Manipulation of Absorption Maxima by Controlling Oxidation Potentials in Bis(tridentate) Ru(II) N-Heterocyclic Carbene Complexes

Abstract

A series of seven Ru(II) complexes bearing NHC ligands have been synthesized. The electronic structures of these complexes were analysed by spectroscopic and electrochemical methods and further examined by theoretical calculations. Data show that absorption maxima are dependent on the HOMO level rather than the HOMO-LUMO gaps.

Introduction

Manipulation of absorption maxima (λmax) as well as oscillator strengths of light harvesting chromophores is crucial in solar energy conversion-related applications.1 In many organic and inorganic systems, λmax values are strongly correlated with HOMO and LUMO levels. Destabilization and stabilization of HOMO and LUMO, respectively, are common approaches to increase λmax values. Such approaches can be applied to Ru(II) polypyridyl complexes, in which the nature of the electronic transitions in low energy region is characteristic of Ru-based occupied molecular orbitals (MOs)-to-ligand-based virtual MOs charge transfer. A strategy for stabilizing virtual MOs includes expanding conjugation of the ligand or incorporating electron withdrawing groups to the periphery of the ligand. One of the representative methods of destabilizing occupied MOs involves applying a strong ligand field.2 Toward these ends, a huge numbers of new ligand systems that include N,3 C,4 or S5 donor moiety have been developed in addition to the conventional polypyridyl ligands such as 2,2'-bipyridine (bpy) and 2,2';6'2''-terpyridine (tpy) derivatives.6

We have previously exploited N-heterocyclic carbene (NHC) compounds, 2-(3-methylimidazolium-1-yl)pyridine (mip) or 2,6-bis-(3-methylimidazolium-1-yl)pyridine (bip), as bidentate or tridentate ligands, respectively, for a new type of ruthenium chromophore.7(a) Electrochemical data as well as theoretical calculation data of [Ru(bip)2]2+, for example, indicate that HOMO of this complex is ca. 0.2 V more destabilized relative to that of [Ru(tpy)2]2+ benchmark, as intended to bring a strong ligand field. However, the degree of destabilization of LUMO is even larger due to a confined electronic delocalization within the pyridyl ring. When one bip ligand is replaced with 2,2';6'2''-terpyridine- 4'-carboxylic acid (CTN, Scheme 1), the tpy-localized virtual orbital becomes the LUMO of the molecule, while a Rubased occupied orbital whose energy is destabilized by the bip ligand still remains a HOMO.7(b) The λmax value was brought back to visible region (463 nm). Dinda et al. reported that using 1,1'-[2,6-pyridinediylbis(methylene)]bis[3-methylimidazolyl] ligand to augment the ligand field gave rise to 0.44 and 0.24 V destabilization of HOMO levels for homoleptic [RuL2]2+ and heteroleptic [Ru(tpy)L]2+ (L=1,1'-[2,6-pyridinediylbis(methylene)]-bis[3-methylimidazolyl]), respectively, relative to that of [Ru(tpy)2]2+.8(a) The λmax values were 429 and 500 nm and molar extinction coefficient (ε) were 12000 and 5200 M−1cm−1, respectively. These results highlight the capacity of NHC compound as a versatile ligand for strong ligand field effects.

Scheme 1

Although the heteroleptic approach mentioned above works for manipulating the λmax values of Ru complexes to some extent, there is a need to find the factors that govern the λmax values of Ru(NHC) complexes to exploit such systems further. Under this background, combined with the lack of information regarding systematic structure-property relationships of Ru(NHC) systems, we have newly synthesized seven Ru complexes that possess a heteroleptic [Ru(tpy)- (NHC)]2+ or a homoleptic [Ru(NHC)2]2+ topology and feature NNC- or NN^C-type NHC structural motif (NN = bipyridyl, C = azolyl, and ^ = methylene). The structures of series complexes are shown in Scheme 2.

The synthesis of NNC- or NN^C-type ligands and their transition metal complexes have been reported in a handful of literature.8 However, the background for employing such ligand systems to build a new series of Ru complexes deserves comments as follow; (1) NHC moiety is connected to bpy with or without a methylene bridge and thus lies at the corner of the tridentate ligand. Due to the presence of bpy moiety in both NNC- and NN^C-type ligands, a minimum conjugation is ensured of at least up to two pyridyl rings. This structural motif prevents an ultimate destabilization of the unoccupied MO energy level derived from the confined electronic delocalization in the single pyridyl ring observed in the bip example.7(a) (2) The presence of the methylene bridge is manipulated to vary the strength of a ligand field induced by NHC moiety. With a methylene bridge, the Npyridine-Ru-Ccarbene bite angle becomes near orthogonal, thus inducing an augmented ligand field.8(a) (3) Employing a benzimidazolyl group in place of a simple imidazolyl one has multi purposes; one of which is to reduce energy mismatch between bpy and NHC thus extends the conjugation over the entire ligand, and the other is to delicately tune the σ-donating power of the NHC moiety. And the last one is to increase the absorption intensity of the complex by increasing the light absorbing cross section.

Scheme 2

During our analyses of experimental and theoretical calculation data of these series complexes, we found that λmax values are strongly correlated to electrochemical oxidation potentials (E1/22+/3+) rather than the commonly accepted HOMO-LUMO gaps. Here we report these results.

 

Experimental

Materials. All reactions were carried out under a nitrogen atmosphere unless otherwise noted. Standard Schlenk techniques were employed to manipulate air-sensitive solutions, while workup procedures were done in air. Tetrahydrofuran (THF) were purchased from Fischer Scientific (HPLC grade) and dried over Na/benzophenone and were subsequently distilled under nitrogen prior to use. Toluene was purchased from Samchun Chemicals and dried over Na/benzophenone and were subsequently distilled under nitrogen prior to use. N,N-dimethylformamide (DMF) and acetonitrile were distilled over CaH2 prior to use. Ethylene glycol (Aldrich) was used without further purification. Tetrakis(triphenylphosphine)-palladium(0) was purchased from Pressure Chemical Co. Copper(II) oxide was purchased from Daejung Chemical Co. 2-Bromopyridine, 2,6-dibromopyridine (98%), tributyltin chloride (96%), n-butyl lithium (2.0 M solution in n-hexane), imidazole, 1-methylimidazole, benzimidazole, 1-methylbenzimidazole, iodomethane, 2,2':6',2''-terpyridine, ammonium hexafluorophosphate (NH4PF6) were purchased from Aldrich Chemical Co. 6-Bromo-2,2'-bipyridine,9(a) 6-chloromethyl-2,2’-bipyridine,9(b) (2,2':6',2''-tepyridine)(trichloro)- Ru(III) (Ru(tpy)Cl3)9(c) were prepared according to literature procedures. Column chromatography was performed on silica gel 60 (230-400 Mesh, Merck).

Instrumentation. 1H and 13C NMR spectra were recorded with Bruker (75 MHz for 13C NMR), Agilent (400 MHz and 100 MHz for 1H and 13C NMR, respectively) and Agilent (500 MHz and 125 MHz for 1H and 13C NMR, respectively) spectrometers. 1H NMR spectra were taken in CDCl3 and DMSO-d6 and were referenced to residual CDCl3 (7.27 ppm) and DMSO-d6 (2.50 ppm), respectively. Chemical shifts of the 13C NMR spectra were measured relative to CDCl3 (77.16 ppm) or DMSO-d6 (39.52). Elemental analyses were done at the National Center for Inter-University Research Facilities located in the Seoul National University. Highresolution mass spectrometry (HRMS) data were obtained at the Korea Basic Science Institute (Daegu). Electronic absorption spectra were recorded on a Beckman Du-650 spectrophotometer. Cyclic voltammograms were obtained with a CH Instrument voltammetric analyzer. Measurements were performed after the acetonitrile (spectroscopic grade) solution was purged with dry nitrogen gas for 30 min. The supporting electrolyte was 0.1 M tetrabutylammonium hexafluorophosphate (TBAPF6). Glassy carbon and Ag/Ag+ (0.01 M AgNO3) were used as working and reference electrodes, respectively. The scan rate was maintained at 100 mV/s.

Synthesis of Ligands.

6-(1H-Imidazol-1-yl)-2,2'-bipyridine (bi): To a flask containing 6-bromo-2,2'-bipyridine (0.47 g, 2 mmol), imidazole (0.16 g, 2.4 mmol), potassium carbonate (0.33 g, 2.4 mmol) in 6 mL DMF solution, copper(II) oxide (16 mg, 10 mol %) was added. The solution was heated at 150 ℃ for 36 h. After the solution was cooled to room temperature, the solvent was removed under reduced pressure and extracted with CH2Cl2 (330 mL). The organic layer was washed with brine and dried with Na2SO4. Purification with silica gel column chromatography (DCM:MeOH = 20:1) gave white solid in 95% yield (0.43 g). 1H NMR (400 MHz, CDCl3) δ 8.67 (ddd, J = 4.8, 1.8, 0.9 Hz, 1H), 8.45 (s, 1H), 8.40 (dt, J = 8.0, 1.0 Hz, 1H), 8.34 (dd, J = 7.8, 0.7 Hz, 1H), 7.89 (t, J = 7.9 Hz, 1H), 7.84−7.79 (m, 1H), 7.72 (t, J = 1.3 Hz, 1H), 7.34− 7.30 (m, 2H), 7.23 (s, 1H); 13C NMR (100 MHz, CDCl3) δ 155.5, 154.7, 149.1, 148.3, 139.9, 136.9, 134.9, 130.6, 124.2, 121.1, 118.9, 116.1, 111.9; HRMS (FAB+), m/z [M+H]+ found (calc): 223.0986 (223.0984).

3-([2,2'-Bipyridin]-6-yl)-1-methyl-1H-imidazol-3-ium hexafluorophosphate (bim): To a flask containing bi (0.22 g, 1 mmol) in dry acetonitrile, methyl iodide (0.13 mL, 2 mmol) was added and heated reflux for 4 h. After the solution was cooled to room temperature, the solvent was removed by rotary evaporator. Crude mixture was dissolved with minimum amount of methanol and dropped to diethyl ether. White precipitate was filtered and redissolved in 10 mL of water. Excess NH4PF6 was added and the resulting solution was stirred for 10 min. White precipitate was filtered, washed with water and dried by vacuo (yield: quantitative). 1H NMR (400 MHz, DMSO) δ 10.21 (s, 1H), 8.75 (d, J = 8.0 Hz, 1H), 8.65 (t, J = 2.0 Hz, 1H), 8.63 (d, J = 8.0 Hz, 1H), 8.53 (d, J = 8.0 Hz, 1H), 8.33 (t, J = 8.0 Hz, 1H), 8.04 (dt, J = 7.0, 2.0 Hz, 2H), 7.98 (t, J = 1.8 Hz, 1H), 7.55 (ddd, J = 7.5, 4.8, 1.1 Hz, 1H), 4.01 (s, 3H); 13C NMR (100 MHz, DMSO) δ 155.0, 153.4, 149.7, 146.1, 141.9, 137.7, 135.8, 125.3, 124.9, 121.4, 121.2, 119.2, 113.9, 36.5; HRMS (FAB+), m/z [M]+ found (calc): 237.1141 (237.1140).

1-([2,2'-Bipyridin]-6-yl)-1H-benzimidazole (bzi): Prepared as described for bi, from benzimidazole (0.28 g, 2.4 mmol), 6-bromo-2,2'-bipyridine (0.47 g, 2 mmol), potassium carbonate (0.33 g, 2.4 mmol) and copper(II) oxide (16 mg, 10 mol %) in 6 mL of DMF. The solution was refluxed for 48 h. After purification on silica gel (DCM:MeOH = 22:1), light yellow solid was obtained (yield: 0.61 g, 94%). 1H NMR (400 MHz, CDCl3) δ 8.64 (dd, J = 4.7, 0.7 Hz, 1H), 8.59 (s, 1H), 8.35 (t, J = 8.2 Hz, 2H), 8.07 (dd, J = 6.6, 2.4 Hz, 1H), 7.90−7.83 (m, 2H), 7.75 (td, J = 7.8, 1.8 Hz, 1H), 7.41 (d, J = 7.8 Hz, 1H), 7.39 −7.32 (m, 2H), 7.26 (ddd, J = 7.4, 4.8, 1.0 Hz, 1H); 13C NMR (100 MHz, CDCl3) δ 155.6, 154.6, 149.0, 148.9, 144.5, 141.2, 139.6, 136.8, 131.9, 124.1, 124.0, 123.1, 120.9, 120.5, 118.6, 113.7, 112.6; HRMS (FAB+), m/z [M+H]+ found (calc): 273.1137 (273.1140).

3-([2,2'-Bipyridin]-6-yl)-1-methyl-1H-benzimidazol-3-ium hexafluorophosphate (bzim): Prepared as described for 1, from bzi (0.27 g, 1 mmol) and methyl iodide (0.13 mL, 2 mmol). Yield: Quantitative. 1H NMR (400 MHz, DMSO) δ 10.57 (s, 1H), 8.80 (ddd, J = 4.7, 1.7, 0.9 Hz, 1H), 8.64 (dd, J = 7.8, 0.6 Hz, 1H), 8.55 (dd, J = 7.5, 1.5 Hz, 1H), 8.48 (dd, J = 4.9, 4.0 Hz, 1H), 8.43 (t, J = 8.0 Hz, 1H), 8.18 (dd, J = 7.4, 1.6 Hz, 1H), 8.12−8.04 (m, 2H), 7.88−7.80 (m, 2H), 7.58 (ddd, J = 7.5, 4.8, 1.1 Hz, 1H), 4.24 (s, 3 H); 13C NMR (100 MHz, DMSO) δ 155.5, 153.5, 149.8, 146.9, 143.1, 141.9, 137.8, 132.4, 129.3, 128.0, 127.2, 125.2, 121.3, 121.1, 116.9, 115.5, 114.1, 33.8; HRMS (FAB+), m/z [M]+ found (calc): 287.1295 (287.1297).

3-(2,2'-Bipyridine-6-ylmethyl)-1-methyl-1H-imidazol-3-ium hexafluorophosphate (b^im): A solution of 6- (chloromethyl)-2,2-bipyridine (0.45 g, 2.2 mmol) and 1-methylimidazole (0.16 g, 2.0 mmol) in 10 mL of dry acetonitrile was refluxed for 48 h. After the solution was cooled to room temperature, the solvent was removed by rotary evaporator. Crude mixture was dissolved by 3 mL of dichloromethane and the solution was dropped to diethyl ether. White precipitate was filtered and redissolved in 10 mL of water. Excess NH4PF6 was added and the resulting solution was stirred for 10 min. White precipitate was filtered, washed with water and dried by vacuo (yield: 0.79 g, quantitative). 1H NMR (400 MHz, CDCl3) δ 10.58 (s, 1H), 8.62 (d, J = 4.6 Hz, 1H), 8.34 (d, J = 7.9 Hz, 1H), 8.27 (d, J = 7.9 Hz, 1H), 7.80 (t, J = 7.7 Hz, 2H), 7.73 (s, 2 H), 7.67 (d, J = 7.5 Hz, 1H), 7.32−7.27 (m, 1H), 5.81 (s, 2H), 4.08 (s, 3H); 13C NMR (100 MHz, DMSO) δ 155.6, 154.9, 153.7, 149.8, 139.2, 137.8, 137.8, 125.0, 124.1, 123.7, 123.0, 121.0, 120.5, 53.4, 36.3; HRMS (FAB+), m/z [M]+ found (calc): 251.1295 (251.1297).

3-(2,2'-Bipyridine-6-ylmethyl)-1-methyl-1H-benzimidazol-3-ium Hexafluorophosphate (b^zim): The same procedure as the synthesis of b^im. 6-(Chloromethyl)-2,2-bipyridine (0.45 g, 2.2 mmol) and 1-methylbenzimidazole (0.16 g, 2.0 mmol) gave b^im in quantitative yield. (0.892 g) 1H NMR (400 MHz, DMSO) δ 9.95 (s, 1H), 8.66 (ddd, J = 4.8, 1.8, 0.9 Hz, 1H), 8.35 (dd, J = 7.9, 0.9 Hz, 1H), 8.10− 8.02 (m, 4H), 7.90−7.85 (m, 1H), 7.74−7.66 (m, 3H), 7.44 (ddd, J = 7.5, 4.8, 1.2 Hz, 1H), 6.01 (s, 2H), 4.17 (d, J = 0.4 Hz, 3H); 13C NMR (100 MHz, DMSO) δ 155.2, 154.4, 152.8, 149.4, 143.6, 138.8, 137.4, 131.9, 131.3, 126.7, 126.5, 124.5, 122.9, 120.4, 120.2, 113.9, 113.7, 50.8, 33.4; HRMS (FAB+), m/z [M]+ found (calc): 301.1455 (301.1453).

General Procedure for Ru Complexes.

Ru(tpy)L: Ru(tpy)Cl3 (132 mg, 0.3 mmol) and L (0.3 mmol) in 5 mL of ethylene glycol was heated at 180 ℃ for 4 h. After the solution was cooled to room temperature, the solution was added dropwisely to a saturated aqueous solution of NH4PF6, causing to precipitation of the compound. After the precipitate was filtered, it was purified by silica gel column chromatography (CH3CN:0.5 M NaNO3 = 9:1).

Ru(tpy)(bim) (1): Orange solid; Yield: 59%. 1H NMR (400 MHz, DMSO) δ 9.01 (s, 1H), 8.99 (s, 1H), 8.88−8.82 (m, 2H), 8.78 (s, 1 H), 8.77 (s, 1H), 8.59 (d, J = 2.3 Hz, 1H), 8.55 (dd, J = 8.3, 1.1 Hz, 1H), 8.53−8.49 (m, 1H), 8.47 (t, J = 8.1 Hz, 1H), 8.07 (td, J = 7.9, 1.5 Hz, 1H), 8.02 (td, J = 7.9, 1.5 Hz, 2H), 7.52 (dd, J = 5.1, 1.0 Hz, 1H), 7.36 (dd, J = 5.6, 0.8 Hz, 2H), 7.32 (ddd, J = 7.5, 5.5, 1.2 Hz, 1 H), 7.28−7.22 (m, 3H), 2.76 (s, 3H); 13C NMR (100 MHz, DMSO) δ 184.5, 157.0, 155.2, 154.5, 154.4, 152.3, 151.9, 150.3, 138.7, 138.2, 137.6, 134.9, 134.9, 127.6, 124.5, 124.5, 124.4, 123.8, 119.6, 118.5, 111.8, 34.9; HRMS (FAB+), m/z [M+PF6]+ found (calc): 716.0695 (716.0708); Anal. calcd for C29H23F12N7P2Ru: C, 40.48; H, 2.69; N, 11.39. Found: C, 40.36; H, 2.82; N, 11.27.

Ru(tpy)(bzim) (2): Light orange solid; Yield: 52%. 1H NMR (400 MHz, CD3CN) δ 8.72 (d, J = 8.2 Hz, 2H), 8.59 (dd, J = 17.0, 8.8 Hz, 3H), 8.48 (t, J = 8.2 Hz, 3H), 8.41 (t, J = 8.2 Hz, 1H), 8.25 (d, J = 8.2 Hz, 1H), 8.01 (td, J = 8.0, 1.5 Hz, 1H), 7.92 (td, J = 8.0, 1.4 Hz, 2H), 7.50 (t, J = 7.3 Hz, 1H), 7.42 (dd, J = 9.1, 6.5 Hz, 2H), 7.28 (dt, J = 7.4, 4.2 Hz, 4H), 7.16−7.09 (m, 2H), 2.98 (s, 3H); 13C NMR (75 MHz, CD3CN) δ 199.7, 158.2, 156.5, 156.3, 155.9, 154.3, 153.5, 151.2, 139.9, 139.2, 138.8, 136.7, 136.4, 133.0, 128.5, 128.3, 125.7, 125.4, 125.3, 124.8, 120.0, 118.3, 113.6, 112.3, 111.7, 33.4; HRMS (FAB+), m/z [M+PF6]+ found (calc): 766.0859 (766.0866); Anal. calcd for C33H25F12N7P2Ru: C, 43.53; H, 2.77; N, 10.77; Found: C, 43.48; H, 2.89; N, 10.69.

Ru(tpy)(b^im) (3): Red solid; Yield: 56%. 1H NMR (400 MHz, DMSO) δ 8.93 (t, J = 8.3 Hz, 3H), 8.72 (t, J = 8.3 Hz, 3H), 8.43 (td, J = 8.0, 4.4 Hz, 2H), 8.16 (d, J = 7.4 Hz, 1H), 8.02 (td, J = 7.9, 1.4 Hz, 2H), 7.97 (td, J = 8.0, 1.4 Hz, 1H), 7.67 (d, J = 4.9 Hz, 2H), 7.50 (d, J = 1.8 Hz, 1H), 7.35−7.28 (m, 3H), 7.02 (d, J = 1.9 Hz, 1H), 7.01 (d, J = 5.7 Hz, 1H), 5.94 (s, 2H), 2.54 (s, 3H); 13C NMR (100 MHz, DMSO) δ 176.8, 157.4, 157.1, 156.5, 155.8, 154.6, 153.0, 147.7, 138.0, 137.9, 137.7, 135.2, 128.0, 127.2, 126.3, 124.3, 124.3, 124.0, 123.8, 123.5, 122.9, 53.5, 35.4; HRMS (FAB+), m/z [M+PF6]+ found (calc): 730.0862 (730.0865); Anal. calcd for C33H25F12N7P2Ru: C, 43.53; H, 2.77; N, 10.77; Found: C, 43.45; H, 2.91; N, 10.67.

Ru(tpy)(b^zim) (4): Dark orange solid; Yield: 63%. 1H NMR (400 MHz, DMSO) δ 9.00−8.94 (m, 3H), 8.75 (t, J = 8.3 Hz, 3H), 8.50 (dd, J = 12.0, 4.3 Hz, 2H), 8.42 (dd, J = 7.7, 1.1 Hz, 1H), 8.04 (d, J = 1.5 Hz, 1H), 8.04−7.99 (m, 3H), 7.76 (dd, J = 5.6, 0.8 Hz, 2H), 7.39−7.34 (m, 3H), 7.33− 7.28 (m, 2H), 7.27−7.22 (m, 1H), 7.04−7.00 (m, 1H), 6.26 (s, 2H), 2.78 (s, 3H); 13C NMR (75 MHz, DMSO) δ 192.2, 157.3, 156.9, 156.6, 155.8, 154.5, 153.3, 147.5, 138.3, 138.2, 138.1, 138.0, 135.9, 134.9, 134.2, 128.1, 127.3, 126.6, 124.4, 124.3, 123.8, 123.6, 123.0, 122.8, 109.9, 49.7, 31.6; HRMS (FAB+), m/z [M+PF6]+ found (calc): 780.1010 (780.1022); Anal. calcd for C34H27F12N7P2Ru: C, 44.17; H, 2.94; N, 10.60; Found: C, 44.11; H, 3.03; N, 10.53 .

RuL2: RuCl3·3H2O (78 mg, 0.3 mmol) and L (0.6 mmol) in 8 mL of ethylene glycol was heated at 180 ℃ for 4 h. After cooling to room temperature, crude solution was dropped to saturated aqueous solution of NH4PF6 caused precipitation of the compound. After filteration, the solid was purified by silica gel column chromatography (CH3CN: 0.5 M NaNO3 = 9:1) gave the desired product.

Ru(bim)2 (5): Orange solid; Yield: 71%. 1H NMR (400 MHz, DMSO) δ 8.81 (m, 4H), 8.55 (d, J = 2.2 Hz, 2H), 8.51−8.43 (m, 4H), 8.08 (dd, J = 11.2, 4.5 Hz, 2H), 7.48 (d, J = 5.0 Hz, 2H), 7.34 (t, J = 8.0 Hz, 2H), 7.29 (d, J = 2.2 Hz, 2H), 2.66 (s, 6H); 13C NMR (100 MHz, DMSO) δ 184.5, 154.8, 154.3, 151.8, 150.5, 138.2, 137.2, 127.5, 124.3, 124.3, 119.3, 117.9, 111.2, 34.8; HRMS (FAB+), m/z [M+H]+ found (calc): 719.0806 (719.0817); Anal. calcd for C28H24F12N8P2Ru: C, 38.94; H, 2.80; N, 12.98; Found: C, 38.89; H, 2.83; N, 12.94.

Ru(bzim)2, (6): Yellow solid; Yield: 68%. 1H NMR (400 MHz, DMSO) δ 8.99−8.83 (m, 6H), 8.58 (t, J = 9.4 Hz, 4H), 8.12 (t, J = 7.3 Hz, 2H), 7.51 (dd, J = 13.4, 6.7 Hz, 4H), 7.48−7.40 (m, 4H), 7.38−7.32 (m, 2H), 2.87 (s, 6H); 13C NMR (75 MHz, DMSO) δ 196.7, 154.7, 154.4, 152.1, 150.4, 138.6, 138.1, 135.2, 131.1, 127.7, 124.6, 124.5, 124.1, 119.4, 112.7, 111.9, 111.1, 32.1; HRMS (FAB+), m/z [M+PF6]+ found (calc): 819.1125 (819.1132); Anal. calcd for C36H28F12N8P2Ru: C, 44.87; H, 2.93; N, 11.63; Found: C, 44.84; H, 3.01; N, 11.59.

Ru(b^im)2 (7): Brown solid; Yield: 66%. 1H NMR (400 MHz, DMSO) δ 8.77 (d, J = 8.0 Hz, 2H), 8.55 (d, J = 8.0 Hz, 2H), 8.32 (t, J = 7.9 Hz, 2H), 7.99 (d, J = 7.6 Hz, 2H), 7.94 (t, J = 7.5 Hz, 2H), 7.55 (d, J = 1.5 Hz, 2H), 7.27 (t, J = 7.9 Hz, 2H), 7.10 (d, J = 1.6 Hz, 2H), 6.87 (d, J = 5.0 Hz, 2H), 5.81 (d, J = 15.4 Hz, 2H), 4.97 (d, J = 15.2 Hz, 2H), 2.45 (s, 6H); 13C NMR (100 MHz, DMSO) δ 180.3, 158.1, 156.9, 154.3, 148.5, 138.0, 137.4, 127.2, 125.2, 123.7, 123.3, 123.2, 122.3, 54.0, 35.3; HRMS (FAB+), m/z [M+PF6]+ found (calc): 747.1126 (747.1130); Anal. calcd for C30H28F12N8P2Ru: C, 40.41; H, 3.17; N, 12.57; Found: C, 40.38; H, 3.23; N, 12.55.

 

Results and Discussion

Synthesis. The syntheses of NNC- and NN^C-type ligands are shown in Scheme 3 and 4, respectively. The synthesis of NNC-type ligands, 3-([2,2'-bipyridin]-6-yl)-1-methyl-1H-imidazol-3-ium hexafluorophosphate (bim) and 3-([2,2'- bipyridin]-6-yl)-1-methyl-1H-benzimidazol-3-ium hexafluorophosphate (bzim), was carried out via copper-catalyzed Ullmann coupling between 6-bromo-2,2'-bipyridine and imidazole or benzimidazole in DMF at 150 ℃ followed by methylation with iodomethane in refluxing acetonitrile. The two-step reactions were facile and straightforward providing the desired NNC ligands (bim, bzim) in near quantitative yields. The reaction of 6-bromo-2,2'-bipyridine with excess amount of 1-methylimidazole or 1-methylbenzimidazole as solvents attempting to generate bim or bzim directly only gave products in poor yield (ca. 20%).

The synthesis of NN^C-type ligands was even more straightforward; reaction of 6-chloromethyl-2,2'-bipyridine with 1-methylimidazole or 1-methylbenzimidazole in refluxing acetonitrile gave b^im and b^zim, respectively, in quantitative yields.

Scheme 3.Synthesis of NNC ligands, bim and bzim.

Scheme 4.Synthesis of NN^C ligands, b^im and b^zim.

Heteroleptic Ru complexes ([Ru(tpy)L]2+) were synthesized by a reaction between [Ru(tpy)Cl3] and a slight excess amount of ligand L in a refluxing ethylene glycol solution for 4 h. For homoleptic Ru complexes (RuL2), RuCl3 and ligand L were heated as the same procedures of Ru(tpy)L. All complexes were obtained in moderate yields (52-71%). Meanwhile, other synthetic approaches, such as Ag(I) transmetalation also provided the target products, but yields were poor (ca. 15%).

Absorption Spectroscopy and Electrochemistry. Figure 1. displays electronic absorption spectra of seven complexes as well as their archetypal benchmark molecules, [Ru(bip)2]2+ and CTN for homoleptic and heteroleptic series, respectively. The corresponding λmax and ε values are listed in Table 1. All complexes exhibit conventional absorption signatures characteristic of intense bands in the ultraviolet region (250- 330 nm) for π-π* ligand centered (LC) transitions and moderately intense bands in the visible region (400-600 nm) for metal-to-ligand charge transfer (MLCT) bands. Further examination using the time-dependent density functional theory (TD-DFT) confirms these attributions, except for the fact that a mixed-metal-ligand-to-ligand CT is more suitable for describing the latter case (Table S1).

The λmax values in the lower energy region reside between 464-484 nm for the heteroleptic series and 455-494 nm for the homoleptic series. The λmax values for the homoleptic series span wider than those for the homoleptic one, whose background will be discussed in the later part of this paper. Interestingly, the absorption signatures of these series complexes display clear trends as follow: (1) Complexes possessing NN^C-type ligand have lower energy λmax values than those possessing NNC-type ligand. (2) Complexes possessing imidazole group in their respective NNC- or NN^C-type ligands have lower energy λmax values than those possessing benzimidazolyl group. (3) Heteroleptic complexes have larger ε values compared to homoleptic counterparts. (4) Complexes possessing NN^C-type ligand have larger ε values relative to those possessing NNC-type ligand within the respective heteroleptic and homoleptic series. Trend (1) is contrary to our anticipation because if a conjugation is disrupted in NN^C-type ligand, the λmax values would shift to the blue. This behaviour can be found in the literature by comparing two separate examples reported in ref. 7(a) and 8(a), respectively. In ref 7(b), λmax of [Ru(tpy-CO2H)(CNC)]2+ (CTN) is 463 nm while that of [Ru(tpy)(C^N^C)]2+ in ref 8(a) is ~500 nm. Although a direct comparison might not be appropriate because of a presence of carboxylic acid group in the tpy ligand in CTN, these two systems roughly indicate that whether the extent of electronic delocalization includes NHC moiety is not a major factor in determining the λmax of Ru(NHC) complexes. Trend (2) combined with the results described above regarding trend (1) indicate that λmax values are more affected by the σ-donating power induced by NHC moiety. This analysis drives us to examine the relationship between λmax and σ-donating power more closely. Since the magnitude of σ-donating power is best indicated by electrochemical redox potentials, we measured redox potentials of the series complexes and considered the dependence of these values on λmax.

Figure 1.Electronic absorption spectra heteroletic (a) and homoleptic (b) complexes along with benchmark molecules, [Ru(bip)2]2+ (b, green) and CTN (a, green). Experimental conditions: solvent = acetonitrile, temperature = 25 ± 1 ℃.

Electrochemical oxidation potentials of the seven complexes in acetonitrile at 23 ℃ were recorded and their values are listed in Table 1. The one-electron reversible reductions of heteroleptic series occur within 0.05 V range (−1.02 ~ −1.07 V vs. NHE) while those of homoleptic ones do within 0.04 V range (−1.08 ~ −1.12 V vs. NHE). These confinements of reduction potentials originate the fact that LUMOs of these series are located either tpy or bpy ligand, both of which have very similar energy levels. We will show the shapes and energies of frontier orbitals of new complexes in the later part of this paper. On the contrary, the one-electron reversible oxidation potentials of complexes significantly differ each other according to the nature of the ligand. The values of heteroleptic series are 1.34, 1.45, 1.46, and 1.53 V vs. NHE for complexes 3, 4, 1, and 2, respectively. Those of homoleptic series are 1.15, 1.43, and 1.52 V vs. NHE for complexes 6, 7, and 5, respectively. The maximum difference of the values is ca. 0.2 V. The trend of oxidation potential values is similar to that of absorption maxima. NN^C-type ligand gave more σ-donating effect than NNC-type one and the imidazole group provide more σ-donating effect than the benzimidazolyl group. As a result, HOMO-LUMO gaps of the series measured by electrochemical method are appeared to be dependent only on the oxidation potential values phenominologically.

Table 1.aExperimental conditions: [compound] = 5 mM; [TBAPF6] = 0.1 M; solvent = acetonitrile; temperature = 25 ± 1 ℃; scan rate = 100 mV/s; reference electrode = Ag/Ag+; working electrode = glassy carbon. All potentials are referenced to a ferrocene/ferrocenium redox couple as an internal standard and converted to NHE by the relation ferrocene/ferrocenium vs. NHE = +0.64 V. bFrom ref. 7(a)

Figure 2.Cyclic voltammogram of seven complexes. Gradual shift of oxidation potential is guided by red dot line. Invariant reduction potential is guided by blue dot line. Experimental conditions: [compound] = 5 mM; [TBAPF6] = 0.1 M; solvent = acetonitrile; temperature = 23 ℃; scan rate = 100 mV/s; reference electrode = Ag/Ag+; working electrode = glassy carbon. All potentials are referenced to a ferrocene/ferrocenium redox couple as an internal standard and converted to NHE by the relation ferrocene/ferrocenium vs NHE = +0.64 V.

Figure 3.Dependence of the electrochemically determined E1/2 2+/3+ (filled) and E1/2 1+/2+ (open) values on λmax, for heteroleptic (blue rectangles) and homoleptic (red circles) series. All these redox potentials are relative to the NHE. Experimental condition for electrochemical measurement: See footnote of Table 1.

Figure 3 displays the relationship between the oxidation potential values and λmax. As mentioned earlier, only E1/2 2+/3+ values are strongly dependent on λmax, yet E1/2 1+/2+ values are essentially invariant with respect to λmax for both homoleptic and heteroleptic series. The degrees of correlation between E1/2 2+/3+ and λmax values for both homoleptic and heteroleptic series are very similar to each other as manifested by the virtually same slopes of each trend line.

In general, λmax values scale with HOMO-LUMO gap; a gradual increase of λmax coincides with a concomitant destabilization of HOMO and a stabilization of LUMO. Therefore, a strong dependence of E1/2 2+/3+ values on λmax is perhaps natural. The prominent dependence of λmax only on E1/2 2+/3+ values is phenomenological behaviour embossed by the silent dependence of λmax values on E1/2 2+/3+ ones. The background of levelling effect of E1/2 1+/2+ values can be rationalized as follow: (1) In the heteroleptic series, most of the electronic population in LUMO is localized at tpy ligand regardless of a second ligand. The LUMO energy is primarily determined by the tpy energy level, which is not affected by Ru(NNC) or Ru(NN^C) moiety. (2) In the homoleptic series, most of the electronic population in LUMO is localized at bpy moiety of one of two NNC-type ligands (5) or equally at two bpy moieties of each ligand (6 and 7). As a result, the LUMO energies of each series are levelled. The absolute level of trend line of the E1/2 1+/2+/λmax dependences for a homoleptic series is slightly higher than that for a heteroleptic one. This pattern mirrors the LUMO energies of bpy and tpy of their own, which are determined by DFT calculation.9 While the slopes of two trend lines of both homoleptic and heteroleptic series are virtually the same, the absolute levels of those two differ by 0.1 V; it is quite natural that the destabilization of HOMO with two NHC ligands is more pronounced than that with only one unit.

The significant destabilization of HOMO level in homoleptic [Ru(b^im)2]2+ is worth noting. Considering the fact that the oxidation potential of [Ru(bip)2]2+ in which four NHC moieties are coordinated to the Ru center is only 1.38 V, an observed value of 1.15 V [Ru(b^im)]2+ is remarkable. Due to this substantial negative shift of oxidation potential, λmax values of [Ru(b^im)2]2+ exhibit substantial bathochromic shift up to 495 nm, thus causing widely spread λmax values of homoleptic series.

Computational Study. The geometry optimizations and electronic structure calculations of seven new complexes were performed via using density functional theory (DFT). The nature of each MO is characterized by percent contribution of each atom summed into several classes; Ru, tpy, bpy, and imidazolyl or benzimidazolyl part for heteroleptic series, and Ru, bpy, and imidazolyl or benzimidazolyl part for homoleptic series (Table 2). The 3-dimensional representations of the isosurfaces of each MO clearly confirm the character of the MO (Figure 4). For all seven complexes, the three highest occupied orbitals of these complexes have their majority of electron populations (56-84%) at Ru metal while three lowest unoccupied MOs have those in tpy or bpy ligands indicating that the lowest energy absorption bands are metal-to-ligand charge transfer in character. It is important to note that substantial amount of the electron population of HOMO is delocalized over azolyl ring plane (18-30%) highlighting the significant contribution of NHC group to the shape and energy of the HOMO. The degree of σ-donation can be modulated by the type of NHC ligands as well as their geometries. Complexes with NN^C-type ligand have geometries more close to the perfect octahedron in terms of N-Ru-C bite angle compared to those with NNCtype one.8(a) For heteroleptic series, the ligand field applied by the NHC moiety is thus more prominent in complex 3 and 4 than in complex 1 and 2. Accordingly, the degree of destabilization of HOMO energy level is thus more prevailing in complex 3 and 4. The same is true for the homoleptic series; complex 7 has higher HOMO energy level compared to complex 5 and 6. When we focus on the nature of NHC ligand, imidazolyl group provide stronger ligand field than benzimidazolyl group since the electronic populations are far more delocalized over the aromatic ring in benzimidazolyl ring, thus weaken the σ-donating effect. As a result, the energy levels of HOMOs with benzimidazolyl group are less destabilized compared to those with simple imidazolyl group. These analyses clearly show how the structure and geometry NHC moiety affect the shapes and energies of HOMOs.

Table 2.aMethylimidazole part of ligand. bMethylbenzimidazole part of ligand. cBipyridyl part of NNC or NN^C ligand. d0 and 1 indicate HOMO and LUMO, respectively. −1, −2, −3 correspond to HOMO-1, HOMO-2, HOMO-3, respectively while 2 and 3 represent LUMO+1 and LUMO+2, respectively.

As shown in Figure 4, the electronic populations in the LUMOs of heteroleptic series are apparently localized in the whole tpy ligand, while those of homoleptic series are separately localized in the two bpy ligands. As mentioned earlier, the energy levels of LUMOs of tpy or bpy are similar each other. No matter what the LUMO is localized in one tpy ligand or two bpy ones, the resulting reduction potentials of such complexes are thus appear to be similar. In the case of HOMOs, however, electronic populations are shared by Ru metal and NHC moiety of NNC or NN^C-type ligands.

Figure 4.Frontier molecular orbitals calculated at the B3LYP/LANL2DZ level.

Figure 5.Frontier molecular orbital (five highest occupied and four lowest unoccupied) energy diagram of complex 1-7, [Ru(tpy)2]2+, and [Ru(bip)2]2+ calculated at the B3LYP/LANL2DZ level. Gradual stabilization of HOMO energy levels and levelling of LUMO energy levels are guided by red dotted lines.

The energies of the frontier MOs are compared in Figure 5 along with those of [Ru(tpy)2]2+ and [Ru(bip)2]2+ benchmarks. The calculation results are in excellent agreement with the experimental electrochemical data as well as the absorption spectroscopic data highlighted by the red dotted guide lines.

 

Conclusion

In sum, we have newly synthesized seven Ru complexes that possess a heteroleptic [Ru(tpy)(NHC)]2+ or a homoleptic [Ru(NHC)2]2+ topology and feature NNC- or NN^C-type NHC structural motif. These complexes have varying degrees of oxidation potentials induced by different ligand field of NHC moiety, yet have levelled reduction potential due to the similar LUMO energy localized in bpy or tpy moiety. We observed that the λmax values of these series complexes are correlated with only the oxidation potentials. Given the structure-property relationship obtained in this study, we could modulate λmax values of Ru(NHC) complexes to some extent.