Hydrogen spillover through Matryoshka-type (ZIFs@)n−1ZIFs nanocubes

Hydrogen spillover phenomenon is well-documented in hydrogenation catalysis but still highly disputed in hydrogen storage. Until now, the existence of hydrogen spillover through metal–organic frameworks (MOFs) remains a topic of ongoing debate and how far the split hydrogen atoms diffuse in such materials is unknown. Herein we provide experimental evidence of the occurrence of hydrogen spillover in microporous MOFs at elevated temperatures, and the penetration depths of atomic hydrogen were measured quantitatively. We have made Matryoshka-type (ZIFs@)n−1ZIFs (where ZIFs = ZIF-8 or ZIF-67) nanocubes, together with Pt nanoparticles loaded on their external surfaces to produce atomic hydrogen. Within the (ZIFs@)n−1ZIFs, the ZIF-8 shell served as a ruler to measure the travelling distance of H atoms while the ZIF-67 core as a terminator of H atoms. In addition to the hydrogenolysis at normal pressure, CO2 hydrogenation can also trace the migration of H atoms over the ZIF-8 at high pressure.


Supplementary
Comments: As shown, Co 2+ and Zn 2+ ions are randomly distributed over the metal nodes sites in the ZIF structure, indicating that they are homogeneously mixed. And the average size of Zn/Co-ZIF is 107 nm which is ranged between the size of ZIF-67 nanocubes (208 nm) and ZIF-8 nanocubes (88 nm) prepared under similar conditions. Figure 14. Photographs of dry powders of ZIF-8, ZIF-67, and Matryoshka-type (ZIFs@)n1ZIFs nanocubes prepared from "large-scale" synthesis via increasing the reaction scale by 100-fold; refer to Methods in the main text for more information.

Supplementary
Comments: The 2-layer ZIFs represents ZIF-67@ZIF-8, 3-layer ZIFs represents ZIF-67@ZIF-8@ZIF-67, and 4-layer ZIFs is denoted as ZIF-67@ZIF-8@ZIF-67@ZIF-8. The FESEM images (at different magnifications) at the bottom panel were obtained from the ZIF-67 sample from "largescale" synthesis. Comments: The result of this experiment indicates that ZIF-67 structure is easily decomposed if it was not separated from Pt, which is different from ZIF-67@ZIF-8/Pt configuration. The purple frame shown in the inset of the image (d) indicates the (111) diffraction peak of Pt (39.8 degrees). It was reported in the literature that even in the inert gas atmosphere, metallic cobalt was obtained. 1 Herein, the decomposed cobalt species would be reduced by the hydrogen. As shown, the metallic cobalt metal was further oxidized to cobalt oxide (such as CoO, Co3O4) during the sample treatment for XRD. Binary CoO and Co3O4 were found.

Comments:
The spectra in Co 2p3/2 region can be deconvoluted into three peaks corresponding to Co 0 , Co 2+ , and the shake-up satellite. 2 The presence of Co 2+ is due to the oxidation of cobalt metal during sample preparation for XPS test. Moreover, as shown in Figure ( Comments: As shown, the thickness of ZIF-8 layer was adjusted as 5 nm, 10 nm, 20 nm, 30 nm, and 50 nm by tuning the synthesis parameters (e.g., the amount of ZIF-67 core and the amount of zinc nitrate solution). flowing (4.6% H2, 120 mL min 1 ). The temperature was raised from room temperature to 260 o C at a ramping rate of 3 o C min 1 under the flowing of N2 gas (120 mL min 1 ). Sample amount 20 mg. Figure 29. A proposed mechanism for consecutive hydrogen atom migration on ZIFs, where the ZIFs structure provides transport tunnels for both electron and proton.

Comments:
The above-proposed mechanism is an extension of a reported work on electron-proton mobility to the metal-organic frameworks. In that work, 3 the energy barrier (Eact) for combined electron-proton mobility on TiO2 is between 0.6 and 0.7 eV, which is energetically more favourable than other possible diffusion mechanism (such as oxygen vacancies). 3 In Supplementary Figure 30, a migrating proton (an H atom can be viewed as a proton and epair) 3 may prefer to bind with (negative) nitrogen of imidazolate linker, yielding a momentary N-H bond. Due to the adding proton, at the same time, the N-Zn bond is weakening or broken, and the electron (e -) from the H atom is now associated with Zn 2+ (i.e., the electron is now transferred to the conduction band of ZIF-8 which has a theoretical bandgap energy of Eg = 5.5 eV) to keep an overall charge balance. It is thought that this transition process would require a localized structural distortion, but this can be done easily as the structure of ZIFs is not rigid. Apparently, there are two main pathways for H atom transport: (i) intra-linker: a proton moves within an imidazolate linker (Supplementary Fig. 29; 4 the diffusion distance is ca. 2.2 Å); and (ii) inter-linkers: a proton transports between two imidazolate linkers ( Supplementary Fig. 29; the distance of two adjacent imidazolate linkers is ca. 3.1 Å). We anticipate that future investigations by theoretical chemists will resolve the actual transport processes of H atoms. Figure 32. XRD patterns and the corresponding TEM images of the spent catalysts after CO2 hydrogenation reaction (220 o C and 30 bar). Scale bars in (b-k) are 40 nm, 10 nm, 100 nm, 50 nm, 20 nm, 10 nm, 100 nm, 10 nm, 10 nm, and 5 nm, respectively.  Table 1. The surface atomic ratios of Zn/Co on the surface of some representative Matryoshka-type (ZIFs@)n1ZIFs nanocubes (n = 4, 5, 6, and 7).
Comments: As shown in Supplementary Table 1, no Co signal was found if the outmost layer of the composite is ZIF-8. Likewise, no Zn signal was observed if the outmost layer is ZIF-67. Therefore, the epitaxial growth of a new ZIFs layer is complete, indicating that the ZIFs in the shell could fully cover the cores.