A partial-trace matrix inequality and Werner-state distillability

In quantum information theory, the distillability problem asks whether all entangled states with a non-positive partial transpose (NPT) can be converted into pure entanglement using local operations and classical communication (LOCC). While it is known that some states require multiple copies to be distilled, the specific distillability of Werner states—a fundamental class of mixed states—has remained an open question for decades. This paper addresses the two-copy distillability of NPT Werner states in arbitrary local dimensions.

The authors utilize a recent theoretical reduction that maps the distillability of Werner states to a specific multipartite matrix inequality involving partial traces. By focusing on matrices with a rank of at most two, they derive a sharp partial-trace inequality. The proof relies on a sophisticated projection decomposition on two copies of the underlying bipartite Hilbert space, allowing them to bound the quadratic forms associated with the partial trace operations.

The primary contribution is the proof of a rank-two partial-trace inequality, which establishes that NPT Werner states are two-copy undistillable for the parameter range α ≥ -1/2. This result resolves a significant open problem in the field. Additionally, the authors demonstrate that two individually one-copy-undistillable NPT Werner states cannot activate each other's distillability, providing a negative result for entanglement activation in this context. They also resolve a long-standing singular-value maximization problem related to 4x4 Werner states.

Understanding the limits of entanglement distillation is crucial for quantum communication and error correction. By proving that certain NPT states cannot be distilled even when using two copies, this work clarifies the boundaries of available quantum resources. The findings provide a rigorous mathematical foundation for the behavior of Werner states and contribute to the broader effort of classifying NPT bound entangled states, which are central to understanding the non-additivity of entanglement and the irreversibility of quantum resource manipulation.