Figure 16. STM images of the initial and final stages in the self-assembly of Fe nanowires obtained by selective adsorption of ferrocene molecules. (a) CaF2 stripes separated by CaFj/Si(111) trenches at a coverage of 1.9 monolayers of CaF2. (b) Linear array of Fe nanowires 3 nm wide and 0.8 nm thick. Reprinted with permission from [422], J. L. Lin et al., Appl. Phys. Lett. 78, 829 (2001). © 2001, American Institute of Physics.

with a size of a few nanometers exhibit unique properties (e.g., high coercivity, superparamagnetism, enhanced surface sensitivity) not shown by their bulk phases, which has generated a great deal of technological and fundamental interest in the size and phase-selective synthesis of iron oxide particles and films. The functional properties of iron oxides are strongly dependent on their chemical composition, microstructure, and phase purity. The control over morphology, microstructure, and composition poses a synthetic challenge and is influenced by the preparation method, specific synthesis conditions, and pyrolytic or hydrolytic properties of the precursor [123] under processing conditions. The production of magnetic nanoparticles has improved enormously in recent years, although the synthesis of high-purity iron oxide nanoparticles a few nanometers in size with controlled size and distribution is still a challenging problem [124-127, 428-430] despite the application of different physical (e.g., molecular beam epitaxy (MBE) [431], reactive sputtering [432], and pulsed laser deposition (PLD) [433]) and chemical (e.g., sol-gel [434], electrochemical synthesis [435], ball milling [436], aerosol synthesis [437], and CVD [438]) methods. In view of the above, the use of iron alkoxide clusters as precursors to nanocrystalline oxides is a convenient route for controlling the composition and morphology of the target oxide. Furthermore, it allows a phase-selective synthesis, which is of significant importance, especially in the case of oxides, for instance, the Fex-Oy system, where several phases (hematite (Fe2O3), maghemite (y-Fe2O3), magnetite (Fe3O4), etc.) can be simultaneously formed in a narrow processing window.

Kiyomura and Gomi [127] has used iron (III) acetylaceto-nate complex (Fig. 17) in a plasma-assisted MOCVD process to grow magnetite films on Si(111) with a uniform morphology at 400 °C; however, the carbon content in the deposited film was rather high (2-6 at.%) because of the thermally robust nature of the precursor.

Armelao and Artigliato [124] have used iron ethoxide [Fe(OEt)3]„ as a single molecular source for the sol-gel synthesis of nanocrystalline and transparent hematite thin films. The crystalline phase was obtained by thermal treatment of a homogeneous and amorphous material. The crystallite dimensions were controlled by varying the temperature, and the coatings were found to be nanostructured, even after

prolonged heating. The iron sites were found to be octa-hedrally coordinated in both the crystalline and amorphous states, which shows the inherent advantage of using a single molecular source, which results in a chemically homogeneous matrix. As a result, the microstructural evolution is consistent and stable [124]. Although iron ethoxide is an interesting precursor for solution methods, its application in gas-phase techniques is limited because of its poor volatility. The attempted volatilization (200 0C/10-2 torr) results in the formation of a polynuclear iron cluster, Fe9O3(OC2H5)21(C2H5OH), with low vapor pressure [439]. The molecular structure of this species reveals a cyclic arrangement of Fe-O rings based on the different coordination of Fe(III) centers (Fig. 18). The oxo-alkoxide species represent a molecular intermediate that is probably involved in the alkoxide-to-oxide conversion.

The thermal instability of iron ethoxide is related to the degree of association of the ethoxide molecule in the solid state. It is known that the corresponding aluminum compound is an infusible solid due to the extensive polymerization of Al(OEt)3 through alkoxy-bridging. The bridging tendency in alkoxide species is governed by the propensity of the metal center to increase their coordination state [49-51], which can be tuned either through an increase in the steric bulk of the alkoxide ligands or through the use of chelating ligands [440, 441]. Indeed, the higher alkoxide analogue [Fe(OBu')3]2 was found to be an excellent precursor for the deposition of different iron oxide phases. The molecular structure of [Fe(OrBu)3]2 exhibits a centrosym-metric dimer where two tetrahedrally coordinated Fe (III) centers are linked via bridging -OtBu groups (Fig. 19).

The decomposition temperature of [Fe(OrBu)3]2 needed to form iron oxide was determined by thermogravimetric analysis performed under nitrogen. The major weight loss occurs in a single decomposition step around 180 0C corresponding to the elimination of the organic ligands to form the inorganic oxide. This demonstrates the inherent advantage of SSPs in obtaining pure compounds with control over the phase purity and particle size in the resulting ceramic material. Such a single-step decomposition behavior is not observed in other precursor systems, for example, the organic derivatives of the inorganic iron salts like glycolates, citrates, j8-diketonates, etc., where the precursor degrades in several

Figure 18. Molecular structure of Fe9O3(OC2H5)21(C2H5OH)

Figure 19. Molecular structure of [Fe(O(Bu)3]2.

steps, because of the stability of intermediate species; consequently higher temperatures are necessary for a complete decomposition of the precursor to form the oxide [442]. The DTA curve shows a small endothermic peak around 150 0C corresponding to the melting temperature of the precursor, and the exothermic transition at 225 0C is possibly due to the oxidative degradation of the iron alkoxide, resulting in the liberation of the organic by-products. The crystallization of hematite is observed as a small exothermal peak around 370 0C. [Fe(OrBu)3]2 was used in a horizontal cold-wall reactor to deposit hematite films at 400 0C, whereas magnetite films were obtained at 450 and 500 0C, as confirmed by the X-ray diffraction (XRD) data (Fig. 20). It was shown that the Fe3O4 phase results from the reduction of a and y ferric oxides [123, 428]. Apparently, the dangling oxygen bonds on the surface of ferric oxide nanocrystals cleave upon heating in a vacuum, thereby eliminating oxygen. However, the reduction of Fe3+ to Fe2+ by dihydrogen liberated during the decomposition of [Fe(OrBu)3]2 cannot be ruled out (Eqs. (15) and (16)).

6Fe2O3 ^ 4Fe3O4 + O2

Subject to the nature of the metal atom, a thermally activated fragmentation of the metal tert-butoxides produces,

0 0

Post a comment