Simulation-Guided Tracer Optimization for Magnetic Particle Spectroscopy: Interplay of Core, Shell and Anisotropy Across Physiological-Range Viscosities for Single- and Dual-Frequency Excitation.

Detection of magnetic nanoparticles (MNP) relies on their nonlinear magnetic response, changing with the local environment in vivo. Employing an effective viscosity model within stochastic coupled Néel-Brownian simulations, we quantify how viscosity (0.8-50 000 mPa·s) affects signal intensity of single-frequency magnetic particle spectroscopy (MPS, 25 kHz) and frequency-mixing MPS (FM-MPS; fH = 40.5 kHz, fL = 0.063 kHz). We investigate single-core, non-interacting MNP with various core sizes (12-28 nm), hydrodynamic shell thicknesses (5-100 nm), and anisotropy (4-200 kJ m- 3), using dominant readouts (MPS: third-harmonic; FM-MPS: fH + 2fL mixing-term). Increasing viscosity progressively suppresses Brownian rotation, with thicker shells affected at lower viscosities, and drives signals toward an immobilization plateau. Under the simulated conditions, MNP appear distinguishable in blood-associated viscosities using MPS, whereas distinguishability is pronounced in cellular/tissue-like viscosities with FM-MPS. Moderate anisotropy (∼20-40 kJ m- 3) enhances signals by ∼10%, whereas higher anisotropies suppress relaxation, finally collapsing the signal. While large cores maximize the signal at fixed particle number, dosage-normalized conditions suggest favorable performance for intermediate cores (∼15-18 nm). These results propose a cohesive framework linking particle properties, excitation scheme, and physiological environment, thereby providing insight into MNP tracer optimization for biosensing, cell tracking, imaging, and in vivo diagnostics.