Synthesis, characterisation and application of silica-magnetite nanocomposites

Abstract Silica-magnetite composites were prepared for eventual applications in biomolecular separations (nucleic acids). Their production on large scale has been optimised and they have been extensively characterised in a physical and chemical context. They perform at least as well, if not better than a commercially available equivalent at adsorbing and eluting DNA. Several methods for the preparation of magnetite were compared in order to select one, which produced particles, possessing high magnetic susceptibility, low rate of sedimentation and good chemical stability. Of the main methods studied: (i) oxidative hydrolysis of iron(II) sulphate in alkaline media, (ii) alkaline hydrolysis of iron(II) and iron(III) chloride solutions, and (iii) precipitation from iron(II) and iron(III) chloride solutions by hydrolysis of urea, method (i) produced the ‘best’ magnetite particles. Silica-magnetite composites were prepared using the ‘best’ magnetite, and, for comparison, two methods for depositing silica were used to coat the silica onto magnetite nanoparticles, from silicic acid at pH 10 and by acid hydrolysis of tetraethoxysilane (TEOS) at 90 °C. The best method for yielding silica-magnetite composites that worked well in DNA adsorption and elution proved to be that involving silicic acid and this material could be made in 20 g batch sizes. Silica-magnetite composites from the two methods proved to have distinct and different physical and chemical properties. All magnetite and silica-magnetite samples were fully characterised for their relative chemical composition using Fourier-transform infrared, XRF and thermo-gravimetric analysis. Their physical characteristics were determined using scanning electron microscopy and N 2 adsorption and Mossbauer spectroscopy was used to confirm the identity of the iron oxides produced. Selected samples were comparatively tested for their ability to adsorb, and subsequently elute, 2-deoxyguanosine-5-monophosphate (GMP) and its non-phosphorylated analogue 2-deoxyguanosine (G) and a range of sequence defined oligonucleotides (NAs) and sheared salmon sperm DNA. It was found that magnetite readily adsorbed GMP via the GMP phosphate anion in water, whereas silica did not, due to electrostatic repulsion between the negatively charged surface of silica and the GMP. Both magnetite and silica magnetite were further tested in adsorption studies of G and GMP in different chaotropic media, 4 M sodium chloride or 4 M ammonium sulphate. The high salt conditions aided binding of GMP silica magnetite but inhibited adsorption to magnetite presumably due to competition for binding sites on the magnetite's surface by the chaotrope anions. Interestingly, the results from NAs binding studies indicated that sequence appeared to play an important role in adsorption of the different species to silica-magnetite composites. This may indicate a contribution by hydrophobic interactions to the binding mechanism. Multiple depositions of silica onto magnetite performed by deposition from silicic acid at pH 10 did not appear to greatly increase the composite percentage represented by silica whilst composite produced by the acid hydrolysis of TEOS at 90 °C did. However, it appeared that the silica deposited by the first method represented a complete coating of the magnetite core whilst the second method yielded a porous or incomplete coating. In comparison with commercially available silica-magnetite composite in DNA adsorption and elution, the material was observed to perform approximately 10% more efficiently. These findings indicate that it is possible to produce a consistent and cheap silica-magnetite nanoparticle on relatively large scale (greater than 20 g batch size) which is at least as good as, if not better than, a commercially available alternative.