Samuel A. Márquez González
4 min
The PPT-squared conjecture asks whether the composition of two positive-partial-transpose (PPT) quantum channels must always result in an entanglement-breaking (EB) channel. While this remains an open problem in general, the author focuses on the class of diagonal orthogonal covariant (DOC) channels. The approach involves isolating the quantum coherence data into correlation matrices, which are then transformed into a triplewise completely positive (TCP) core using a toroidal mixing technique. By absorbing the coherence data into this core, the remaining problem reduces to a purely classical entrywise inequality involving the diagonal-sector matrices of the channels.
The author derives a dimension-dependent sufficient condition for the EB property of composed DOC channels. For PPT DOC maps with diagonal-sector matrices A and D, the composition is guaranteed to be EB if the entrywise bound (AD){ij} >= 2 * floor(sqrt(d)) * sqrt(A{ii} * A_{jj} * D_{ii} * D_{jj}) holds for all i, j. In dimension four, this coefficient becomes four. The paper also constructs a continuous family of PPT channels that are not EB individually, yet their pairwise compositions are proven to be EB, providing an explicit set of nontrivial PPT-squared instances. A rank-sensitive refinement is further introduced to show how lower coherence rank can improve the universal coefficient, allowing the certificate to succeed in cases where the standard bound might fail.
This work offers a new, analytically tractable mechanism for certifying the EB property in composed quantum channels. By separating the classical population mixing from the quantum coherence data, the author provides a robust certificate that is independent of the specific phases of the coherence matrices. This approach complements existing methods like semidefinite hierarchies and factor-width arguments, offering a distinct perspective on how classical mixing can wash out the entanglement-carrying effects of coherent sectors in quantum dynamics.
The positive-partial-transpose-squared (PPT-squared) conjecture asks whether the composition of two positive-partial-transpose (PPT) quantum channels must be entanglement breaking (EB). The problem remains open in general and, within the diagonal orthogonal covariant (DOC) class, the first unresolved deterministic dimension is four. A sufficient condition for PPT-squared composition is derived for DOC maps in arbitrary finite dimension. The proof isolates the quantum coherence data into two correlation matrices, mixes each with the identity until it lies in the convex hull of rank-one correlation matrices, and converts the resulting objects into an explicit triplewise completely positive (TCP) core. The remaining contribution is purely classical and is TCP whenever a simple entrywise mixing inequality is satisfied. For PPT DOC maps $Φ_{A,B,C}$ and $Φ_{D,E,F}$ on $M_d$, the entrywise bound $(AD)_{ij}\ge 2\lfloor\sqrt{d}\rfloor\sqrt{A_{ii}A_{jj}D_{ii}D_{jj}}$ for all $i,j$ guarantees that $Φ_{A,B,C}\circΦ_{D,E,F}$ is EB. In dimension four the universal coefficient becomes four. A continuous family of bistochastic PPT channels is then constructed; every member with $a\ne1$ is not EB, while every pairwise composition within the family is proved to be EB. A rank-sensitive refinement shows explicitly how low coherence rank can improve the universal coefficient. The resulting certificate is sufficient rather than necessary and is complementary to factor-width and semidefinite-hierarchy approaches.
Sam: The paper also tests a continuous family of channels in dimension four, with a parameter a. Does the criterion capture the transition?
Alex: It captures the threshold. At a equal to one, the channel sits on the entanglement-breaking boundary. Moving away from that point, the entrywise inequality in the mixing criterion eventually fails.
Sam: Presumably because the matrix product has to clear the inequality entry by entry. If a drifts, I'd guess some off-diagonal entries dip below the threshold.
Alex: That's a reasonable reading. The inequality acts as a filter. It requires the off-diagonal coherence to be small enough that the classical populations dominate the output.
Sam: I'd be careful about how the paper positions this. It doesn't resolve the PPT-squared conjecture. It sidesteps the hardest part by working in a regime where the answer can be certified directly.
Alex: Agreed. The practical gain is that you replace a hierarchy of semidefinite programs with an entrywise check, and that holds across a continuous family of channels. You give up necessity in exchange for a tractable, deterministic certificate.
Sam: And it suggests that noise-induced decoherence can be used as a structural handle for certification, rather than only managed as a nuisance.
Alex: Within diagonal orthogonal covariant channels, the route to proving entanglement-breaking behavior runs through the classical residual. It's a limited tool, but a usable one.
Sam: If you want the figures and the method choices we skipped, you can generate a deep dive of this paper. The paper has the rest either way.
Alex: Thanks for listening.