We synthesized HfO2 nanocrystals from HfCl4 using a surfactant-free solvothermal process in benzyl alcohol and found that the resulting nanocrystals could be transferred to nonpolar media using a mixture of carboxylic acids and amines. Using solution (1)H NMR, FTIR, and elemental analysis, we studied the details of the transfer reaction and the surface chemistry of the resulting sterically stabilized nanocrystals. As-synthesized nanocrystals are charge-stabilized by protons, with chloride acting as the counterion. Treatment with only carboxylic acids does not lead to any binding of ligands to the HfO2 surface. On the other hand, we find that the addition of amines provides the basic environment in which carboxylic acids can dissociate and replace chloride. This results in stable, aggregate-free dispersions of HfO2 nanocrystals, sterically stabilized by carboxylate ligands. Moreover, titrations with deuterated carboxylic acid show that the charge on the carboxylate ligands is balanced by coadsorbed protons. Hence, opposite from the X-type/nonstoichiometric nanocrystals picture prevailing in literature, one should look at HfO2/carboxylate nanocrystals as systems where carboxylic acids are dissociatively adsorbed to bind to the nanocrystals. Similar results were obtained with ZrO2 NCs. Since proton accommodation on the surface is most likely due to the high Brønsted basicity of oxygen, our model could be a more general picture for the surface chemistry of metal oxide nanocrystals with important consequences on the chemistry of ligand exchange reactions.

Ligand exchange is central in the processing of inorganic nanocrystals (NCs) and requires understanding of surface chemistry. Studying sterically stabilized HfO2 and ZrO2 NCs using (1) H solution NMR and IR spectroscopy as well as elemental analysis, this paper demonstrates the reversible exchange of initial oleic acid ligands for octylamine and self-adsorption of oleic acid at NC surfaces. Both processes are incompatible with an X-type binding motif of carboxylic acids as reported for sulfide and selenide NCs. We argue that this behavior stems from the dissociative adsorption of carboxylic acids at the oxide surface. Both proton and carboxylate moieties must be regarded as X-type ligands yielding a combined X2 binding motif that allows for self-adsorption and exchange for L-type ligands.

Surface chemistry bridges the gap between nanocrystal synthesis and their applications. In this respect, the discovery of complex ligand binding motifs on semiconductor quantum dots and metal oxide nanocrystals opens a gateway to new areas of research. The implications are far-reaching, from catalytic model systems to the performance of solar cells.

Although TiO2, ZrO2 and HfO2 nanocrystals are often synthesized in tri-noctylphosphine oxide (TOPO), it is unclear whether TOPO also serves as ligand. Using liquid and solid state 1 H and 31 P nuclear magnetic resonance spectroscopy and X-ray fluorescence spectroscopy, we show that the nanocrystal surface is capped by several derivatives of TOPO. In the 31 P NMR spectrum, di-noctylphosphinate (δ = 57 ppm) and P,P'-(di-n-octyl) pyrophosphonate (δ = 20 ppm) are found coordinated to the nanocrystal. In addition, hydrogen chloride associates with the metal oxide nanocrystal surface and protonates TOPO. The resulting hydroxyl-tri-n-octylphosphonium, [HO-PR3] + , is tightly associated with the nanocrystal surface (δ(31 P) = 73 ppm) due to electrostatic interactions and hydrogen bonding. To simplify the complex surface composition, we exchange the original surface species for carboxylate or phosphonate ligands. The protonation of TOPO is an unexpected example of lyophilic ion pairing between an acidic metal oxide nanocrystal and a weakly basic ligand molecule that is formed in nonpolar solution. Our results contrast with the classically envisaged L-type binding motif of TOPO to surface metal ions. The generality of this stabilization mode and its relevance to catalysis is discussed.