Phenylthiol-BODIPY-based supramolecular metallacycles for synergistic tumor chemo-photodynamic therapy

Significance

Pt-containing complexes have been widely investigated for their potential for antitumor efficacy. Here we report two triangular metallacycles via the coordination of diplatinum(II) acceptors and phenoxy-boron dipyrromethene (BODIPY) or phenylthiol-BODIPY ligands. The triangular metallacycle with phenylthiol-BODIPY shows enhanced antitumor efficacy by the synergistic chemotherapy of platinum(II) ligand and photodynamic therapy of heavy-atom-functionalized BODIPY photosensitizer compared to individual single therapy. This study provides a foundation for further studies on the in vivo efficacy of synergistic chemo-photodynamic therapy for solid tumors.

Abstract

The development of more effective tumor therapy remains challenging and has received widespread attention. In the past decade, there has been growing interest in synergistic tumor therapy based on supramolecular coordination complexes. Herein, we describe two triangular metallacycles (1 and 2) constructed by the formation of pyridyl boron dipyrromethene (BODIPY)–platinum coordination. Metallacycle 2 had considerable tumor penetration, as evidenced by the phenylthiol-BODIPY ligand imparting red fluorescent emission at ∼660 nm, enabling bioimaging, and transport visualization within the tumor. Based on the therapeutic efficacy of the platinum(II) acceptor and high singlet oxygen (¹O₂) generation ability of BODIPY, 2 was successfully incorporated into nanoparticles and applied in chemo-photodynamic tumor therapy against malignant human glioma U87 cells, showing excellent synergistic therapeutic efficacy. A half-maximal inhibitory concentration of 0.35 μM was measured for 2 against U87 cancer cells in vitro. In vivo experiments indicated that 2 displayed precise tumor targeting ability and good biocompatibility, along with strong antitumor effects. This work provides a promising approach for treating solid tumors by synergistic chemo-photodynamic therapy of supramolecular coordination complexes.

The high morbidity and mortality of cancer have motivated the search for effective tumor therapy, which remains as a challenge in medicine. Currently, tumors are treated mainly by surgery, chemotherapy, and radiotherapy (1–4). Among the recent developments in diagnostic equipment and treatment technology, photodynamic therapy (PDT), which involves the use of photosensitizers and specific light activation, has emerged as an effective treatment modality (5–11). PDT relies on photosensitizing dyes to produce highly active singlet oxygen (¹O₂), which surrounds biological macromolecules and is toxic to the surrounding diseased tissue, thereby killing cancer cells in solid tumors (12). Compared with traditional chemotherapy, PDT shows potential as an efficient approach for tumor therapy because of its fewer side effects, minimal damage to normal tissues, and negligible drug resistance (13). Photosensitizers play a vital role in determining the therapeutic effect of PDT. Among these, boron dipyrromethene (BODIPY) derivatives have received wide attention compared to other fluorescent dyes for use in PDT because of their ¹O₂ generation ability and long fluorescence emission wavelength after simple modifications or coordination-driven self-assembly (2, 14–19). For example, introducing sulfur, and platinum(II) to BODIPY acting as heavy-atom/metal cannot only red-shift the absorption and fluorescence emission, but also promote ¹O₂ generation, and boost the effect of PDT (14–20).

In recent decades, coordination-driven self-assembly has become a powerful tool for constructing supramolecular coordination complexes (SCCs) (20–25). Numerous discrete SCCs have been synthesized via the formation of metal-organic coordination bonds between electron-poor acceptors containing transition metals (Fe, Cu, Co, Zn, Pt, Pd, etc.) and rigid electron-rich donors (such as carboxylate, pyridine), including metallacycles and metallacages (26–35). By rational structure design and careful ratio control of acceptors and donors, SCCs with various geometries and different properties have been developed, which attracted great interest and showed wide application prospects in supramolecular polymers, host-guest chemistry, sensors, catalysts, amphiphilic self-assembly, and biomedicines (36–46). Furthermore, because of the excellent antitumor effects of Pt-containing complexes, SCCs based on Pt-coordination show potential for tumor therapy (47–49). For instance, a previous study described an elegant light-emitting metallacage, assembled using Pt as the metal nodes, with higher antitumor efficacy compared with clinical Pt-based drugs (50). Yu et al. (51) developed a fluorescent, self-assembled metallacycle to encapsulate doxorubicin as the chemotherapeutic drug for cell imaging and synergistic cancer therapy. Zhou et al. (52) designed a self-assembled BODIPY-Pt supramolecular metallacycle for chemo-photodynamic therapy and reported excellent synergistic effects in A2780cis cells in vitro; however, no antitumor effect was observed in vivo. Thus, considering the superior antitumor effects of platinum(II) and ¹O₂, the development of Pt-based SCCs using ¹O₂ generating heavy-atom-functionalized BODIPY photosensitizers as functional building blocks with the goal of combining chemotherapy and PDT may enhance antitumor efficacy compared to individual single therapy.

In this work, we report two triangular metallacycles (1 and 2; Fig. 1) with ¹O₂ generation ability, constructed via the coordination of diplatinum(II) acceptors and phenoxy-BODIPY or phenylthiol-BODIPY ligands. Because of its superior ¹O₂ generation capability and longer maximum fluorescence emission wavelength, metallacycle 2 was further encapsulated by an amphiphilic polymer to prepare nanoparticles (NPs) for tumor therapy. The BODIPY ligands serve as the photosensitizer to generate ¹O₂ to kill cancer cells under light irradiation, whereas the platinum(II) acceptors in the metallacycle enhance the therapeutic effect through strong antitumor effects. In vitro and in vivo experiments demonstrated that the synergistic combination of PDT and chemotherapy greatly enhanced the antitumor efficacy compared with single-treatment approaches. Moreover, the characteristic fluorescence of the BODIPY core within the metallacycle enabled imaging of the transport of therapeutic metallacycles by confocal microscopy.

Fig. 1.

Chemical structures and synthetic procedures of metallacycles (1 and 2), BODIPY ligands (3 and 4), and diplatinum acceptor (5).

Results and Discussion

By applying a coordination-driven self-assembly method, the triangular metallacycles (1 and 2) were obtained in near quantitative yields (98% for both) by mixing 60° diplatinum(II) acceptor (5) and the corresponding BODIPY ligand (3 or 4) in a 1:1 ratio in dimethyl sulfoxide (DMSO) and stirring at 50 °C for 12 h. The detailed procedures for the synthesis of 3 and 4 are presented in SI Appendix, Figs. S1–S30. The structures of the triangular metallacycles were characterized using multinuclear NMR spectroscopy (¹H and ³¹P) and electrospray ionization time-of-flight mass spectrometry (ESI-TOF-MS). A sharp singlet peak was observed in the ³¹P NMR spectra of metallic triangles 1 and 2 at 13.70 ppm (¹J_Pt–P = 2660 Hz; Fig. 1, spectrum B) and 13.80 ppm (¹J_Pt–P = 2668 Hz; Fig. 2, spectrum C), indicating a single phosphorus environment. Relative to free platinum(II) acceptor 5 (Fig. 2, spectrum A), the peaks displayed upfield chemical shifts of 5.83 and 5.73 ppm for 1 and 2, respectively, suggesting the coordination of the pyridyl groups to platinum(II) as well as the formation of a highly symmetrical structure. In the ¹H NMR spectrum of metallacycle 1 (Fig. 2, spectrum E), the peaks related to the α-pyridyl protons H_b displayed a downfield shift (from 8.62 to 8.88 ppm) compared to that of free BODIPY ligand 3 (Fig. 2, spectrum F). Similarly, a downfield chemical shift (from 8.66 to 8.91 ppm) of the α-pyridyl protons H_c was observed in the ¹H NMR spectrum of 2 (Fig. 2, spectrum G) in contrast to that of the free BODIPY ligand 4 (Fig. 2, spectrum H). Additionally, the signals corresponding to H_a on the platinum(II) acceptor 5 (Fig. 2, spectrum D) shifted downfield (from 8.22 to 8.49 and 8.51 ppm) after the assembly of the metallacycle. These results support the formation of the triangular metallacycles. ESI-TOF-MS provided further evidence of the successful production of triangular metallacycles (53). For example, the mass spectrum of metallacycle 1 (Fig. 2, spectrum I) displayed four sets of mass peaks, from 3+ to 6+, representing the [3 + 3] assembled metallacycles with loss of different numbers of triflate counterions (OTf^–). A peak at m/z = 1152.72 representing the loss of five triflate counterions from 1 ([1 – 5OTf]⁵⁺) was compared with the expected isotope pattern (Fig. 2, spectrum I), supporting the stoichiometric composition of the metallacycle. In the mass spectrum of metallacycle 2, similar peaks were found and their isotropic distribution matched well with the calculated data (Fig. 2, spectrum J). In the 2D diffusion-ordered ¹H NMR spectroscopy (DOSY) experiments, a single band was observed, confirming the formation of a single product (SI Appendix, Figs. S22 and S28). The diffusion coefficients of 1 and 2 were measured to be the same (3.43 × 10⁻¹¹ m²/s) because of their similar molecular size. Overall, these results indicated the successful self-assembly of the discrete metallacycles, 1 and 2.

Fig. 2.

(A–C) Partial ³¹P NMR (CDCl₃/CD₂Cl₂; 297 K) spectra and (D–H) ¹H NMR spectra (DMSO-d₆; 297 K) of (A and D) the free 60° diplatinum(II) acceptor (5), (F) phenoxy-boron dipyrromethene (BODIPY) (3), (H) phenylthiol-BODIPY (4), (B and E) metallacycle 1, and (C and G) metallacycle 2. Experimental (red) and calculated (blue) ESI-TOF-MS spectra of (I) [1 – 5OTf]⁵⁺ and (J) [2 – 5OTf]⁴⁺. Absorption and fluorescence spectra of (K and L) metallacycle 1 and (M and N) metallacycle 2 in different solvents (λ_ex: 570 nm for 1 and 610 nm for 2; c = 10 μM).

The optical properties of the BODIPY-based metallacycles were also investigated, including the UV-Vis absorption and fluorescence emission characteristics in different solvents. In the absorption spectra, similar absorption bands, at ∼570 and ∼610 nm for metallacycle 1 and 2, respectively, were observed in different solvents (Fig. 2, spectra K and M, SI Appendix, Figs. S24 and S25), due to the π-π* transition of BODIPY moiety. In the emission spectra, the fluorescence maxima of 1 and 2 were observed at ∼610 and 645 nm (Fig. 2, spectrum L and N), respectively. However, the emission intensities varied according to the solvents used (Fig. 2, spectrum H and I). For example, in dimethylformamide and DMSO, metallacycle 1 and 2 displayed small red shifts in the fluorescence wavelength and a decrease in the fluorescence intensities, which may be attributed to the stability of the molecular excited-state and/or effect of strong polar solvents on electron transfer between the metal centers and ligands. The quantum yields (Φ_F) of the BODIPY ligands and the metallacycles in different solvents were also measured (SI Appendix, Table S1; rhodamine B; Φ_F = 50% in ethanol as the standard). The difference in the substituents greatly impacted the maximum fluorescence emission wavelength of the BODIPY ligands and metallacycles. 2, with a structure based on phenylthiol-BODIPY, therefore has a longer fluorescence emission wavelength and lower quantum yield because of the introduction of a sulfur atom (a heavy atom), consistent with previous reports (14).

The ¹O₂ generation ability of the metallacycles was then evaluated using electron paramagnetic resonance (EPR) with 2,2,6,6-tetramethyl-4-piperidone (TEMP) as the ¹O₂ capturer. Characteristic triplet signals representing the oxidized TEMP were observed in the EPR spectra after light irradiation, indicating successful production of ¹O₂ by the metallacycles (Fig. 3B). Metallacycle 2 displayed a much stronger ¹O₂ generation ability than 1 under the same conditions. The generation of ¹O₂ requires the triplet state of the photosensitizer; and the sulfur atom may promote the triplet transition of the metallacycle, thereby producing more ¹O₂. The singlet oxygen quantum yield of metallacycle 2 was further measured using rose bengal (39%) as the standard (54, 55). The singlet oxygen quantum yield of metallacycle 2 was calculated to be 12.3% (SI Appendix, Fig. S31). Therefore, 2 was identified as a candidate for bioapplication.

Fig. 3.

(A) Nanoencapsulation method for preparing nanoparticles (NPs); CT, chemotherapy; DSPE-PEG-MAL, 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[maleimide(polyethylene glycol)-2000]; ROS, reactive oxygen species. (B) EPR spectra of metallacycles, 1 and 2; the signals at both sides of the spectra represent the internal marker, Mn²⁺. (C) TEM image and (D) dynamic light scattering (DLS) analysis of NPs (Inset: Tyndall effect of the aqueous NP solution).

NPs composed of polymers with good biocompatibility are a promising platform for disease diagnosis and treatment. Because of the hydrophobicity of the metallacycle and hydrophilic environment of living organisms, amphiphilic liposome 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[maleimide(polyethylene glycol)-2000] (DSPE-PEG-MAL) was used to encapsulate metallacycle 2. The metallacycle-loaded NPs were obtained via the precipitation method (Fig. 3A). The aqueous NP solution exhibited a typical Tyndall effect (Fig. 3D; inset), demonstrating the formation of nanoaggregates. The size of NPs plays a pivotal role in their accumulation and retention in tumors. As such, the size and morphology of the NPs were determined using a transmission electron microscope (TEM) and dynamic light scattering. TEM revealed spherical, regular, and well-separated NPs with an average size of ∼50 nm (Fig. 3C and D). Dynamic light scattering indicated an average NP size of 49 nm in water, in agreement with the TEM measurement. This dimension corresponds with the reported optimal particle size (around 50 nm) for enhanced in vivo anticancer properties, e.g., stronger tumor inhibition and deeper tissue penetration. Additionally, the size distribution showed negligible changes after one week (Fig. 3D), suggesting good stability of the NPs in an aqueous environment.

The cytotoxicities of the NPs containing metallacycle 2 or BODIPY 4 were investigated in human embryonic kidney 293 cells (HEK293; Fig. 4A) and U87 glioblastoma cells (Fig. 4B) under light irradiation using a 3-(4′,5′-dimethylthiazol-2′-yl)-2,5-diphenyl tetrazolium bromide assay. Different concentrations of 2 and 4 in a NP solution displayed low toxicity toward U87 and HEK293 cells in the dark. However, under light irradiation, the viability of both cell types decreased gradually with increased concentrations of 2 and 4. The half-maximal inhibitory concentrations (IC₅₀) of 4 and 2 against U87 cells were 2.49 and 0.35 μM, respectively. Thus, NPs containing metallacycle 2 displayed greater antitumor activity. Under the same conditions, the viability of U87 cells was lower than that of HEK293 cells, likely because of disassembly of the metallacycle in the acidic internal environment of U87 cancer cells, which resulted in release of the highly cytotoxic platinum acceptor.

Fig. 4.

Cell viability of (A) human embryonic kidney 293 (HEK293) cells and (B) U87 cells after treatment with 4 and 2 in the absence or presence of laser irradiation. (C) Confocal laser scanning microscopy images of untreated (blank) cells and cells treated with 4 or 2 in the presence of the singlet oxygen detector, dichlorofluorescin diacetate (DCF-DA) under light irradiation. Confocal laser scanning microscopy images and mean fluorescence intensities of human primary glioblastoma (U87) cells after incubation with (D) ligand 4 or (E) metallacycle 2.

To evaluate the generation of ¹O₂ in cells after the internalization of NPs, dichlorofluorescein diacetate was selected as an oxidant-sensitive fluorescent dye, which exhibited green emission after oxidization to dichlorofluorescein by reactive oxygen species. NPs containing the metallacycle 2 or BODIPY 4 were internalized by U87 cells, which displayed typical green fluorescence attributed to dichlorofluorescein following light irradiation; in contrast, the blank group displayed no green fluorescence, suggesting efficient ¹O₂ generation by 2 and 4 (Fig. 4C). A more pronounced green fluorescence was observed in the U87 cells treated with NPs containing the metallacycle 2 compared to other cells, indicating a higher intracellular concentration of ¹O₂, possibly because of the higher intracellular accumulation of the metallacycles.

Cellular uptake and localization experiments were further conducted in vitro. Human malignant glioma cells (U87) were incubated at 37 °C with phenylthiol-BODIPY ligand 4 and metallacycle 2 for 24 h, respectively. Subsequently, confocal laser scanning microscopy was used to determine whether 2 and 4 were internalized by U87 cells, and counterstaining with 4′,6-diamidino-2-phenylindole (DAPI) was performed to differentiate nuclear and cytoplasmic areas. In the confocal laser scanning microscopy images (Fig. 4D and E), the cytoplasm of U87 cells showed strong red fluorescence after incubation, confirming that 2 and 4 were located mainly in the cytoplasm. U87 cells incubated with metallacycle 2 showed a strong red fluorescence, indicating a high cellular uptake rate of metallacycle 2 in U87 cells. This may be attributed to the electrostatic interaction between the positive metallacycle and negative cell membrane surface, which facilitated entry of the metallacycles.

The tumor accumulation efficacies were then monitored by recording fluorescence imaging of tumor sites at different times point (1 h, 2 h, 4 h, 8 h, 16 h, 24 h, 48 h) after intratumoural injection BODIPY 4 NPs and metallacycle 2 NPs, respectively. The fluorescence signal of the metallacycle 2 NPs at tumor tissues gradually increased with time. In vivo fluorescence imaging at 1 h postinjection showed that the metallacycle 2 NPs had higher fluorescence signal at tumor site than BODIPY 4 NPs. This signal persisted, even at postinjection times as long as 48 h, with a significant biodistribution of the metallacycle 2 NPs throughout the specimen (tumor and liver). In contrast, the fluorescence signal in the body annihilated much faster for the mice after administration of the BODIPY 4 NPs (Fig. 5A and SI Appendix, Fig. S32). Thus, metallacycle 2 can be applied for imaging-induced tumor therapy due to its strong and long-time fluorescence emission.

Fig. 5.

(A) In vivo and ex vivo fluorescence images of the mice and their major organs at 24 h after the last injection (H: heart; L: liver; S: spleen; L': lung; K: kidney; T: tumor). (B) Protocol of tumor therapy in mice inoculated with glioblastoma cells (U87); Variations in the (C) tumor volume and (D) body weight of mice during therapy; PBS, phosphate-buffered saline (control group); (E) gross appearance of the excised tumors at the end of the 21-d treatment period.

In order to evaluate the antitumor efficacy, U87 tumor-bearing mice were intraperitoneal injected with different formulations, including PBS, NPs containing BODIPY 4 and NPs containing metallacycle 2 (8 injections in 21 d; Fig. 5B). The tumor sites were irradiated using a 600–605 nm (60 mW cm⁻²) laser for 10 min after 24 h of NPs injection. Control mice injected with PBS showed rapid tumor growth over time (Fig. 5C and D), with an average tumor volume of nearly 1000 mm³ at the end of the experiment, whereas treatment with NPs containing 2 or 4 inhibited tumor growth. Accordingly, the tumor inhibition rate of NPs containing metallacycle 2 was calculated to be 66.6%, which was higher than that of NPs containing metallacycle 4 (27.8%), suggesting a better antitumor activity of NPs containing metallacycle 2. In comparison with the structure of the metallacycle 2 and BODIPY 4, it can be inferred that organoplatinum(II) in metallacycle 2 played a crucial role as the chemical antitumor drug, which was consistent with literature reports (3, 50). Thus, an effective chemo-photodynamic synergistic effect was achieved in NPs containing metallacycle 2. The therapeutic efficacy of NPs containing metallacycle 2 was also investigated by the hematoxylin-eosin (H&E) staining, nuclear nonhistone protein (Ki67) staining, and terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL). After treatment of NPs containing metallacycle 2, the greatest degree of nucleus–cytoplasm separation, the fewest green fluorescent dots representing the antigens secreted during cell proliferation, and the most abundant DNA disruption representing apoptosis was observed in the tumor cells, indicating a large inhibition of proliferative activity (Fig. 6).

Fig. 6.

H&E staining, Ki67 staining, and TUNEL images of tumor tissue sections from mice treated with phosphate-buffered saline (PBS), ligand 4, or metallacycle 2 NPs.

Biosafety is an important concern in the application of NP-based therapy. Neither death nor any sign of toxicity was observed during the 3-wk experiment, and the body weight of the mice showed negligible changes during the treatment period (Fig. 5E). Moreover, the blood chemistry test results of mice in the metallacycle treatment group showed no significant variation in liver and kidney functional indices (SI Appendix, Fig. S33). Histological examinations of the major organs (heart, liver, spleen, lung, and kidney) showed no obvious abnormalities or lesions after treatment with NPs containing 2 (SI Appendix, Fig. S34). These data demonstrate that NPs containing the metallacycle 2 were nontoxic and safe for in vivo applications.

Conclusions

Two triangular metallacycles (1 and 2) were constructed via coordination-driven self-assembly using phenoxy-BODIPY 3 and phenylthio-BODIPY 4 as the building blocks, respectively. The structures of 1 and 2 were confirmed using multinuclear NMR (¹H and ³¹P), ESI-TOF-MS, UV-Vis spectroscopy, and fluorescence spectroscopy. Because of the excellent antitumor ability of the organometallic platinum 5 and the efficient ¹O₂ generation ability of the fluorescent BODIPY 4, synergistic chemo-photodynamic therapy as well as bioimaging were achieved using 2. In vitro experiments demonstrated the higher rates of cellular uptake and ¹O₂ generation, and lower IC₅₀ value of 2 versus 4. Furthermore, 2 exerted a stronger antitumor effect in vivo than 4, indicating that the combination of chemotherapy and PDT results in a synergistic effect. This study provides an alternative strategy for the construction of antitumor drugs with enhanced efficacy through coordination-driven self-assembly.

Materials and Methods

All reagents and deuterated solvents were commercially available and used as supplied without further purification. Compounds 5 and 6 were synthesized according to the published literature procedures (3, 56). NMR spectra were recorded on a Bruker Advance 500 MHz spectrometer. ¹H and ¹³C NMR chemical shifts were reported relative to residual solvent signals, and ³¹P{¹H} NMR chemical shifts were referenced to an external unlocked sample of 85% H₃PO₄ (δ = 0.0 ppm). Mass spectra were recorded on Micromass Quattro II triple-quadrupole mass spectrometer and 6530 Q-TOF LC/MS. The melting points were collected on a YRT-3 automatic melting point apparatus. The UV-Vis absorption spectra were measured by a Hitachi U-5300 absorption spectrophotometer. The fluorescent emission spectra were recorded on a Hitachi F-7000 fluorescence spectrophotometer. Transmission electron microscopy (TEM) was performed on a Hitachi F-7700. Dynamic light scattering experiments were performed using a Nano ZS90 instrument with a He-Ne laser (633 nm) and 90° collecting optics. Quantum yields were measured via relative method by using rhodamine B (Φ_F = 50% in ethanol, 297 K) as the standard.

The triangular metallacycles (1 and 2) were obtained in near quantitative yields (98% for both) by mixing 60° diplatinum(II) acceptor (5) and the corresponding BODIPY ligand (3 or 4) in a 1:1 ratio in DMSO and stirring at 50 °C for 12 h.

Metallacycle 1.

¹H NMR (500 MHz, DMSO-d₆, 297 K): 8.83 (t, 18H), 8.09 (s, 12H), 7.93 (m, 9H), 7.83 (m, 21H), 7.71 (m, 12H), 7.59 (m, 12H), 7.29 (m, 12H), 7.04 (m, 18H), 4.00 (s, 18H), 2.56 (s, 9H), 1.31 (s, 72H), 1.08 (m, 108H). ³¹P NMR (202 MHz, DMSO-d₆, 297 K): 13.70 ppm (s, ¹⁹⁵Pt satellites, ¹J_Pt–P = 2669.4 Hz). ¹⁹F NMR (471 MHz, DMSO-d₆, 297 K): -77.72 (s), -144.43 (m). ESI-TOF-MS: m/z 1152.72 [1 – 5OTf]⁵⁺, 1478.16 [1 – 4OTf]⁴⁺.

Metallacycle 2.

¹H NMR (500 MHz, DMSO-d₆, 297 K): 8.88–8.81 (m, 18H), 8.10 (s, 12H), 8.89–8.80 (m, 32H), 7.72 (m, 6H), 7.53 (m, 6H), 7.37 (m, 6H), 7.21 (m, 24H), 7.13 (m, 9H), 4.01 (s, 18H), 2.50 (s, 9H), 1.34 (m, 72H), 1.11 (m, 108H). ³¹P NMR (202 MHz, DMSO-d₆, 297 K): 13.80 ppm (s, ¹⁹⁵Pt satellites,¹J_Pt–P = 2674.9 Hz). ¹⁹F NMR (471 MHz, DMSO-d₆, 297 K): -77.72 (s), -138.28 (m). ESI-TOF-MS: m/z 1171.89 [2 – 5OTf]⁵⁺, 1502.13 [2 – 4OTf]⁴⁺.

Data Availability

All study data are included in the article and SI Appendix.