View PDF VersionPrevious ArticleNext Article

Open Access Article
This Open Access Article is licensed under a
Creative Commons Attribution 3.0 Unported Licence
DOI: 10.1039/D4DT02708F (Paper) Dalton Trans., 2025, 54, 728-744

Distinguishing between aquo and hydroxo coordination in molecular copper complexes by 1H and 17O ENDOR spectroscopy†

Julia Haak ab and George E. Cutsail III *ab
aMax Planck Institute for Chemical Energy Conversion, Stiftstraße 34-36, D-45470 Mülheim an der Ruhr, Germany. E-mail: george.cutsail@cec.mpg.de
bInstitute for Inorganic Chemistry, University of Duisburg-Essen, Universitätsstraße 5–7, 45117 Essen, Germany

Received 24th September 2024 , Accepted 12th November 2024

First published on 21st November 2024

Abstract

Aquo and hydroxo ligands play an essential role in the chemistry of many copper enzymes and small molecule catalysts. The formation of a series of copper complexes with H2O and OH− ligands in various positions, including [Cu(bpy)(OAc)(H2O)2,ax]+ (Cu-I), [Cu(bpy)(OH)2,eq(HxO)2,ax] (Cu-III), [Cu(OH)4,eq(HxO)2,ax]2− (Cu-IV), [Cu(bpy)(H2O)2,eq(H2O)2,ax]2+ (Cu-V) and [Cu(bpy)2(H2O)ax]2+ (Cu-VI), were investigated through Electron Paramagnetic Resonance (EPR) and UV-Vis spectroscopy in aqueous copper bipyridine solutions in the dependence of the pH and the copper-to-bipyridine ratio (bpy = 2,2′-bipyridine). 2H- and 17O-enrichment of the copper complexes allowed us to determine the 1H and 17O nuclear hyperfine interactions of their HxO ligands via Q-band Electron Nuclear Double Resonance (ENDOR) spectroscopy. These techniques gave direct insight into the metal–ligand covalencies and geometries and were further supported by Density Functional Theory (DFT) calculations. It is shown that 1H and 17O ENDOR spectroscopy can aid in (1) determining the coordination position, thereby differentiating between equatorial and axial HxO ligands and (2) distinguishing equatorial aqua and hydroxo ligands, particularly through their anisotropic dipolar components. We further studied the influence of trans coordinating ligands on the hyperfine parameters of aquo and hydroxo ligands, enabled through contrasting the coordination environments in the examined complexes, supported by quantum chemical computations.

Introduction

The ability of monocopper active sites to catalyze both the activation of molecular oxygen, on the one hand, and the formation of the O–O bond in dioxygen, on the other hand, is of fundamental importance to numerous catalytic processes. In nature, dopamine β-monooxygenase (DβM) and peptidylglycine α-amidating monooxygenase (PHM) use molecular oxygen to selectively hydroxylate secondary C–H bonds (under production of water) at uncoupled dinuclear copper sites, where the activation of dioxygen occurs categorically at a single copper center.1 Other prominent monooxygenases such as lytic polysaccharide monooxygenase (LPMO) that catalyzes the oxidative depolymerization of polysaccharides2,3 and particulate methane monooxygenase (pMMO) that oxidizes methane to methanol4 have been demonstrated to use dioxygen as a co-substrate at monocopper active sites, two reactions that play crucial roles in the production of renewable energy and thereby the reduction of greenhouse gases.

In materials science research, the intriguing redox chemistry of copper enzymes has drawn considerable interest, which led to the development of a variety of synthetic heterogeneous and homogeneous copper catalysts over the past decades, for example for oxidative C–H functionalizations.5 Copper-exchanged zeolites have proved to be especially interesting for the conversion of methane to methanol, mimicking pMMO's reactivity.6 More recent studies have targeted copper active site designs in metal–organic frameworks (MOFs) to mimic biologically active sites and control catalysis.7,8

By means of O–O bond formation, a variety of heterogeneous and homogeneous catalysts for electrochemical water oxidation have also been designed.9–18 The first homogeneous copper water oxidation catalyst (WOC) was reported in 2012 by the group of Mayer, demonstrating the catalytic activity of the copper bipyridine system (1[thin space (1/6-em)]:[thin space (1/6-em)]1 M[thin space (1/6-em)]:[thin space (1/6-em)]L ratio) at a high pH value.10 At pH values between pH 11.8 and 13.3 electrochemical water oxidation occurs with high turnover frequencies of ∼100 s−1. This work has inspired more mechanistic and theoretical studies of the parent copper bipyridine system and its substituted modifications, which are generally all catalytically active in aqueous solutions and have exchangeable water or hydroxo ions that occupy an open coordination site.13,15,17,19–21 These exchangeable ligands are readily displaced upon reduction of the Cu(II) center to Cu(I) to allow for O2 or H2O2 binding, accompanied by re-oxidation of the metal center to a Cu(II) state. For various copper WOCs, the coordinated hydroxo ion is suggested to play a direct role in the formation of O2 and its incorporation into the product.15,19

EPR spectroscopy allows one to obtain information on the general coordination environment, the electronic configuration and the nuclearities (e.g. dimerization) of paramagnetic metal centers. For monocopper(II) active sites, a d9S = 1/2 electronic configuration yields an EPR spectrum that directly shows the ground state configuration of the singly occupied molecular orbital (SOMO), distinguishing dx2−y2 from dz2 ground-states by the g-tensor (dx2−y2: g∥ > g⊥; dz2: g⊥ > g∥). The copper hyperfine interaction, which is commonly observed by EPR spectroscopy, is a reporter of the ligand field of the copper center and gives estimates on the degree of spin delocalization onto ligands. More detailed electronic and geometric information is encoded in the magnitude and orientation of nuclear superhyperfine interactions of the ligand nuclei, which sometimes may be resolved by EPR spectroscopy directly. In copper bpy complexes, the 14N couplings of the bpy ligands are often at least partly observed, which allows for estimates on the magnitude of hyperfine interactions.22 The determination of the ligand's full hyperfine tensor from EPR alone can be challenging and in many cases unfeasible, requiring the application of more advanced EPR techniques. The EPR linewidth of metal centers, including copper, is often significantly broadened by numerous possible influences, including other numerous weaker hyperfine interactions and/or strain effects, which effectively conceal the hyperfine interactions of weakly coupled nucleus like protons. To resolve such couplings, the use of hyperfine techniques like ENDOR spectroscopy is required. The understanding and benchmarking of 1H couplings are critical for the distinction between various ligands, specifically water and hydroxo, and even their coordination positions (i.e. equatorial vs. axial).

The interaction of the copper center with exchangeable oxygen ligands (water and hydroxo ions) holds great information on the exact nature of the ligands, covalencies, distances and geometries. While there are various past studies and assignments on the protons of these HxO ligands,23–30 direct studies of the oxygen hyperfine interaction are less common and are less systematically understood. Oxygen has only one NMR-active isotope, 17O, with a nuclear spin of I = 5/2, which has a very low natural abundance (0.038%), thus requiring any potential EPR and hyperfine measurements to use 17O enrichment. In terms of 17O hyperfine studies of copper systems, they are still relatively rare in comparison with other nuclei, despite the numerous catalytic processes that invoke oxygen ligands.30–34

Herein, we have set out to understand the ligand hyperfine interactions of aqua and hydroxo copper complexes via both 1H and 17O ENDOR spectroscopy to understand the hyperfine and structural relationship of these ligands. To systematically understand these interactions, we have characterized a range of well-studied copper bpy complexes to survey how changes in the ligand sphere change the spectroscopic footprint of the respective HxO ligand. This system was chosen not simply for its relevance to water oxidation catalysis, but also for its ease of preparation (low H217O requirements) and variety of oxygen ligand conformers.

Based on previous studies,22,35–47 it was expected that the bis-aqua complex [Cu(bpy)(H2O)2]2+ forms upon mixing of bpy and a copper source near neutral pH.10 As the pH is then increased, this species is suggested to transform via the EPR-silent dimeric [Cu(bpy)(OH)]22+ at intermediate pH values into the bis-hydroxo complex [Cu(bpy)(OH)2] at further elevated pH, which was assumed to be the water oxidation catalyst (Scheme 1).


image file: d4dt02708f-s1.tif
Scheme 1 Proposed pH dependent structures of copper bpy (1[thin space (1/6-em)]:[thin space (1/6-em)]1) complexes.

In this study, we applied UV-Vis, EPR, ENDOR and ELDOR-Detected NMR (EDNMR) spectroscopy, in combination with DFT calculations to shed further light on the nature of the pH-dependent equilibria and the evolving complexes. Most importantly, we take advantage of the simplicity of the copper bpy system to purposefully study the ligand superhyperfine interactions in various water-derived ligands. Employment of 17O-enriched water allows not only the study of 1H couplings, but also the direct interaction of the 17O nucleus of H2O and OH− ligands with the metal center.

Methods and materials

Sample preparation

Aqueous solutions of copper bpy complexes were prepared by mixing one equivalent of 2,2′-bipyridine (Sigma Aldrich) with one equivalent of Cu(OAc)2·H2O (Sigma Aldrich) in water, or in the case of Cu-VI by mixing both components in a 2[thin space (1/6-em)]:[thin space (1/6-em)]1 ratio. If not mentioned otherwise, NaOAc·3H2O (Alfa Aesar) was added to a final concentration of 0.1 M. For Cu-V 30% v/v glycerol was used instead of sodium acetate. 17O- and 2H-enriched samples were prepared in H217O (90 atom%, Sigma Aldrich) or D2O (99.95 atom%, Deutero). pH values were adjusted with solutions of NaOH (prepared from pellets; Alfa Aesar) either with an InLab Versatile Pro or an InLab Micro Pro-ISM pH electrode (Mettler-Toledo), calibrated by four-point calibration in the range of pH 4 to 13.

UV-Vis spectroscopy

UV-Vis spectra were recorded with a Cary 60 UV-Vis spectrophotometer from Agilent using Hellma quartz cuvettes (quartz glass of high performance purity) with a path length of 1 cm. Spectra were recorded on copper bpy solutions of concentrations of 10 mM.

Cyclic voltammetry

Cyclic voltammograms were obtained using a VersaSTAT4 potentiostat with a platinum auxiliary electrode, a 3 mm diameter glassy carbon working electrode and an Ag/AgCl (3 M KCl) reference electrode (all electrodes from BASi). 0.1 M NaOAc was used as an electrolyte.

CW X-band EPR spectroscopy

Continuous-wave X-band EPR spectra were recorded using a Bruker MS5000 spectrometer. The samples that were recorded at room temperature contained 10 mM copper bpy solutions, filled in 50 μL capillaries (Hirschmann). Samples between pH 6.1 and 12.7 were produced via pH titrations (pH adjustments with NaOH, see above). The spectrometer conditions applied for all samples are reported in Table S1.† For measurements at cryogenic temperatures a liquid nitrogen cryostat or a finger dewar was employed and custom quartz EPR tubes with 4 mm outer diameter (OD) were filled with 1 mM solutions and frozen. The spectrometer conditions for these measurements are reported in Table S2.†

Pulsed Q-band EPR and ENDOR spectroscopy

Q-band EPR and ENDOR spectra were recorded using 1 mM solutions in 2.8 mm OD custom quartz tubes with a Bruker Elexsys E580 spectrometer equipped with home-built up/down Q-band conversion accessories,48 an Oxford CF935 helium flow cryostat and temperature controller, and a cylindrical TE011 microwave resonator.49 Two-pulse (Hahn) echo detected Q-band EPR spectra were recorded with a π/2–τ–π–τ–echo pulse sequence and Davies50 ENDOR experiments employed a π–TRF–twait–π/2–τ–π–τ–echo sequence, where the radio frequency (RF) pulse is applied during time TRF and the RF was randomly hopped51 without phase cycling. The spectrometer conditions for the individual experiments are reported in Tables S3–S5.†

Pulsed W-band EPR and EDNMR spectroscopy

W-band pulsed spectra were recorded using 1 mM solutions filled in 0.9 OD custom quartz tubes and collected on a Bruker Elexsys E680 spectrometer equipped with a closed cycle helium cryostat system. For EPR measurements, a two-pulse (Hahn) sequence of π/2–τ–π–τ–echo was employed. The magnet was swept up and down and the offsets were averaged to correct for sweep delays. EDNMR spectra52 were recorded using a tHTA–twait–π/2–τ–π–τ–echo sequence, where tHTA is the length of the high turning angle pulse and twait is the time between the HTA pulse and the echo detection. The spectrometer conditions for the individual experiments are reported in Tables S6 and S7.†

Raw data were processed and analyzed in Matlab R2023a and simulated using the EasySpin package (v. 6.0.0-dev. 51).53 EDNMR simulations were obtained using the function horseradish.54

DFT calculations

All calculations were performed with the Orca quantum chemistry software (v. 5.0),55,56 taking dispersion effects into account by utilizing Grimme's D3 correction with Becke–John damping.57,58 Solvation effects (water) were included with the conductor-like polarizable continuum model (CPCM)59 and relativistic effects were treated at the level of zeroth order regular approximation (ZORA).60–62 Geometry optimizations were carried out using the BP86 functional,63,64 while the B3LYP functional65,66 was used for the determination of the EPR parameters. The all-electron def2-TZVP basis set67,68 was employed, with the exception of the calculation of EPR parameters, where the epr-III basis set69,70 was used for all hydrogens and the aug-cc-pVTZ-J basis set for copper.71,72 For all calculations automatically generated auxiliary basis sets were applied (AutoAux).73

Results and discussion

Catalytic activity and UV-Vis and EPR spectroscopy

Copper bpy complexes were prepared as described for the water oxidation experiments reported by Barnett et al. by mixing one equivalent of bipyridine ligand with one equivalent of copper acetate.10 To understand the system as previously reported, we prepared all samples with 0.1 M sodium acetate electrolyte, unless noted otherwise. We did find that the addition of the electrolyte aided as a glassing agent during the freezing of EPR samples (Fig. S1†). Spectrophotometric pH titrations of the 1[thin space (1/6-em)]:[thin space (1/6-em)]1 Cu[thin space (1/6-em)]:[thin space (1/6-em)]bpy solution have a similar result to what has previously been reported.22 Briefly, the color of the solution changes from light blue (pH 6.5, λmax = 668 nm, and ε = 37 M−1 cm−1) to a darker blue at a high pH value (pH = 12, λmax = 615 nm, ε = 41 M−1 cm−1), as shown in Fig. 1. The blue shift is analogous to the spectra obtained by Garribba et al. and consistent with the equilibrium shift from [Cu(bpy)(H2O)2]2+ to [Cu(bpy)(OH)2].22
image file: d4dt02708f-f1.tif
Fig. 1 Spectroscopic investigations of copper bpy solutions at several pH values showing (a) the shifting d–d transitions in the UV-Vis spectra and (b) the CW X-band (∼9.5 GHz) EPR spectra recorded at room temperature. The relative amount of paramagnetic species at the individual pH points was obtained from the double integral of the EPR spectra and is added as an inset. Experimental conditions are reported in the Experimental section.

At even higher pH values, the spectral character does not change significantly (pH = 12.7, λmax = 605 nm, and ε = 52 M−1 cm−1); however, the overall background intensity increases, leading to a higher effective absorptivity and the lack of clear isosbestic points (Fig. S2†). This is most likely related to the observable formation of a dark precipitate at high pH values, which was reported before.74 The formation of these insoluble copper species (possibly copper oxides) and their consecutive deposition on the cuvette then leads to diffuse scattering and a systematic increase (and error) of the measured absorptivity. However, Barnett et al. showed that the active catalyst in their water oxidation reactions is a homogeneous one, indicating that the formed precipitate does not play a significant role in catalysis, but rather represents a decomposition product.10

At pH ≳ 12 catalytically active solutions were obtained, as seen by large, irreversible oxidation waves (Fig. S3†), in good agreement with what has been described by Barnett et al.10

The Cu(II) complexes were further investigated by X-band (∼9.43 GHz) EPR spectroscopy. The spectra recorded at room temperature are shown in Fig. 1b. The approximately isotropic spectra of the rapidly tumbling molecular solutions at low to neutral pH (pH 6.1–7.0) exhibit a signal centered at giso ∼ 2.15, split to a quartet caused by the electron-nuclear hyperfine interaction of the 63,65Cu isotopes (I = 3/2), aiso(63,65Cu). As the pH is further increased, the quartet signal intensity decreases, before the appearance of a new quartet signal at high pH values with a slightly lower giso value of ∼2.12 and a larger |aiso|-value. The double integration of the collected spectra quantifies a maximum signal reduction of 90% at pH 8.9, Fig. 1b. This is explained by the formation of an EPR silent species. The three species identified at neutral, intermediate and high pH values are termed Cu-I, Cu-II and Cu-III, and are assumed to be [Cu(bpy)(H2O)2]2+, the EPR silent dimer [Cu(bpy)(μ-OH)]22+ and [Cu(bpy)(OH)2], as assigned previously (Scheme 1).10,22,37

The room temperature spectra are not fully isotropic, but exhibit a slightly anisotropic character due to the slow tumbling rate on the timescale of the EPR experiment (10−11 s). To accurately simulate the room temperature spectra, the complete spin Hamiltonian must be known, which was estimated from the frozen solution (solid state, powder) EPR spectra, shown in Fig. 2. All frozen solutions exhibit axial EPR spectra, typical of type-2 copper with a dx2−y2 ground state, with some of them also exhibiting resolved nitrogen superhyperfine interaction (I(14N) = 1/2) of the bpy ligand along g⊥. Collecting EPR spectra at several frequencies (X-, Q- and W-band frequency) and globally fitting the multifrequency EPR spectra (Fig. S4–S9†) yields simulations that excellently reproduce the spectra, including the nitrogen superhyperfine interaction – when available (Fig. 2 and Table 1). The 14N superhyperfine assignments were further confirmed by orientation-selective 14N Davies ENDOR experiments (see below).


image file: d4dt02708f-f2.tif
Fig. 2 CW X-band (∼9.5 GHz) EPR spectra of copper bpy solutions at several pH values collected at 77 K with an enlargement of the copper hyperfine in the g∥ region in black with simulations in color. Spectra of the solutions at room temperature (r.t.) with simulations are added as inserts for three pH values. Experimental conditions are reported in the Experimental section. Simulation parameters are listed in Table 1.
Table 1 Summary of spin Hamiltonian parameters for Cu-I, Cu-III and Cu-IV obtained through EPR simulations together with the estimated uncertainties for some parameters. The full parameters are reported in Tables S8 (frozen solution) and S10† (room temperature). Within the scope of this study the three complexes are assigned to be [Cu(bpy)(OAc)(H2O)2,ax]+, [Cu(bpy)(OH)2,eq(HxO)2,ax] and [Cu(OH)4,eq(HxO)2,ax]2−, respectively
  Cu-I Cu-III Cu-IV
a The isotropic components of the g- and A-tensors are calculated as follows: image file: d4dt02708f-t1.tif and image file: d4dt02708f-t2.tif. b Copper hyperfine couplings are reported for the 63Cu isotope. Simulations consider also the 65Cu isotope with hyperfine couplings that are scaled by the gyromagnetic ratio of both nuclei: image file: d4dt02708f-t3.tif.
Simulation parameters of frozen solution spectra
g = [g1, g2, g3] [2.278 ± 0.001, 2.068 ± 0.003, 2.055 ± 0.001] [2.254 ± 0.001, 2.055 ± 0.001, 2.050 ± 0.002] [2.272 ± 0.0005, 2.053 ± 0.0005, 2.050 ± 0.0005]
g iso[thin space (1/6-em)]a 2.134 2.120 2.125
A (63Cu) = [A1, A2, A3] in MHzb [525 ± 7, 43 ± 10, 36 ± 10] [560 ± 5, 64 ± 10, 62 ± 10] [582 ± 4, 90 ± 5, 92 ± 5]
Simulation parameters of room temperature spectra
g = [g1, g2, g3] [2.284, 2.074, 2.061] [2.254, 2.055, 2.050] [2.277, 2.058, 2.055]
g iso[thin space (1/6-em)]a 2.140 2.120 2.130
A (63Cu) = [A1, A2, A3] in MHzb [−525, +43, −36] [−560, −64, −62] [−582, −40, −40]
a iso(63Cu) in MHza −173 −229 −221
Rotational correlation time in s (isotropic) (3.5 ± 0.3) × 1 × 10−11 (2.5 ± 0.4) × 1 × 10−11 (2 ± 0.6) × 1 × 10−11

The low temperature EPR spectra of copper bpy solutions at moderate pH (pH 6–7) can be simulated with a mostly axial g-tensor (g∥ = g1 > g⊥ = g2 = g3) with a small rhombic splitting, gCu-I = [2.278, 2.068, 2.055], and an axial hyperfine tensor (A = [A1, A2, A3]) of ACu-I(63Cu) = [525, 43, 36] MHz. At pH 12, a single species with a larger A∥(63Cu) value of 560 MHz and overall lower g-values is observed (gCu-III = [2.254, 2.055, 2.050]).

With g- and A(63Cu)-tensor estimates from the frozen solution EPR spectra, we are now able to simulate the slow-motional regime room temperature EPR spectra (Fig. 2 and Table 1). The room temperature spectra at pH ≤ 6.5, and also at pH 12.4, are well reproduced by simulation by considering only a single set of spin Hamiltonian parameters for each. The EPR spectra at pH values between pH 6.5 and pH 12.0 can be reproduced as mixtures of the Cu-I and Cu-III components (Fig. S10†). However, at intermediate pH values larger deviations between the simulation and experiment are observed, which implies the formation of additional EPR active species together with dimer formation. Nevertheless, our simulations still are in reasonable agreement with the experimental spectra, which indicates that Cu-I and Cu-III are the two major EPR active species between pH 7 and pH 12.

While the solid-state simulations do not provide the possibility to determine the absolute or relative signs of the individual hyperfine components, the relative hyperfine signs are determinable from the simulation of the room temperature EPR spectrum. By making A∥(63Cu) negative, as expected for a type-2 copper, the simulation parameters of the room temperature spectrum of Cu-I reveal a slight but necessary rhombic splitting of the A⊥(Cu) values in order to accurately reproduce the experiment and yield the observed isotropic hyperfine splitting pattern, Fig. S10.† Therefore, the copper hyperfine of Cu-I is either ACu-I(63Cu) = [−525, +43, −36] MHz or [−525, −43, +36] MHz, where the former is only slightly favored, Fig. S11.†

The spin Hamiltonian parameters of both complexes are in good agreement with the ones obtained from simulations of the room temperature spectra (Table 1). For Cu-I, slightly lower individual g- and giso-values at low temperatures are observed compared to the ones determined at room temperature. There is no difference observed between the room temperature and low temperature samples of Cu-III. Generally, Cu-III exhibits a smaller giso and larger hyperfine splitting than Cu-I. The trend of a reduced g∥ and increased A∥ for Cu-III compared to Cu-I reflects the findings of Garribba et al. that showed the same trends for the species they assign to be [Cu(bpy)(H2O)2]2+ (g∥ = 2.307 and A∥ = 510 MHz) and [Cu(bpy)(OH)2] (g∥ = 2.232 and A∥ = 588 MHz),22 where ethylene glycol was used as a glassing agent rather than the sodium acetate reported here.

Notably, when the pH is further increased, the CW X-band EPR spectra of the frozen solutions show the formation of an additional species at higher pH values (Fig. 2). The amount of this species increases as the pH is taken up further until it is the only detected species at pH 13.7 (Cu-IV). The spectrum of Cu-IV is best simulated with an axial g-tensor (gCu-IV = [2.272, 2.053, 2.050]) and large copper hyperfine (ACu-IV(63Cu) = [–582, –90, –92] MHz). It lacks any resolved 14N hyperfine coupling in the g⊥ region and exhibits significantly narrower lines, particularly along g∥. This is especially pronounced at the copper hyperfine transition at 2680 G which is much narrower than in Cu-I and Cu-III which allows to resolve two sharp features, corresponding to transitions of the 63Cu and 65Cu isotopes. The dramatic EPR linewidth sharpening and the loss of 14N hyperfine coupling are attributed to the loss of the bpy ligand. The formation of the tetra hydroxide complex [Cu(OH)4]2− is therefore expected, and the spectrum is in very good agreement with the spectrum of [Cu(OH)4]2− reported by Chao et al. (g = [2.252, 2.050, 2.046]; A∥ = 558 MHz; A⊥ = 89 MHz).75 The spectrum of Cu-IV obtained at room temperature is well reproduced by a simulation with spin Hamiltonian parameters, which are in good agreement with the ones obtained at low temperature (Fig. S10†). However, a slightly lower copper hyperfine is applied (Table 1).

1H ENDOR spectroscopy

To characterize the HxO ligands of the presented copper complexes, Q-band pulsed Davies ENDOR experiments were conducted and the ligand superhyperfine couplings were determined. The 1H Davies ENDOR spectra of the copper complexes are displayed in Fig. S12–S16.† The spectra exhibit a multitude of doublets that are centered at the nuclear Larmor frequency νn and split by the effective, orientation-selective hyperfine coupling A, as expected for weakly coupled nuclei (|A| < 2|νn|; νn(1H) ∼ 51 MHz at 12[thin space (1/6-em)]000 G). The complex 1H ENDOR pattern originates from a large number of exchangeable protons of the water/hydroxo ligands and non-exchangeable protons of the bpy ligand. To identify the exchangeable HxO protons, solvent-exchanged samples in D2O were prepared and the difference in the spectra was calculated (Fig. S12–S16†). A selection of difference spectra along with their simulations are displayed in Fig. 3. We note that the 1H ENDOR spectra of Cu-IV in H2O are dramatically simplified compared to all other complexes because of the loss of the bpy ligand. Hence, the corresponding 1H ENDOR spectra of the D2O solvent exchanged sample had an extremely 1H weak ENDOR response, as all of the protons in Cu-IV are exchangeable.
image file: d4dt02708f-f3.tif
Fig. 3 Q-band (∼34.0 GHz) Davies ENDOR responses of the exchangeable protons found in Cu-I (a), Cu-III (b), Cu-IV (c), Cu-V (d), and Cu-VI (e) in black with simulations in red. The spectra were either recorded through direct measurement (c) or through subtraction of spectra recorded in H2O and D2O. Experimental conditions are described in the Experimental section. Simulation parameters are listed in Table 2. The full patterns along with the simulations are displayed in Fig. S17–S21.†

In the 1H difference spectra of Cu-I, a single sharp 1H doublet of A⊥ = 3 MHz is observed along g⊥ that broadens to a single doublet of approximately A∥ ∼ 7 MHz. The 2D field-frequency ENDOR pattern of the exchangeable 1H signal of Cu-I is well reproduced by simulation of a single proton hyperfine tensor (ACu-I(1H) = [+7, −3, −3] MHz, Table 2). The small isotropic (aiso = +0.3 MHz) and dipolar couplings (t = +3.3 MHz; T = A − aiso; T = [−t, −t, 2t]) of this proton are in good agreement with the values observed for the axially coordinating water molecules in other copper complexes and proteins, e.g. in the prion protein (aiso ∼ 0 MHz, t ∼ 2 MHz)25 or [Cu(H2O)6]2+ in Mg(NH4)2(SO4)2·6H2O (aiso < +0.2 MHz, t ∼ +3.7 MHz),23 reflecting the minimal overlap (Fermi contact) of the axial waters with the dx2−y2 SOMOs. As is, this signal cannot be exclusively assigned to an equatorially or axially coordinated ligand; however, we will demonstrate experimentally (vide infra) that it is indeed from axial H2O ligand(s), and no equatorial waters are present.

Table 2 1H hyperfine couplings of the exchangeable protons for copper bpy complexes obtained from simulation
  A (1H) = [A1, A2, A3] in MHz Euler anglesa [α, β, γ] in ° a iso in MHzb T in MHzb t in MHzb
a The reported Euler angles describe the angles between the calculated g-tensor and the respective hyperfine tensor and are reported in a zyz’ convention. b The isotropic and anisotropic components of the hyperfine component aiso and t are calculated as image file: d4dt02708f-t4.tif and A = aiso + T with T ≈ [−t, −t, 2t].
Cu-I [+7, −3, −3] [15, 0, 0] +0.3 [+6.7, −3.3, −3.3] +3.3
Cu-III (eq) [−16.5, −8.0, +7.5] [0, 0, 0] −5.6 [−10.9, −2.4, +13.1] +6.6
Cu-III (ax) [+4.9, −4.1, −3] [25, 0, 0] −0.7 [+5.6, −3.4, −2.3] +2.8
Cu-IV (eq) [−13.5, −5, +8.5] [0, 0, 0] −3.2 [−10.2, −1.7, +11.8] +5.9
Cu-IV (ax) [+4.3, −3.6, −3.4] [20, 0, 0] −0.9 [+5.2, −2.7, −2.5] +2.6
Cu-V (eq) [−5.5, −9.5, +7.0] [30, 0, 0] −2.6 [−2.9, −6.9, +9.6] +4.9
Cu-V (ax) [+6, −3, −3;] [15, 0, 0] 0 [+6, −3, −3] +3.0
Cu-VI [+6, −3, −2.8] [50, 0, 0] +0.1 [+5.9, −3.1, −2.9] +3.0

While a single exchangeable proton is observed for Cu-I, the inclusion of two unique proton signals with very different magnitudes for Cu-III and Cu-IV is required to reproduce their observed 1H ENDOR pattern. For both samples this includes a proton with larger couplings (set 1: ACu-III(1H) = [−16.5, −8.0, +7.5] MHz; ACu-IV(1H) = [−13.5, −5.0, +8.5] MHz) and another proton of smaller couplings. The latter one is comparable in size to the one presented for Cu-I (set 2: ACu-III(1H) = [+4.9, −4.1, −3] MHz; ACu-IV(1H) = [+4.3, −3.6, −3.4] MHz) with small isotropic (|aiso| ∼ ≤1 MHz) and also rather small dipolar contributions (t ∼ +3 MHz), which led us to assign the set 2 protons to axial HxO ligands. The shape, however, is less axial and the isotropic component of the hyperfine is larger (Table 2). A possible reason for this is the interaction of solvent waters with the nearby hydroxo ligands, creating a different solvation shell environment compared to the one found for Cu-I with water redirected towards the xy-plane (g⊥). The potential overlapping signals of axial and second shell waters preclude a detailed assessment of their individual couplings, including the determination of the nature (OH− and H2O) of such axial ligands. While we expect the coordination of water ligands to the neutral/negatively charged [Cu(bpy)(OH)2]/[Cu(OH)4]2− cores, we cannot conclusively exclude OH− coordination in the axial positions and will consequently keep referring to those ligands as HxO. More importantly, the patterns of both set 1 proton couplings are well resolved yielding tensors exhibiting a considerable amount of rhombic distortion with large anisotropic character that is approximately twice as high as the one found for the axial water in Cu-I. Furthermore, a larger isotropic contribution to the hyperfine tensor, aiso = −5.6 MHz (Cu-III) and −3.2 MHz (Cu-IV), is observed compared to the small +0.3 MHz isotropic coupling of the axial water in Cu-I. The increased isotropic character, together with the orientation of tmax along g⊥, is consistent with an equatorial ligand that has σ overlap with the dx2−y2 SOMO, yielding more spin delocalization to the proton via Fermi contact interaction. These findings are in excellent agreement with the 1H coupling determined for the hydroxo proton found in the [Cu(OH)]+ unit of zeolites, determined to be [−13.0, −4.5, +11.5] MHz76, equivalent to an aiso value of −2.0 MHz. The dipolar couplings exhibit an averaged t value of +6.8 MHz, similar to the ones observed for Cu-III (+6.6 MHz) and Cu-IV (+5.9 MHz), all three with clear deviations from axial symmetry. Therefore, the protons associated with set 1 are tentatively assigned to equatorially coordinating OH− ligands. An overview of 1H and 17O couplings in relevant copper studies found in the literature is given in Tables S13 and S14,† respectively.

To further verify the identity of the axially coordinated ligands in the presented complexes, the bis-bpy complex [Cu(bpy)2]2+ (Cu-VI) was prepared as described before.22,41 The steric demands of the ligands force the complex into a pseudo-tetrahedral configuration,22 precluding equatorial coordination of solvent molecules, allowing additional coordination only in the axial position(s). The 1H ENDOR difference spectra of Cu-VI show indeed only a single splitting, which can be simulated with a tensor that is in excellent agreement with the tensor found for Cu-I (ACu-VI(1H) = [+6, −3, −2.8] MHz; Table 2). This confirms the axial coordination of water in this structure and in Cu-I, and also reinforces the assignment of HxO ligands in the axial position of Cu-III and Cu-IV. Of note, ENDOR cannot quantify the number of equivalent ligands and can therefore not differentiate between one or two axial HxO ligands. Similarly, the EPR spectra cannot differentiate between both possibilities.

The resemblance of the exchangeable protons’ 1H hyperfine tensors of Cu-VI and Cu-I strongly suggests that Cu-I has no coordinating equatorial waters. One possible reason for this is the blockage of the equatorial positions through the coordination of the employed sodium acetate at low pH. To test its role, we prepared another 1[thin space (1/6-em)]:[thin space (1/6-em)]1 Cu[thin space (1/6-em)]:[thin space (1/6-em)]bpy complex at neutral pH and used glycerol instead of sodium acetate as a glassing agent. A slight shift of the spin Hamiltonian parameters is observed upon change of the glassing agent (Fig. S7†) with an increase of g∥ (|Δ| = 0.032) and a minor decrease of A∥(63Cu)(|Δ| = 45 MHz) in between Cu-I and Cu-V, reflecting a change in the coordination sphere. The 14N ENDOR experiments, however, do not exhibit any significant changes, confirming highly similar interactions between the copper center and the bpy ligand, as seen for all bpy complexes (Fig. S22†). The 1H ENDOR difference spectra of Cu-V (Fig. 3 and S15†) show two sets of signals. One 1H ENDOR signal of A ∼ 3 MHz is observed along g⊥, strikingly similar to the one observed in Cu-I. At this field position, there is also a clear new 1H doublet with a larger observed coupling of 9 MHz. Moving to a lower magnetic field, the 1H ENDOR pattern of Cu-V becomes more complex, with large effective splitting still observed halfway through the pattern that collapses into several overlapping signals at the lowest field. This ENDOR pattern is once again satisfactorily reproduced with the inclusion of two different proton hyperfine patterns. Again, a smaller axial coupling (set 2: ACu-VI(1H) = [+6, −3, −3] MHz) is observed and assigned to axially coordinating waters. The second set of signals is best simulated with, again, a more rhombic and overall larger tensor of ACu-VI(1H) = [−5.5, −9.5, +7.0] MHz (set 1), similar to what has been observed for the equatorial water ligands found in copper bis(oxazoline) (A(1H) = [−10.8, −5.02, +8.6] MHz)29 and [Cu(H2O)6]2+ in Mg(NH4)2(SO4)2·6H2O (A(1H) ∼ [+9, −8, −4] MHz, Table S13†).23 Hence, we attribute the set 1 proton of Cu-VI to equatorial waters. This tensor is still dominated by its anisotropic component (t = +4.9 MHz) but has an appreciable isotropic component, aiso = −2.6 MHz, as expected for an equatorial ligand.

A summary of the resolved exchangeable 1H signals is detailed in Table 2. Overall, smaller isotropic and anisotropic contributions of the hyperfine couplings can be observed for the equatorial ligands in Cu-V (aiso = −2.6 MHz; t = +4.9 MHz) compared to Cu-III (aiso = −5.6 MHz; t = 6.6 MHz) and Cu-IV (aiso = −3.2 MHz; t = +5.9 MHz). This reflects the reduced covalent interaction between the copper center and the H2O ligand, compared to the OH− ligand, as seen by the reduction in aiso. Furthermore, the expected increased Cu–O (and thereby Cu–H) distance of Cu–OH2 compared to Cu–OH manifests itself in the smaller value for through-space (distance dependent) anisotropic coupling t. While it is assumed that both Cu-III and Cu-IV have hydroxo ligands, significant differences can also be observed between the 1H couplings of their exchangeable protons, with overall larger isotropic and anisotropic contributions for Cu-III, making Cu-IV more similar to Cu-V (|Δ(aiso)|IV–V = 0.6 MHz; |Δ(t)|IV–V = 1.0 MHz) than to Cu-III (|Δ(aiso)|IV–III = 2.4 MHz; |Δ(t)|IV–III = 0.7 MHz). This shows that a clear differentiation between H2O and OH− ligands by 1H ENDOR alone is not readily feasible. Hence, we further characterized the different HxO ligands using 17O ENDOR and DFT calculations (see below). An overview of the characterized complexes is given in Chart 1.


image file: d4dt02708f-c1.tif
Chart 1 Structures of copper complexes examined in this study. All complexes show resonances of axially coordinating HxO ligands, which are omitted in the structures for clarity.

17O and 14N ENDOR spectroscopy

Cu-I, Cu-III, Cu-IV, Cu-V and Cu-VI were further characterized for their 17O superhyperfine interaction. At the Q-band frequency, 17O resonances of HxO ligands have the potential to overlap with the 14N resonances of the bpy ligand (I(14N) = 1, 99.6%), both of which fall in the strong coupling regime (|A| > 2|νn|) and therefore yield signal pairs centered at A/2 split by 2νn (νn(14N) ∼ 4 MHz and νn(17O) ∼ 7 MHz at 12[thin space (1/6-em)]000 G). The 17O resonances are distinguished from the overlapping 14N signals of the bpy ligand by overlaying the ENDOR spectra of the 17O labeled and unlabeled samples and subtraction of the two spectra (Fig. 4). The spectra of Cu-I, Cu-III, Cu-V and Cu-VI exhibit signal intensities between 15 and 25 MHz (Fig. S23–S27†), which can tentatively be assigned to the ν+ features of the bpy ligand. The lower frequency feature (ν−), on the other hand, is cut off due to the lower limit of the employed resonator (as will also be seen for the 17O resonances below). The 14N ENDOR spectra are simulated in conjunction with the respective X-band EPR spectrum of the sample, where both simulations utilize two strongly coupled nitrogen nuclei. For all complexes, the hyperfine tensors of the two nuclei are set to be identical in size, but differ in their orientation, orienting the maximal component of one of the tensors along gx and the maximal component of the other one along gy for the approximately square planar complexes. It is important to note that the simulated 14N ENDOR pattern is fairly insensitive to the orientation of the hyperfine tensor in the x and y direction due to their equivalencies in the axial EPR spectrum. While the ENDOR spectrum could be simulated using only one 14N nucleus, the 14N superhyperfine splitting pattern of the EPR spectrum confirms that two equal nuclei are present. Comparing the 14N ENDOR spectra of the above-mentioned complexes (Fig. S4, S5, S8 and S9†) shows only negligible differences (Fig. S28†). The patterns can be simulated with mostly isotropic hyperfine tensors of similar sizes and shapes (A(14N) ∼ [30, 30, 40] MHz), indicating minimal alterations of the bpy ligand as the pH or the Cu[thin space (1/6-em)]:[thin space (1/6-em)]bpy ratio is varied (Tables S8 and S9†).
image file: d4dt02708f-f4.tif
Fig. 4 17O Q-band (∼34.0 GHz) Davies ENDOR responses of Cu-I (a), Cu-III (b), Cu-IV (c), Cu-V (d) and Cu-VI (e) with simulations in red. The spectra were either recorded through direct measurement (c) or through subtraction (black) of spectra recorded in 17O-enriched H2O (purple) and H2O of natural abundance (blue). Experimental conditions are described in the Experimental section. Simulation parameters are listed in Table 3. The full ENDOR patterns together with their simulations are depicted in Fig. S29–S31.†

Additionally, the 14N ENDOR responses near and at the g⊥ magnetic field positions are further split by 3P due to quadrupole interactions. The axial nature of the g-tensor makes the determination of an exact and unique solution to the 14N ENDOR field-frequency pattern rather challenging for copper active sites.28,77,78 Fig. S4, S5, S8 and S9† show simulations of the data including quadrupole interactions, where the largest component, Pmax, is aligned along the maximal hyperfine component, which reproduces the peak positions satisfactorily. In fact, the quadrupole tensors agree well with that observed for the imidazole coordination in a lytic polysaccharide monooxygenase, for which a rhombic g-tensor allowed for complete orientation selective ENDOR analysis.28 Ultimately, the size of the nitrogen hyperfine interaction is sufficiently well estimated from the EPR and ENDOR patterns together, which shows no substantial variation among Cu-I, Cu-III, Cu-V and Cu-VI. For Cu-IV, on the other side, no 14N resonances are observed (Fig. S25†) confirming the loss of the bpy ligand at pH > 12, as already observed by EPR and suggested by 1H ENDOR spectroscopy.

Subsequently, ENDOR spectra of 17O-enriched samples were collected and the spectra of unlabeled samples were subtracted to directly depict the 17O resonances (Fig. S23–S27†). For Cu-I, no strongly coupled 17O resonances were observed (Fig. 4), although strong couplings (∼30–70 MHz) for equatorial waters are expected.30,32–34 This ultimately confirms the blockage of the copper center by the added sodium acetate as suggested by the 1H ENDOR experiments above. However, the 17O ENDOR difference spectra of Cu-V show clear resonances characteristic of strongly coupled oxygen nuclei (Fig. 4 and S26†). At the ‘single crystal-like position’, along g∥, only a single ν+17O feature is observed at approximately 26.7 MHz. We were unable to detect any ENDOR responses below a radio frequency of ∼17 MHz during this experiment, and this lower limit of detection is clearly seen in the 14N responses that fall off in intensity below this frequency. Because of this limitation of the resonator used, the ν− feature of the strongly coupled 17O doublet is not observed, nor could weakly coupled 17O interactions near the oxygen Larmor frequency be measured. The observed ν+ feature at 26.7 MHz corresponds to a coupling of approximately 41.0 MHz. Along g⊥, a ν+ signal is observed at ∼38.7 MHz, with its ν− feature expected at ∼38.7–13.6 = 25.1 MHz as given by ν− = ν+ − 2vn (νn(17O) = 6.8 MHz at 11[thin space (1/6-em)]814 G), corresponding to an observed hyperfine coupling of A = 63.8 MHz. However, the feature at ∼27 MHz appears broader and with a more asymmetric line shape compared to its ν+ partner. Therefore, it is suspected that a second ν+ feature may overlap at this position with a corresponding coupling of approximately A = 40 MHz. This field position is directed along g⊥ and probes both A2 and A3 of the 17O hyperfine tensor making the overlap of two different couplings probable. This pattern therefore yields an axial hyperfine tensor for the coordinated 17O nuclei, where the observed pattern is well simulated by a tensor of ACu-V(17O) = [−39, −39, −66] MHz. The signs of all 17O tensors are based on the negative nuclear gn value of the 17O nucleus and DFT calculations (see below), in agreement with what has been suggested in the literature previously (Table S13†). The simulations shown in Fig. 4 include a second equivalent 17O nucleus whose A-tensor is rotated by 90° about the z-axis (g∥/A1) to effectively form ACu-V(17O) = [−39, −66, −39] MHz to account for both directions. However, due to the highly axial nature of the g-tensor this inclusion is rather a matter of principle and does not significantly improve the fit, similar to what has been discussed above for the 14N ENDOR simulations. No resolved 17O superhyperfine splitting of the EPR spectra is observed, but only an increased line broadening.

The 17O resonances of Cu-III and Cu-IV are quite similar in character to Cu-V, and can also be simulated with two sets of axial 17O hyperfine tensors, that were set to be identical, but again rotated 90° around the z-axis. Cu-III and Cu-IV have slightly larger hyperfine couplings than Cu-V (Table 3).

Table 3 17O hyperfine couplings of copper bpy complexes obtained from simulation. Other simulation parameters are reported in Table S15†
  A (17O) = [A1, A2, A3] in MHz a iso in MHza ρ s [%] t in MHza,c ρ p [%]
a The isotropic and anisotropic components of the hyperfine tensors, aiso and t, are calculated as follows: image file: d4dt02708f-t5.tif and A = T + aiso with T = [−t, −t, 2t]. The same signs for all three hyperfine components are assumed. b s and p orbital spin populations (ρs,p) were estimated from image file: d4dt02708f-t6.tif and image file: d4dt02708f-t7.tif, where a0 and b0 are the isotropic and anisotropic hyperfine coupling constants (a0(17O) = −4622.83 MHz and b0(17O) = −372.18·2/5 = 148.87 MHz).79 c Based on the DFT calculated Cu–O distances and Mulliken spin populations (Cu-III: 1.90 Å; ρCu = 0.60 (DFT-III); Cu-IV: 1.97 Å; ρCu = 0.62 (DFT-IV); Cu-V: 2.02 Å; ρCu = 0.67 (DFT-I)/2.05 Å; ρCu = 0.69 (DFT-II)), the non-local contribution to the dipolar hyperfine coupling t is estimated to be only <1 MHz, and therefore neglected in the calculation of p orbital spin density.
Cu-III [−29, −72, −29; −29, −29, −72] −43 0.9 −14 9.4
Cu-IV [−34, −34, −70; −34, −70, −34] −46 1.0 −12 8.1
Cu-V [−39, −66, −39; −39, −39, −66] −48 1.0 −9 6.1

Similar to Cu-I, no strong 17O resonances were detected for Cu-VI, although the 1H ENDOR experiments indicated the coordination of axial water ligands for these and all other studied complexes. In the past, axial water ligands exhibited weak 17O couplings of ∼1.5–4 MHz,4,24,80 with the exception of comparably large couplings of A = [−8, −8, −16] MHz found for copper-exchanged zeolites.30 However, couplings of this size appear too low for the lower limit of the resonator employed here.

W-band (∼94 GHz) ELDOR-detected NMR spectra of Cu-I, Cu-III and Cu-IV were recorded to measure the weakly coupled oxygen nuclei (Fig. S32–S36†). Analogous to the 14N/17O ENDOR experiments, the spectra of 17O-enriched samples were subtracted from the spectra of samples of natural abundance, allowing the creation of difference spectra that only display the 17O resonances (Fig. S37 and S38†). The spectra of Cu-III and Cu-IV exhibit features of strongly coupled 17O resonances. Simulations with the 17O hyperfine tensors obtained from ENDOR spectroscopy reproduce the EDNMR spectra quite well (Fig. S39 and S40†), further validating the hyperfine tensor estimates from the ENDOR spectroscopy. Additionally, the EDNMR spectra of Cu-I, Cu-III and Cu-IV all exhibit a weakly coupled signal centered at the 17O Larmor frequency (νn(17O) ∼ 19 MHz at 33[thin space (1/6-em)]000 G), indicating the coordination of axial HxO ligands,81 and/or other solvent-shell 17O water nuclear interactions. Given the lower resolution of the EDNMR experiment, the data preclude accurate hyperfine estimates for these weak couplings, a known disadvantage of the technique.81

The hyperfine tensors of the strongly coupled 17O nuclei in Cu-III, Cu-IV and Cu-V can be divided into their isotropic and dipolar components again, as shown in Table 3. The isotropic components of all three species are in a similar range of −43 to −48 MHz. This is in good agreement with the equatorial water ligands found in copper-exchanged zeolites, with 17O hyperfine couplings between −41 and −51 MHz.30,32 In the [Cu(H2O)6]2+ doped Tutton salt two sets of equatorial ligands along gx and gy can be differentiated and yield isotropic hyperfine components of −44 and −35 MHz, respectively.33 The aiso(17O) values of Cu-III, Cu-IV and Cu-V correspond to minimal s orbital spin population (<1%), while the dipolar components of −9 to −14 MHz indicate a considerable amount of p orbital spin density ρp, with larger values for the two OH− coordinated complexes Cu-III (t = −14 MHz, ρp = 9.4%) and Cu-IV (t = −12 MHz, ρp = 8.1%), compared to the H2O-coordinated complex Cu-V (t = −9 MHz, ρp = 6.1%). The dipolar component of the 17O hyperfine tensors of the equatorial waters in copper-exchanged zeolites30,32 and [Cu(H2O)6]2+ in the Tutton salt33 ranges between −6.5 and −9.5 MHz, with the exception of a larger coupling for one of the 17O species in zeolites (t = −12 MHz).30 While the other couplings are in good agreement with the coupling found for Cu-V, the large coupling rather resembles the 17O couplings of the hydroxide ligands in Cu-III and Cu-IV. The authors discussed the possibility of a hydroxide coordinating instead of water; however, they found the 1H coupling of the respective HxO ligand to be largely overestimated by DFT (∼20 MHz), contradicting their experimental findings. Rather, differences in the hyperfine couplings were attributed to geometrical differences, leading to different degrees of overlap between the Cu dx2−y2 and the O 2p orbitals. While this may be true for the presented case, we also observe general DFT overestimations of the 1H couplings of hydroxide ligands (see below), suggesting that this may not be an ideal tool to discriminate between the two species.

DFT calculations

To shed further light on the speciation of copper bpy complexes depending on pH, DFT calculations on [Cu(bpy)(H2O)2]2+ (DFT-I), [Cu(bpy)(OH)2] (DFT-III) and [Cu(OH)4]2− (DFT-IV) were carried out, as well as on the corresponding complexes with additional axial water molecules ([Cu(bpy)(H2O)2,eq(H2O)2,ax]2+ (DFT-II), [Cu(bpy)(OH)2,eq(H2O)2,ax] (DFT-V) and [Cu(OH)4,eq(H2O)2,ax]2− (DFT-VI)). The DFT calculated spin Hamiltonian parameters of all examined complexes are listed in Tables 4 and S16.† The geometry optimizations of DFT-V and DFT-VI yielded structures without the axial water molecules, caused by their interaction with the hydroxyl ions nearby and the lack of an explicit solvation model. Therefore, these complexes are not further discussed below. The optimized structures of DFT-I, DFT-II, DFT-III and DFT-IV are depicted in Fig. 5.
image file: d4dt02708f-f5.tif
Fig. 5 DFT-optimized structures of DFT-I, DFT-II, DFT-III and DFT-IV. Cu: light blue, O: red, N: dark blue, C: dark gray, and H: light gray.
Table 4 Representativea, calculated ligand hyperfine interactions and the respective Mulliken spin densities of copper bpy complexes
  Ligand hyperfine interactions in MHz Spin populations ρ
A (1H) = [A1, A2, A3] A (17O) = [A1, A2, A3] a iso(17O) t(17O) Cu Sum(O) Sum(N)
a The hyperfine interactions of the individual ligands vary slightly. For clarity, one coupling per ligand is listed. All tensors are listed in Table S16.†
DFT-I [−7.7, +9.0, −3.1] [−34, −61, −35] −43/−46 −9 0.67 0.10 0.23
DFT-II (eq.) [−8.0, +8.2, −3.9] [−32, −32, −59] −41/−40 −8/−9 0.69 0.09 0.22
DFT-II (ax.) [+5.7, −2.8, −2.7] [1.0, 2.5, 2.4] 2.0/2.2 −0.5
DFT-III [−22.3, −10.8, +7.4] [−20, −81, −20] −40 −20 0.60 0.26 0.14
DFT-IV [−16.4, −7.2, +7.8] [−30, −73, −29] −44/−43 −14/−15 0.62 0.36 [—]

As expected, the DFT calculations yielded fairly axial g- and A(63Cu)-tensors with g∥ > g⊥ and A∥ ≫ A⊥ for all examined complexes (Table S16†). For the water coordinated structures slightly higher g∥ and negligibly lower A∥ values are observed when considering axial water molecules (DFT-II) compared to the structure with equatorial waters only (DFT-I) due to the slightly larger induced ligand field. Overall, the calculated g-shift trends of g∥(DFT-I, DFT-II) > g∥(DFT-IV) > g∥(DFT-III) are in good agreement with the experimental observations (g∥(Cu-V) > g∥(Cu-IV) > g∥(Cu-III)). The changes in A∥ are rather small in the experimental and calculated values. However, the calculations do reproduce the slightly smaller A∥ value of Cu-V (DFT-I, DFT-II) compared to the two hydroxo complexes.

The DFT calculated 1H and 17O ligand hyperfine couplings are in relatively good agreement with the experimentally determined values and trends observed (Table 4). The axial water molecules in DFT-II have rather small, axial 1H hyperfine interactions of A(1H) ∼ [+6, −3, −3] MHz and 17O couplings of 1–3 MHz, which reproduces the experimental observations of all examined complexes quite well, further supporting the observation of the axially coordinated water by 1H ENDOR and 17O EDNMR spectroscopy, respectively.

For the equatorial water couplings in DFT-I and DFT-II, larger, rhombic 1H couplings are calculated, as seen in the 1H ENDOR experiments above. The corresponding 17O hyperfine tensors have isotropic components of −40 to −46 MHz and dipolar couplings of t(17O) ∼ −9 MHz. For the two hydroxo complexes DFT-III and DFT-IVaiso(17O) is found to be in the same range. Their dipolar components, however, are calculated to be significantly higher with t(17O) ∼ −20 MHz for DFT-III and t(17O) ∼ −14 MHz for DFT-IV. This is a significant overestimation compared to their experimental values (Cu-III: t(17O) ∼ −14 MHz; Cu-IV: t(17O) ∼ −12 MHz), especially for DFT-III. While the shape of the corresponding 1H couplings of the hydroxo ligands is overall quite well reproduced, the couplings are again overestimated (ADFT-III(1H) = [−22.3, −10.8, +7.4] MHz, ADFT-IV(1H) ∼ [−16.4, −7.2, +7.8] MHz) compared to the experimental values of Cu-III ([−16.5, −8.0, +7.5] MHz) and Cu-IV ([−13.0, −5.0, +8.5] MHz).

Differences between the hyperfine spectroscopic footprints of the individual complexes can be explained by changes in the spin densities of the ligands. The DFT-calculated Mulliken spin populations of all complexes are listed in Table 4. Between DFT-I and DFT-II the spin populations on the oxygen (ρsum(17O) ∼ 0.1/0.09) and also nitrogen (ρsum(14N) ∼ 0.22/0.23) nuclei stay constant, as the additional axial water molecules do not bear a significant amount of spin themselves or significantly influence the existing spin distribution on the rest of the complex. However, as the bis-aqua complex is converted into the bis-hydroxo complex (DFT-III), significantly more spin density is found on the OH− ligands (ρsum(17O) ∼ 0.26), while the contrary observation can be made for the bpy ligand (ρsum(14N) ∼ 0.14). This leads to a larger anisotropic hyperfine interaction for the 17O nuclei for DFT-III (t ∼ −20 MHz) compared to DFT-I (t ∼ −9 MHz). The effects of the OH− ligands in DFT-III slightly outweigh the effects of the bpy ligand, leading to an overall reduced spin density on the copper center (DFT-I: 0.67; DFT-III: 0.60). The tetra-hydroxo model DFT-IV bears approximately the same spin population on the copper center (0.62), and also on the ligands. However, the distribution of the spin density over four instead of two OH− ligands significantly reduces the density per ligand, and thereby also the hyperfine coupling (t ∼ −14/−15 MHz).

The relatively large difference in the 17O ligand hyperfine couplings between Cu-III and Cu-IV inspired us to further explore the variability of this property, including the effects of ligand charge. Hence, we optimized the geometries of several square planar copper complexes with the neutral tetrahydrofuran (thf) ligand in various thf[thin space (1/6-em)]:[thin space (1/6-em)]HxO ratios (complexes thf-I–thf-X, shown in Chart 2). The aquo complexes thf-I–thf-IV with n = 1 to 3 water and 4-n thf molecules exhibit highly similar 17O hyperfine couplings with isotropic components between −47 and −49 MHz and anisotropic components ranging between −10 and −11 MHz (Table 5). For the hydroxo complexes thf-V–thf-VIII, on the other hand, significant variations in the 17O hyperfine are observed with aiso values between −18 and −51 MHz and t values between −10 and −25 MHz. It is noticeable that the 17O couplings of OH− ligands, that have a second OH− ligand in trans position exhibit significantly smaller anisotropic hyperfine components (t ∼ −10 to −12 MHz), approximately half of the couplings found for OH− ligands positioned trans to a thf ligand (t ∼ −21 to −25 MHz).


image file: d4dt02708f-c2.tif
Chart 2 Overview of the DFT calculated structures thf-I–thf-X.
Table 5 Calculated 17O hyperfine couplings of the HxO ligands in complexes thf-I–thf-X, together with the optimized Cu–O bond distances
  A = [A1, A2, A3] in MHz a iso in MHz t in MHz Cu–O distance in Å Mulliken spin density
thf-I [−37, −67, −36] −47 −10 1.991 0.06
thf-II [−38, −69, −38] −49 −10 1.997 0.06
[−68, −36, −36] −47 −10 1.998 0.06
thf-III [−36, −68, −35] −47 −11 1.985 0.06
[−36, −68, −35] −47 −11 1.988 0.06
thf-IV 1-O: [−68, −36, −37] 1-O: −48 1-O: −10 1.986 0.06
2-O: [−35, −71, −35] 2-O: −48 2-O: −11 1.986 0.07
3-O: [−39, −69, −37] 3-O: −49 3-O: −10 1.983 0.06
thf-V [−71, −5, −3] −28 −21 1.847 0.15
thf-VI [−8, −76, −9] −32 −22 1.876 0.15
[−10, −81, −10] −35 −23 1.882 0.15
thf-VII [−41, −4, −5] −18 −12 1.859 0.08
[−41, −4, −5] −18 −12 1.859 0.08
thf-VIII 1-O: [−23, −102, −24] 1-O: −51 1-O: −25 1.940 0.16
2-O: [−52, −19, −20] 2-O: −32 2-O: −10 1.896 0.07
3-O: [−51, −18, −18] 3-O: −30 3-O: −10 1.891 0.07
thf-IX 1-O: [−68, −1, 0] 1-O: −26 1-O: −21 1.843 0.15
2-O: [−51, −83, −52] 2-O: −62 2-O: −10 2.045 0.05
thf-X 1-O: [−73, −3, −5] 1-O: −29 1-O: −22 1.842 0.15
2-O: [−41, −22, −23] 2-O: −29 2-O: −6 2.017 0.03

The mixed H2O/OH− complexes thf-IX and thf-X with two thf molecules and the HxO ligands either in cis (thf-IX) or trans (thf-X) positions to each other (Chart 2) were examined as well. The OH− ligands of thf-IX and thf-X show comparable hyperfine couplings to the ones determined for the trans-thf positioned OH− couplings in thf-V–thf-VIII. The 17O coupling of the H2O ligand in thf-IX also exhibits couplings comparable to the ones found for thf-I–thf-IV. In thf-X, however, where the H2O ligand is positioned trans to the OH− ligand, the isotropic and anisotropic components are again approximately halved compared to the one found in thf-IX. This again indicates that the OH− ligand has a major influence on ligands that are trans-positioned, as their spin density is approximately halved (Table 5).

A potential reason for variations in the hyperfine couplings is the distortion from an ideal square planar geometry, leading to a reduced overlap of the O 2p orbitals with the dx2−y2 SOMO. To examine this influence, the optimized structure of DFT-IV was stepwise distorted by simultaneously constraining the O1–Cu–O2 and O3–Cu–O4 (O–Cu–Ocis) angles from 90° to 109.5° and reoptimizing the rest of the structure (Fig. S41a†). The subsequently calculated 17O hyperfine couplings are listed in Table S17,† showing only minor changes in t, while aiso is reduced from −44 MHz (O–Cu–Ocis: 90°, O–Cu–Otrans: 133°) to −25 MHz (O–Cu–Ocis: 109.5°, O–Cu–Otrans: 133°) (Fig. S41b†). This can in part be the reason for variations in the isotropic 17O couplings of thf-I–thf-X; however, the observed distortions in these complexes are rather small with O–Cu–Otrans angles of between 169° and 179°, suggesting that distortions from square planar have only minor effects on hyperfine couplings. In particular, the minimal influence of the distortion on t shows, that the 17O hyperfine variations found in thf-I–thf-X are better explained by the electronic effects of trans positioned hydroxide than geometry distortions.

Altogether, we conclude that the spectroscopic “footprints” of HxO ligands do not only depend on their inherent nature (OH−vs. H2O), but are also heavily influenced by other ligands trans to them, with a clear reduction of its anisotropic components when the trans positioned ligand is another hydroxo ligand.

Conclusion

The ability to accurately identify the coordination sphere of a catalytic metal center is essential for understanding and assigning mechanisms. Advanced EPR techniques, such as ENDOR and EDNMR spectroscopy (both employed here) and other hyperfine approaches such as electron stimulated echo envelope modulation (ESEEM) and hyperfine sublevel correlation (HYSCORE) spectroscopy have unique abilities to characterize frozen, randomly oriented (powder like) paramagnetic catalytic centers with high resolution. The extraction of structural information and accurate ligand identification requires both clear standards and an understanding of various influential factors of hyperfine couplings.

We have examined aqueous copper bpy solutions between pH 7 and pH 14 by employing UV-Vis and EPR spectroscopy. In solutions with added sodium acetate, [Cu(bpy)(OAc)(H2O)2,ax]+ (Cu-I) forms at neutral pH and no waters coordinate in the equatorial positions. Using glycerol instead of sodium acetate as a glassing agent for EPR spectroscopy generates the acetate-free [Cu(bpy)(H2O)2,eq(H2O)2,ax]2+ (Cu-V). This in part demonstrates the powerful ability of EPR and advanced EPR spectroscopy to confirm the coordination environment of a catalytic center under conditions used in catalysis. At pH 12, the signal is completely recovered and the only detectable species is the bis-hydroxo complex [Cu(bpy)(OH)2,eq(HxO)2,ax] (Cu-III). At even further elevated pH, the complex loses the bpy ligand and the tetra-hydroxo complex [Cu(OH)4,eq(HxO)2,ax]2− (Cu-IV) emerges.

1H and 17O ENDOR spectra of the various copper complexes studied here reveal both clear trends and ambiguity in the differentiation of equatorial vs. axial waters, and differentiation difficulties between water and hydroxo ligands. The equatorial HxO ligand shows relatively large and rhombic 1H couplings, whereas the couplings are generally larger for the OH− compared to the H2O ligands. Their 17O nuclei exhibit axial hyperfine couplings, with the maximum hyperfine interaction directed along the Cu–O bonds. Analogous to the proton couplings, the 17O nuclei of the hydroxo ligands in Cu-III and Cu-IV exhibit larger dipolar couplings than the coordinating water molecules in Cu-V. Additionally, in between Cu-III and Cu-IV the 1H and 17O couplings show clear differences, with generally larger couplings for Cu-III. We showed computationally that the trans positioned ligand heavily influences the anisotropic component of the 17O hyperfine couplings with a clear reduction in t(17O) when a OH− is at the trans position. The observations and influences identified here are generally applicable to the study of copper WOC systems by EPR spectroscopy, but also more broadly to other monocopper active sites, such as LPMOs, where the trans coordinating N-terminal amino group ligand is thought to tune reactivity.

Author contributions

JH performed all EPR, ENDOR, EDNMR and UV-Vis experiments and analysis under the supervision of GEC. JH wrote the original draft with contributions from GEC. GEC conceptualized the experiments. GEC acquired funding.

Data availability

All data supporting the findings and conclusions are available in the ESI.†

Conflicts of interest

There are no conflicts of interest to declare.

Acknowledgements

The Max Planck Society is acknowledged for generous financial support.

References

  1. J. P. Klinman, The Copper-Enzyme Family of Dopamine β-Monooxygenase and Peptidylglycine α-Hydroxylating Monooxygenase: Resolving the Chemical Pathway for Substrate Hydroxylation, J. Biol. Chem., 2006, 281(6), 3013–3016,  DOI:10.1074/jbc.r500011200.
  2. G. Vaaje-Kolstad, B. Westereng, S. J. Horn, Z. Liu, H. Zhai, M. Sørlie and V. G. H. Eijsink, An Oxidative Enzyme Boosting the Enzymatic Conversion of Recalcitrant Polysaccharides, Science, 2010, 330(6001), 219–222,  DOI:10.1126/science.1192231.
  3. R. J. Quinlan, M. D. Sweeney, L. Lo Leggio, H. Otten, J. C. N. Poulsen, K. S. Johansen, K. B. R. M. Krogh, C. I. Jørgensen, M. Tovborg, A. Anthonsen, T. Tryfona, C. P. Walter, P. Dupree, F. Xu, G. J. Davies and P. H. Walton, Insights into the Oxidative Degradation of Cellulose by a Copper Metalloenzyme That Exploits Biomass Components, Proc. Natl. Acad. Sci. U. S. A., 2011, 108(37), 15079–15084,  DOI:10.1073/pnas.1105776108.
  4. M. O. Ross, F. MacMillan, J. Wang, A. Nisthal, T. J. Lawton, B. D. Olafson, S. L. Mayo, A. C. Rosenzweig and B. M. Hoffman, Particulate Methane Monooxygenase Contains Only Mononuclear Copper Centers, Science, 2019, 364(6440), 566–570,  DOI:10.2210/pdb3rgb/pdb.
  5. A. E. Wendlandt, A. M. Suess and S. S. Stahl, Copper-Catalyzed Aerobic Oxidative C-H Functionalizations: Trends and Mechanistic Insights, Angew. Chem., Int. Ed., 2011, 50(47), 11062–11087,  DOI:10.1002/anie.201103945.
  6. P. Vanelderen, J. Vancauwenbergh, B. F. Sels and R. A. Schoonheydt, Coordination Chemistry and Reactivity of Copper in Zeolites, Coord. Chem. Rev., 2013, 257(2), 483–494,  DOI:10.1016/j.ccr.2012.07.008.
  7. J. Aguila-Rosas, D. Ramos, C. T. Quirino-Barreda, J. A. Flores-Aguilar, J. L. Obeso, A. Guzmán-Vargas, I. A. Ibarra and E. Lima, Copper(II)-MOFs for Bio-Applications, Chem. Commun., 2023, 59(79), 11753–11766,  10.1039/d3cc03146b.
  8. X. Gao, H. Wang and J. Wang, Functional Metal-Organic Framework – a Review of Research on Cu-MOFs, J. Coord. Chem., 2024, 77, 1307–1323,  DOI:10.1080/00958972.2024.2374881.
  9. H. Lee, X. Wu and L. Sun, Copper-Based Homogeneous and Heterogeneous Catalysts for Electrochemical Water Oxidation, Nanoscale, 2020, 12(7), 4187–4218,  10.1039/c9nr10437b.
  10. S. M. Barnett, K. I. Goldberg and J. M. Mayer, A Soluble Copper-Bipyridine Water-Oxidation Electrocatalyst, Nat. Chem., 2012, 4(6), 498–502,  DOI:10.1038/nchem.1350.
  11. M. K. Coggins, M. T. Zhang, Z. Chen, N. Song and T. J. Meyer, Single-Site Copper(II) Water Oxidation Electrocatalysis: Rate Enhancements with HPO42- as a Proton Acceptor at PH 8, Angew. Chem., Int. Ed., 2014, 53(45), 12226–12230,  DOI:10.1002/anie.201407131.
  12. F. Chen, N. Wang, H. Lei, D. Guo, H. Liu, Z. Zhang, W. Zhang, W. Lai and R. Cao, Electrocatalytic Water Oxidation by a Water-Soluble Copper(II) Complex with a Copper-Bound Carbonate Group Acting as a Potential Proton Shuttle, Inorg. Chem., 2017, 56(21), 13368–13375,  DOI:10.1021/acs.inorgchem.7b02125.
  13. T. Zhang, C. Wang, S. Liu, J.-L. Wang and W. Lin, A Biomimetic Copper Water Oxidation Catalyst with Low Overpotential, J. Am. Chem. Soc., 2014, 136(1), 273–281,  DOI:10.1021/ja409267p.
  14. L. Z. Fu, T. Fang, L. L. Zhou and S. Z. Zhan, A Mononuclear Copper Electrocatalyst for Both Water Reduction and Oxidation, RSC Adv., 2014, 4(96), 53674–53680,  10.1039/c4ra07211a.
  15. M. Gil-Sepulcre, P. Garrido-Barros, J. Oldengott, I. Funes-Ardoiz, R. Bofill, X. Sala, J. Benet-Buchholz and A. Llobet, Consecutive Ligand–based Electron Transfer in New Molecular Cu–based Water Oxidation Catalysts, Angew. Chem., Int. Ed., 2021, 60, 18639–18644,  DOI:10.1002/anie.202104020.
  16. S. J. Koepke, K. M. Light, P. E. Vannatta, K. M. Wiley and M. T. Kieber-Emmons, Electrocatalytic Water Oxidation by a Homogeneous Copper Catalyst Disfavors Single-Site Mechanisms, J. Am. Chem. Soc., 2017, 139(25), 8586–8600,  DOI:10.1021/JACS.7B03278.
  17. D. L. Gerlach, S. Bhagan, A. A. Cruce, D. B. Burks, I. Nieto, H. T. Truong, S. P. Kelley, C. J. Herbst-Gervasoni, K. L. Jernigan, M. K. Bowman, S. Pan, M. Zeller and E. T. Papish, Studies of the Pathways Open to Copper Water Oxidation Catalysts Containing Proximal Hydroxy Groups during Basic Electrocatalysis, Inorg. Chem., 2014, 53(24), 12689–12698,  DOI:10.1021/ic501018a.
  18. P. Garrido-Barros, I. Funes-Ardoiz, S. Drouet, J. Benet-Buchholz, F. Maseras and A. Llobet, Redox Non-Innocent Ligand Controls Water Oxidation Overpotential in a New Family of Mononuclear Cu-Based Efficient Catalysts, J. Am. Chem. Soc., 2015, 137(21), 6758–6761,  DOI:10.1021/jacs.5b03977.
  19. Q.-Y. Mao, Y.-J. Pang, X.-C. Li, G.-J. Chen and H.-W. Tan, Theoretical Study of the Mechanisms of Two Copper Water Oxidation Electrocatalysts with Bipyridine Ligands, ACS Catal., 2019, 9, 8798–8809,  DOI:10.1021/acscatal.9b01022.
  20. M. S. Ali Akbari, S. Nandy, K. H. Chae, R. Bikas, A. Kozakiewicz-Piekarz and M. M. Najafpour, Water Oxidation by a Copper(II) Complex with 6,6′-Dihydroxy-2,2′-Bipyridine Ligand: Challenges and an Alternative Mechanism, Langmuir, 2023, 39(15), 5542–5553,  DOI:10.1021/acs.langmuir.3c00322.
  21. T. Ghosh, P. Ghosh and G. Maayan, A Copper-Peptoid as a Highly Stable, Efficient, and Reusable Homogeneous Water Oxidation Electrocatalyst, ACS Catal., 2018, 8(11), 10631–10640,  DOI:10.1021/acscatal.8b03661.
  22. E. Garribba, G. Micera, D. Sanna and L. Strinna-Erre, The Cu(II)-2,2-Bipyridine System Revisited, Inorg. Chim. Acta, 2000, 299, 253–261,  DOI:10.1016/s0020-1693(99)00508-3.
  23. N. M. Atherton and A. J. Horsewill, Proton ENDOR of Cu(H2O)62+ in Mg(NH4)2(SO2)4·6H2O, Mol. Phys., 1979, 37(5), 1349–1361,  DOI:10.1080/00268977900100991.
  24. D. Kim, N. H. Kim and S. H. Kim, 34 GHz Pulsed ENDOR Characterization of the Copper Coordination of an Amyloid β Peptide Relevant to Alzheimer's Disease, Angew. Chem., Int. Ed., 2013, 52(4), 1139–1142,  DOI:10.1002/anie.201208108.
  25. C. S. Burns, E. Aronoff-Spencer, C. M. Dunham, P. Lario, N. I. Avdievich, W. E. Antholine, M. M. Olmstead, A. Vrielink, G. J. Gerfen, J. Peisach, W. G. Scott and G. L. Millhauser, Molecular Features of the Copper Binding Sites in the Octarepeat Domain of the Prion Protein, Biochemistry, 2002, 41(12), 3991–4001,  DOI:10.1021/bi011922x.
  26. A. Pöppl and L. Kevan, A Practical Strategy for Determination of Proton Hyperfine Interaction Parameters in Paramagnetic Transition Metal Ion Complexes by Two-Dimensional HYSCORE Electron Spin Resonance Spectroscopy in Disordered Systems, J. Phys. Chem., 1996, 100(9), 3387–3394,  DOI:10.1021/jp9525999.
  27. R. J. Jodts, M. O. Ross, C. W. Koo, P. E. Doan, A. C. Rosenzweig and B. M. Hoffman, Coordination of the Copper Centers in Particulate Methane Monooxygenase: Comparison between Methanotrophs and Characterization of the CuC Site by EPR and ENDOR Spectroscopies, J. Am. Chem. Soc., 2021, 143(37), 15358–15368,  DOI:10.1021/jacs.1c07018.
  28. J. Haak, O. Golten, M. Sørlie, V. Eijsink and G. Cutsail III, pH-Mediated Manipulation of the Histidine Brace in LPMOs and Generation of a Tri-Anionic Variant, Investigated by EPR, ENDOR, ESEEM and HYSCORE Spectroscopy, Chem. Sci., 2024 10.1039/d4sc04794j.
  29. M. E. Owen, E. Carter, G. J. Hutchings, B. D. Ward and D. M. Murphy, Influence of Counterions on the Structure of Bis(Oxazoline)Copper(II) Complexes; an EPR and ENDOR Investigation, Dalton Trans., 2012, 41(36), 11085–11092,  10.1039/c2dt31273e.
  30. P. C. Bruzzese, E. Salvadori, S. Jäger, M. Hartmann, B. Civalleri, A. Pöppl and M. Chiesa, 17O-EPR Determination of the Structure and Dynamics of Copper Single-Metal Sites in Zeolites, Nat. Commun., 2021, 12(1), 4638,  DOI:10.1038/s41467-021-24935-7.
  31. Y. K. Liao, P. C. Bruzzese, E. Salvadori and M. Chiesa, 17O Hyperfine Spectroscopy in Surface Chemistry and Catalysis, J. Magn. Reson. Open, 2023, 16–17, 100101,  DOI:10.1016/J.JMRO.2023.100101.
  32. A. Actis, E. Salvadori and M. Chiesa, Framework Coordination of Single-Ion Cu2+ Sites in Hydrated 17O-ZSM-5 Zeolite, Catal. Sci. Technol., 2021, 11(15), 5191–5199,  10.1039/d1cy00838b.
  33. D. Getz and B. L. Silver, ESR of Cu2+(H2O)6. I. The Oxygen-17 Superhyperfine Tensors in 63Cu2+ Doped Zinc Tutton's Salt at 20°K, J. Chem. Phys., 1974, 61(2), 630–637,  DOI:10.1063/1.1681939.
  34. B. L. Silver, D. Getz and Z. Luz, 17O Superhyperfine Interaction in the ESR Spectrum of Hydrated Cu2+ Ion, Chem. Phys. Lett., 1972, 13(2), 113–114,  DOI:10.1016/s0009-2614(72)90001-2.
  35. T. Wagner-Jauregg, B. E. Hackley, J. T. A. Lies, O. O. Owens and R. Proper, Model Reactions of Phosphorous-Containing Enzyme Inactivators. IV. The Catalytic Activity of Certain Metal Salts and Chelates in the Hydrolysis of Diisopropyl Fluorophosphate, J. Am. Chem. Soc., 1955, 77(4), 922–929,  DOI:10.1021/ja01609a037.
  36. A. E. Martell, S. Chaberek, J. R. C. Courtney, S. Westerback and H. Hyytiainen, Hydrolytic Tendencies of Metal Chelate Compounds. I. Cu(II) Chelates, J. Am. Chem. Soc., 1957, 45, 3036–3041,  DOI:10.1021/ja01512a005.
  37. L. B. Ryland, G. S. Ronay and F. M. Fowkes, Equilibria in Aqueous Solutions of Copper(II) Chelates with Alpha,Alpha’-Dipyridyl, o-Phenanthroline and Ethylenediamine, J. Phys. Chem., 1958, 62, 798–801,  DOI:10.1021/j150565a006.
  38. R. L. Gustafson and A. E. Martell, Hydrolytic Tendencies of Metal Chelate Compounds. V. Hydrolysis and Dimerization of Copper(II) Chelates of 1,2-Diamines, J. Am. Chem. Soc., 1959, 81(3), 525–529,  DOI:10.1021/ja01512a005.
  39. D. D. Perrin and V. S. Sharma, Complex Formation by Hydrolysed Copper(II) Ions in Aqueous Solution, J. Inorg. Nucl. Chem., 1966, 28, 1271–1278,  DOI:10.1016/0022-1902(66)80454-2.
  40. M. Noack and G. Gordon, Oxygen-17 NMR and Copper EPR Linewidths in Aqueous Solutions of Copper(II) Ion and 2,2′-Dipyridine, J. Chem. Phys., 1968, 48(6), 2689–2699,  DOI:10.1063/1.1669503.
  41. F. A. Walker, H. Sigel and D. B. McCormick, Spectral Properties of Mixed-Ligand Copper(II) Complexes and Their Corresponding Binary Parent Complexes, Inorg. Chem., 1972, 11(11), 2756–2763,  DOI:10.1021/ic50117a036.
  42. H. Sigel, What Is the Reason for the Cis Configuration of Diaquo-Cu2+-(2,2′-Bipyridyl)2 in Aqueous Solution?, Inorg. Chim. Acta, 1972, 6(2), 195,  DOI:10.1016/s0020-1693(00)91785-7.
  43. G. Arena, R. Calì, E. Rizzarelli and S. Sammartano, Thermodynamics of Copper(II) 2,2′-Dipyridyl Complexes in Aqueous Solution, Thermochim. Acta, 1976, 17(2), 155–164,  DOI:10.1016/0040-6031(76)85022-8.
  44. I. N. Marov, V. K. Belyaeva, E. B. Smirnova and I. F. Dolmanova, Electron Spin Resonance Spectra of Copper(II) Bipyridyl Complexes, Inorg. Chem., 1978, 17(6), 1667–1669,  DOI:10.1021/ic50184a058.
  45. L. Banci, A. Bencini, D. Gatteschi and C. Zanchini, EPR Spectra of Triplet States with Large Zero-Field Splitting. [Cu(Bipy)(OH)]2SO4·4H2O, J. Magn. Reson., 1982, 48, 9–19,  DOI:10.1016/0022-2364(82)90233-5.
  46. L. Banci, A. Bencini and D. Gatteschi, Correlation between Anisotropic Exchange and Structure of Di-Mu-Hydroxo Bridged Copper(II) Complexes, J. Am. Chem. Soc., 1983, 105(4), 761–764,  DOI:10.1021/ja00342a017.
  47. I. Fábián, Hydrolytic Reactions of Copper(II) Bipyridine Complexes, Inorg. Chem., 1989, 28(20), 3805–3807,  DOI:10.1021/ic00319a011.
  48. M. Judd, G. Jolley, D. Suter, N. Cox and A. Savitsky, Dielectric Coupler for General Purpose Q-Band EPR Cavity, Appl. Magn. Reson., 2022, 53(7–9), 963–977,  DOI:10.1007/s00723-021-01404-4.
  49. E. Reijerse, F. Lendzian, R. Isaacson and W. Lubitz, A Tunable General Purpose Q-Band Resonator for CW and Pulse EPR/ENDOR Experiments with Large Sample Access and Optical Excitation, J. Magn. Reson., 2012, 214, 237–243,  DOI:10.1016/j.jmr.2011.11.011.
  50. E. R. Davies, A New Pulse Technique, Phys. Lett., 1974, 47A(1), 1–2,  DOI:10.1016/0375-9601(74)90078-4.
  51. B. Epel, D. Arieli, D. Baute and D. Goldfarb, Improving W-Band Pulsed ENDOR Sensitivity-Random Acquisition and Pulsed Special TRIPLE, J. Magn. Reson., 2003, 164, 78–83,  DOI:10.1016/s1090-7807(03)00191-5.
  52. P. Schosseler, T. Wacker and A. Schweiger, Pulsed ELDOR Detected NMR, Chem. Phys. Lett., 1994, 224, 19–324,  DOI:10.1016/0009-2614(94)00548-6.
  53. S. Stoll and A. Schweiger, EasySpin, a Comprehensive Software Package for Spectral Simulation and Analysis in EPR, J. Magn. Reson., 2006, 178, 42–55,  DOI:10.1016/j.jmr.2005.08.013.
  54. N. Wili, S. Richert, B. Limburg, S. J. Clarke, H. L. Anderson, C. R. Timmel and G. Jeschke, ELDOR-Detected NMR beyond Hyperfine Couplings: A Case Study with Cu(II)-Porphyrin Dimers, Phys. Chem. Chem. Phys., 2019, 21(22), 11676–11688,  10.1039/c9cp01760g.
  55. F. Neese and J. Wiley, The ORCA Program System, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2012, 2(1), 73–78,  DOI:10.1002/wcms.81.
  56. F. Neese, Software Update: The ORCA Program System—Version 5.0, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2022, 12(5), e1606,  DOI:10.1002/wcms.1606.
  57. S. Grimme, J. Antony, S. Ehrlich and H. Krieg, A Consistent and Accurate Ab Initio Parametrization of Density Functional Dispersion Correction (DFT-D) for the 94 Elements H-Pu, J. Chem. Phys., 2010, 132(15), 154104,  DOI:10.1063/1.3382344/926936.
  58. S. Grimme, S. Ehrlich and L. Goerigk, Effect of the Damping Function in Dispersion Corrected Density Functional Theory, J. Comput. Chem., 2011, 32(7), 1456–1465,  DOI:10.1002/jcc.21759.
  59. M. Cossi, N. Rega, G. Scalmani and V. Barone, Energies, Structures, and Electronic Properties of Molecules in Solution with the C-PCM Solvation Model, J. Comput. Chem., 2003, 24(6), 669–681,  DOI:10.1002/jcc.10189.
  60. E. Van Lenthe, J. G. Snijders and E. J. Baerends, The Zero–order Regular Approximation for Relativistic Effects: The Effect of Spin–Orbit Coupling in Closed Shell Molecules, J. Chem. Phys., 1996, 105(15), 6505–6516,  DOI:10.1063/1.472460.
  61. E. Van Lenthe, E. J. Baerends and J. G. Snijders, Relativistic Total Energy Using Regular Approximations, J. Chem. Phys., 1994, 101(11), 9783–9792,  DOI:10.1063/1.467943.
  62. E. Van Lenthe, E. J. Baerends and J. G. Snijders, Relativistic Regular Two–component Hamiltonians, J. Chem. Phys., 1993, 99(6), 4597–4610,  DOI:10.1063/1.466059.
  63. J. P. Perdew, Density-Functional Approximation for the Correlation Energy of the Inhomogeneous Electron Gas, Phys. Rev. B: Condens. Matter Mater. Phys., 1986, 33(12), 8822–8824,  DOI:10.1103/physrevb.33.8822.
  64. A. D. Becke, Density-Functional Exchange-Energy Approximation with Correct Asymptotic Behavior, Phys. Rev. A, 1988, 38(6), 3098–3100,  DOI:10.1103/physreva.38.3098.
  65. A. D. Becke, Density–functional Thermochemistry. III. The Role of Exact Exchange, J. Chem. Phys., 1993, 98(7), 5648–5652,  DOI:10.1063/1.464913.
  66. C. Lee, W. Yang and R. G. Parr, Development of the Colle-Salvetti Correlation-Energy Formula into a Functional of the Electron Density, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37(2), 785,  DOI:10.1103/physrevb.37.785.
  67. F. Weigend, Accurate Coulomb-Fitting Basis Sets for H to Rn, Phys. Chem. Chem. Phys., 2006, 8(9), 1057–1065,  10.1039/b515623h.
  68. F. Weigend and R. Ahlrichs, Balanced Basis Sets of Split Valence, Triple Zeta Valence and Quadruple Zeta Valence Quality for H to Rn: Design and Assessment of Accuracy, Phys. Chem. Chem. Phys., 2005, 7(18), 3297–3305,  10.1039/b508541a.
  69. V. Barone, Structure, Magnetic Properties and Reactivities of Open-Shell Species From Density Functional and Self-Consistent Hybrid Methods, in Recent Advances in Density Functional Methods, ed. D. P. Chong, World Scientific, Singapore, 1995, pp. 287–334.  DOI:10.1142/9789812830586_0008.
  70. N. Rega, M. Cossi and V. Barone, Development and Validation of Reliable Quantum Mechanical Approaches for the Study of Free Radicals in Solution, J. Chem. Phys., 1996, 105(24), 11060–11067,  DOI:10.1063/1.472906.
  71. N. B. Balabanov and K. A. Peterson, Systematically Convergent Basis Sets for Transition Metals. I. All-Electron Correlation Consistent Basis Sets for the 3d Elements Sc-Zn, J. Chem. Phys., 2005, 123(6), 064107,  DOI:10.1063/1.1998907/932754.
  72. E. D. Hedegård, J. Kongsted and S. P. A. Sauer, Optimized Basis Sets for Calculation of Electron Paramagnetic Resonance Hyperfine Coupling Constants: Aug-Cc-PVTZ-J for the 3d Atoms Sc-Zn, J. Chem. Theory Comput., 2011, 7(12), 4077–4087,  DOI:10.1021/ct200587k.
  73. G. L. Stoychev, A. A. Auer and F. Neese, Automatic Generation of Auxiliary Basis Sets, J. Chem. Theory Comput., 2017, 13(2), 554–562,  DOI:10.1021/acs.jctc.6b01041.
  74. S. Barnett, Development and Mechanistic Study of Copper-Based Water Oxidation Catalysts, Ph. D. Dissertation, University of Washington, Seattle, WA, 2013. https://digital.lib.washington.edu/researchworks/items/77d25f7b-15b9-4efc-8727-4dce8d5558d3 (accessed 21.09.2024) Search PubMed.
  75. Y. Y. H. Chao and D. R. Kearns, Electron Spin Resonance Investigation of the Soluble Blue Copper(II) Hydroxide Complex, J. Phys. Chem., 1977, 81(7), 666–668,  DOI:10.1021/j100522a016.
  76. P. C. Bruzzese, E. Salvadori, B. Civalleri, S. Jäger, M. Hartmann, A. Pöppl and M. Chiesa, The Structure of Monomeric Hydroxo-CuII Species in Cu-CHA. A Quantitative Assessment, J. Am. Chem. Soc., 2022, 144(29), 13079–13083,  DOI:10.1021/jacs.2c06037.
  77. M. M. Werst, C. E. Davoust and B. M. Hoffman, Ligand Spin Densities in Blue Copper Proteins by Q-Band 1H and 14N ENDOR Spectroscopy, J. Am. Chem. Soc., 1991, 113(5), 1533–1538,  DOI:10.1021/ja00005a011.
  78. J. E. Roberts, J. F. Cline, V. Lum, H. Freeman, H. B. Gray, J. Peisach, B. Reinhammar and B. Hoffman, Comperative ENDOR Study of Six Blue Copper Proteins, J. Am. Chem. Soc., 1984, 106, 5324–5330,  DOI:10.1021/ja00330a048.
  79. J. A. J. Fitzpatrick, F. R. Manby and C. M. Western, The Interpretation of Molecular Magnetic Hyperfine Interactions, J. Chem. Phys., 2005, 122(8), 084312,  DOI:10.1063/1.1851501.
  80. N. Cox, W. Lubitz and A. Savitsky, W-Band, ELDOR-Detected NMR (EDNMR) Spectroscopy as a Versatile Technique for the Characterisation of Transition Metal-Ligand Interactions, Mol. Phys., 2013, 111(18–19), 2788–2808,  DOI:10.1080/00268976.2013.830783.
  81. N. Cox, W. Lubitz and A. Savitsky, W-Band, ELDOR-Detected NMR (EDNMR) Spectroscopy as a Versatile Technique for the Characterisation of Transition Metal-Ligand Interactions, Mol. Phys., 2013, 111(18–19), 2788–2808,  DOI:10.1080/00268976.2013.830783.
  82. A. R. Lorenz, J. Ammeter and H. H. Günthard, 17O Hyperfine Structure in ESR Spectra of Cu(II) Hydroxyquinolinate: Phthalimide, Chem. Phys. Lett., 1972, 16(2), 275–277 CrossRef CAS.
  83. A. R. Lorenz, J. H. Ammeter and H. H. Günthard, 17O Hyperfine Structure in Single Crystal ESR Spectra of Cu(II) Picolinate:Zn(II) Picolinate·4H2O, Chem. Phys. Lett., 1973, 23(4), 463–466 CrossRef CAS.

Footnote

† Electronic supplementary information (ESI) available: Additional experimental details including spectrometer conditions, UV-Vis spectra, cyclic voltammograms, multi-frequency EPR spectra, 14N ENDOR spectra, room temperature EPR pH titration, 1,2H ENDOR subtractions with simulations, 14N/17O ENDOR subtractions, EDNMR spectra, and computational details including coordinates of DFT optimized structures. Additional citations are made within the ESI.82,83 See DOI: https://doi.org/10.1039/d4dt02708f
This journal is © The Royal Society of Chemistry 2025
Click here to see how this site uses Cookies. View our privacy policy here.