Efficient Singlet Oxygen Generation by Luminescent 2-(2′-Thienyl)Pyridyl Cyclometalated Platinum(II) Complexes and their Calixarene Derivatives

Abstract

Luminescent cyclometalated platinum(II) complexes, namely [Pt(Thpy)(PPh₃)X]ⁿ⁺ (HThpy = 2-(2′-thienyl)pyridine; X = Cl⁻ (1), n = 0; X = CH₃CN (2), pyridine (3), n = 1) and [Pt(Thpy)(HThpy)Y]ⁿ⁺ (Y = Cl⁻ (4), n = 0; Y = pyridine (5), n = 1), exhibit structured emission with peak maximum at ~556 and 598 nm in degassed acetonitrile and with emission quantum yield and lifetime of up to 0.38 and 26 μs respectively. These complexes are efficient photosensitizers of singlet oxygen with yields up to > 90%. Complex 5 exhibited photocytotoxicity towards cancer cells and fluorescence microscopic images of cells incubated with 5 reveal substantial uptake at the nucleus and mitochondria.

INTRODUCTION

Photosensitized production of singlet molecular oxygen, ¹O₂, has been widely investigated for non-radical aerobic oxidation, photodynamic therapy (PDT) and polymer science (1–16). Apart from organic photosensitizers commonly used for the generation of singlet oxygen, transition metal complexes have recently been shown to be attractive alternatives due to their strong absorptions in the UV-visible region, long-lived triplet metal-to-ligand charge transfer (MLCT) states, high emission quantum efficiencies of the triplet state, and high photostability (17–23). In the past decades, Che and coworkers investigated the photoluminescent properties of cyclometalated platinum(II) complexes, namely [Pt(C^N^N)R¹]ⁿ⁺ bearing tridentate C-deprotonated 6-phenyl-2,2′-bipyridine [C^N^N]⁻ and related ligands (24–27), and [Pt(Thpy)R²R³]ⁿ⁺ with bidentate Thpy ligand (HThpy = 2-(2′-thienyl)pyridine) (28–30). Their spectroscopic characteristics were reported to vary significantly with the electronic properties and structure of the ancillary ligands (R^1–3).

Under aerobic conditions, the MLCT excited states of cyclometalated platinum(II) complexes are quenched by triplet dioxygen, which may lead to formation of singlet oxygen in various amounts (20,21). Previous studies have established that the amount of singlet oxygen photosensitized by these complexes depends on the excited state redox potential, and steric bulk of the cyclometalated ligand (20,21). However, other variables may be involved in determining whether or not a late transition cyclometalated complex is capable of producing singlet oxygen in high yield. Despite the fact that the photophysical properties of literally hundreds of cyclometalated complexes have been investigated under anaerobic conditions, there is a surprising paucity of data for their behavior in the presence of dioxygen and light. Thus, the effects of ancillary ligand(s), solvent, neighbouring metal ion(s) (in polynuclear complexes), as well as electronic charge on the metal center have not been investigated towards the production of singlet oxygen in these complexes.

Herein we describe the photophysical properties and production of singlet oxygen from a series of cyclometalated [Pt(Thpy)(PPh₃)X]ⁿ⁺ and [Pt(Thpy)(HThpy)Y]ⁿ⁺ complexes (1–5; Scheme 1). We therefore envisage, through varying the ancillary X or Y group, the emission properties can be systematically modified and probed to maximize the efficiency of ¹O₂ production. With reference to previous studies on luminescent biological probes based on cyclometalated iridium(III) and platinum(II) complexes with auxiliary ligands modified with amino acids (20,31), we anticipate that the 2-(2′-thienyl)pyridyl platinum(II) complexes which display visible emission with long emission lifetime and high quantum yields for ¹O₂ production (21), are photosensitizers for the generation of singlet oxygen of biological interest. In addition, the red-shifts in absorption and emission of these new photosensitizers are advantageous towards shifting the operational window of PDT into the tissue transparency window at 700–1100 nm.

Scheme 1.

Structures of 2-(2′-thienyl)pyridyl platinum(II) complexes 1–7.

We also envision that 2-(2′-thienyl)pyridyl platinum(II) complexes incorporated into upper-rim phosphinated calix[4]arenes, as reported previously by us (30), could be developed as photosensitizing hosts towards target guests. Therefore we describe the production of singlet oxygen by complexes [(PtThpyCl)₂L] (6) and [{PtThpy(CH₃CN)}₂L](ClO₄)₂ (7), where L is 5,17-bis(diphenylphosphino)-25,26,27,28-tetra-n-butoxycalix[4]arene. The selectivity of binding would be affected by the moieties at the lower rim of the calixarenes, and/or their conformation and size.

MATERIALS AND METHODS

Materials and General Procedures

[Pt(Thpy)(HThpy)Cl] (4) was prepared by the literature method (28–29). Acetonitrile for photophysical measurements was distilled over potassium permanganate and calcium hydride. Dichloromethane for photophysical studies was washed with concentrated sulfuric acid, 10 % sodium hydrogen carbonate, and water, dried by calcium chloride, and distilled over calcium hydride. All other solvents were of analytical grade and purified according to conventional methods (32).

Instrumentation and Physical Measurements

Fast atom bombardment (FAB) mass spectra were obtained on a Finnigan Mat 95 mass spectrometer with a 3-nitrobenzyl alcohol matrix, whereas electrospray mass spectra were obtained on a LCQ quadrupole ion trap mass spectrometer. High-resolution ESI mass spectra were obtained from a Waters Micromass Q-Tof Premier quadrupole time-of-flight tandem mass spectrometer. ¹H (500 MHz), ¹³C (126 MHz) and ³¹P (202 MHz) NMR spectra were performed on DPX 500 Bruker FT-NMR spectrometer with chemical shifts (in ppm) relative to tetramethylsilane (¹H and ¹³C) and H₃PO₄ (³¹P) as references. Elemental analyses were performed by the Institute of Chemistry at the Chinese Academy of Sciences, Beijing. UV-vis spectra were recorded on a Perkin Elmer Lambda 19 UV/vis spectrophotometer.

Emission and Lifetime Measurements

Steady-state emission spectra were recorded on a Fluorolog-3 Model FL3-21 spectrophotometer. Solution samples for measurements were degassed with at least four freeze-pump-thaw cycles. Low-temperature (77 K) emission spectra for glassy solutions and solid-state samples were recorded in 5 mm diameter quartz tubes, which were placed in a liquid nitrogen Dewar equipped with quartz windows. The emission spectra were corrected for monochromator and photomultiplier efficiency and for xenon lamp stability. Emission lifetime measurements were performed with a Quanta Ray DCR-3 pulsed Nd:YAG laser system (pulse output 355 nm, 8 ns). The emission signals were detected by a Hamamatsu R928 photomultiplier tube and recorded on a Tektronix TDS 350 oscilloscope. Errors for λ values (±1 nm), τ (±10 %), Φ (±10 %) were estimated.

Luminescence quantum yields were determined using the method of Demas and Crosby (33) with [Ru(bpy)₃]Cl₂ in degassed acetonitrile as a standard reference solution (Φ_r = 0.062) and calculated according to the following equation: Φ_s = Φ_r(B_r/B_s)(n_s/n_r)²(D_s/D_r), where the subscripts s and r refer to sample and reference standard solution respectively, n is the refractive index of the solvents, D is the integrated intensity, and Φ is the luminescence quantum yield. The quantity B was calculated by B = 1 – 10^−AL, where A is the absorbance at the excitation wavelength and L is the optical path length. The UV-visible absorption and emission spectral data of 1–5 are listed in Table S1 (see Supplementary Materials).

Synthesis

Preparation and characterization of [Pt(Thpy)PPh₃Cl] (1) and [Pt(Thpy)PPh₃(CH₃CN)]ClO₄ (2) were documented by us previously (30). Using similar synthetic procedures adapted for 2 except pyridine was used as solvent, [Pt(Thpy)PPh₃(py)]ClO₄ (3) and [Pt(Thpy)(HThpy)py]ClO₄ (5) were obtained by using [Pt(Thpy)PPh₃Cl] (1) and [Pt(Thpy)(HThpy)Cl] (4) as precursors, respectively.

[Pt(Thpy)PPh₃(py)]ClO₄, 3

A mixture of [Pt(Thpy)PPh₃Cl] (1) (0.10 g, 0.15 mmol) and AgClO₄ (0.04 g, 0.19 mmol) in pyridine (10 mL) stirred at room temperature for 15 min afforded a pale yellow solution and white precipitate of AgCl. The pale yellow solution was separated from the white solid and was centrifuged for 3 min to remove residual AgCl solid. Addition of diethyl ether yielded a yellow solid, which was recrystallized by vapor diffusion of diethyl ether into a CH₂Cl₂/CH₃OH (2:1 v/v) solution of the crude product to afford yellowish orange crystals. Yield: 0.10 g, 82 %. ¹H NMR (CD₂Cl₂): δ 5.92 (d, J = 4.9 Hz, 1H), 6.99–7.06 (m, 3H), 7.30 (t, J = 7.0 Hz, 2H), 7.35–7.38 (m, 6H), 7.50 (t, J = 7.4 Hz, 3H), 7.58–7.63 (m, 7H), 7.82 (t, J = 7.8 Hz, 1H), 7.90 (t, J = 7.7 Hz, 1H), 8.38 (d, J = 6.4 Hz, 2H). ¹³C{¹H} NMR (CD₂Cl₂): δ 118.6, 121.6 (d, J = 3.4 Hz), 127.4, 127.5, 127.7, 128.0, 128.8 (d, J = 11.4 Hz), 131.9 (d, J = 2.3 Hz), 134.6 (d, J = 11.4 Hz), 135.7 (d, J = 4.6 Hz), 139.3 (d, J = 8.8 Hz), 139.7, 141.6, 143.2, 146.8, 151.4, 160.7. ³¹P{¹H} NMR (CD₃CN): δ 19.02 (t, with ¹⁹⁵Pt satellites, ¹J_PtP = 4124 Hz). ESI-MS: m/z 696.2 [M⁺], 617.1 [M – py]⁺. Anal. Calcd for C₃₂H₂₆N₂O₄PSPtCl: C, 48.28; H, 3.29; N, 3.52. Found: C, 48.42; H, 3.42; N, 3.29 %.

[Pt(Thpy)(HThpy)py]ClO₄, 5

The procedure for synthesis of 3 was applied using [Pt(Thpy)(HThpy)Cl] (4) (0.09 g, 0.16 mmol) to yield 0.096 g (85 %) of a yellow orange crystalline solid. ¹H NMR (CD₂Cl₂): δ 5.33 (s, 1H), 6.35 (d, J = 4.8 Hz, 1H), 6.93 (t, J = 6.7 Hz, 1H), 7.16 (t, J = 4.4 Hz, 1H), 7.31 (d with broad ¹⁹⁵Pt satellites, 1H, ³J_HH = 5.9 Hz), 7.47–7.54 (m, 5H), 7.64 (d, J = 5.0 Hz, 1H), 7.68 (d, J = 3.7 Hz, 1H), 7.78 (d, J = 7.3 Hz, 1H), 7.87 (t, J = 7.8 Hz, 1H), 7.95 (t, J = 7.8 Hz, 1H), 8.05 (t, J = 7.9 Hz, 1H), 8.24 (broad s, 1H), 9.26 (d with broad ¹⁹⁵Pt satellites, 1H, ³J_HH = 5.9 Hz). ¹³C{¹H} NMR (CD₂Cl₂): δ 118.7, 120.8, 125.5, 127.2, 127.3, 128.2, 129.1, 129.9, 130.4, 130.9, 139.5, 139.7, 139.8, 140.7, 140.8, 142.9, 148.7, 150.8, 154.1, 155.6, 163.2. ESI-MS: m/z 595.1 [M⁺], 516.1 [M–py]⁺. Anal. Calcd for C₂₃H₁₈N₃O₄S₂PtCl: C, 39.74; H, 2.61; N, 6.05. Found: C, 39.73; H, 2.63; N, 6.09 %.

X-ray Crystallography

Crystals of [Pt(Thpy)PPh₃(CH₃CN)]ClO₄ (2) were obtained by vapour diffusion of diethyl ether into an acetonitrile solution of the complex, whereas crystals of [Pt(Thpy)PPh₃(py)]ClO₄•(C₂H₅)₂O (3•(C₂H₅)₂O) and [Pt(Thpy)(HThpy)py]ClO₄ (5) were grown by vapour diffusion of diethyl ether into a CH₂Cl₂/CH₃OH (2:1 v/v) solution of the respective complex. The crystal data and details of data collection and refinement are summarized in Table S2 (see Supplementary Materials).

Diffraction data for 2, 3•(C₂H₅)₂O, and 5 were collected at 301 K on a Bruker Smart CCD 1000 diffractometer using graphite monochromatized Mo-Kα radiation (λ = 0.71073 A). For 3•(C₂H₅)₂O and 5, raw frame data were integrated with SAINT (34) program and semi-empirical absorption correction with SADABS (35) was applied. The structures were solved by the direct method employing SHELXS-97 (36) program on a PC. Pt, P, S, Cl and many non-H atoms were located according to the direct methods. The positions of other non-hydrogen atoms were found after successful refinement by full-matrix least-squares using the program SHELXL-97 (36) on a PC. For 2, one crystallographic asymmetric unit consists of one formula unit. One perchlorate anion was also located. The absolute structure was assisted by the Flack absolute structure parameter, which is equal to 0.013(7). For 3•(C₂H₅)₂O, one crystallographic asymmetric unit consists of one formula unit. One perchlorate anion and one diethyl ether solvent molecule located. For 5, one crystallographic asymmetric unit consists of one formula unit. One perchlorate anion was also located.

Emission Quenching Measurements for Cyclometalated Complexes

Luminescence quenching studies were conducted in dilute solutions (concentration of complexes 1–7 = 1 × 10⁵ M) in acetonitrile or dichloromethane that were purged for five minutes with N₂, air or dioxygen. The concentration of oxygen for samples in air (0.00183 M⁻¹) and dioxygen (0.00868 M⁻¹) was estimated from published oxygen solubility data (37). Samples were excited with a 405 nm pulsed diode laser having a pulse duration of ca. 1.2 ns and an energy of 500 nJ pulse⁻¹.

Set-up for Time-Resolved Detection of ¹O₂ NIR Emission

Singlet oxygen quantum yields and luminescence quenching rates of ¹O₂ were determined using a time-resolved Nd:YAG laser set-up (Minilase II, New Wave Research Inc., excitation pulse 1 ns) and a liquid N₂ cooled Ge photodetector (Applied Detector Corporation Model 403 S). A Schott color glass filter (model RG850; cut-on 850 nm; Newport, USA) was taped to the sapphire entrance of the detector to block any additional ultraviolet and visible light from entering. The port opening to the detector contained the remaining filters: a long wave pass filter (silicon filter model 10LWF ~ 1000; Newport, USA) which transmits in the range of 1100–2220 nm and blocks in the range of 800–954 nm; a band pass filter (model BP-1270–080-B*; CWL 1270 nm; Spectrogen, USA) which blocks in the UV, visible, and IR regions and only transmits in the range of 1200 to 1310 nm with a maximum transmission of 60% at 1270 nm. Signals were digitized on a LeCroy 9350 CM 500 MHz oscilloscope and analyzed using Origin software. All of those experiments were carried out at ambient temperature under air. A 250 W tungsten lamp (Oriel Instruments) equipped with a 10 cm H₂O filter and 492 nm cutoff filter was used for steady-state generation of ¹O₂.

¹O₂ Quantum Yield Measurements

Samples were prepared in acetonitrile or dichloromethane with absorbances between 0.1–0.3 at 355 nm or 532 nm excitation wavelengths. The absorbance of the reference sensitizers (C₆₀, perinaphthenone and methylene blue) and the metal complex were matched within 80%. The initial ¹O₂ intensity was extrapolated to t = 0. The data points of the initial 0–5 μs were not used due to electronic interference signals from the detector.

¹O₂ Quenching Measurements

The quenching rates of ¹O₂ were measured by Stern-Volmer analysis using Methylene Blue and C₆₀ as external sensitizers in CD₃CN and CD₂Cl₂, respectively. The concentration of the metal complexes used in the measurements ranged between 0.1–10 mM.

Binding Studies on Calf Thymus DNA and Bovine Serum Albumin (BSA)

Calf thymus DNA and bovine serum albumin (BSA) were purchased from Sigma Chemical Co. A 20 mM phosphate buffered saline (PBS) solution of pH 7.2 was prepared by dissolving 1.25 g of Na₂HPO₄, 0.35 g of NaH₂PO₄, and 8.0 g of NaCl in 1 L of doubly distilled water. Complex 5 (1.78 × 10⁻³ g) was first dissolved in 0.256 mL dimethyl sulfoxide (DMSO) to afford a solution with concentration of 1 × 10⁻² M. Dilution of the complex solution in DMSO with Tris buffer solution was achieved to afford the solution with complex concentration of 1 × 10⁻⁵ M. The concentration of DMSO in the resultant Tris buffer solution should be ~0.1 %. The emission spectral titrations were carried out on 3-mL solutions of 5 (complex concentration = 1 × 10⁻⁵ M) with increasing amounts of calf thymus DNA or BSA to give [DNA]/[Pt] between 3 and 18 and [BSA]/[Pt] between 0.1 and 0.6 respectively. DNA was added in 11 μL increments to the complex solution (1 × 10⁻⁵ M) in Tris buffer, whereas BSA was added in 0.3 μL increments to the complex solution (1 × 10⁻⁵ M) in PBS solution. The solution was mixed well and was allowed to stand for 5 minutes. The emission spectra were recorded upon each addition of 3 equivalents of DNA or 0.1 equivalent of BSA to the solution of 5, wherein the dilution of complex concentrations upon addition of DNA or BSA was negligible.

Photocytotoxicity

HeLa cells (Human cervical carcinoma) were seeded in a 96-well flat-bottomed microplate at 1 × 10⁴ cells/well in 100 μL MEM (minimum essential medium) containing 10% fetal bovine serum (FBS) and incubated for 24 h. Complex 5 was dissolved in DMSO as a 1 × 10⁻² M stock solution, and subsequently diluted with the medium and added to each well with final concentration of DMSO ≤ 1%. For determination of photocytotoxicity, cells were irradiated at room temperature for 30 min on a translucent plastic platform using broadband visible light (~1 mW/cm²). After further incubation for 48 h, 10 μL of (3-(4,5-dimethyl-2-thiazoyl)-2,5-diphenyltetrazolium bromide (MTT, 5 mg/mL) was added to each well and incubation was continued for 4 h. 100-μL Solubilization solution (10% SDS in 0.01 M HCl) was added to dissolve the formazan and the absorbance at 550 nm was measured with a microplate reader.

Fluorescence Microscopy

HeLa cells were seeded in 35 nm glass bottom dishes at 2 × 10⁵/dish in 2 mL MEM containing 10% FBS and then incubated for 24 h. Cells were treated with 5 μM of 5 for 1 h and examined with a fluorescence microscope (Zeiss Axiovert 200M and FluoArc) using FITC (fluorescein isothiocyanate) filter. Nucleus and mitochondria were stained with 5 μM Hoechst 33342 (Invitrogen) and 50 nM MitoTracker Red (Invitrogen), respectively, for 10 min.

RESULTS

Synthesis and characterization of 1, 2, and 4 have been reported previously (28–30), whereas complexes 3 and 5 were prepared by similar synthetic procedures as developed for 2, except pyridine was used as solvent in the reactions using precursors 1 and 4, respectively. Complexes 3 and 5 were characterized by ¹H and ¹³C{¹H} spectroscopy, ESI mass spectrometry, and X-ray crystallography (see Supplementary Materials). They are air- and moisture-stable at room temperature. We have investigated the UV-visible absorption and emission properties of 1–5; the spectral data and spectra are given in the Supplementary Materials. The absorption spectra of 1–5 in acetonitrile show high-energy, intense absorption band with λ_max at 228–330 nm (ε = 17600–41100 M⁻¹ cm⁻¹), which are attributed to spin-allowed intraligand ¹IL: π(Thpy) → π*(Thpy) transitions. The moderately intense, low-energy band with λ_max at 389–419 nm (ε = 2400–5600 M⁻¹ cm⁻¹) is assigned to a spinallowed metal-to-ligand charge transfer ¹MLCT: Pt(5d) → π*(Thpy) transition (28–30).

The emission spectra of 1–3 in degassed acetonitrile at 298 K exhibit predominantly vibronically structured bands with peak maximum at ~556 and 600 nm and a minor broad peak at λ_max = 467 nm. Comparison of the emission data among complexes 1–3 reveals an increase in both of the emission intensity and lifetime upon replacing the coordinated chloride ligand in 1 with the neutral nitrogen donor ligands, namely acetonitrile (in 2) and pyridine (in 3), respectively. This ligand modification causes an increase in the cationic charge of the platinum(II) complexes, consequently leading to lowering in energy of the occupied Pt(5d) orbitals. Thus the energy of d-d excited state increases accounting for the increase in emission intensity and lifetime. The low-energy emission bands (λ > 550 nm) show a dramatic decrease in intensity upon exposure to air, while the high-energy peak at 467 nm is insensitive to the exposure to air with no change in the emission intensity (38). Similarly, the emission lifetime at λ_max > 550 nm shows substantial diminution upon aeration of the solution. For example, the emission lifetime of [Pt(Thpy)PPh₃(py)]ClO₄ (3) in degassed CH₃CN at λ_max 556 nm is 6.73 μs, whereas upon exposure to air, the emission lifetime decreases to 0.26 μs. The low-energy emission bands with λ_max > 550 nm were previously assigned to triplet excited states with mixed ³MLCT and ³IL character (29,39,40). Therefore, the triplet mixed ³MLCT/³IL emissive excited states of [Pt(Thpy)(PPh₃)X]ⁿ⁺ complexes can be efficiently quenched by triplet dioxygen. The emission spectra of 4 and 5 in acetonitrile show low-energy structured bands with peak maxima at ~558 and 602 nm, which can be similarly assigned to having mixed ³MLCT and ³IL parentage.

The quantum yields for singlet oxygen production (Φ_Δ) photosensitized by [Pt(Thpy)(PPh₃)X]ⁿ⁺ and [Pt(Thpy)(HThpy)Y]ⁿ⁺ complexes were obtained by time-resolved laser measurements of near-infrared luminescence signal (1270 nm) intensity of ¹O₂ as a function of absorption at a 355-nm excitation wavelength in deuterated acetonitrile and dichloromethane (Table 1). Figure 1 depicts a typical linear graph of ¹O₂ emission intensity against absorption of selected sensitizers of 1 and 7 in CD₃CN; the singlet oxygen quantum yields were calculated with reference to methylene blue. Similar plots were obtained in CD₂Cl₂ (see Supplementary Materials) using C₆₀ (Φ_Δ = 1) as a reference compound. Triplet-triplet annihilation was negligible at these concentrations, as evidenced by the fact that the ¹O₂ intensity showed a linear correlation with the absorption of the complexes. The values of ¹O₂ quantum yields for complexes 1–7 in CD₃CN range from 0.42 ± 0.03 to 0.95 ± 0.05 as listed in Table 1.

Table 1.

Summary of spectroscopic data: luminescence lifetime (τ), emission quantum yield (Φ), ¹O₂ quantum yield (Φ_Δ), and quenching rate of ¹O₂ (k_T).

Complex	aτ/μs		aΦ (λmax/nm)	bΦΔ		c kT/107 M−1s−1
	N2	Air		CD3CN	CD2Cl2
[Pt(Thpy)(PPh3)Cl], 1	6.8	–d	0.020 (558)	0.44±0.02	0.12±0.03	1.40±0.04
[Pt(Thpy)(PPh3)CH3CN]ClO4, 2	26.2	–d	0.17 (555)	0.75±0.04	0.27±0.02	1.26±0.05
[Pt(Thpy)(PPh3)py]ClO4, 3	6.73	0.26	0.056 (556)	0.82±0.06	0.55±0.05	1.46±0.04
[Pt(Thpy)(HThpy)Cl], 4	12.0	–d	0.19 (556)	0.56±0.04	0.31±0.03	3.64±0.04
[Pt(Thpy)(HThpy)py]ClO4, 5	24.0	0.33	0.38 (554)	0.95±0.05	0.27±0.03	5.54±0.03
[(PtThpyCl)2Calix(OnBu)], 6	1.38	–d	0.021 (559)	0.42±0.03	0.15±0.03	0.91±0.11e
[{PtThpy(CH3CN)}2Calix (OnBu)](ClO4)2, 7	25.2	0.35	0.20 (556)	0.81±0.06	0.26±0.02	1.69±0.02

^aMeasured in CH₃CN.

^bDetermined with 355 nm excitation using methylene blue (0.57) in CD₃CN and C₆₀ (1.00) in CD₂Cl₂ as references.

^cMeasured in CD₃CN unless otherwise stated.

^dτ < 0.1 μs.

^eDetermined in CD₂Cl₂.

Figure 1.

A plot of ¹O₂ emission intensity against absorption with excitation wavelength of 355 nm using sensitizers (1 and 7) in CD₃CN compared to that of methylene blue as a reference.

Singlet oxygen quenching by ground-state sensitizer may limit the usefulness of the singlet oxygen sensitizer. We therefore determined the total rate constants (k_T) of ¹O₂ quenching by the ground state of 2-(2′-thienyl)pyridyl platinum(II) complexes as summarized in Table 1. The second order quenching rate constants were obtained from the slopes of linear least-squares fits to the data. The rate constants of the quenching of ¹O₂ by 1–7 are in the range of (9.1 × 10⁶ – 5.56 × 10⁷) M⁻¹ s⁻¹, where relatively higher quenching rate constants were observed from [Pt(Thpy)(HThpy)Y]ⁿ⁺ systems (4 and 5). The rate constants of singlet oxygen quenching by [Pt(Thpy)]⁺-modified calixarene receptors (6 and 7) were observed to be 9.0 × 10⁶ and 1.70 × 10⁷ M⁻¹ s⁻¹ respectively, which are comparable to those for their mononuclear congeners, 1 (k_T = 1.40 × 10⁷ M⁻¹ s⁻¹) and 2 (k_T = 1.26 × 10M⁻¹ s⁻¹). Overall the values for k_T are quite similar to those of standard organic photosensitizers (e.g. TPP, k_T = 2 × 10 M⁻¹ s⁻¹).

To evaluate the potentials of these cyclometalated platinum(II) complexes in photobiology, studies on their cytotoxicity, photocytotoxicity, DNA and protein binding, have been performed. Complexes [Pt(Thpy)(PPh₃)py]ClO₄ (3) and [Pt(Thpy)(HThpy)py]ClO₄ (5) with relatively high quantum yields of singlet oxygen production were chosen for cytotoxicity studies towards HeLa cancer cells (human cervical epithelioid carcinoma). The half maximal inhibitory concentration (IC₅₀) of 3 in the dark is 21.07 μM. After exposure to visible light, its cytotoxicity was significantly enhanced with the IC₅₀ value decreased to 3.29 μM. Similarly, photocytotoxicity was observed for 5 with the IC₅₀ in the dark of 6.12 μM and 1.78 μM after visible light irradiation (Fig. 2).

Figure 2.

Plot of percent viability of HeLa cancer cells as a function of concentration of 5 in the dark (black line) and after irradiation with visible light (red line).

The emission spectrum of 5 (λ_ex = 400 nm; concentration = 1 × 10⁻⁵ mol dm⁻³) in Tris buffer solution at room temperature shows a minor band at 464 nm and a predominant slightly vibronic structured band with peak maxima at 553 and 596 nm. Upon gradual addition of calf thymus DNA (3 equivalents per interval up to 18 equivalents) to a Tris buffer solution of 5, the emission intensities at both λ_max of 464 and 553 nm increase. Similar emission spectral traces of 5 at different equivalents of bovine serum albumin (BSA) were performed in PBS solutions at 298 K (λ_ex = 400 nm, concentration = 1 × 10⁻⁵ M). With addition of 0.1 equivalent of BSA per interval, the emission at 462 nm shows minute increase in intensity, whereas the lower-energy emission bands are slightly red-shifted to 557 and 601 nm and exhibit enhancement in intensity by ~6 times at 0.6 equivalent of BSA from that before the addition of BSA.

We also examined the cellular uptake and localization of 5 in live cells. Fluorescence microscopic images of HeLa cancer cells after incubation with 5 for 1 h revealed noticeable appearance of discrete intracellular fluorescence signals (Figs. 3a and 4a). Staining of cells with fluorescent probe for nuclear DNA (Hoechst 33342, Fig. 3b) or mitochondria (MitoTracker Red, Fig. 4b), demonstrated that 5 was colocalized to the nucleus as well as to mitochondria.

Figure 3.

Live cell images of HeLa cells incubated with 5 (5 μM) at 37 °C for 1 h showing (a) fluorescence emitted by 5; (b) nucleus stained with Hoechst 33342. (c) Overlay of fluorescence images of (a) and (b).

Figure 4.

Live cell images of HeLa cells incubated with 5 (5 μM) at 37 °C for 1 h showing (a) fluorescence emitted by 5; (b) mitochondria stained with MitoTracker Red. (c) Overlay of fluorescence images of (a) and (b).

DISCUSSION

Previous studies on singlet oxygen production by cyclometalated complexes have primarily focused on the nature of the encounter complex between triplet dioxygen and the excited metal complex. In many cases, the excited state redox potential of the metal complex may play a major role in determining how much singlet oxygen is produced (20, 21). This study shows that ancillary ligands can also drastically increase the amount of singlet oxygen produced by cyclometalated complexes: the quantum yield of ¹O₂ production (Φ_Δ) depends on the quantum yield of triplet-state formation. Through comparisons of emission quantum yields (Φ) of complexes 1–7 with their respective ¹O₂ quantum yields (Φ_Δ), a general trend of increase in Φ_Δ can be obtained with increasing Φ. The ¹O₂ quantum yields (Φ_Δ) of 1–3 follow the order of X = Cl⁻ < CH₃CN ~ pyridine, increasing from (0.44 ± 0.02) in 1 to (0.75 ± 0.04) in 2 and (0.82 ± 0.06) in 3. Similar observations have been made for 6 and 7, where the respective Φ_Δ in CD₃CN is enhanced from 0.42 ± 0.03 (6) to 0.81 ± 0.06 (7), when the Cl⁻ ligand is changed to CH₃CN. Increasing the strength of Y by changing Cl⁻ to pyridine in complexes 4 and 5, formulated as [Pt(Thpy)(HThpy)Y]ⁿ⁺, results in a significant increase in the ¹O₂ quantum yield (Φ_Δ) from 0.56 ± 0.04 (4) to 0.95 ± 0.05 (5) in CD₃CN. On the other hand, replacing the PPh₃ ligand in [Pt(Thpy)(PPh₃)Cl] (1) with the HThpy ligand (Complex 4, [Pt(Thpy)(HThpy)Cl]) gives only a minor enhancement in Φ_Δ in CD₃CN from 0.44 ± 0.02 (1) to 0.56 ± 0.04 (4). Similar observations were made for the complexes (3) and [Pt(Thpy)(PPh₃)py]⁺ [Pt(Thpy)(HThpy)py]⁺ (5), where again the two complexes differ from each other by 3 bearing the PPh₃ ligand and 5 having the HThpy ligand. The Φ_Δ of 3 and 5 in CD₃CN are comparable being 0.82 ± 0.06 (3) and 0.95 ± 0.05 (5), respectively.

Φ_Δ can also be influenced by the solvent; for example, Pt(Thpy)(HThpy)py]ClO₄ (5) shows substantial variation in Φ_Δ when measured in CD₃CN (0.95 ± 0.05) and CD₂Cl₂ (0.27 ± 0.03). Incidentally, the photoluminescent characteristics of 5 is also solvent sensitive; the high-energy absorption maximum shows a bathochromic shift from λ_max at 291 nm (ε = 24200 M⁻ cm⁻¹) in CH₂Cl₂ to 305 nm (ε = 24100 M⁻ cm⁻¹) in DMF, whereas the low-energy absorption band at λ > 390 nm displays similar solvatochromism from 393 nm (ε = 5240 M⁻¹ cm⁻¹) in CH₂Cl₂ to 403 nm (ε = 7750 M⁻ cm⁻¹) in DMF. In general, the ¹O₂ quantum yields (Φ_Δ) of 1–7 recorded in CD₂Cl₂ show diminished value of (0.15 ± 0.03 –0.55 ± 0.05) as compared to those in CD₃CN (0.42 ± 0.03 – 0.95 ± 0.05).

In summary, the ability of [Pt(Thpy)(PPh₃)X]ⁿ⁺ and [Pt(Thpy)(HThpy)Y]ⁿ⁺ complexes to generate singlet oxygen varies with the auxiliary ligand X or Y. Complex 5 with high Φ (0.95 ± 0.05) exhibited binding to DNA and protein, noticeable photocytotoxicity, and predominant cellular localization in the nucleus and mitochondria, revealing the potential for new development of target-specific photosensitizers (41). Furthermore, in acetonitrile, the calixarene derivatives 6 and 7 still produce singlet oxygen in moderate to high yield. Reports on the host-guest complex formation of calixarenes or cyclodextrins with neutral hydrophobic guests are well documented in the literature (42–44). Hence it may be possible to bind a variety of substrates to the calixarenes in complexes 6 and 7, and subsequently react these substrates with singlet oxygen produced by the cylometalated Pt(II) moieties. Studies in this direction are in progress.