Extreme points of absolutely PPT states with exactly three distinct eigenvalues

1. Problem Statement

The paper addresses a long-standing open problem in quantum information theory: determining whether the set of Absolutely Separable (AS) states coincides with the set of Absolutely Positive Partial Transpose (AP) states for two-qutrit systems ($3 \times 3$ dimensions).

• Context: While AS and AP sets are known to be convex and compact, their equality depends on whether they share the same set of extreme points.
• Previous Work: It was recently proven that AS and AP coincide for qubit-qudit systems. For two-qutrit systems, the extreme points of AP states with exactly two distinct eigenvalues were fully characterized in a prior study (Song & Chen, 2025).
• Gap: The characterization of extreme points for two-qutrit AP states with exactly three distinct eigenvalues ($a > b > c > 0$) remained incomplete. This paper fills that gap.

2. Methodology

The authors employ a rigorous algebraic and geometric approach to characterize the boundary and extreme points of the set $AP_{3,3}$.

• Mathematical Framework:

  • PPT Criterion: A full-rank two-qutrit state $\rho$ with eigenvalues $\lambda = (\lambda_1, \dots, \lambda_9)$ belongs to $AP_{3,3}$ if and only if two specific $3 \times 3$ real symmetric matrices, $L_1(\lambda)$ and $L_2(\lambda)$, are positive semi-definite (PSD).
  • Boundary Conditions: A state is on the boundary of $AP_{3,3}$ if at least one of the determinants $\det(L_1)$ or $\det(L_2)$ equals zero.
  • Extreme Point Test: Based on Lemma 4, a boundary point is an extreme point if a specific system of linear equations (derived from the null space of the PSD matrices and the unitary equivalence constraints) admits only the trivial solution.

• Analytical Strategy:

  • Eigenvalue Multiplicity Classification: The authors classify states based on the multiplicity of the largest eigenvalue ($\mu(a)$). They investigate cases where $\mu(a) = 1, 2, \dots, 7$.
  • Parametric Construction: For each case, they solve the system of equations defined by the boundary conditions ($l_1(\lambda)=0$ or $l_2(\lambda)=0$) and the normalization constraint ($\sum \lambda_i = 1$).
  • Inequality Simplification: A key technical lemma (Lemma 6) proves that for full-rank states with three distinct eigenvalues, the condition that $L_1, L_2$ are PSD implies the inequality $4\lambda_7\lambda_9 - (\lambda_1 - \lambda_8)^2 \geq 0$. This allows the authors to ignore this inequality in subsequent derivations, simplifying the algebraic complexity.
  • Software Assistance: The authors utilize Mathematica to solve high-degree polynomial equations and determine valid intervals for the parameters.

3. Key Contributions and Results

A. Characterization of Extreme Points

The paper provides an explicit analytical construction of all extreme points of $AP_{3,3}$ with exactly three distinct eigenvalues.

• General Form: The extreme points are denoted as $\nu_{\mu(a), k, 9-\mu(a)-k}$, representing diagonal states with specific multiplicities for eigenvalues $a, b,$ and $c$.
• Parameterization: Most of these extreme points depend on a single continuous parameter (typically the smallest eigenvalue $c$) within a specific interval.
• Limits: As the parameter approaches the boundaries of these intervals, the three-eigenvalue extreme points converge to the known extreme points with two distinct eigenvalues (denoted as $\zeta_1$ through $\zeta_8$ in the paper).

B. The Unique Exception (Non-Extreme Boundary Point)

A major finding is the identification of a specific boundary point that is not an extreme point, despite having three distinct eigenvalues.

• The Point: $\nu_{1,5,3} = \text{diag}(3c, b, b, b, b, b, c, c, c)$ where $3c > b > c > 0$.
• Proof: The authors prove (Lemma 9) that this specific state can be expressed as a convex combination of two known extreme points with two distinct eigenvalues ($\zeta_1$ and $\zeta_6$).
  $$\nu_{1,5,3} = x\zeta_1 + (1-x)\zeta_6, \quad x \in (0,1)$$
• Significance: This is the only exception found where a boundary point of $AP_{3,3}$ with three distinct eigenvalues is not an extreme point.

C. Comprehensive Classification Tables

The authors present detailed tables (Tables I–VII) and figures (Figures 1–7) classifying the extreme points based on the multiplicity of the maximum eigenvalue:

• Multiplicity 1 ($\mu(a)=1$): 6 distinct families of extreme points (e.g., $\nu_{1,1,7}, \nu_{1,2,6}, \dots$).
• Multiplicity 2 ($\mu(a)=2$): 7 distinct families (including split cases where $l_1=0$ or $l_2=0$).
• Multiplicities 3–7: Detailed classifications for $\nu_{3,k,6-k}$ through $\nu_{7,1,1}$.
• Umbrella Model: Figure 8 visualizes the hierarchical structure of these points, showing how the "blue" points (three eigenvalues) connect to the "black" points (two eigenvalues) at the limits, forming an umbrella-like structure.

4. Significance and Implications

1. Advancing the AS vs. AP Conjecture: By fully characterizing the extreme points for the three-eigenvalue case, the paper brings the community closer to resolving whether $AS_{3,3} = AP_{3,3}$. Since the extreme points of $AP_{3,3}$ are now largely mapped, the next step is to determine if these specific points are also separable.
2. Geometric Insight: The "Umbrella Model" provides a new geometric intuition for the structure of the set of absolutely PPT states, illustrating how states with more distinct eigenvalues form continuous manifolds connecting the discrete extreme points of lower-rank spectra.
3. Methodological Framework: The paper establishes a robust methodology for analyzing extreme points in high-dimensional quantum state spaces using linear algebra constraints and spectral analysis. This framework can be extended to systems with more than three distinct eigenvalues or higher dimensions.
4. Open Questions: The authors conclude by posing four critical questions for future research:
   • Is the converse true? (If interior points are separable, are the boundary limits separable?)
   • Can the complex algebraic expressions in the tables be simplified?
   • Can these techniques be extended to states with $>3$ distinct eigenvalues?
   • What is the maximum number of distinct eigenvalues possible for an extreme two-qutrit AP state?

Conclusion

This paper represents a significant step forward in the structural analysis of quantum entanglement. By explicitly constructing the extreme points of full-rank two-qutrit AP states with three distinct eigenvalues and identifying the single non-extreme boundary exception, the authors provide a near-complete map of the boundary of the AP set, laying the groundwork for the final resolution of the absolute separability problem in $3 \times 3$ systems.