Floquet-Resolved Dissipation Selects Entanglement Beyond Population Spectroscopy

Abstract

Reliable quantum-state engineering in periodically driven devices requires more than reproducing their excitation spectrum: the environment must resolve the transitions of the driven system. We show that two dissipative descriptions of the same parametrically coupled qubits can yield closely similar period-averaged populations yet qualitatively different asymptotic entanglement. Both retain the complete periodically driven Hamiltonian, the same microscopic bath couplings, and the same physical observables; they differ in the dynamical representation used to resolve the dissipative channels and the corresponding system operators: the static dressed basis in the partial harmonic decomposition approach (PHDA) and the Floquet basis, including all relevant drive sidebands, in Floquet-Born-Markov (FBM) theory. The population maps preserve the same resonance skeleton and remain closely similar over broad low-to-moderate drive regions, while clearer differences emerge as the modulation becomes stronger. Phase-resolved single- and two-qubit coherence observables reveal a substantially stronger redistribution of amplitudes and phases. Concurrence amplifies this hidden state-level discrepancy: Floquet-resolved dissipation selects stronger and more extended Bell-like entangled regions and a larger overlap with a fixed Bell-like reference, while PHDA generally underestimates the entanglement and its thermal persistence. A channel-resolution test verifies the secular, completely positive FBM construction throughout the relevant parameter domain. Our results establish population agreement as an insufficient benchmark for open Floquet quantum-state engineering and identify coherence-sensitive observables as the decisive validation test.