Multi-block exceptional points in open quantum systems

This paper investigates the relationship between exceptional points in non-Hermitian Hamiltonians and a novel "multi-block" structure in no-jump Liouvillian superoperators, demonstrating how quantum jump terms modify this structure and affect population dynamics in open quantum systems like qubits and qutrits.

The Gist

Imagine you are watching a complex dance performance. In a perfect, closed world (like a stage with no audience), the dancers move in predictable, symmetrical patterns. But in the real world, the environment is noisy. The dancers get tired, they stumble, and sometimes they are "measured" by the audience, causing them to jump to a new position unexpectedly.

This paper is about understanding the special moments in this chaotic dance where the rules of physics seem to break down, creating a unique kind of "singularity" called an Exceptional Point (EP).

Here is the breakdown of the paper's story, using simple analogies:

1. The Two Ways to Watch the Dance

The authors look at open quantum systems (like atoms interacting with their environment) in two different ways:

• The "No-Jump" View (The NHH): Imagine watching the dancers only when they are gliding smoothly, ignoring the moments they stumble or get bumped by the audience. This is a simplified, "idealized" view where the system loses energy but doesn't jump around randomly.
• The "Full" View (The Liouvillian): This is the real deal. It includes the smooth gliding plus the random "quantum jumps" (stumbles) caused by the environment measuring the system.

The paper asks: If we find a special "perfect moment" (an EP) in the simplified "No-Jump" view, what happens when we add the real-world stumbles back in?

2. The Big Discovery: The "Multi-Block" Surprise

In the simplified "No-Jump" world, the authors found something strange. When the system hits a special point (an EP), it doesn't just merge two dancers into one. Instead, it creates a "Multi-Block" structure.

The Analogy:
Think of a standard EP like two people holding hands and spinning as one unit.
But in this "No-Jump" world, the authors discovered that an EP is actually like a tower of blocks.

• If you have a 2nd-order EP (a simple merge), the "No-Jump" tower looks like a 3-block tower sitting on top of a 1-block tower.
• If you have a 3rd-order EP, the tower becomes a 5-block, a 3-block, and a 1-block stacked together.

They call these Multi-Block Exceptional Points. It's like finding that a single "magic moment" is actually a whole pyramid of magic moments hidden inside each other.

3. What Happens When You Add the "Stumbles"? (Quantum Jumps)

Now, imagine you start the music and the audience starts throwing things at the dancers (the "Quantum Jumps"). What happens to our tower of blocks?

• The Generic Case: Usually, when you shake a tower of blocks, it falls apart completely. The big tower splits into many small, separate pieces. The "magic" is diluted.
• The Special Case: Sometimes, if the "stumbles" happen in a very specific, symmetrical way, the tower doesn't fall apart completely. Instead, the big blocks might merge into slightly smaller, but still connected, groups.

The paper shows that how the tower breaks depends entirely on the type of stumble. Some stumbles destroy the magic completely; others just rearrange it.

4. Why Should We Care? (The Real-World Impact)

Why do we care about these towers of blocks? Because they change how the system behaves over time.

• The Slow-Motion Effect: In a normal system, if you push a pendulum, it swings and stops quickly (exponential decay). But near these "Multi-Block" points, the system behaves like it's moving in slow motion. It doesn't just stop; it decays with a "polynomial" tail (like $t^2$ or $t^3$).

  • Analogy: Imagine a car braking. Usually, it stops in a straight line. Near an EP, the car seems to "drift" for a long time before finally stopping.
  • Use: This could help keep quantum computers (which are very fragile) alive longer by slowing down their decay.

• Super-Sensitive Sensors: These points are incredibly sensitive to tiny changes. If you tweak a parameter slightly, the "tower" of blocks reacts violently. This makes them perfect for building sensors that can detect the tiniest changes in the environment (like a single virus or a tiny magnetic field).

5. The "Geometric Map" (The Quantum Geometric Tensor)

Finally, the authors introduce a tool called the Quantum Geometric Tensor (QGT).

• The Analogy: Imagine you are trying to find a hidden treasure (the EP) on a foggy island. You can't see it directly. But you have a compass (the QGT). As you walk closer to the treasure, the compass needle starts spinning wildly and the "distance" on your map goes to infinity.
• The paper shows that by looking at this "compass," you can pinpoint exactly where these Multi-Block EPs are, even when the system is messy and full of jumps. It tells you exactly which "dancers" (energy levels) are involved in the magic.

Summary

This paper reveals that when we look at open quantum systems, the "idealized" moments of perfection (EPs) are actually complex, multi-layered structures (Multi-Block EPs). When the real world interferes, these structures can either shatter or rearrange, but they leave behind a unique signature: slower decay and super-sensitivity. By understanding these hidden "towers of blocks," we can build better sensors and potentially longer-lasting quantum computers.

Technical Summary

1. Problem Statement

Open quantum systems are typically described by the Lindblad master equation, which governs the dynamics of the reduced density matrix. This dynamics can be decomposed into two parts:

1. Coherent non-unitary evolution: Described by an effective Non-Hermitian Hamiltonian (NHH).
2. Incoherent quantum jumps: Described by jump operators representing the continuous measurement of the system by its environment.

While Exceptional Points (EPs)—spectral singularities where eigenvalues and eigenvectors coalesce—are well-studied in NHHs (Hamiltonian EPs or HEPs) and full Liouvillian superoperators (Liouvillian EPs or LEPs), the precise structural relationship between them remains unclear. Specifically, previous conjectures suggested that an $n$-th order HEP corresponds to a $(2n-1)$-th order LEP. However, this view fails to capture the multi-block structure of the Liouvillian spectrum in the absence of quantum jumps (the "no-jump" Liouvillian). The paper addresses the gap in characterizing how HEPs transform into specific multi-block structures in the no-jump Liouvillian and how these structures are modified (split or merged) when quantum jump terms are reintroduced.

2. Methodology

The authors employ a Hybrid-Liouvillian formalism to bridge the gap between NHHs and full Liouvillians:

• Model Construction: They consider an $N$-level system coupled to an environment where the ground state acts as a population sink. The system is reduced to an effective $(N-1)$-level subsystem (excited states) governed by an effective NHH ($\hat{H}_{\text{eff}}$).
• Vectorization: The density matrix is vectorized to represent the Liouvillian superoperator ($\mathcal{L}$) as a matrix. The full Liouvillian is decomposed into a block-diagonal form, isolating the effective no-jump Liouvillian ($\mathcal{L}'_{\text{eff}}$) which contains only the coherent non-unitary terms.
• Mathematical Derivation:
  • They analyze the spectral properties of $\mathcal{L}'_{\text{eff}} = (-i\hat{H}_{\text{eff}}) \otimes \mathbb{1} + \mathbb{1} \otimes (i\hat{H}^*_{\text{eff}})$.
  • Using the properties of Kronecker sums and Jordan normal forms, they derive the exact correspondence between the Jordan blocks of $\hat{H}_{\text{eff}}$ and $\mathcal{L}'_{\text{eff}}$.
  • They treat quantum jump terms as perturbations to the no-jump Liouvillian, utilizing the Lidskii-Vishik-Lyusternik theorem and Newton diagram techniques to analyze how these perturbations split the multi-block EPs.
• Case Studies: The theory is illustrated using two prime examples:
  1. A Non-Hermitian Qubit (derived from a 3-level system).
  2. A Non-Hermitian Qutrit (derived from a 4-level system).
• Diagnostics: The authors use Population Dynamics (time evolution of state probabilities) and the Quantum Geometric Tensor (QGT) to detect and characterize the EPs.

3. Key Contributions

A. Discovery of Multi-Block EPs

The paper establishes a rigorous general relation between HEPs and no-jump Liouvillian EPs (njLEPs).

• Main Result: An $n$-th order HEP in $\hat{H}_{\text{eff}}$ does not simply become a single $(2n-1)$-th order EP in the Liouvillian. Instead, it manifests as a multi-block EP in $\mathcal{L}'_{\text{eff}}$.
• Block Structure: An $n$-th order HEP results in a decomposition of the Liouvillian spectrum into Jordan blocks of sizes:
  $$(2n-1), (2n-3), \dots, 3, 1$$
  For example, a 2nd-order HEP (qubit) creates blocks of size 3 and 1. A 3rd-order HEP (qutrit) creates blocks of size 5, 3, and 1.
• Correction to Literature: This refines the previous conjecture that an $n$-th order HEP corresponds to a single $(2n-1)$-th order LEP, highlighting that the "order" is distributed across multiple blocks.

B. Perturbation Analysis of Quantum Jumps

The authors analyze how adding quantum jump terms (dissipation within the excited manifold) affects these multi-block structures:

• Generic Perturbations: Typically, a block of size $m$ splits into $m$ branches following a complex root behavior $\propto \Gamma^{1/m}$ (where $\Gamma$ is the perturbation strength), consistent with the Lidskii-Vishik-Lyusternik theorem.
• Symmetry-Protected Perturbations: Specific types of jumps (e.g., combined bit-phase-flip errors) can preserve parts of the degeneracy. In the qubit case, a size-3 block might split only partially into a size-2 block and a size-1 block, rather than fully splitting into three distinct eigenvalues.

C. Dynamical Signatures

The paper links the algebraic structure of EPs to physical observables:

• Population Dynamics: In the presence of an EP of order $n$ (largest Jordan block size $n$), the time evolution of populations acquires polynomial prefactors $t^{n-1}$ multiplying the exponential decay.
• Sensitivity: The order of the polynomial prefactor determines the speed of relaxation to the steady state. Higher-order EPs (larger blocks) lead to slower decay. Perturbing the system (adding jumps) reduces the block sizes, thereby accelerating the decay and changing the polynomial order.

D. Quantum Geometric Tensor (QGT) as a Detector

The authors demonstrate that the QGT, computed for the vectorized eigenmatrices of the Liouvillian, serves as a sensitive indicator:

• The QGT diverges at the locations of EPs.
• By analyzing the QGT for different levels, one can identify exactly which levels participate in the formation of specific EPs and distinguish between different types of EPs (e.g., distinguishing between a multi-block EP at $\Gamma=0$ and the split EPs at $\Gamma>0$).

4. Key Results

• Qubit Example (2nd Order HEP):
  • $\hat{H}_{\text{eff}}$ at EP $\rightarrow$ $\mathcal{L}'_{\text{eff}}$ has blocks $J_3 \oplus J_1$.
  • With generic jumps ($\Gamma$): The $J_3$ block splits into 3 branches $\propto \Gamma^{1/3}$.
  • With symmetric jumps: The $J_3$ block splits into $J_2 \oplus J_1$, creating a Diabolical Point (DP) and a 2nd-order EP at different eigenvalues.
• Qutrit Example (3rd Order HEP):
  • $\hat{H}_{\text{eff}}$ at EP $\rightarrow$ $\mathcal{L}'_{\text{eff}}$ has blocks $J_5 \oplus J_3 \oplus J_1$.
  • With generic jumps: The $J_5$ block splits into 5 branches $\propto \Gamma^{1/5}$.
  • With symmetric jumps: The blocks reorganize into $J_3 \oplus J_2$ and $J_1 \oplus J_2$, preserving some degeneracy.
• Block Merging: The authors theoretically show that specific perturbations could merge these blocks (e.g., merging $J_5$ and $J_3$ into a single $J_8$), which would significantly enhance the system's sensitivity to perturbations ($\propto \Gamma^{1/8}$), though realizing this physically within standard Liouvillian constraints is challenging.

5. Significance

• Theoretical Advancement: The paper provides the first complete characterization of the "no-jump" Liouvillian spectrum, correcting the misconception that HEPs map to single high-order LEPs. It introduces the concept of multi-block EPs as a fundamental feature of open quantum systems.
• Experimental Relevance: The findings explain why experimental signatures of EPs in open systems (like superconducting qubits or optical cavities) might differ from pure NHH predictions. The presence of quantum jumps can partially protect or split EPs in non-trivial ways.
• Applications in Sensing and Computing:
  • Sensing: The multi-block structure implies that the sensitivity of EP-based sensors depends on the specific block sizes. Merging blocks (if achievable) could lead to ultra-high sensitivity scaling as $\Gamma^{1/m}$ with very large $m$.
  • Quantum Control: The polynomial prefactors in population dynamics suggest that EPs can be used to engineer slower decay rates (longer lifetimes) for excited states, which is beneficial for quantum memory and state preparation.
• Diagnostic Tool: The extension of the Quantum Geometric Tensor to Liouvillian superoperators offers a robust, experimentally accessible method to map the complex landscape of EPs in parameter space.

In summary, this work fundamentally redefines the understanding of exceptional points in open quantum systems by revealing their inherent multi-block nature in the no-jump limit and detailing how environmental coupling (quantum jumps) reshapes this landscape, with direct implications for quantum sensing and control.