*3.3. Photocatalysis*

Given the extraordinarily large-scale utilization of solar energy, POM@MOF materials have become particularly attractive for their use in visible-light-driven photocatalytic reactions. In particular, their use as catalysts for proton reduction has received considerable attention in recent years. Within this context, Lin's group reported on the integration of the two required components, namely the photosensitizer [Ru(bpy)3] 2+ or [Ir(ppy)2(bpy)]+ and hydrogen evolution catalyst, into Zr-based MOFs to perform proton reduction [47,152]. For example, by using a one-pot self-assembly synthesis strategy, a transition metal Ni-based anionic POM was embedded into the highly cationic MOF by using a pre-functionalized [Ir(ppy)2(bpy)]+-derived dicarboxylate ligand [47]. In contrast to the homogeneous mixture of POM and the Ir-functionalized ligand, which only produced trace amounts of H2 (TON = 2), a TON of 1476 was observed for the hierarchically-organized POM@MOF assembly, which allowed facile electron transfer due to the proximity of the Ni4P2 to multiple photosensitizers in Ni4P2@MOF. Another simple strategy for accommodating antenna molecules into MOFs was demonstrated in the work of Wang et al., in which several transition metal-substituted (V, Ni, and Co) Wells-Dawson-type POM@MIL-101(Cr) were prepared using one-pot synthesis [154]. Prior to the photocatalytic evaluation, the cationic photosensitizer [Ru(bpy)3] 2+ was adsorbed onto the POM@MOF, for which they observed that the adsorption ability was significantly enhanced upon increasing the POM loading. The photocatalytic performance of the three POM@MOF frameworks was significantly higher in comparison to their homogeneous counterpart and remained nearly unchanged after three additional cycles. Another very nice work in which [Ru(bpy)3] 2+ was also used as a photosensitizer was reported by Li et al. [155]. They reported the first water-soluble supramolecular MOF, denoted as SMOF-1, which was built by a self-assembly process from the hexaarmed [Ru(bpy)3] <sup>2</sup>+-based precursor and cucurbit uril (CB) (see Figure 14). The resulting polycationic SMOF-1 exhibited only a weak gas adsorption ability, but was able to accommodate the bulky redox active [P2W18 O62] 6− anion. The hydrogen production of the resulting WD-POM@SMOF-1 was about four times higher than that of its heterogeneous system. More specifically, in acidic media and using methanol as a sacrificial electron donor, the TON and H2 production rate was 392 and 3.553 μmol g<sup>−</sup><sup>1</sup> <sup>h</sup>−1, respectively. The authors attributed this high activity to the unique one-cage-one-gues<sup>t</sup> encapsulation pattern, which allowed (i) a quick diffusion and close contact of the hydronium and methanol molecules and (ii) facile electron transfer from the excited [Ru(bpy)3]<sup>2</sup>+ to the WD-POM. In addition to this, the catalyst could be recovered by evaporation of the solvent and could be reused at least six times without a significant decrease in TON.

**Figure 14.** The building blocks used for the synthesis of the metal-cored supramolecular organic framework, SMOF-1. Reprinted with permission from [155]. Copyright (2016), Springer Nature.

It is, however, important to note here that in the previous studies, the high-cost noble [Ru(bpy)3]<sup>2</sup>+ and [Ir(ppy)2(bpy)]+ have been used as photosensitizers. The first noble metal-free photoactive POM@MOF catalyst was reported by Dolbecq et al. [145]. In this work, a redox active Co-based POM was embedded in a light-harvesting porphyrinic MOF, denoted as MOF-545, for the visible-light-driven oxidation of water (Figure 15) or, in other words, both the photosensitizer and the catalyst were incorporated into the same porous material. The authors observed that the O2 production started upon exposure to light and increased linearly over time, until a plateau was reached after 1 h of catalysis. The authors stated that the unique activity of this "three in one" photoactive catalyst was the result of (i) the immobilization of the porphyrin ligand in the MOF, which increased its oxidizing power, and (ii) the confinement of the Co-POM in the pores of MOF-545, which resulted in an increased stabilization of the POM catalytic sites. Nevertheless, the reuse of this POM@MOF catalyst was hampered due to a partial loss of the powder during centrifugation. To overcome this issue, the authors deposited a thin film of the latter POM@MOF on indium tin oxide, which served as a conducting support to allow better electronic transport, but also permitted easier reuse [146]. The films obtained through drop casting not only exhibited a significantly better performance in photocatalytic water oxidation (TON = 1600 and TOF = 0.467 s<sup>−</sup>1) in comparison to the POM@MOF in suspension (TON = 70 and TOF = 0.04 s<sup>−</sup>1), but also outperformed the previously homogeneous P2W18Co4-based photosystems (TON = 75) [162].

**Figure 15.** A fully noble metal-free POM@MOF catalyst for the photocatalytic oxidation of water. Reprinted with permission from [145]. Copyright (2018), American Chemistry Society.

However, it is important to note here that in the previously presented studies, a sacrificial donor or acceptor was required for the photocatalytic process. In a very recent work by Niu and co-workers, the assembling of a photosensitizer, electron donor, and acceptor into one single framework was reported [149]. For the synthesis of this Zn-based framework, the photosensitizer *<sup>N</sup>*,*N*-di(4-pyridyl)-1,4,5,8-naphthalenetetracarboxydiimide (DPNDI) was used as the organic ligand, while pyrrolidine-2-yl-imidazole and the [BW12O40]<sup>5</sup>− anion were introduced, respectively, as an electron donor and electron acceptor (see Figure 16). The resulting Zn-DPNDI-PYI catalyst was examined in the oxidative coupling of benzylamine, exhibiting a conversion of 99% after 16 h of reaction. This high activity was not only the result of the consecutive photo-induced electron transfer (conPET) process, but was also assigned to the long-lived charge separated state.

**Figure 16.** Zn-DPNDI-PYI as photocatalyst for the coupling of primary amines and oxidation of olefins with air under visible light. Reprinted from [149]. Copyright (2019), with permission from Elsevier.

#### **4. Summary and Outlook**

Metal–organic frameworks exhibiting well-defined cages, large surface areas, and a high thermal and chemical stability are excellent hosts for encapsulating polyoxometalates. More than 100 studies on such POM@MOF hybrids have appeared in the last decade. In this review, we mainly focused on the common synthetic aspects and catalytic applications of POM@MOF hybrids in organocatalysis, electrocatalysis, and photocatalysis. More specifically, the activity, recyclability, stability, and interesting synergetic functions of POM@MOF were discussed.

The size of the pores and the aperture of the pore windows are very critical parameters in the design of a POM@MOF. The embedding of POMs into MOFs not only allows the shortcomings of POMs to be overcome, but also ensures the use of the unique advantages of MOFs. The rise of POM@MOF systems is mainly attributed to the excellent dispersion and subsequent stability of the POM in the MOF host. The unique cages and windows and the tunable chemical environment of MOFs enable interesting interactions and synergic e ffects between POM and MOFs, thus creating excellent novel heterogeneous catalysts.

Although POM@MOF hybrid materials have made tremendous progress in recent years, many challenges still need to be addressed. First of all, the interaction between POMs and MOFs is often limited to weak electrostatic interactions, which can result in POM leaching during the catalysis. To this end, stronger interactions, such as covalent bonds between the MOF host and the encapsulated POM, would allow a further increase of the POM@MOF reusability in catalysis. Secondly, at this moment, there is still too much 'trial and error' involved to obtain a good control on the position and distribution of POMs inside MOF cages/channels. Thirdly, very little is known about the synergetic e ffects and electron transfer mechanism in catalytic reactions. To address this problem, theoretical calculations combined with in-situ and ex-situ characterization techniques would provide a better understanding of the synergetic e ffects and electron transfer mechanism. Finally, up until now, only some well-known archetypical POMs have been encapsulated in MOFs. New and innovative types of POMs (such as V-centered POMs) with a proven excellent performance in oxidation and photocatalytic reactions should be combined with MOFs to further enhance the application range of these hybrids. We have no doubt that several exciting new (catalytic) applications will be reported in the next months and years in this strongly growing field of research.

**Author Contributions:** Conceptualization and supervision: P.V.D.V. Writing-review and editing: J.S., S.A., Y.-Y.L. and K.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was financially supported by Ghent University BOF doctoral gran<sup>t</sup> 01D04318, the Research Foundation Flanders (FWO-Vlaanderen) gran<sup>t</sup> no. G000117N, the International S&T Cooperation Program of China (2016YFE0109800) and the China Scholarship Council (CSC).

**Conflicts of Interest:** The authors declare no conflicts of interest.
