A host–guest semibiological photosynthesis system coupling artificial and natural enzymes for solar alcohol splitting

Development of a versatile, sustainable and efficient photosynthesis system that integrates intricate catalytic networks and energy modules at the same location is of considerable future value to energy transformation. In the present study, we develop a coenzyme-mediated supramolecular host-guest semibiological system that combines artificial and enzymatic catalysis for photocatalytic hydrogen evolution from alcohol dehydrogenation. This approach involves modification of the microenvironment of a dithiolene-embedded metal-organic cage to trap an organic dye and NADH molecule simultaneously, serving as a hydrogenase analogue to induce effective proton reduction inside the artificial host. This abiotic photocatalytic system is further embedded into the pocket of the alcohol dehydrogenase to couple enzymatic alcohol dehydrogenation. This host-guest approach allows in situ regeneration of NAD+/NADH couple to transfer protons and electrons between the two catalytic cycles, thereby paving a unique avenue for a synergic combination of abiotic and biotic synthetic sequences for photocatalytic fuel and chemical transformation.

catalyst, ADH and PNQ was added into an EtOH/H 2 O solution (v:v = 3:2, pH 4.5, 5.0 mL) containing NAD + with a magnetic stir bar. The flask was sealed with a septum and protected from air by Ar. The samples were irradiated by a 300 W Xenon lamp.
The reaction was maintained at 25ºC by using a water filter to absorb heat.
The generated hydrogen was characterized by GC 7890T instrument analysis using a 5 Å molecular sieve column, thermal conductivity detector, and argon used as carrier gas. The amount of hydrogen generated was determined by the external standard method 1 . The generated aldehyde was characterized by an Agilent 6890N GC using a FFAP capillary column, flame ionization detector, and nitrogen used as carrier gas. The amount of aldehyde generated was determined by the external standard (1) The quadratic equation can be rearranged to Supplementary Equation (5) The corresponding solution was shown in Supplementary Equation (6) The measurements are performed under the conditions where the emission intensity of the free host in such a concentration is F 0 ; after addition of a given amount [G 0 ], the fluorescent intensity can be defined by Supplementary Equation (7):

Single Crystal X-ray Crystallography
Intensities of the Co 3 TPS 2 were collected at 180 (2)  In the structural refinement of Co 3 TPS 2 , all the non-hydrogen atoms were refined anisotropically. Hydrogen atoms within the ligand backbones, a DMF molecule and three Et 4 N + cations were fixed geometrically at calculated distances and allowed to ride on the parent non-hydrogen atoms. To assist the stability of refinements, two amide groups on the ligands, a DMF molecule and two Et 4 N + cations were limited to the desired position with rational thermal parameters by several restrains. One methyl on an Et 4 N + cation, and a carbon atom and nitrogen atom on the DMF were disordered into two parts with s.o.f of each part being refined using free variables. The thermal parameters on adjacent atoms in all Et 4 N + cations, a DMF molecule and some parts of ligands were restrained to be similar. In addition, the SQUEEZE subroutine in PLATON was used for refinements 14 .
Despite rapid handling times and a low-temperature collection, the quality of data was less than ideal. We have tried our best to gain a better data, but failed. The reported herein are the best possible result obtained after an optimal exposure time.
The crystal is poorly diffracting due to the low equality of the crystal, which leads to an Alert level A in checkcif report.

General information for theoretical 'docking study'
Docking calculations were performed with the AutoDock program 4.2.
The Co 3 TPS 2 , PNQ and NADH were downloaded from the CCDC database.
The structure of enzyme alcohol dehydrogenase (PDB code: 5ENV) was downloaded from the PDB database. The cage Co 3 TPS 2 was used to perform the docking calculation after energy minimization. The models of the enzyme were refined by removing hydrogen atoms. Polar hydrogens were then added, followed by assignment of Kollman charges, fragmental volumes, and atomic solvation parameters to adhesive by means of AutoDock Tools. For the ligand, the molecule was refined by removing and subsequently adding hydrogen atoms in a similar manner to that for adhesive. Next, Gasteiger partial charges were assigned to the ligands, and nonpolar hydrogens were merged. All torsions were allowed to rotate during docking. The Lamarckian genetic algorithm was used to determine the appropriate binding positions, orientations, and conformations of the ligands. Default parameters were used, except for the number of generations which was set to 300. The blind docking strategy was used with a 50 Å × 78 Å × 114 Å grid box which ensured sufficient spaced to cover the entire surface of the enzyme. The Lamarckian genetic algorithm was chosen with default parameters except for the number of generations, which was set to 100 for more accurate docking results. The best docking mode of the