Coordination Chemistry of Ru(II) Complexes of an Asymmetric Bipyridine Analogue: Synergistic Effects of Supporting Ligand and Coordination Geometry on Reactivities

Abstract

The reactivities of transition metal coordination compounds are often controlled by the environment around the coordination sphere. For ruthenium(II) complexes, differences in polypyridyl supporting ligands affect some types of reactivity despite identical coordination geometries. To evaluate the synergistic effects of (i) the supporting ligands, and (ii) the coordination geometry, a series of dicarbonyl–ruthenium(II) complexes that contain both asymmetric and symmetric bidentate polypyridyl ligands were synthesized. Molecular structures of the complexes were determined by X-ray crystallography to distinguish their steric configuration. Structural, computational, and electrochemical analysis revealed some differences between the isomers. Photo- and thermal reactions indicated that the reactivities of the complexes were significantly affected by both their structures and the ligands involved.

1. Introduction

Polypyridines with multiple covalently bonded pyridine groups exhibit unique photophysical and redox properties [1,2]. Among the polypyridines, bipyridine analogues play an important role in the formation of various transition metal complexes as bidentate ligands with two nitrogen donor atoms [3]. Bipyridine analogues function not only as supporting ligands for stabilizing metal complexes, but also as electron pool sites. For example, in addition to catalysts for the production of useful resources, such as in multi-electron reductions of carbon dioxide and water–gas shift reactions [4,5,6,7,8,9,10,11], bipyridine analogues are also utilized as photosensitizers [12] and phosphorescence materials [13]. A variety of studies that imparted selectivity to various reactions by strictly controlling the coordination sphere have been reported for many ruthenium complexes. Various molecular structures can be readily designed for a ruthenium center because ruthenium can take various oxidation states and can easily interact with the ligand. These structures include examples where differences in the coordination geometry dramatically changed both the electrical and photochemical properties of the complex, and therefore its catalysis for chemical reactions [14,15,16]. Many examples of the relationship between the supporting bipyridine ligands and the reactivity of metal complexes are known. For example, the relationship between the number of heteroatoms involved in the supporting ligand and the reactivity of the complex has been reported in a ruthenium complex containing bipyridine analogues [17].

Recently, we reported the synthesis of ruthenium complexes with 2,2’-bipyridine (bpy) and an analogue, the asymmetric bidentate ligand 2-(2-pyridyl)-1,8-naphthyridine (pynp, Chart 1). The supporting ligands affected the properties of the complexes, and diastereomers with these ligands showed different reactivities [18]. As is the case in some of the studies mentioned above, the difference in reactivities based on coordination geometry or the effect of the supporting ligand on the properties of the complex have been reported independently, but there are very few examples that investigated the synergistic effects of both. In this study, we prepared a series of Ru(II) complexes (Chart 2) containing the asymmetric bidentate pynp and a symmetric bidentate ligand N^N, [Ru(pynp)(N^N)(CO)₂]²⁺ (N^N = bpy or 1,10-phenanthroline (phen) as a more rigid ligand, Chart 1), as a platform for investigating these synergistic effects [19]. In addition to characterizing the above complexes, we performed photo- and thermal reactions on the complexes, along with the bis–pynp complex [Ru(pynp)₂(CO)₂]²⁺ as a reference compound. We comprehensively evaluated the synergistic effects of the coordination geometry and supporting ligands on the reactivity.

2. Results and Discussion

2.1. Diastereoselective Synthesis and Characterization of Ruthenium Complexes with the Asymmetric pynp Ligand

2.1.1. Synthesis and Structure of Complexes

Unlike unsubstituted 2,2’-bipyridine and 1,10-phenanthroline, the pynp ligand has no C2 symmetry axis (Chart 1), so metal complexes containing pynp(s) may have several diastereomers. We previously reported that for ruthenium carbonyl complexes containing both bpy and pynp, formation of the desired isomers can be controlled by the order of introduction of the two ligands [18,20]. Using this method, dicarbonylruthenium complexes containing both pynp and phen were successfully prepared. As in the previous study, the distal isomer is produced preferentially when phen is introduced first (Figure 1a). In the case of N^N = bpy, the product obtained by this method is almost all the d-form, whereas in the phen system a small quantity of the corresponding p-form (10–20%) is also formed. This may be due to energy differences between the d- and p-isomers. Density functional theory (DFT) calculations for each complex system suggest that the p-isomers are more stable, but the energy difference between the d- and p-isomers in the phen system is smaller than that of the corresponding bpy system (N^N = bpy: 4.3 kcal/mol, N^N = phen: 1.9 kcal/mol). The DFT calculations indicate that the minor formation of the p-isomer in addition to the d-isomer is due to the relatively small energy difference in the phen system.

In contrast, initial introduction of the asymmetric pynp ligand leads to the selective formation of p-isomers (Figure 1b). Considering the molecular structure of an isolated and analyzed intermediate cis(solv)-[Ru(pynp)(CO)₂(solv)₂]²⁺; Supplementary Materials Figure S1), it is reasonable to propose that an intramolecular interaction between the non-coordinating nitrogen atom in the pynp ligand and the adjacent coordinated carbonyl promotes the formation of p-isomers [18]. In addition, only the dp-isomer was selectively isolated in the case of [Ru(pynp)₂(CO)₂]²⁺ (Figure 2), even though the bis-pynp complex could form three diastereomers (pp, dp, and dd; Chart 3). According to DFT calculations for the three diastereomers, the total energy of the experimentally formed dp-isomer was 3.79 kcal/mol higher than that of the most stable pp-isomer, which appears to contradict the experimental result. We then performed total energy calculations for the two possible intermediates in the coordination of the second pynp ligand to the ruthenium center by assuming that the aqua ligand in the trans position of the coordinated pynp is substituted earlier than the aqua ligand in the cis position (see the structure diagram in Supplementary Materials Figure S1). The results indicate that both intermediates have almost the same energy (ΔE = 0.08 kcal/mol; Figure S2). Therefore, the second pynp ligand may coordinate through a kinetically favorable pathway rather than a thermodynamic one to give the selective dp-isomer.

Table 1. Selected bond and interatomic distances (Å) and angles (°) for dicarbonyl complexes.

Parameter	[1d]2+ 2	[1p]2+ 3	[2d]2+	[2p]2+	[3dp]2+
Ru–C	1.905 (3)	1.925 (11)	1.903 (9)	1.917 (4)	1.919 (10)
	1.900 (3)	1.910 (10)	1.888 (9)	1.911 (4)	1.886 (8)
			1.909 (9)
			1.895 (10)
C–O	1.134 (4)	1.123 (13)	1.145 (11)	1.133 (5)	1.132 (10)
	1.124 (4)	1.119 (12)	1.123 (11)	1.122 (5)	1.120 (11)
			1.133 (11)
			1.121 (11)
Ru–C–O	176.9 (3)	177.6 (8)	177.0 (8)	176.0 (4)	175.7 (7)
	176.6 (3)	174.3 (9)	175.5 (8)	171.6 (4)	175.6 (8)
			177.4 (8)
			177.1 (8)
C…N 1	-	2.671 (14)	-	2.607 (5)	2.754 (14)

¹ Distance between the non-coordinating nitrogen in pynp and the CO carbon atoms. ² [18]. ³ [20].

summarizes the structural parameters of a series of [Ru(pynp)(N^N)(CO)₂]²⁺ complexes. The data around the ruthenium center and the {Ru–CO}²⁺ moieties are within the typical range of dicarbonylruthenium(II) complexes that have polypyridines as supporting ligands [4,21,22]. In addition, the interatomic distances between the carbonyl carbon and the non-coordinating nitrogen of pynp (2.607(5) to 2.754(14) Å) in the three p-forms are much shorter than the sum of van der Waals radii (3.25 Å) [23]. The shortening suggests that pynp and the adjacent CO ligand interact considerably in these complexes.

2.1.2. Characterization of Complexes

Spectroscopic and electrochemical analyses were performed on the synthesized complexes (

Table 2. Spectral and electrochemical data for dicarbonyl complexes.

Technique	[1d]2+	[1p]2+	[2d]2+	[2p]2+	[3dp]2+
IR (νCO/cm−1)	2097	2097	2099	2093	2094
	2044	2045	2045	2044	2040
UV–vis (λmax/nm)	339 (1.78)	343 (2.90)	342 (1.64)	349 (2.05)	336 (2.93)
(ε/×104 M−1 cm−1)	314 (2.22)	315 (2.63)	273 (3.19)	272 (3.79)
CV (E/V vs. Fc+/Fc)	−1.09 1	−1.07 1	−1.12 1	−1.09 1	−1.11 1
	−1.45 1	−1.48 1	−1.59 1	−1.49 1	−1.29 1
	−0.98 2		−1.05 2	−1.03 2	−1.03 2
	(−0.78) 3		(−0.87) 3		(−0.79) 3

¹E_pc. ²E_pa. ³ Data in parentheses represent oxidation waves caused by the second reduction.

). Two strong IR bands assignable to νCO were observed around 2100 and 2040 cm⁻¹ in all complexes. These values were similar to those in other dicarbonyl-ruthenium(II) complexes [4,21,22], and no clear differences were observed between the isomers. Given that the complexes contain two carbonyl ligands of the highest order in the spectrochemical series, no obvious absorption was observed in the visible region (Supplementary Materials Figure S3). However, intense polypyridyl-centered π–π* intraligand transitions were observed in the UV region [24].

Although spectroscopic measurements did not show any marked differences between isomers, electrochemistry clearly showed different behavior. The cyclic voltammograms (CV) of all the complexes showed multiple ligand-based reduction waves in the range of −1 to −2 V vs. Fc⁺/Fc. Since pynp is more easily reduced than bpy or phen, the first reduction peak at ca. −1 V was attributed to the reduction of the pynp ligand [25]. When the cathodic scan was immediately reversed after the first peak potential, the coupled oxidation wave was reversible in the d-isomers (Figure 3a, dotted line); however, those of the corresponding p-isomers were either irreversible or quasi-reversible at the first reduction wave (Figure 3b). Although the first reduction wave of [2p]²⁺ in Figure 3b appears to be reversible, the coupled anodic current decreased as the scan rates slowed (Supplementary Materials Figure S4). The basicity of the free nitrogen atom of the 1,8-naphthyridine moiety in pynp increased due to one-electron reduction of pynp, thus making intramolecular nucleophilic attack to the adjacent carbonyl carbons possible, leading to the formation of a metallacyclic compound (Scheme 1) [26]. Due to this structural change, only the p-isomer exhibited irreversible one-electron reduction behavior.

2.2. Reactivities of Complexes

2.2.1. Photochemical Reactions

Polypyridylruthenium complexes are generally photoreactive. When cis-[Ru(bpy)₂(CO)₂]²⁺ in acetonitrile is irradiated with light, two carbonyl ligands simultaneously dissociate and the corresponding solvent complex (cis-[Ru(bpy)₂(CH₃CN)₂]²⁺) is produced [27,28]. However, when pynp is substituted for one of the two bpy ligands, the two carbonyl groups dissociate stepwise due to their electronic non-equivalence [18]. Since no comparisons between the d- and p-isomers were made in the previous report, we compared the photoreactivities of both isomers. As previously reported, there are two reaction steps for both isomers (Scheme 2) [18,29]. In the first step, one carbonyl ligand was dissociated, and, at the same time, the acetonitrile used as the solvent became coordinated (the first step in Scheme 2 and Figure 4a). Steric retention of the complex was supported by structural analysis of the isolated species at this step (Figure 5a). In the subsequent step, photoisomerization from the d- to p-isomer subsequent slow dissociation of the second CO ligand and solvent coordination occurred (the second step in Scheme 2a and Figure 4b). Reaction analysis and DFT calculations suggest that the formation of photoexcited states of the complex promote such an isomerization to a more stable p-form in similar complexes containing pynp [30,31]. In the bis-pynp complex ([3dp]²⁺), a similar two-step photoreaction (Scheme 2a) was confirmed from structural analyses of both products (dp-form in Figure 5b and pp-form in Supplementary Materials Figure S5). On the other hand, the final structure of the p-isomers was unchanged from spectroscopic analyses (Scheme 2b).

From comparisons of the first CO-dissociation rates (

Table 3. Rates of the first photoreaction step (298 K) of the dicarbonyl complexes.

Rate	[1d]2+	[1p]2+	[2d]2+	[2p]2+	[3dp]2+
kobs/sec−1	12 × 10−3	6.5 × 10−3	2.4 × 10−3	1.3 × 10−3	7.6 × 10−3

), it was found that the photoreactions of the d-isomers were always faster than those of the corresponding p-isomers (ca. two times in both complexes). This difference was interpreted to be due to their geometries, based on photoreaction behavior seen in similar diastereomers [32]. The pynp ligand has a naphthyridine unit in place of the pyridine unit in bpy (or phen), which is both more delocalized and a π-acceptor. The superior charge acceptor properties of the naphthyridine unit in the d-isomers leads to better labilization of the trans-CO, and thus the d-isomers exhibit faster CO release compared with the corresponding p-isomers, despite having lower extinction coefficients between 300 and 400 nm (Table 2 and Supplementary Materials Figure S3). We also found that the overall photoreaction of the bpy system proceeded more smoothly than that of the corresponding phen system. This was probably because phen compounds were significantly more stable than their bpy analogues, based on the rigidity of phen [33]. Despite prolonged photoirradiation, incomplete dissociation of the second CO ligand in the phen system consequently prevented isolation of the single disubstituted complex ([Ru(pynp)(phen)(CH₃CN)₂]²⁺).

2.2.2. Thermochemical Reactions

As shown in Scheme 3, in thermal reactions one carbonyl ligand of a dicarbonylruthenium(II) complex undergoes nucleophilic attack from solvent molecules [34,35]. Thus, when the d-isomers of the dicarbonyl complexes, which are expected to be more reactive (see Section 2.2.1.), were heated in water/acetonitrile or alcohol (methanol or ethanol)/acetonitrile mixtures, one of the coordinated CO moieties underwent nucleophilic attack from the solvent to the CO ligand at the trans position of pynp. As expected, molecular structures of the isolated complexes showed the formation of the hydroxycarbonyl (–C(O)OH in Figure 6a) and the methoxy- or ethoxycarbonyl (–C(O)OC₂H₅ in Figure 6b and Supplementary Materials Figure S6) complexes. Given that the pynp ligand is more π-acidic than bpy or phen, the CO carbon in the trans position of pynp has a more positive charge.

We next investigated other thermochemical reactions using monocarbonyl complexes that do not undergo nucleophilic attack on the coordinated carbonyl. When the monocarbonyl complex (d-[Ru(pynp)(phen)(CO)(CH₃CN)]²⁺), which was produced by the photoreaction described in Section 2.2.1, was heated in acetone, it isomerized from the d- to the p-isomer (Figure 7). Notably, this thermal isomerization reaction was accelerated more than 70 times when water was added to the solution. A similar isomerization was observed in d-[Ru(pynp)(bpy)(CO)(CH₃CN)]²⁺. In contrast, such thermal isomerization could not be confirmed for other monocarbonyl complexes with anionic ligands (–COR⁻; R⁻ = OH, OCH₃, or OC₂H₅). These results strongly suggest that the thermal isomerization reaction involves a donor-acceptor interaction between the solvent and the complex. In ruthenium(II) complexes, terminal CO ligands that interact with the donor solvent tend to exhibit νCO IR frequencies over 2000 cm⁻¹ [36,37]. In this study, the CO stretching frequency of the thermally isomerized acetonitrile complexes exceeded 2000 cm⁻¹ (2008 and 2009 cm⁻¹), while those of complexes containing an anionic ligand were 1950–1960 cm⁻¹. It would therefore be expected that weak interactions between the coordinated carbonyl and the solvent (acetone or water) could induce thermal isomerization. That is, the interaction between the coordinated CO and the solvent significantly changes the electronic state of the ruthenium center, resulting in lowering the activation energy for isomerization of the d-isomer to the thermally more stable p-isomer [38].

3. Materials and Methods

3.1. General Remarks

All chemicals were purchased from commercial sources and used without further purification unless otherwise stated. All solvents used for the syntheses were anhydrous. Acetonitrile for electrochemical measurements was purified by passing through solvent purification columns (Glass Contour, Laguna, CA, USA). Then 2-(2-pyridyl)-1,8-naphthyridine (pynp), [Ru(CO)₂Cl₂]_n, trans-(Cl)-[Ru(bpy)(CO)₂Cl₂], trans-(Cl)-[Ru(phen)(CO)₂Cl₂], cis(OH₂)-[Ru(bpy)(CO)₂(OH₂)₂](CF₃SO₃)₂, cis-(OH₂)-[Ru(phen)(CO)₂(OH₂)₂](CF₃SO₃)₂, and cis-(OH₂)-[Ru(pynp)(CO)₂(OH₂)₂](CF₃SO₃)₂ were prepared according to previously reported procedures [18,39,40].

IR spectra were obtained using a JASCO FT–IR 4100 spectrometer (Tokyo, Japan). Electrospray ionization mass spectrometry (ESI–MS) data were obtained using a Bruker Daltonics micrOTOF spectrometer. Electronic spectra were obtained on a JASCO V-560 spectrophotometer. The ¹H and ¹³C{¹H}-NMR spectra were acquired on a JEOL JMN-AL300 spectrometer (Tokyo, Japan) operating at ¹H and ¹³C frequencies of 300 and 75.5 MHz, respectively. Elemental analysis data were obtained on a PerkinElmer 2400II series CHN analyzer (Yokohama, Japan). Electrochemical measurements were performed on an electrochemical analyzer (ALS/CHI model 660E, Tokyo, Japan) with a solution of the complex in acetonitrile (1 mM) and n-Bu₄NClO₄ (0.1 M) as a supporting electrolyte in a cell consisting of a glassy carbon working electrode (ϕ = 1.6 mm), a Pt wire counter electrode, and Ag/AgNO₃ (0.01 M) as the reference electrode. All potentials are reported in volts versus the ferrocenium/ferrocene couple (Fc⁺/Fc) under Ar at 25 °C. DFT calculations were performed using the quantum computation software Gaussian 09 [41]. The geometries of the Ru complexes were fully optimized using a restricted DFT method employing the B3LYP function [42,43], with a 6-31G(d) basis set for the light elements (H, C, N, and O) [44,45] and a LanL2DZ basis set [46] for the Ru atom. The solvent effect of acetonitrile was evaluated using an implicit solvent model and a polarizable continuum model. The vibrational analyses were performed at the same calculation level employed for geometry optimization.

3.2. Synthesis of Complexes

3.2.1. Synthesis of the Distal Isomers ([1d](PF₆)₂ and [2d](PF₆)₂)

[Ru(bpy)(CO)₂(OH₂)₂](CF₃SO₃)₂ (53 mg, 0.082 mmol) and pynp (20 mg, 0.098 mmol) were added to methanol (10 mL). The mixture was refluxed with stirring for 1.5 h. The reaction vessel was cooled to room temperature. The reaction mixture was condensed to 3 mL under reduced pressure. A light orange precipitate formed on addition of a saturated aqueous solution of KPF₆ to the mixture. The product was collected by filtration, washed with cold water and diethyl ether, and dried in vacuo to obtain the product (56 mg, 84%). Anal. Calcd for [1d](PF₆)₂: C₂₅H₁₇N₅O₂F₁₂P₂Ru: C, 37.05; H, 2.11; N, 8.64. Found: C, 37.40; H, 2.03; N, 8.51. ESI–MS (CH₃CN): m/z 260.5 ([M]²⁺), 246.5 ([M–CO]²⁺). IR (KBr): 2097, 2044 cm⁻¹ (νCO). ¹H-NMR (acetone-d₆): δ 9.76–9.71 (m, 2H), 9.25 (d, J = 8.4 Hz, 1H), 9.09–8.99 (m, 2H), 8.77–8.71 (m, 2H), 8.64 (dd, J = 6.6, 1.8 Hz, 2H), 8.57 (td, J = 6.3, 1.5 Hz, 1H), 8.49 (dd, J = 1.8, 1.5 Hz, 1H), 8.28–8.25 (m, 2H), 8.17–8.13 (m, 1H), 7.76–7.69 (m, 2H), 7.66–7.63 (m, 1H). ¹³C{¹H}-NMR (acetone-d₆): δ 192.08 and 191.49 (CO), 158.89, 158.73, 157.02, 156.71, 156.61, 156.16, 150.19, 145.33, 143.16, 142.76, 141.93, 139.61, 131.27, 130.70, 129.41, 129.07, 128.95, 127.20, 126.18, 125.92, 125.49, 125.39, 122.23. A similar reaction between [Ru(phen)(CO)₂(OH₂)₂](CF₃SO₃)₂ (60 mg, 0.089 mmol) and pynp (23 mg, 0.109 mmol) gave a mixture of both [2d](PF₆)₂ and [2p](PF₆)₂. Yield: 63 mg (84%). The crude product was recrystallized from a mixture of acetonitrile, acetone, and diethyl ether. Anal. Calcd for [2d](PF₆)₂: C₂₇H₁₇N₅O₂F₁₂P₂Ru·CH₃CN·0.5C₂H₆CO: C, 40.50; H, 2.56; N, 9.29. Found: C, 40.38; H, 2.29; N, 9.14. ESI–MS (CH₃CN): m/z 272.6 ([M]²⁺), 258.6 ([M–CO]²⁺). IR (KBr): 2099, 2045 cm⁻¹ (νCO). ¹H-NMR (acetone-d₆): δ 10.16 (dd, J = 5.1, 1.2 Hz, 1H), 9.80 (dd, J = 5.8, 0.9 Hz, 1H), 9.28 (dd, J = 7.6, 0.9 Hz, 1H), 9.18 (dd, J = 8.5, 1.2 Hz, 1H), 8.99 (s, 2H), 8.87 (dd, J = 8.3, 1.2 Hz, 1H), 8.80 (td, J = 8.1, 1.5 Hz, 1H), 8.52–8.47 (m, 2H), 8.39–8.23 (m, 4H), 8.08 (dd, J = 5.4, 1.5 Hz, 1H), 7.95 (dd, J = 8.1, 5.4 Hz, 1H), 7.60 (dd, J = 8.4, 4.5 Hz, 1H). ¹³C{¹H}-NMR (acetone-d₆): δ 192.33 and 191.47 (CO), 160.51, 159.49, 158.91, 157.29, 156.00, 155.95, 151.31, 147.34, 147.26, 145.17, 143.21, 141.68, 140.94, 139.41, 132.07, 132.06, 131.24, 129.16, 129.10, 128.46, 127.63, 127.59, 125.94, 125.83, 122.27.

3.2.2. Synthesis of the Proximal Isomers ([1p](PF₆)₂ and [2p](PF₆)₂)

Using a protocol similar to that described for the synthesis of [1d](PF₆)₂, the reaction between [Ru(pynp)(CO)₂(OH₂)₂](CF₃SO₃)₂ (106 mg, 0.161 mmol) and bpy (31 mg, 0.197 mmol) or phen (39 mg, 0.197 mmol) gave [1p](PF₆)₂ or [2p](PF₆)₂ in 81% and 86% yield, respectively. Anal. Calcd for [1p](PF₆)₂: C₂₅H₁₇N₅O₂F₁₂P₂Ru: C, 37.05; H, 2.11; N, 8.64. Found: C, 37.22; H, 2.01; N, 8.60. ESI–MS (CH₃CN): m/z 260.5 ([M]²⁺), 246.5 ([M–CO]²⁺). IR (KBr): 2097, 2045 cm⁻¹ (νCO). ¹H-NMR (acetone-d₆): δ 9.64 (d, J = 5.1 Hz, 1H), 9.49 (d, J = 3.9 Hz, 1H), 9.36 (d, J = 8.4 Hz, 1H), 9.17–9.11 (m, 2H), 9.02–8.99 (m, 2H), 8.86 (d, J = 8.1 Hz, 1H), 8.69 (t, J = 3.8 Hz, 1H), 8.50 (t, J = 4.7 Hz, 1H), 8.35 (t, J = 4.8 Hz, 1H), 8.22–8.16 (m, 2H), 8.02 (d, J = 6.3 Hz, 1H), 7.84–7.77 (m, 2H), 7.50 (t, J = 7.4 Hz, 1H). ¹³C{¹H}-NMR (acetone-d₆): δ 193.12 and 192.20 (CO), 161.36, 158.14, 157.00, 156.88, 156.71, 155.91, 153.77, 151.45, 151.15, 145.42, 143.02, 142.94, 142.87, 140.57, 130.70, 130.65, 129.66, 128.19, 127.20, 126.92, 126.76, 126.03, 122.99. Anal. Calcd for [2p](PF₆)₂: C₂₇H₁₇N₅O₂F₁₂P₂Ru·0.5CH₃CN·0.5C₂H₆CO: C, 40.08; H, 2.45; N, 8.72. Found: C, 39.90; H, 2.17; N, 8.68. ESI–MS (CH₃CN): m/z 272.6 ([M]²⁺), 258.6 ([M–CO]²⁺). IR (KBr): 2093, 2044 cm⁻¹ (νCO). ¹H-NMR (acetone-d₆): δ 10.04 (dd, J = 5.1, 1.2 Hz, 1H), 9.48 (dd, J = 3.9, 1.8 Hz, 1H), 9.40 (d, J = 8.4 Hz, 1H), 9.30 (dd, J = 8.4, 1.2 Hz, 1H), 9.18 (d, J = 8.7 Hz, 1H), 9.10 (d, J = 7.8 Hz, 1H), 9.03 (dd, J = 7.8, 1.8 Hz, 1H), 8.96 (dd, J = 8.4, 0.9 Hz, 1H), 8.56–8.38 (m, 4H), 8.23–8.17 (m, 2H), 7.88 (d, J = 4.8 Hz, 1H), 7.82 (dd, J = 9.0, 5.4 Hz, 1H), 7.60 (dd, J = 7.5, 1.2 Hz, 1H). ¹³C{¹H}-NMR (acetone-d₆): δ 192.08 (CO), 161.61, 158.85, 157.05, 156.69, 153.96, 152.34, 151.63, 147.09, 146.35, 145.47, 142.88, 142.05, 141.81, 140.59, 132.99, 132.47, 130.39, 129.46, 129.38, 128.88, 128.13, 127.84, 127.19, 126.77, 123.01.

3.2.3. Synthesis of [3dp](PF₆)₂

A similar reaction between [Ru(pynp)(CO)₂(OH₂)₂](CF₃SO₃)₂ (58 mg, 0.083 mmol) and pynp (20 mg, 0.098 mmol) in methanol (20 mL) gave [3dp](PF₆)₂. Yield: 67 mg (94%). Anal. Calcd for [3dp](PF₆)₂: C₂₈H₁₈N₆O₂F₁₂P₂Ru·H₂O: C, 38.23; H, 2.29; N, 9.56. Found: C, 37.95; H, 1.94; N, 9.32. ESI–MS (CH₃CN): m/z 286.1 ([M]²⁺), 272.1 ([M–CO]²⁺). IR (KBr): 2094, 2040 cm⁻¹ (νCO). ¹H-NMR (acetone-d₆): δ 9.85 (d, J = 4.8 Hz, 1H), 9.49 (dd, J = 2.2, 1.8 Hz, 1H), 9.28 (d, J = 8.4 Hz, 2H), 9.06–9.01 (m, 3H), 8.85–8.82 (m, 2H), 8.78 (t, J = 1.5 Hz, 1H), 8.53 (dd, J = 6.6, 1.8 Hz, 1H), 8.39–8.16 (m, 3H), 7.88 (d, J = 4.8 Hz, 1H), 7.75 (t, J = 1.5 Hz, 1H), 7.66 (dd, J = 2.2, 1.8 Hz, 1H), 7.54–7.49 (m, 1H). ¹³C{¹H}-NMR (acetone-d₆): δ 192.58 and 192.27 (CO), 160.76, 160.25, 158.70, 157.76, 157.19, 155.92, 155.86, 151.01, 145.17, 144.22, 142.97, 142.70, 140.27, 139.57, 139.51, 131.42, 130.27, 129.16, 127.64, 127.58, 126.57, 126.11, 125.68, 125.63, 122.23, 122.03.

3.3. Photochemical Reactions

Photochemical reactions of the complexes were conducted in the degassed CH₃CN. The solution was irradiated with UV–visible light (λ = 300–400 nm) through a cutoff filter using an Asahi Spectra MAX-302 (Tokyo, Japan) with a xenon lamp (0.1 mW cm⁻²). In a typical reaction, an acetonitrile solution of [1d](PF₆)₂ (0.10 mM) was placed in a quartz flask with a stopper and irradiated with a xenon lamp for 30 min. The reaction mixture was condensed to 1 mL under reduced pressure. Addition of diethyl ether to the solution resulted in the formation of a precipitate of the mono- or disubstituted complexes in moderate yields (65–75%).

d-[Ru(pynp)(bpy)(CO)(CH₃CN)](PF₆)₂: ESI–MS (CH₃CN): m/z 246.5 ([M–CH₃CN]²⁺). IR (KBr): 2009 cm⁻¹ (νCO). Electronic spectrum (CH₃CN): λ_max/nm (ε/M⁻¹ cm⁻¹) 250 (31,200), 312 (18,500), 341 (19,400), 480 (950). p-[Ru(pynp)(bpy)(CH₃CN)₂](PF₆)₂: ESI–MS (CH₃CN): m/z 232.5 ([M–(CH₃CN)₂]²⁺). Electronic spectrum (CH₃CN): λ_max/nm (ε/M⁻¹ cm⁻¹) 239 (26,400), 288 (33,600), 407 (4400), 473 (6500). d-[Ru(pynp)(phen)(CO)(CH₃CN)](PF₆)₂: ESI–MS (CH₃CN): m/z 258.5 ([M–CH₃CN]²⁺). IR (KBr): 2008 cm⁻¹ (νCO). Electronic spectrum (CH₃CN): λ_max/nm (ε/M⁻¹ cm⁻¹) 268 (33,400), 341 (16,700), 440 (1600). dp-[Ru(pynp)₂(CO)(CH₃CN)](PF₆)₂: ESI–MS (CH₃CN): m/z 272.0 ([M–CH₃CN]²⁺), 292.5 ([M]²⁺). IR (KBr): 2007 cm⁻¹ (νCO). Electronic spectrum (CH₃CN): λ_max/nm (ε/M⁻¹ cm⁻¹) 241 (40,200), 340 (32,000), 420_sh (2600). pp-[Ru(pynp)₂(CH₃CN)₂](PF₆)₂: ESI–MS (CH₃CN): m/z 258.0 ([M–(CH₃CN)₂]²⁺), 278.5 ([M–CH₃CN]²⁺). Electronic spectrum (CH₃CN): λ_max/nm (ε/M⁻¹ cm⁻¹) 315 (40,900), 496 (6400). Kinetic measurements for the photoreaction were followed at an appropriate wavelength for each complex (476 nm for [1d]²⁺, 478 nm for [1p]²⁺, 475 nm for [2d]²⁺ and [2p]²⁺, and 497 nm for [3dp]²⁺), and the logarithm of the complex concentration versus time plots were generated for the first step of the reaction in each measurement.

3.4. Thermochemical Reactions

3.4.1. Reaction of the Coordinated CO in [1d]²⁺ and [2d]²⁺ with Solvents

The [1d](PF₆)₂ (50 mg, 0.0617 mmol) was dissolved in acetonitrile (5 mL) and a suitable cosolvent (water, ethanol, or methanol) (30 mL) and refluxed for 24 h. The volume was reduced to 1 mL using a rotary evaporator. Addition of diethyl ether to the solution resulted in precipitation of the reaction products ([Ru(pynp)(bpy)(CO)(C(O)R)]PF₆, R = OH, OC₂H₅, or OCH₃) in moderate to high yields (50%–82%). [Ru(pynp)(bpy)(CO)(C(O)OH)]PF₆: ESI–MS (CH₃CN): m/z 538.1 ([M]⁺). IR (KBr): 1957 cm⁻¹ (νCO). [Ru(pynp)(bpy)(CO)(C(O)OC₂H₅)]PF₆: ESI–MS (CH₃CN): m/z 566.1 ([M]⁺). IR (KBr): 1954 cm⁻¹ (νCO). [Ru(pynp)(bpy)(CO)(C(O)OCH₃)]PF₆: ESI–MS (CH₃CN): m/z 552.1 ([M]⁺). IR (KBr): 1954 cm⁻¹ (νCO).

[Ru(pynp)(phen)(CO)(C(O)R)](PF₆) complexes were also isolated in 50 to 75% yields using a similar procedure starting from [2d](PF₆)₂. [Ru(pynp)(phen)(CO)(C(O)OH)]PF₆: ESI–MS (CH₃CN): m/z 562.0 ([M]⁺). IR (KBr): 1960 cm⁻¹ (νCO). [Ru(pynp)(phen)(CO)(C(O)OC₂H₅)]PF₆: ESI–MS (CH₃CN): m/z 590.2 ([M]⁺). IR (KBr): 1961 cm⁻¹ (νCO). [Ru(pynp)(phen)(CO)(C(O)OCH₃)]PF₆: ESI–MS (CH₃CN): m/z 576.1 ([M]⁺). IR (KBr): 1960 cm⁻¹ (νCO).

3.4.2. Thermal-induced Isomerization Using Monocarbonyl Complexes d-[Ru(pynp)(N^N)(CO)(CH₃CN)]²⁺ (N^N = bpy or phen) and dp-[Ru(pynp)₂(CO)(CH₃CN)]²⁺

In a typical reaction, d-[Ru(pynp)(phen)(CO)(CH₃CN)](PF₆)₂ (4 mg, 0.005 mmol) was dissolved in acetone (20 mL) and refluxed until the solution became dark red (~72 h). The solution was condensed using a rotary evaporator, and addition of diethyl ether resulted in the formation of a precipitate. After cooling overnight, the isomerized product was collected by filtration, washed with diethyl ether, and then dried in vacuo. The yield was 3 mg (75%). A change of the solvent (acetone/water, 1:20 v/v) brought about a considerable reduction in reaction time (1 h). Similar reactions using d-[Ru(pynp)(bpy)(CO)(CH₃CN)](PF₆)₂ or dp-[Ru(pynp)₂(CO)(CH₃CN)](PF₆)₂ gave the corresponding p- or pp-isomer, respectively. These products were characterized by spectroscopic and X-ray crystallographic analyses.

p-[Ru(pynp)(bpy)(CO)(CH₃CN)](PF₆)₂: ESI–MS (CH₃CN): m/z 246.5 ([M–CH₃CN]²⁺). IR (KBr): 1992 cm⁻¹ (νCO). Electronic spectrum (CH₃CN): λ_max/nm (ε/M⁻¹ cm⁻¹) 251 (31,300), 312 (19,600), 343 (21,400), 383 (4700), 490 (1200). p-[Ru(pynp)(phen)(CO)(CH₃CN)](PF₆)₂: ESI–MS (CH₃CN): m/z 258.5 ([M–CH₃CN]²⁺). IR (KBr): 1994 cm⁻¹ (νCO). Electronic spectrum (CH₃CN): λ_max/nm (ε/M⁻¹ cm⁻¹) 268 (34,500), 332_sh (11,000), 521 (2200). pp-[Ru(pynp)₂(CO)(CH₃CN)](PF₆)₂: ESI–MS (CH₃CN): m/z 272.0 ([M–(CH₃CN)]²⁺), 292.5 ([M]²⁺). IR (KBr): 1994 cm⁻¹ (νCO). Electronic spectrum (CH₃CN): λ_max/nm (ε/M⁻¹ cm⁻¹) 245 (31,200), 269 (26,900), 343 (28,800), 423_sh (3000), 499 (1000).

3.5. X-ray Crystallographic Analyses

Single crystals of [2d]₂(PF₆)₃(CF₃SO₃)·H₂O, [2p](PF₆)₂·CH₃CN, [3dp](CF₃SO₃)₂, and d-[Ru(pynp)(phen)(CO)(CH₃CN)](PF₆)₂ were obtained by vapor diffusion of diethyl ether into an acetonitrile/acetone solution of the complex. p-[Ru(pynp)(phen)(CO)(CH₃CN)](PF₆)₂·CH₃OH crystals were obtained by vapor diffusion of diethyl ether into an acetonitrile/acetone/methanol solution of the complex. [Ru(pynp)(CO)₂(OH₂)(CH₃CN)](CF₃SO₃)₂·C₂H₅OH crystals were obtained by vapor diffusion of diethyl ether into an acetonitrile/acetone/ethanol solution of the complex. Single crystals of dp-[Ru(pynp)₂(CO)(CH₃CN)](PF₆)₂·(CH₃)₂CO and [Ru(pynp)(bpy)(CO)(C(O)OH)]PF₆·(CH₃)₂CO were obtained by vapor diffusion of diethyl ether into an acetone solution of the complex. Single crystals of pp-[Ru(pynp)₂(CH₃CN)₂](PF₆)₂ were obtained by vapor diffusion of diethyl ether into an acetonitrile solution of the complex. Single crystals of d-[Ru(pynp)(bpy)(CO)(C(O)OC₂H₅)]PF₆·(CH₃)₂CO and d-[Ru(pynp)(phen)(CO)(C(O)OC₂H₅)]CF₃SO₃ were obtained by layering diethyl ether over an acetone/ethanol solution of the complex. Crystal structure determination and refinement data for the complexes are given in Figure 1, Figure 2, Figure 5, Figure 6 and Figure 7, Supplementary Materials Figures S1, S5 and S6, and

Table 4. Crystallographic data for dicarbonyl complexes.

Parameter	[2d]	[2p]	[3dp]
Chemical formula	C55H34F21N10O8P3Ru2S	C29H20F12N6O2P2Ru	C30H18F6N6O8RuS2
Formula weight	1689.02	875.51	869.69
Temperature (K)	93	93	93
Crystal system	monoclinic	triclinic	triclinic
Space group	P21/c	P-1	P-1
a (Å)	13.2730(3)	8.57310(10)	9.68729(18)
b (Å)	22.5773(6)	12.7558(3)	13.0592(3)
c (Å)	20.2047(5)	15.8850(4)	13.6489(3)
α (°)	90	82.856(6)	83.1931(7)
β (°)	93.7571(8)	75.919(6)	81.5465(7)
γ (°)	90	74.843(6)	72.0749(7)
V (Å3)	6041.7(3)	1622.90(8)	1620.10(6)
Z	4	2	2
Calcd density (g/cm3)	1.857	1.792	1.783
μ (Mo Kα) (mm–1)	0.744	0.691	0.709
No. unique reflns	62,024	16,768	16,909
No. obsd reflns	13,818	7331	7304
Refinement method	Full-matrix least-squares on F2
Parameters	901	470	478
R (I > 2σ(I)) 1	0.0941	0.0552	0.1119
wR (all data) 2	0.2536	0.1483	0.2633
S	1.155	1.077	1.189

¹R = Σ(||F_o| − |F_c||)/Σ|F_o|; ² wR = {Σ_w(F_o² − F_c²)²/Σ_w(F_o²)²}^1/2.

,

Table 5. Crystallographic data for [Ru(pynp)(N^N)(CO)(CH₃CN)]²⁺ (N^N = phen or pynp).

Parameter	d-(phen)	p-(phen)	dp-(pynp)
Chemical formula	C28H20F12N6OP2Ru	C29H20F12N6O2P2Ru	C32H27F12N7O2P2Ru
Formula weight	847.50	875.51	932.61
Temperature (K)	93	93	93
Crystal system	orthorhombic	triclinic	monoclinic
Space group	Pna21	P-1	P21/n
a (Å)	13.8734(9)	11.9822(4)	10.067(4)
b (Å)	24.7954(15)	12.5252(3)	32.588(12)
c (Å)	9.0869(5)	13.2975(4)	11.329(5)
α (°)	90	84.1114(11)	90
β (°)	90	87.5838(8)	109.861(5)
γ (°)	90	63.326(2)	90
V (Å3)	3125.9(3)	1773.87(9)	3496(2)
Z	4	2	4
Calcd density (g/cm3)	1.801	1.639	1.772
μ (Mo Kα) (mm−1)	0.712	0.632	0.648
No. unique reflns	30,815	18,631	35,054
No. obsd reflns	7100	8064	7963
Refinement method	Full-matrix least-squares on F2
Parameters	452	436	508
R (I > 2σ(I)) 1	0.0506	0.0832	0.0745
wR (all data) 2	0.1237	0.2503	0.1623
S	1.028	1.049	1.113

¹R = Σ(||F_o| − |F_c||)/Σ|F_o|; ² wR = {Σ_w(F_o² − F_c²)²/Σ_w(F_o²)²}^1/2.

and

Table 6. Crystallographic data for [Ru(pynp)(bpy)(CO)(COR)]⁺ (R = OH or OC₂H₅).

Parameter	R = OH	R = OC2H5
Chemical formula	C28H24F6N5O4PRu	C30H28F6N5O4PRu
Formula weight	740.56	768.62
Temperature (K)	93	93
Crystal system	triclinic	monoclinic
Space group	P-1	P21/n
a (Å)	8.35369(15)	17.2474(5)
b (Å)	12.9208(3)	8.1493(2)
c (Å)	13.3343(3)	22.1009(6)
α (°)	88.9616(7)	90
β (°)	79.0700(7)	102.3249(8)
γ (°)	88.2204(7)	90
V (Å3)	1412.36(5)	3034.77(14)
Z	2	4
Calcd density (g/cm3)	1.741	1.682
μ (Mo Kα) (mm−1)	0.697	0.652
No. unique reflns	12,245	30,159
No. obsd reflns	4934	6960
Refinement method	Full-matrix least-squares on F2
Parameters	406	463
R (I > 2σ(I)) 1	0.0583	0.0475
wR (all data) 2	0.1331	0.1132
S	1.219	1.067

¹R = Σ(||F_o| − |F_c||)/Σ|F_o|; ² wR = {Σ_w(F_o² − F_c²)²/Σ_w(F_o²)²}^1/2.

and Supplementary Materials Table S1. CCDC-1966877–1966887 contains the supplementary crystallographic data for this paper.

4. Conclusions

This study successfully formed non-equivalent CO-ligand environments by lowering the molecular symmetry of the well-known benchmark complex, [Ru(bpy)₂(CO)₂]²⁺. In addition to the diastereoselective synthesis of the desired complexes, the relationship between reaction selectivity and coordination geometry was demonstrated using redox properties. We found that the monoacetonitrile complexes undergo thermal isomerization in the d-isomers, whereas the diacetonitrile complexes undergo photoisomerization to give the corresponding p-form. This study will conduct further research on synthetic chemistry, stereochemistry, and structure–reactivity relationships in metal complexes.