The paper in Nature Communications is here: go.nature.com/2MO3GC7

Sustainable solvent usage is a topic of growing interest in both the research community and chemical industry due to a growing awareness of the impact of solvents on pollution, energy usage, and contributions to air quality and climate change. Solvent losses represent a major portion of organic pollution, and solvent removal represents a large proportion of process energy consumption. A vast majority of chemical transformations have to be performed in the presence of solvent, whereby, the irreplaceable role of solvents not only brings the equilibrium and reaction rates under control, but also can perturb the reaction, leading to divergent reaction outcomes. For example, a wide range of reactions exhibits unique performance, being operated in polar aprotic solvents, such as dimethyl sulfoxide (DMSO), 1-methyl-2-pyrrolidinone (NMP), and ionic liquids. However, separation of the product from these solvents is extremely complicated, and even worse, side reactions are often accompanied with this process. To counter these concerns, the development of alternatives that could mimic the solvation environment of these solvents to aid the accomplishment of those transformations, is highly desirable. We envisioned that creating a suitable solvation environment in close proximity to the active sites in heterogeneous catalysts and performing the reaction using a separate-friendly solvent instead, for mass and heat transfer, may offer prospective solutions to the aforementioned challenges pertaining to the separation of the products from high boiling point solvents or nonvolatile ionic liquids. 

However, to transfer such complex systems from liquid phase to the realm of solid state, it necessitates precise control of the spatial continuity and dynamic interactions between the immobilized partners. These challenges have sparked our interest in the exploration of porous organic polymers (POPs) to fulfill this task. This class of porous materials has quickly moved to the forefront of materials research due to their high internal surface areas, robustness under various chemical conditions, and facile chemical tunability. More significantly, the flexibility and modularity of polymer chains enable the provision of a solvent-like reaction environment, onto which catalytically active sites can be grafted in a predefined way that can be tuned with precision at the molecular scale. With these merits, we reasoned that such a catalytic system can be targeted by constructing judiciously designed reaction participants into porous organic polymers (Figure 1). 

To implement this strategy, we first modify the solvent moieties with specific functionality for the potential construction of porous frameworks. With respect to the polymer synthesis, our research group has established a powerful strategy by polymerization of the vinyl-functionalized monomers under solvothermal conditions, with the advantages of excellent functional group compatibility, adjustable composition, high yield, and tunable pore structures (Nano Today, 2009, 4, 135-142; J. Am. Chem. Soc., 2012, 134, 16948-16950; J. Am. Chem. Soc., 2015, 137, 5204-5209; Chem, 2016, 1, 628-639). Our initial step was to construct monomeric solvent analogues into highly porous frameworks. Various solvent moieties (NMP, DMSO, and imidazolium-type ionic liquid,) were functionalized with styryl to afford V-NMP, V-DMSO, and V-IL. The resulting functionalized solvent analogues were polymerized according to the method developed by our group, yielding corresponding porous solid solvents (PSSs) named as PSS-NMP, PSS-DMSO, and PSS-IL, respectively (Table 1).

Given the excellent spatial continuity of these solvent moieties in the three dimensional nanospace of PSSs, we therefore suggest that the solvation effect should also be considered in their solid state. To prove the existence of the solvation effect in the PSSs, we postulated that with the densely populated hydrogen bond acceptors, they should be effective in disrupting and breaking the intramolecular hydrogen-bonding network present in the carbohydrates, thus leading to the decreased crystallinity. To test this hypothesis, we mixed them with fructose, and mechanically grinded to facilitate their interaction. It is very interesting to find that after grinding fructose with PSSs, the XRD peaks associated with fructose disappeared completely in the resultant composites, whereas these peaks were retained in the composite of other porous polymers without a hydrogen bond acceptor, thereby indicating the retention of the solvation effect of their liquid analogues after being constructed into porous polymers.
 
After proving the solvation ability of PSSs, we proceeded to introduce catalytically active species with the sulfonic acid group chosen to demonstrate the possibility of PSSs as an alternative to the corresponding solvent in regulating the performance of the acid sites. To test the efficiency of the resultant catalysts, we evaluated their performance in the selective dehydration of fructose to produce 5-hydroxymethylfurfural (HMF). The choice of this transformation is based on the consideration that it is highly solvent dependent, which performs well in the presence of solvents like DMSO, NMP, or ionic liquids, however, it remains a substantial challenge to separate HMF in an energy-friendly manner from those solvents. To our delight, the resultant catalysts exhibit exceptional conversions of fructose to 5-hydroxymethylfurfural in THF as a readily separable solvent, far outperforming those grafted on more conventional supports and corresponding homogeneous catalysts. More importantly, the catalytic systems afforded HMF as a sole product, reaching a yield that exceeds the best performance of all reaction systems reported so far and even surpassing their performance in the corresponding desired solvent media. We therefore deduce that a high density of solvent moieties well-oriented around the acid sites could boost the synergistic effect in the confined nanospace, which leads to the high catalytic performance. Given the compositional tunability of the materials synthesis, this strategy may open a new avenue to design efficient and green chemical processes as well as intrigue important insights into the design and construction of sophisticated reaction environments to control the performance of active species in heterogeneous catalysts.

Figure 1 |Schematic illustration of reactant as represented by fructose in various systems. (left) solvent containing hydrogen bond acceptors. (right) a microenvironment with solvent-like behavior. 

ToC: