Two-photon-absorbing ruthenium complexes enable near infrared light-driven photocatalysis

The study addresses the challenge that conventional one-photon (UV/visible) photocatalysts suffer from limited light penetration, substrate absorption competition, and incompatibility with light-sensitive functionalities. Near-infrared (NIR) light could mitigate these issues, but single-photon NIR excitation generally yields low-energy, short-lived excited states. Two strategies to use lower-energy photons are triplet–triplet annihilation (TTA) upconversion and direct two-photon absorption (TPA). TTA often requires precise sensitizer–annihilator energy matching, limiting generality. Direct TPA has seen broad use in bioimaging and photodynamic therapy but is rarely applied to homogeneous photocatalysis due to typically small TPA cross sections of standard photocatalysts (e.g., [Ru(bpy)3]2+). Prior work indicates that strong intramolecular charge transfer and superpolarizability, especially in octupolar architectures and extended π-conjugation ligands, enhance TPA. Motivated by reports of bisstyryl-substituted bipyridine Ru complexes with much larger TPA cross sections, the authors hypothesize that ruthenium polypyridyl complexes bearing extended π-conjugated ligands can serve as effective NIR-driven TPA photosensitizers for both energy-transfer (e.g., 1O2 generation) and photoredox catalysis.

• Conventional photocatalysts (Ru, Ir complexes, organic dyes) rely on UV/visible excitation; NIR use is rare due to low-energy excited states.
• TTA upconversion enables NIR-driven catalysis but is constrained by sensitizer/annihilator matching and interactions; several demonstrations exist but with limitations.
• Direct TPA is established in bioimaging/PDT; however, typical photocatalysts have very small TPA cross sections (e.g., [Ru(bpy)3]2+ σ2 ~4.3 GM at 880 nm), limiting applicability. Few catalytic TPA systems require visible light and/or intense lasers.
• Design principles: strong donor–π–acceptor character and extended conjugation yield higher σ2; organic D–π–A dyes and transition metal complexes with elongated π systems have been explored.
• Octupolar bisstyryl-bpy metal complexes show superpolarizability and σ2 two orders of magnitude above [Ru(bpy)3]2+ in NIR, and have been used as PDT agents under 700–800 nm excitation. These precedents motivate testing analogous Ru complexes for homogeneous NIR photocatalysis using inexpensive LEDs (>700 nm), which had not been demonstrated via direct simultaneous two-photon absorption by a single molecule.

Synthesis:

• Prepared bisstyryl-bipyridine ligands: bpyvp-H, bpyvp-F, bpyvp-OMe via base-promoted condensation of 4,4'-dimethyl-2,2'-bipyridine with appropriate benzaldehydes in dry DMF, workup by precipitation and Soxhlet washing.
• Synthesized Ru complexes by refluxing Ru precursors with ligands and performing anion exchange with NH4PF6 to obtain: [Ru(bpy)2(bpyvp-H)]2+ (1), [Ru(bpy)(bpyvp-H)2]2+ (2), [Ru(bpyvp-H)3]2+ (3), [Ru(bpyvp-F)3]2+ (4), [Ru(bpyvp-OMe)3]2+ (5).

Characterization:

• UV–vis absorption and emission spectra collected; excitation spectra matched absorption confirming emissive species.
• Electrochemistry (CV/SWV) in DMF/0.1 M LiClO4 versus Fc+/0; observed one oxidation (+0.62 to +0.77 V) and three reductions (−1.62 to −2.12 V), all diffusion-controlled.
• Excited-state lifetimes (deaerated and aerated CH3CN) measured by time-resolved emission; lifetimes of 3, 4, 5: 431, 628, 877 ns (deaerated), ~130 ns in air, consistent with 1O2 quenching.
• Femtosecond transient absorption (pump–probe): excitation at 480 nm and at 800 nm (off-resonance one-photon). Identical transient features for both, and signal scaled linearly with the square of 800 nm power, confirming TPA populates the same MLCT excited state.

Computation:

• DFT (B3LYP) with SDD (Ru) and 6-31G* (H, C, N, O, F) in CPCM (CH3CN) for ground-state geometry and frontier orbitals. HOMOs mainly Ru 4d; LUMOs on bpyvp-type ligands; trends matched redox potentials.
• TD-DFT for singlet excited states, showing MLCT and mixed MLCT/ligand-centered charge transfer (LCCT) transitions. For 5: S1 ~551 nm (MLCT), strong S5 ~512 nm (MLCT), and mixed MLCT/LCCT in 350–450 nm region; EDDMs analyzed.

Photocatalysis under NIR LEDs:

• Light source: 740 nm LED (PR160L-740-C, 8.18 W) at room temperature unless noted; deaerated or O2-saturated as appropriate.
• Model 1O2 reaction: benzyl amine oxidative coupling to N-benzylidenebenzylamine; monitored yields vs time; controls (dark, no PS, Ar vs O2). Scavenger tests (ascorbic acid, D-mannitol, sodium azide) to probe oxidants. LED power dependence and penetration depth comparison between 456 nm and 740 nm using a row of tubes.
• Additional 1O2 reactions: thioanisole sulfoxidation; anthracene [4+2] Diels–Alder with O2; cyclooctene allylic hydroperoxidation; HMF oxidation to MA and HKPA.
• Photoredox reactions: reductive dehalogenation of phenacyl bromide with TEOA (sacrificial donor); redox-neutral C–H cyanation of tetrahydroisoquinoline with TsCN (no sacrificial reagents); Ni-cocatalyzed allylation of benzaldehydes with allyl acetate using DIPEA as sacrificial donor. Reaction workups by chromatography and product characterization by NMR.

Photophysics/electrochemistry and TPA validation:

• Increasing bpyvp content red-shifts MLCT absorption and increases extinction coefficient: e.g., 1 (465 nm), 2 (475 nm), 3 (∼477–488 nm band); 5 shows further red-shift and higher ε due to OMe donors.
• Emission maxima 660–673 nm across 1–5. Deaerated lifetimes: 3 = 431 ns, 4 = 628 ns, 5 = 877 ns; lifetimes drop to ~130 ns in air, indicating 1O2 formation.
• Electrochemistry: one oxidation from +0.62 (5) to +0.77 V vs Fc+/0 (1) and three reductions from −1.62 to −2.12 V, consistent with extended π systems lowering LUMOs.
• Transient absorption of 5 shows same spectral features for 480 nm and 800 nm excitation; 800 nm signal scales linearly with the square of excitation power (R2 = 0.999), confirming direct TPA to the same MLCT excited state. Similar behavior observed for other complexes.

1O2 energy-transfer catalysis (740 nm, O2):

• Benzyl amine oxidative coupling: performance increases with more bpyvp ligands. After 2 h: 1 gives 16% yield; near-complete conversion with 2 in 100 min and 3 in 50 min. Using 1 mol% PS, CH3CN, O2 atmosphere. Controls: light, O2, and PS all necessary; under air, 43% yield in 30 min vs 89% under O2. Complex 5 outperforms 3 (96% vs 89% in 30 min), attributed to longer lifetime and higher TPA cross section; 4 gives 42% in 30 min.
• Scavengers: Ascorbic acid (O2 scavenger) and D-mannitol (•OH scavenger) minimally affect yield; sodium azide (1O2 scavenger) drops yield from 65% to 5% (10 min), implicating 1O2 as primary oxidant.
• Power dependence is nonlinear with LED power, consistent with TPA kinetics. Penetration test: 456 nm LED affords ~90% yield only in first tube; 740 nm LED gives nearly 100% in first tube and 91–72% in tubes 2–4, with 26% even in tube 10, demonstrating deeper NIR penetration.
• Other 1O2 reactions with 5: thioanisole to sulfoxide 95% yield; anthracene [4+2] product 99%; cyclooctene hydroperoxide 91%.
• HMF oxidation with 5: 92% conversion in 5 h to maleic anhydride (55% yield) and 5-hydroxy-4-keto-2-pentenoic acid (37%).

Photoredox catalysis under 740 nm:

• Reductive hydrodehalogenation of phenacyl bromide with 0.2 mol% 5 and 10 eq TEOA in CH3CN (Ar): acetophenone 86% yield after 8 h. [Ru(bpy)3]2+ inactive at 740 nm.
• Redox-neutral C–H cyanation of tetrahydroisoquinoline with TsCN (Ar, no sacrificial donor): 96% yield after 24 h using 1 mol% 5; mechanism involves oxidative quenching of 5* by TsCN, Ru(III)-mediated amine oxidation, radical steps to iminium, and CN− addition.
• Ni-assisted allylation of benzaldehydes with allyl acetate using 5 (5 mol%), Ni(OTf)2 (15 mol%), phen (20 mol%), DIPEA, CH3CN, Ar, 72 h: yields 57–87% depending on substituent (lower for p-OMe due to decreased aldehyde electrophilicity). Proposed cycle involves two single-electron transfers between reduced Ru and Ni species.

Structure–activity trends:

• More bpyvp ligands generally enhance TPA photocatalysis (3 > 2 > 1), correlating with larger σ2.
• Electron-donating OMe substituents on terminal phenyls (5) improve performance over H (3) and F (4), likely via stronger intramolecular charge transfer and longer excited-state lifetime.

The findings demonstrate that ruthenium polypyridyl complexes bearing extended π-conjugated bisstyryl-bipyridine ligands can serve as effective two-photon-absorbing photosensitizers for homogeneous photocatalysis using inexpensive NIR LEDs (740 nm). Direct TPA was verified spectroscopically, and the same MLCT excited state is accessed under both one-photon (visible) and two-photon (NIR) excitation, enabling the complexes to operate analogously to conventional one-photon photocatalysts. The complexes efficiently mediate energy-transfer processes to generate singlet oxygen for diverse oxidations and enable electron-transfer photoredox reactions, including reductive dehalogenation, redox-neutral C–H cyanation, and Ni-assisted allylation. Performance correlates with electronic structure: increasing the number of bpyvp ligands and installing electron-donating substituents increase intramolecular charge transfer character, improve TPA cross section, and extend excited-state lifetime, boosting catalytic efficiency. The practical advantages of NIR light, notably deeper penetration and reduced competition with substrate absorption, were directly demonstrated, highlighting the potential of TPA-enabled NIR photocatalysis for scalable and substrate-diverse transformations.

Five Ru polypyridyl complexes with extended π-conjugated bisstyryl-bipyridine ligands exhibit robust NIR (740 nm) two-photon-activated photocatalysis, achieving both energy-transfer (1O2) and photoredox transformations under ambient conditions. Key contributions include: (i) experimental validation of direct TPA populating MLCT excited states under NIR excitation; (ii) broad catalytic scope spanning 1O2 oxidations, reductive and redox-neutral photoredox reactions, and metallaphotoredox C–C coupling; (iii) elucidation of structure–activity relationships showing that a higher count of bpyvp ligands and electron-donating para substituents (OMe) enhance TPA-driven activity, aided by longer excited-state lifetimes. The work provides design guidelines for improved TPA photosensitizers and establishes NIR-driven photocatalysis as a viable approach with practical advantages in penetration depth and substrate compatibility. Future efforts can focus on further increasing TPA cross sections through ligand engineering, expanding reaction scope, and optimizing reaction conditions for efficiency and scalability.