Integrating molecular photon upconversion via triplet-triplet annihilation (TTA-UC) driectly into a solar cell offers a means of harnessing sub-bandgap, near infrared (NIR) photons and surpassing the Shockley-Quessier limit. However, all...

No abstract

Incorporating photon upconversion, via triplet–triplet annihilation (TTA-UC), directly into a solar cell is an intriguing strategy for harnessing sub-band gap photons and surpassing the Shockley–Queisser limit. A majority of TTA-UC solar cells to date rely on difficult to synthesize and expensive platinum and/or palladium porphyrin sensitizers. Here, we present, as far as we know, the first TTA-UC solar cell that integrates quantum dot (QD) sensitizers directly into the photocurrent generation mechanism. The photoanodes are composed of a nanocrystalline TiO2 substrate, 4,4′-(anthracene-9,10-diyl)­bis­(4,1-phenylene)­diphosphonic acid (A) as the annihilator molecule, and CdSe QDs as the sensitizer in an inorganic–organic-inorganic layered architecture (TiO2-A-QD). The TiO2-A-QD devices generate a photocurrent that is more than 1.4 times the sum of its parts and does so via a TTA-UC mechanism as demonstrated by intensity dependence, IPCE, and spectroscopic measurements. The maximum efficiency onset threshold (i.e., the I th value) of 0.9 mW cm–2 (1.9 × 1015 ex s–1 cm–2) is below solar flux and on par with some of the lowest I th values reported to date. However, the J sc for the QD sensitized device (29 μA cm–2) is still lower than comparable molecular sensitized devices (185 μA cm–2) due in part to lower sensitizer surface loadings, less than unity energy transfer yields (∼40–80%), slow regeneration kinetics, and competitive QD* quenching by the CoII/III(phen)3 redox mediator. Nonetheless these results demonstrate that multilayer assemblies containing QD sensitizers is an effective strategy to harness UC in a TTA-UC solar cell.

Four Cu(I) bis(phenanthroline) photosensitizers formulated from a new ligand structural motif (Cu1−Cu4) coded according to their 2,9-substituents were synthesized, structurally characterized, and fully evaluated using steady-state and time-resolved absorption and photoluminescence (PL) measurements as well as electrochemistry. The 2,9-disubstituted-3,4,7,8-tetramethyl-1,10-phenanthroline ligands feature the following six-membered ring systems prepared through photochemical synthesis: 4,4-dimethylcyclohexyl (1), tetrahydro-2H-pyran-4-yl (2), tetrahydro-2H-thiopyran-4-yl (3), and 4,4-difluorocyclohexyl (4). Universally, these Cu(I) metal-to-ligand charge transfer (MLCT) chromophores display excited-state lifetimes on the microsecond time scale at room temperature, including the three longest-lived homoleptic cuprous phenanthroline excited states measured to date in de-aerated CH 2 Cl 2 , τ = 2.5−4.3 μs. This series of molecules also feature high PL quantum efficiencies (Φ PL = 5.3−12% in CH 2 Cl 2 ). Temperature-dependent PL lifetime experiments confirmed that all these molecules exhibit reverse intersystem crossing and display thermally activated delayed PL from a 1 MLCT excited state lying slightly above the 3 MLCT state, 1050−1490 cm −1 . Ultrafast and conventional transient absorption measurements confirmed that the PL originates from the MLCT excited state, which remains sterically arrested, preventing an excessive flattening distortion even when dissolved in Lewis basic CH 3 CN. Combined PL and electrochemical data provided evidence that Cu1−Cu4 are highly potent photoreductants (E ox * = −1.73 to −1.62 V vs Fc +/0 in CH 3 CN), whose potentials are altered solely based on which heteroatoms or substituents are resident on the 2,9-appended ring derivatives. It is proposed that longrange electronic inductive effects are responsible for the systematic modulation observed in the PL spectra, excited-state lifetimes, and the ground state absorption spectra and redox potentials. Cu1−Cu4 quantitatively follow the energy gap law, correlating well with structurally related cuprous phenanthrolines and are also shown to triplet photosensitize the excited states of 9,10diphenylanthracene with bimolecular rate constants ranging from 1.61 to 2.82 × 10 8 M −1 s −1 . The ability to tailor both photophysical and electrochemical properties using long-range inductive effects imposed by the 2,9-ring platforms advocates new directions for future MLCT chromophore discovery.

A sonochemical-based hydrosilylation method was employed to covalently attach a rhenium tricarbonyl phenanthroline complex to silicon(111). fac-Re(5-(p-Styrene)-phen)(CO)3Cl (5-(p-styrene)-phen = 5-(4-vinylphenyl)-1,10-phenanthroline) was reacted with hydrogen-terminated silicon(111) in an ultrasonic bath to generate a hybrid photoelectrode. Subsequent reaction with 1-hexene enabled functionalization of remaining atop Si sites. Attenuated total reflectance–Fourier transform infrared spectroscopy confirms attachment of the organometallic complex to silicon without degradation of the organometallic core, supporting hydrosilylation as a strategy for installing coordination complexes that retain their molecular integrity. Detection of Re(I) and nitrogen by X-ray photoelectron spectroscopy (XPS) further support immobilization of fac-Re(5-(p-styrene)-phen)(CO)3Cl. Cyclic voltammetry and electrochemical impedance spectroscopy under white light illumination indicate that fac-Re(5-(p-styrene)-phen)(CO)3Cl undergoes two electron reductions. Mott–Schottky analysis indicates that the flat band potential is 239 mV more positive for p-Si(111) co-functionalized with both fac-Re(5-(p-styrene)-phen)(CO)3Cl and 1-hexene than when functionalized with 1-hexene alone. XPS, ultraviolet photoelectron spectroscopy, and Mott–Schottky analysis show that functionalization with fac-Re(5-(p-styrene)-phen)(CO)3Cl and 1-hexene introduces a negative interfacial dipole, facilitating reductive photoelectrochemistry.