Electron Paramagnetic Resonance Characterization of Dioxygen

Dec 5, 2016 - ... solids contain many distinct metal-ion sites, have a varying amount of defect sites within the lattice, and can be composed of multi...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/IC

Electron Paramagnetic Resonance Characterization of DioxygenBridged Cobalt Dimers with Relevance to Water Oxidation Troy A. Stich,† J. Gregory McAlpin,† Ryan M. Wall,† Matthew L. Rigsby,‡ and R. David Britt*,† †

Department of Chemistry, University of California, 1 Shields Avenue, Davis, California 95616-0935, United States Department of Chemistry, University of Wisconsin−Madison, 1101 University Avenue, Madison, Wisconsin 53706-1322, United States

‡

S Supporting Information *

ABSTRACT: A variety of metal oxides can catalyze the oxidation of water to molecular oxygen when polarized by a sufficiently high electrochemical potential. Minimizing the overpotential and increasing the rate of the oxygen-evolving reaction (OER) are key goals in making such materials a component of viable energy storage devices. However, the structural factors that imbue the metal oxides with their catalytic power are difficult to assess as these solids contain many distinct metal-ion sites, have a varying amount of defect sites within the lattice, and can be composed of multiple phases. In the present study, we determined the magnetic properties for a series of dimeric cobalt complexes in which the two metal centers are bridged by a dioxygen moiety. Our spectroscopically validated electronic structure description indicates that each species is best described as two Co(III) ions that are bound to a μ−η1η1 superoxide ligand. Intriguingly, we found evidence that the two compounds that possess oxygen-evolving activity coordinate the superoxide ion in an unusual, nonplanar fashion. It appears as if the intermediately long Co···Co distance of 3.9 Å is responsible for the unusual superoxide binding geometry. This structural factor may be an important element in the design of solid-state OER catalysts.

■

INTRODUCTION Cobalt oxides (CoOx) are promising materials for energy production, capable of oxidizing water to molecular oxygen (the so-called oxygen evolution reaction, OER) under mild conditions and modest potentials.1,2 The nature of the site(s) responsible for O−O bond formation are unknown; yet, identifying geometric and electronic structure factors that are involved in the splitting of water by CoOx is expected to aid in the design of better catalysts. Electron paramagnetic resonance (EPR) spectroscopy3,4 and X-ray absorption near edge spectroscopy (XANES)5 have detected the presence of Co(IV) sites in anodically deposited CoOx films when they are poised at potentials near those able to oxidize water. Results from electronic structure calculations of models of CoOx suggested that a peroxide-bridged Co(III) dimer moiety is generated as an intermediate in the OER.6−8 These findings motivated our investigation of a series of Co(III,III) bridging peroxo complexes for possible OER activity (complexes 1−4, Scheme 1).9 The bipyridine (bpy) and terpyridine (tpy) ligands are resistant to oxidation, allowing oxidizing equivalents to accumulate on the metal centers that also bind the substrate waters. In previous studies, all four species were found to produce O2; however, in the cases of complexes 1 and 2, water oxidation was found to be driven not by the ligand-supported metal complexes but by free cobalt ions liberated by © XXXX American Chemical Society

Scheme 1

Received: August 15, 2016

A

DOI: 10.1021/acs.inorgchem.6b01954 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Table 1. Magnetic Parameters for Cobalt−Oxygen (and Related) EPR-Active Species gx, gy, gz

complex O2−:MgO29 +

1 1+opt 2+ 2+opt 3+ 3+xtal of 3 3+opt 4+ 4+xtal of 4 4+opt [(NH3)4Rh-μ(OH,O2)Rh(NH3)4]4+,30−32 [Co2(en)4O2]5+ Co/MgO(O2−)57 CoPi3 [Co4O4(C5H5N)4(CH3CO2)4]+,55

2.0016, 2.0083, 2.0770 1.993, 2.017, 2.072 1.995, 2.010, 2.039 2.002, 2.016, 2.087 2.000, 2.014, 2.052 1.995, 2.002, 2.067 2.004, 2.050, 2.203 1.985, 2.033, 2.097 1.994, 2.000, 2.064 1.997, 2.075, 2.244 1.984, 2.036, 2.102 2.004, 2.020, 2.088 1.992, 2.020, 2.088 2.010, 2.023, 2.079 1.990, 2.033, 2.141 2.01, 2.19, 2.48 2.06, 2.32, 2.35

decomposition of the respective complexes.9 At this pH (2.1), CoOx cannot form; instead, free cobalt ions are able to catalyze O2 evolution.10 In contrast, the bispyridylpyrazolate (bpp) ligand employed in 3 and 4 was shown to provide a kinetically stable ligand framework that maintains the dimeric cobalt complex throughout the OER (aqueous solution, pH 1 triflic acid).9 X-ray crystal structures of 3 and 4 are consistent with a peroxide ligand bridging two low-spin (S = 0) Co(III) ions.9 A recent time-resolved Fourier transform infrared (FTIR) spectroscopy study that monitored water oxidation by Co3O4 identified a transient IR band at 1013 cm−1 and assigned it to a Co(III)-superoxo moiety that is possibly hydrogen-bonded to an adjacent Co(III)-hydroxo fragment.11 When we oxidized the peroxo-bridged dicobalt complex 3 by one electron using ceric ammonium nitrate (CAN), a continuous wave (CW) EPR spectrum develops that has hints of 59Co (I = 7/2, 100% natural abundance) hyperfine structure.9 The presence of this hyperfine structure could indicate formal oxidation of one of the metal centers to the 4+ oxidation state or could result from interaction of the magnetic 59Co nuclei with the unpaired electron borne by an adjacent superoxide moiety.12,13 In the present study, we compare the EPR properties of the four complexes illustrated in Scheme 1, each oxidized by one electron, to determine the proper electronic structure description. These findings are discussed in the context of the OER: in particular, we note geometrical features of 3+ and 4+ that may be important for enhancing product release.

■

Ax, Ay, Az 59Co(MHz) (avg) n/a 0, 10, −20 −8, −27, 42 max = 55 MHz 66, −42, 116 20, −30, 85 max = 55 MHz 66, −39, 114 19, −31, 89

42, 20, 93 −20, 77, −5

Oxford 935CF cryostat. Pulse X-band EPR studies were performed using a dielectric MD-5 resonator (Bruker). All CW EPR data were acquired under nonsaturating, slow-passage conditions. All simulations were performed using the EasySpin 4.0 toolbox15 in MatLab (The Mathworks Inc., Natick, MA). DFT Calculations. Coordinates from crystal structures were used as starting points for geometry optimizations carried out using Gaussian0916 using the TPSS functional.17 Magnetic properties were calculated in ORCA using the TPSSh18 functional, and scalar relativistic effects were included with the zeroth-order regular approximation (ZORA)19,20 along with the resolution of identity (RI) and chain-of-spheres (COSX) approximations to Coulomb and exact exchange.21 The conductor-like screening model (COSMO)22 was used to model the dielectric effects of solvent water on the system.

■

RESULTS AND DISCUSSION Dioxygen-bridged dinuclear metal complexes exhibit varied electronic structures depending on the geometry of the O2 bridge, the nature of the metal ions, and the overall oxidation state of the metal−oxygen core. Many dioxygen-bridged dicobalt complexes have been generated by treating a mononuclear or dinuclear Co(II) precursor with molecular oxygen to give a peroxo-bridged Co(III,III) dimer.23 Ligands that are strong σ-donors and poor electron acceptors are optimal for enhancing the binding of elemental O2 to cobaltous complexes.12 Several of these peroxo dimers have been chemically oxidized by one electron giving an EPR-active S = 1/2 species. The relevant magnetic parameters for a select number of these systems are given in Table 1. The consensus electronic structure description finds that the unpaired electron is borne mostly on the O2 moiety, yielding a superoxidebridged Co(III,III) dimer.12,13,24−26 Yet, OER activity has not been reported for any of these systems. The low-temperature X- and Q-band CW EPR spectra of the four Ce(IV)-oxidized peroxo-bridged Co(III,III) dimers studied here are relatively similar to one another (Figure 1). For the oxidized complexes 2+, 3+, and 4+, the Q-band (34 GHz) data reveal a near axial g-tensor centered at g ≈ 2 with modest ganisotropy (gmin − gmax < 0.09; see left side of Figure 1 and magnetic parameters listed in Table 1). For the oxidized complexes 3+ and 4+, the g-tensors are noticeably more axial, whereas the spectrum for 1+ appears to be more rhombic. This degree of magnetic axiality is unusual for metal-bound

EXPERIMENTAL SECTION

Sample Preparation. Solutions of peroxo-bridged cobalt dimers (1−4) were prepared as described previously.9 The one-electron oxidation of each system was performed by addition of an equimolar amount of ceric ammonium nitrate (CAN). Aliquots from each solution were added to 4 mm OD quartz tubes that were then immediately frozen and stored in liquid nitrogen. EPR Spectroscopy. X-band (9.33 GHz) CW EPR spectra were recorded using a Bruker (Billerica, MA) E-500 spectrometer equipped with a SHQE resonator. Cryogenic temperatures (10 K) were achieved and controlled using an Oxford Instruments ESR900 liquid helium cryostat in conjunction with an Oxford Instruments ITC503 temperature and gas flow controller. CW and pulse Q-band studies (ca. 34 GHz) at 5 K were performed using a Bruker E-580 EleXsys spectrometer using a laboratory-built probe14 modified to fit into an B

DOI: 10.1021/acs.inorgchem.6b01954 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

however, possesses a multitude of hyperfine splittings (Figure 1, right). Inspection of the numerical derivative of the CW spectrum (Figure S1) reveals that 15 of these lines centered at g = 1.9945(5) have regular spacings of approximately 1.8 mT (corresponding to a hyperfine coupling of ≈50 MHz) and appear in an intensity pattern of ≈1:2:3:4:5:6:7:8:7:6:5:4:3:2:1. These hyperfine lines are attributed to the unpaired electron spin interacting in near equivalent fashion with two 59Co nuclei. Similar features are evident in the X-band spectra of all four complexes (Figure 1, right). The relatively weak 59Co hyperfine interaction (HFI) that is observed has two main contributions: (i) through-space dipolar interaction of the unpaired electron within the 2pπgx orbital with each cobalt center and (ii) a modest polarization of the Co 3d orbitals induced by weak bonding (covalency) interactions with the superoxo group. The through-space contribution can be calculated from the geometry of the four-atom system. These atoms are too close to one another to accurately employ the point−point dipole approximation. Therefore, the unpaired electron in the 2pπgx orbital is split into four portions, each one placed at the center of one of the four lobes of this molecular orbital, 0.54 Å from the oxygen nucleus.27−29 The dipolar part of each 59Co HFI was determined by adding the four pairwise interactions with each lobe of the 2pπgx orbital. If the four atoms Co−OO−Co are coplanar, the lobes of the halfoccupied 2pπgx orbital lay in a plane orthogonal to the core, and the dipolar 59Co hyperfine is computed to be just 1.5 MHz. As the Co−OO−Co torsion angle (φ) deviates from planarity, the 2pπgx lobes come closer to the cobalt nuclei, and this through-space coupling increases to 4 MHz for φ = 45°. The covalency of the metal−oxygen bonds is expected to be small. This is particularly evident when comparing the g-values of a superoxo-bridged Co(III,III) dimer to those of an isoelectronic [Rh-μ(OH,O2)Rh]4+ analogue (Table 1).30−32 The spin−orbit coupling constant of rhodium is at least factor of 2 larger than that for cobalt, but the g-shifts of the two compounds are essentially identical, suggesting that little spin density lives on the metal center. Typically, the isotropic 59Co hyperfine (Aiso) coupling in superoxide-bridged Co(III,III) dimers is on the order of 25−40 MHz,24,33 which corresponds to only 0.4−0.7% cobalt 4s character.34 The largest element of the 59Co hyperfine tensor for 4+ is 55 MHz which, using the lower range of Aiso from above, corresponds to an anisotropic coupling of only 15 MHz or 6% cobalt 3d-character for the spin-carrying molecular orbital. Additionally, the X- and Q-band three-pulse ESEEM spectra of 4+ (Figure 2), which probe the hyperfine coupling of the ligating 14N atoms, reveal only modest coupling. The Fourier transform (Figure 2, top-right) of the X-band ESEEM spectra gives a peak at 1.5 MHz with a shoulder at 2.0 MHz assigned to two of the transitions expected for an 14N nucleus (I = 1) experiencing nearly equal nuclear Zeeman and hyperfine fields.35,36 At ≈347 mT, this “cancellation” behavior implies that the hyperfine interaction is Aiso(14N) ≈ 2 MHz. The frequencies of these features are largely independent of where in the EPR spectrum they are sampled, suggesting little orientation selection; however, the magnitude of the 59Co hyperfine coupling likely broadens the EPR envelope such that all orientations are probed independent of the field position at which the ESEEM experiment is conducted. At the 4-fold higher field of the Q-band ESEEM experiment, the g-tensor becomes more resolved (cf. left and right sides of Figure 1), yet the observed nuclear spin-flip transition frequencies are still

Figure 1. CW EPR spectra of complexes 1+, 2+, 3+, and 4+ acquired at Q-band (34.11 GHz, left panel) and X-band (9.33 GHz, right panel) excitation frequencies. Spectrometer settings, Q-band: temperature = 5.0 K; power = 12.5 μW; modulation amplitude = 0.8 mT; modulation frequency = 100 kHz; sweep rate = 1.43 mT/s. Spectrometer settings, X-band: temperature = 10.0 K; power = 10 μW; modulation amplitude = 0.1 mT; modulation frequency = 100 kHz; sweep rate = 0.75 mT/s.

superoxides, which typically display some rhombicity in the gtensor owing to two superoxide π* orbitals having different energies (vide infra, Scheme 2). Scheme 2

At Q-band, the expected 59Co hyperfine splitting is not observed owing to heterogeneity in the g-values, leading to field-dependent broadening of the EPR spectrum. However, the non-Gaussian line shape for the lowest field resonance (especially evident in the spectrum of 1+, Figure 1, top left) results from two nearly equivalent 59Co hyperfine couplings of no more than 30 MHz that manifests as a triangular shaped peak. The spectrum of compound 3+ acquired at X-band, C

DOI: 10.1021/acs.inorgchem.6b01954 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 2. Top: X-band ESE EPR (left panel) and orientation-selected three-pulse ESEEM spectra (center panel) of 4+. Corresponding cross-term averaged spectra are presented to the right. Spectrometer settings: microwave frequency = 9.76 GHz; temperature = 10 K; tπ/2 = 16 ns; τ = 144 ns; T0 = 68 ns; dT = 12 ns. Bottom: Q-band ESE EPR (left panel) and orientation-selected three-pulse ESEEM spectra (center panel) of 4+. Corresponding cross-term averaged spectra are presented to the right. Spectrometer settings: microwave frequency = 34.10 GHz; temperature = 10 K; tπ/2 = 48 ns; τ = 212 ns; T0 = 100 ns; dT = 28 ns.

rather insensitive to the field: two peaks at 2.2 and 5 MHz are visible in all data sets except those collected closer to gx,y. At this high-field extremum, no regular modulations are evident in the ESEEM spectrum. Weak 14N and 59Co hyperfine couplings support the assignment of species 1+−4+ as being formally superoxide radicals weakly interacting, in equivalent fashion, with two cobalt ions. Cohen and Känzig derived the following equations using perturbation theory to interpret the g-shifts in terms of a few molecular parameters of the superoxide moiety (Scheme 2):37 gz = ge + 2l[λ 2 /(λ 2 + Δ2 )]1/2

depend on E, the energy separation between the 2pπgx (πv*) and 2pσ orbitals. Performing the Cohen and Känzig analysis on the g-values of 1+ and 2+ gives typical values for E (cf. values for O2−:MgO, Table 2). Δ is seen to increase relative to that observed for the Table 2. Cohen and Känzig Parameters (eV)

(1)

g x = ge[λ 2 /(λ 2 + Δ2 )]1/2 − λ /E{1 − [λ 2 /(λ 2 + Δ2 )]1/2 − [Δ2 /(λ 2 + Δ2 )]1/2 }

λ

Δ

E

0.0115 0.0241 0.0239 0.0010 0.0008

0.3067 0.688 0.565 0.031 0.027

3.04 3.00 3.02 3.08 3.08

electrostatic Mg2+O2− pair, indicating some covalency between the cobalt ions in 1+ and 2+ and the superoxide, leading to stabilization of the 2pπgy orbital. For complexes 3+ and 4+, the Cohen and Känzig analysis again gives typical values for E, but both the spin−orbit coupling constant and the splitting between the two π* orbitals approach zero (Table 2). As Δ approaches 0 eV, eqs 1−3 break down due to the perturbative treatment being no longer valid. Relativistic electronic structure calculations were performed to better understand the origin of the unusual magnetic properties of 3+ and 4+ compared to those of 1+ and 2+ and most other superoxo-bridged Co(III,III) dimers. Crystal structure coordinates are only available for the peroxo-bridged version of these complexes, so geometry optimizations were performed using these structures as starting points (key geometric parameters are presented in Table 3). In all four cases, the removal of one electron leads to shortening of the O−O bond from ca. 1.41 Å to ca. 1.34 Å, consistent with the oxidation event being localized to the O2 moiety and not the

(2)

g y = ge[λ 2 /(λ 2 + Δ2 )]1/2 − λ /E{[λ 2 /(λ 2 + Δ2 )]1/2 − [Δ2 /(λ 2 + Δ2 )]1/2 − 1}

complex O2−:MgO 1+ 2+ 3+ 4+

(3)

The largest g-shift is observed for gz, which is oriented along the O−O bonding vector (eq 1). Angular momentum is mixed in the ground state (via the Lz operator) from the energetically near 2pπgy orbital (Δ = the energy difference between 2pπgx and 2pπgy). These levels are degenerate in the gas phase, but as superoxide binds to another species, they can split significantly. The spin−orbit coupling constant λ for the X2Πg state of the O2− ion has been measured to be 160 cm−1 (0.0198 eV).38 In effect, however, a significant reduction from this free-ion value has been observed for superoxide bound to metal systems (to as low as 80 cm−1, 0.0099 eV) due to charge donation and/or covalency. Eqs 2 and 3 give relations for gx and gy that also D

DOI: 10.1021/acs.inorgchem.6b01954 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Table 3. Structural Parameters for Select Dioxygen-Bridged Metal Dimersa complex

O−O (Å)

O2 (3Σg) O2−MgO 1 1opt 1+opt 2(ClO4−) 2(NO3−) 2opt 2+opt 3(PF6−) 3(PF6− F3CSO3−) 3opt 3+opt 4 4opt 4+opt 1 2+ 4 opt 3 2+ 4 opt Δ,Λ-[(en)2Co-μ(OH,O2)Co(en)2]3+b Δ,Λ-[(en)2Co-μ(OH,O2)Co(en)2]4+ Δ,Λ-[(en)2Rh-μ(OH,O2)Rh(en)2]3+ [(NH3)4Rh-μ(OH,O2)Rh(NH3)4]3+ [(NH3)4Rh-μ(OH,O2)Rh(NH3)4]4+ [(TPA)Co-μ(OH,O2)Co(TPA)]c μ(OH,O2)-[Co(enN4)]23+d μ(OH,O2)-[Co(enN4)]24+e

1.234 1.35 1.419 1.409 1.332 1.415 1.412 1.396 1.340 1.363 1.397 1.377 1.336 1.391 1.375 1.335 1.366 1.328 1.465 1.339 1.521 1.479 1.337 1.412 1.402 1.352

avg MO (Å)

M····M (Å)

avg. M−O−O (deg)

M−O−O−M (deg)

ref

1.875 1.940 1.931 1.873 1.866 1.884 1.872 1.912 1.883 1.916 1.902 1.890 1.915 1.899 1.915 1.902 1.862 1.873 1.991 1.990 2.000

4.491 4.641 4.640 3.304 3.270 3.357 3.353 3.832 3.797 3.862 3.912 3.785 3.841 3.890 3.879 3.900 3.272 3.261 3.448 3.446 3.418 3.265 3.178 3.279

112.6 118.4 118.4 112.9 112.2 112.8 121.2 115.5 113.8 115.3 112.0 113.2 115.5 112.1 111.7 112.5 110.3 119.9 107.2 n/a n/a

180.0 151.9 172.6 56.4 57.3 61.1 24.9 87.0 89.9 88.3 103.6 90.1 86.4 102.2 98.9 102.0 60.7 22.0 68.7 62 0.0

109.4 n/a

61.6 n/a

58 DFT DFT 59 60 DFT DFT 9 40 DFT DFT 9 DFT DFT DFT DFT 39 39 31 32 32 61 45 45

1.849 1.881

a Shaded rows denote formal superoxo-containing species. ben = ethylenediamine. cTPA = tris(2-pyridylmethyl)amine). dBis(μ-,N′-ethane-1,2diylbis(1-(pyridin-2-yl)methanimine))-(μ-OH)-(μ-O2)-dicobalt. eResults from TZV/B3PW91 geometry optimization reported in ref 45.

according to DFT results, these two π* (spin-up) orbitals remain approximately degenerate (Δ = 0.057 eV, cf. spin-up orbitals 209 and 210 in Figure 3) for complex 4+, as predicted by the Cohen and Känzig analysis (Table 2).42 In the oxidation of 2 to 2+, the electron is liberated from the πv* orbital, as predicted by ligand field theory and DFT (Figure S2).13 The inplane πh* orbital of 2+ remains doubly occupied as it is participating in bonding interactions with the two triply positive charged cobalt ions. In the case of oxidation of 4, however, it is the πh* orbital that is predicted by DFT to give up an electron. The tilted geometry of the O2 fragment leads to nearly equal interactions of the two orthogonal π* orbital with the two cobalt ions. The Löwdin spin population analysis on the model for 4+ shows that each oxygen carries 41% of the unpaired electron, while each cobalt possesses about 7% (Figure 3, top), similar to the predictions made based on the observed 59Co hyperfine coupling (vide supra).43 The DFT-predicted g-matrix for 2+ matches the experimental values very well, especially gy (Table 1), but not so for 3+ or 4+. The error in the predicted gy value likely stems from the neardegeneracy of the superoxide π* orbitals calculated for 3+ and 4+, leading to unquenched orbital angular momentum in the ground state that is not replicated by the single determinant DFT wave function. Further evidence of this is gleaned from the increased spin contamination calculated for the models of 3+ (contamination = ⟨S2⟩ − 0.750 = 0.042) and 4+ (0.048) relative to those computed for models of 1+ (0.005) and 2+ (0.009). The presence of significant spin contamination indicates that higher multiplicity wave functions are needed to mix into the desired doublet ground state wave function to minimize the SCF energy.

metal centers. For the μ-hydroxo, μ-peroxo-bridged complex 2, oxidation causes the Co−OO−Co dihedral angle φ to go from 61 to 25°. Such flattening has been observed crystallographically in the cases of Δ,Λ-[(en)2Coμ(OH,O2)Co(en)2]3+39 and [(NH3)4Rhμ(OH,O2)Rh(NH3)4]3+32 and is predicted on the basis of ligand field arguments.12 In contrast, for complexes 3 and 4, DFT results predict that oxidation opens up φ from 90 to 103°. In the peroxide form, that φ = 90° implies that both π* orbitals interact equally with the coordinating Co(III) ions. Thus, the change in the torsion angle upon oxidation is expected as the odd number of electrons (three) in the pair of π* orbitals necessarily results in a Jahn−Teller distortion to lower the energy of the doubly occupied orbital. Analysis of the geometry optimized structures of the bppbridged complexes points to a combination of factors that keep the Co−OO−Co core from going planar upon oxidation. The bpp ligand keeps the metal centers approximately 3.8 Å apart, significantly farther (0.5 Å) than most other cis peroxobridged Co(III) dimers (e.g., 2). While the bpp(tpy)2 ligand set does possess significant flexibility, allowing the cobalt ions to be as far apart as 4.343 Å for the hydroxo-water-bridged dicobaltic complex,40 3.953 Å for the chloride-bridged dicobaltous,41 and 3.797 Å for the peroxide-bridged form,9 this intermetal distance is still much longer than that measured for dimers bridged by hydroxide, for example. This long intermetal distance and the flexible bpp scaffold allows for the two cobalt centers to have askew 3d-orbitals, which leads to a bonding interaction of one cobalt with each of the fully occupied peroxide π* orbitals. Upon one-electron oxidation to give the superoxide form, this geometry is maintained for the most part (Table 3) and, E

DOI: 10.1021/acs.inorgchem.6b01954 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

(μ−OH)-(μ-O2)-dicobalt) has been recently synthesized by Rose and co-workers.45 The X-ray structure of μ(OH,O2)[Co(enN4)]23+ reveals that φ is only 61.6°, more similar to 2 than 3 or 4, and that the two cobalt ions are separated by only 3.12 Å. The linking alkane group could be lengthened and oxidized to give a longer rigid alkene, offering an appealing means to test the hypothesis raised in the present study because having the cobalt ions separated by ≈3.8 Å may be a factor in affording or enhancing OER activity. Model systems for intermediates in CoOx-catalyzed water oxidation are rather few. In Figure 4, we compare the Q-band

Figure 4. Comparison of CW EPR spectra for cobalt oxide in phosphate buffer poised at 1.5 V vs SHE (top); electrochemically oxidized [Co4O4py4ac4] (middle); and oxidized complex 4 (bottom). Spectrometer settings were reported earlier.

EPR spectra of two such models to that for the “resting” Co(IV) species in CoPi. In contrast to the dioxygen-bridged dimers described above, the g-matrix of the “resting” Co(IV) species in CoOx made in phosphate buffer (CoPi) possesses much more significant anisotropy.3,4 Previously, this has been interpreted as an indication of the localized nature of the unpaired electron spin within cobalt 3d orbitals (cobalt having a large spin−orbit coupling constant relative to oxygen). A cuboidal model of the CoOx film, [Co4O4py4ac4],46 can oxidize alcohols47 and may possess water-oxidizing activity.48−54 Advanced EPR studies of the one-electron oxidized form of this cube prepared in acetonitrile show that the new electron hole is shared equally over all eight atoms of the core.55 This hole delocalization scheme is in stark contrast to that for 3+ and 4+ studied here, in which the hole is highly localized on the O2 moiety. We suspect that the more acute Co−OCo bonding angles enhance the delocalization pathway in the cube. This leads to spin density on each of the eight core atoms and results in a much narrower EPR spectrum compared to that for the Co(IV)-containing CoPi (cf. Figure 4). Thus, far, no EPR data have been acquired for the oxidized cube in the presence of water. If, as some calculations predict,6,56 an acetate or pyridine ligand dissociates allowing water to bind, we would expect dramatic changes in the EPR spectrum due to loss of molecular symmetry and possibly localization of the electron hole onto a single metal-oxo (or hydroxo) unit. In conclusion, the tethering together of the two cobalt ions by the bpp ligand provides a kinetically stable framework that

Figure 3. DFT-computed spin population (top) and qualitative molecular orbital diagram (bottom) involving the cobalt 3d and oxygen 2p atomic orbitals of complex 4+. Percentages given in parentheses correspond to total contribution of the five 3d orbitals from both cobalt ions and total contribution of the three 2p orbitals from both oxygen atoms, respectively.

It is now clear that the canted geometry of the superoxide fragment leads to the unusual magnetic properties observed for complexes 3+ and 4+. This geometry may also have an important impact on OER activity. A similarly long intermetal distance (3.797 Å) was observed for a peroxo-bridged copper dimer that also features a pyrazolate-based scaffold chelating both metal centers.44 Like the bpp-bridged cobalt dimers above, the Cu−O2-Cu complex featured a tilted peroxo moiety (φ = 65.2°) with each π* orbital interacting with one metal, leading to O−O bond activation and enhanced reactivity. In the copper dimer case, this O2 binding motif is credited with easing the transition from 3O2 to 1O2. We invoke a similar reasoning for the cobalt dimer and suggest that one-electron oxidation of the superoxo-bridged 3+ or 4+ directly yields 3O2 without need of an intersystem crossing. Indeed, geometry optimizations of complex 4 with two electrons removed, thus corresponding to a bound neutral O2 molecule, yield a total energy for the triplet (SCF energy = −5121.40472 Eh) 60 kcal/mol lower than that of the singlet (−5121.30862 Eh). Another peroxo-bridged cobalt dimer with tethering ligands (μ(OH,O2)-[Co(enN4)]23+ = bis(μ-,N′-ethane-1,2-diylbis(1-(pyridin-2-yl)methanimine))F

DOI: 10.1021/acs.inorgchem.6b01954 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

(7) Mattioli, G.; Giannozzi, P.; Bonapasta, A. A.; Guidoni, L. Reaction Pathways for Oxygen Evolution Promoted by Cobalt Catalyst. J. Am. Chem. Soc. 2013, 135, 15353−15363. (8) Mavros, M. G.; Tsuchimochi, T.; Kowalczyk, T.; McIsaac, A.; Wang, L. P.; Van Voorhis, T. What Can Density Functional Theory Tell Us about Artificial Catalytic Water Splitting? Inorg. Chem. 2014, 53, 6386−6397. (9) Rigsby, M. L.; Mandal, S.; Nam, W.; Spencer, L. C.; Llobet, A.; Stahl, S. S. Cobalt analogs of Ru-based water oxidation catalysts: overcoming thermodynamic instability and kinetic lability to achieve electrocatalytic O2 evolution. Chemical Science 2012, 3, 3058−3062. (10) Brunschwig, B. S.; Chou, M. H.; Creutz, C.; Ghosh, P.; Sutin, N. Mechanisms Of Water Oxidation To Oxygen - Cobalt(IV) As An Intermediate In The Aquocobalt(II)-Catalyzed Reaction. J. Am. Chem. Soc. 1983, 105, 4832−4833. (11) Zhang, M.; de Respinis, M.; Frei, H. Time-resolved observations of water oxidation intermediates on a cobalt oxide nanoparticle catalyst. Nat. Chem. 2014, 6, 362−367. (12) Lever, A. B. P.; Gray, H. B. Electronic-Spectra Of MetalDioxygen Complexes. Acc. Chem. Res. 1978, 11, 348−355. (13) Miskowski, V. M.; Robbins, J. L.; Treitel, I. M.; Gray, H. B. Electronic-Structure And Spectra Of μ-Superoxo-Dicobalt(III) Complexes. Inorg. Chem. 1975, 14, 2318−2321. (14) Calvo, R.; Abresch, E. C.; Bittl, R.; Feher, G.; Hofbauer, W.; Isaacson, R. A.; Lubitz, W.; Okamura, M. Y.; Paddock, M. L. EPR Study of the Molecular and Electronic Structure of the Semiquinone Biradical QA−•QB−• in Photosynthetic Reaction Centers from Rhodobacter sphaeroides. J. Am. Chem. Soc. 2000, 122, 7327−7341. (15) Stoll, S.; Schweiger, A. EasySpin, a comprehensive software package for spectral simulation and analysis in EPR. J. Magn. Reson. 2006, 178, 42−55. (16) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J.. Gaussian 09, revision B.01; Gaussian, Inc.: Wallingford CT, 2009. (17) Tao, J. M.; Perdew, J. P.; Staroverov, V. N.; Scuseria, G. E. Climbing the density functional ladder: Nonempirical meta-generalized gradient approximation designed for molecules and solids. Phys. Rev. Lett. 2003, 91, 146401. (18) Staroverov, V. N.; Scuseria, G. E.; Tao, J. M.; Perdew, J. P. Comparative assessment of a new nonempirical density functional: Molecules and hydrogen-bonded complexes. J. Chem. Phys. 2003, 119, 12129−12137. (19) van Lenthe, E.; Ehlers, A.; Baerends, E. J. Geometry optimizations in the zero order regular approximation for relativistic effects. J. Chem. Phys. 1999, 110, 8943−8953. (20) Pantazis, D. A.; Chen, X.-Y.; Landis, C. R.; Neese, F. All-electron scalar relativistic basis sets for third-row transition metal atoms. J. Chem. Theory Comput. 2008, 4, 908−919. (21) Neese, F.; Wennmohs, F.; Hansen, A.; Becker, U. Efficient, approximate and parallel Hartree-Fock and hybrid DFT calculations. A ’chain-of-spheres’ algorithm for the Hartree-Fock exchange. Chem. Phys. 2009, 356, 98−109. (22) Klamt, A.; Schuurmann, G. Cosmo - A New Approach To Dielectric Screening In Solvents With Explicit Expressions For The Screening Energy And Its Gradient. J. Chem. Soc., Perkin Trans. 2 1993, 799−805.

can withstand the high potentials needed for water oxidation. We suggest that the bpp ligand also keeps the metal centers distant enough from one another to allow the O2 fragment to bind in tilted geometry, making the πh* and πv* orbitals nearly degenerate, a feature reflected by the unusual Cohen and Känzig parameters (Table 2) that were derived from the gtensor analysis. In an O2-activating dicopper complex, this geometric property was postulated to help minimize the 3O2 → 1 O2 barrier.44 When the microscopic reversibility is invoked, the tilting of the O2 moiety by the bpp-ligated cobalt dimers may aid in 1O2 → 3O2 conversion. Most useful, however, is that we now have an electronic structure description of a functional model for the Co(III)-superoxo moiety recently observed during water oxidation by Co3O4.11

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.6b01954. Coordinates for DFT models and supplementary EPR spectra (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Daniel L. M. Suess for fruitful discussions and acknowledge William H. Casey for his support of J.G.M. via the United States Department of Energy Office of Basic Energy Science under grant DE-FG03-02ER15693 (to W.H.C.). The research in this study was funded by the National Science Foundation under the NSF Center CHE-1305124. The EPR spectrometers at the CalEPR facility used in this study were funded by the National Institutes of Health (Grant S10RR021075) and the NSF (Grant CHE-1048671).

■

REFERENCES

(1) Kanan, M. W.; Nocera, D. G. In situ formation of an oxygenevolving catalyst in neutral water containing phosphate and Co2+. Science 2008, 321, 1072−1075. (2) Nocera, D. G. The Artificial Leaf. Acc. Chem. Res. 2012, 45, 767− 776. (3) McAlpin, J. G.; Surendranath, Y.; Dinca, M.; Stich, T. A.; Stoian, S. A.; Casey, W. H.; Nocera, D. G.; Britt, R. D. EPR Evidence for Co(IV) Species Produced During Water Oxidation at Neutral pH. J. Am. Chem. Soc. 2010, 132, 6882−6883. (4) Gerken, J. B.; McAlpin, J. G.; Chen, J. Y. C.; Rigsby, M. L.; Casey, W. H.; Britt, R. D.; Stahl, S. S. Electrochemical Water Oxidation with Cobalt-Based Electrocatalysts from pH 0−14: The Thermodynamic Basis for Catalyst Structure, Stability, and Activity. J. Am. Chem. Soc. 2011, 133, 14431−14442. (5) Friebel, D.; Bajdich, M.; Yeo, B. S.; Louie, M. W.; Miller, D. J.; Casalongue, H. S.; Mbuga, F.; Weng, T.-C.; Nordlund, D.; Sokaras, D.; Alonso-Mori, R.; Bell, A. T.; Nilsson, A. On the chemical state of Co oxide electrocatalysts during alkaline water splitting. Phys. Chem. Chem. Phys. 2013, 15, 17460−17467. (6) Wang, L.-P.; Van Voorhis, T. Direct-Coupling O2 Bond Forming a Pathway in Cobalt Oxide Water Oxidation Catalysts. J. Phys. Chem. Lett. 2011, 2, 2200−2204. G

DOI: 10.1021/acs.inorgchem.6b01954 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry (23) McLendon, G.; Martell, A. E. Stabilities Of Binuclear Cobalt Oxygen Adducts - Linear Free-Energy Relationship. J. Chem. Soc., Chem. Commun. 1975, 223−225. (24) Mori, M.; Weil, J. A.; Kinnaird, J. K. Preparation And Electron Paramagnetic Resonance Spectroscopy Of Dicobalt Peroxo Anions. J. Phys. Chem. 1967, 71, 103−110. (25) Weil, J. A.; Kinnaird, J. K. 17O Hyperfine Splitting In A μAmido-μ-Peroxodicobalt Complex. J. Phys. Chem. 1967, 71, 3341. (26) Pickens, S. R.; Martell, A. E. Electronic-Structures Of μ-PeroxoBis (Pentadentate Polyamine)Cobalt(III) Complexes. Inorg. Chem. 1980, 19, 15−21. (27) Lindsay, D. M.; Herschbach, D. R.; Kwiram, A. L. ESR of Matrix-Isolated Alkali Superoxides. Chem. Phys. Lett. 1974, 25, 175− 181. (28) Gordy, W. Theory and applications of electron spin resonance; Wiley: New York, 1980. (29) Chiesa, M.; Giamello, E.; Paganini, M. C.; Sojka, Z.; Murphy, D. M. Continuous wave electron paramagnetic resonance investigation of the hyperfine structure of 17O2− adsorbed on the MgO surface. J. Chem. Phys. 2002, 116, 4266−4274. (30) Springborg, J.; Zehnder, M. Synthesis And Crystal-Structure Of A Peroxo-Bridged Binuclear Rhodium(III) Complex And Its Superoxo-Bridged Analog. Helv. Chim. Acta 1984, 67, 2218−2225. (31) Springborg, J.; Zehnder, M. Synthesis And Crystal-Structure Of Meso-Delta,Delta-μ-Peroxo-μ-Hydroxobis Bis(Ethylenediamine)Rhodium(III) Trifluoromethane Sulfonate. Helv. Chim. Acta 1986, 69, 199−209. (32) Springborg, J.; Zehnder, M. Syntheses, Solution Properties And Crystal-Structures Of Peroxo-Bridged And Superoxo-Bridged Dinuclear Ammine Rhodium(III) Complexes. Acta Chemica Scandinavica Series a-Physical and Inorganic Chemistry 1988, 41, 484−495. (33) Schrauzer, G. N.; Lee, L. P. Cobaloximes(II) And Vitamin-B12r As Oxygen Carriers - Evidence For Monomeric And Dimeric Peroxides And Superoxides. J. Am. Chem. Soc. 1970, 92, 1551−1557. (34) Morton, J. R.; Preston, K. F. Atomic Parameters for Paramagnetic-Resonance Data. J. Magn. Reson. 1978, 30, 577−582. (35) Flanagan, H. L.; Singel, D. J. The Impact of ExcitationFrequency on the Nuclear Modulation of Electron-Spin Echoes - N-14 Hyperfine and Quadrupole Interactions of Dpph in Disordered Solids. Chem. Phys. Lett. 1987, 137, 391−397. (36) Flanagan, H. L.; Singel, D. J. Analysis of N-14 Eseem Patterns of Randomly Oriented Solids. J. Chem. Phys. 1987, 87, 5606−5616. (37) Kanzig, W.; Cohen, M. H. Paramagnetic Resonance Of Oxygen In Alkali Halides. Phys. Rev. Lett. 1959, 3, 509−510. (38) Huber, K. P.; Herzberg, G. Constants of Diatomic Molecules; Van Nostrand Reinhold: New York, 1979; Vol. 4. (39) Thewalt, U.; Struckmeier, G. Cobalt Complexes With O2 Bridges - Structure Of Cations μ-Hydroxo-μ-Peroxo-Bis Bis(Ethylenediamine)Cobalt(III) 3+ And μ-Hydroxo-μ-Superoxo-Bis Bis(Ethylenediamine)Cobalt(III) 4+. Z. Anorg. Allg. Chem. 1976, 419, 163−170. (40) Fukuzumi, S.; Mandal, S.; Mase, K.; Ohkubo, K.; Park, H.; Benet-Buchholz, J.; Nam, W.; Llobet, A. Catalytic Four-Electron Reduction of O2 via Rate-Determining Proton-Coupled Electron Transfer to a Dinuclear Cobalt-μ-1,2-peroxo Complex. J. Am. Chem. Soc. 2012, 134, 9906−9909. (41) Mandal, S.; Shikano, S.; Yamada, Y.; Lee, Y.-M.; Nam, W.; Llobet, A.; Fukuzumi, S. Protonation Equilibrium and Hydrogen Production by a Dinuclear Cobalt−Hydride Complex Reduced by Cobaltocene with Trifluoroacetic Acid. J. Am. Chem. Soc. 2013, 135, 15294−15297. (42) DFT results on the model for 2+ predict a Δ of 0.346 eV, also in very good agreement with the value determined from the Cohen and Känzig analysis of the corresponding g-values. (43) Optimization of the geometry following in silico removal of one electron has a dramatic effect on not only the structure (Table 3) but also the predicted 59Co hyperfine coupling as well (Table 1). Clearly, the modest reorientation of the superoxide moiety reduces the

covalency with the metal centers to give rise to metal hyperfine couplings that are more in line with experimental results. (44) Dalle, K. E.; Gruene, T.; Dechert, S.; Demeshko, S.; Meyer, F. Weakly Coupled Biologically Relevant CuII2(μ-η1:η1-O2) cis-Peroxo Adduct that Binds Side-On to Additional Metal Ions. J. Am. Chem. Soc. 2014, 136, 7428−7434. (45) Cho, Y. I.; Joseph, D. M.; Rose, M. J. Criss-Crossed” Dinucleating Behavior of an N4 Schiff Base Ligand: Formation of a μ-OH,μ-O2 Dicobalt(III) Core via O2 Activation. Inorg. Chem. 2013, 52, 13298−13300. (46) Chakrabarty, R.; Bora, S. J.; Das, B. K. Synthesis, structure, spectral and electrochemical properties, and catalytic use of cobalt(III)-Oxo cubane clusters. Inorg. Chem. 2007, 46, 9450−9462. (47) Chakrabarty, R.; Sarmah, P.; Saha, B.; Chakravorty, S.; Das, B. K. Catalytic Properties of Cobalt(III)-Oxo Cubanes in the TBHP Oxidation of Benzylic Alcohols. Inorg. Chem. 2009, 48, 6371−6379. (48) McCool, N. S.; Robinson, D. M.; Sheats, J. E.; Dismukes, G. C. A Co4O4 ″Cubane″ Water Oxidation Catalyst Inspired by Photosynthesis. J. Am. Chem. Soc. 2011, 133, 11446−11449. (49) Stracke, J. J.; Finke, R. G. Electrocatalytic Water Oxidation Beginning with the Cobalt Polyoxometalate [Co4(H2O)2(PW9O34)2]10−: Identification of Heterogeneous CoOx as the Dominant Catalyst. J. Am. Chem. Soc. 2011, 133, 14872−14875. (50) Smith, P. F.; Kaplan, C.; Sheats, J. E.; Robinson, D. M.; McCool, N. S.; Mezle, N.; Dismukes, G. C. What Determines Catalyst Functionality in Molecular Water Oxidation? Dependence on Ligands and Metal Nuclearity in Cobalt Clusters. Inorg. Chem. 2014, 53, 2113− 2121. (51) Berardi, S.; La Ganga, G.; Natali, M.; Bazzan, I.; Puntoriero, F.; Sartorel, A.; Scandola, F.; Campagna, S.; Bonchio, M. Photocatalytic Water Oxidation: Tuning Light-Induced Electron Transfer by Molecular Co4O4 Cores. J. Am. Chem. Soc. 2012, 134, 11104−11107. (52) La Ganga, G.; Puntoriero, F.; Campagna, S.; Bazzan, I.; Berardi, S.; Bonchio, M.; Sartorel, A.; Natali, M.; Scandola, F. Light-driven water oxidation with a molecular tetra-cobalt(III) cubane cluster. Faraday Discuss. 2012, 155, 177−190. (53) Evangelisti, F.; Guettinger, R.; More, R.; Luber, S.; Patzke, G. R. Closer to Photosystem II: A Co4O4 Cubane Catalyst with Flexible Ligand Architecture. J. Am. Chem. Soc. 2013, 135, 18734−18737. (54) Ullman, A. M.; Liu, Y.; Huynh, M.; Bediako, D. K.; Wang, H.; Anderson, B. L.; Powers, D. C.; Breen, J. J.; Abruna, H. D.; Nocera, D. G. Water Oxidation Catalysis by Co(II) Impurities in Co(III)4O4 Cubanes. J. Am. Chem. Soc. 2014, 136, 17681−17688. (55) McAlpin, J. G.; Stich, T. A.; Ohlin, C. A.; Surendranath, Y.; Nocera, D. G.; Casey, W. H.; Britt, R. D. Electronic Structure Description of a [Co(III)3Co(IV)O4] Cluster: A Model for the Paramagnetic Intermediate in Cobalt-Catalyzed Water Oxidation. J. Am. Chem. Soc. 2011, 133, 15444−15452. (56) Fernando, A.; Aikens, C. M. Reaction Pathways for Water Oxidation to Molecular Oxygen Mediated by Model Cobalt Oxide Dimer and Cubane Catalysts. J. Phys. Chem. C 2015, 119, 11072− 11085. (57) Giamello, E.; Sojka, Z.; Che, M.; Zecchina, A. Spectroscopic Study Of Superoxide Species Formed By Low-Temperature Adsorption Of Oxygen Onto CoO-MgO Solid-Solutions - An Example Of Synthetic Heterogeneous Oxygen Carriers. J. Phys. Chem. 1986, 90, 6084−6091. (58) Ramprasad, D.; Gilicinski, A. G.; Markley, T. J.; Pez, G. P. Cobalt(II) Polypyridine Complexes and Their Reversible Reactivity with Dioxygen. Inorg. Chem. 1994, 33, 2841−2847. (59) Ludovici, C.; Frohlich, R.; Vogtt, K.; Mamat, B.; Lubben, M. Caged O2 - Reaction Of Cytochrome bo3 Oxidase With Photochemically Released Dioxygen From A Cobalt Peroxo Complex. Eur. J. Biochem. 2002, 269, 2630−2637. (60) Howard-Jones, A. R.; Adam, V.; Cowley, A.; Baldwin, J. E.; Bourgeois, D. Cryophotolysis of a caged oxygen compound for use in low temperature biological studies. Photochem. Photosynth. Sci. 2009, 8, 1150−1156. H

DOI: 10.1021/acs.inorgchem.6b01954 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry (61) Wang, H. Y.; Mijangos, E.; Ott, S.; Thapper, A. Water Oxidation Catalyzed by a Dinuclear Cobalt-Polypyridine Complex. Angew. Chem., Int. Ed. 2014, 53, 14499−14502.

I

DOI: 10.1021/acs.inorgchem.6b01954 Inorg. Chem. XXXX, XXX, XXX−XXX