Near-field mechanical fingerprints for terahertz sensing of hidden nanoparticles in complex media

Terahertz (THz) spectroscopy holds transformative potential for noninvasive sensing, yet characterizing individual nanoparticles in complex biological environments remains challenging due to the far-field diffraction limit. While near-field dipolar theory is well established, its application to characterizing/identifying nanoparticles immersed in complex media at THz frequencies is largely unexplored. This work utilizes numerical simulations of magneto-optical (MO) heterodimers—comprising n -doped indium antimonide ( [ n -InSb, GaSe ] ) and isotropic or birefringent particles (e.g., SiO 2 , GaSe )—under counterpropagating, circularly polarized THz illumination. The analytical framework is supported by a numerical validation against full-wave simulations. We demonstrate that while far-field observables like absorption cross sections are often dominated by the MO-active particle, mechanical variables—specifically induced binding forces and spin/orbital torques—exhibit superior sensitivity for detecting “hidden” neighboring components. Because these mechanical signatures depend directly on near-field interactions, they provide higher information density regarding interparticle coupling. Key findings reveal material-specific spectral “hot spots” and “zeros” that serve as robust calibration markers even within dispersive biological surrogates. We show that the spin torque on non-MO particles is significantly modified by MO-neighbor proximity, a phenomenon controllable via static magnetic fields. While experimental implementation in complex media presents significant challenges, our results establish a theoretical roadmap for using optomechanical signatures as high-resolution detection channels for diagnostics, signal transduction, and low-energy nanocircuit control.