Interfacial coupling between CoFe2O4 (CFO) nanoparticles and oxidatively functionalized multi-walled carbon nanotubes (MWCNTs) enables controlled modulation of structural, optical, and spin dynamic properties in CFO–MWCNT nanocomposites. The solvothermal synthesis promotes nucleation of CFO on –COOH/–OH functional groups, ensuring uniform anchoring along the nanotube surface. X-ray diffraction confirms a cubic spinel phase with lattice expansion from 8.385 Å to 8.410 Å and crystallite growth from 18 nm to 25 nm, reflecting strain transfer and partial nanoparticle coalescence at the carbon interface. The observed bandgap narrowing from 2.72 eV to 2.50 eV, confirmed via Tauc plot analysis, is attributed to localized defect states induced by charge delocalization and orbital hybridization at the interface of the CFO–MWCNT boundary. DC magnetometry reveals a reduction in saturation magnetization from 46 emu/g to 35 emu/g due to diamagnetic dilution and interfacial spin canting, while coercivity decreases from 852 Oe to 841 Oe, indicating modified pinning and domain-wall dynamics associated with exchange-coupled interfaces. Ferromagnetic resonance measurements show a resonance field shift from 3495 G to 3500 G and an increase in the Landé g-factor from 1.97 to 2.00, signifying altered spin–orbit coupling and enhanced local magnetic perturbations. The spin–lattice relaxation time increases from 1.41 ns to 1.59 ns, demonstrating suppressed phonon-mediated relaxation and improved spin coherence across the hybrid network. Spin density rises from 3.72 × 1022 to 4.58 × 1022 spins/g, confirming an increase in unpaired electrons generated by orbital asymmetry at the interface. The anisotropy field and effective magnetocrystalline anisotropy constant exhibit pronounced modulation, evidencing strengthened exchange stiffness and altered Co2+/Fe3+ superexchange pathways. These results establish CFO-MWCNT nanocomposites as tuneable platforms for spintronic logic elements, high-frequency microwave attenuation, and magneto-optical device architectures.
Kumar et al. (2026) studied this question.