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.

Soluble, redox-active, organic materials hold promise as charge-storage species for flow batteries; however, their stability during extended operation remains a key challenge. While a number of spectroscopic and electrochemical techniques are currently used to probe these complex and often ill-defined decay pathways, each technique has limitations, including accessibility and direct evaluation of practical electrolytes without preparatory steps. Here, we use microelectrode voltammetry to directly observe nonaqueous flow battery electrolytes, simultaneously identifying the rate of charged materials decay (reversible material loss) and total material decay (irreversible material loss). We validate this technique using ferrocene as a stable model redox couple, examine and address sources of error, and finally, demonstrate its capability by assessing the decay of a well-studied and moderately-stable substituted dialkoxybenzene [2,5-di-tert-butyl-1,4-bis(2methoxyethoxy)benzene]. These results suggest that microelectrodes may have utility for rapid assessment of redox electrolyte state-of-charge and state-of-health, both in-operando and postmortem.

The lack of suitable membranes for nonaqueous electrolytes limits cell capacity and cycle lifetime in organic redox flow cells. Using soluble, stable materials, we sought to compare the best performance that could be achieved with commercially available microporous separators and ion-selective membranes. We use organic species with proven stability to avoid deconvoluting capacity fade due to crossover and/or cell imbalance from materials degradation. We found a trade-off between lifetime and coulombic efficiency: non-selective separators achieve more stable performance but suffer from low coulombic efficiencies, while ion-selective membranes achieve high coulombic efficiencies but experience capacity loss over time. When electrolytes are pre-mixed prior to cycling, coulombic efficiency remains high, but capacity is lost due to cell imbalance, which can be recovered by electrolyte rebalancing. The results of this study highlight the potential for gains in nonaqueous cell performance that may be enabled by suitable membranes. File list (2) download file view on ChemRxiv UK-MIT04 MS for submission.pdf (1.45 MiB) download file view on ChemRxiv UK-MIT04 SI for submission.pdf (3.59 MiB)

In an effort to further develop earth-abundant photosensitizers for use in solar fuels production, a series of heteroleptic, visible-light-absorbing Cu(I) photosensitizers (PSs) of the design [Cu(PP)(NN)]PF 6 were prepared (PP = bidentate diphosphine, Xantphos = 4,5-bis-(diphenylphosphino)-9,9-dimethylxanthene, DPEphos = bis-[(2-diphenylphosphino)phenyl]ether, Nixantphos = 4,6-bis-(diphenylphosphino)-10H-phenoxazine; NN = bidentate diimine, biq = 2,2′-biquinoline). The Xantphos ligand was modified to impart flexibility (PP = DPEphos) and redox activity (PP = Nixantphos) to the diphosphine backbone to probe their influence on the light-absorbing properties and excited state dynamics using steady-state and time-resolved (nanosecond) spectroscopic techniques. Following visible light excitation to populate redox-active Cu(dπ) → biq(π*) metal-to-ligand charge transfer excited states (MLCT ES), intermolecular electron transfer via reductive quenching by the N,N-dimethylaniline (DMA) electron donor produces the one-electron reduced [Cu(PP)(biq − )] species. The strong reducing potential of [Cu(PP)(biq − )] PS in DMF solvent (E 1/2 (PS 0/− ) = −1.57, −1.62, and −1.58 V vs Fc + /Fc for PP = Xantphos, DPEphos, and Nixantphos, respectively) permits thermodynamically favorable ground state (GS) electron transfer to reduce the Rh(III)-based lowest unoccupied molecular orbitals (E p c (Rh III/II/I ) = −1.46 V) within the cis-[Rh III (Me 2 bpy) 2 Cl 2 ] + (Me 2 bpy = 4,4′-dimethyl-2,2′-bipyridine) water reduction catalyst (CAT). In combination with nanosecond transient absorption (nsTA) and GS electronic absorption measurements, the initial photophysical and photochemical processes are proposed that contribute to formation of the [Rh I (Me 2 bpy) 2 ] + intermediate for H 2 O reduction to H 2 . Subsequently, DMF solutions of concentration-matched Cu(I) PSs (360 μM) were photolyzed at λ irr = 447.5 nm in the presence of Rh(III) CAT (36 μM), DMA electron donor (1.55 M), and H 2 O substrate (1.09 M), whereby Cu(I) PSs with PP = Xantphos or Nixantphos produced comparable amounts of H 2 (48 ± 3 vs 33 ± 1 μmol) after 3 h. Conversely, the PP = DPEphos ligand produced significantly lower levels of H 2 (2.6 ± 0.6 μmol after 3 h), indicating that the structurally flexible DPEphos ligand hinders photosensitizer stability and results in more rapid decomposition to produce the photocatalytically inactive homoleptic [Cu(biq) 2 ] + complex.

No abstract