Researchers have long been interested in replicating the reactivity that occurs in photosynthetic organisms. To mimic the long-lived charge separations characteristic of the reaction center in photosynthesis, researchers have applied the Marcus theory to design synthetic multistep electron-transfer (ET) systems. In this Account, we describe our recent research on the rational design of ET control systems, based on models of the photosynthetic reaction center that rely on the Marcus theory of ET. The key to obtaining a long-lived charge separation is the careful choice of electron donors and acceptors that have small reorganization energies of ET. In these cases, the driving force of back ET is located in the Marcus inverted region, where the lifetime of the charge-separated state lengthens as the driving force of back ET increases. We chose porphyrins as electron donors and fullerenes as electron acceptors, both of which have small ET reorganization energies. By linking electron donor porphyrins and electron acceptor fullerenes at appropriate distances, we achieved charge-separated states with long lifetimes. We could further lengthen the lifetimes of charge-separated states by mixing a variety of components, such as a terminal electron donor, an electron mediator, and an electron acceptor, mimicking both the photosynthetic reaction center and the multistep photoinduced ET that occurs there. However, each step in multistep ET loses a fraction of the initial excitation energy during the long-distance charge separation. To overcome this drawback in multistep ET systems, we used designed new systems where we could finely control the redox potentials and the geometry of simple donor-acceptor dyads. These modifications resulted in a small ET reorganization energy and a high-lying triplet excited state. Our most successful example, 9-mesityl-10-methylacridinium ion (Acr(+)-Mes), can undergo a fast photoinduced ET from the mesityl (Mes) moiety to the singlet excited state of the acridinium ion moiety (Acr(+)) with extremely slow back ET. The high-energy triplet charge-separated state is located deep in the Marcus inverted region, and we have detected the structural changes during the photoinduced ET in this system using X-ray crystallography. To increase the efficiency of both the light-harvesting and photoinduced ET, we assembled the Acr(+)-Mes dyads on gold nanoparticles to bring them in closer proximity to one another. We can also incorporate Acr(+)-Mes molecules within nanosized mesoporous silica-alumina. In contrast to the densely assembled dyads on gold nanoparticles, each Acr(+)-Mes molecule in silica-alumina is isolated in the mesopore, which inhibits the bimolecular back ET and leads to longer lifetimes in solution at room temperature than the natural photosynthetic reaction center. Acr(+)-Mes and related compounds act as excellent organic photocatalysts and facilitate a variety of reactions such as oxygenation, bromination, carbon-carbon bond formation, and hydrogen evolution reactions.

The photocatalytic water oxidation to evolve O 2 was performed by photoirradiation (l > 420 nm) of an aqueous solution containing [Ru(bpy) 3 ] 2+ (bpy ¼ 2,2 0 -bipyridine), Na 2 S 2 O 8 and water-soluble cobalt complexes with various organic ligands as precatalysts in the pH range of 6.0-10. The turnover numbers (TONs) based on the amount of Co for the photocatalytic O 2 evolution with [Co II (Me 6 tren)(OH 2 )] 2+ (1) and [Co III (Cp * )(bpy)(OH 2 )] 2+ (2) [Me 6 tren ¼ tris(N,N 0 -dimethylaminoethyl) amine, Cp * ¼ h 5 -pentamethylcyclopentadienyl] at pH 9.0 reached 420 and 320, respectively. The evolved O 2 yield increased in proportion to concentrations of precatalysts 1 and 2 up to 0.10 mM. However, the O 2 yield dramatically decreased when the concentration of precatalysts 1 and 2 exceeded 0.10 mM. When the concentration of Na 2 S 2 O 8 was increased from 10 mM to 50 mM, CO 2 evolution was observed during the photocatalytic water oxidation. These results indicate that a part of the organic ligands of 1 and 2 were oxidized to evolve CO 2 during the photocatalytic reaction. The degradation of complex 2 under photocatalytic conditions and the oxidation of Me 6 tren ligand of 1 by [Ru(bpy) 3 ] 3+ were confirmed by 1 H NMR measurements. Dynamic light scattering (DLS) experiments indicate the formation of particles with diameters of around 20 AE 10 nm and 200 AE 100 nm during the photocatalytic water oxidation with 1 and 2, respectively. The particle sizes determined by DLS agreed with those of the secondary particles observed by TEM. The XPS measurements of the formed particles suggest that the surface of the particles is covered with cobalt hydroxides, which could be converted to active species containing high-valent cobalt ions during the photocatalytic water oxidation. The recovered nanoparticles produced from 1 act as a robust catalyst for the photocatalytic water oxidation.

A heterodinuclear iridium-ruthenium complex [Ir(III)(Cp*)(H(2)O)(bpm)Ru(II)(bpy)(2)](SO(4))(2) {1(SO(4))(2), Cp* = eta(5)-pentamethylcyclopentadienyl, bpm = 2,2'-bipyrimidine, bpy = 2,2'-bipyridine} acts as the most effective catalyst for selective production of hydrogen from formic acid in an aqueous solution at ambient temperature among catalysts reported so far. An unusually large tunneling effect was observed for the first time for the catalytic hydrogen production in H(2)O vs D(2)O.