Size-dependent structural transformations of hematite nanoparticles. 1. Phase transition

Using Fourier Transform InfraRed (FTIR) spectroscopy, Raman spectroscopy, X-ray diffraction (XRD), and Transmission Electron Microscopy (TEM), we characterize the structure and/or morphology of hematite (alpha-Fe(2)O(3)) particles with sizes of 7, 18, 39 and 120 nm. It is found that these nanoparticles possess maghemite (gamma-Fe(2)O(3))-like defects in the near surface regions, to which a vibrational mode at 690 cm(-1), active both in FTIR and Raman spectra, is assigned. The fraction of the maghemite-like defects and the net lattice disorder are inversely related to the particle size. However, the effect is opposite for nanoparticles grown by sintering of smaller hematite precursors under conditions when the formation of a uniform hematite-like structure throughout the aggregate is restricted by kinetic issues. This means that not only particle size but also the growth kinetics determines the structure of the nanoparticles. The observed structural changes are interpreted as size-induced alpha-Fe(2)O(3)<-->gamma-Fe(2)O(3) phase transitions. We develop a general model that considers spinel defects and absorbed/adsorbed species (in our case, hydroxyls) as dominant controls on structural changes with particle size in hematite nanoparticles, including solid-state phase transitions. These changes are represented by trajectories in a phase diagram built in three phase coordinates-concentrations of spinel defects, absorbed impurities, and adsorbed species. The critical size for the onset of the alpha-->gamma phase transition depends on the particle environment, and for the dry particles used in this study is about 40 nm. The model supports the existence of intermediate phases (protohematite and hydrohematite) during dehydration of goethite. We also demonstrate that the hematite structure is significantly less defective when the nanoparticles are immersed in water or KBr matrix, which is explained by the effects of the electrochemical double layer and increased rigidity of the particle environment. Finally, we revise the problem of applicability of IR spectroscopy to the lattice vibrations of hematite nanoparticles, demonstrating that structural comparison of different samples is much more reliable if it is based on the E(u) band at about 460 cm(-1) and the spinel band at 690 cm(-1), instead of the A(2u)/E(u) band at about 550 cm(-1) used in previous work. The new methodology is applied to analysis of the reported IR spectra of Martian hematite.