DOI: 10.1002/cphc.201300161 Experimental and First-Principles Characterization of Functionalized Magnetic Nanoparticles Georgios S. E. Antipas,* Eleftherios Statharas, Philippos Tserotas, Nikolaos Papadopoulos, and E. Hristoforou [a] 1. Introduction Potential applications of iron oxide nanoparticles (NPs) are di- verse owing to their magnetic, catalytic, and conducting nature, as well as their promise for mediation of biological functions. [1] Among iron oxides of interest, magnetite (Fe 3 O 4 ) and maghemite (g-Fe 2 O 3 ) are of elevated importance as their fundamental properties of superparamagnetism, [2] magnetic di- polar interactions, [3] single-electron transfer, [4] and magnetore- sistance [5] are applied to magnetic resonance imaging, [6] nano- device functionalization, [7] high-density data storage, pigment synthesis, [8] and mediation of mineral separation. [9] Most proba- bly, however, the future of NPs lies within the realm of medical diagnosis and therapy [10] through facilitating cell separation, [11] targeted drug delivery, [12] or inducing cell apoptosis through hyperthermia. [13] Apart from their promise, magnetic NPs also present a number of challenges in their routine integration into appli- cations. Notably, NP aggregation due to strong magnetic dipole–dipole interactions, typically associated with magnet- ite, [14] may compromise biocompatibility. Most biological appli- cations require monodisperse NPs of size between 5 and 20 nm [12a, 15] but size variance with respect to the synthesis route induces unpredictable magnetic behavior, [16] possibly due to alterations of the NP lattice. Also, modification of the NP surface is normally required to avoid immune response trig- gering; this is achieved by coating the NP with various bio- compatible polymers, [17] which in turn raise the issue of stabili- ty of the coated NP in the context of bond strength between core and coating (surfactant). Hence, the stability of the func- tionalized NP translates into surfactant binding avidity to the NP surface and, correspondingly, into the potential modifica- tions brought to the surface itself as a result of particle size. To this extent, magnetic NPs have been reported to assume an oxidized-like state with an elementary unit cell that is similar to magnetite but includes only Fe 3 + , [18] whereas, in another in- stance, the magnetite (001) surface has been reported as pref- erentially terminated by tetrahedral Fe atoms, based on results from charge-ordered density functional theory (DFT) calcula- tions. [19] NP synthesis itself is achieved by an array of methods, including reduction of hematite by CO/CO 2 , [20] coprecipitation from a solution of ferrous/ferric salt mixtures in an alkaline medium, [21] hydrolysis, [22] sol–gel-based techniques, and oxida- tion of Fe(OH) 2 by H 2 O 2 . [23] Recently, template-confined meth- ods have also been adopted to develop nanocrystalline Fe 3 O 4 NPs; [24] one of the major drawbacks of this technique is that complete removal of the template is a tedious task that affects the purity of the NPs. [25] Magnetic iron oxide nanoparticles synthesized by coprecipita- tion and thermal decomposition yield largely monodisperse size distributions. The diameters of the coprecipitated particles measured by X-ray diffraction and transmission electron mi- croscopy are between approximately 9 and 15 nm, whereas the diameters of thermally decomposed particles are in the range of 8 to 10 nm. Coprecipitated particles are indexed as magnetite-rich and thermally decomposed particles as maghe- mite-rich; however, both methods produce a mixture of mag- netite and maghemite. Fourier transform IR spectra reveal that the nanoparticles are coated with at least two layers of oleic acid (OA) surfactant. The inner layer is postulated to be chemi- cally adsorbed on the nanoparticle surface whereas the rest of the OA is physically adsorbed, as indicated by carboxyl O ÀH stretching modes above 3400 cm À1 . Differential thermal analy- sis (DTA) results indicate a double-stepped weight loss process, the lower-temperature step of which is assigned to condensa- tion due to physically adsorbed or low-energy bonded OA moieties. Density functional calculations of Fe–O clusters, the inverse spinel cell, and isolated OA, as well as OA in bidentate linkage with ferrous and ferric atoms, suggest that the higher- temperature DTA stage could be further broken down into two regions: one in which condensation is due ferrous/fer- rous– and/or ferrous/ferric–OA and the other due to condensa- tion from ferrous/ferric– and ferric/ferric–OA complexes. The latter appear to form bonds with the OA carbonyl group of energy up to fivefold that of the bond formed by the ferrous/ ferrous pairs. Molecular orbital populations indicate that such increased stability of the ferric/ferric pair is due to the contri- bution of the low-lying Fe 3 + t 2g states into four bonding orbi- tals between À0.623 and À0.410 a.u. [a] Dr. G. S. E. Antipas, E. Statharas, P. Tserotas, Dr. N. Papadopoulos, Prof. Dr. E. Hristoforou School of Mining Engineering and Metallurgy National Technical University of Athens Zografou Campus, Athens 15780 (Greece) E-mail : gantipas@metal.ntua.gr  2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim ChemPhysChem 2013, 14, 1934 – 1942 1934 CHEMPHYSCHEM ARTICLES