Synthesis, crystal structure, electrochemical properties, and photophysical characterization of ruthenium(II) 4,4′-dimethoxy-2,2′-bipyridine polypyridine complexes

Abstract A series of ruthenium(II) polypyridine complexes of the type [Ru(tpy)((CH3O)2bpy)(4-R-py)]2+, where tpy = 2,2′;6′,2″-terpyridine, (CH3O)2bpy = 4,4′-dimethoxy-2,2′-bipyridine, and 4-R-py = pyridine (py, 2), 4-methoxypyridine (4-CH3O-py, 3), 4-aminopyridine (4-NH2-py, 4), were synthesized and their crystal structures, electronic absorption, luminescence, and electrochemical properties were investigated. The effect of adding electron-donating groups to the bidentate and monodentate ligand was investigated and compared with [Ru(tpy)(bpy)(py)]2+ (1) where bpy = 2,2′-bipyridine. While anticipated trends were not observed for the Ru-N(6) bond length as 4-R-py was varied, noticeable modifications of the measured photophysical properties were observed. A red-shift of the metal-to-ligand charge transfer (1MLCT) is observed from 466 nm in 1 to 474 nm, 478 nm, and 485 nm for 2–4, respectively. Additionally, a red-shift in the luminescence maxima is observed in 2–4 as compared to 1, with 4 exhibiting the greatest shift of more than 100 nm. Complexes 2–4 exhibited luminescence quantum yields of 2.7 × 10−4, 7.2 × 10−4, and 7.4 × 10−4, respectively, which are increased compared to the quantum yield of 2.0 × 10−4 in 1. These findings demonstrate systematic tuning of absorbance and luminescence properties of ruthenium polypyridine complexes by addition of π-donating substituents to the monodentate and bidentate ligand. Graphical Abstract

Complexes with greater photostability and reduced susceptibility to photo-induced ligand dissociation are imperative for applications in sensing and dye-sensitized solar cells [9,10]. Complexes of the type [Ru(tpy)(bpy)(4-R-py)] 2þ , where R¼ electron donating groups, such as methoxy and amino, have been shown to shut down the excited state ligand dissociation pathway thereby increasing luminescence [11]. [Ru(tpy)(bpy)(4-aminopyridine)] 2þ has an approximate 1.8-fold decrease in the rate of photoinduced ligand dissociation as compared to [Ru(tpy)(bpy)(py)] 2þ in CH 3 CN with k irr. ¼ 450 nm [11]. Additionally, the p-donating amino group induces a 10 nm red-shift in the lowest energy electronic band, which corresponds to a metal-to-ligand charge transfer ( 1 MLCT). The phosphorescence maximum (k em ) of [Ru(tpy)(bpy)(4-aminopyridine)] 2þ is red-shifted by 78 nm from 618 to 696 nm as compared to [Ru(tpy)(bpy)(py)] 2þ , and the quantum yield of emission increases by fifty percent. The example of [Ru(tpy)(bpy)(4-aminopyridine)] 2þ , with amino as a p-donating group on the py ligand, demonstrates that p-donating substituents in the 4 position of the py ligand induce a red-shift in absorption and emission, a decreased photoinduced ligand dissociation, and an increased emission quantum yield.
In this work, a series of complexes of the type [Ru(tpy)((CH 3 O) 2 bpy)(4-R-py)] 2þ , where (CH 3 O) 2 bpy ¼ 4,4 0 -methoxy-2,2 0 -bipyridine and 4-R-py ¼ py (2), 4-methoxypyridine (3), and 4-aminopyridine (4), were synthesized and their structures and photophysical properties were investigated (Figure 1). In light of the previous work described above, the (CH 3 O) 2 bpy ligand was chosen to examine the effect of more strongly p-donating methoxy groups on the bpy ligand, in addition to the p-donating groups of methoxy and amino on the pyridine ligand, in hopes of discovering complexes that will provide excellent low energy absorption and emission qualities while being stable against unwanted ligand loss.

X-Ray crystallography
Single crystals of 2-4 were obtained through similar procedures. Each compound was dissolved in a mixture of acetonitrile, methanol, and chloroform and then layered with diethyl ether. The solutions were then stored in a À25 C freezer and allowed to slowly evaporate until small dark red crystals formed.
Crystals were placed onto the tips of glass optical fibers and mounted on a Bruker SMART platform diffractometer equipped with an APEX II CCD area detector for data collection [18]. For each crystal a preliminary set of cell constants and an orientation matrix were calculated from reflections harvested from three orthogonal wedges of reciprocal space. Full data collections were carried out using MoKa radiation (0.71073 Å, graphite monochromator) with frame times ranging from 45 to 60 seconds and at detector distances of approximately four cm. Randomly oriented regions of reciprocal space were surveyed: four or five major sections of frames were collected with 0.50 steps in x at four or five different u settings and a detector position of À38 in 2h. The intensity data were corrected for absorption [19]. Final cell constants were calculated from the xyz centroids of approximately 4000 strong reflections from the actual data collections after integration [20].
Structures were solved using SHELXT [21] and refined using SHELXL [22]. Space groups were determined based on systematic absences or just intensity statistics. Direct-methods solutions were calculated which provided most non-hydrogen atoms from the E-map. Full-matrix least squares/difference Fourier cycles were performed which located the remaining non-hydrogen atoms. All non-hydrogen atoms were refined with anisotropic displacement parameters. In 4, the amino group's hydrogen atoms were found from the difference Fourier map and refined freely. All other hydrogen atoms were placed in ideal positions and refined as riding atoms with relative isotropic displacement parameters. Full matrix least squares refinements on F 2 were run to convergence. In 4, highly disordered solvent was found in channels along [001]. Reflection contributions from this solvent were fixed and added to the calculated structure factors using the SQUEEZE routine of the program Platon [23], which determined there to be 52 electrons in 221 Å 3 treated this way. Because the exact amount and identity of the solvent was not known, no solvent was included in the atom list or molecular formula. Thus all calculated quantities that derive from the molecular formula (e.g. F(000), density, molecular weight, etc.) are known to be inaccurate.
Crystallographic data for 2-4 are listed in Table 1.

Structure
The structure of 1 was previously reported [24], and the structures of 2-4 are shown in Figure 2. The three new complexes exhibit distorted octahedral geometry about the metal center. The Ru-N bond lengths from 2-4 show no significant difference compared to the structure of 1 ( Table 2). Due to the large deviation in bond lengths for 1, 2-4 were analyzed separately for structural trends. The shortest Ru-N bond in all three complexes occurs to the central tpy nitrogen, with the two terminal tpy Ru-N bond lengths being 0.11-0.12 Å longer. The Ru-N(2) bond of 2 is only slightly longer than 3, with no significant difference observed between 3 and 4. The Ru-N(3) bond of 4 is only slightly longer than 2, with no significant differences observed between 3 and either 2 or 4. The Ru-N(1) bond shows no difference between 2, 3, and 4. The asymmetric coordination of the bipyridine results in a typical bpy Ru-N bond length cis to the tpy ligand and a longer Ru-N bond cis to the pyridine ligand. This elongation of the Ru-N(4) bond has been ascribed to steric crowding between the pyridine ligand and the bpy ligand [24]. However, on investigating the bond length in comparison to torsion between pyridine and bipyridine, no clear trend is observed ( Figure 3 and Table 2). However, we do notice slight trends with respect to Hammett parameters (Table  S2) of the pyridine substituents and Ru-N(5) and Ru-N(4) bond lengths ( Figure 3). There is a decrease in bond length of Ru-N(5) as the Hammett parameter increases and there is an increase in the Ru-N(4) bond length as the Hammett parameter increases. The Ru-N bond to the pyridine ligand is the longest in all four complexes. We had expected to see a trend in the Ru-N(6) bond lengths of 2-4 when compared to the Hammett parameters; however, while 3 is longer than 4, neither are significantly different from unsubstituted 2 and there is no trend observed (Figure 3).
Differences due to substitution at the 4,4 0 -postitions of the bipyridine were also analyzed by comparing 2 to previously reported [Ru(tpy)(4,4 0 -dimethylbpy) (pyridine)](PF 6 ) 2 , 5, and [Ru(tpy)(4,4 0 -bis(trifluoromethyl)bpy)(pyridine)](PF 6 ) 2, 6 (Table 2) [25]. As the Hammett parameter of the substituent (Table S2) on the bipyridine ligand becomes more positive, the Ru-N(4) bond decreases in length. The unusual length of the Ru-N(4) bond has previously been attributed to sterics [24]; however, in this series we see that bond length increases with increasing dihedral angle between the bipyridine and pyridine ( Figure S10). This may indicate that the Ru-N(4) bond length is  impacted by both electronics and sterics. As is typical for these types of complexes, the Ru-N(2) bond is the shortest in all these complexes, where 6 has the longest Ru-N(2) bond of the three complexes, possibly due to the trans influence, though 2 and 5 do not differ significantly from each other ( Figure S10).
The Ru-N(5) and Ru-N(6) bond lengths, respectively, show no significant differences across 2, 5, and 6. Finally, when analyzing the Ru-N(1) and Ru-N(3) bonds, 5 stands out the most for its differences in lengths. There is no obvious explanation for this, however, in viewing the distortion of the tpy ligand by analysis of the C15-N3-Ru1-N4 and C1-N1-Ru1-N4 dihedrals (Table S3), it may be possible that distortions accrue because of crystal packing.

Electronic absorption, emission, and electrochemistry
The absorption maxima, molar extinction coefficients, emission maxima and quantum yields, and redox potentials are summarized in Table 3. The cyclic voltammogram of 1 has been previously reported [11]. The oxidation potential of þ1.25 V versus SCE in CH 3 CN was attributed to Ru III/II couple, and the reduction potential of À1.25 V versus SCE in CH 3 CN was attributed to the reduction of the tpy ligand. The reduction waves of 2-4 are in the range of À1.29 to À1.33 V versus SCE in CH 3 CN, which is similar to 1 and other tpy containing Ru(II) poylpyridine complexes and are attributed to the reduction of the tpy ligand. Complex 2 exhibits an oxidation wave at þ1.11 V versus SCE in CH 3 CN, which is similar to 1 and other Ru(II) polypyridyl complexes and are also assigned to the Ru II/III couple. Complexes 3 and 4 exhibit an oxidation wave at þ1.07 and þ1.00 V versus SCE, respectively, which are also attributed to the Ru II/III couple. The observed cathodic shift of 2-4 from þ1.11 to 11.00 V versus SCE has a linear correlation with Hammett constants of the R-groups in 2-4 ( Figure S14). This shift is due to the increasing strength of the p-donating groups, which destabilize the Ru(dp) orbitals.
The electronic absorption spectra of 1-4 are shown in Figure 4. The electronic absorption of 1 has been previously reported and exhibits characteristic p-p Ã transitions at 288 and 312 nm localized in the tpy and bpy ligands [11]. Electronic absorptions of 2-4 exhibit transitions at 272 and 315 nm that can be attributed to p-p Ã transitions localized on the tpy and (CH 3 O) 2 bpy ligands. Complex 1 exhibits a metal-to-ligand charge transfer ( 1 MLCT) at 466 nm, which has been reported as Ru(dp)!tpy/bpy(p Ã ) [11]. Similarly, the 1 MLCT of 2 is observed at 474 nm and is attributed to the Ru(dp)!tpy/(CH 3 O) 2 bpy(p Ã ). This represents an 8 nm red-shift due to methoxy substituents on the (CH 3 O) 2 bpy. The 1 MLCT is observed at 478 and 485 nm for 3 and 4, respectively. A linear correlation is observed with the 1 MLCT absorption maxima (k abs ) and Hammett parameters of the 4-R-py ligand in 2-4 ( Figure S14). This red-shift of 4 nm in   -E red ) versus k abs of 1-4 ( Figure S15) shows the decrease in DE redox corresponds to a red-shift in the 1 MLCT of 1-4, respectively. Methoxy groups on the bpy in 2-4 and the methoxy and amino groups on the py in 3 and 4, respectively, increase the p-donating ability of the ligands, which causes a destabilization of the Ru(dp) HOMO orbitals. Conversely, the reduction potential of the tpy(p Ã ) LUMO is relatively stable. Taken together, this will decrease the energy gap between the ground state and the 1 MLCT, which explains the observed red-shift of the 1 MLCT in 1-4. This is in agreement with previous studies [11]. The overall 19 nm red-shift of the 1 MLCT from 466 nm in 1 to 485 nm in 4 demonstrates the combined effect of electron-donating groups added to the bpy and py ligands.
The luminescence spectra of 1-4 in CH 3 CN at 298 K are shown in Figure 5. The luminescence of 1 has been previously reported and exhibits an emission maximum (k em ) at 618 nm (U em ¼ 0.00020) [11], which has been assigned to the Ru ! tpy/bpy 3 MLCT state. Emission maxima of 2-4 are observed at 697, 695, and 720 nm, respectively, and are assigned to the Ru ! tpy/(CH 3 O) 2 bpy 3 MLCT state. The 79 nm red-shift from 1 to 2 demonstrates the effect of increased electron-donation of the (CH 3 O) 2 bpy ligand compared to unsubstituted bpy. The shift caused by the addition of two methoxy groups to the bpy ligand is over four times greater than the shift of 17 nm observed from only the addition of one methoxy group to the py ligand in [Ru(tpy)(bpy)(4-CH 3 O-py)] 2þ [11]. This suggests that the addition of p-donating groups on the bpy ligand is more effective at destabilizing Ru(dp) orbitals, causing a large red-shift in k em . Conversely, the U em increased 35% from 0.00020 in 1 to 0.00027 in 2, but increased 115% from 1 to [Ru(tpy)(bpy)(4-CH 3 O-py)] 2þ (0.00043) [11]. Previous reports suggest that more p-donating ligands cause an increase in U em by destabilizing the Ru(r Ã ) orbitals thereby increasing the energy of the triplet ligand field ( 3 LF) state, widening the DE between the 3 MLCT and the 3 LF, which makes a competing pathway to the phosphorescence less likely [11]. Taken together, these data suggest that the addition of CH 3 O groups on the bpy ligand has a greater effect on the Ru(dp) orbitals and the red shift of the k em , whereas, the addition of CH 3 O on the py ligand has a greater effect on the Ru(r Ã ) orbitals and U em . This is further confirmed by emission data from 3, which exhibits a similar red-shift of its emission maximum as 2 compared to 1, however the U em increases 160% to 0.00072 in 3 as compared to 2. Complex 4 has the strongest r-donating ligand in 4-NH 2 -py and further red-shifts the k em to 720 nm and has the highest U em at 0.0074. This seems to indicate that the stronger p-donating groups will continue to increase the red-shift of the k em , however, the effect on U em may plateau. This plateau could be a result of non-radiative decay from the triplet state coming into play due to the decrease in DE between the ground state and triplet excited state. Given the relationship of the r-donating ligands and k em discussed above, we were surprised to observe that a linear correlation is not observed in the plot of k em versus the Hammett parameters in 2-4 ( Figure S14). The results presented here show that the k em and U em can be independently tuned by placement of the p-donating groups on either the bidentate or monodentate ligand.

Conclusion
The synthesis and structure of three new complexes, [Ru(tpy)((CH 3 O) 2 bpy)(py)] 2þ (2), [Ru(tpy)((CH 3 O) 2 bpy)(4-MeO-py)] 2þ (3), and [Ru(tpy)((CH 3 O) 2 bpy)(4-MeO-py)] 2þ (4), are reported, and their properties were investigated and compared to [Ru(tpy)(bpy)(py)] 2þ (1). The addition of electron-donating methoxy groups to the bpy as well as methoxy and amino groups to the py generally showed a red-shift of the absorbance and emission maxima and an increase in quantum yield of emission. However, the p-donating groups on the bpy ligand have a greater effect on the emission maxima, and the p-donating groups on the py ligand have a greater effect on the quantum yield of emission. These results demonstrate the ability to tune the photophysical properties of Ru(II) polypyridine complexes, which may be useful in studying and developing dye-sensitized solar cells and luminescent sensors.