arXiv:2610.34322
Two-qubit binegativity can increase under local amplitude damping
Xiao-Lan Zong, Wei Song, Ming Yang
Abstract
It has been conjectured that the binegativity of two-qubit states is an entanglement monotone. We present a full-rank state as a counterexample: applying amplitude damping to only one of the qubits can increase the binegativity of the state. The output state remains full rank, and the amount of increase can be calculated exactly. In addition, we construct a family of counterexamples that vary continuously with a parameter. In the corresponding one-way local operations and classical communication protocol, the probability-weighted average of the binegativity over the branches after the local measurement is also higher than that of the initial state.
AI Summary
The paper addresses the conjecture that two-qubit binegativity is an entanglement monotone under LOCC and PPT-preserving channels. It disproves both deterministic monotonicity and monotonicity on average using explicit two-qubit counterexamples. The first counterexample is a full-rank state subjected to amplitude damping on one qubit with damping probability $253/289$. The output remains full rank, while binegativity increases from $54/47125$ to $1248/1047625$, a relative increase of approximately $3.96\%$, without postselection. The calculation uses the identity $N_2(\rho)=N(\rho)[1+C(v)]/2$, where $v$ is the eigenvector associated with the unique negative eigenvalue of the partial transpose. Exact eigenvalues and eigenvectors show that increased concurrence of this eigenvector outweighs decreased negativity. Continuity gives nearby full-rank counterexamples under the same channel. A second construction uses $\rho_p=p|\Phi^+\rangle\langle\Phi^+|+(1-p)|10\rangle\langle10|$, with $0<p<3/7$, and a local diagonal filter followed by conditional resetting. An analytic inequality establishes increased binegativity throughout this parameter range. The local measurement alone also increases probability-weighted average binegativity; a separate full-rank example is provided. The paper also explains why inputs satisfying $N_2(\rho)=N(\rho)$ cannot exhibit an increase above their initial value under PPT channels with two-qubit outputs. This restriction covers states used in earlier noise studies, but does not establish monotonicity between intermediate evolution times. The results concern binegativity specifically, rather than a failure of negativity or concurrence monotonicity.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 4 pages, 1 figures