Hafnium sulfate prenucleation clusters and the Hf(18) polyoxometalate red herring

In prior studies, aqueous Hf sulfate-peroxide solutions were spin-coated, dehydrated, patterned by electron-beam lithography, ion-exchanged (OH(-) for SO4(2-)), and finally converted to HfO2 hard masks via annealing. The atomic-level details of the underlying aqueous chemistries of these processes are complex and yet to be understood. Yet a thorough understanding of this specific chemical system will inspire development of design rules for other aqueous-precursor-to-solid-state metal oxide systems. Often-observed crystallization of the Hf18 polyoxometalate from aqueous Hf sulfate-peroxide precursor solutions has led us to believe that Hf18 may represent an important intermediate step in this process. However, via detailed solution studies described here (small-angle X-ray scattering, electrospray ionization mass spectrometry, and Raman spectroscopy), we ascertained that Hf18 is in fact not a prenucleation cluster of Hf sulfate coatings. Rather, the Hf tetramers, pentamers, and hexamers that are the core building blocks of Hf18 are robustly persistent over variable compositions and aging time of precursor solutions, and therefore they are likely the rudimentary building blocks of the deposited thin-film materials. These Hf clusters are capped and linked by sulfate and peroxide anions in solution, which probably prevents crystallization of Hf18 during the rapid dehydration process of spin-coating. In fact, crystallization of Hf18 from the amorphous gel coating would be detrimental to formation of a high-density conformal coating that we obtain from precursor solutions. Therefore, this study revealed that the well-known Hf18 polyoxometalate is not likely to be an important intermediate in the thin-film process. However, its subunits are, confirming the universal importance of deriving information from the solid state, albeit judiciously and critically, to understand the solution state.