Local Stability and Hopf Bifurcation in a Three-Dimensional Photocatalytic Microplastic Reactor Model with Adaptive Gain

Adaptive feedback can destabilize a loop that would be stable under any fixed gain, so the speed at which the gain adapts is itself a design parameter. We study this effect in a minimal three-dimensional model motivated by the photocatalytic degradation of microplastics: a pollutant concentration is driven toward a setpoint by an ultraviolet (UV) actuator whose gain adapts online. The model has a single bilinear nonlinearity, so the local analysis can be carried out in closed form. Under an explicit feasibility condition, the system has a unique positive equilibrium. The Routh–Hurwitz criterion shows that this equilibrium is locally asymptotically stable below an explicit critical adaptation speed κc and unstable above it. At κ=κc, a purely imaginary eigenvalue pair crosses the imaginary axis transversally, and a Hopf bifurcation occurs, with an explicit onset frequency. The first Lyapunov coefficient is computed in closed form; it separates a supercritical onset, for well-damped actuators, from a subcritical onset with hysteresis, for weakly damped actuators. Numerical experiments confirm the predicted limit cycle and the classification. All the stability results established here are local.