Eigenrate-Based Thermodynamic Decomposition of Competing Release Mechanisms in Polymeric Nano- and Microspheres: The...

Simultaneous operation of Fickian diffusion, swelling-mediated relaxation, and polymer erosion in polymeric nano- and microspheres produces release profiles whose mechanistic origins cannot be resolved by empirical models without system-specific refitting. I present the Modified Multicomponent Interactive Release for Spheres (MMIR-S) framework, a thermodynamically self-consistent mechanistic model for homogeneously loaded spherical particles that determines mechanism-specific release weights from tabulated molecular descriptors, without fitting to release data. The theoretical core is Thermodynamic Eigenrate Decomposition (TED), in which each mechanism is assigned a commensurable first-order depletion rate: the diffusion eigenrate (π²D_eff,app/R²) from the dominant eigenvalue of Fick's second law in spherical coordinates; the swelling eigenrate governed by the Flory-Rehner criterion, closing when χ_PM > 0.5; and the erosion rate accommodating surface, first-order, or bulk erosion kinetics. Normalizing these rates yields mechanism weights summing to unity, connected to Hansen solubility parameters, Flory-Huggins interaction parameters, and partition coefficients without empirical intermediaries. The burst release term distinguishes physically adsorbed drug, governed by a first-order desorption rate constant, from chemically adsorbed drug governed by an Arrhenius rate constant parameterized by binding free energy, producing a biexponential burst profile. A polydispersity correction integrates the release function over a log-normal size distribution via Gauss-Hermite quadrature. Validation against the six-compound, two-pH, multi-polymer dataset of Stiepel et al. [1] demonstrates TED weight distributions consistent with effective diffusivity trends recovered by regression and machine learning. The framework recovers the Higuchi, Hopfenberg, and first-order models as limiting cases, providing a foundation for a priori prediction of release profiles from molecular structure alone.