In some references that I have read, a crucial assumption in deriving the Lindblad master equation is that the system and environment remain separable for all time. Hence, the system and environment cannot entangle with each other at any time. Hence, decoherence cannot occur.
However, there are papers which discuss decoherence dynamics using the Lindblad master equation. How can this be the case? An example of such a paper is "Decoherence by Lindblad Motion" by Klaus Dietz.