NPT bound entanglement and the rank-five frontier

Problem

Does there exist a finite-dimensional bipartite state with non-positive partial transpose (NPT) that is undistillable by local operations and classical communication (LOCC)? For a density operator $\rho_{AB}$, partial transposition on $B$ is denoted by $\rho_{AB}^{T_B}:=(\operatorname{id}_A\otimes T_B)(\rho_{AB})$. The finite-copy criterion makes the counterexample sought by the general problem precise:

\begin{equation} \exists\,\rho_{AB}:\quad \rho_{AB}^{T_B}\not\succeq0, \qquad \forall\,n\geq1\ \ \forall\,\lvert\psi_n\rangle \text{ with }\operatorname{SR}_{A^n:B^n}(\lvert\psi_n\rangle)\leq2, \quad \langle\psi_n\rvert (\rho_{AB}^{T_B})^{\otimes n} \lvert\psi_n\rangle\geq0. \tag{1} \end{equation}

Here $\operatorname{SR}_{A^n:B^n}$ is Schmidt rank across the indicated cut. A state satisfying Eq. (1) would be NPT bound entangled; equivalently, the question is whether every NPT state instead has a negative expectation on some Schmidt-rank-at-most-two vector at some finite copy number.

The nested low-rank frontier asks whether the counterexample in Eq. (1) can additionally satisfy

\begin{equation} \operatorname{rank}(\rho_{AB})=5. \tag{2} \end{equation}

Thus Eq. (2) asks for the lowest rank not already excluded by known one-copy results.

Source

Krüger and Werner explicitly list NPT bound entanglement as an open problem, while Chen and Đoković isolate rank five as the next frontier after proving all rank-at-most-four NPT states distillable [KW05], [CD16].

Progress

Horodecki, Horodecki, and Horodecki established the finite-copy Schmidt-rank criterion used in Eq. (1) and proved that states with positive partial transpose are undistillable. Their theorem does not prove the converse for NPT states [HHH98].
Chen and Đoković proved the sharp low-rank obstruction

\begin{equation} \operatorname{rank}(\rho_{AB})\leq4, \quad \rho_{AB}^{T_B}\not\succeq0 \quad\Longrightarrow\quad \rho_{AB}\ \text{is 1-distillable}. \tag{3} \end{equation}

Equation (3) makes Eq. (2) the first possible rank for a one-copy-undistillable NPT state [CD16].
The same work constructed a parameterized family $\rho_\epsilon$ of rank-five two-qutrit NPT states and proved the fixed-copy statement, but not the all-copy statement,

\begin{equation} \underbrace{\forall\,n\geq1\ \exists\,\delta_n>0\ \forall\,\epsilon\in(0,\delta_n]:\quad \rho_\epsilon\ \text{is $n$-undistillable}}_{\text{proved}} \quad\not\Longrightarrow\quad \underbrace{\exists\,\epsilon_*>0\ \forall\,n\geq1:\quad \rho_{\epsilon_*}\ \text{is $n$-undistillable}}_{\text{needed}}. \tag{4} \end{equation}

The authors conjectured the missing right-hand statement in Eq. (4) [CD16].
For a different one-parameter family of symmetric rank-five two-qutrit states, Lei, Song, Chen, and Liu proved 1-undistillability on the remaining NPT interval $[(24\sqrt{2}-33)/7,(33-12\sqrt{6})/25)$. For two copies, they proved that any Schmidt-rank-at-most-two negative-expectation vector must have a component outside a specified $17$-dimensional subspace. This narrows the search but proves neither two-copy nor all-copy undistillability [LSC+26].
A principal unrestricted candidate family is formed by the $d\times d$ Werner states

\begin{equation} \rho_\alpha=\frac{I+\alpha F}{d^2+\alpha d}, \qquad F\lvert x\rangle\lvert y\rangle =\lvert y\rangle\lvert x\rangle, \qquad -\frac12\leq\alpha<-\frac1d,\quad d\geq3. \tag{5} \end{equation}

The parameter window in Eq. (5) is exactly the NPT, one-copy-undistillable window in this convention [DSS+00].
Costa Rico reformulated finite-copy Werner distillability through partial-trace inequalities and obtained new finite-copy bounds [CR25]. Two separate July 2026 preprints subsequently proved that $\rho_\alpha$ is two-copy distillable exactly when $\alpha<-1/2$, in every local dimension. Hence the full candidate window in Eq. (5) is two-copy undistillable, but these results do not control all tensor powers [FHPV26], [BGH26].

Comment

The unrestricted question is posed explicitly in the source collection [KW05]. The rank-five formulation is a nested but stronger route to a counterexample: a rank-five example would settle the general problem, whereas excluding rank five would leave higher ranks open. The decisive quantifier is the existence of one fixed state that is $n$-undistillable for every $n$; separately choosing a state or parameter for each $n$, as on the left of Eq. (4), is insufficient.

References

[HHH98]: M. Horodecki, P. Horodecki, and R. Horodecki, “Mixed-State Entanglement and Distillation: Is There a “Bound” Entanglement in Nature?” Physical Review Letters 80, 5239–5242 (1998).DOI arXiv
[CD16]: L. Chen and D. Ž. Đoković, “Distillability of Non-Positive-Partial-Transpose Bipartite Quantum States of Rank Four,” Physical Review A 94, 052318 (2016).DOI arXiv
[LSC+26]: Y. Lei, Z. Song, L. Chen, and M. Liu, “Entanglement Distillation of Some Rank-Five Symmetric NPT States in Two-Qutrit Systems,” arXiv:2608.03710 (2026).DOI arXiv
[DSS+00]: D. P. DiVincenzo, P. W. Shor, J. A. Smolin, B. M. Terhal, and A. V. Thapliyal, “Evidence for Bound Entangled States with Negative Partial Transpose,” Physical Review A 61, 062312 (2000).DOI arXiv
[CR25]: P. Costa Rico, “New Partial Trace Inequalities and Distillability of Werner States,” Letters in Mathematical Physics 115, 47 (2025).DOI arXiv
[FHPV26]: T. C. Fraser, F. Huber, B. Pozsgay, and I. Vona, “On the Two-Copy Distillability of Werner States and a New Partial Trace Inequality,” arXiv:2607.24309 (2026).DOI arXiv
[BGH26]: K. Bharti, R. Gajjala, and T. Haug, “Two-Copy Nondistillability of Werner States: Sharp Partial-Trace Inequalities and Finite-Copy Extensions,” arXiv:2607.24479 (2026).DOI arXiv
[KW05]: O. Krüger and R. F. Werner (eds.), “Some Open Problems in Quantum Information Theory,” arXiv:quant-ph/0504166 (2005).DOI arXiv

NPT bound entanglement and the rank-five frontier

Problem

Source

Progress

Comment

References

Cite this problem

BibTeX

Plain text

Problem

Source

Progress

Comment

References

Related problems