Hierarchical porous photosensitizers with efficient photooxidation

The study addresses the challenge of designing highly efficient and stable photosensitizers (PSs) for converting light into chemical energy, particularly for photo-oxidation processes that generate singlet oxygen (1O2) or reactive oxygen species (ROS). Organic PSs offer advantages such as high extinction coefficients, tunable energy levels, and ease of handling, yet suffer from stability issues, deactivation, leaching, and poor recyclability during oxidative reactions. Efficient and sustained ROS generation requires rapid transport of active species away from catalytic centers to mitigate self-deactivation. Inspired by biological pores that enable selective capture, transport, and release through gating, porous materials decorated with PSs have been explored. However, PSs on porous surfaces tend to aggregate, reducing activity over time, and embedding PSs within pores can block O2 permeation and yield unreliable catalysis. Hydrogen-bonded macrocycles provide uniform pore sizes and strong recognition between hydrogen donors and acceptors. The authors hypothesize that co-assembling a hydrogen-donating PS with a hydrogen-accepting macrocyclic unit can yield hierarchically ordered porous architectures that prevent chromophore aggregation while maintaining accessible pathways for O2 transport, thereby enhancing photo-oxidation efficiency, stability, and selectivity.

Prior strategies to improve PS performance include heavy-atom effects, minimizing singlet–triplet gaps, and suppressing non-radiative decay. Nevertheless, oxidative species can degrade PSs, causing instability, deactivation, leaching, and low recyclability. Porous architectures are attractive for facilitating O2 transport, as seen in biological systems (e.g., uniporters with open–closed pore gating). Artificial porous materials with surface-bound PSs provide large surface areas but suffer from chromophore aggregation and declining activity. Embedding PSs into porous backbones complicates morphology control and can block O2 diffusion. Hydrogen-bonded macrocycles and laminates derived from C3-symmetric amide units form well-defined pores via lateral cross-linking and exhibit selective recognition for hydrogen donors/acceptors. This suggests a route to integrate PSs into ordered porous assemblies via specific hydrogen bonding to disperse chromophores and preserve pore accessibility.

Design and synthesis: A thienyl-substituted diketopyrrolopyrrole (TDPP) photosensitizer (compound 2) was synthesized by substituting one edge of a C3-symmetric N,N',N"-tris(3-methylpyridyl) trimesic amide (compound 1) to create a hydrogen-donating PS building block. Compound identities were confirmed by 1H and 13C NMR spectroscopy and TIMS-TOF mass spectrometry. Self-assembly characterization: Self-association of 1 and 2 in acetone was probed by vapor pressure osmometry (VPO) to estimate aggregate weights (indicative of dimers), and FT-IR to identify hydrogen-bonded N–H and N=C–H red-shifts. AFM imaged nanosheets from acetone solutions (3 mM), revealing thicknesses and lateral dimensions. X-ray diffraction (SAXS/WAXS) of aggregates deposited on membranes elucidated periodicities and packing (e.g., herringbone ordering, interdigitated monolayers for 1; 3D orthorhombic structure for 2). Co-assembly studies: Equimolar mixtures of 1 and 2 in acetone were examined by 1H NMR for hydrogen-bonded N–H resonances and chemical shift perturbations of pyridine and amide protons; NOESY established proximity between thiophene (2) and pyridine (1), indicating intercalation via hydrogen bonding. UV–vis and fluorescence titrations monitored spectral shifts upon adding 1 to 2, evidencing a transition from H-type to looser J-type aggregates and emission changes at higher 1 loadings (50–83 mol%). Structural analysis of porous layers: 2D XRD of co-assemblies captured retention of laminate integrity with altered lattice constants upon guest 1 addition. HR-TEM with negative staining (uranyl acetate) and selected-area electron diffraction visualized in-plane ordered pores: base-centered rectangularly perforated layers (RPL) at ~1:1 co-assembly (interdomain spacings ~1.2 and 2.7 nm) and hexagonally perforated layers (HPL) at ~1:5 (pore distance ~1.5 nm). VPO provided molecular weights of co-assembled porous units, informing stoichiometries (RPL unit from trimeric 1 and 2; HPL unit from pentameric 1 and monomeric 2). Molecular dynamics simulations of six-molecule hydrogen-bonded clusters modeled bilayer cyclic pores, matching experimental lattice parameters. Association constants: NMR dilution experiments (0.1–3.0 mM) yielded self-association constants K2 in acetone: 3264 M−1 for 1 and 2098 M−1 for 2, quantifying hydrogen-bonding propensities. Framework formation: Perforated laminates were cross-linked via pyridine groups using 1,4-bis(bromomethyl)benzene (BBMB) to form hydrogen-bonded porous frameworks HOF1 (from 1:1 co-assembly) and HOF2 (from 1:5 co-assembly). PXRD verified periodic diffraction (2θ=6.8°) and retention of perforated structures. DSC heating/cooling scans characterized thermal transitions and phase behavior. Porosity and adsorption: Nitrogen sorption activation proved challenging due to dynamic hydrogen bonding; permeability was assessed via benzaldehyde adsorption from water. Uptakes and Langmuir isotherm analysis provided monolayer capacities and calculated surface areas (S0). Photosensitization and photocatalysis: Singlet oxygen generation was quantified by ABDA bleaching under xenon lamp (AM 1.5 G, 100 mW cm−2) using equal PS contents (0.1 μmol), comparing 2, RPL, and HPL (measuring time to degrade 100 nmol ABDA and photostability over 4 h). Photocatalytic degradation tests of pyrene, 1-bromopyrene (BP), 1-nitropyrene, 1-aminopyrene, and 1-hydroxypyrene (HP) were conducted in H2O:acetone (1:1 v/v) suspensions of 2, HOF1, and HOF2 with equal PS content, monitoring absorbance changes. Selectivity tests degraded HP in the presence of varying BP concentrations, quantifying degradation ratios after 30 min irradiation. A synthetic application purified a low-conversion (60%) PBr3 bromination of pyrenol by post-reaction irradiation with 2, HOF1, or HOF2 without additional purification, evaluating residual HP and BP yields. Instrumentation: TEM (JEM-2010HR, JEM-ARM200P at 200 kV), AFM (tapping mode on mica), 400 MHz NMR for titrations and NOESY, UV–vis for adsorption and photocatalysis monitoring.

• Co-assembly via specific hydrogen bonding between hydrogen-acceptor macrocycle 1 and hydrogen-donor PS 2 transforms dense chromophore stacks into uniformly perforated laminates: rectangularly perforated layers (RPL) at ~1:1 and hexagonally perforated layers (HPL) at ~1:5 ratios. XRD and TEM confirm in-plane ordered pores with lattice parameters matching simulations (RPL: ~1.3 and ~1.8 nm; HPL: pore spacing ~1.4–1.5 nm).
• Self-association constants (acetone): K2(1)=3264 M−1; K2(2)=2098 M−1, rationalizing donor–acceptor exchange and porous structure regulation.
• Singlet oxygen generation (ABDA bleaching, 0.1 μmol PS, 100 nmol ABDA, AM 1.5 G, 100 mW cm−2): 2 requires 16.5 min; RPL 7 min; HPL 4 min, demonstrating accelerated 1O2 generation by porous architectures that prevent chromophore aggregation.
• Photostability under continuous AM 1.5 G irradiation for 4 h: 2 significantly quenched; RPL and HPL retain ~80% and ~90% of initial 1O2 generation efficiency, respectively.
• Cross-linked frameworks (HOF1 from 1:1; HOF2 from 1:5) preserve perforation: PXRD shows periodic 2θ=6.8°. DSC reveals thermal transitions (HOF1: 81 °C first heating, additional at 102 °C on second; HOF2: 127 °C first, depression at 75 °C on second), consistent with donor–acceptor phase behavior.
• Permeability/adsorption: Benzaldehyde uptake—HOF1: 151 mg g−1; HOF2: 435 mg g−1. Langmuir-derived surface areas S0—HOF1: ~122.0 m2 g−1; HOF2: ~348.8 m2 g−1.
• Photocatalytic degradation of pyrene derivatives (equal PS content): For pyrene, 2 degrades 35% in 120 min, whereas HOF1 and HOF2 achieve much faster degradation (70 and 50 min, respectively). 1-bromopyrene shows negligible degradation with 2, but HOF1 and HOF2 degrade 15% and 17% in 1 h. Electron-rich substrates show rapid, complete degradation: HOF1—aminopyrene 80 min, hydroxypyrene 90 min; HOF2—aminopyrene 40 min, hydroxypyrene 50 min.
• Selectivity: In mixtures of HP and BP, HOF2 selectively degrades HP to 99% within 30 min while minimally affecting BP (≈2% at 5 mg L−1 BP; ≈5% at 45 mg L−1 BP), demonstrating strong selectivity toward electron-rich HP.
• Synthetic purification: After PBr3 bromination of pyrenol (60% conversion), post-irradiation with 2 or HOF1 undesirably degrades BP product by ~39% and ~20% while removing HP. Using HOF2 yields BP up to 96% with no HP residue, enabling purification without post-processing.

The findings validate that precise hydrogen-bond recognition between a hydrogen-accepting macrocycle and a hydrogen-donating TDPP photosensitizer directs the formation of hierarchically ordered porous laminates. These architectures disperse chromophores, suppressing π–π aggregation, and create accessible transport pathways for O2. As donor–acceptor stoichiometry shifts, 2D laminates inflate into rectangular and then hexagonal perforated layers, tuning pore geometry and PS isolation. The resulting porous HOFs deliver faster singlet oxygen generation and enhanced photostability compared with aggregated PSs, directly addressing the challenges of deactivation and declining activity. The improved mass transport and stabilized active sites translate into significantly higher photocatalytic degradation rates for aromatic wastes, with HOF2 outperforming HOF1 due to more optimized pore organization and greater chromophore isolation. The frameworks’ selective recognition of electron-rich substrates enables targeted degradation in mixtures, culminating in a practical purification step for aryl bromination that eliminates additional processing. Overall, the integrative effect of ordered porosity and controlled chromophore dispersion provides a robust solution to the transport and stability limitations in PS-based photocatalysis.

This work introduces a hydrogen-bond-driven co-assembly strategy to construct hierarchically ordered porous photosensitizers. By pairing a hydrogen-accepting macrocycle (1) with a hydrogen-donating TDPP chromophore (2), the authors achieve rectangularly and hexagonally perforated laminates that, upon cross-linking, form stable HOFs (HOF1 and HOF2). These materials exhibit rapid and sustained singlet oxygen generation, superior photostability, and markedly improved photocatalytic performance, including selective degradation of electron-rich aromatics and streamlined purification of aryl-bromination products. The approach provides a generalizable path for designing porous PS systems that balance chromophore isolation with mass transport. Future research can explore expanding chromophore types, tuning pore sizes and surface chemistries for substrate specificity, integrating stimuli-responsive gating, and applying the strategy to diverse photocatalytic reactions and membrane-based separations.

The HOFs show difficulty in nitrogen sorption activation due to the dynamic nature of hydrogen bonding, which may complicate conventional surface area assessments and gas-phase adsorption measurements. Thermal analyses indicate phase behavior changes between heating cycles, suggesting possible donor–acceptor phase separation or rearrangements under thermal stimuli. Photocatalytic evaluations were performed under specific solvent systems (H2O:acetone) and simulated solar conditions; broader operational windows (e.g., purely aqueous media, different light spectra) and long-term cycling tests beyond 4 h would further establish generalizability and durability.