Synthesis and Characterization of Ru(II) Complexes with π-Expansive Imidazophen Ligands for the Photokilling of Human Melanoma Cells

Abstract

Ru(II) complexes were synthesized with π-expanding (phenyl, fluorenyl, phenanthrenyl, naphthalen-1-yl, naphthalene-2-yl, anthryl, and pyrenyl groups) attached at a 1H-imidazo[4,5-f][1,10]phenanthroline ligand and 4,4′-dimethyl-2,2′-bipyridine (4,4′-dmb) co-ligands. These Ru(II) complexes were characterized by 1D and 2D NMR, and mass spectroscopy, and studied for visible-light and dark toxicity to human malignant melanoma SK-MEL-28 cells. In the SKMEL-28 cells, the Ru(II) complexes are highly phototoxic (EC₅₀ = 0.2–0.5 μM), and have low dark toxicity (EC₅₀ = 58–230 μM). The highest phototherapeutic index (PI) of the series was found with the Ru(II) complex bearing the 2-(pyren-1-yl)-1H-imidazo[4,5-f][1,10]phenanthroline ligand. This high PI is in part attributed to the π-rich character added by the pyrenyl group, and a possible low-lying and longer-lived ³IL state due to equilibration with the ³MLCT state. While this pyrenyl Ru(II) complex possessed a relatively high quantum yield for singlet oxygen formation (Φ_Δ = 0.84), contributions from type-I processes (oxygen radicals and radical ions) are competitive with the type-II (¹O₂) process based on effects of added sodium azide and solvent deuteration.

INTRODUCTION

A number of contributions have been made to understanding metal-organic dyad sensitizers in photodynamic therapy (PDT) and synergy in cancer cell killing (1–4), where ligand substitution on the metal serves to increase light absorption and triplet lifetime (5). In systems based on Ru(II) as well as other metals (6,7), excited states with intraligand (IL) character have been shown to increase the triplet lifetime (5). Whether these photosensitizers involve predominantly a type II reaction producing ¹O₂ as the key cytotoxic agent, or whether there is competition with the type I reaction (oxygen radicals or radical ions) (8,9) has not been fully examined. It is plausible that both pathways contribute to the observed photocytotoxic effects due to the various types of excited states that are accessible in metal complexes.

Depending on the natue of the metal complex, some of these excited state configurations include: metal-to-ligand charge transfer (MLCT), metal centered (MC), metal-to-metal charge transfer (MMCT), in the case of multimetallic complexes, IL, intraligand charge transfer (ILCT), and ligand-to-ligand charge transfer (LLCT). We previously found that the ligand centered (³IL) state (10) in the π-expansive organic unit extended the triplet lifetime 10-fold due to an equilibrium with ³MLCT state (11). When the energy of the ³IL state is sufficiently low compared to the ³MLCT state, its interaction with the ³MLCT state becomes almost negligible, thereby forming a pure ³IL state with an even longer lifetime. For example, a triplet lifetime of >100 μs was found by Kozlov et al. when pyrenylethynylene-conjugated bipyridine was incorporated in a Ru complex (12). In a similar vein, Ru(II) complexes that incorporated pyrenylethynylene conjugated to phenanthroline into the dyad were reported to extend the lifetime of the triplet state to 240 μs, which may be the longest lifetime reported to date for Ru complexes (13). Previous reports on Ru(II) dyads derived from 2,2′-dimethyl-2,2′-bipyridine (4,4′-dmb) showed high phototoxicity with low dark toxicity (14,15), making these metal-organic dyads attractive photosensitizers for PDT.

In the present study, we explore complexes of the type [Ru(4,4′-dmb)²(ip-R)]²⁺, where imidazo[4,5-f][1,10]phenanthroline (ip) is fused to an aromatic R group. We hypothesized that increasing the π-conjugation in the R group will systematically lower the energy of the ³IL state relative to the ³MLCT state and thus improve phototoxic effects toward human melanoma SKMEL-28 cells, particularly when the ³IL energy drops below that of the ³MLCT. The R groups investigated were: 2-phenyl (1), 2-(9H-fluoren-2-yl) (2), 2-(phenanthren-9-yl) (3), 2-(naphthalen-1-yl) (4), 2-(naphthalene-2-yl) (5), 2-(anthracene-10-yl) (6), and 2-(pyren-1-yl) (7) along with ruthenium 4,4′-dimethyl-2,2′-bipyridine (4,4′-dmb) co-ligands (Figure 2). We determined whether complexes 1–7: (i) could be synthesized in high yield, (ii) would lead to high phototherapeutic indicies (PI) in malignant melanoma SK-MEL-28 cells, (iii) would have high singlet oxygen quantum yields (Φ_Δ), and (iv) would show evidence in both type I and type II sensitized oxidation mechanisms. Our results have identified 7, with pyrenyl as the R group, as a potentially useful photosensitizer against melanoma cells.

Figure 2.

Structures of Ru(II) complexes 1–7 bearing functional ligands (ip-R).

RESULTS AND DISCUSSION

Synthesis and Characterization.

A two-step synthesis was carried out to reach Ru(II) complexes 1–7 (Figure 3). Step 1: 1,10-phenanthroline-5,6-dione reacted with NH₄OAc, in glacial acetic acid, and an appropriately substituted aldehyde under microwave irradiation at 180 °C for 15 min to obtain ip–R_1–7 in 50–80% yields. Step 2: [Ru(4,4′-dmb)₂Cl₂]∙2H₂O reacted with ip–R_1–7 under microwave irradiation in ethylene glycol at 180 °C for 15 min to obtain a crude chloride salt of Ru(II) complexes 1–7. These chloride salts were purified by column chromatography using acetonitrile and water with KNO₃, which led to their isolation as mixtures of Cl⁻ and NO₃⁻ salts. The mixtures were then converted to the corresponding PF₆⁻ salts by introducing a saturated KPF₆ solution. The Ru(II) complexes 1–7 as PF₆⁻ salts were formed in 23–55% yields as red or orange solids.

Figure 3.

Two-step synthesis of Ru(II) complexes 1–7.

Proton NMR assignments for the phenanthroline unit of ip have been reported earlier (16). However, 2-(phenanthren-9-yl)-1H-imidazo[4,5-f][1,10]phenanthroline (pcip) was not previously characterized. Here, we assign the pcip protons a and d as the most downfield protons because these are connected to a carbon which is directly attached to nitrogen (Figure 4 and Figure S1). The close proximity to imidazolium nitrogen brings f and c signals downfield compared to the b and e signals. The singlet for g appeared at 8.48 ppm, and a doublet at 8.19 was tentatively assigned for k. Protons b and e appeared as a multiplet within 7.92–7.73 ppm. The other protons o, l, n, m, i, and j appeared as a multiplet.

Figure 4.

Proton NMR assignments of 2-(phenanthren-9-yl)-1H-imidazo[4,5-f][1,10]phenanthroline (pcip).

The characterization of Ru(II) complex 1 and 7 has been previously reported (16). Therefore, the characterization of Ru(II) complexes 2-6 are reported here. Based on HPLC data, the purity of 2 was 98%, of 3 was 98%, of 4 was 96%, of 5 was 99%, and of 6 was 90%. The LCMS data indicated that 2 contained a fluorenyl group [(ESI+) m/z for C₅₀H₄₀N₈Ru: calcd 427.1209, found 427.1204], 3 contained a phenanthrenyl group [(ESI+) m/z for C₅₁H₄₀N₈Ru: calcd 433.1209, found 433.1209], 4 contained a naphthyl group [(ESI+) m/z for C₄₇H₃₈N₈Ru: calcd 408.1131, found 408.1126], 5 contained a naphthyl group [(ESI+) m/z for C₄₇H₃₈N₈Ru: calcd 408.1131, found 408.1126], and 6 contained an anthryl group [(ESI+) m/z for C₅₁H₄₀N₈Ru: calcd 433.1209, found 433.1204].

For Ru(II) complexes 1–7, protons c and f in phenanthroline moiety were assigned as the most downfield protons at 9.13–8.89 ppm, because they are far from Ru electron cloud and closer to imidazolium nitrogens (Figures S2–S7). However, the signals for c and f were slightly upfield shifted for 6 and appeared within 8.37–8.32 ppm. The protons b and e in phenanthroline moiety of complexes 1–7 that appeared within 7.93–7.63 ppm were assigned based on their correlation with c and f. For all complexes, a and d appeared as a doublet within 8.26–7.97 ppm. Downfield chemical shifts were observed for the aromatic protons attached to the 4,4′-dmb unit in the Ru(II) complexes in the following order: 3′,3 > 6′,6 > 5′,5 as reported previously in the literature (17). For 2, methylene proton h appeared as a singlet at 4.14–4.13 ppm. Other aromatic protons on the fluorinone moiety of these complexes appeared in the range 8.55–7.38 ppm. Doublet for m attached to napthalene chromophore in 5 appeared at 8.85 ppm, although this proton does not show an interaction with other protons in ¹H-¹H COSY, this doublet could be the result of long-range four bond coupling. A singlet for g in 3 appeared at 8.55 ppm, and indeed all other aromatic protons of the phenanthrene unit in 3 appeared within 8.29–7.75 ppm. Protons o, m, and k were assigned as the most downfield protons and appeared as doublet, doublet, and singlet at 9.29 (for 3), 9.18 (for 4), and 8.88 ppm (for 6), respectively. Protons k does not show correlation with another proton in 6; whereas, o and m showed correlation with n and l, respectively.

Photokilling of melanoma SK-MEL-28 cells.

Table 1 shows the phototoxic effects of Ru(II) complexes 1–7 in human malignant melanoma SK-MEL-28 cells. The cells were incubated with 1 nM to 300 μM complex 1–7 prior to irradiation with visible light (400–700 nm) in a 16 h drug-to-light interval. The phototoxicity results were obtained with a fluence of 100 J cm⁻², in which the highest concentration tested was 300 μM. Dark toxicities of Ru(II) complexes 1–7 were fairly low (ranging from 59–230 μM). We remind the reader that compounds with EC₅₀ dark toxicities of >300 μM are generally considered to be nontoxic (4). In our case, the phototoxicities of Ru(II) complexes 1–7 were found to be in the tenths of micromolar range (0.24–0.46 μM). Ru(II) complex 7 had a particularly low dark toxicity and a fairly good phototoxicity resulting in the highest PI (of 657) in the series. The PI obtained for Ru(II) complex 6 is 409, which had an anthracene substituent. Subtle structural differences between naphth-1-yl and naphth-2-yl led to quite large differences in phototoxicities in the SK-MEL-28 cells, where the PI of 4 was about double that of 5. The PI value for Ru(II) complex 1 was 558, which was second highest in the series. The relationship between the size of the substituent and phototoxicity of Ru(II) complexes 1–7 was only marginally correlated. For example, the pyrenyl-substituted Ru(II) complex 7 had a PI of only ~100 greater than the phenyl-substituted Ru(II) complex 1. Another example is the slightly improved dark toxicity of 7 compared to 4, where the phototoxicity is slightly better for 4 than 7.

Table 1.

SK-MEL-28 cell photocytotoxicities of the new Ru(II) complexes 1–7 and Ru(II) complexes 8-11 previously reported in the literature.

Ru(II) complexes	dark EC50 (μM)a	light EC50 (μM)a	PIb
1	143 ± 10	0.26 ± 0.01	558
2	82 ± 1	0.46 ± 0.02	179
3	59 ± 4	0.25 ± 0.01	232
4	133 ± 8	0.26 ± 0.04	505
5	96 ± 8	0.35 ± 0.02	277
6	98 ± 21	0.24 ± 0.01	409
7	230 ± 11	0.35 ± 0.02	657
8c,d	>300	14.9 ± 0.2	>20
9c,d	140 ± 10	0.07 ± 0.01	2071
10d	58.5 ± 0.5	0.35 ± 0.01	650
11e	49.9 ± 0.3	0.125 ± 0.003	400

^a

Experiments were conducted in triplicate. Dark EC₅₀ and light EC₅₀ values were derived from eq 1 using a dose-response curve of Graph Pad Prism 6.0, which are shown in the Supporting Information.

$$y=y_i+\frac{y_i-y_f}{1+{10}^{\left({\text{logEC}}_{50}-x\right)\times (\text{Hillslope})}}$$ (1)

^bPhototherapeutic indicies (PIs) were calculated as the ratio of the dark EC₅₀ and light EC₅₀ values.

^cReference 19.

^dReference 20.

^eReference 21.

We believe there is value in comparing the new Ru(II) complexes 1–7 to several Ru(II) complexes reported in the literature (19–21). Compounds 8-11 are Ru-based dyads containing π-expansive ligands combined with relatively small and substitutionally inert coligands bpy or 4,4′-dmb. As seen in Table 1, changing either the coligand and/or the π-expansive functional ligand can have a marked effect on the dark and phototoxicities as well as the PIs, but not necessarily in a systematic way. Ru(II) complex 7 ranks among the least dark toxic in this comparison (surpassed only by 8) and displays similar phototoxic effects as 10 and 11 (surpassed only by 9). While its PI is not as large as that of 9, the PI measured for 7 compares favorably with 10, being among the two largest in the comparison. Ru(II) complex 7 does have a clear advantage over its counterpart with a similar PI in that 7 is almost four times less dark toxic, which is of consideration for a phototherapeutic agent.

Quantum yield for singlet oxygen formation (Φ_Δ).

Next, the singlet oxygen quantum yields for Ru(II) complexes 1–7 were measured. Table 2 shows that the Φ_Δ values ranged from a high of 0.87 for 6 to a low of 0.58 for 1, which compare favorably to literature values for 8 (0.17) (19,20), 9 (0.79) (19,20), 10 (0.78) (20), and 11 (0.74) (21). For the new Ru(II) complexes, interestingly, we found that the quantum yields increase in the order 1 ≈ 2 ≈ 3 < 4 ≈ 5 < 6 ≈ 7 and very roughly correlate with the decrease of the triplet energies (E_T) of the π-group (R₁–R₇) alone. The triplet energies of the π-groups (R₁–R₇) are in the range of 3.66–1.82 eV. These π-groups (R₁–R₇) decrease in the order benzene > fluorene > phenanthrene ≈ naphthalene > pyrene > anthracene. For comparison, to the π-groups (R₁–R₇), the energy of the MLCT state for [Ru(bpy)₃]²⁺ complex is reported to be 2.1 eV (18). While the high Φ_Δ’s for Ru(II) complexes 1–7 and low E_T’s for the π-groups (R₁–R₇) point to a type-II process, contributions from type-I processes may also exist, as was investgated next with Ru(II) complex 7.

Table 2.

Singlet oxygen quantum yields (Φ_Δ) for Ru(II) complexes 1–7, and literature triplet energies (E_T) of the π-group (R₁–R₇) alone.

Ru(II) complex	ΦΔa	π-group (R1–R7) alone	ET (eV)b
1	0.58	benzene	3.66
2	0.59	fluorene	2.92
3	0.60	phenanthrene	2.69
4	0.69	naphthalene	2.64
5	0.67	naphthalene	2.64
6	0.87	anthracene	1.82
7	0.84	pyrene	2.08

^a

Singlet oxygen quantum yields (Φ_Δ) for 1–7 were measured in aerated acetonitrile using eq 2 relative to [Ru(bpy)₃](PF₆)₂ (Φ_Δ = 0.57 in aerated acetonitrile) (22,23) and contains ±5% error. See Materials and Methods section for more information on the Φ_Δ measurement.

$${\mathrm{\varphi }}_s={\mathrm{\varphi }}_r\left(\frac{I_s}{A_s}\right)\left(\frac{A_r}{I_r}\right)\left(\frac{{\mathrm{\eta }}_s^2}{{\mathrm{\eta }}_r^2}\right)$$ (2)

^bE_T values of benzene (24), fluorene, phenanthrene, napthalene (25), pyrene (26), and anthracene (27) were obtained from the literature.

Mechanistic considerations.

We next carried out experiments with Ru(II) complex 7 to assess whether its mechanism is dominated by type I (oxygen radicals or radical ions) or type II (¹O₂) processes. Data were collected with 9,10-anthracene dipropionate dianion 12 as a trapping agent for ¹O₂ by a [4 + 2] cycloaddition reaction, in which the product of the trapping reaction is 9,10-endoperoxide 13 (Figure 6) (28). Anthracene 12 has been used to verify the presence of ¹O₂ in aqueous solution and is a good trapping agent based on a comparison to trans-2-methyl-2-pentenoate anion that undergoes an ‘ene’ reaction and is considered to be a fingerprint for ¹O₂ (28,29). Here, anthracene 12 was used as it requires UV-vis instead of NMR and also a much lower concentration than the trans-2-methyl-2-pentenoate anion, another popular trap for ¹O₂ detection. Our studies were carried out in H₂O or D₂O, and in the presence or absence of sodium azide (Table 3). We find that the efficiency of ¹O₂ trapping by 12 increased by changing the solvent from H₂O and in D₂O. However, the increase in yield of endoperoxide 13 was 2-fold in D₂O compared to H₂O. Because the lifetime of ¹O₂ (τ_Δ) is 30-fold longer in D₂O vs H₂O (30), a greater difference in ¹O₂ trapping in D₂O vs H₂O would have been expected if the Ru(II) complex 7 was mainly serving as a type-II sensitizer. We also found only a modest decrease in the ¹O₂ trapping by 12 when sodium azide (a known ¹O₂ quencher) was added to the reaction. Based on these results, we propose that Ru(II) complex 7 contains contributions from both type-I and type II processes.

Figure 6.

Reactive oxygen intermediates are generated by Ru(II) complex 7 upon illumination with white light in D₂O. The ¹O₂ produced is detected by its reaction with 9,10-anthracene dipropionate dianion 12 to reach endoperoxide 13.

Table 3.

H₂O/D₂O solvent and sodium azide quenching photochemical effects with Ru(II) complex 7 on the trapping of ¹O₂ by antracene 12.

solvent	yield 13 (μM)a	yield 13 (μM)b (solvent+NaN3)	ratio 13D2O/13H2O	ratio 13D2O/13D2O+NaN3	ratio 13H2O/13H2O+NaN3
D2O	28.1±0.9	19.8±1.0	2.0	1.5	0.80
H2O	13.9±3.7	16.7±2.6

^aIrradiation (15 min) of a biphasic system containing 82 μM of 12 in D₂O or H₂O, and 63.7 μM of Ru(II) complex 7 in CDCl₃.

^bEndoperoxide 13 yield in presence of 1-mM sodium azide (NaN₃) added as a singlet oxygen quencher.

What are the mechanistic implications of these results?

Our finding for Ru(II) complex 7 is similar to an earlier study that showed Ru(II) complexes with α-oligothiophene groups contained contributions from both type I and type II processes (14). While type-I and type II processes are oxygen dependent, contributions from Ru(II) complexes may also include oxygen-independent reactions. However in the current study, we did not explore the photokilling of the SK-MEL-28 cells under oxygen-free conditions to assess its potential contribution.

CONCLUSION

Herein, we report on Ru(II) dyads tethered to R groups that differ in π-conjugation in an attempt to increase the PI in a rational manner. Specifically, we synthesized Ru(II) complexes 1–7 by covalent attachment of phenyl, fluorenyl, phenanthrenyl, naphthalen-1-yl, naphthalene-2-yl, anthryl, and pyrenyl groups to ip, which led to stable Ru(II) complexes 1–7. These complexes were synthesized in modest yields (33–78%). As expected, the higher ¹O₂ quantum yields roughly correlated with lower triplet-state energies of the R groups. We found Ru(II) complex 7 to be the most promising dyad of the series, with a relatively high value for Φ_Δ (0.84) and a large PI toward SK-MEL-28 cells, which included contributions from both type I and type II processes based on the effects of added sodium azide and solvent deuteration.

MATERIALS AND METHODS

Materials.

1,10-Phenanthroline-5,6-dione, NH₄OAC, glacial acetic acid, benzaldehyde, 9H-fluorene-2-carbaldehyde, phenanthrene-9-carbaldehyde, 1-napthaldehyde, 2-napthaldehyde, 9-anthraldehyde, pyrene-1-carbaldehyde, 9H-fluorene-2-carbaldehyde, phenanthrene-9-carbaldehyde, 4,4′-dimethyl-2,2′-bipyridine (4,4′-dmb), potassium nitrate (KNO₃), acetonitrile (MeCN), potassium hexafluoro phosphate (KPF₆), dichloromethane (DCM), methanol (MeOH), and deuterated solvents (DMSO-d₆ and CD₃CN) were purchased from Sigma-Aldrich and used without further purification. Fetal bovine serum (FBS), RPMI 1640 (Corning Cellgro), and Eagle’s Minimum Essential Medium (EMEM) (Corning Cellgro) were purchased from VWR. Human malignant melanoma cells (SK-MEL-28) were obtained from the American Type Culture Collection (ATCC). Prior to use, FBS was divided into 40 mL aliquots, heat inactivated (30 min, 55 °C), and then stored at −20 °C. Deionized water of 18 MΩ cm resistivity from a Barnstead filtration system was used for the biological experiments.

Instrumentation.

Microwave reactions were performed in a CEM Discover microwave reactor. NMR spectra were collected using Bruker AVANCE 500 (Dalhousie University Nuclear Magnetic Resonance Research Resource) or 300 MHz spectrometers (Acadia Centre for Microstructural Analysis), and ESI mass spectra were obtained using a Bruker microTOF focus mass spectrometer (Dalhousie University Mass Spectrometry Laboratory). HPLC analyses were performed with an Agilent 1100 series instrument (ChemStation Rev. A 10.02 software) using a Hypersil GOLD C18 reverse phase column with an A-B gradient (98% → 5% A; A = 0.1% formic acid in H₂O, B = 0.1% formic acid in MeOH).

Synthesis.

The ligands 2-phenyl-1H-imidazo[4,5-f][1,10]phenanthroline (bip), 2-(naphthalen-1-yl)-1H-imidazo[4,5-f][1,10]phenanthroline (1-nip), 2-(naphthalen-2-yl)-1H-imidazo[4,5-f][1,10]phenanthroline (2-nip), 2-(9H-fluoren-2-yl)-1H-imidazo [4,5-f][1,10] phenanthroline (fip), 2-(pyren-1-yl)-1H-imidazo[4,5-f][1,10]phenanthroline (ippy), and 2-(anthracene-10-yl)-1H-imidazo[4,5-f][1,10]phenanthroline (aip) were synthesized according to literature procedures (31–36). Compounds [Ru(4,4′-dmb)₂Cl₂]∙2H₂O (37), Ru(4,4′-dmb)₂(bip)](PF₆)₂ 1 (38), Ru(4,4′-dmb)₂(ippy)](PF₆)₂ 7 (22), and 9,10-anthracene dipropionate dianion 8 (28,29) were synthesized as reported in the literature. For our in vitro study, Ru(II) complexes 1–7 were converted to their corresponding Cl⁻ salts.

Synthesis of 2-(phenanthren-9-yl)-1H-imidazo [4,5-f][1,10] phenanthroline (pcip).

Yield 247 mg (78%): purity >80%. A mixture of phen-dione (166.6 mg, 0.8 mmol), phenanthrene-9-carbaldehyde (165 mg, 0.8 mmol), NH₄OAC (1.23 g, 16 mmol), and glacial acetic acid (4 mL) was subjected to microwave irradiation for 15 min at 180 °C. The resulting solution was diluted with water in a beaker, and ammonium solution was added dropwise until the pH of the solution reached 8. A yellow precipitate was formed, which was filtered through a fine glass-sintered frit. Single spot in TLC indicated pure product was formed. ¹H NMR (300 MHz, DMSO-d₆) δ 9.08 (d, J = 5.0 Hz, 2H; a,d), 8.99 (m, 4H; c,f,o,l), 8.48 (s, 1H; g), 8.19 (d, J = 7.7 Hz, 1H; k), 7.92–7.73 (m, 6H; b,e,n,m,i,j).

Synthesis of [Ru(4,4′-dmb)₂(fip)](PF₆)₂ 2.

Yield 55 mg (48%): purity 98%. In a microwave vessel, 63.4 mg (0.11 mmol) of [Ru(4,4′-dmb)₂Cl₂]∙2H₂O was mixed with 38.4 mg (0.10 mmol) of FIP. Ethylene glycol (3.0 mL) was added to the vessel. The mixture was microwaved for 15 min at 180 °C. The resulting dark red color solution was directly loaded on to a silica column and eluted using 100% MeCN followed by 5% H₂O in MeCN to remove the excess starting material and other impurities. Finally, Ru(II) complex 2 was eluted as an orange band with 5–10% H₂O and 0.5–2.5% KNO₃ in MeCN to give an orange solid as a mixture of Cl⁻ and NO₃⁻ salts. This mixture was converted to the PF₆⁻ salt by adding 5–10 mL of water and 2–4 mL of saturated KPF₆. Evaporation of DCM under reduced pressure gave the pure PF₆⁻ salt as an orange solid: R_f = 0.51 (10% H₂O + 2.5% KNO₃ in MeCN). HPLC retention time: 25.6 min. ¹H NMR (300 MHz, acetonitrile-d₃) δ 9.00 (s, 2H; f, c), 8.52 (s, 1H; g), 8.40 (s, 2H; 3’), 8.36 (s, 2H; 3), 8.32 (s, 1H, n), 8.09 (d, J = 7.9 Hz, 1H; m), 8.04 (d, J = 5.2 Hz, 2H; a,d), 7.97 (d, J = 6.7 Hz, 1H; l), 7.79 (dd, J = 8.3, 5.2 Hz, 2H; b,e), 7.67 (m, 3H; 6’, i), 7.51 – 7.38 (m, 4H; 6, k, j), 7.30 (d, J = 5.8 Hz, 2H; 5’), 7.06 (d, J = 7.1 Hz, 2H; 5), 4.13 (s, 2H; h), 2.59 (s, 6H, 4’-Me), 2.48 (s, 6H, 4-Me). MS (ESI+) m/z: 999.13 [M-PF₆]⁺, 853.47 [M-PF₆-H]⁺, 427.40 [M–2PF₆]²⁺. HRMS (ESI+) m/z for C₅₀H₄₀N₈Ru: calcd 427.1209, found 427.1204.

Synthesis of [Ru(4,4′-dmb)₂(pcip)](PF₆)₂ 3.

Yield 38 mg (33%): purity 98%. In a microwave vessel, 63.4 mg (0.11 mmol) of [Ru (4,4′-dmb)₂Cl₂]∙2H₂O was mixed with 39.6 mg (0.10 mmol) of PCIP. Ethylene glycol (3.0 mL) was added to the vessel. The mixture was microwaved at 180 °C for 15 min. The resulting dark red color solution was directly loaded on to a silica column and eluted using 100% MeCN followed by 5% H₂O in MeCN to remove the excess starting material and other impurities. Finally, Ru(II) complex 3 was eluted as an orange band with 5–10% H₂O and 0.5–2.5% KNO₃ in MeCN to give an orange solid as a mixture of Cl⁻ and NO⁻₃ salts. This mixture was converted to the PF₆⁻ salt by adding 5–10 mL of water and 2–4 mL of saturated KPF₆. Evaporation of DCM under reduced pressure gave the pure PF₆⁻ salt as an orange solid: Rf = 0.64 (10% H₂O + 2.5% KNO₃ in MeCN). HPLC retention time: 24.9 min.¹H NMR (500 MHz, acetonitrile-d₃) δ 9.29 (d, J = 8.2 Hz, 1H; o), 9.10 (d, J = 8.2 Hz, 2H; f, c), 8.93 (d, J = 8.2 Hz, 1H; l), 8.86 (d, J = 8.4 Hz, 1H; k), 8.55 (s, 1H; g), 8.43 (s, 2H; 3’), 8.38 (s, 2H; 3), 8.15 (d, J = 7.8 Hz, 1H; h), 8.07 (dd, J = 5.2, 1.3 Hz, 2H; a, d), 7.87–7.75 (m, 6H; b, e, n, m, i, j), 7.71 (d, J = 5.7 Hz, 2H; 6’), 7.46 (d, J = 5.9 Hz, 2H; 6), 7.33 (dd, J = 6.4, 1.5 Hz, 2H; 5’), 7.10 (dd, J = 5.8, 1.5 Hz, 2H; 5), 2.61 (s, 6H; 4’-Me), 2.51 (s, 6H; 4-Me). MS (ESI+) m/z: 1011.00 [M-PF₆]⁺, 865.33 [M–PF₆–H]⁺, 433.33 [M–2PF₆]²⁺. HRMS (ESI+) m/z for C₅₁H₄₀N₈Ru: calcd 433.1209, found 433.1209.

Synthesis of [Ru(4,4′-dmb)₂(1-nip)](PF₆)₂ 4.

Yield 44 mg (40%): purity 96%. In a microwave vessel, 63.4 mg (0.11 mmol) of [Ru(4,4′-dmb)₂Cl₂]∙2H₂O was mixed with 34.63 mg (0.10 mmol) of 1-NIP. Ethylene glycol (3 mL) was added to the vessel. The mixture was microwaved for 15 min at 180 °C. The resulting dark red color solution was directly loaded on to a silica column and eluted using 100% MeCN followed by 5% H₂O in MeCN to remove the excess starting material and other impurities. Finally, Ru(II) complex 4 was eluted as an orange band with 5–10% H₂O and 0.5–2.5% KNO₃ in MeCN to give an orange solid as a mixture of Cl⁻ and NO₃⁻ salts. This mixture was converted to the PF₆⁻ salt by adding 5–10 mL of water and 2–4 mL of saturated KPF₆. Evaporation of DCM under reduced pressure gave the the pure PF₆⁻ salt as an orange solid: R_f = 0.53 (10% H₂O + 2.5% KNO₃ in MeCN). HPLC retention time: 23.7 min.¹H NMR (500 MHz, acetonitrile-d₃) δ 9.18 (d, J = 8.6 Hz, 1H; m), 9.05 (s, 2H; c, f), 8.42 (s, 2H, 3’), 8.38 (s, 2H; 3), 8.18 (t, J = 10 Hz, 2H; g, i), 8.12–8.07 (m, 3H; k, a, d), 7.82 (dd, J = 8.3, 5.3 Hz, 2H; b, e), 7.77 (t, J = 10 Hz, 1H; h), 7.75–7.68 (m, 4H; 6’, l, j), 7.44 (d, J = 5.8 Hz, 2H; 6), 7.33 (dd, J = 6.4, 1.5 Hz, 2H; 5’), 7.08 (dd, J = 5.8, 1.4 Hz, 2H; 5), 2.61 (s, 6H; 4’-Me), 2.51 (s, 6H; 4-Me). MS (ESI+) m/z: 961.2 [M-PF₆]⁺, 815.2 [M–PF₆–H]⁺, 408.1 [M–2PF₆]²⁺. HRMS (ESI+) m/z for C₄₇H₃₈N₈Ru: calcd 408.1131, found 408.1126.

Synthesis of [Ru(4,4′-dmb)₂(2-nip)](PF₆)₂ 5.

Yield 50 mg (45%): purity 99%. In a microwave vessel, 63.4 mg (0.11 mmol) of [Ru(4,4′′-dmb)₂Cl₂]∙2H₂O was mixed with 34.63 mg (0.10 mmol) of 2-NIP. Ethylene glycol (3 mL) was added to the vessel. The mixture was placed in a microwave for 15 min at 180 °C. The resulting dark red color solution was directly loaded on to a silica column and eluted using 100% MeCN followed by 5% H₂O in MeCN to remove the excess starting material and other impurities. Finally, Ru(II) complex 6 was eluted as an orange band with 5–10% H₂O and 0.5–2.5% KNO₃ in MeCN to give an orange solid as a mixture of Cl⁻ and NO₃⁻ salts. This mixture was converted to the PF₆⁻ salt by adding 5–10 mL of water and 2–4 mL of saturated KPF₆. Evaporation of DCM under reduced pressure gave the pure PF₆⁻ salt as an orange solid: R_f = 0.58 (10% H₂O + 2.5% KNO₃ in MeCN). HPLC retention time: 24.4 min. ¹H NMR (500 MHz, acetonitrile-d₃) δ 9.05 (d, J = 8.2 Hz, 2H; c, f), 8.85 (d, J = 1.6 Hz, 1H; m), 8.45 (dd, J = 8.6, 1.8 Hz, 1H; g), 8.42 (s, 2H; 3’), 8.38 (s, 2H; 3), 8.17 (d, J = 8.7 Hz, 1H; h), 8.15–8.12 (m, 1H; l), 8.06 (dd, J = 5.2, 1.2 Hz, 2H; a, d), 8.05–8.03 (m, 1H; i), 7.82 (dd, J = 8.2, 5.3 Hz, 2H; b, e), 7.70–7.67 (m, 4H; 6’, j, k), 7.43 (d, J = 5.8 Hz, 2H; 6), 7.32 (dd, J = 6.1, 1.1 Hz, 2H; 5’), 7.07 (dd, J = 5.8, 1.0 Hz, 2H; 5), 2.61 (s, 6H; 4’-Me), 2.50 (s, 6H; 4-Me). MS (ESI+) m/z: 961.2 [M-PF₆]⁺, 815.3 [M-PF₆–H]⁺, 408.1 [M-2PF₆]²⁺. HRMS (ESI+) m/z for C₄₇H₃₈N₈Ru: calcd 408.1131, found 408.1126.

Synthesis of [Ru(4,4′-dmb)₂(aip)](PF₆)₂ 6.

Yield 64 mg (55%): purity 90%. In a microwave vessel, 63.4 mg (0.11 mmol) of [Ru(4,4′′-dmb)₂Cl₂]∙2H₂O was mixed with 39.61 mg (0.10 mmol) of AIP. Ethylene glycol (3 mL) was added to the vessel. The mixture was microwaved for 15 min at 180 °C. The resulting dark red color solution was directly loaded on to a silica column and eluted using 100% MeCN followed by 5% H₂O in MeCN to remove the excess starting material and other impurities. Finally, the desired Ru(II) complex was eluted as an orange band with 5–10% H₂O and 0.5–2.5% KNO₃ in MeCN to give an orange solid as a mixture of Cl⁻ and NO₃⁻ salts. This mixture was converted to the PF₆⁻ salt by adding 5–10 mL of water and 2–4 mL of saturated KPF₆. Evaporation of DCM under reduced pressure gave the pure PF₆⁻ salt as an orange solid: R_f = 0.48 (10% H₂O + 2.5% KNO₃ in MeCN). HPLC retention time: 24.0 min. ¹H NMR (500 MHz, acetonitrile-d₃) δ 8.88 (s, 1H; k), 8.44 (s, 2H; 3’), 8.40 (s, 2H; 3), 8.37 (d, J = 1.6 Hz, 1H; f), 8.32 (d, J = 1.6 Hz, 1H; c), 8.26 (d, J = 8.5 Hz, 2H; a, d), 8.11 (dd, J = 5.3, 1.3 Hz, 2H; o, l), 7.90 (t, J = 8.6 Hz, 1H; m), 7.85 (dd, J = 8.8, 1.1 Hz, 2H; g, j), 7.81 (t, J = 6.8 Hz, 1H; n), 7.73 (d, J = 5.7 Hz, 2H; 6’), 7.63 (m, 2H; b, e), 7.56 (m, 2H; h, i), 7.49 (d, J = 4.7 Hz, 2H; 6), 7.34 (dd, J = 5.7, 1.0 Hz, 2H; 5’), 7.14 (dd, J = 6.1, 1.2 Hz, 2H; 5), 2.62 (s, 6H; 4’-Me), 2.54 (s, 6H; 4-Me). MS (ESI+) m/z: 1011.2 [M–PF₆]⁺, 433.1 [M–2PF₆]²⁺. HRMS (ESI+) m/z for C₅₁H₄₀N₈Ru: calcd 433.1209, found 433.1204.

Cell Assays for Dark and Phototoxicity.

Solutions of the Ru(II) complexes were prepared as follows: 5 mM stock solutions of the Ru(II) chloride complexes were prepared in 10% DMSO-H₂O and stored at −20 °C prior to use. Dilutions of the stock solutions were carried out with Dulbeco’s phosphate buffered saline (DPBS) (pH = 7.4). The SK-MEL-28 cells were cultured according to the procedure described in the literature (16). Cell viability experiments were conducted in triplicate using a 96-well plate, which had 200 μL of DPBS buffer added. Aliquotes (50 μL) of cells in the log phase (SK-MEL-28 cells: 550,000 cells mL⁻¹) were transferred to the inner well containing 25 μL of warm culture medium. The plate was placed in a 5% CO₂ water-jacketed incubator at 37 °C for 3 h to allow the cells to equilibrate. The Ru(II) complexes were diluted with DPBS several times and prewarmed at 37 °C before adding 25 μL of aliquots of appropriate concentration to the cells. Ru(II) complexes 1–7 treated microplates were incubated under 5% CO₂ at 37 °C for 16 h drug-to-light intervals. Control cells that are not treated with light were kept in the dark. Cells in the microplate, those receiving light treatment, were treated with visible light (400 −700 nm, 34.2 mW cm⁻²) using a 190 W BenQ MS 510 overhead projector for 49 min to apply the total light doses of 100 J cm⁻². Control cells and light treated cell in the microplate both were kept inside the incubator for another 48 h. Then, 10 μL aliquots of Alamar blue was added to each well plate and incubated for another 15–16 h. Viability of the cells was quantified based on the fluorescence of Alamar blue dye that emits from live cells, whereas it does not fluoresce in dead cells. Fluorescence was quantified using a Cytofluor 4000 fluorescence microplate reader. Excitation filter and emission filter of the microplate reader were set in 530 ± 25 nm and 620 ± 40 nm, respectively. EC₅₀ values of the dark toxicity and photocytotoxicity were derived from eq 1 from the dose response curve of Graph Pad Prism 6.0. The y_i and y_f represent the initial and final fluorescence signal intensities, respectively. EC₅₀ values are reproducible within the range of ±25% in the submicromolar, ±10% below 10 μM, and ±5% above 10 μM region. PIs were calculated from the ratio of the dark EC₅₀-to-light EC₅₀ values.

Singlet oxygen quantum yield measurements.

The quantum yields for ¹O₂ production (Φ_Δ) were measured relative to [Ru(bpy)₃](PF₆)₂ according to eq 2 (see footnote “a” in Table 2), where I, A, and η are the integrated emission intensity, the absorbance at the excitation wavelength, and the refractive index of the solvent, respectively, in which the subscript “s” denotes the standard. The reference value for Φ_Δ used for [Ru(bpy)₃](PF₆)₂ was 0.56 in aerated MeCN (39).

Supplementary Material

SI

Figure S1: ¹H NMR spectrum of 2-(phenanthren-9-yl)-1H-imidazo [4,5-f][1,10] phenanthroline (pcip).

Figure S2: ¹H NMR spectrum (upper) and ¹H-¹H COSY spectrum (lower) of 1.

Figure S3: ¹H NMR spectrum of 2.

Figure S4: ¹H NMR spectrum (upper) and ¹H-¹H COSY spectrum (lower) of 3.

Figure S5: ¹H NMR spectrum (upper) and ¹H-¹H COSY spectrum (lower) of 4.

Figure S6: ¹H NMR spectrum (upper) and ¹H-¹H COSY spectrum (lower) of 5.

Figure S7: ¹H NMR spectrum (upper) and ¹H-¹H COSY spectrum (lower) of 6.

Figure S8: In vitro PDT dose?response curves for Ru(II) complexes 1–7 in SK-MEL-28 cells.

Figure 1.

Type I and type II photosensitized oxidations. The type I reaction forms oxygen radicals and radical ions, and the type II reaction mainly forms singlet oxygen by energy transfer from an excited photosensitizer to triplet oxygen. Both reactions lead to oxidized and/or oxygenated products. Oxygenated products can include hydroperoxides and endoperoxides.

Figure 5.

Structures of Ru(II) complexes 8–11.