MOFs (Metal-organic frameworks) have drawn growing interests in the past decades. These microporous crystalline structures are known for their extraordinary tunability and applications in fields across gas storage, catalysis, chromatography and drug delivery. While the highly porous structures of MOFs have been constantly reported to be prospective as carrier materials, analytical techniques that are capable of probing the spatial distribution of guest molecules inside MOFs directly are scarce. In this work, we explored the potential of in situ gas cluster ion beams (GCIBs) and mono-atomic beams sputter to construct the molecular depth profile with Time-of-flight secondary ion spectrometry (ToF-SIMS). Fixed acquisition phase (pulsed C₆₀⁺ beam) was used to evaluate the difference between sputter parameters. Three different MOFs were prepared, namely ZIF-8, UiO-66 and HKUST-1. Although cluster beams of lower energy per atom (E/n) could better preserve organic structures, preferential sputtering is often observed for organic-inorganic hybrid structure of MOF, i.e. ion sputtering tends to remove the organic linkers of MOFs faster than the inorganic nodes. As a result, artifacts in depth profile may be observed. In order to obtain more balanced removal rate of each component while preserving organic structure, low-energy Ar⁺ cosputter may be employed to mitigate the preferential sputtering in MOFs. In this work, Ar-GCIB with different cluster sizes and different kinetic energies have been explored to sputter MOFs with and without Ar⁺ cosputter. Optimization of acceleration voltage and cluster size of Ar-GCIB and the current density of Ar⁺ has been performed to constructing authentic depth profiles of MOFs. The results showed that ZIF-8 could be profiled using single Ar-GCIB. On the other hand, the inorganic parts of UiO-66 and HKUST-1 could not be removed without Ar⁺ cosputter. This difference is attributed to the strong oxo-bonds of Zr₆O₄(OH)₄-cluster and Cu₂O₈-dimer nodes in UiO-66 and HKUST-1, respectively, and weaker ZnN₄ bonds in ZIF-8. Furthermore, to validate if the optimized parameter can be used to determine the distribution of molecules inside MOFs, caffeine and acetaminophen were chosen to be the model drug molecules. Finally, the interactions between molecules and MOFs were discussed according to the validated depth profiles.

Peng-Hsuan Chiang

Appearances