Discrete Schrodinger equation on graphs: An effective model for branched quantum lattice

C. Trunk; D. Matrasulov; J. Yusupov; M. Akramov

REVIEW 2 major objections 5 minor 43 references

Discrete Schrodinger equation on graphs: An effective model for branched quantum lattice

T0 review · 2 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A single determinant condition gives the spectrum of the discrete Schrödinger equation on any graph, with star-graph eigenvalues converging to the continuous limit as the lattice step shrinks.

desk verdict A useful secular-determinant recipe for discrete quantum graphs, with a real degeneracy gap at k=0 and k=±2/a that the paper never mentions; otherwise the math is standard but sound. read the letter →

arxiv 2411.14397 v1 pith:CTZYDHSE submitted 2024-11-21 quant-ph cond-mat.mes-hall

M. Akramov , C. Trunk , J. Yusupov , D. Matrasulov This is my paper · ORCID

classification quant-phcond-mat.mes-hall MSC 81Q3505C5039A12

keywords discreteSchrödingerequationquantumgraphssecularbranchedlatticesstargraphconductingpolymersself-adjointvertexconditionscontinuumlimit

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper proposes a way to solve the discrete Schrödinger equation on a graph whose edges are finite one-dimensional lattices rather than continuous wires. On each edge it writes the wave function as a combination of two exact lattice modes, $g_\pm(a)^n$, and joins the edges at vertices by imposing continuity of the wave function plus a discrete current-conservation rule. The central result is a secular equation, $\det M(k)=0$, of the same size and structure as in continuous quantum-graph theory, whose roots are the eigenvalues of the branched lattice. A three-edge star graph is worked out explicitly, and its first five eigenvalues converge to the continuous star-graph values as the step size $a$ goes to zero. If correct, this provides an exact spectral tool for branched molecular chains and conducting polymers, where previously only continuous or approximate treatments were available.

What carries the argument

The carrying object is the exact lattice solution $g_\pm(a)=1+\frac{-k^2a^2\pm ka\sqrt{k^2a^2-4}}{2}$, with $g_+(a)^n$ and $g_-(a)^n$ playing the roles that $e^{\pm ikx}$ play on a continuous edge. The paper forms the adapted combination (33), $f_{i,j}(n)=g_+^n-g_-^n$, which automatically satisfies the continuity conditions when written as $A_{i,j}f_{i,j}(N_{i,j})$, and feeds the discrete current balance (32). This combination converts the vertex rules into a homogeneous linear system whose coefficient matrix $M(k)$ yields the determinant condition (37). The same $g_\pm$ also gives the continuum limit $g_\pm^n\to e^{\pm ikx}$ as $a\to 0$, which is why the spectra converge.

What would settle it

Take a small graph, e.g., two edges with a few lattice points, impose conditions (31)–(32), compute the roots of $\det M(k)=0$, and then diagonalize the full matrix $H_d$ (including the vertex rows defined by these conditions) and compare the two spectra; any mismatch between the root set and the true eigenvalues falsifies the secular-equation claim.

Watch

Extended reading notes

Core claim

The central claim is that the spectrum of the discrete Schrödinger operator on an arbitrary graph with vertex conditions (31)–(32) coincides with the roots of $\det M(k)=0$, where $M(k)$ is assembled from the edge functions $f_{i,j}(n)=g_+(a_{i,j})^n-g_-(a_{i,j})^n$. The paper shows that the edge solution (33) is an exact solution of the finite-difference equation (28), that its $a\to 0$ limit reproduces the continuous edge solution, and that the discrete vertex rules reduce to the same algebraic system as the continuous ones. For the star graph with three edges of rationally independent lengths and Kirchhoff vertex conditions, the first five eigenvalues are computed and shown numerically to approach the continuous star-graph eigenvalues as $a$ decreases (Table 2).

Load-bearing premise

Everything rests on the assumption that the discrete matching rules at the vertices—wave-function values agree and a discrete current is conserved—make the graph's energy operator behave like its continuous counterpart; the paper borrows this from the continuous case by analogy and does not prove the boundary terms cancel.

Editorial extensions

If this is right

For any graph with discrete edges, the spectrum can be obtained by solving a single algebraic equation, $\det M(k)=0$, rather than by diagonalizing the full lattice Hamiltonian.
As each step size $a_{i,j}\to 0$, the discrete eigenvalues tend to the continuous quantum-graph eigenvalues given by Eq. (27), giving a controlled discretization scheme for branched quantum wires.
For a finite chain the eigenvalues are exactly $k_m=\pm\frac{1}{a}\sqrt{2-2\cos(\pi m/N)}$ under Dirichlet conditions and $\pm\frac{1}{a}\sqrt{2-2\cos(\pi m/(N-1))}$ under Neumann conditions, reproducing $\pi m/L$ in the limit.
The determinant structure matches the continuous case, so existing computational strategies for continuous quantum graphs can be adapted to discrete networks.
The model offers a starting point for computing band spectra and charge transport in branched conducting polymers, as discussed in the paper.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If the self-adjointness gap is filled, the determinant condition should extend to graphs with loops and multiple edges by adding the corresponding continuity and current equations; the paper currently excludes these cases.
The dispersion relation encoded in $g_\pm$ is the tight-binding relation $2-2\cos q$, so the graph secular equation can be read as a scattering condition for tight-binding waves, suggesting a direct link to transfer-matrix or Green's-function methods for disordered networks.
A small-graph numerical comparison—directly diagonalizing the full matrix $H_d$ under the vertex rules versus rooting $\det M(k)=0$—would settle the self-adjointness question and is a natural first validation before applying the model to polymer spectra.
The paper avoids degeneracies by choosing rationally independent edge lengths; a limiting analysis of the degenerate equal-length star could provide a discrete analogue of symmetry-induced spectral multiplicity.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Punchline: the paper gives a workable way to write down eigenvalues for discrete Schrödinger equations on star-type networks, with an explicit determinant and a clean continuum limit. It is not a breakthrough: the "new exact solution" is just the standard solution of a constant-coefficient recurrence, and the graph construction is a discrete analogue of Kottos–Smilansky. But as an effective tool for branched lattice models it is honest and mostly correct.

What is good: the single-edge eigenvalue formulas (12) and (17) are right, and the continuum limits are computed carefully. Table 2 shows genuine convergence for the star graph, and the determinant in (41) has the right structure. There are no fitted parameters; the derivation is straightforward and readable.

The main soft spot is the exceptional-point degeneracy. At k=0 and k=±2/a, the characteristic roots coincide, so the basis functions f_ij(n)=g_+^n−g_-^n vanish identically on every edge. The ansatz (33) cannot represent nonzero solutions at those wavenumbers, yet det M(k)=0 can still hold with only a zero wavefunction. A concrete example: the Neumann chain has a constant zero mode at k=0, but feeding k=0 into (35)–(36) forces all φ_i=0 and gives a trivial wavefunction. The determinant is zero, but not because of a genuine eigenfunction. The same happens at k=±2/a, where the determinant has spurious roots. The low-lying eigenvalues in Table 2 are far from these exceptional values, so the numerics are unaffected, but the central claim that the spectrum is exactly the roots of det M(k)=0 is not correct as stated. This needs either an explicit exclusion of those k values, a separate treatment using the generalized solution (A+Bn)r^n, or a modified basis that does not collapse. A referee should ask for this.

The self-adjointness gap the reader flagged is less serious. The paper asserts by analogy, but I checked: with continuity (31) and the discrete current conservation (32) using real λ_i, the boundary form (30) does telescope to zero. The proof is missing, not the result. It should be added, but it is not a load-bearing flaw.

The novelty is overstated in the abstract, and the paper does not compare with existing discrete Schrödinger operator literature, but neither is fatal. For someone modeling branched conducting polymers or discrete lattices, this is a usable recipe once the exceptional-point issue is fixed. It deserves a serious referee, not a desk reject, but I would not cite the exact bijection as stated until the degeneracy is addressed.

Referee Report

2 major / 5 minor

Summary. The paper proposes an exact-solution approach to the discrete Schrödinger equation on finite graphs. It first derives the general solution in Eq. (3) for a one-dimensional chain, obtains Dirichlet and Neumann spectra in Eqs. (12) and (17), and shows their continuum limits. It then recalls the standard continuous quantum graph formulation and, for a discrete graph, imposes continuity (31) and discrete current conservation (32) at vertices, writes the edge solution as Eq. (33), and claims that the spectrum of the discrete graph Hamiltonian is given by the roots of det(M(k))=0 in Eq. (37). The method is specialized to a three-edge star graph with Kirchhoff conditions, yielding the explicit matrix M(k) in Eq. (41) and numerical eigenvalues in Table 2 that converge to the continuous-star-graph values. The paper concludes with a discussion of conducting polymers as a possible application.

Significance. The manuscript has concrete strengths: the discrete solution and secular equations are derived analytically, the continuum limits are computed explicitly, there are no fitted parameters, and the star-graph calculation provides a transparent, machine-checkable example. If the central equivalence between eigenvalues and roots of det(M(k))=0 were fully established, the approach would be a useful exact tool for branched lattice models. However, two load-bearing mathematical points are not addressed: the self-adjointness of the discrete vertex conditions is asserted but not proved, and the edge ansatz degenerates at k=0 and k=±2/a, so the claimed bijection between eigenvalues and determinant roots fails at those exceptional wavenumbers. These gaps do not invalidate the numerical low-lying eigenvalues but do mean the central claim, as stated for arbitrary graphs, is incomplete.

major comments (2)

[Branched lattices, arbitrary branching topology, Eqs. (30)–(32)] The paper does not prove that the discrete vertex conditions (31) and (32) make Hd self-adjoint. Eq. (30) gives the boundary form, and the text asserts that Hd is self-adjoint if and only if this form vanishes, but no domain is specified and no calculation shows that (31)–(32) force (30) to vanish for all admissible Φ. Since Hd is finite-dimensional, symmetry would suffice, but the domain of Hd—which functions on the edges are allowed and which vertex conditions are built into the domain—is never defined. Without this, the statement that the eigenvalues of Hd are exactly the roots of det(M(k))=0 is not supported.
[Branched lattices, arbitrary branching topology, Eqs. (33) and (37)] The central equivalence 'spectrum = roots of det M(k)=0' fails at the exceptional wavenumbers k=0 and k=±2/a. At these values g_+(a)=g_-(a), so f_{i,j}(n)=g_+(a)^n-g_-(a)^n vanishes for every n, and both basis functions in Eq. (33) are identically zero on every edge. The ansatz therefore cannot represent nonzero solutions of the discrete Schrödinger equation at these wavenumbers. A concrete example is a single edge with two vertices and λ_1=λ_2=0, which is the discrete Neumann chain; it has a constant eigenfunction at k=0, but Eq. (35) forces φ_1=φ_2=0 and Eq. (33) gives a zero wavefunction. Thus a genuine eigenvalue is missed, while det(M(0))=0 can be a spurious root. The same degeneracy at k=±2/a may produce additional spurious roots or miss eigenvalues for certain edge lengths. Because the numerical examples in Table 2 use low nonzero eigenvalues far from k=0 and k=±2/a, they do not detect this gap. The exceptional cases must be treated separately by a limiting or Jordan-form argument, or the claim must be restricted to non-exceptional k.

minor comments (5)

[Abstract] The word 'explicitely' should be 'explicitly'.
[Eq. (15)] The printed form of the second row of the determinant is ambiguous: the terms should read g_+(a)^N - g_+(a)^{N-1} and g_-(a)^N - g_-(a)^{N-1}, not 'g+(a)N − g+(a)N − 1' as written.
[Text before Eq. (34)] The phrase 'as it is shown in the section' is vague and should refer explicitly to the earlier single-chain section.
[Possible experimental realization] The word 'pol-conjugated' appears to be a typo for 'π-conjugated' in the context of conducting polymers.
[Table 2] The table would be more informative if the caption stated the method used to solve det(M(k))=0 and the error measure used to establish convergence to the continuous values.

Circularity Check

1 steps flagged · score 1.0 of 10

Central derivation is self-contained and analytic; only a mild by-construction element is the choice of discrete vertex conditions that mirror the continuous ones.

other [Branched lattices section, Eqs. (31)-(32), and Conclusion]
"Similar as in the previous section, we impose a continuity condition and some current conservation rule. The continuity condition reads as ... The discrete version of the current conservation ... It is shown that the secular equation derived for star graph reproduces its well-know continuum counterpart."

The discrete vertex conditions (31)-(32) are taken over, by analogy, from the continuous graph conditions (22)-(23) rather than derived from the vanishing of the discrete boundary form (30). Consequently, the later statement that the discrete secular equation reproduces its continuous counterpart, and the convergence shown in Table 2, are partly guaranteed by this choice of identical vertex rules rather than being an independent output of the discrete model. This is mild design-by-analogy, not a fitted-parameter circularity: the solution (33) and the secular determinant (37) are still obtained by exact substitution and algebraic solution.

full rationale

The paper's core derivation is self-contained and parameter-free. The solution (3) is the general solution of the constant-coefficient recurrence (2), and the Dirichlet and Neumann secular equations (8) and (16) are obtained by direct substitution; the eigenvalues (12) and (17) follow by standard algebra. For graphs, the edge solution (33) is again an exact linear combination of fundamental solutions, and the secular system (35)-(36) and determinant condition (37) are derived by substitution into the stated vertex conditions. No fitted quantity is renamed as a prediction, and the continuum limits (34) and Table 2 are independent consistency checks of a finite-difference approximation. Self-citations (e.g., Refs. [9,12-14,34,40]) appear in the introduction and in the application discussion and are not load-bearing for the spectral equations; the quantum-graph formulation is attributed to the external reference [6]. The only mild caveat is that the discrete vertex conditions are chosen to mirror the continuous ones, making the continuum limit partly by construction; this is a modeling assumption rather than a circular derivation. Separately, for completeness as a correctness risk, the ansatz basis functions in (33) vanish identically when g_+ = g_-, i.e., at k = 0 and k = ±2/a, so the equivalence 'spectrum = roots of det M(k)' may be incomplete at those exceptional points; this is a mathematical gap, not a circularity.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

No parameters are fitted to data. The step size a_{i,j} is a discretization parameter that vanishes in the continuum limit, and the star graph lengths (0.8, 1.1, 1.5) are chosen for the example, not fitted. The main assumptions are the finite-difference discretization, the Kirchhoff-type vertex conditions, and the implicit claim that these conditions make the discrete graph Hamiltonian self-adjoint.

assumptions (3)

standard math The finite-difference equation (2) is the correct discrete Schrödinger equation, and the solution space is spanned by g_+^n and g_-^n.
This is the standard constant-coefficient recurrence; the paper presents it as new but it follows from the ansatz Ψ_n = g^n.
domain assumption The vertex conditions (31)-(32), continuity plus discrete current conservation, make the discrete graph Hamiltonian Hd self-adjoint.
The paper asserts this by analogy with the continuous case (Eq. 30 ff.) but gives no proof that the boundary form (30) vanishes under (31)-(32).
ad hoc to paper The edge lengths and step sizes are such that the secular determinant has non-degenerate roots (rationally independent lengths are chosen for the star graph).
The paper states that rationally independent lengths avoid degenerate eigenvalues; the general graph case is not analyzed.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Discrete Schrodinger equation on graphs: An effective model for branched quantum lattice." pith.science (2026). https://pith.science/paper/CTZYDHSE

@misc{pith2026241114397,
  author       = {Pith},
  title        = {Pith review of: Discrete Schrodinger equation on graphs: An effective model for branched quantum lattice},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CTZYDHSE}},
  note         = {Machine review of arXiv:2411.14397}
}

read the original abstract

We propose an approach to quantize discrete networks (graphs with discrete edges). We introduce a new exact solution of discrete Schrodinger equation that is used to write the solution for quantum graphs. Formulation of the problem and derivation of secular equation for arbitrary quantum graphs is presented. Application of the approach for the star graph is demonstrated by obtaining eigenfunctions and eigenvalues explicitely. Practical application of the model in conducting polymers and branched molecular chains is discussed.

Figures

Figures reproduced from arXiv: 2411.14397 by the authors.

**Figure 2.** Discrete star graph. It satisfies hHdΨ, Φi − hΨ, HdΦi = X V i,j=1 i<j Ci,j Ψ (ai,j ) i,j (0)Φ(ai,j )∗ i,j (ai,j ) − Ψ (ai,j ) i,j (ai,j )Φ(ai,j )∗ i,j (0) + Ψ(ai,j ) i,j (Li,j )Φ(ai,j )∗ i,j (Li,j − ai,j ) − Ψ (ai,j ) i,j (Li,j − ai,j )Φ(ai,j )∗ i,j (Li,j ) . (30) In fact, Hd is a matrix, and, hence, it is self-adjoint if and only if (30) is zero. Similar as in the previous section, we impose a continuity condit… view at source ↗

**Figure 3.** Conducting polymer chains. Possible experimental realization of the model. Here we briefly discuss the possible practical application of our model to so-called conducting polymers. These latter are the pol-conjugated (polymer) molecular chains exhibiting semiconducting electronic properties and therefore are called organic semiconductors. Conducting polymers are the basic functional materials for organic electronic… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 43 canonical work pages

[1]

Pauling, J

L. Pauling, J. Chem. Phys. 4, 673 (1936)

work page 1936
[2]

Ruedenberg, C.W

K. Ruedenberg, C.W. Scherr, J. Chem. Phys. 21, 1565 (1953)

work page 1953
[3]

Alexander, Phys

S. Alexander, Phys. Rev. B 27, 1541 (1985)

work page 1985
[4]

Exner, P

P. Exner, P. Seba, P. Stovicek, J. Phys. A: Math. Gen. 21, 4009 (1988)

work page 1988
[5]

Kostrykin, R

V. Kostrykin, R. Schrader J. Phys. A: Math. Gen. 32, 595 (1999)

work page 1999
[6]

Kottos, U

T. Kottos, U. Smilansky, Ann. Phys., 76, 274 (1999)

work page 1999
[7]

Exner, H

P. Exner, H. Kovarik, Quantum waveguides , (Springer, 2015)

work page 2015
[8]

Hul et al , Phys

O. Hul et al , Phys. Rev. E 69, 056205 (2004)

work page 2004

Show all 43 references

[9]

Matrasulov, J.R

D.U. Matrasulov, J.R. Yusupov, K.K. Sabirov, J. Phys. A 52, 155302 (2019)

2019
[10]

Berkolaiko, P

G. Berkolaiko, P. Kuchment, Introduction to Quantum Graphs, Mathematical Surveys and Monographs AMS (2013)

2013
[11]

Kurasov, Spectral Geometry of Graphs , Springer- Verlag, Berlin (2024)

P. Kurasov, Spectral Geometry of Graphs , Springer- Verlag, Berlin (2024)

2024
[12]

Yusupov, K.K

J.R. Yusupov, K.K. Sabirov, M. Ehrhardt, D.U. Matra- sulov, Phys. Lett. A 383, 2382 (2019)

2019
[13]

Matrasulov, K

J. Matrasulov, K. Sabirov, Physica A 608, 128279 (2022)

2022
[14]

Sabirov, J

K.K. Sabirov, J. Yusupov, D. Jumanazarov, D. Matra- sulov, Phys. Lett. A 382, 2856 (2018)

2018
[15]

Dreisow, M

F. Dreisow, M. Heinrich, R. Keil, et al , Phys. Rev. Lett., 105, 143902 (2010)

2010
[16]

Exner, O

P. Exner, O. Turek, J. Phys. A: Math. Theor. 50 455201 (2017)

2017
[17]

E Korotyaev, N Saburova, J. Math. Anal. Appl. 436 104 (2016)

2016
[18]

V.Barrera-Figueroa, V.S.Rabinovich, J. Phys. A: Math . Theor. 50 215207 (2017)

2017
[19]

L. Alon, R. Band, G. Berkolaiko, Commun. Math. Phys., 362 909–948 (2018)

2018
[20]

Exner, J

P. Exner, J. Lipovsk´ y, Phys. Lett. A, 384 (18), 126390, (2020)

2020
[21]

Rezapour, C.T

S. Rezapour, C.T. Deressa, A. Hussain, S. Etemad, R. George, B. Ahmad, Mathematics. 10(4), 568 2022

2022
[22]

Baleanu, S

D. Baleanu, S. Etemad, H. Mohammadi, S. Rezapour, Communications in Nonlinear Science and Numerical Simulation, 100, 105844 (2021)

2021
[23]

Etemad, S

S. Etemad, S. Rezapour, Advances in Diﬀerence Equa- tions volume 2020, 276 (2020)

2020
[24]

Baleanu, S.M

D. Baleanu, S.M. Aydogn, H. Mohammadi, S. Rezapour, Alexandria Engineering Journal, 59, 3029 (2020)

2020
[25]

H. Khan, K. Alam, H. Gulzar, S. Etemad, S. Rezapour, Mathematics and Computers in Simulation, 198, 455 (2022)

2022
[26]

Hussain, E.N

S. Hussain, E.N. Madi, H. Khan, H. Gulzar, S. Etemad, S. Rezapour, M.K.A. Kaabar, Journal of Function Spaces 9(23), 4320865 (2022)

2022
[27]

Ahmad, A

M. Ahmad, A. Zada, M. Ghaderi, R. George, S. Rezapour, Fractal and Fractional 6(4), 203 (2022)

2022
[28]

H. Khan, J. Alzabut, A. Shah, Z.-Y. He, S. Etemad, S. Rezapour, A. Zada, Fractals 31 (04), 2340055 (2023)

2023
[29]

Aydogan, D

S.M. Aydogan, D. Baleanu, H. Mohammadi, S. Rezapour, Advances in Diﬀerence Equations 2020, 382 (2020)

2020
[30]

A Kvitsinsky, J

A. A Kvitsinsky, J. Phys. A. 27, 215 (1994)

1994
[31]

E.Tarasov, Phys

V. E.Tarasov, Phys. Lett. A, 380 68 (2016)

2016
[32]

E Korotyaev, N Saburova, Math. Annal. 377 723 (2020)

2020
[33]

E Korotyaev, N Saburova, J. Math. Anal. Appl. 420 576 (2014)

2014
[34]

Yusupov, Kh.Sh

J.R. Yusupov, Kh.Sh. Matyokubov, K.K. Sabirov, D.U. Matrasulov, Chem. Phys., 537, 110861 (2020)

2020
[35]

Akramov, J.R

M.E. Akramov, J.R. Yusupov, I.N. Askerzade, D.U. Ma- trasulov, Physica Scripta, 98, 115238 (2023)

2023
[36]

Heeger, R

A.J. Heeger, R. Pethig, Philos. Trans. R. Soc. Lond. A 314, 17 (1985)

1985
[37]

Heeger, Rev

A.J. Heeger, Rev. Mod. Phys. 73, 681 (2001)

2001
[38]

Heeger, S

A.J. Heeger, S. Kivelson, J.R. Schrieﬀer, W.-P. Su, Rev . Mod. Phys. 60, 781 (1988)

1988
[39]

Soganci, O

T. Soganci, O. Gumusay, H.C. Soyleyici, M. Ak, Polymer 134, 187 (2018)

2018
[40]

Sabirov, J.R

K.K. Sabirov, J.R. Yusupov, Kh.Sh. Matyokubov, NANOSYSTEMS: Phys., Chem., Math. 11, 183 (2020)

2020
[41]

Huang, X

J. Huang, X. Chen, X. Li, J. Wang, AAPPS Bulletin 33, 14 (2023)

2023
[42]

B. Lu, L. Liu, J.-Y. Song, K. Wen, Ch. Wang, AAPPS Bulletin 33, 7 (2023)

2023
[43]

Y. Li, T. Xin, Ch. Qiu, K. Li, G. Liu, J. Li, Y. Wan, D. Lu, Fundamental Research 3, 229 (2023). p-7

2023

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

Pauling, J

L. Pauling, J. Chem. Phys. 4, 673 (1936)

work page 1936

[2] [2]

Ruedenberg, C.W

K. Ruedenberg, C.W. Scherr, J. Chem. Phys. 21, 1565 (1953)

work page 1953

[3] [3]

Alexander, Phys

S. Alexander, Phys. Rev. B 27, 1541 (1985)

work page 1985

[4] [4]

Exner, P

P. Exner, P. Seba, P. Stovicek, J. Phys. A: Math. Gen. 21, 4009 (1988)

work page 1988

[5] [5]

Kostrykin, R

V. Kostrykin, R. Schrader J. Phys. A: Math. Gen. 32, 595 (1999)

work page 1999

[6] [6]

Kottos, U

T. Kottos, U. Smilansky, Ann. Phys., 76, 274 (1999)

work page 1999

[7] [7]

Exner, H

P. Exner, H. Kovarik, Quantum waveguides , (Springer, 2015)

work page 2015

[8] [8]

Hul et al , Phys

O. Hul et al , Phys. Rev. E 69, 056205 (2004)

work page 2004