The synthesis of two new heteroleptic Cu(I) photosensitizers (PS), [Cu(Xantphos)(NN)]PF (NN = biq = 2,2'-biquinoline, dmebiq = 2,2'-biquinoline-4,4'-dimethyl ester; Xantphos = 4,5-bis(diphenylphosphino)-9,9-dimethylxanthene), along with the associated structural, photophysical, and electrochemical properties, are described. The biquinoline diimine ligand extends the PS light absorbing properties into the visible with a maximum absorption at 455 and 505 nm for NN = biq and dmebiq, respectively, in CHCl solvent. Following photoexcitation, both Cu(I) PS are emissive at low energy, albeit displaying stark differences in their excited state lifetimes (τ = 410 ± 5 (biq) and 44 ± 4 ns (dmebiq)). Cyclic voltammetry indicates a Cu-based HOMO and NN-based LUMO for both complexes, whereby the methyl ester substituents stabilize the LUMO within [Cu(Xantphos)(dmebiq)] by ∼0.37 V compared to the unsubstituted analogue. When combined with HO, N,N-dimethylaniline (DMA) electron donor, and cis-[Rh(NN)Cl]PF (NN = Mebpy = 4,4'-dimethyl-2,2'-bipyridine, bpy = 2,2'-bipyridine, dmebpy = 2,2'-bipyridine-4,4'-dimethyl ester) water reduction catalysts (WRC), photocatalytic H evolution is only observed using the [Cu(Xantphos)(biq)] PS. Furthermore, the choice of cis-[Rh(NN)Cl] WRC strongly affects the catalytic activity with turnover numbers (TON = mol H per mol Rh catalyst) of 25 ± 3, 22 ± 1, and 43 ± 3 for NN = Mebpy, bpy, and dmebpy, respectively. This work illustrates how ligand modification to carefully tune the PS light absorbing, excited state, and redox-active properties, along with the WRC redox potentials, can have a profound impact on the photoinduced intermolecular electron transfer between components and the subsequent catalytic activity.

Two diimine-bridged Ru(II),Mn(I) complexes with a [(bpy)Ru(BL)Mn(CO)Br] architecture, where bpy = 2,2'-bipyridine and BL = 2,3-bis(2-pyridyl)pyrazine (dpp; Ru(dpp)Mn) or 2,2'-bipyrimidine (bpm; Ru(bpm)Mn), were designed to both dissociate multiple equivalents of CO and produce O when irradiated with visible light. Analysis of the complexes by Fourier transform infrared (FTIR) spectroscopy and cyclic voltammetry suggest a stronger π-accepting ability for bpm compared to that of dpp. Both complexes absorb light throughout the UV and visible regions with lowest energy absorption bands comprising overlapping Ru(dπ)→BL(π*) and Mn(dπ)→BL(π*) singlet metal-to-ligand charge transfer (MLCT) and Br(p)→dpp(π*) singlet halide-to-ligand charge transfer (XLCT) transitions. This lowest energy band is centered at 510 nm (ε = 12 000 Mcm) for Ru(dpp)Mn and 553 nm (ε = 3240 Mcm) for Ru(bpm)Mn, and the absorption band extends to nearly 700 nm in each case. Irradiation with visible light (both 470 and 627 nm) releases all three CO ligands, as observed by a combination of UV-vis, FTIR, and gas chromatography. The exchange of the first CO ligand with a solvent molecule occurs more efficiently for Ru(dpp)Mn (Φ = 0.22 ± 0.03 in HO; 0.37 ± 0.06 in CHCN) than for Ru(bpm)Mn (Φ = 0.049 ± 0.008 in HO and 0.16 ± 0.03 in CHCN), and the CO dissociation efficiency is unaffected by irradiation wavelength. The differences between Ru(dpp)Mn and Ru(bpm)Mn are proposed to result from the variation in electron density distribution across each formally reduced BL in the Mn(dπ)→BL(π*) MLCT excited state based on the nature of BL. Exhaustive photolysis causes the decomplexation of oxidized Mn(II), and the resulting [(bpy)Ru(BL)] complexes produce O with quantum yields (Φ) of 0.37 ± 0.03 and 0.16 ± 0.01 for Ru(dpp) and Ru(bpm), respectively, with 460 nm irradiation. This bimetallic architecture presents the opportunity to use visible light to codeliver both CO and O, both of which have biological relevance in photoactivated therapeutics, with spatiotemporal control.