Article pubs.acs.org/IC
Copper Complexes with NH-Imidazolyl and NH-Pyrazolyl Units and Determination of Their Bond Dissociation Gibbs Energies Alexander Wilting,†,‡ Merle Kügler,†,‡ and Inke Siewert*,† †
Georg-August-University Göttingen, Tammannstr. 4, D-37077 Göttingen, Germany S Supporting Information *
ABSTRACT: We synthesized two dinuclear copper complexes, which have ionizable N imidazole and N pyrazole protons in the ligand, respectively, and determined the BDFE of the hypothetical H atom transfer reactions CuII(LH−1) + H• ↔ CuI(L) in MeOH/H2O (BDFE: bond dissociation Gibbs (free) energy). The ligands have two adjacent N,N′,O-binding pockets, which differ in one Nheterocycle: La has an imidazole unit and Lc, a pyrazole unit. The copper(II) complexes of La and Lc have been characterized, and the substitution pattern has only little influence on the structural properties. The BDFEs of the hypothetical PCET reactions have been determined by means of the species distribution and the redox potentials of the involved species in MeOH/H2O (80/20 by weight). The pyrazole copper complex 3 exhibits a lower BDFE than the isoelectronic imidazole copper complex 1 (1, 292(3) kJ mol−1; 3, 279(1) kJ mol−1). The difference is mainly caused by the higher acidity of the N pyrazole proton of 3 compared to the N imidazole proton of 1. The redox potentials of 1 and 3 are very similar.
■
2-yl)pyridine)ruthenium.4d The deprotonation of all four imidazolyl NH-protons led to a shift of the redox potential of the RuIII/II couple of about 1.28 V toward lower potentials. A similar shift of roughly 300 ± 70 mV per NH unit was observed in further MII complexes with NH imidazolyl units, namely, FeII, CoII, and RuII as metal ions,4 and this was attributed to higher Coulombic effects.4d Mayer and co-workers determined a bond dissociation Gibbs energy (BDFE) of 260 kJ mol−1 (62 kcal mol−1) in acetonitrile for the PCET reaction of RuII(acac)2(pyim) leading to RuIII(acac)2(pyimH−1).4g Consequently, it was easily oxidized by the TEMPO-radical (cf. BDFE of TEMPO: 278 kJ mol−1 (66.5 kcal mol−1) in MeCN;5 pyim = 2-(1H-imidazol-2yl)pyridine). A BDFE of 293 ± 8 kJ mol−1 (70 ± 2 kcal mol−1) has been determined for the PCET reaction (TPP)FeII(imMe) ↔ (TPP)FeIII(imMeH−1) + H• in acetonitrile.6 (TPP)FeIII(imMeH−1) undergoes PCET reactions with H atom transfer reagents such as ascorbate or TEMPOH (TPP = tetraphenylporphyrin, imMe = 4-methylimidazole). To the best of our knowledge, investigations with biologically relevant Cu complexes are not known. Therefore, we wanted to explore the thermodynamic coupling of the hypothetical PCET reaction as shown in Scheme 1 involving CuII/CuI ions and NH-imidazole units. The vertical arrows in Scheme 1 represent the acid/base equilibrium of the reduced (left) and oxidized (right) species, the horizontal arrows, the redox process of the protonated (top) and deprotonated species (bottom), and the diagonal arrow represents the PCET step and is described by the BDFE.
INTRODUCTION
Electron transfer (ET) reactions are ubiquitous in nature. Numerous examples are well-known to involve proton-coupled electron transfer processes (PCET), in particular in redox reactions, which are relevant for energy conversion reactions or oxygenation reactions.1 Iron and copper ions in the active center of enzymes often play a crucial role in such reactions, and thus, it is natural to investigate 3d metal complexes in order to study PCET reactions.2 Since imidazole units are very common N-donor sides for copper ions in metalloenzymes,2a,c we were interested in the thermodynamic coupling of CuII/I complexes bearing ionizable NH-imidazolyl and NH-pyrazolyl units (Figure 1).3 There was some experimental evidence that NH-imidazolyl units in metal complexes are strongly coupled to the redox state of the metal center.4 Williams et al. showed that the deprotonation of the ligand has a great influence on the redox potential of the RuIII/II couple in bis(2,6-di(1H-imidazol-
Received: September 11, 2015
Figure 1. Presentation of the copper complexes. © XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.5b02084 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
referenced by adding ferrocene to the electrolyte solution after each batch of measurements. The potential of the SCE was referenced to NHE by adding 0.244 V. We used different custom-made gastight cells for the measurements. Spectrophotometric Titrations. UV/vis spectra of HLa, HLc, 1, and 3 were collected in 2.5 mL samples with 1 cm optical length cuvettes. The solutions were prepared in MeOH/H2O mixtures (80/ 20 by weight) with a fixed ionic strength (I = 0.1 M KCl). The pH was adjusted by the addition of concentrated KOH and HCl solutions (pH 2−12). The pH was measured by a combined glass electrode (Metrohm 6.0234.100) filled with 0.1 M NaCl in MeOH/H2O (80/20 by weight). Details about the fitting procedure can be found in the Supporting Information. X-ray Crystallography. X-ray data were collected with a STOE IPDS II diffractometer (graphite monochromated Mo Kα radiation, λ = 0.71073 Å) by use of ω scans at −140 °C. The structures were solved by direct methods and refined on F2 using all reflections with SHELX-2013.13 Most non-hydrogen atoms were refined anisotropically. Most hydrogen atoms were placed in calculated positions and assigned to an isotropic displacement parameter of 1.2/1.5 Ueq(C). The nitrogen bound hydrogen atoms were refined by using DFIX restraints (d(N−H) = 0.87 Å). Face-indexed absorption corrections were performed numerically with the program X-RED.14 Crystals of 2 were found to be nonmerohedrally twinned (twin-law: 0.96 0.08 0, −0.17 1.03 −0.01, −0.02 0.03 1; twin ratio: 0.316(1):0.684(1)). A HKLF 5 format file was used for the refinement of the structure. dmf molecules have been disordered over two positions and an inversion center and were refined using SAME restraints and fixed occupancy factors of two times 0.5 and one time 0.25. Compound 3 contained 5.25 dmf molecules per formula unit. Some of the molecules were disordered over several positions including symmetry related positions and refined using SAME, SADI, and FLAT restraints and by applying fixed occupancy factors of 0.5 and 0.25. The tert-butyl group was disordered over two positions (occupancy factor = 0.64(4)/0.36(4)). CCDC-1445308 (2) and CCDC-1445307 (3) contain the supplementary crystallographic data for this paper (see also Table 1). These data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/ cif. General Procedure for the Complex Synthesis. A solution of copper bromide (2 equiv) in acetonitrile was added to a solution of the ligand (1 equiv) and triethylamine (ex.) in thf. The reaction mixture was stirred for 16 h at room temperature and a precipitate formed. The solid was collected by filtration, washed with thf and Et2O, and dried in vacuo. Recrystallization from dmf/ether yielded the analytically pure complexes. Spectroscopic Data of 1, [LaCu2Br3]. Brown powder, yield 72%. MALDI-MS (DCTB): m/z 718.9 ([M−Br]+). IR (KBr): ν (cm−1) 3437 (s), 3055 (m), 1647 (m), 1617 (s), 1560 (m), 1495 (s), 1473 (s), 1451 (m), 1375 (w), 1277 (w), 1255 (s), 1180 (w), 1164 (w), 1101 (m), 1012 (w), 985 (w), 894 (w), 862 (w), 820 (m), 767 (m), 745 (w), 724 (w), 715 (w), 696 (w), 668 (w), 644 (m), 562 (w). C26H23Br3Cu2N6O·0.5H2O calcd.: C, 38.5; H, 2.98; N, 10.4. Found: C, 38.5; H, 3.04; N 10.4. Spectroscopic Data of 2, [LbCu2Br3]. Brown powder, yield 85%. MALDI-MS (DCTB): m/z 746.9 ([M−Br]+). IR (KBr): ν (cm−1) 3436 (s), 2361 (w). 1636 (m), 1603 (s), 1569 (m), 1540 (m), 1498 (s), 1473 (m), 1416 (w), 1373 (w), 1282 (w), 1253 (m), 1196 (w), 1108 (w), 1034 (w), 960 (w), 817 (m), 742 (w), 718 (w), 687 (w), 645 (w), 520 (w). C28H27Br3Cu2N6O·2H2O calcd.: C, 38.8; H, 3.61; N, 9.7. Found: C, 39.1; H, 3.58; N, 9.7. Spectroscopic Data of of 3, [LcCu2Br3]. Brown powder, yield 82%. MALDI-MS (DCTB): m/z 718.9 ([M-Br]+). IR (KBr): ν (cm−1) 3420 (s), 2962 (m), 1607 (s), 1570 (s), 1560 (m), 1479 (s), 1465 (s), 1452 (s), 1394 (m), 1363 (m), 1273 (m), 1253 (s), 1156 (w), 1097 (s), 1076 (w), 1017 (w), 822 (m), 799 (m), 710 (w), 646 (m), 583 (w). C26H23Br3Cu2N6O·2dmf·Et2O calcd.: C, 42.3; H, 4.63; N, 11.0. Found: C, 42.2; H, 4.33; N, 11.0.
Scheme 1. Thermodynamic Square Scheme of the Metal− Ligand PCET Reaction of One of the Copper Complexesa
a
The copper ions are also coordinated by halogenido ligands, which have been omitted for clarity.
Investigations on metal complexes coordinated by pyrazole ligands are nearly unknown,7 and therefore we also aimed to quantify the difference in isoelectronic Cu-imidazole and Cupyrazole systems. We expected a higher acidity of the pyrazole N proton upon copper ion binding compared to the imidazole N proton, because the distance between the proton and the copper ion is shorter in the former case.8,9 The effect might be even more enhanced due to the higher dipole moment of pyrazole (μ = 3.86 D, cf. imidazole μ = 2.42 D).10 The redox potential of the Cu−pyrazolyl complex was expected to be slightly higher than the one of the Cu−imidazolyl complex, because imidazole is a stronger σ donor than pyrazole11 (πbonding interactions do not play a significant role in Cu2+− imidazole/pyrazole complexes).12
■
EXPERIMENTAL SECTION
General Information. Manipulations of air-sensitive reagents were carried out in an MBraun glovebox, or by means of Schlenk-type techniques involving the use of a dry argon or nitrogen atmosphere. Dry and degassed solvents used for cyclic voltammetry were purified by usual methods (dmf over CaH2, methanol over Mg) and degassed by three pump and freeze cycles prior to use. All other reagents were purchased as reagent grade or with higher quality and used without further purification. Microanalyses were performed on an Elementar Vario El II elemental analyzer. IR spectra were recorded using samples prepared as KBr pellets with a Bruker VERTEX 70 FTIR-spectrometer. Mass spectra were recorded on a Bruker APEX IV micrOTOF or a Bruker Autof lex Speed mass spectrometer. The matrix substance for MALDI-MS was trans-2-[3-(4-tert-butylphenyl)-2-methyl-2propenylidene]malononitrile (DCTB) in thf. UV/vis spectra were recorded with a Varian Cary50 Scan. The ligands HLa, HLb, and HLc were synthesized according to previous reported literature procedures.3 Electrochemistry. Electrochemical measurements were recorded with a Metrohm Autolab PGSTAT 101 or Gamry Instruments Reference 600 using dried and distilled dmf or dried and distilled methanol and bidistilled, degassed water. A common three-electrode setup consisted of a glassy carbon electrode as a working electrode (CH Instruments, A = 7 mm2), a platinum wire as a counter electrode, and a silver wire as pseudo reference (organic solvents) or an SCE (aq. solution) as reference electrode. The measurements in organic solvents were B
DOI: 10.1021/acs.inorgchem.5b02084 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry Table 1. Crystal Data and Refinement Details for 2 and 3 compound
2
3
empirical formula fw T [K] cryst size [mm3] cryst syst space group a [Å] b [Å] c [Å] α [deg] β [deg] γ [deg] V [Å3] Z ρ [g/cm3] F(000) μ [mm−1] Tmin/Tmax Θ range [deg] hkl range measured reflns unique reflns [Rint] obs. reflns (I > 2σ(I)) data/res/ params goodness-offit (F2) R1, wR2 (I > 2σ(I)) R1, wR2 (all data) resid el. dens [e/Å3]
C35.50H44.50Br3Cu2N8.50O3.50
C41.75H59.75Br3Cu2N11.25O6.25
1013.10 133(2) 0.300 × 0.240 × 0.080
1186.06 133(2) 0.456 × 0.045 × 0.039
triclinic P1̅ 9.7087(4) 13.8549(6) 17.1249(7) 82.158(3) 84.692(3) 71.183(3) 2157.13(16) 2 1.560 1016 3.808 0.4768/0.6981 1.563−25.976
monoclinic P21/c 18.573(4) 17.051(3) 16.993(3) 90 102.78(3) 90 5248.1(19) 4 1.501 2408 3.148 0.4846/0.7341 1.640−26.947
±11, ±16, ±21 47244
±23, ± 21, ± 21 56187
47244
11015 [0.1959]
32213
5864
47244/80/492
11015/199/602
0.983
1.109
0.0875, 0.2227
0.1120, 0.2036
0.1177, 0.2441
0.2004, 0.2432
−1.803/1.372
−1.149/1.165
Scheme 2. Synthesis of 1, 2, and 3
■
RESULTS AND DISCUSSION Complex Synthesis and Characterization. We recently synthesized ligands with two N,N′-donor sides, which are linked by a phenol unit.3 The ligands differ by the isoelectronic heterocycles pyrazole and imidazole. The coordination behavior of the ligands is similar.3 The reaction of the ligand precursors HLa, HLb, and HLc with copper bromide in the presence of triethylamine led to the formation of dinuclear metal complexes with the same structural motif (Scheme 2). The complexes were isolated and characterized by the usual methods such as mass spectrometry and elemental analysis. Complexes 2 and 3 have been characterized by single crystal Xray diffraction experiments, too. The quality of the crystal of 3 was very low, and therefore the solution does not allow for an extensive discussion on bond lengths. The results are depicted in Figure 2, and selected bond lengths are shown in Table 2. The metal ions in 2 and 3 are coordinated by two bromide ions, the phenol O atom, the pyridine N atom, and the imidazole and pyrazole N atom, respectively. The metal ions in 2 show a slightly distorted square pyramidal coordination geometry (τ = 0.01 (Cu1), τ = 0.32 (Cu2)15). The metal−metal distance of 3.196 Å is very similar to the one observed in a structurally related copper complex having a similar N,N′,O-ligand
Figure 2. Molecular structures of 2·2.5dmf (top) and 3·5.25dmf (bottom). Solvent molecules and hydrogen atoms were omitted for clarity. Thermal ellipsoids are set at the 50% level.
geometry.16 The distances between the Cu atoms and the O1 atom are within the range that is typically observed for phenoxide ligands in equatorial positions of CuII complexes having a SPY-5 geometry.17 The distances between the axial Br C
DOI: 10.1021/acs.inorgchem.5b02084 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry Table 2. Selected Bond Lengths [Å] for 2 and 3 with Estimated Standard Deviations in Parentheses atoms
2·2.5dmf
3·5.25dmf
Cu1−Cu2 Cu1−O1 Cu1−N1 Cu1−N2 Cu1−Br1 Cu1−Br2/3 Cu2−O1 Cu2−N4 Cu2−N5 Cu2−Br2 Cu2−Br3
3.2456(24) 1.921(7) 2.027(7) 1.938(8) 2.3943(15) 2.8072(16) 1.941(6) 2.027(8) 1.966(7) 2.5751(14) 2.4537(15)
3.1960(16) 1.936(8) 2.016(8) 1.960(9) 2.465(2) 2.6043(18) 1.933(7) 2.019(9) 1.972(9) 2.4906(18) 2.504(2)
Table 3. Spectral Results and Assignments of the UV/Vis Bands in MeOH/H2O (80/20 by weight, I = 0.1 M KCl) ILCT (weak LMCT bands) 1 λmax /nm ε/M−1 cm−1 3 λmax /nm ε/M−1 cm−1
LMCT
257
288
335
371
∼425
36 × 103
18 × 103
18 × 103
11 × 103
∼1.4 × 103
243
294
327
361
∼425
58 × 103
28 × 103
21 × 103
16 × 103
∼3.6 × 103
and likely the shifts arise from the lower flexibility around the Cpy−Cim/Cpz bond upon complexation.22 λmax of this two bands is red-shifted by 5−10 nm in 1 compared to the bands in 3, and this might be a consequence of the slight energy difference of the three highest occupied MOs in pyrazole and imidazole.23 Very weak LMCT bands, which might arise from π → Cu(II) transitions of the heterocycles, were expected to be in the same area and hence were masked by the more intense ILCT bands.20,22,24 The absorption coefficients of the ILCT bands at 288 and 371 nm (1) and at 294 and 361 nm (3) are slightly higher than the coefficients in the respective ligands (Table 3, Table S1). Proton Coupled Electron Transfer Reaction. A concerted PCET reaction could be favored, when a large shift of the redox potential is observed upon deprotonation.25 Thus, we examined the redox potential of 1 and 3 with respect to the protonation state of the ligand. Compound 3 showed a large separation of the reduction and the corresponding oxidation wave of 0.62 V vs the Fc/Fc+ potential26 in dmf (Figure 3). The reduction of one copper center occurred at a potential of −0.65 V and the corresponding oxidation at −0.03 V vs the Fc/Fc+ potential (scan rate 100 mVs−1). The separation of the redox waves of 1 was even larger: The reduction wave occurred at a potential of −0.91 V and the corresponding oxidation wave at 0.07 V vs Fc/Fc+ in dmf (scan rate 100 mVs−1, see Table 4).
atoms and the Cu atoms are slightly longer than the distances of the Cu atoms to the equatorial bromide ions since the antibonding a1 (dz2) orbital along the apical axis is doubly occupied, whereas the b1 orbital (dx2−y2) is only singly occupied (cf. d(Cu1−Br1) = 2.3943(15), d(Cu1−Br3) = 2.8072(16), d(Cu2−Br3) = 2.4537(15), d(Cu2−Br2) = 2.5751(14)).18 The ligand is not planar. The imidazolyl−pyridinyl units are twisted against each other with an angle of 46° (angle between the planes defined by the imidazolyl-pyridinyl units). π-Stacking interactions of the imidazolyl-pyridinyl units are present in the crystal lattice, which led to a chain-like arrangement of the molecules along the ac plane (Figure S1). The distances of ∼3.6 Å between the centroids of the heterocycles match the distance, which is typically observed for such interactions.19 A solid state structure of 1 could not be obtained, but we know from previous investigations with a similar ligand system that the methylation of the N-imidazolyl atom has virtually no influence on the structural parameters.4h The Cu ions in 3 exhibit a distorted square pyramidal coordination sphere (τ = 0.47 (Cu1) and τ = 0.36 (Cu2)), and the ligand is also twisted. The angle between the two planes defined by the pyridine rings of the side arm is 53°, which is very similar to the related zinc complex.3 Within the crystal lattice, the ligand molecules show a one-dimensional chain arrangement along the c-axis due to πstacking interactions (Figure S2). The average Cu−N distance of 2 is the same as in 3. In both cases, the bond distances between the pyridine N atoms and the Cu atoms are slightly longer than distances between the Cu atoms and the N atoms of the imidazolyl and pyrazolyl units, respectively. Solution UV/vis spectra of 1 and 3 were similar, which displays the similar structure of the complexes (Figure S3). A solution of 3 in dmf showed a weak, broad absorption band around 700 nm, which was assigned to d−d transitions (Figure S3). The low symmetry around the metal center causes the broadening of the band. The d−d band of 1 is blue-shifted with respect to 3, and it partially overlaps with a LMCT band, which indicates that the pyrazolyl ligand induces a slightly weaker ligand field (Figure S3).20 Around 425 nm, both complexes showed a weak absorption band in MeOH/H2O (80/20 by weight, I = 0.1 M KCl), which may be attributed to a phenolate → CuII LMCT.21 The UV spectra of 1 and 3 in MeOH/H2O (80/20 by weight, I = 0.1 M KCl) showed four intense ILCT bands between 240 and 380 nm as deduced from the corresponding ligand spectra (Figure S4, Table 3, Table S1). The bands around 250 nm have a similar λmax in 1 and 3. The two ILCT-bands lowest in energy are red-shifted upon metal binding (λmax = 335, 371 nm (1) and λmax = 327, 361 nm (3)),
Figure 3. CV data of 1 (top) and 3 (bottom) before (orange) and after (blue) the addition of two equivalents of potassium tert-butoxide. Measurements were carried out in a 0.1 M solution of (nBu4N)PF6 in dmf, scan rate = 100 mVs−1. D
DOI: 10.1021/acs.inorgchem.5b02084 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
HLc are very similar to the first and third pKa’s of HLa, these might be ascribed to the pyridinyl units. The distinct protonation side for each step cannot be unambiguously identified in HLa, because the pKa ranges of pyridinyl and imidazolyl units overlap. Deprotonation of the pyridinyl heterocycles can occur in a wide pH range; e.g., pyridin-2ylmethanamine exhibits a pKa of 2.46, 31 whereas the deprotonation of the pyridine moiety in pyridine-2,6-bis(monothiocarboxylic acid) occurs at a pKa of 5.48.32 Structural analogues containing imidazolyl heterocycles exhibit pKa values in this range, too (bis(imidazol-2-yl)methane pKa = 4.74 and pKa = 6.93 in water;33 N-phenylalanylbis(imidazol-2-yl)methylamine pKa = 3.09 and pKa = 5.28 in water34). Though, all dissociation constants most likely belong to N-pyridinyl and N-imidazolyl sides, and the dissociation constants of the phenolic protons were too high to be determined in the experimental range of pH 2−12. The pKa of pyridines decreases in MeOH/H2O mixtures with respect to water, whereas the pKa of phenols increases.35 4-tert-Butylphenol exhibits a pKa of 10.2 in water,36 and it shifts to ∼11.5 in MeOH/H2O (80/20 by weight) as determined by UV/vis titrations (Figure S9). The deprotonation leads to a characteristic red-shift of the band at λmax = 275 nm in the UV/vis spectra, and such a characteristic shift was not observed in HLa or HLc. The protonation of pyridine, imidazole, and pyrazole does not lead to characteristic shifts of the respective UV/vis bands (Figure S10). In order to investigate the complex formation in dependency of a solution’s pH, potentiometric titrations of HLa and HLc were performed with 2 equiv of CuII ions (MeOH/H2O 80/20 by weight). Evaluation of the titration curves indicates that both ligands have a very high binding affinity to CuII ions (Figures S11, S12). Three deprotonation steps are observed upon raising the pH (cf. HLa and Cu2+: pKa = 6.57, pKa = 8.01, and pKa = 9.13; HLc and Cu2+: pKa = 3.31, pKa = 6.68, and pKa = 10.74), which means that likely water or methanol ligands are bound to the metal centers in solution, which also are deprotonated. Indeed, the formation of aqua and methanol species was observed by ESI-MS studies of the solutions obtained by potentiometric studies at pH ∼ 3 (Figure S13). This raised the questions of whether the aqua species are present only in potentiometric titrations (the complexes are formed in situ in this method) or whether the halogenido ligands are always substituted by aqua or methanol ligands? Therefore, we reinvestigated the dissociation processes by UV/vis experiments. Analytically pure 1 and 3 were dissolved in MeOH/H2O (80/20 by weight) solutions at pH ∼ 2, and aliquots of KOH solution were added (I = 0.1 M KCl). From the trend of the absorbance vs pH, it becomes obvious that no further changes in the UV/vis spectra occurred beyond a pH of 8 (1) and a pH of 9 (3), respectively (Figures S14, S15). The UV/vis data have been fitted with SpectFit,37 and the results are depicted in Figure 4, Table 5, Figures S14−S16, and Tables S3 and S4. Compounds 1 and 3 exhibited only two deprotonation steps in the experimental range of pH 2−12, and the third deprotonation step at rather high pH, which was determined by potentiometric studies, was not observed by UV/vis titration. The UV/vis data of 1 showed a broad band around 380 nm upon raising the solution’s pH, and this is characteristic of a deprotonation of imidazole yielding imidazolato units in the presence of CuII ions (Figures S14, S15).24a Although the absence of the third pKa already indicated that the two pKa’s determined by UV/vis titrations may be associated with the
Table 4. Potential of the Reduction and Oxidation Wave of 1 and 3 vs. Fc/Fc+ in dmf, Scan Rate 100 mVs−1 E′red/V E′ox/V
1
1H−2
3
3H−2
−0.91 0.07
−1.23 −0.18
−0.65 −0.03
−0.96 −0.05
The potential of the reduction process differs by about 250 mV in the two complexes, whereas the potential of the preceding oxidation process was very little affected by the coordination environment. Since π-bonding interactions do not play a significant role,12 the lower reduction potential in dmf can be explained by the stronger σ-donor abilities of imidazole with respect to pyrazole.11 Scanning to higher potential revealed two further irreversible oxidation processes in 1 and 3 (Figure S27), which correspond to ligand oxidation as deduced from the CV data of the corresponding zinc complexes.3 The first wave may arise from oxidation of the phenol to the phenoxide radical (∼255 mV (1) and ∼330 mV (3) vs Fc/Fc+).27 Subsequently, we added two equivalents of potassium tertbutoxide in order to investigate the deprotonated species. The CuII → CuI reduction wave of 1 shifted by about −320 mV upon deprotonation and the wave of the corresponding oxidation process by about −250 mV. So, as expected, the deprotonation led to a shift as previously observed for related Fe, Ru, and Co complexes.4 The 2-fold deprotonation of 3 shifted the CuII → CuI reduction wave about −310 mV, and the corresponding oxidation wave occurred at a potential of −50 mV vs Fc/Fc+. The shift of the redox potential upon deprotonation of 3 may indicate a thermodynamic coupling also in the case of the pyrazolyl complex, which we further examined by determination of the BDFE. Scheme 1 shows the thermodynamic square schemes of 1. The bond dissociation Gibbs energy can be determined according to eq 1, where R is the gas constant, T is temperature, and F is the Faraday constant.28a The first two terms describe the redox potential of the complex and its acid/ base equilibrium, cG is equivalent to the H+/H• standard reduction potential in the respective solvent s (SI and ref 28). BDFE = 2.301RT pK a + FE°(Ms−) + cG
(1)
Since we isolated the dicopper(II) complexes, we focused on the upper right part of the thermodynamic square scheme in order to determine the BDFE (Scheme 1), that is the reduction potential of the fully protonated CuIICuII species and the acid dissociation constant of the CuIICuII species. It is noteworthy to say that Ka and E° of the complexes must be determined under identical conditions. We always used MeOH/H2O in a ratio of 80/20 by weight at a fixed ion strength of 0.1 M KCl for all measurements, because all species of 1 and 3 were soluble in this mixture. Determination of the Dissociation Constants. We performed speciation studies in order to determine the Ka of the ligands and complexes. The ligand HLa exhibits three deprotonation steps in the experimental range of pH = 2−12 (cf. pKa = 2.09, pKa = 3.80, and pKa = 5.14; Table S2, Figure S5). HLc exhibits only two protonation events in the measured pH range (cf. pKa = 2.18 and pKa = 4.99, Table S2, Figure S7). These are most likely ascribed to the deprotonation of the pyridinyl units, because pyrazole has a rather low pKa,29 and likely the protonation of the pyrazolyl moieties appears below pH 2.30 The pKa’s of HLa and HLc were confirmed by UV/vis titration experiments (Figures S6, S8). Since the two pKa’s of E
DOI: 10.1021/acs.inorgchem.5b02084 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
intensity, have been observed in the case of 1, too (Figure S14). In the case of [Cu(pypz)3]2+, the changes are less prominent. The two main bands at 243 and 288 nm (shoulder at 300 nm) split up to three bands at ∼245, 275, and 305 nm upon raising the solution’s pH. The intensity does not change significantly (∼20%), and the red shift was little. A similar behavior was observed in the titration of 3, that is, minor red shifts of the bands and small changes in the intensity. On the other hand, the deprotonation of aqua ligands in 2,2′-bipyridine-diaquacopper acetate leads to a significant decrease in the absorption, and no characteristic shifts of the ILCT bands have been observed in water. In MeOH/H2O, a strong blue shift of the characteristic ILCT bands has been observed and, again, a decrease of the extinction (Figure S26). Both trends are opposite of the trends observed in the titration experiments of 1 and 3. From these combined experiments, we assume that the dissociation constants determined by UV/vis experiments likely belong to the deprotonation of the heterocycles and not to bound water or methanol ligands. This is underpinned by ESIMS studies: Acidic HCl/KCl solutions of 1 and 3 exclusively show the presence of chlorido species and no methanol or water species (Figure S17). Determination of the Redox Potentials. The redox potential of the “pure” electron transfer is the basis for the BDFE calculations. In other words, the redox potential must be determined from the pH independent region of the protonated or deprotonated species. The oxidized, fully protonated CuII species are present at rather acidic pH in MeOH/H2O, and cyclic voltammetry revealed a constant standard potential E1 for the CuII/I redox couple of 3 in the range of pH = 2.69−3.30 and of 1 in the range of pH = 2.00−3.00 (Figure S28). Coulometry in MeOH/H2O (80/20 by weight) showed that the redox process corresponds to a 1e− reduction in both cases. Simulation of the scan rate dependent CV data led to very similar standard potentials for the CuII/I couples of 340 mV for 1 and 341 mV for 3 vs NHE (Figures 5, S30, and S31; simulation parameter, Table S5).39
Figure 4. Species distributions of 1 (top) and 3 (bottom) in dependence of the pH. Solvent: MeOH/H2O (80/20 by weight); I = 0.1 M KCl.
Table 5. Complex Stability Constants log β of 1 and 3 and Proton Dissociation Constants pKa of the Complexes at I = 0.1 M KCla log β 1 log β 1H−1 log β 1H−2
14.6(4) 8.9(4) 0.8(4)
log β 3 log β 3H−1 log β 3H−2
13.0(1) 9.6(1)
pKa 1 pKa 1H−1
5.7 8.1
pKa 3 pKa 3H−1
3.4 6.0
3.6(1)
a
Solvent: MeOH/H2O (80/20 by weight). Standard deviations of calculated values are given in parentheses.
deprotonation of the NH units, we aimed to substantiate this assumption. In order to distinguish between NH deprotonation equilibria and possibly formed aqua species, we determined the characteristic UV/vis bands of NH/Cu- and N-ligated Cu/OH2 species and their shifts upon deprotonation. The UV/vis spectra of Cu(BF)4 and HLa in the absence of halogenido ions differ from those of 1 dissolved in MeOH/H2O at pH = 3.3 and pH = 7 (I = 0.1 M KCl). The aqua species had a small but distinct wave at 297 nm at pH = 7, which is not present in the presence of halogenido ions, and a slightly less prominent feature at 354 nm (Figure S20). At pH = 3.3, the aqua species has a slightly more prominent feature at 292 nm and a less prominent feature at 256 nm. The UV/vis spectra of 3 (I = 0.1 M KCl) and Cu2+/HLc differ significantly at pH = 2.4, but at pH = 4.7 no significant differences of the aqua and proposed halogendio species can be observed (Figure S21). In order to get a spectroscopic signature for NH deprotonation, we investigated 2-(1H-imidazol-2-yl)pyridine, pyim, and 2-(1Hpyrazol-3-yl)pyridine, pypz, respectively, in the presence of Cu2+ ions by pH dependent UV/vis titrations. According to the literature, 2-(1H-imidazol-2-yl)pyridine forms 3:1 complexes with copper(II) ions (Figures S22, S23).38 Upon raising the solution’s pH, the three ILCT bands of [Cu(pyim)3]2+ lowest in energy shifted toward higher wavelengths, and the bands at 295 and 335 nm got more prominent (Figure S24). The same trends, that is, the red shift of the bands and the increasing
Figure 5. CV data of 1 (top, pH = 2.7) and 3 (bottom, pH = 2.5) and simulated data, 0.1 M solution of KCl in MeOH/H2O (80/20 by weight); scan rate = 50 mVs−1. Simulation parameters and further scan rates are provided in the SI (Table S5, Figures S30, S31). F
DOI: 10.1021/acs.inorgchem.5b02084 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry In order to determine the BDFE of the hypothetical PCET reaction according to eq 1, we derived the constant cG according to the work of Tilset and Parker, and from tabulated data, because it has not been tabulated in MeOH/H2O (cG = 226.3 kJ mol−1 in MeOH/H2O (80/20 by weight); SI).28a,e,40 The BDFE1 of 1 was calculated to be 292(3) kJ mol−1 (70 kcal mol−1) and the one of 3 to be 279(1) kJ mol−1 (67 kcal mol−1) in MeOH/H2O (80/20 by weight; see Table 6).
N proton upon metal binding. The shorter distance of the Cu ion and the NH unit in 3 leads to a greater shift of pKa, and the slightly different σ-donor abilities of the heterocycles have virtually no influence on the BDFE in protic solvents. Future work will focus on the reactivity of such complexes with HAT reagents.42
Table 6. Thermodynamic Data for the PCET Reaction Shown in Scheme 1a
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.5b02084. π-Stacking interactions of 2 and 3; UV/vis data of the ligands, 1, and 3; data of the spectrophotometric titrations; species distribution in MeOH/H2O; ESI-MS data; CV data; detailed derivation of the constant cG; estimation of BDFEs in MeOH/H2O from tabulated values in MeCN (PDF) Crystallographic information (CIF)
1 3 1 relative to 3
E1 vs NHE/V
pKaox
BDFE1/kJ mol−1
0.340(4) 0.341(4) −0.001
5.7(6) 3.4(1) 2.3
292(3) 279(1) 13
■
ASSOCIATED CONTENT
S Supporting Information *
a 1 E and pKaox have been determined in 0.1 M solution of KCl in MeOH/H2O (80/20 by weight).
The bond dissociation energy of the imidazole-Cu entity is higher than the one of the pyrazole-Cu entity, and both have considerably weak bonds.5 The difference in the BDFE is caused by the lower pKa of the N proton in the latter case, whereas the redox potentials of 1 and 3 are virtually the same. The enhanced acidity of the N pyrazole proton upon Cu2+-ion binding is caused by the shorter distance between the copper ion and the N proton in the pyrazole-Cu entity.8,12 The BDFE of 3 is similar to the estimated BDFE of TEMPOH in MeOH/ H2O (cf. BDFETEMPOH ≈ 280 kJ mol−1 in MeOH/H2O, SI). The BDFE of 1 is ∼30 kJ mol−1 larger than the one obtained by Mayer and co-workers for RuII(acac)2(pyim) in MeCN.5 (TPP)FeII(imMe) and 1 exhibit a very similar BDFE (even if we consider a small correction of −2 kJ mol−1 for the BDFE of 1 due to the different solvents; SI), although the coordination environment and the d-electron configuration of the metal ions are very different.6 Metalloproteins with the respective metal ions in a similar coordination sphere have similar functions in nature, for example, hemoglobin and hemocyanin, which both act as oxygen carriers, or tyrosinase and peroxidase (heme-Fe), which both can carry out the hydroxylation of phenols and oxidation of catechols.41 Although PCET reactions involving the imidazole units are not discussed in a respective reaction pathway of the proteins, the similar BDFE may reflect the similar reactivity of the units.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Author Contributions ‡
These authors contributed equally.
Funding
I.S. is grateful to the DFG (Emmy Noether program, SI 1577/ 2-1) and the Fonds der Chemischen Industrie for funding. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We gratefully acknowledge the help of Dr. Sebastian Dechert with the X-ray crystallography. I.S. thanks Prof. Dr. Franc Meyer for helpful discussions and his support.
■
REFERENCES
(1) Migliore, A.; Polizzi, N. F.; Therien, M. J.; Beratan, D. N. Chem. Rev. 2014, 114, 3381−465. (2) (a) Sundberg, R. J.; Martin, R. B. Chem. Rev. 1974, 74, 471−517. (b) Bertini, I.; Gray, H. B.; Stiefel, E. I.; Selverstone, J. Biological Inorganic Chemistry: Structure & Reactivity; University Science Books: Sausalito, CA, 2007. (c) Stephanos, J. J.; Addison, A. W. Chemistry of Metalloproteins: Problems and Solutions in Bioinorganic Chemistry; John Wiley & Sons: Hoboken, NJ, 2014. (3) Wilting, A.; Kügler, M.; Siewert, I. Z. Anorg. Allg. Chem. 2015, 641, 2498−2505. (4) (a) Carina, R. F.; Verzegnassi, L.; Bernardinelli, G.; Williams, A. F. Chem. Commun. 1998, 2681−2682. (b) Brewer, C.; Brewer, G.; Luckett, C.; Marbury, G. S.; Viragh, C.; Beatty, A. M.; Scheidt, W. R. Inorg. Chem. 2004, 43, 2402−2415. (c) Lambert, F.; Policar, C.; Durot, S.; Cesario, M.; Yuwei, L.; Korri-Youssoufi, H.; Keita, B.; Nadjo, L. Inorg. Chem. 2004, 43, 4178−4188. (d) Stupka, G.; Gremaud, L.; Bernardinelli, G.; Williams, A. F. Dalton Trans. 2004, 407−412. (e) Stupka, G.; Gremaud, L.; Williams, A. F. Helv. Chim. Acta 2005, 88, 487−495. (f) Yamaguchi, T.; Harada, K.; Sunatsuki, Y.; Kojima, M.; Nakajima, K.; Matsumoto, N. Eur. J. Inorg. Chem. 2006, 2006, 3236−3243. (g) Wu, A.; Masland, J.; Swartz, R. D.; Kaminsky, W.; Mayer, J. M. Inorg. Chem. 2007, 46, 11190−11201. (h) Kügler, M.; Gałęzowska, J.; Schendzielorz, F.; Dechert, S.; Demeshko, S.; Siewert, I. Eur. J. Inorg. Chem. 2015, 2015, 2695−2706. (5) Warren, J. J.; Tronic, T. A.; Mayer, J. M. Chem. Rev. 2010, 110, 6961−7001.
■
CONCLUSION We synthesized two dinuclear copper(II) complexes having a N,N′,O-coordination sphere and ionizable N imidazole or N pyrazole protons and investigated the thermodynamic coupling of the metal centered reduction event and the protonation of the heterocycles. The complexes are structurally very similar, and the change of one N-donor unit from imidazole to pyrazole does not lead to substantial changes. The potential of the CuII/ CuI redox couple of 1 is lower than the one of 3 in dmf, and in both cases, the potential shifts upon deprotonation of the heterocyclic units. Interestingly, the redox potentials of the CuII/CuI couple of 1 and 3 are nearly the same in MeOH/H2O (80/20 by weight), which shows that the different σ-donor abilities of the heterocycles do not play a crucial role in protic solvents. The analysis of the hypothetical PCET reaction CuII(LH−1) + H• ↔ CuI(L) revealed a 13 kJ mol−1 higher BDFE for 1 compared to 3. The difference in 1 and 3 is attributed to the more prominent increase of the acidity of the G
DOI: 10.1021/acs.inorgchem.5b02084 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry (6) Warren, J. J.; Mayer, J. M. J. Am. Chem. Soc. 2008, 130, 2774− 2776. (7) (a) Baitalik, S.; Flörke, U.; Nag, K. J. Chem. Soc., Dalton Trans. 1999, 719−727. (b) Baitalik, S.; Bag, P.; Nag, K. Polyhedron 2002, 21, 2481−2488. (8) Muller, P. Pure Appl. Chem. 1994, 66, 1077−1184. (9) Johnson, C. R.; Henderson, W. W.; Shepherd, R. E. Inorg. Chem. 1984, 23, 2754−2763. (10) Shagun, V. A.; Tarasova, O. A.; Trofimov, B. A. Russ. J. Org. Chem. 2006, 42, 1039−1044. (11) Astley, T.; Canty, A. J.; Hitchman, M. A.; Rowbottom, G. L.; Skelton, B. W.; White, A. H. J. Chem. Soc., Dalton Trans. 1991, 1981− 1990. (12) (a) Canty, A. J.; Lee, C. V. Inorg. Chim. Acta 1981, 54, L205− L206. (b) Canty, A. J.; Lee, C. V. Organometallics 1982, 1, 1063−1066. (13) Sheldrick, G. M. Acta Crystallogr., Sect. A: Found. Crystallogr. 2008, 64, 112−122. (14) X-Red; STOE & Cie GmbH: Darmstadt, Germany, 2002. (15) Addison, A. W.; Rao, T. N.; Reedijk, J.; van Rijn, J.; Verschoor, G. C. J. Chem. Soc., Dalton Trans. 1984, 1349−1356. (16) (a) Eduok, E. E.; O’Connor, C. J. Inorg. Chim. Acta 1984, 88, 229. (b) Popov, L. D.; Levchenkov, S. I.; Shcherbakov, I. N.; Aleksandrov, G. G.; Tsaturyan, A. A.; Beloborodov, S. S.; Maevskii, O. V.; Kogan, V. A. Russ. J. Gen. Chem. 2013, 83, 2314−2319. (17) (a) Rao, S. P. S.; Manohar, H.; Aoki, K.; Yamazaki, H.; Bau, R. J. Chem. Soc., Chem. Commun. 1986, 4−6. (b) Reddy, P. A. N.; Nethaji, M.; Chakravarty, A. R. Inorg. Chim. Acta 2002, 337, 450−458. (c) Philibert, A.; Thomas, F.; Philouze, C.; Hamman, S.; Saint-Aman, E.; Pierre, J. L. Chem. - Eur. J. 2003, 9, 3803−3812. (d) Gebbink, R. J. M. K.; Watanabe, M.; Pratt, R. C.; Stack, T. D. P. Chem. Commun. 2003, 630−631. (e) Choudhary, M.; Patel, R. N.; Rawat, S. P. J. Mol. Struct. 2014, 1060, 197−207. (18) Rossi, A. R.; Hoffmann, R. Inorg. Chem. 1975, 14, 365−374. (19) (a) Hunter, C. A. Chem. Soc. Rev. 1994, 23, 101−109. (b) Janiak, C. J. Chem. Soc., Dalton Trans. 2000, 3885−3896. (20) Bernarducci, E.; Schwindinger, W. F.; Hughey, J. L.; KroghJespersen, K.; Schugar, H. J. J. Am. Chem. Soc. 1981, 103, 1686−1691. (21) Arora, H.; Philouze, C.; Jarjayes, O.; Thomas, F. Dalton Trans. 2010, 39, 10088−10098. (22) Nakamoto, K. J. Phys. Chem. 1960, 64, 1420−1425. (23) The HOMO and HOMO−1 show π-symmetry in imidazole and pyrazole, but the relative energy of the π1−MO is higher, the energy of the π2−MO is lower, and the energy of the lone pair is slightly higher in imidazole than in pyrazole.20 (24) (a) Fawcett, T. G.; Bernarducci, E. E.; Krogh-Jespersen, K.; Schugar, H. J. J. Am. Chem. Soc. 1980, 102, 2598−2604. (b) Godlewska, S.; Jezierska, J.; Baranowska, K.; Augustin, E.; Dołȩga, A. Polyhedron 2013, 65, 288−297. (25) Jencks, W. P. J. Am. Chem. Soc. 1972, 94, 4731−4732. Sinha, A.; Bruice, T. C. J. Am. Chem. Soc. 1984, 106, 7291−7292. (26) The potential of the Fc/Fc+ couple is 0.40 V vs. SCE in acetonitrile Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877− 910. (27) (a) Rotthaus, O.; Jarjayes, O.; Thomas, F.; Philouze, C.; SaintAman, E.; Pierre, J. L. Dalton Trans. 2007, 889−895. (b) Zats, G. M.; Arora, H.; Lavi, R.; Yufit, D.; Benisvy, L. Dalton Trans. 2011, 40, 10889−10896. (28) (a) Tilset, M.; Parker, V. D. J. Am. Chem. Soc. 1989, 111, 6711− 6717. (b) Bordwell, F. G.; Cheng, J. P.; Ji, G. Z.; Satish, A. V.; Zhang, X. J. Am. Chem. Soc. 1991, 113, 9790−9795. (c) Parker, V. D.; Handoo, K. L.; Roness, F.; Tilset, M. J. Am. Chem. Soc. 1991, 113, 7493−7498. (d) Wayner, D. D. M.; Parker, V. D. Acc. Chem. Res. 1993, 26, 287−294. (e) Liu, W. Z.; Bordwell, F. G. J. Org. Chem. 1996, 61, 4778−4783. (29) Acheson, R. M. An Introduction to the Chemistry of Heterocyclic Compounds, 3rd ed.; John Wiley & Sons: New York, 1976. (30) (a) Behle, L.; Neuburger, M.; Zehnder, M.; Kaden, T. A. Helv. Chim. Acta 1995, 78, 693−702. (b) Prokofieva, A.; Prikhod’ko, A. I.; Enyedy, E. A.; Farkas, E.; Maringgele, W.; Demeshko, S.; Dechert, S.;
Meyer, F. Inorg. Chem. 2007, 46, 4298−4307. (c) Müller, H.; BauerSiebenlist, B.; Csapo, E.; Dechert, S.; Farkas, E.; Meyer, F. Inorg. Chem. 2008, 47, 5278−5292. (31) Couturier, Y.; Petitfaux, C. Bull. Soc. Chim. Fr. 1974, 855−862. (32) Stolworthy, J. C.; Paszczynski, A.; Korus, R.; Crawford, R. L. Biodegradation 2001, 12, 411−418. (33) Várnagy, K.; Sóvágó, I.; Á goston, K.; Likó, Z.; Süli-Vargha, H.; Sanna, D.; Micera, G. J. Chem. Soc., Dalton Trans. 1994, 2939−2945. (34) Ö sz, K.; Várnagy, K.; Süli-Vargha, H.; Sanna, D.; Micera, G.; Sóvágó, I. Inorg. Chim. Acta 2002, 339, 373−382. (35) Rosés, M.; Rived, F.; Bosch, E. J. Chromatogr. A 2000, 867, 45− 46. (36) Lee, Y.; Yoon, J.; Von Gunten, U. Environ. Sci. Technol. 2005, 39, 8978−8984. (37) Binstead, R. A.; Zuberbühler, A. D.; Jung, B. SpectFit, version 3.0.38, Windows; Spectrum Software Associates: Chapel Hill, NC, 2006. (38) Boggess, R. K.; Martin, R. B. Inorg. Chem. 1974, 13, 1525−1527. We used a ligand to Cu2+ ion ratio of 4:1 in order to suppress the formation of aqua species. (39) DigiElch Professional, version 7; ElchSoft. (40) The solvation Gibbs energy of H2 has not been determined in MeOH/H2O mixtures, and thus we estimated the solubility based on the values for MeOH, EtOH, and EtOH/H2O mixtures. (41) (a) Sadler, A.; Subrahmanyam, V. V.; Ross, D. Toxicol. Appl. Pharmacol. 1988, 93, 62−71. (b) Itoh, S.; Fukuzumi, S. Acc. Chem. Res. 2007, 40, 592−600. (c) Connor, K. L.; Colabroy, K. L.; Gerratana, B. Biochemistry 2011, 50, 8926−8936. (42) Mayer, J. M. Acc. Chem. Res. 2011, 44, 36−46.
H
DOI: 10.1021/acs.inorgchem.5b02084 Inorg. Chem. XXXX, XXX, XXX−XXX
