---
title: Additional Energy Degeneracy Point
url: https://www.emergentmind.com/topics/additional-energy-degeneracy-point-aedp
type: topic
---

# Additional Energy Degeneracy Point

Searching arXiv for the cited AEDP papers and closely related work to ground the article.
arxiv_search(query="all:2401.05140 OR all:\"Current circulation near additional energy degeneracy points in quadratic Fermionic networks\" OR all:AEDP", max_results=10)
arxiv_search(query="2401.05140", max_results=5)
arxiv_search(query="2509.05621", max_results=5)
arxiv_search(query="2507.17485", max_results=5)
arxiv_search(query="1910.12144", max_results=5)
Additional Energy Degeneracy Point (AEDP) denotes a parameter value or Brillouin-zone point at which energy levels become degenerate in a way that is central to the structure, transport, or topology of the system under study. In the arXiv literature, the term is used in closely related but not identical senses: in open quadratic Fermionic networks, an AEDP is a parameter value at which a pairwise level spacing $\Delta E \to 0$, producing resonantly enhanced off-diagonal correlations and current-circulation effects; in parameter-dependent quantum Hamiltonians, it is a $k$-fold degeneracy point; and in Bloch-band theory it is a point where more bands meet than would be enforced by ordinary space-group symmetries alone [2401.05140] [2507.17485] [1910.12144] [2509.05621].

## 1. Terminology and formal scope

The term AEDP appears in at least three technically precise settings.

| Context | AEDP definition | Principal consequence |
|---|---|---|
| Open quadratic Fermionic transport | Parameter value where $\Delta E \to 0$ for some pair $(l,m)$ | Resonant enhancement and often sign change of current contribution |
| Parameter-dependent quantum Hamiltonians | Point $p_0$ where one eigenvalue has algebraic and geometric multiplicity $k \ge 2$ | Multifold degeneracy; generic perturbations yield Weyl points |
| Bloch-band theory | $\mathbf k_0$ where more bands meet than ordinary space-group symmetries alone enforce | High-degeneracy band crossings protected by site-permutation symmetries |

For a parameter-dependent Hamiltonian
$$
H: M \to \mathrm{Herm}(n), \qquad p \mapsto H(p),
$$
a point $p_0 \in M$ is called an additional energy degeneracy point, or strictly a $k$-fold degeneracy point, if $H(p_0)=A_0$ has one eigenvalue $\lambda_0$ of algebraic and geometric multiplicity $k \ge 2$. Equivalently, at $p_0$ there exists $\lambda_0$ satisfying
$$
\det(H(p_0)-\lambda_0 I_n)=0
$$
with $\lambda_0$ a root of multiplicity $k>1$ of the characteristic polynomial. In this usage, two-fold degeneracies are Weyl points, while $k \ge 3$ are multifold AEDPs [2507.17485].

In Bloch-band language, an AEDP is a point $\mathbf k_0$ in the Brillouin zone at which more bands meet than would be enforced by the ordinary space-group symmetries alone. The site-permutation-symmetry literature treats such crossings as symmetry-protected rather than merely accidental [1910.12144].

In open-system transport, the usage is more operational. There, an AEDP is precisely a parameter value at which $\Delta E \to 0$ for some pair $(l,m)$, and the transport signature is not just spectral coincidence but an anomalous current response in the steady-state correlation matrix [2401.05140].

## 2. AEDPs in open quadratic Fermionic networks

For an $N$-site quadratic Fermionic Hamiltonian,
$$
H_s=\sum_{i\ne j} v_{ij}\,c_i^\dagger c_j + \Omega \sum_n c_n^\dagger c_n,
$$
or in matrix form $H_s=c^\dagger M c$ with $M$ symmetric, coupling site $1$ and site $\ell \equiv N_U+2$ to two baths at temperatures $T_L \ne T_R$ via a dissipative Lindbladian master equation yields a closed equation for the single-particle correlation matrix $C_{ij}=\mathrm{Tr}[c_i^\dagger c_j \rho]$:
$$
\dot C = i[M^T,C] + G - \tfrac12\{G+R,C\}.
$$
In steady state, the problem reduces to the Lyapunov equation
$$
A C + C A^\dagger = G, \qquad A=\tfrac12(G+R)-iM.
$$
If $A^\dagger P = P D$ with $D$ diagonal, and $\tilde C=P^\dagger C P$, $\tilde G=P^\dagger G P$, then
$$
\tilde C_{lm}=\frac{\tilde G_{lm}}{D_l^*+D_m}.
$$
Any local particle current obeys $J_{ij}\propto 2\,\mathrm{Im}\,C_{ij}$, so the imaginary parts of off-diagonal correlations are the relevant transport objects [2401.05140].

In the weak-coupling expansion
$$
A^\dagger=iM+\tfrac12(G+R)=M_0+\kappa M_1, \qquad \kappa \ll 1,
$$
first-order perturbation theory yields, for each pair of levels $(l,m)$, a current contribution of the form
$$
F_{lm}(\Delta E,E_S^{(1)})=
\frac{a_{lm}\,\Delta E-b_{lm}\,E_S^{(1)}}{\Delta E^2+E_S^{(1)2}},
$$
where $\Delta E=E_m-E_l$ is the bare-Hamiltonian level spacing, $E_S^{(1)}=\kappa(E_l^{(1)}+E_m^{(1)})$ is the sum of first-order Lamb-shifts, and $a_{lm},b_{lm}$ are overlap-factors coming from $G$ and $P$. An AEDP is precisely a parameter value at which $\Delta E \to 0$ for some pair $(l,m)$. As $\Delta E \to 0$, one gets a resonant enhancement, and often sign change, of $F_{lm}$. In multi-branched geometries, this can invert one branch enough to force circulation [2401.05140].

A common misconception is that degeneracy alone is sufficient. The transport analysis instead identifies two conditions: an AEDP in the one-particle spectrum and a geometric or coupling asymmetry that makes the overlap factors $a_{lm},b_{lm}$ non-zero. Without the latter, the degeneracy need not produce circulating currents.

## 3. SSH and unequal-hopping tight-binding realizations

The Su-Schrieffer-Heeger model with periodic boundary conditions is defined by
$$
H_s=t\sum_{n=1}^N [1+(-1)^n\delta]\,(c_{n+1}^\dagger c_n+\mathrm{h.c.})+\Omega\sum_n c_n^\dagger c_n,
$$
with $N=N_U+N_D+2$ and baths at $n=1$ and $n=N_U+2$. Its spectrum is
$$
\lambda_k^\pm = \pm \sqrt{2}\,t\,\sqrt{1+\delta^2+(1-\delta^2)\cos(4\pi k/N)}, \qquad k=1,\dots,N/2.
$$
At $\delta=0$ many $\pm$ pairs coincide, giving an AEDP at $\delta=0$. If $N_U=N_D$, the geometry is left-right symmetric and, although $\delta=0$ is an AEDP, all branch-currents remain equal and there is no current circulation. If $N_U\ne N_D$, one still has the $\delta=0$ AEDP, but the overlap factors need not cancel; then a narrow window $|\delta|\lesssim \delta_C$ around zero sees one branch current exceed the total and current circulation develops. Near $\delta=0$, $\Delta E\propto \delta$, so the $F$-function peaks like a Lorentzian of width $\sim \kappa E^{(1)}$. Beyond $\delta=0$ the levels reorder and current circulation reverses sign, clockwise to anticlockwise or vice versa [2401.05140].

The unequal-hopping tight-binding model is
$$
H_s=t\sum_{n=1}^N v_n\,(c_{n+1}^\dagger c_n+\mathrm{h.c.})+\Omega\sum_n c_n^\dagger c_n,
$$
with
$$
v_n=
\begin{cases}
1-\Delta & \text{for } n \text{ on the upper branch},\\
1+\Delta & \text{for } n \text{ on the lower branch}.
\end{cases}
$$
At $\Delta=0$ the two branches are identical, so many $E_n^{\mathrm{upper}}=E_m^{\mathrm{lower}}$, giving an AEDP at $\Delta=0$. More generally, level crossings occur whenever
$$
(1-\Delta)\cos\!\left(\frac{2\pi n}{N_U+1}\right)
=
(1+\Delta)\cos\!\left(\frac{2\pi m}{N_D+1}\right)
$$
for some $n,m$; each solution $\Delta=\Delta^*$ is an AEDP. Feeding these energies into the perturbative current formula gives, for small $|\Delta|$, a resonant current-circulation peak of width $\sim \kappa$ around each $\Delta^*$ [2401.05140].

The model comparison is sharp. In the SSH geometry, having unequal number of Fermionic sites in the upper and lower branches is enough for generating current circulation. In the unequal-hopping tight-binding model, this asymmetry is not adequate; unequal hopping strengths in the upper and lower branches are required to induce current circulation. This distinction is one of the clearest demonstrations that AEDP-triggered transport is model-dependent even when the spectral mechanism is similar.

## 4. Exact transport, LLME, and distinct onset of particle and heat circulation

The quadratic-network analysis compares local Lindbladian master equation (LLME) results with exact results obtained via the Non-Equilibrium Green Function (NEGF) formalism. Within the flat-band, weak-coupling regime, $t\lesssim \Omega$ and $\kappa \ll t$, LLME branch-currents $J_U(\delta)$ and $J_D(\delta)$, including their sign reversals at the AEDP, agree almost quantitatively with the NEGF results. For stronger inter-site hopping, $t\sim \Omega$, or non-flat bath spectra, LLME begins to deviate in magnitude and in the precise lineshape, although the AEDP locations and current-circulation reversal still coincide. The discrepancies are attributed to the local-dissipator and secular approximations inherent in LLME, neglect of Lamb-shifts beyond first order, and differences in microscopic bath models [2401.05140].

The same framework distinguishes particle from heat circulation. Particle current involves only nearest-neighbor correlations $\mathrm{Im}\,C_{i,i+1}$, whereas heat current also samples next-nearest terms $\mathrm{Im}\,C_{i,i+2}$. In odd-$N$ SSH chains, many AEDPs occur at $\delta=0$, but only two of those level-pairs strongly affect the nearest-neighbor block; consequently particle current circulation is tightly localized around $\delta=0$. By contrast, more AEDPs contribute to the next-nearest correlations, so heat current circulation extends over a broader $|\delta|$ window [2401.05140].

This establishes that the onset point of particle and heat current circulation need not be the same. Parameter ranges can therefore exist in which heat circulates but particle currents remain parallel, or vice versa. The distinction reflects how energy-spectrum degeneracies feed differently into the two current operators, rather than any failure of the AEDP mechanism itself.

## 5. Modified quantum Wheatstone bridge and metrological use

A metrological realization of the AEDP mechanism is provided by a four-site fermionic ring coupled at site $1$ to a left bath and at site $4$ to a right bath. The system Hamiltonian is
$$
H_0
=
\mu\sum_{n=1}^4 c_n^\dagger c_n
+\sum_{n=1}^4 J_n(c_n^\dagger c_{n+1}+c_{n+1}^\dagger c_n),
$$
with periodic boundary $c_5\equiv c_1$, fixed $J_1$ and $J_4$, unknown $J_2=\lambda$, and tunable $J_3$. In this model one defines
$$
E_0^2=J_1^2+\lambda^2+J_3^2+J_4^2, \qquad \xi=J_1J_3-\lambda J_4.
$$
Two levels become degenerate precisely when
$$
\xi=0
\quad\Longrightarrow\quad
J_1J_3=\lambda J_4
\quad\Longleftrightarrow\quad
\frac{J_1}{J_4}=\frac{\lambda}{J_3}.
$$
This is the balanced Wheatstone-bridge condition, and the associated degeneracy is the AEDP. Equivalently,
$$
J_3^{\rm AEDP}=\frac{J_4}{J_1}\lambda,
\qquad
\lambda^{\rm AEDP}=\frac{J_1}{J_4}J_3.
$$
The detectable signature is a circulation-current reversal as $J_3$ is tuned through the AEDP [2509.05621].

The physical origin of the circulation is geometric asymmetry: the upper branch $1\to2\to3\to4$ has three links, whereas the lower branch $1\to4$ has one link. Near the avoided crossing, this asymmetry induces a circulating current inside the loop even with zero magnetic field. At $\xi<0$ the local currents in the upper and lower branches flow in one sense, whereas for $\xi>0$ they reverse; exactly at $\xi=0$ the circulation current vanishes [2509.05621].

In the NEGF formalism,
$$
I_U=\int dE\,T_U(E)\bigl[f_L(E)-f_R(E)\bigr],\qquad
I_D=\int dE\,T_D(E)\bigl[f_L(E)-f_R(E)\bigr],
$$
with
$$
T_U(E)=2\gamma J_1\,\mathrm{Im}[G_{11}^R G_{12}^A],\qquad
T_D(E)=2\gamma J_4\,\mathrm{Im}[G_{11}^R G_{14}^A].
$$
In the low-temperature, low-bias limit one finds
$$
G_U=\frac{16\,\gamma^2\,\xi\,\lambda^2\,J_1J_3}{(4\xi^2+\gamma^2\lambda^2)^2},
\qquad
G_D=-\frac{16\,\gamma^2\,\xi\,\lambda^3\,J_4}{(4\xi^2+\gamma^2\lambda^2)^2},
$$
so that a circulation current may be defined by
$$
J_{\rm circ}\equiv I_U-I_D\propto \xi,
$$
with positive, zero, and negative values for $\xi>0$, $\xi=0$, and $\xi<0$, respectively. Under moderately strong dephasing and particle losses the device remains functional, but extreme environmental effects eventually degrade performance. The efficiency
$$
\eta=1-\Delta J_3/J_3^0
$$
remains $\gtrsim 80\%$ for $\gamma_P,\Gamma_0\lesssim \gamma$ and falls to zero once environmental rates become too large [2509.05621].

## 6. Multifold AEDPs, determinantal varieties, and Weyl-point counting

In singularity-theoretic treatment, the geometric degeneracy variety is
$$
\Sigma=\{A\in\mathbb C^{n\times n}\mid \exists\,\lambda\in\mathbb C\text{ with }\mathrm{rank}(A-\lambda I_n)\le n-2\}.
$$
Equivalently, $A\in\Sigma$ if and only if $A$ has some eigenvalue $\lambda$ of geometric multiplicity at least $2$. One also introduces the lifted variety
$$
\widetilde\Sigma=\{(A,\lambda)\in\mathbb C^{n\times n}\times\mathbb C\mid \mathrm{rank}(A-\lambda I_n)\le n-2\},
$$
defined as the zero-locus of all $(n-1)\times(n-1)$ minors $M_{ij}(A-\lambda I)$. The projection $(A,\lambda)\mapsto A$ exhibits $\Sigma$ as the projection of a determinantal variety of codimension $4$ in $\mathbb C^{n\times n}\times\mathbb C$, while $\Sigma$ itself has codimension $3$ in $\mathbb C^{n\times n}$. Locally at a strictly $k$-fold degenerate $A_0$, with all other eigenvalues simple, the analytic germ $(\Sigma;A_0)$ splits into branches $\Sigma_{\lambda_0}$ parametrized by each multiple eigenvalue $\lambda_0$, each branch irreducible of dimension $n^2-3$ [2507.17485].

Let
$$
f:(\mathbb C^3,0)\to(\mathbb C^{n\times n},A_0)
$$
be the holomorphic germ of $H$ in local coordinates centered at the degeneracy, and let $\lambda_0$ be the $k$-fold eigenvalue of $A_0$. Form the pull-back ideal
$$
J=\langle M_{ij}(f(x,y,z)-(\lambda+\lambda_0)I_n):1\le i,j\le n\rangle\subset \mathcal O_4.
$$
If $f$ is isolated with respect to $(\Sigma;A_0;\lambda_0)$ and is perturbed to a transverse holomorphic $f_t$, then the number of complex Weyl points emerging from the singularity is
$$
\#\,\mathrm{cWP}=\dim_{\mathbb C}(\mathcal O_4/J).
$$
Under genericity, or Cohen-Macaulay, assumptions this count is invariant under perturbations [2507.17485].

In the special case of a strictly $k$-fold eigenvalue with no other degeneracies at the same point, the local analytic set-germ is analytically equivalent to the corank $\ge 2$ determinantal germ in $\mathbb C^{k\times k}$ at $0$, and
$$
\mathrm{mult}(\Sigma;A_0;\lambda_0)=\frac{k^2(k^2-1)}{12}.
$$
Hence the universal upper bound is
$$
\#\,\mathrm{cWP}=\frac{k^2(k^2-1)}{12},
\qquad
\#\,\mathrm{WP}\le \#\,\mathrm{cWP}.
$$
For the spin-1 Hamiltonian with $k=3$, one finds $\dim_{\mathbb C}(\mathcal O_4/J)=6=3^2(3^2-1)/12$; for a generic four-fold crossing with $k=4$, numerical algebraic elimination gives $20=4^2(4^2-1)/12$ [2507.17485].

This framework resolves a frequent ambiguity in the interpretation of multifold crossings. A multifold AEDP is not, under generic perturbation, expected to remain a single multifold object; rather, it is typically dissolved into a set of Weyl points, and the determinantal multiplicity gives a rigorous upper bound on their number.

## 7. Site-permutation symmetry and engineered high-degeneracy band crossings

In spinless tight-binding Bloch bands, high-degeneracy AEDPs can be protected by site-permutation symmetries. For a unit cell with $N$ basis sites and one orbital per site, the Bloch Hamiltonian has matrix elements
$$
H_{ij}(\mathbf k)=\sum_{\mathbf R} t_{ij}(\mathbf R)\,e^{i\mathbf k\cdot(\mathbf R+\mathbf r_j-\mathbf r_i)}.
$$
At the special point $\Gamma\equiv \mathbf k=0$, if the lattice has $N$ equivalent sites, then
$$
H_N=t\,n_1\,\mathbb J_N+(m-t\,n_1)\,\mathbb I_N,
$$
with eigenvalues
$$
\lambda_{\rm HD}=m-t\,n_1 \quad (\text{degeneracy }N-1),
\qquad
\lambda_{\rm sing}=m+(N-1)t\,n_1 \quad (\text{nondegenerate}).
$$
Thus an $(N-1)$-fold degeneracy is guaranteed at $\Gamma$ by uniform connectivity of the $N$ sites [1910.12144].

Let $P^{(ij)}$ exchange site labels $i\leftrightarrow j$ within the unit cell. These permutations generate the symmetric group $S_N$, and for the uniform-connectivity Hamiltonian one has
$$
P^{(ij)}H_NP^{(ij)}=H_N
\quad\Longrightarrow\quad
[P^{(ij)},H_N]=0.
$$
Because there are $N-1$ independent generators of site permutations, the spectrum splits into a singlet plus one $(N-1)$-dimensional irreducible subspace of $S_N$ [1910.12144].

Explicit realizations illustrate the construction. For a three-site cell,
$$
H^{(3)}(0)=2t(\mathbb J_3-\mathbb I_3),
$$
with secular equation
$$
(E+2t)^2(E-4t)=0,
$$
giving a twofold degeneracy at $E=-2t$. For a four-site cell,
$$
H^{(4)}(0)=t(\mathbb J_4-\mathbb I_4),
$$
with
$$
(E+3t)^3(E-3t)=0,
$$
giving a triply degenerate level at $E=-3t$. For a five-site cell,
$$
H^{(5)}(0)=t(\mathbb J_5-\mathbb I_5),
$$
with
$$
(E+4t)^4(E-4t)=0,
$$
giving a fourfold flat band at $E=-4t$ plus one dispersive singlet [1910.12144].

The low-energy theory around such AEDPs is Dirac-like. For the threefold case, projection onto the degenerate subspace yields a pseudospin-1 Hamiltonian,
$$
H_{\rm eff}(\mathbf q)=v(q_xS_x+q_yS_y),
$$
with dispersions $(-v|\mathbf q|,0,+v|\mathbf q|)$. For the fivefold case, the effective form is pseudospin-2,
$$
H_{\rm eff}(\mathbf q)=v(q_xS_x^{(2)}+q_yS_y^{(2)}),
$$
with branches $E_m(\mathbf q)=v\,m\,|\mathbf q|$ for $m=-2,-1,0,+1,+2$ [1910.12144].

This symmetry-based construction addresses another recurrent misconception: high-degeneracy points in spinless systems are not necessarily accidental. The site-permutation framework shows that threefold, fourfold, and fivefold degeneracy points can be symmetry-protected even when ordinary space-group arguments alone would not require them.

Source: https://www.emergentmind.com/topics/additional-energy-degeneracy-point-aedp