Utilizing Magnetic and Diffusive Properties to Improve Size Homogeneity of Superparamagnetic Nanoparticles

Superparamagnetic iron oxide nanoparticles (SPIONs) have enormous potential in biomedical applications, including drug delivery, hyperthermia, and as a contrast agent for magnetic resonance imaging (MRI). SPIONs with diameters greater than about 20 nm are generally synthesized with multi-core structures held together by an external matrix. This typically yields particles with a very broad size distribution, having a dynamic light scattering polydispersity index (PdI) of about 0.23. This broad size distribution hinders clinical translation of these particles due to safety and performance variability. While several size fractionation techniques have been employed, including magnetic fractionation, centrifugation, gel chromatography, and vacuum filtration to improve nanoparticle size homogeneity, PDI values remain much higher than gold, silica and other nanoparticles.

A new size fractionation method was based on the difference in particle mobility between different sized particles from a controlled periodic magnetic field. Separated SPIONs (SePs) had an average size distribution at least three times narrower than all four conventional methods. The method had an overall PdI 0.072 ± 0.0056, independent from five different scale-up conditions. A strong relationship was observed between a theoretically based mathematic model and experimental results, which confirmed the methodology toward a true homogenous size separation. Starch coated SPIONs with a PdI of 0.23 were separated into eight fractions with average sizes of 70.7, 72.3, 75.5, 78, 86, 100.6, 109.2, and 116.9 with PdIs of 0.061, 0.08, 0.07, 0.085, 0.074, 0.088, 0.073, and 0.094 respectively. Results indicated that the PdI of the SPION suspension could have a dramatic effect on MRI imaging and tissue penetration of these particles. SePs significantly increased the particle penetrating through 1 µm membrane and had distinct differences in MRI contrast enhancement and relaxivities from their polydispersed original sample and different conventional separation methods.