RESUMO
We describe the use of a modified Stöber method for coating maghemite (γ-Fe(2)O(3)) nanocrystals with silica shells in order to built magnetic fluorescent sensor nanoparticles in the 50-70nm diameter range. In detail, the magnetic cores were coated by two successive silica shells embedding two fluorophores (two different silylated dye derivatives), which allows for ratiometric pH-measurements in the pH range 5-8. Silica coated magnetic nanoparticles were prepared using maghemite nanocrystals as cores (5-10nm in diameter) coated by tetraethoxyorthosilicate via hydrolysis/condensation in ethanol, catalyzed by ammonia. In the inner shell was covalently attached a sulforhodamine B, which was used as a reference dye; while a pH-sensitive fluorescein was incorporated into the outer shell. Once synthesized, the particles were characterized in terms of morphology, size, composition and magnetization, using dynamic light scattering (DLS), transmission electron microscopy (TEM), X-ray diffraction (XRD) and vibrating sample magnetometry (VSM). TEM analysis showed the nanoparticles to be very uniform in size. Wide-angle X-ray diffractograms showed, for uncoated as well as coated nanoparticles, typical peaks for the spinel structure of maghemite at the same diffraction angle, with no structural changes after coating. When using VSM, we obtained the magnetization curves of the resulting nanoparticles and the typical magnetization parameters as saturation magnetization (M(s)), coercivity (H(c)), and remanent magnetization (M(r)). The dual-dye doped magnetic-silica nanoparticles showed a satisfactory magnetization that could be suitable for nanoparticle separation and localized concentration of them. Changes in fluorescence intensity of the pH indicator in the different pH buffered solutions were observed within few seconds indicating an easy accessibility of the embedded dye by protons through the pores of the silica shell. The relationship between the ratio in fluorescence (sensor/reference dyes) and pH was adjusted to a sigmoidal fit using a Boltzmann type equation. Finally, the proposed method was statistically validated against a reference procedure using samples of water and physiological buffer with 2% (w/v) of horse serum added, indicating that there are no significant statistical differences at a 95% confidence level.
