arXiv:2610.03676
How I Learned to Stop Worrying and Love the Redfield Equation: Completely-Positive Resummation of Non-Markovian Dynamics
Greg Kaplanek, Jason Pollack
Abstract
The Redfield equation with time-dependent coefficients (or TCL2 equation) gives one of the simplest perturbative descriptions of open quantum system dynamics derived from a microscopic theory. Its status as an approximation is nevertheless subtle: while perturbation theory fixes the reduced dynamics order-by-order in the system-environment coupling, solving the Redfield equation yields a particular resummation that need not preserve complete positivity. Here we compare Redfield evolution with an exponentiated (Magnus) resummation of the same second-order perturbative map, which is completely positive and trace preserving (CPTP) for a broad class of physical conditions. We also construct a simple CPTP resolvent resummation, illustrating the non-uniqueness of resummations of the same perturbative map. The Redfield and Magnus prescriptions agree to second order but differ through time-ordering corrections at higher orders. We exhibit a weak-coupling criterion which guarantees that Redfield evolution, even when non-Markovian, remains close to a CPTP map. We illustrate the criterion in two examples, focusing on finite-dimensional (qubit) systems. Our results clarify when, and why, Redfield evolution can remain reliable even when it is highly non-Lindbladian.
AI Summary
The paper asks when finite-time Redfield/TCL2 evolution can remain close to a completely positive map despite a non-GKLS instantaneous generator. It treats solutions of truncated master equations as nonunique resummations of the same microscopic second-order reduced map, $\mathcal I+\lambda^2\Phi_2(t)$, assuming an initially factorized system–environment state. Using positivity of the environmental two-time correlation kernel, it writes $\Phi_2(t)$ in GKLS form at each fixed physical time. This establishes that both the exponential $\Phi_{\rm Mag}=e^{\lambda^2\Phi_2}$ and resolvent $\Phi_{\rm res}=(\mathcal I-\lambda^2\Phi_2)^{-1}$ are CPTP completions. The exponential construction connects to existing cumulant and coarse-graining approaches, without requiring stationary correlations. Redfield evolution differs from it through Magnus time-ordering corrections, beginning with $\lambda^4\Omega_4=\frac{\lambda^4}{2}\int ds\,[\dot\Phi_2(s),\Phi_2(s)]$. Mutually commuting unequal-time generators make the two maps identical. The proposed diagnostic is $\epsilon_{\rm Mag}=\lambda^2\|\Omega_4\|/\|\Phi_2\|$. Analytical calculations and numerical complete-positivity tests examine a qubit coupled to another qubit or a Lorentzian oscillator bath. Near resonance in the two-qubit example, the regime $\lambda^2/\omega\ll\Omega-\omega\lesssim\lambda$ separates small relative time-ordering corrections from loss of fixed-order perturbative control. Bath examples give diagnostic scalings $\lambda^2/(\omega\gamma)$ for short correlation times and $\lambda^2/|\Omega^2-\omega^2|$ for long ones. The diagnostic controls only the first Magnus correction: it neither bounds the full series nor guarantees Redfield complete positivity or accuracy against exact dynamics. Examples retain small positivity violations; infinite-dimensional systems require additional care.
Central claim may not hold: The abstract promises a weak-coupling proximity guarantee, but the body provides only a leading-Magnus diagnostic conditional on unestablished control of higher corrections. The proposed criterion bounds a relative exponent correction rather than a channel distance, and the paper supplies neither an absolute map-error bound nor uniform-in-time remainder control. The oscillator section incorrectly represents its nonnegative-frequency correlator using Lorentzian mode functions integrated over all real frequencies; the decomposition must instead integrate from zero to infinity.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 18 pages + appendices; 4 figures