for odd NNN.
Even-length results are largely recalled from earlier work, with an explicit N=4N=4N=4 construction and Jordan-chain verification. The approach is restricted to two boundary parameters; extensions to more parameters remain unresolved. General large-NNN Jordan-form verification is discussed as requiring numerical checks rather than supplied explicitly.
Up to these days, the popular PT-symmetric imaginary cubic oscillator did not find any consistent probabilistic quantum-mechanical interpretation because its Hamiltonian has been shown, by mathematicians, intrinsic-exceptional-point (IEP) singular. In the paper we explain why there is even no reasonable small-perturbation-based regularization of the similar unacceptable (i.e., IEP-singular) quantum models. The explanation is based on a partial formal analogy of the IEP singularity with the conventional exceptional point (EP). What is important is that we are able to construct a simplified NNN by NNN-matrix (and exactly solvable) toy-model Hamiltonian admitting the asymptotic (i.e., high-excitation) EP-related wave-function degeneracy which, in some sense (i.e., in the limit of large NNN) mimics several aspects of its IEP analogue. In this comparison, the difference is that the regularization of the EP singularities is possible (using an ad hoc perturbation of size O(1/N){\cal O}(1/N)O(1/N)) while an analogous regularization of the IEP singularity is not (we have to consider N→∞N \to \inftyN→∞).
A family of non-Hermitian real and tridiagonal-matrix candidates H(N)(λ)=H0(N)+λW(N)(λ) for a hiddenly Hermitian (a.k.a. quasi-Hermitian) quantum Hamiltonian is proposed and studied. Fairly weak assumptions are imposed upon the unperturbed matrix [the square-well-simulating spectrum of H0(N) is not assumed equidistant)] and upon its maximally non-Hermitian N-parametric antisymmetric-matrix perturbations [matrix W(N)(λ) is not even required to be PT-symmetric]. Despite that, the “physical” parametric domain D[N] is (constructively) shown to exist, guaranteeing that in its interior, the spectrum remains real and non-degenerate, rendering the quantum evolution unitary. Among the non-Hermitian degeneracies occurring at the boundary ∂D[N] of the domain of stability, our main attention is paid to their extreme version corresponding to Kato’s exceptional point of order N (EPN). The localization of the EPNs and, in their vicinity, of the quantum-phase-transition boundaries ∂D[N] is found feasible, at the not too large N, using computer-assisted symbolic manipulations, including, in particular, the Gröbner-basis elimination and the high-precision arithmetics.
In spite of its unbroken PT symmetry, the popular imaginary cubic oscillator Hamiltonian H(IC)=p2+ix3 does not satisfy all of the necessary postulates of quantum mechanics. This failure is due to the “intrinsic exceptional point” (IEP) features of H(IC) and, in particular, to the phenomenon of a high-energy asymptotic parallelization of its bound-state-mimicking eigenvectors. In this paper, it is argued that the operator H(IC) (and the like) can only be interpreted as a manifestly unphysical, singular IEP limit of a hypothetical one-parametric family of certain standard quantum Hamiltonians. For explanation, ample use is made of perturbation theory and of multiple analogies between IEPs and conventional Kato’s exceptional points.