Dissipation driven phase transition in the non-Hermitian Kondo model
Pith reviewed 2026-05-24 04:01 UTC · model grok-4.3
The pith
The non-Hermitian Kondo model exhibits a novel ~YSR phase between Kondo and unscreened phases due to dissipation.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
The non-Hermitian Kondo model is characterized by two renormalization group invariants, a generalized Kondo temperature T_K and a loss strength parameter alpha. It exhibits the Kondo phase for 0 < alpha < pi/2, the ~YSR phase for pi/2 < alpha < pi where the Kondo cloud shrinks to a bound state screening the impurity, an intermediate regime for pi < alpha < 3pi/2 where the impurity is unscreened in the ground state but screened by the bound state, and the unscreened phase for alpha > 3pi/2. The transition at alpha = pi/2 is driven by dissipation with associated different time scales.
What carries the argument
Bethe Ansatz solution that maps the model to two RG invariants T_K and alpha, classifying ground states and phase boundaries.
Load-bearing premise
The inclusion of the loss term allows the Bethe Ansatz to remain exactly solvable and correctly classify the ground states of the non-Hermitian Kondo Hamiltonian.
What would settle it
An experiment measuring the screening of the impurity or the presence of the bound state at intermediate values of the loss parameter alpha in a non-Hermitian setup would confirm or refute the phase boundaries.
Figures
read the original abstract
Non-Hermitian Hamiltonians capture several aspects of open quantum systems, such as dissipation of energy and non-unitary evolution. An example is an optical lattice where the inelastic scattering between the two orbital mobile atoms in their ground state and the atom in a metastable excited state trapped at a particular site and acting as an impurity, results in the two body losses. It was shown in \cite{nakagawa2018non} that this effect is captured by the non-Hermitian Kondo model. which was shown to exhibit two phases depending on the strength of losses. When the losses are weak, the system exhibits the Kondo phase and when the losses are stronger, the system was shown to exhibit the unscreened phase where the Kondo effect ceases to exist, and the impurity is left unscreened. We re-examined this model using the Bethe Ansatz and found that in addition to the above two phases, the system exhibits a novel $\widetilde{YSR}$ phase which is present between the Kondo and the unscreened phases. The model is characterized by two renormalization group invariants, a generalized Kondo temperature $T_K$ and a parameter `$\alpha$' that measures the strength of the loss. The Kondo phase occurs when the losses are weak which corresponds to $0<\alpha<\pi/2$. As $\alpha$ approaches $\pi/2$, the Kondo cloud shrinks resulting in the formation of a single particle bound state which screens the impurity in the ground state between $\pi/2<\alpha<\pi$. As $\alpha$ increases, the impurity is unscreened in the ground state but can be screened by the localized bound state for $\pi<\alpha<3\pi/2$. When $\alpha>3\pi/2$, one enters the unscreened phase where the impurity cannot be screened. We argue that in addition to the energetics, the system displays different time scales associated with the losses across $\alpha=\pi/2$, resulting in a phase transition driven by the dissipation in the system.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript re-examines the non-Hermitian Kondo model (with two-body loss term) via the Bethe Ansatz. It claims that, in addition to the previously reported Kondo and unscreened phases, an intermediate ~YSR phase exists; the model is characterized by two RG invariants (generalized T_K and loss parameter α), with dissipation-driven transitions at α = π/2, π, 3π/2 that separate regimes of screened, partially screened, and unscreened impurity ground states.
Significance. If the Bethe-Ansatz mapping and phase classification hold, the work would supply an exactly solvable example of a dissipation-driven transition in an open Kondo system, together with two RG invariants that organize the non-Hermitian phases; this would be a concrete advance for non-Hermitian extensions of strongly correlated models.
major comments (2)
- [Abstract] Abstract: the central claim that the Bethe Ansatz yields the phase boundaries at α = π/2, π, 3π/2 (and the existence of the ~YSR phase) is stated without any explicit Bethe equations, bound-state conditions, or wave-function construction; this derivation is load-bearing for the reported phase diagram.
- [Abstract] Abstract and introduction: the mapping of the non-Hermitian loss term onto the two RG invariants (generalized T_K and α) presupposes that the algebraic structure of the Hermitian Bethe equations survives the addition of the imaginary loss; no discussion is given of how the two-particle S-matrix or the rapidity equations are altered.
minor comments (1)
- The symbol ~YSR is introduced without a clear definition of what physical feature the tilde denotes; a brief parenthetical explanation would improve readability.
Simulated Author's Rebuttal
We thank the referee for the careful reading and constructive comments. We address the two major points below and revise the manuscript to improve clarity on the Bethe-Ansatz derivation.
read point-by-point responses
-
Referee: [Abstract] Abstract: the central claim that the Bethe Ansatz yields the phase boundaries at α = π/2, π, 3π/2 (and the existence of the ~YSR phase) is stated without any explicit Bethe equations, bound-state conditions, or wave-function construction; this derivation is load-bearing for the reported phase diagram.
Authors: The explicit Bethe equations (modified by the imaginary loss), bound-state conditions that fix the critical values α = π/2, π, 3π/2, and the wave-function construction for the ~YSR phase appear in Sections III and IV. The abstract is concise by design, but we agree a short reference to these results will strengthen it. We will add one sentence to the abstract summarizing the key rapidity equations and phase boundaries. revision: yes
-
Referee: [Abstract] Abstract and introduction: the mapping of the non-Hermitian loss term onto the two RG invariants (generalized T_K and α) presupposes that the algebraic structure of the Hermitian Bethe equations survives the addition of the imaginary loss; no discussion is given of how the two-particle S-matrix or the rapidity equations are altered.
Authors: Section II derives the two-particle S-matrix with the imaginary phase shift induced by the loss term; this modification preserves the nested algebraic structure while rendering the rapidities complex, from which the two RG invariants follow directly. We will expand the introduction with an explicit paragraph stating the form of the altered S-matrix and confirming that the Bethe-ansatz hierarchy remains intact. revision: yes
Circularity Check
Bethe-Ansatz solution yields independent phase boundaries; no reduction to inputs
full rationale
The derivation maps the non-Hermitian Kondo Hamiltonian onto two RG invariants (generalized T_K and loss parameter alpha) via the Bethe-Ansatz equations and then classifies ground states by bound-state conditions at the reported thresholds. Alpha enters as an external dissipation strength rather than a fitted output; the thresholds pi/2, pi, 3pi/2 are stated to emerge from the algebraic structure of the Bethe equations once the imaginary loss term is included. No self-definitional loop, no fitted quantity renamed as prediction, and the sole external citation (nakagawa2018non) is to the Hermitian starting model, not to the phase classification itself. The central claim therefore remains independent of its own outputs.
Axiom & Free-Parameter Ledger
free parameters (2)
- alpha
- T_K
axioms (1)
- domain assumption The non-Hermitian Kondo Hamiltonian remains integrable and solvable by Bethe Ansatz after inclusion of the loss term.
invented entities (1)
-
~YSR phase
no independent evidence
Lean theorems connected to this paper
-
IndisputableMonolith/Foundation/RealityFromDistinction.leanreality_from_one_distinction unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
The model is characterized by two renormalization group invariants, a generalized Kondo temperature TK and a parameter α that measures the strength of the loss. The Kondo phase occurs when ... 0<α<π/2 ... π/2<α<π ... π<α<3π/2 ... α>3π/2
-
IndisputableMonolith/Cost/FunctionalEquation.leanwashburn_uniqueness_aczel unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
Like its Hermitian counterpart, the Hamiltonian Eq.(1) is integrable [1] with its Bethe Ansatz equations being the analytical continuation of those of the Hermitian case
What do these tags mean?
- matches
- The paper's claim is directly supported by a theorem in the formal canon.
- supports
- The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
- extends
- The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
- uses
- The paper appears to rely on the theorem as machinery.
- contradicts
- The paper's claim conflicts with a theorem or certificate in the canon.
- unclear
- Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.
Forward citations
Cited by 1 Pith paper
-
Arrow of Time as an indicator of Measurement-Induced Phase Transitions
The arrow of time exhibits nonanalytic behavior at the critical point of measurement-induced phase transitions, with an identified critical exponent, in an exactly solved model of random quantum circuits with non-proj...
Reference graph
Works this paper leans on
-
[1]
M. Nakagawa, N. Kawakami, and M. Ueda, Non- hermitian kondo effect in ultracold alkaline-earth atoms, Physical Review Letters 121, 203001 (2018)
work page 2018
- [2]
-
[3]
N. Goldman, J. C. Budich, and P. Zoller, Topological quantum matter with ultracold gases in optical lattices, Nature Physics 12, 639 (2016)
work page 2016
-
[4]
N. Goldman, G. Juzeli¯ unas, P. ¨Ohberg, and I. B. Spiel- man, Light-induced gauge fields for ultracold atoms, Re- ports on Progress in Physics 77, 126401 (2014)
work page 2014
-
[5]
M. Lewenstein, A. Kubasiak, J. Larson, C. Menotti, G. Morigi, K. Osterloh, and A. Sanpera, Travelling to ex- otic places with ultracold atoms, in AIP Conference Pro- ceedings, Vol. 869 (American Institute of Physics, 2006) pp. 201–211
work page 2006
-
[6]
N. E. Myerson-Jain, S. Yan, D. Weld, and C. Xu, Con- struction of fractal order and phase transition with ryd- berg atoms, Physical Review Letters 128, 017601 (2022)
work page 2022
-
[7]
D. I. Tsomokos, S. Ashhab, and F. Nori, Using supercon- ducting qubit circuits to engineer exotic lattice systems, Physical Review A 82, 052311 (2010)
work page 2010
- [8]
-
[9]
G. Lindblad, On the generators of quantum dynamical semigroups, Communications in mathematical physics 48, 119 (1976)
work page 1976
- [10]
- [11]
- [12]
-
[13]
D. Zou, T. Chen, W. He, J. Bao, C. H. Lee, H. Sun, and X. Zhang, Observation of hybrid higher-order skin- topological effect in non-hermitian topolectrical circuits, Nature Communications 12, 7201 (2021)
work page 2021
- [14]
-
[15]
W.-C. Wang, Y.-L. Zhou, H.-L. Zhang, J. Zhang, M.-C. Zhang, Y. Xie, C.-W. Wu, T. Chen, B.-Q. Ou, W. Wu, et al. , Observation of pt-symmetric quantum coherence in a single-ion system, Physical Review A 103, L020201 (2021)
work page 2021
-
[16]
M. Sakhdari, M. Hajizadegan, Q. Zhong, D. Christodoulides, R. El-Ganainy, and P.-Y. Chen, Experimental observation of p t symmetry breaking near divergent exceptional points, Physical review letters 123, 193901 (2019)
work page 2019
-
[17]
J. Schnabel, H. Cartarius, J. Main, G. Wunner, and W. D. Heiss, Pt-symmetric waveguide system with evi- dence of a third-order exceptional point, Physical Review A 95, 053868 (2017)
work page 2017
-
[18]
C. Shi, M. Dubois, Y. Chen, L. Cheng, H. Ramezani, Y. Wang, and X. Zhang, Accessing the exceptional points of parity-time symmetric acoustics, Nature communica- tions 7, 11110 (2016)
work page 2016
-
[19]
L. Riegger, N. D. Oppong, M. H¨ ofer, D. R. Fernandes, I. Bloch, and S. F¨ olling, Localized magnetic moments with tunable spin exchange in a gas of ultracold fermions, Physical review letters 120, 143601 (2018)
work page 2018
-
[20]
M. Kan´ asz-Nagy, Y. Ashida, T. Shi, C. P. Moca, T. N. Ikeda, S. F¨ olling, J. I. Cirac, G. Zar´ and, and E. A. Dem- ler, Exploring the anisotropic kondo model in and out of equilibrium with alkaline-earth atoms, Physical Review B 97, 155156 (2018)
work page 2018
-
[21]
D. H. White, T. A. Haase, D. J. Brown, M. D. Hooger- land, M. S. Najafabadi, J. L. Helm, C. Gies, D. Schu- mayer, and D. A. Hutchinson, Observation of two- dimensional anderson localisation of ultracold atoms, Na- ture communications 11, 4942 (2020)
work page 2020
-
[22]
The YSR phase was first discovered in a BCS supercon- ductor by Yu, Shiba and Rusinov who pointed out the existence of these bound modes
- [23]
-
[24]
P. R. Pasnoori, N. Andrei, C. Rylands, and P. Azaria, Rise and fall of yu-shiba-rusinov bound states in charge-conserving s-wave one-dimensional superconduc- tors, Phys. Rev. B 105, 174517 (2022)
work page 2022
- [25]
-
[26]
N. Andrei, Integrable models in condensed matter physics, Low-Dimensional Quantum Field Theories for Condensed Matter Physicists (World Scientific Publish- ing Co, 2013) pp , 457 (1992)
work page 2013
- [27]
-
[28]
A. Tsvelick and P. Wiegmann, Exact results in the theory of magnetic alloys, Advances in Physics 32, 453 (1983)
work page 1983
-
[29]
Our Bethe equations are related from those of [1] by a change of variables Λ γ → 1 + Λγc
-
[30]
Such a complex solution of Bethe Ansatz describing a dissipative mode was first found in [40]
-
[31]
Andrei, Diagonalization of the kondo hamiltonian, Physical Review Letters 45, 379 (1980)
N. Andrei, Diagonalization of the kondo hamiltonian, Physical Review Letters 45, 379 (1980)
work page 1980
- [32]
-
[33]
The other sectors re- quire adding a hole, which may be sent to −∞ without changing the results
We consider the |U ⟩ and |B⟩ states only in even N e and odd N e parity sector respectively. The other sectors re- quire adding a hole, which may be sent to −∞ without changing the results
- [34]
-
[35]
Y. LUH, Bound state in superconductors with paramag- netic impurities, Acta Physica Sinica 21, 75 (1965)
work page 1965
-
[36]
Shiba, Classical Spins in Superconductors, Progress of Theoretical Physics 40, 435 (1968)
H. Shiba, Classical Spins in Superconductors, Progress of Theoretical Physics 40, 435 (1968)
work page 1968
-
[37]
A. I. Rusinov, Superconductivity near a Paramagnetic Impurity, Soviet Journal of Experimental and Theoreti- cal Physics Letters 9, 85 (1969)
work page 1969
-
[38]
P. R. Pasnoori, N. Andrei, C. Rylands, and P. Azaria, Rise and fall of yu-shiba-rusinov bound states in charge-conserving s-wave one-dimensional superconduc- tors, Physical Review B 105, 174517 (2022)
work page 2022
-
[39]
Since the |B⟩ and |U ⟩ states do not have the same fermionic parity, one needs to add a spinon in bulk to (say) the |B⟩ state
-
[40]
P. Kattel, P. R. Pasnoori, and N. Andrei, Exact solu- tion of a non-hermitian-symmetric spin chain, Journal of Physics A: Mathematical and Theoretical 56, 325001 (2023). 7 In these appendixes , we provide additional details regarding the solution of Bethe Ansatz equations in all three phases. Moreover, in the Kondo phase, we do a more careful analysis of ...
work page 2023
-
[41]
The Kondo Phase When sin(ϕ) < c 2, we can solve Eq.(A3) in the thermodynamic limit by computing the density of the roots defined as σ(Λ) = 1 Λγ+1−Λγ describing the number of solution in the interval (Λ , Λ + dΛ) of the solution rather than the solution Λγ themselves. In terms of the density, Eq.(A3) can be written as N eΘ (2Λγ − 2) + Θ 2(Λγ − 1 + e−iϕ = Z...
-
[42]
The excitation energy is given by ∆Et = D Z ∆σ(Λ) [Θ(2Λ − 2) − π] dΛ = 2D tan−1 e(π/c)(Λh 1 −1) + tan−1 e(π/c)(Λh 2 −1) , (A17) which shows that this is a sum of two terms carrying spin- 1 2, which gives a total spin-one state. To get a spin one-half state, we need to add a hole with an electron, which gives a total energy ∆Ed = 2D tan−1 e(π/c)(Λh 1 −1) +...
-
[43]
and in the thermodynamic limit, the energy of the singlet is equal to the energy of the triplet excitations. Thus, starting from the ground state, all the excited states are constructed by exciting the charge degree of freedom, adding an even number of spinons, or adding string solutions with an appropriate number of spinons. d. Analytical expression for ...
-
[44]
Bound mode phase When c 2 < sin(ϕ) < 3c 2 , there is a new solution of the Bethe equation in the thermodynamic limit of the form ΛIS = 1 − cos(ϕ) + i 2(2 sin(ϕ) − c). (A39) In this regime, the solution of Eq.(A6), which gives the distribution of the continuous root distribution, can be written in the Fourier space as ˜σ0(p) = 1 2 e−ipN esech cp 2 + (ecp −...
-
[45]
However, as shown in Eq.(A72), the energy of the impurity string solution vanishes
Unscreened phase Notice that the impurity string solution still exists in this region. However, as shown in Eq.(A72), the energy of the impurity string solution vanishes. The energy of the string solution when 3 c < 2 sin(ϕ) is calculated as ∆Eist = D Z ∆σist(Λ)[Θ(2Λ − 2) − π]dΛ + D (Θ (2Λstp − 2) − π) = 2iD tanh−1 c − 2ie−iϕ c − 2iD tanh−1 c c − 2ie−iϕ −...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.