N-Heterocyclic carbene-based C-centered Au(I)-Ag(I) clusters with intense phosphorescence and organelle-selective translocation in cells

The study investigates how N-heterocyclic carbene (NHC) ligands, when integrated into carbon-centered hexagold clusters (CAu6) and further coordinated with Ag(I), modulate photophysical properties and cellular behavior. Prior CAu6 clusters with phosphine ligands showed solid-state luminescence but poor solution emission; bidentate pyridyl-phosphine ligands enabled solution phosphorescence with protective encapsulation. NHC ligands, being strong electron donors with high designability, have recently stabilized and tuned metal nanostructures. The hypothesis is that NHC ligation and heterometal incorporation (Ag) can synergistically enhance radiative decay, suppress non-radiative pathways, and enable organelle-selective cellular translocation. The goal is to design bidentate NHC–pyridyl ligands to build CAu6Ag2 clusters with high solution-phase phosphorescence and to elucidate their intracellular uptake routes and organelle specificity.

- Classical CAu6 clusters with phosphine ligands emit in the solid state but not in solution (Schmidbaur et al.). - Bidentate pyridyl-phosphine ligation to CAu6 with additional metal ions afforded heterometallic clusters with strong solution-phase red phosphorescence via full surface protection (Wang and co-workers). - Strategies to enhance cluster luminescence include alloying, supramolecular networking, surface hardening, and ligand electronic tuning. - NHC ligands are robust, strongly donating, and have stabilized metal surfaces, nanoparticles, and precise nanoclusters; NHC-protected Au clusters show enhanced stability and tunable photophysics. - Prior work on NHC-protected CAu6 showed that imidazolylidene vs benzimidazolylidene profoundly shifts emission (red-shift vs large blue-shift), indicating ligand structure strongly controls photophysics. - Heterometallic phosphine-protected CAu6Ag2 clusters have been used for specific cellular labeling (e.g., nucleolus), suggesting potential for bioimaging with ligand-controlled localization.

- Ligand design and synthesis: Four bidentate ligands were prepared, each comprising an NHC (imidazolylidene or benzimidazolylidene) tethered to a pyridyl donor: 1a (N-isopropyl-N-2-(5-methylpyridyl)benzimidazolylidene), 1b (N-isopropyl-N-2-pyridylbenzimidazolylidene), 1c (N-isopropyl-N-2-(5-methylpyridyl)imidazolylidene), 1d (N-isopropyl-N-2-pyridylimidazolylidene). Synthetic details for imidazolium/benzimidazolium salts (e.g., 1a-HI, 1b-HI, 1c-HI) included base-mediated N-arylation with 2-bromopyridines, followed by alkylation with 2-iodopropane and crystallization. - Cluster synthesis: CAu6 clusters [(C)(Au-L)](BF4)2 (2a–d) were prepared by generating the carbene in situ from 1a–d with tht-AuCl/K2CO3, followed by anion exchange (NaBF4), base (KOH), AgBF4/H2O treatment, and stabilization (Me3SiCHN2/NEt3). Single crystals were grown by CH2Cl2 layering with Et2O. Yields: 8–52%. - Heterometallic clusters: CAu6Ag2 clusters [(C)(Au-L)6Ag2](BF4)4 (3a–d) were obtained by adding AgBF4 to solutions of 2a–d in CH2Cl2/MeOH, followed by crystallization (yields 64–93%). - Structural characterization: Single-crystal X-ray diffraction (Rigaku XtaLAB Synergy-DW, CuKα, 93 K) solved 2a–d and 3a–d, revealing bicapped octahedral cores with two opposite Ag(I) sites coordinated by three pyridyl donors and interacting with neighboring Au atoms. Intramolecular C–H···Au contacts were identified; key distances were measured. Solution stability was confirmed by 1H/13C NMR and ESI-MS. - Spectroscopy and photophysics: UV–vis absorption spectra in CH2Cl2 (or CH2Cl2/MeOH for 3d) were recorded (300–450 nm bands); steady-state emission measured in solid state and solution; quantum yields (Φ) determined; time-resolved phosphorescence lifetimes (τ) recorded; radiative (kr) and non-radiative (knr) rate constants derived (kr ≈ Φ/τ, knr ≈ (1−Φ)/τ). Comparative data were obtained for phosphine-protected analogue 4. - Theoretical calculations: TD-DFT computed absorption spectra, molecular orbitals (HOMO/LUMO; triplet SOMO/SOMO−1), and ligand contributions (Mulliken partition). Radiative rate constants and lifetimes were calculated using ZORA with spin–orbit coupling (ADF). Minimum energy crossing points (MECP) between T1 and S0 were located using the Harvey method to assess non-radiative decay barriers. - Cell imaging and uptake mechanism: Confocal luminescence microscopy of HeLa, HEK293T, and COS7 cells incubated with clusters (typically 1–2 μM; higher 5–10 μM assessed for comparison). Co-localization with ER-tracker Red identified organelle accumulation. Time-lapse imaging monitored uptake dynamics (0–30 min) and stability (up to 36 h). Phosphorescence lifetime imaging microscopy (PLIM) quantified intracellular lifetimes. Uptake pathways were probed by temperature dependence (4 °C vs 37 °C) and pharmacological inhibitors: wortmannin (macropinocytosis), sucrose (clathrin-mediated), genistein (caveolin-dependent). Cytotoxicity and morphological changes were monitored by microscopy.

- Structure: 3a–d adopt bicapped octahedral CAu6Ag2 cores with two Ag(I) ions on opposite faces, each bound by three pyridyl donors and interacting with three Au atoms. Strong intramolecular C–H···Au interactions were observed; shortest H–Au distances: 3a 2.690 Å, 3b 2.818 Å, 3c 2.765 Å, 3d 2.730 Å. Compared to phosphine analogue 4, NHC complexes have shorter C–Au and Au–Ag distances, yielding more compact cores. - Photophysics (solid state): 3a–d show yellow emission with λem,max 559–578 nm; parent CAu6 (2a–d) emit at 482–490 nm. - Photophysics (solution, CH2Cl2 unless noted): 3a and 3b (benzimidazolylidene NHCs) display intense yellow phosphorescence at λmax ≈ 562 nm with very high quantum yields Φ = 0.88 (3a) and 0.86 (3b), and lifetimes τ = 1.85 μs (3a) and 1.66 μs (3b). 3c and 3d (imidazolylidenes) show much lower Φ = 0.14 (3c) and 0.01 (3d; measured in CH2Cl2/MeOH 9:1) and shorter τ = 0.32 μs (3c) and 0.16 μs (3d). Rate constants: 3a kr = 4.8×10^5 s−1, knr = 0.6×10^5 s−1; 3b kr = 5.2×10^5 s−1, knr = 0.8×10^5 s−1; 3c kr = 4.3×10^5 s−1, knr = 26.5×10^5 s−1; 3d kr = 0.6×10^5 s−1, knr = 61.9×10^5 s−1. Phosphine analogue 4: Φ = 0.31, λmax = 650 nm, τ = 3.74 μs, kr = 0.8×10^5 s−1, knr = 1.9×10^5 s−1. - Absorption: Multiple bands 300–450 nm; benzimidazolylidene-protected 3a and 3b exhibit higher molar absorptivities (ε336 ≈ 8.3–8.4×10^4 M−1 cm−1) than imidazolylidenes 3c and 3d (ε333–385 ≈ 1.6–3.4×10^4 M−1 cm−1). - Theory: TD-DFT reproduces absorption energies (first peaks ~387–388 nm) and intensity trends (3b > 3d). Computed phosphorescence energies: 3b 2.09 eV (592 nm), 3d 2.08 eV (596 nm), agreeing with experiment (2.21 and 2.17 eV). NHC ligands reduce ligand participation in frontier orbitals and modify spin–orbit coupled states, increasing kr (3b > 3d). MECP barriers from T1 to S0: 11.6 kcal/mol (3b), 10.8 kcal/mol (3d), consistent with differing knr; small difference suggests additional factors influence non-radiative decay. Solvent (MeOH) lowers kr for 3d. - Cellular behavior: At 1–2 μM, 3a and 3b rapidly enter cells (≤10 min) and selectively accumulate in the endoplasmic reticulum (ER), confirmed by co-localization with ER-tracker Red. Accumulation in ER is visible by 10 min; nuclear region accumulation increases with longer incubation. ER localization of 3a persists for at least 36 h. Phosphine analogue 4 distributes uniformly in cytosol and lights nucleoli at higher concentration (10 μM). PLIM shows intracellular lifetime for 3a ≈ 0.15 μs, separable from autofluorescence. - Uptake mechanisms: Uptake is energy-dependent (abolished at 4 °C). For 3a, caveolin-dependent endocytosis is implicated (genistein inhibits ER accumulation), while 4 is affected by multiple inhibitors (macropinocytosis, clathrin- and caveolin-dependent), indicating nonspecific energy-dependent uptake and rapid cytosolic release. - Cytotoxicity/morphology: At 1–2 μM no cytotoxicity observed by confocal analysis; at 5–10 μM, nucleolar accumulation and cell rounding/nuclear envelope stress were observed.

The findings demonstrate that subtle ligand design—specifically NHC versus phosphine and benzimidazolylidene versus imidazolylidene—profoundly modulates both the electronic structure and excited-state dynamics of CAu6Ag2 clusters. NHC ligation enhances radiative decay via altered spin–orbit coupling of low-lying states while benzimidazolylidenes suppress non-radiative pathways, yielding record-high solution-phase phosphorescence quantum yields for Au clusters. The same ligand changes also dictate intracellular trafficking: NHC-protected clusters follow caveolin-dependent endocytosis and selectively accumulate in the ER, enabling organelle-specific phosphorescence imaging and PLIM readouts, whereas phosphine-protected clusters distribute nonspecifically in cytosol. Thus, ligand engineering provides a route to simultaneously tune photophysical performance and biological fate, addressing the study’s goal to create functional, organelle-targeted phosphorescent metal clusters.

This work introduces a rational design for NHC-based, carbon-centered CAu6Ag2 clusters with outstanding solution-phase phosphorescence (Φ up to 0.88) and microsecond lifetimes, and reveals that ligand structure governs organelle-selective translocation (ER targeting) and uptake pathways. The combined experimental and theoretical analyses clarify that NHC ligands enhance radiative decay through spin–orbit coupling effects and, with benzimidazolylidenes, suppress non-radiative decay. These insights establish strategies for designing metal clusters with tunable photophysics and subcellular targeting, informing development of imaging agents and metallodrugs. Future directions include expanding ligand libraries to program targeting to other organelles, dissecting the molecular basis of caveolin-mediated uptake and ER retention, optimizing solvent and medium effects on excited-state dynamics, assessing in vivo behavior and toxicity, and integrating these clusters into functional sensing and therapeutic platforms.

- Solvent dependence affects photophysics: e.g., 3d required CH2Cl2/MeOH and exhibits low kr attributed to MeOH effects. - Non-radiative decay origins are not fully resolved: MECP barrier differences between 3b and 3d are small, implying additional factors control knr. - Biological scope is limited to in vitro cell lines (HeLa, HEK293T, COS7); in vivo relevance and broader cell-type generality remain to be established. - Higher concentrations (5–10 μM) induce cellular stress and morphological changes, indicating a limited therapeutic window. - The precise molecular targets/receptors mediating caveolin-dependent uptake and ER retention were not identified. - Only a limited set of NHC ligands was explored; generalizability across broader ligand structures is untested.