Introduction

Quinoline-based compounds have important biological applications ranging from anti-cancer agents [1], anti-inflammatory [2] and antimalarial agents [3]. When used as antimalarials, quinoline-based drugs such as quinine, chloroquine and mefloquine are believed to inhibit hemozoin formation by interacting with the prosthetic heme group of hemoglobin [4], as well as the hemozoin crystals [5]. The mechanism of the hemozoin inhibition process has been the subject of interest in recent years [6]. The heme group is the most common target of many antimalarial drugs especially during the pathogenic asexual blood stage level of the life cycle of the Plasmodium parasite [7]. As a result, efforts have been made to fully understand the mechanism by which heme activates antimalarials.

In the literature, several spectroscopic techniques [8,9,10,11,12], computational methods [13, 14] and X-ray crystallography [11] have been used to shed light on the role of the antimalarial drugs in inhibiting hemozoin. In this work, we were interested in preparing relatively stable synthetic heme model-antimalarial adducts to help us characterize them spectroscopically, determine their solid-state structures, and study their electrochemical behavior. Ruthenium(II) porphyrin carbonyls are good candidates for this purpose, since generally they are low-spin, diamagnetic and relatively more stable than the iron species [15]. Due to the larger d-orbitals of Ru compared to Fe, a stronger metal-to-ligand backbonding is expected in ruthenium(II) porphyrin carbonyls, thus allowing for easy isolation and characterization of products. Although chemically different from the iron(II) counterpart, this robust nature of the ruthenium(II) complexes has allowed researchers to use Ru in place of Fe to model the properties of (por)Fe active sites [16,17,18].

Traditionally, ruthenium(II) porphyrin complexes (e.g. ruthenium porphyrin nitrosyls) [19] are synthesized, first, by inserting the Ru-CO fragment into porphyrins. The resulting precursor ruthenium(II) carbonyl porphyrin complex, often formulated as (por)Ru(CO), possesses a weakly coordinated solvent molecule trans to CO. The solvent molecule in the resulting (por)Ru(CO)(solvent), which is introduced during workup, can be replaced by stronger π-donor and/or σ-donor ligands. For example, the six-coordinate ruthenium porphyrin carbonyls, (OEP)Ru(CO)(py) and (OEP)Ru(CO)(Im) have been prepared by this method [20]. We were interested in expanding the study on the reaction of (por)Ru(CO) with quinoline-based ligands. To this end, we have prepared the ruthenium(II) porphyrin complex, (OEP)Ru(CO) (OEP = 2,3,7,8,12,13,17,18-octaethylporphyrinato dianion) as a synthetic heme active site and studied its reactivity with quinine (QN) in forming the (OEP)Ru(CO)(QN) adduct. We have examined the site of binding of the QN ligand to the ruthenium center (OEP)Ru(CO) by comparing the spectroscopic properties of the adducts formed between (OEP)Ru(CO) and quinoline (Qnl) (Fig. 1). We have obtained the X-ray crystal of (OEP)Ru(CO)(Qnl) to assist in understanding the site of binding of the quinoline-based drugs with the Ru center. We have also studied the redox behaviors of the (OEP)Ru(CO)(Q) adducts.

Fig. 1
figure 1

Structures of the ligands used in this work: a quinoline; b Quinine

Full size image

Experimental section

General procedures

Unless otherwise stated, all reactions and manipulations were performed under an atmosphere of nitrogen using standard Schlenk glassware. Reagents were purchased from commercial sources (see below) and used as received, except noted otherwise. Dichloromethane and n-hexane were deaerated by a three-cycle freeze–pump–thaw and dried over a 4 Å molecular sieves before use.

The free-base porphyrin, 2,3,7,8,12,13,17,18-Octaethyl-21H,23H-porphine [(OEP)H2, 97%], triruthenium dodecacarbonyl [Ru3(CO)12, 99%], decalin (≥ 99%), dichloromethane (CH2Cl2, ≥ 99.8%), n-hexane (95%), quinine (QN, 98%), tetrabutylammonium hexafluorophosphate (NBu4PF6, ≥ 99%) and ferrocene (Fc, 98%) were purchased from Sigma-Aldrich. Quinoline (Qnl, 99%) and Chloroform-d (CDCl3, 99.8 atom %D, 1 v/v% TMS) were purchased from Thermo Fisher Scientific.

Infrared spectra were collected at room temperature on a Perkin-Elmer Spectrum 65 FT-IR Spectrophotometer equipped with ATR accessory. The 1H NMR spectra were obtained on a Varian Mercury 400 MHz spectrometer at 25 °C and the signals were referenced to the residual signal of the solvent employed (CHCl3 at 7.26 ppm). Coupling constants are reported in Hz. UV–Visible spectra were collected on a Thermo Scientific Evolution 201 UV–Visible Spectrophotometer.

ESI mass spectra were obtained on a Micromass Q-TOF mass spectrometer. Elemental analyses were obtained by staff of Atlantic Microlab, Norcross, GA. X-ray diffraction data were collected using a D8 Quest κ-geometry diffractometer with a Bruker Photon II cmos area detector [21, 22] and an Incoatec I µ s microfocus Mo Kα radiation (λ = 0.71073 Å). Cyclic voltammetry measurements were performed using a Gamry Interface 1000B Potentiostat/Galvanostat/ZRA. In all the electrochemical experiments, a three-electrode cell was utilized and consisted of a 3.0 mm diameter Pt disk working electrode, a Pt wire counter electrode, and a Ag/AgCl wire as reference electrode. Solutions were deaerated before use by passing a stream of N2 gas through the solution for a minimum of 10 min. A blanket of N2 was maintained over the solution while performing the experiments. The electrochemical experiments were performed in solutions containing 0.1 M NBu4PF6 and 1.0 mM of the analyte. Ferrocene, Fc (1.0 mM) was used as internal standard for the electrochemical experiments and potentials were referenced to the Fc/Fc+ couple at 0.00 V.

Synthesis of compounds

Synthesis of (OEP)Ru(CO) (1)

The compound, (OEP)Ru(CO) (1) was prepared by a slight modification of the method reported in the literature [23]. In this work, the free-base porphyrin, (OEP)H2 (503.9 mg, 0.942 mmol) and Ru3(CO)12 (613.6 mg, 0.960 mmol) were refluxed in decalin (100 mL) for 4 h. The resulting dark-red solution was concentrated, then purified by column chromatography (alumina) using 100% dichloromethane to a 2:1 dichloromethane/acetone solvent gradient as eluent. The product was crystallized from n-hexane and dried in vacuo to give a bright-red precipitate characterized as 1 (583.0 mg, 93% isolated yield). IR (ATR, cm−1):υCO = 1918 (vs), 1922 (s) (in CH2Cl2). Also: 2966 (m), 2930 (m), 2867 (m), 1683 (w), 1590 (m), 1539 (w), 1464 (m), 1447 (m), 1375 (m), 1319 (w), 1272 (s), 1227 (m), 1144 (m), 1110 (m), 1056 (m), 1017 (s), 991 (m), 962 (s), 922 (w), 840 (s), 745 (s), 712 (m). 1H NMR (CDCl3, 400 MHz): δ (ppm) 9.96 (s, 4 H, pyrrole-H of OEP), 4.04 (q, J = 7.7 Hz, 16 H, CH2), 1.94 (t, J = 7.6 Hz, 24 H, CH3). UV–vis: (CH2Cl2):λmax, nm (log ε) 392 (4.80) Soret, 547 (4.02), 515 (3.65), 302 (3.71), 246 (3.80). Suitable crystals for X-ray crystallography were obtained by slow evaporation of a CH2Cl2/n-hexane (2:1) solution of 1 and were characterized as the six-coordinate monohydrate complex, (OEP)Ru(CO)(H2O).

Preparation of (OEP)Ru(CO)(Qnl) (2)

A 50 mL Schlenk tube equipped with a magnetic stir bar was charged with 1 (101 mg, 0.153 mmol) and CH2Cl2 (10 mL). The mixture was stirred at room temperature, under N2 atmosphere resulting in a red solution. An excess amount of quinoline (90 µL, 0.746 mmol) was added to the solution and stirred for 48 h. During this period, the solution changed color from red to dark-magenta. After 48 h of stirring, the resulting solution was slowly reduced in vacuo to approximately 3 mL. n-Hexane (10 mL) was then added. The solvent was slowly removed under reduced pressure, and the resulting solid was carefully washed with cold n-hexane (3 mL). The product was dried in vacuo to give a red precipitate that was characterized as 2 (99.8 mg, 83% isolated yield). IR (ATR, cm−1): υCO = 1938 (vs), 1932 (s) (in CH2Cl2). Also: 2964 (m), 2930 (m), 2868 (m), 1896 (w), 1634 (w), 1599 (w), 1536 (w), 1510 (w), 1464 (m), 1371 (m), 1316 (w), 1271 (m), 1230 (m), 1147 (m), 1109 (m), 1056 (m), 1017 (s), 991 (m), 959 (s), 860 (w), 838 (m), 800 (m), 783 (m), 746 (m), 711 (m). 1H NMR (CDCl3, 400 MHz): δ (ppm) 9.92 (s, 4 H, pyrrole-H), 7.76 (d, J = 8.6 Hz, 1 H, Qnl-H), 7.49 (app d, J = 7.7 Hz, 1 H, Qnl-H), 7.35 (app d, J = 6.9 Hz, 1 H, Qnl-H), 7.32 (app d, J = 4.9 Hz, 1 H, Qnl-H), 7.28 (app d, J = 6.5 Hz, 1 H, Qnl-H H), 6.86 (app d, J = 4.8 Hz, 1 H, Qnl-H), 6.25 (br s, 1 H, Qnl-H), 4.01 (dq, J = 7.7 Hz, 3.1 Hz, 16 H, CH3CH2), 1.92 (t, J = 7.7 Hz, 24 H, CH3CH2). UV–Vis: (CH2Cl2): λmax, nm (log ε) 392 (5.34) Soret, 547 (4.51), 515 (4.18), 314 (4.32), 302 (4.32), 229 sh (4.71), 227 (4.73). ESI mass spectrum (TOF), m/z: 833.2400 for [M + MeCN + H]+; 764.1798 for [M – CO + H]+; 741.1968 for [M – Qnl + DMSO + H]+. Anal. Calcd. for C46H51N5ORu∙0.5H2O: C, 69.06; H, 6.55; N, 8.75. Found: C, 68.92; H, 6.63; N, 8.37. Suitable crystals for X-ray crystallography were obtained by slow evaporation of a CH2Cl2/ n-hexane solution of the compound at room temperature and characterized as structure 2A. In a separate crystallization process, a second batch of single crystals were obtained and were determined to be the same structure as 2A, but with quite different parameters (vide infra). We refer to this second crystal structure as 2B.

Preparation of (OEP)Ru(CO)(QN) (3)

A 50 mL Schlenk tube equipped with a magnetic stir bar was charged with 1 (50.2 mg, 0.0758 mmol) followed by CH2Cl2 (10 mL). The solution was stirred under N2 atmosphere resulting in a red solution. The red solution was treated with a slight excess of quinine, QN (40.1 mg, 0.123 mmol) and the solution was allowed to stir for 48 h. After 48 h, the resulting dark-red solution was slowly reduced to 3 mL under vacuum. n-Hexane (10 mL) was added and the mixture was stirred vigorously for 1 min. After several hours of sitting undisturbed, precipitates began to form. The supernatant was discarded. The red precipitate was washed twice (two 5 mL portions) with n-hexane, and the supernatant discarded each time. The precipitate was dried under vacuum to give a red solid characterized as 3 (52.4 mg, 70% isolated yield). IR (ATR, cm−1): 1931 (vs), 1930 (s) (in CH2Cl2). Also, 3152 br (w), 2963 (m), 2930 (m), 2870 (m), 1688 (w), 1623 (m), 1591 (w), 1510 (m), 1467 (m), 1451 (m), 1431 (m), 1374 (m), 1272 (m), 1241 (m), 1229 (m), 1150 (m), 1108 (m), 1056 (m), 1032 (w), 1019 (s), 992 (m), 960 (m), 919 (w), 840 (m), 744 (m), 717 (m), 647 (m). 1H NMR (CDCl3, 400 MHz): δ (ppm) 9.81 (s, 4 H, pyrrole-H), 7.10 (dd, J = 6.2 Hz, J = 3.2 Hz, 2 H, qnl-H), 6.60 (br s, 1 H, qnl-H), 5.89 (br s, 1 H, qnl-H), 5.36 (m, J = 8.0 Hz, 1 H, CH = CH2), 4.76 (overlapping d, J = 10.8 Hz, 2 H, CH2 = CH), 4.70 (s, 1 H, CH(OH)), 4.26 (br s, 1 H qnl-H), 3.92 (app q, J = 6.7 Hz, 16 H, CH2CH3), 3.23 (s, 3 H, OCH3), 1.84 (t, J = 7.6 Hz, 24 H, CH3CH2), 2.26 – 1.26 (overlapping m, 11 H, quinuclidine-H’s). UV–Vis (CH2Cl2): λmax, nm (log ε) 392 (5.11) Soret, 547 (4.28), 515 (3.95), 331 (4.19), 320 (4.19), 232 (4.64), 227 (4.59). ESI mass spectrum (TOF), m/z: 1333.5198 for [M + QN + Na]+; 663.2582 for [M – QN + H]+; 325.1913 for [QN + H]+. Anal. Calc. for C56H68N6O3Ru∙0.5H2O: C, 68.79; H, 6.99; N, 8.44 Found: C, 68.96; H, 7.08; N, 8.08.

X-ray crystallography

A red plate-shaped crystal of 1 and of dimensions 0.110 × 0.172 × 0.202 mm was selected for structural analysis. Intensity data for this compound were collected using a D8 Quest κ geometry diffractometer with a Bruker Photon II cmos area detector [21, 22] and an Incoatec Iμs microfocus Mo Kα source (λ = 0.71073 Å). The sample was cooled to 100(2) K. Cell parameters were determined from a least-squares fit of 9810 peaks in the range 2.81 < θ < 31.55°. A total of 107,368 data were measured in the range 2.304 < θ < 31.585° using φ and ω oscillation frames. The data were corrected for absorption by the empirical method [24] giving minimum and maximum transmission factors of 0.5989 and 0.6482. The data were merged to form a set of 10,990 independent data with R(int) = 0.0320 and a coverage of 99.7%.

The triclinic space group P1 was determined by statistical tests and verified by subsequent refinement. The structure was solved by dual-space methods and refined by full-matrix least-squares methods on F2 [25, 26]. The positions of hydrogens bonded to carbons were initially determined by geometry and were refined using a riding model. Hydrogens bonded to O2 were located on a difference map, and their positions were refined with a riding model. Non-hydrogen atoms were refined with anisotropic displacement parameters. Hydrogen atom displacement parameters were set to 1.2 times the isotropic equivalent displacement parameters of the bonded atoms. A total of 397 parameters were refined against 10,990 data to give wR(F2) = 0.0634 and S = 1.004 for weights of w = 1/[σ2 (F2) + (0.0300 P)2 + 1.4000 P], where P = [Fo2 + 2Fc2] / 3. The final R(F) was 0.0232 for the 10,374 observed, [F > 4σ(F)], data. The largest shift/s.u. was 0.002 in the final refinement cycle. The final difference map had maxima and minima of 1.239 and -0.692 e/Å3, respectively. Structural diagrams were prepared using Mercury 2021.3.0 program [27]. A summary of the crystal and structure refinement data is shown in Table 1.

Table 1 Crystal data and structure refinement for 1, 2A and 2B
Full size table

A red block-shaped crystal, 2A of dimensions 0.074 × 0.126 × 0.264 mm was selected for structural analysis. Intensity data for this compound were collected and the structure solved as described for 1 (Table 1). Crystal structure 2B with monoclinic space group P21/n was determined similarly as indicated by the crystal data and structure refinement data in Table 1.

Results and discussion

Preparation, spectroscopy and X-ray crystallography

The ruthenium(II) porphyrin complexes, (OEP)Ru(CO)(Q) (Q = Qnl (2) and QN (3)) were prepared by reacting a dichloromethane solution of (OEP)Ru(CO) (1) with quinoline (Qnl) and quinine (QN), respectively (Scheme 1). The reactions occurred at room temperature over a 24 – 48 h period, producing red solids at 70 – 83% isolated yields. The compounds are air-stable as solids at room temperature, and showed no evidence of decomposition over several days as judged by IR and 1H NMR spectroscopy.

Scheme 1
scheme 1

Synthesis of (OEP)Ru(CO)(Q) [Q = Quinoline, Qnl (2); QN (3)] from (OEP)Ru(CO) (1)

Full size image

Compounds 1 – 3 were characterized by IR spectroscopy both in the solid form and in CH2Cl2. Table 2 lists the υCO bands for compounds 1 – 3. These IR bands are in the range of the υCO of ruthenium(II) carbonyls. For example, solid samples of compound 2 have a strong υCO band at 1938 cm−1, which is in the same range as those of other N-bound complexes, (OEP)Ru(CO)(t-Bu(py)) (υCO = 1935 cm−1) [28], (OEP)Ru(CO)(py) and (OEP)Ru(CO)(Im) (υCO = 1933 cm−1) [20], but higher (by 20 cm−1) than that of the precursor complex, 1. Similarly, a CH2Cl2 solution of 2 displayed υCO bands at 1932 cm−1 in the IR spectrum. Thus, in 2, based on the IR spectral data, Qnl is coordinated to the axial position of the ruthenium(II) center through the quinoline-N.

Table 2 The υNO bands of the compounds 1 – 3
Full size table

The υCO band in 3 was observed at 1931 cm−1, which is 13 cm−1 higher than that of the precursor compound 1. This υCO band is in the same range as that of 2, and is an indication that the site of binding is though the quinoline-N, and not the other basic sites (i.e., hydroxyl group, the quinuclidine-N, or the olefin). The υCO band in 3 is, as expected, lower than that of 2. This parallels the stronger donor effect of quinine compared to quinoline. Quinine donates more electron density for backdonation into the RuCO moiety. The IR spectrum of 3 also displays medium intensity peaks at 1623, 1510, 1431, and 1241 cm−1 due to vibrations related to the QN ligand. The proposed binding sites of the ligands to Ru in compounds 2 and 3 are supported by 1H NMR data (vide infra).

The 1H NMR spectrum of 1 showed only the expected peaks for the OEP porphyrin. Even as solids, ruthenium(II) porphyrin carbonyls such as (TPP)Ru(CO), and possibly 1 could easily pick up a water molecule if they are exposed to air [29]. Indeed, the X-ray crystal structure of 1 (vide infra) revealed H2O is coordinated to Ru, although there were no peaks associated with H2O in its 1H NMR spectrum. The 1H NMR spectroscopic data for compounds 1–3 revealed the peaks associated with the porphyrin macrocycle. The porphyrin pyrrole-H peaks were observed at 9.96, 9.92 and 9.81 for 1, 2 and 3, respectively. Similarly, the peaks associated with the ethyl groups, CH3CH2 were displayed at the expected chemical shifts of ~ 4.0 (CH2) and ~ 1.9 (CH3).

In addition to the signals due to the OEP macrocycle, the 1H NMR spectra of compounds 2 and 3 displayed new peaks associated with the Q ligands. As expected, the associated Q proton peaks are shifted upfield due to ring current effect from the OEP macrocycle [30]. For example, the 1H NMR peaks for the bound Qnl in compound 2 appeared at δ 7.76 d (1 H), 7.49 app d (1 H), 7.35 app d (1 H), 7.32 app d (1H), 7.28 app d (1 H), 6.85 d (1 H) and 6.25 br s (1 H). The 1H NMR spectrum for the uncoordinated free-base Qnl ligand collected at similar experimental conditions and in the same solvent displayed the following peaks: δ 8.90 dd (1 H), 8.11 dd (1 H), 8.10 d (1 H), 7.78 d (1 H), 7.71 – 7.67 dddd (1 H), 7.53–7.50 dddd (1 H), 7.36 dd (1 H). Similar upfield δ shifts of ligand peaks have been recorded in (OEP)Ru(CO)(py) [20] and (OEP)Ru(CO)(Im) [28], as well as in O-bound six-coordinate ruthenium nitrosyl porphyrins [17, 18].

We then turned our attention to understanding the site of binding of QN to ruthenium in 3. Since the hydroxyl group, the quinuclidine N, the quinoline N and the olefin are possible sites of binding, we examined the effects of the ring current on the chemical shifts of the various protons in QN via the 1H NMR spectrum of 3. We anticipated that this information would help us determine the site of binding of QN in 3 in solution. The 1H NMR of 3 displayed new peaks associated with QN ligand and were observed more upfield relative to the free-base QN. The 1H NMR signals of the quinoline moiety in 3 were observed at 7.10 dd (2 H), 6.60 br s (1 H), 5.89 br s (1 H), 4.26 br s (1 H). Other QN signals were observed at 5.36 m (1 H) and 4.76 app d (2 H), and were assigned to the vinyl-H on the quinuclidine ring. The respective vinyl-H signals in the free-base QN ligand are 5.78 m (1 H) and 4.95 app d (2H). We note that the quinoline-H’s in 3 experienced the largest upfield δ shifts suggesting that the quinoline group is probably closest to the Ru center, and as a result experienced the largest effect of the porphyrin ring current. The proposed site of binding of QN is also supported by the fact that the H’s on the quinuclidine ring did not shift as much as those of the quinoline moiety. Signals due to the quinuclidine ring in 3 were observed as several overlapping multiplets between δ 2.26 and 1.26, thus, supporting binding of Ru through the quinoline N.

The UV–visible spectra of compounds 1 – 3 collected in dichloromethane revealed the expected Soret band at 392 nm, which is due to porphyrin a1u(π)–eg*(π) transitions. Additionally, the characteristic α and β bands (Q bands) attributed to a2u(π)–eg*(π) transitions were recorded at 547 and 515 nm, respectively [31]. There are however, distinguishing characteristic bands in the 220 – 350 nm region depending on the sixth ligand on the (OEP)Ru(CO) unit (Fig. 2). For example, the following peaks were observed in 3: 331, 320, 232 and 232 nm, and are associated with the quinine ligand [32]. In the quinoline complex, 2 bands at 314, 302, 229 (sh), and 227 nm were recorded and these were similarly associated with the Qnl ligand. Hence, UV–visible spectra of the compounds parallels the observed UV–visible spectrum of (OEP)Ru(CO)(Py) recorded in a similar solvent [33]. The UV–visible spectra of 2 and 3 both point to the presence of a quinoline functionality.

Fig. 2
figure 2

UV–visible spectra of compounds 1 (2.39 × 10–5 M), 2 (3.06 × 10–6 M) and 3 (4.51 × 10–6 M) collected in CH2Cl2 at room temperature. Showing signals in the 220 – 350 nm wavelength range. See full spectrum in the S12

Full size image

In order to gain insight into the solid-state structures of these complexes, we resorted to X-ray crystallography analysis of single crystal samples of 1 and 2. We obtained two crystal structures of 2 (2A and 2B) from different crystallization batches. Structures 2A and 2B have quite different cell parameters and the relative orientations of the ethyl groups relative to the porphyrin ring are different for these two compounds. The molecular structure of compound 2A is shown in Fig. 3a and that of 2B is shown in Fig. S6 (supplementary material). Selected bond lengths and angles of 2A, 2B and 1 are listed in Table 3.

Fig. 3
figure 3

a Molecular structure of 2A b Relative positions of adjacent porphyrin macrocycles in the crystal structure of 2A. Hydrogen atoms have been omitted for clarity

Full size image
Table 3 Selected bond lengths (Å) and bond angles (°) of 1, 2A and 2B
Full size table

As shown in Fig. 3, compound 2A is a six-coordinate ruthenium(II) porphyrin complex with a Qnl ligand and CO at the axial positions of the porphyrin macrocycle. The Qnl ligand is bound to Ru via its nitrogen atom. The Ru atom is displaced by only 0.06 Å from the 24-atom mean plane towards the Qnl ligand. The Qnl ligand plane is essentially perpendicular to the 24-mean plane of the porphyrin macrocycle with the angle between the two planes being 81.89°. In the crystal packing of 2A, the two Qnl groups of adjacent porphyrins have parallel-displaced configurations. The two Qnl planes are separated by 3.30 Å suggesting a relatively strong π -π interaction between the two quinoline ligands and mimics the interlayer structure of graphite [34]. These two adjacent porphyrins have a mean plane separation of 9.29 Å, and their relative positioning are laterally shifted at 3.55 Å (Fig. 3b). The porphyrin macrocycle is mildly ruffled with the ring slightly out-of-plane from the 24-mean porphyrin plane and specifically away from the Qnl ring (Fig. 4a,b).

Fig. 4
figure 4

a Molecular structure of 2A viewing from the Qnl ligand perpendicular to the 24-mean porphyrin plan. Hydrogen atoms have been omitted for clarity b: Perpendicular atom displacements (in Å ~ 100) of the porphyrin core from the porphyrin 24-atom mean plane

Full size image

Similarly, in the X-ray crystal structure of 2B the mean quinoline group plane is perpendicular to the 24-mean porphyrin plane at an angle of 82.05° (Fig. S8). The Ru atom is displaced by only 0.020 Å towards the Qnl ligand. In the crystal packing diagram, the two closest adjacent porphyrins have a mean plane separation of 3.79 Å and they are oriented head-to-head at the CO ends. As a result, their relative lateral shift is 7.06 Å, more than twice that of 2A. We note the porphyrin macrocycle in 2B is similarly mildly ruffled (Fig. S9b) and that several parts of 2B were disordered. This disorder is likely associated with small amounts of isoquinoline in the quinoline reagent that co-crystallized with quinoline in the crystallization of 2. We detected ~ 0.6% (by 1 HNMR spectroscopy) of isoquinoline in the quinoline sample used for the analysis. We plan to study the isoquinoline porphyrin complex in future work.

The Ru–Naxial(Naxial = quinoline N) bond distances in 2A is 2.3408(13) Å, which is longer than its average Ru–Np bond distance of 2.0574(12) Å and 2.055(3) Å, respectively. The Ru–Naxial bond distances are also longer than those recorded in (OEP)Ru(CO)(Im) (2.192(4) Å) [35], (OEP)Ru(CO)(Py) (2.239(2) Å) [36], and (OEP)Ru(py)2 (2.100 Å) [37], thus, reflecting the trans influence of CO on the lower trans effect axial ligands, Qnl, Im and Py respectively. The Ru–CO bond distance in 2A recorded as 1.8081(15) Å is shorter than that in (OEP)Ru(CO)(Im) (1.829(5) Å, but comparable to that of (OEP)Ru(CO)(Py) (1.812(6) Å). The C–O bond length in 2A is 1.155(2) Å and is similarly comparable to those of (OEP)Ru(CO)(Im) (1.156(5) Å) and (OEP)Ru(CO)(Py) (1.158(6) Å). The ∠Ru–C–O angle in 2A is essentially linear, as is the case for (OEP)Ru(CO)(Im) and (OEP)Ru(CO)(Py), but we note that the ∠Ru–C–O in 2A (176.51(14)o) is slightly lower than that of 2B (179.0(4)o).

The X-ray diffraction studies of a single crystal sample of 1 revealed the monohydrate complex, (OEP)Ru(CO)(H2O) (Fig. 5a). The average Ru–Np bond lengths in 1 is 2.0554(9) Å and is within range of those of other ruthenium(II) porphyrin carbonyl complexes. We note that in 1 because one water hydrogen (H2B) forms a hydrogen bond to a symmetry-related water oxygen, the occupancy of H2B was set to 0.5. The water hydrogen H2C was also assigned an occupancy of 0.5 to give the water the correct number of hydrogens.

Fig. 5
figure 5

a Molecular structure of 1 as a monohydrate complex. b Relative positions of adjacent porphyrin macrocycles in the crystal structure of 1. Hydrogen atoms have been omitted for clarity

Full size image

The Ru–O bond distance in 1 is 2.248(3) Å and it compares quite well with the 2.242 Å recorded in both (OEP)Ru(CO)(EtOH)[36] and (OEP)Ru(CO)(THF) [38], but quite longer than the aryloxide Ru – OC6H4-p-Me (1.964(11) Å) and Ru – O(Ru) (1.1789(11) Å) bond distances in [(TPP)Ru(OC6H4-p-CH3)]2O [39]. The ∠Ru–C–O of 1 is essentially linear with bond angle of 178.12(11) Å.

The porphyrin mean plane of separation between two adjacent porphyrins in 1 is 5.17 Å, and shows the molecules are substantially closer in 1 than in 2, which is expected due to the bulkier Qnl ligand in 2. A lateral shift of 3.65 Å between two adjacent porphyrins was determined for 1. We note that the Ru atom in 1 is displaced by 0.15 Å from the 24-atom mean porphyrin plane towards the CO ligand, but was displaced in the opposite direction in 2. Efforts to grow suitable single crystals for X-ray crystallography of 3 were unsuccessful, instead precipitates were obtained. The precipitates had similar spectral features as before suggesting 3 is stable in solution under N2 over several days.

Compounds 2 and 3 were also characterized by mass spectrometry. The MS-TOF data for 2 showed peaks at m/z 833.2400, 764.1798 and 741.1968 and assigned to [(OEP)Ru(CO)(Qnl) + MeCN + H]+, [(OEP)Ru(Qnl) + H]+ and [(OEP)Ru(CO) + DMSO + H]+, respectively. The MS data for 3 displayed peaks at m/z 890.2352 [(OEP)Ru(CO)(8-HOQ) + 2MeCN + H]+ and 741.1964 [(OEP)Ru(CO) + DMSO + H]+. The quinine complex, 3 displayed peaks at m/z 1333.5198 assigned to [M + QN + Na]+.

Cyclic voltammetry

The redox behavior of compounds 1 – 3 were investigated by cyclic voltammetry in an anhydrous dichloromethane solution containing NBu4PF6 as support electrolyte. The electrochemical cell was composed of Pt working electrode, Ag/AgCl reference electrode and a Pt wire auxiliary electrode.

We begin by examining the redox behaviors of compounds 2. The cyclic voltammogram of reveals that 2 undergoes a reversible one-electron first oxidation at Eo’ =  + 0.21 V vs. Fc/Fc+ couple (Fig. 6a). Compound 3 similarly undergoes a reversible first oxidation at + 0.19 V vs. Fc/Fc+ couple (Fig. 6b). Oxidation potentials in this range are associated with porphyrin centered oxidations [40]. Thus, the π-radical cations [2]•+ and [3]•+, respectively are generated after first oxidation of 2 and 3.

Fig. 6
figure 6

Cyclic voltammogram of 2 (a) and 3 (b) showing the two oxidations. Conditions: 1 mM analyte, 200 mV/s scan rate, 0.1 M NBu4PF6 support electrolyte, room temperature

Full size image

We note that the cathodic-to-anodic current peak ratios (ipc/ipa was determined to be ~ 1.0 and the plots of the ipa vs square root of the scan rate (υ1/2) for the first oxidations of compounds 2 and 3 show a linear relationship, thus, suggesting a diffusion-control redox process. The peak separations, ΔEp =|Epa – Epc| in peak potentials for the first oxidations each of compounds 2 (62 mV) and 3 (141 mV) at a scan rate of 200 mVs−1 are near equivalent to that of their respective Fc/Fc+ internal standard values of 62 mV and 148 mV, indicating that the first oxidations are single-electron reversible electron-transfer processes.

We note that the second oxidation of 2 is reversible and occurs at Eo’ =  + 0.74 V. Similarly, the ΔEp values of the second oxidation of 2 is identical to that of the Fc/Fc+ couple, also suggesting a reversible one-electron transfer process. The second oxidation of 3, however, is irreversible and occurs at an anodic peak potential, Epa =  + 0.66 V and the cathodic peak potential, Epc being + 0.74 V. In addition, there was a pre-wave after the first oxidation of 3 which occurred at Epa =  + 0.62 V. Table 4 summarizes the redox potentials of the compounds referenced with Fc/Fc+ couple as internal standard.

Table 4 Redox potentials of the compounds at a Pt working electrode.[a]
Full size table

In comparison, the cyclic voltammetry of precursor compound 1 revealed reversible first and second oxidations at Eo’ =  + 0.22 and + 0.73 V, respectively (Fig. 7). Additionally, the ΔEp values of both redox processes are comparable to that of the Fc/Fc+ couple, indicating a reversible one-electron transfer process. In summary, compounds 1, 2 and 3 undergo porphyrin-centered first and second oxidations, and that these redox potentials are well within the E1/2(1) =  ~ 0.81 vs SCE obtained for the related (OEP)Ru(CO)(py)[41] and (TPP)Ru(CO)(py) [40] compounds. (Note: 0.81 V converts to ~ 0.35 V vs Fc/Fc+ couple [42]). A second oxidation potential, E1/2(2) was recorded at 1.36 V vs SSCE (~ 0.90 V vs. Fc/Fc+ couple) for (TPP)Ru(CO)(py) [40]. Similar redox behaviors were observed previously in some six-coordinate (por)Ru complexes [17, 18, 43].

Fig. 7
figure 7

Cyclic voltammogram of 1 showing the two oxidations. Conditions: 1 mM analyte, 200 mV/s scan rate, 0.1 M NBu4PF6 support electrolyte, room temperature

Full size image

Conclusion

We have prepared two ruthenium(II) octaethylporphyrin complexes of the form (OEP)Ru(CO)(Q), (Q = Qnl (2), and QN (3) from (OEP)Ru(CO) and characterized them by IR, MS, UV–vis and 1H NMR spectroscopy. We have determined the X-ray crystal structure of 2 that shows CO coordination to Ru at the axial position. The sixth coordination site is occupied by quinoline through the Qnl-N. In the crystal packing of one batch of the single crystals of 2 (i.e. 2A) the axially bound Qnl ligands of adjacent complexes are close to each other at a distance of 3.55 Å. In another batch of single crystal sample, the CO groups on adjacent complexes are on the same side. We also determined the X-ray crystal structure of the precursor complex (OEP)Ru(CO) (1) that shows H2O coordination at the axial position. The X-ray crystal structures of 1 and 2, together with analysis of the spectroscopic data obtained suggest the site of binding of QN to Ru in the quinine complex 3 is likely through the quinoline N of QN. Electrochemical studies of the compounds via cyclic voltammograms revealed oxidations that are porphyrin-centered.