Paramagnetic Resonance of Cobalt(II) Trispyrazolylmethanes and

Dec 15, 2016 - This allowed us to trace the physical and electronic dynamics ... X-ray intensity data were measured on a Bruker CCD-based ...... The S...
0 downloads 0 Views 1MB Size
Download PDF
Article pubs.acs.org/IC

Paramagnetic Resonance of Cobalt(II) Trispyrazolylmethanes and Counterion Association Amy R. Marts,† Joshua C. Kaine,† Robert R. Baum,† Vivien L. Clayton,† Jami R. Bennett,† Laura J. Cordonnier,† Robert McCarrick,† Abed Hasheminasab,‡ Laura A. Crandall,‡ Christopher J. Ziegler,‡ and David L. Tierney*,† †

Department of Chemistry and Biochemistry, Miami University, Oxford, Ohio 45056, United States Department of Chemistry, University of Akron, Akron, Ohio 44325, United States

‡

S Supporting Information *

ABSTRACT: Paramagnetic resonance studies (EPR, ESEEM, ENDOR, and NMR) of a series of cobalt(II) bis-trispyrazolylmethane tetrafluoroborates are presented. The complexes studied include the parent, unsubstituted ligand (Tpm), two pyrazolesubstituted derivatives (4Me and 3,5-diMe), and tris(1pyrazolyl)ethane (Tpe), which includes a methyl group on the apical carbon atom. NMR and ENDOR establish the magnitude of 1H hyperfine couplings, while ESEEM provides information on the coordinated 14N. The data show that the pyrazole 3-position is more electron rich in the Tpm analogues, that the geometry about the apical atom influences the magnetic resonance, and that apical atom geometry appears more fixed in Tpm than in Tp. NMR and ENDOR establish that the BF4− counterion remains associated in fluid solution. In the case of the Tpm3,5Me complex, it appears to associate in solution, in the same position it occupies in the X-ray structure.

■

temperature-dependent, fluid-solution NMR and multifrequency, frozen-solution EPR and ENDOR to map spin densities and electron dynamics in Tp2Co. This allowed us to trace the physical and electronic dynamics responsible for 1H NMR paramagnetic relaxation enhancements (PREs) to unquenched orbital angular momentum in the electronic ground state, without the need for electron spin transitions. We were also able to identify and characterize a small dynamic Jahn−Teller effect in solution, operating on a picosecond time scale with an energy barrier of ∼200 cm−1. This report marks the beginning of similar investigations of the Tpm analogues; we present here solution EPR, NMR, ESEEM, and ENDOR studies of [Co(Tpm)2](BF4)2 and a series of substituted analogues, including two pyrazole-substituted derivatives (4Me and 3,5-diMe) and tris(1-pyrazolyl)ethane (Tpe), which incorporates a pendant methyl group on the apical carbon atom. The studies below use NMR and ENDOR to establish the magnitude of 1H hyperfine couplings, and ESEEM to interrogate 14N couplings, and they offer some insight into the lack of similar pulsed-EPR studies in the literature. They also show that the geometry about the apical atom of the ligand influences the magnetic resonance, while NMR and ENDOR establish that the BF4− counterion remains associated in fluid solution. In the case of the Tpm 3,5Me complex, the

INTRODUCTION The scorpionates represent one of the most versatile ligand scaffolds ever developed. Since they were first reported in the late 1960s,1−3 they have found utility in catalytic,4 molecular magnetic,5 and enzyme model systems.6−9 They also presented some of the first examples of spin-crossover complexes,10,11 a property that remains a topic of current research.12 Both trispyrazolylborate (Tp) and the related trispyrazolylmethane (Tpm) present three pyrazole nitrogens as donors, constrained to a facial arrangement. Both scaffolds are easily modified at the ring carbons and in the apical positions,2,3,13 and this ease of substitution leads, in part, to their unusual versatility.4 The key difference between the two ligands is the identity of the apical atom (boron in Tp vs carbon in Tpm), which leads to an anionic Tp ligand and neutral Tpm. In the absence of steric bulk at the pyrazole 3-position,6,7 divalent metal ions readily form bis-homoleptic complexes of the type Tp2M and [M(Tpm)2]2+, and bis-Tp and -Tpm chelates of nearly every first-row transition metal are known.2,3 The Co(II) complexes of both Tp and Tpm were among the first reported.14−17 Both exhibit D3d global symmetry, imposed by a trigonal distortion on the order of 10% of the bond length along the B··Co··B or C··Co··C axes.18,19 Jesson’s early work on these compounds showed they had virtually identical electronic structures, based on their optical and EPR spectra,14 and both complexes give simple four-line NMR spectra.16 In a recent series of reports,20−22 we have examined Tp2Co and a series of alkyl-substituted derivatives using field- and © XXXX American Chemical Society

Received: October 19, 2016

A

DOI: 10.1021/acs.inorgchem.6b02520 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

by 2νN.24,25 For nuclei with I = 1 (14N), each hyperfine level is further split into a doublet separated by 3/2P, by the nuclear quadrupole interaction. An ENDOR experiment will probe these transitions directly, using RF excitation,24,25 while an ESEEM experiment will detect them indirectly, through time-dependent modulation of the ESE intensity.26,27 Consequently, the ESEEM experiment detects not only principal frequencies but also combination frequencies that can often be used as diagnostic. When available, the ENDOR response is often easier to interpret, while the ESEEM response, as we will show below, can be much easier to detect, but more difficult to analyze directly. At the low- and high-field edges (“single-crystal-like” positions) of an EPR signal, an ESEEM or ENDOR experiment probes a minimal subpopulation of molecular orientations, while at intermediate fields, a larger, but well-defined, subset of orientations is interrogated.25 The observed EPR, ESEEM, and ENDOR signals shown here come from the MJ = ±1/2 ground doublet of the J = 5/2 spin−orbit manifold, as described by Jesson,14 but they are discussed in terms of an effective spin, S′ = 1/2. The ENDOR spectra of Figure 5 were simulated using the program DIPSIM,28 which generates the spectrum from user input distances and angles and values of the isotropic (contact) coupling. The ESEEM simulations in Figure 3 were generated using the saf fron subroutine of EasySpin,29,30 with cross-term-averaged fast Fourier transforms calculated using its caf f t function. Example input is included in the Supporting Information.

tetrafluoroborate anion appears to associate in solution in the same position it occupies in the X-ray structure.

■

EXPERIMENTAL SECTION

Synthesis. The ligand tris(1-pyrazolyl)methane and its methylsubstituted derivatives, Tpm4Me, Tpm3,5Me, and tris(1-pyrazolyl)ethane, were prepared according to published procedures13,23 and recrystallized from boiling hexanes. Mixing the recrystallized ligands with Co(BF4)2·6H2O in a 2:1 ratio in THF led immediately to yellow precipitates, which were collected and washed with hexanes. The hexafluorophosphate analogue of the parent complex was prepared by anion metathesis between [Co(Tpm)2](NO3)2 and (NH4)PF6 in acetonitrile. All compounds were verified for composition by ESI-MS. Crystals of the parent complexes, [Co(Tpm)2](BF4)2 and [Co(Tpm)2](PF6)2, were grown using vapor diffusion of diethyl ether into a solution of the complex in nitromethane, affording translucent yellow crystals. X-ray intensity data were measured on a Bruker CCD-based diffractometer with dual Cu/Mo ImuS microfocus optics (Cu Kα radiation, λ = 1.541 78 Å, Mo Kα radiation, λ = 0.710 73 Å). Crystals were mounted on a cryoloop using Paratone oil and placed under a steam of nitrogen at 100 K (Oxford Cryosystems). The data were corrected for absorption with the SADABS program. The structures were refined using the Bruker SHELXTL software package (version 6.1) and were solved using direct methods until the final anisotropic full-matrix, least-squares refinement of F2 converged. Experimental details for the crystal structures of [Co(Tpm)2](BF4)2 and [Co(Tpm)2](PF6)2 can be found in the Supporting Information. NMR Spectroscopy. Samples for NMR were prepared from microcrystalline material in d3-acetonitrile (identical spectra were obtained in CD3NO2). NMR of 1H (200 MHz) and 19F (188 MHz) were measured on a Bruker Avance 200 spectrometer. The 1H and 19F NMR spectra (Figures 4 and 6, respectively) consist of 1024 transients of 8k data points over a 75 kHz spectral window (375 ppm for 1H; 400 ppm for 19F; tAQ = 54 ms) and were collected using a 4 μs (1H) or 6.4 μs (19F) RF pulse. Prior to Fourier transformation, all FIDs were smoothed by exponential multiplication, which incorporated an additional line width of 5 Hz. Continuous Wave (CW) EPR. Frozen-solution X-band EPR spectra were recorded on a Bruker EMX EPR spectrometer equipped with an ER-4116DM dual-mode resonator. Temperature was maintained using an Oxford Instruments ESR-900 liquid He cryostat and temperature controller. Samples were prepared by dissolving microcrystalline compounds in nitromethane and adding acetone (∼3:1 final ratio) as a glassing agent. The spectra presented in Figure 2 were recorded using the following conditions: T = 4.5 K; νMW = 9.68 GHz (0.2 mW); 10 G field modulation (100 kHz); time constant/ conversion time = 81.92 ms; 4 scans each. Pulsed EPR/ESEEM/ENDOR. X-band electron spin−echo (ESE) detected EPR, ESEEM, and Mims pulsed ENDOR spectra were acquired on a Bruker Elexsys E680 spectrometer, operating at 9.76 GHz, equipped with an Oxford ESR-900 liquid He cryostat and temperature controller and an EN 4118X-MD4 resonator. Samples were 1 mM in 3:1 nitromethane/acetone. Deuterated solvents were used for ENDOR measurements, to ensure the absence of matrix 1H ENDOR, while protio solvents were used for ESEEM measurements, to exclude matrix contributions from 2H. The ENDOR spectra used MW pulse lengths = 10 ns, τ = 188 ns, RF pulse length = 6 μs (100 W), repetition rate = 1000 Hz. Each spectrum consists of 1024 points (6 kHz point spacing), with each point the average of 8000 transients (160 scans at 50 averages per scan). The ESEEM measurements used MW pulse lengths = 10 ns, repetition rate = 1000 Hz, and consisted of 1024 points, using either 4 ns spacing and two-step phase cycling (two-pulse) or 8 ns spacing and four-step phase cycling (three-pulse). Briefly, for a single molecular orientation, the first-order ENDOR/ ESEEM response from a nucleus with I = 1/2(1H) is a doublet with frequencies ν± = |νN ± A/2|, where A is the orientation-dependent hyperfine coupling and νN is the nuclear Larmor frequency; for A/2 < νN (e.g., 1H) this corresponds to a doublet centered at νN and split by A, while A/2 > νN (e.g., 14N) leads to a doublet centered at A/2 and split

■

RESULTS To simplify the comparisons below, all of the complexes examined here were prepared as the tetrafluoroborate salts. While the PF6− derivative of the parent complex is discussed in brief, the magnetic resonance studies focus exclusively on the BF4− complexes. Two of them, the 4-methyl (Tp4Me) and apical methyl (Tpe) derivatives, are being reported for the first time. Although we were unable to obtain diffraction-quality crystals of these, analytically pure material suitable for spectroscopy was readily available. To aid in the presentation of the data that follow, we will use the color palette in Scheme 1 whenever Scheme 1

possible. In short, the pyrazole 3-, 4-, and 5-positions are represented in green, blue, and purple, respectively, while the apical position is represented in red. When data are presented comparing the set of complexes, they will be presented in a color that is consistent with the substitution pattern, according to Scheme 1. Solid-State Structures. ORTEPs of the parent compound, [Co(Tpm)2](BF4)2 (vide inf ra), and of [Co(Tpm3,5Me)2](BF4)213 are shown in Figure 1, with views both normal to the C··Co··C vector (top) and along it (bottom). Similar views for the hexafluorophosphate complex are presented in Figure S1; details for both structures are available in the Supporting Information. Selected metrics are collected in Table 1, along with those for Tp2Co.20 As in Tp2Co, all three Tpm structures show a Co2+ ion held in a pseudo-octahedral local environment (CoN6), trigonally distorted along the C··Co··C vector. The intraligand N−N distance, rN−N(intra), is ∼10% less than the interligand N−N distance, rN−N(inter) (Table 1), and there is a B

DOI: 10.1021/acs.inorgchem.6b02520 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Table 1. Crystallographic Distances and Anglesa complex d

Tp2Co [Co(Tpm)2][BF4]2 [Co(Tpm)2][PF6]2 [Co(Tp3,5Me)2]2+e

Co−N 2.12 2.11 2.12 2.13

(52) (51) (51) (51)

r

b

N−N

(intra)

2.89 2.84 2.85 2.87

r

b

N−N

(inter)

Δr

3.13 3.13 3.14 3.15

c

N−N

(intra)

+0.006 +0.007 +0.012 −0.022

Co··C 3.20 3.11 3.12 3.12

Co··CH 4.35 4.11 4.12 4.07

(0) (0) (0) (0)

Co··5H 5.04 5.04 5.06 5.85

(33) (31) (31) (27)

Co··4H 5.21 5.24 5.24 5.23

(61) (59) (59) (58)

Co··3H 3.44 3.48 3.47 3.95

(84) (83) (83) (91)

Average distances (Å, ±0.001) and azimuthal angles (θ), relative to the C··Co··C vector. bAverage intra- and interligand N··N distances in Å. Average difference in Co−N bond lengths (Å) within a ligand. dCo··B and Co··BH; from Myers et al., Inorg. Chem. 2008, 47, 6701−6710.20 eFrom Reger et al., Inorg. Chem. 2002, 41, 4453−4460.13 a c

small tetragonal distortion in the solid state (ΔrN−N(intra) in Table 1). These values are similar to those shown by Tp2Co, and the pyrazole 3-, 4-, and 5-proton distances and azimuthal angles (relative to the C··Co··C vector) vary minimally relative to Tp2Co. However, the Co··C and Co··CH distances are substantially shorter than the corresponding Co··B and Co··BH distances (0.09 and 0.24 Å, respectively). This is the combined effect of a 0.15 Å shorter C−H bond and 0.09 Å shorter C−N bonds. The result is inversion of the apical atom’s geometry. That is, in Tp2Co, the geometry about the apical B is elongated along the B−H bond, leading to larger average H−B−N angles (110.4°) than N−B−N angles (108.5°). In contrast, the geometry about the apical C in [Co(Tpm)2](BF4)2 is compressed along the C− H bond, leading to smaller H−C−N angles (108.1°) than N− C−N angles (110.8°). The three-dimensional structure of [Co(Tpm3,5Me)2](BF4)2 shows a small tetragonal compression in the solid state, with one Co−N bond from each Tpm ligand 0.022 Å shorter than the other two.13 This is unlike the parent complex, which has one pair of Co−N bonds elongated by 0.007−0.012 Å (almost identical to the neutral bis-Tp analogue). Compared to the parent complex, it appears the sterics introduced by the methyl groups in the 3- and 5-positions lead to a different placement of the counterion within the unit cell. In the Tpm structure, the counterion packs nearly along one Co−N bond direction, ∼52° off the 3-fold axis, and sitting ∼10° out of the pyrazole plane (see Figure 1). The counterion in the PF6− analogue sits almost perfectly on one trigonal face of the coordination octahedron, defined by two pyrazoles from one Tpm ligand and one pyrazole from the other, making a 61° angle with the molecular 3-fold (Figure S1) axis. Meanwhile, in the Tpm3,5Me derivative, the BF4− counterion is almost directly above the apical C (on the 3-fold axis).13 As we will show below, the counterions for all five complexes appear to remain associated in both fluid and frozen solution. EPR Spectroscopy. Frozen solution X-band EPR spectra, shown in Figure 2A, are consistent with trigonal symmetry in all of the tetrafluoroborate examples. The trends are the same as those we showed previously for the (RTpx)2Co series,20 with g∥ ≈ 8 and remarkably insensitive to substitutions and g⊥ modulated by the substitution pattern. The effect on g⊥ is greatest for substitutions near the axis directions (the pyrazole 3-proton and the apical CH), moving to lower field (higher g) with substitutions at the 3-position and higher field (lower g) with substitution at the apical position. All four complexes afford well-resolved 59Co hyperfine structure at g∥ (Figure 2B), with splittings identical to Tp2Co, as summarized in Table 2. Pulsed EPR and ESEEM. X-band two-pulse ESE-detected EPR spectra are shown in Figure S2. Consistent with the CW spectra in Figure 2, there is a strong response near g⊥ that attenuates rapidly as the field is decreased, with no discernible

Figure 1. ORTEP diagrams for [Co(Tpm)2](BF4)2 (left) and [Co(Tpm3,5Me)2](BF4)2 (right). Boron and fluorine atoms of the counterions are shown in black and orange, respectively. Within the complexes, cobalt is shown in dark blue, nitrogen in blue, and carbon in gray, and the protons follow the color palette in Scheme 1.

ESE intensity at g∥. ESE relaxation rates in these highly anisotropic, high-spin systems are exceptionally short (T1 ≈ 2 μs and T2 ≈ 1.4 μs at 4.5 K), independent of solvent (H or D), for all four derivatives. Two- and three-pulse ESEEM data for [Co(Tpm)2][BF4]2, acquired at the peak of the ESE-EPR (Figure S2), are shown in Figure 3. At this field, the constitutive protons (20 within ∼5 Å) do not contribute appreciable modulation (νH ≈ 37 MHz). However, the presence of 6 equivalent coordinated nitrogens and 6 more noncoordinated nitrogens only one bond further away leads to exceptionally deep two-pulse modulation. The uncorrected time trace is shown in the inset to Figure 3A. With modulation depths approaching 100%, the Fourier transform of the two-pulse decay is largely uninformative (not shown). However, there is substantial τ-dependence in the three-pulse ESEEM (Figure 3A). Given the complexity of the time domain patterns, we found it more instructive to attempt to simulate the corresponding frequency domain spectra (Figure 3B). With 12 14N within two bonds of the metal ion, one should expect substantial cross-coupling and combination frequencies that are difficult to account for (not shown). To make the problem more tractable, we generated independent simulations for coordinated 14N (red lines in Figure 3B) and noncoordinated 14N (blue lines in Figure 3B). An expanded view of Figure 3B is shown in the Supporting Information (Figure S3). Using our previous ENDOR studies of Tp2Co20,21 as starting estimates for the hyperfine and quadrupole couplings in [Co(Tpm)2]2+, the final values obtained here (4.5 MHz for C

DOI: 10.1021/acs.inorgchem.6b02520 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 2. (A) Frozen solution X-band CW EPR spectra for [Co(Tpmx)2](BF4)2. (B) Expansion of the g∥ region. Spectra are presented in the same order, colored according to the pattern in part A.

Table 2. EPR Parameters for [Co(Tpmx)2](BF4)2 complex

g⊥

g∥

A∥ (G)

[Co(Tpm)2]2+ [Co(Tpe)2]2+ [Co(Tp4Me)2]2+ [Co(Tp3,5Me)2]2+ Tp2Co

0.79 0.71 0.78 0.86 1.02

8.37 8.37 8.31 8.27 8.48

93 95 94 92 93

NMR spectra. They consist of single resonances from the pyrazole 3-, 4-, and 5-protons and the apical CH protons. As illustrated in Figure 4, the methyl substitutions lead to minimal

Figure 4. 200 MHz room-temperature [Co(Tpmx)2](BF4)2.

1

H NMR spectra for

perturbations, excluding the positions that were substituted. The similarity of the NMR spectra provides the basis for the comparative analysis that follows.

Figure 3. X-band three-pulse ESEEM (left) and corresponding Fourier transforms (right). Values of τ as indicated. Inset: Two-pulse ESEEM time trace. Arrows correspond to the two-pulse ESE at τ = 188, 490, 766, and 1324 ns. Simulations corresponding to coordinated (red) and noncoordinated (blue) 14N are shown below the Fourier transforms. An expanded view of the simulations is shown in the Supporting Information.

δ HC =

⎛ Δδobs − δ HP⎜1 − ⎝

C δMe =0

the coordinated 14N; 2 MHz for the noncoordinated 14N) do not differ greatly from those starting values (6 and 2 MHz, respectively). As is apparent in Figure 3B, the predicted frequency spectrum, for both types of nitrogen, is highly dependent on the value of τ. Final hyperfine couplings were chosen to match the apparent double-quantum frequency (Figure 3B), while the quadrupole parameters were adjusted to best match the low-frequency part of the spectrum. The quadrupole tensors were assumed to be directed at the Co(II) ion for the coordinated 14N and directed at the coordinated 14N for the noncoordinated 14N. While it is possible that these orientations are incorrect, the present data are insufficient to be definitive. However, we note that the simulations were relatively insensitive to the quadrupole directions (aside from β, the tip away from gz), and any increase (or decrease) in the quadrupole couplings very quickly led to substantially less satisfactory results. 1 H NMR Spectroscopy. To interrogate the proton hyperfine couplings, we first turn to solution NMR. In 3-fold symmetry, all six coordinating pyrazoles and the two apical carbons are symmetry equivalent, resulting in four-line 1H

3

( ) RH RMe

3⎞

( ) ⎟⎠ RH RMe

P Δδobs = δ HP − δMe

(1)

Previously, we used the relationship between proton and methyl chemical shift to estimate the relative contributions of contact (isotropic) and dipolar coupling to the hyperfine shifts, in substituted derivatives of Tp2Co, based on eq 1.20 In eq 1, δCH and δCMe are the contact shift of a proton or methyl in a given position, Δδobs is the change in paramagnetic shift from proton to methyl (indicated by the arrows in Figure 4), δPH is the total paramagnetic shift (contact + dipolar) of the proton in the given position, and RH and RMe are the average metal−nucleus distances for the proton and methyl, respectively. When both crystal structures are available, the only unknown parameter in eq 1 is δCH. From the structures of the Tpm and Tpm3,5Me complexes, the 3H and 5H positions are well-determined, and perhaps not surprisingly, the proton and methyl metrics are very similar to those of Tp2Co and (Tp3.5Me)2Co. We therefore felt comfortable assuming the 4Me protons of the Tpm4Me complex could be found at an average distance of 6.19 Å and 59° off-axis, similar to their disposition in (Tp4Me)2Co. The apical methyl protons were assumed to lie at an average distance of 5.01 Å and 7° off-axis, assuming a similar 7° D

DOI: 10.1021/acs.inorgchem.6b02520 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Table 3. Contact and Dipolar Proton Hyperfine Couplings in [Co(Tpm)2](BF4)2 [Co(Tpm)2]2+

Δδobs

δPHa

δCHb

δDHc

Aisod (NMR)

Aisoe (ENDOR)

Aisof Tp2Co

CH 5H 4H 3H

40 46 47 −18

135 90 39 −113

37 21 52 −26

98 69 −13 −87

1.47 0.83 2.06 −1.04

2.3 0.8 2.3 −1.6

1.27 1.07 1.94 −0.27

Total hyperfine shift, corrected for diamagnetism (5−7 ppm). bContact shift, from eq 1. cPseudocontact shift, δPH − δCH. dFrom eq 2. eFrom the ENDOR simulations in Figure 4. fFrom Myers et al., Inorg. Chem. 2008, 47, 6701−6710.20 a

shown in Figure 5. A consequence of the rapid ESE relaxation in these systems is that only short interpulse delays ( 5 MHz, are well outside the breadth of the 1H ENDOR envelope, so the spectra should represent reasonably undistorted line shapes. Interestingly, as indicated by the simulations in Figure 5A, only one of the two expected hyperfine features was observed, for any given proton. This is an uncommon occurrence that implies an implicit triple effect, known to arise when ESE modulation by 14N is unusually large, leading to a dominant ENDOR line that indicates the sign of the intrinsic hyperfine coupling.35 Without knowledge of the intrinsic sign of A(14N), we cannot define the sign of A(1H) from the ENDOR. However, the pattern is consistent with the NMR analysis, with the 3H coupling opposite in sign of the other three. The simulations in Figure 5A were arrived at using the metal− proton metrics in Table 1, along with the NMR-derived values for Aiso (Table 3) as a starting point and then adjusting Aiso to match the observed resonance position. As summarized in Table 3, the room-temperature NMR-derived couplings did not differ greatly from the ground-state hyperfine couplings measured by ENDOR, with the largest difference seen for the apical position (discussed below). The assignment of features in Figure 5A is fully supported by comparison with the analogous ENDOR spectra from the methyl derivatives, presented in Figure 5B. As in the NMR, we anticipate, in each case, the loss of a feature seen in the spectrum of the parent complex (from the proton that was

(2)

The contact shifts were then used to calculate the contact (isotropic) coupling, hereafter referred to as Aiso, based on eq 2.31,32 In eq 2, gav is taken as (2g⊥ + g∥)/3 = 3.32, and the remaining constants take their usual values. It has been shown for these trigonal systems, where the spin−orbit coupling defines the electronic structure, that use of S = 3/2 in eq 2 will underestimate the true value of Aiso by as much as 50%, owing to unaccounted for excited-state contributions.33 Using J = 5/2, in this case, seems more appropriate, and the values in Table 3 have been adjusted accordingly, as they were in our original work on Tp2Co.20 In accord with the contact shifts, the isotropic couplings obtained for the Tpm complex are generally within 10−20% of those in the Tp complex, with the exception of the pyrazole 3H proton, which is nearly 4-fold larger in the Tpm complex. ENDOR Spectroscopy. The NMR-derived couplings, which nominally should translate into ground-state couplings, should allow prediction of the ENDOR spectra, which measure the ground-state couplings directly. X-band Mims 1H ENDOR, taken at the high field (low-g) edge of the EPR spectrum, are E

DOI: 10.1021/acs.inorgchem.6b02520 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

(−150.1). The 3,5Me complex, whose counterion was ENDOR-observable at g⊥, shifts its counterion 19F NMR signal by 8.0 ppm at room temperature (−151.6). This is 80-fold greater than the shift observed for the parent complex. While it is difficult to speculate on the origin of these differences without X-ray structures for the 4Me and Tpe derivatives, as we will show below, we believe lack of a significant chemical shift effect is insufficient to rule out close association in solution.

replaced) and the potential addition of a new feature from the methyl group. As indicated by the accompanying simulations in Figure 5B, the spectra from the methyl-substituted derivatives support the assumption of a methyl proton Aiso ≈ 0, with the exception of the apical methyl of the Tpe complex (red line in Figure 5B), which requires an Aiso of either −3 or +4.4 MHz (dashed red line). In contrast, the apparent change in 3H proton ENDOR, from Tpm to Tpm3,5Me, is exactly as predicted (compare the solid green simulation to the dashed green simulation in Figure 5B). Similarly, substitution at the 4position, predicted to show the largest proton Aiso, leads to loss of a broad envelope of intensity near +0.5 MHz, nicely consistent with the simulation based on the NMR-derived coupling. Meanwhile, any gains in intensity from the added 4Me could only have occurred at frequencies approaching ν−νH = 0, consistent with a 4Me Aiso ≈ 0. Substitution of the 5H proton is predicted to lead to a minimal difference in the overall pattern; field-dependent measurements will be necessary to interrogate this position in more detail. In addition to signals from 1H, the Tp3,5Me complex also shows a sharp, intense doublet centered at the 19F Larmor frequency, which must arise from the BF4− counterion, suggesting that the counterion, at least in this complex, is closely associated in frozen solution. The splitting of 0.16 MHz can be considered a purely dipolar interaction, and its observation at g⊥ suggests it resides very near g⊥, as it does in the crystal structure. Using eq 3,36 where gN, ge, βN, and βe take their usual values and R is the Co··F distance, and taking into account the observed g-value (g/ge) and a 10% reduction in spin density on Co(II) (ρCo = 0.9),21 the observed splitting corresponds to a Co··F distance of ∼6.5 Å. At 0.6 Å closer than the closest approach in the crystal structure,13 this clearly indicates the BF4− counterion is as closely associated with the Tpm3,5Me complex in frozen solution as it is in the solid state. Adipolar = −

g βNg βe 1 T = − ρCo N 3 e ge 2 R

■

DISCUSSION The bis chelates of both Tp and Tpm were carefully examined by Jesson shortly after they were first reported in the late 1960s, and not surprisingly, the two complexes were shown to have very similar electronic structures.14 Later it was shown that they share similar physical structures.18,19 Their EPR and NMR properties are generally similar, as well, with each complex showing an axial EPR spectrum (g∥ ≈ 8, g⊥ ≈ 1) and a four-line NMR spectrum from six symmetry-equivalent pyrazoles and two symmetry-equivalent apical protons.16,20 Pulsed-EPR, ESEEM, and ENDOR. The ENDOR data of Figure 5 mark only the fourth report of the technique being applied to a high-spin (hs) Co(II)-containing complex20,21,37 and only the second report of pulsed-ENDOR37 (there are a number of reports on low-spin Co(II)-containing systems). Similarly, the ESEEM spectra of Figure 3 mark only the fourth report of this technique being applied to any hs Co(II) system38−40 and the first of a well-characterized, structurally homogeneous molecular species, the others being in situ generated aqueous complexes of NAD+,38 a series of fluxional aza-acetates related to EDTA,39 and Co(H2O)62+.40 The data presented here are consistent with those earlier reports and can begin to shed light on the paucity of reports applying advanced EPR techniques to hs Co(II)-containing systems. With relaxation times at 4.5 K of 2 μs or less, there is insufficient phase memory in the electron spin system to allow for facile spectral detection, let alone spectral editing. Common RF pulse lengths in Mims ENDOR range from a few microseconds to 10 or more, depending on the spectrometer and other factors, meaning the ESE intensity will be heavily attenuated at the end of the RF pulse. Hence, we were able to obtain Mims ENDOR only at the shortest τ available, and no 14N ENDOR response was observed under these conditions. Coupled with the inability to obtain a Davies echo, again due to relaxation, it was impossible to detect 14N directly, requiring us to turn to ESEEM. The ESEEM data showed remarkably deep 14N modulation, similar to that shown for the aza-acetates,39 although the data cannot be compared directly due to differences in inherent electronic structure (and g-values). The ESE shows very high sensitivity to τ in three-pulse ESEEM, and the pattern is consistent with 14N couplings similar to those we observed by ENDOR in the (RTpx)2Co series.20 The large g-anisotropy of these systems presents the unusual situation where proton frequencies (νH ≈ 37 MHz) are well-separated from the 14N frequencies (νN ≈ 2 MHz) at g⊥ ≈ 0.8, which is largely inaccessible at frequencies above X-band. This eliminates the complication of proton modulation,39 meaning the signals in Figure 3 can be fully attributed to 14N, but reproducing them proved difficult. While we could draw some conclusions based on simulating the two types of 14N independently, any attempt to include both in a single simulation led to a number of predicted frequencies that were not observed. It is important to

g⊥

(3)

19

F NMR Spectroscopy. The observation of a strongly coupled 19F in the ENDOR of the Tpm3,5Me complex led us to examine the counterion association in fluid solution. Roomtemperature 19F NMR spectra for the set of compounds are shown in Figure 6. The fluorines of the counterion of the parent complex, at −143.7 ppm, are shifted only 0.1 ppm relative to free tetrafluoroborate (NH4BF4, −143.6 ppm), which could imply very weak association in solution. For comparison, the counterion of the 4Me complex (−144.9) is shifted 1.2 ppm, and that of the Tpe complex is shifted 6.5 ppm

Figure 6. 188 MHz room-temperature [Co(Tpmx)2](BF4)2.

19

F NMR of (NH4)BF4 and F

DOI: 10.1021/acs.inorgchem.6b02520 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry note that while rigorous simulation of the ESEEM was not possible, acquisition of the data was relatively easy. Apical Atom Geometry. The method used to separate the paramagnetic shift into its contact and pseudocontact components was developed for the methyl-substituted trispyrazolylborates.20 There, the NMR-derived couplings nicely reproduced the X-band 1H ENDOR pattern at the high-field edge of the EPR spectrum (g ≈ 1), while Q-band ENDOR data obtained at the low-field edge of the EPR (g ≈ 8) showed the BH proton to have a value for Aiso that is ∼2-fold smaller than predicted by NMR.21 In the present study, the NMR-derived couplings again nicely reproduced the X-band 1H ENDOR of the parent complex and the methylpyrazole deivatives, but the assumptions involved fail to describe the methyl substitution at the apical position. A survey of the literature shows surprisingly few examples of bis-Tpm complexes of Co(II) and fewer crystal structures aside from the parent complex19 and its appearance in much more complex extended structures.41 All examples of the parent complex show a trigonally compressed geometry about the apical carbon, with N−C−N angles (110.8°) larger than H− C−N (108.1°) angles. As shown in Table 4, this is in stark

Co··B(BF4−) angle is just 4°. An examination of the roomtemperature (297 K) 19F NMR showed that all of the complexes had some effect on the counterion 19F chemical shift, requiring that all of the complexes retain some level of interaction with their counterions in fluid solution. Previous NMR studies also suggest that the PF6− anion associates strongly with [Co(Tpm)2]2+ in solution42 and that organic bases, such as anilines and pyridines, associate similarly with Tp2Co.43 It is tempting to try to use the chemical shifts as a marker for the strength of that interaction. However, it is important to remember that the pseudocontact (dipolar) shift is both distance- and orientation-dependent.31,32 While the Tpm3,5Me counterion sits nearly on-axis in the solid state, the counterion of the parent Tpm complex occupies a position between two of the coordinated pyrazoles, making a CH··Co··B(BF4−) angle of 52°, very close to the magic angle of 54.7°. The simplest expression describing the pseudocontact shift is given in eq 4.31,32 For the two structurally characterized tetrafluoroborate examples, the anisotropy factors differ little (g∥2 − g⊥2), but the geometric factors differ almost 15-fold. Solving eq 4 for the metal−nucleus distance, R, and using the known values of g∥ and g⊥ (Table 2) and θ = 4° (Tpm3,5Me) or 52° (Tpm), we obtain Co··F distances of 20 and 35 Å, respectively.

Table 4. Geometry about the Apical Atom (X) in [Co(Tpmx)2]2+ and (RTpx)2Co complex

H−X−N

N−X−N

[Co(Tpm)2](BF4)2 [Co(Tpm)2](PF6)2 [Co(Tp3,5Me)2](BF4)2a (nBuTp)2Cob Tp2Cob (Tp4Me)2Cob (Tp3Me)2Cob (Tp3,5Me)2Cob

108.1 108.3 107.8 112.1 110.5 110.4 109.7 109.6

110.8 110.7 111.1 106.6 108.4 108.5 109.2 109.4

2 ⎛ μ ⎞ μ S(S + 1) ⎛ 3 cos2 θ − 1 ⎞ 2 Δν 2 = ⎜ 0⎟ B ⎟(g − g⊥ ) ⎜ ⎝ 4π ⎠ kT ν0 ⎠ ⎝ R3

(4)

While distances greater than 20 Å would imply the counterion is largely dissociated near room temperature, we know it is tightly in place at low temperature in the Tpm3,5Me complex, and it appears associated in the same position it occupies in the crystal.13 The sharpness of the 19F doublet in Figure 5B can only come from a well-ordered species, which argues against them coming back together by diffusion over large distances during the freezing process. Field-dependent ENDOR studies will be necessary to see if the counterion associates equally strongly in the other compounds, where we would expect the highest transition probability at a field corresponding to the CH··Co··B(BF4−) angle. To document this process more precisely, a careful examination of the temperature-dependent 1H and 19F PREs will follow.

a

From Reger et al., Inorg. Chem. 2002, 41, 4453−4460.13 bFrom Myers et al., Inorg. Chem. 2008, 47, 6701−6710.20

contrast to the analogous RTpx series, where the geometry about the apical boron proved sensitive to the substitution pattern, and substitutions at the pyrazole 3-position served to pyramidalize the boron. The only comparable example available is the tetrafluoroborate salt of [Co(Tpm3,5Me)2]2+,13 which shows a nearly indistinguishable geometry about the apical carbon (Table 4). In the absence of a crystal structure of the Tpe complex, which includes an apical methyl group, we were forced to assume that the geometry about the apical carbon was unchanged. However, the apical methyl of the Tpe complex showed a substantially larger hyperfine coupling than predicted, which suggests a substantial rearragement of the electronic structure around the carbon atom. Given that the B−H bond was shown to be electron rich in a trigonally elongated geometry,21 and a larger than expected hyperfine coupling requires a substantially more electron rich C−CH3 bond, we believe this may indicate an inversion of configuration at the apical C in the Tpe complex. Further experiments are in progress to test this. Counterion Association. The obervation of a strong 19F signal from the BF4− counterion in ENDOR of the Tpm3,5Me complex must arise from a tightly associated counterion in frozen solution. The appearance of a 19F doublet along g⊥ is consistent with the counterion being associated in a location similar to that shown in the crystal structure, where the CH··

■

SUMMARY We have presented here paramagnetic resonance studies (EPR, ESEEM, ENDOR, and NMR) of a series of related tetrafluoroborate salts of cobalt(II) bis-trispyrazolylmethanes (Tpm), including two pyrazole-substituted derivatives (4Me and 3,5-diMe) and tris(1-pyrazolyl)ethane, which includes a methyl group on the apical carbon atom. NMR and ENDOR establish the magnitude of 1H hyperfine couplings, while ESEEM provides information on the coordinated and noncoordinated 14N. The data show that the geometry about the apical atom influences the magnetic resonance, while NMR and ENDOR establish that the BF4− counterion remains associated in fluid solution. In the case of the Tpm3,5Me complex, it appears to associate in solution in the same position it occupies in the X-ray structure. The studies presented here suggest that solution studies of counterion interactions can prove valuable and pave the way for further pulsed EPR, ESEEM, and ENDOR studies of high-spin Co(II)-containing systems. G

DOI: 10.1021/acs.inorgchem.6b02520 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

■

Spectral Study of Several {Fe[tris(pyrazolyl)methane]2}(BF4)2 Complexes: Observation of an Unusual Spin-State Crossover. Inorg. Chem. 2001, 40, 1508−1520. (11) Long, G. J.; Grandjean, F.; Reger, D. L. Spin Crossover in Pyrazolylborate and Pyrazolylmethane Complexes. Top. Curr. Chem. 2004, 233, 91−122. (12) Gardinier, J. R.; Treleven, A. R.; Meise, K. J.; Lindeman, S. V. Accessing Spin-Crossover Behavior in Iron(II) Complexes of NConfused Scorpionate Ligands. Dalton Trans. 2016, 45, 12639−12643. (13) Reger, D. L.; Grattan, T. C.; Brown, K. J.; Little, C. A.; Lamba, J. J. S.; Rheingold, A. L.; Sommer, R. D. Synthesis of Modified Tris(pyrazolyl)methane Ligands: Backbone Functionalization. Synth. 2002, 3, 350−356. (14) Jesson, J. P. Optical and Paramagnetic Resonance Spectra of Some Trigonal Co(II) Chelates. J. Chem. Phys. 1966, 45, 1049−1056. (15) Jesson, J. P. Theory of Isotropic Nuclear Resonance Shifts in Octahedral Co2+ Systems. J. Chem. Phys. 1967, 47, 579−581. (16) Jesson, J. P. Isotropic Nuclear Resonance Shifts in Some Trigonal Co(II) and Ni(II) Chelate Systems. J. Chem. Phys. 1967, 47, 582−591. (17) Jesson, J. P.; Trofimenko, S.; Eaton, D. R. Spectra and Structure of Some Transition Metal Poly(1-pyrazolyl)borates. J. Am. Chem. Soc. 1967, 89, 3148−3158. (18) Churchill, M. R.; Gold, K.; Maw, C. E., Jr. The Crystal Structure and Molecular Geometry of Bis[hydrotris(1-pyrazolyl)borato]cobalt(II). Inorg. Chem. 1970, 9, 1597−1604. (19) Astley, T.; Gulbis, J. M.; Hitchman, M. A.; Tiekink, E. R. T. Structure, spectroscopic and angular-overlap studies of tris(pyrazol-1yl)methane complexes. J. Chem. Soc., Dalton Trans. 1993, 509−515. (20) Myers, W. K.; Duesler, E. N.; Tierney, D. L. Integrated Paramagnetic Resonance of High-Spin Co(II) in Axial Symmetry: Chemical Separation of Dipolar and Contact Electron-Nuclear Couplings. Inorg. Chem. 2008, 47, 6701−6710. (21) Myers, W. K.; Scholes, C. P.; Tierney, D. L. Anisotropic Fermi Couplings due to Large Unquenched Orbital Angular Momentum: Qband 1H, 14N and 11B ENDOR of Bis(trispyrazolylborate) Cobalt(II). J. Am. Chem. Soc. 2009, 131, 10421−10429. (22) Tierney, D. L. Jahn-Teller Dynamics in a Series of HighSymmetry Co(II) Chelates Determine Paramagnetic Relaxation Enhancements. J. Phys. Chem. A 2012, 116, 10959−10972. (23) Reger, D. L.; Grattan, T. C.; Brown, K. J.; Little, C. A.; Lamba, J. J. S.; Rheingold, A. L.; Sommer, R. D. Syntheses of tris(pyrazolyl)methane ligands and {[tris(pyrazolyl)methane]Mn(CO)3}SO3CF3 complexes: comparison of ligand donor properties. J. Organomet. Chem. 2000, 607, 120−128. (24) Hoffman, B. M. ENDOR of Metalloproteins. Acc. Chem. Res. 2003, 36, 522−529. (25) Hoffman, B. M.; DeRose, V. J.; Doan, P. E.; Gurbiel, R. J.; Houseman, A. L. P.; Telser, J. EMR of Paramagnetic Molecules; Plenum Press: New York, 1993; pp 151−218. (26) Britt, R. D. Electron spin echo methods: A tutorial. ACS Symp. Ser. 2003, 858 (Paramagnetic Resonance of Metallobiomolecules), 16−54. (27) McCracken, J. Electron spin echo envelope modulation (ESEEM) spectroscopy. In Applications of Physical Methods to Inorganic and Bioinorganic Chemistry; Scott, R. A.; Lukehart, C. M., Eds.; WileyBlackwell: Hoboken, NJ, 2007; pp 57−77. (28) DIPSIM is available free of charge from Dr. Peter Doan, Northwestern University, upon request. (29) Stoll, S.; Britt, R. D. General and Efficient Simulation of Pulse EPR Spectra. Phys. Chem. Chem. Phys. 2009, 11, 6614−6625. (30) Stoll, S.; Schweiger, A. EasySpin, a Comprehensive Software Package for Spectral Simulation and Analysis in EPR. J. Magn. Reson. 2006, 178, 42−55. (31) Bertini, I.; Luchinat, C. NMR of Paramagnetic Molecules in Biological Systems; Benjamin/Cummings: Menlo Park, CA, 1986. (32) Bertini, I.; Luchinat, C.; Parigi, G. Solution NMR of Paramagnetic Molecules; Elsevier: Amsterdam, 2001.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.6b02520. Structural data (CIF) Table of crystallographic parameters (Table S1), sample ESEEM simulation input files, three figures showing ORTEP diagrams (S1), a comparison of CW and ESEdetected EPR of the series (S2), and an expanded view of the ESEEM data (S3) (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Robert R. Baum: 0000-0002-9824-944X Christopher J. Ziegler: 0000-0002-0142-5161 David L. Tierney: 0000-0002-5597-799X Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors wish to thank Dr. Peter Doan (Northwestern University) for helpful discussions and the U.S. National Science Foundation for their support of this work (CHE1152755 to D.L.T.).

■

REFERENCES

(1) Trofimenko, S. Boron-Pyrazole Chemistry. II. Poly(1-pyrazolyl)borates. J. Am. Chem. Soc. 1967, 89, 3170−3177. (2) Trofimenko, S. Scorpionates: The Coordination Chemistry of Polypyrazolylborate Ligands; Imperial College Press: London, 1999. (3) Trofimenko, S. Recent Advances in Poly(pyrazolyl) borate (Scorpionate) Chemistry. Chem. Rev. 1993, 93, 943−980. (4) Otero, A.; Fernández-Baeza, J.; Lara-Sánchez, A.; Sánchez-Barba, L. F. Metal Complexes with Heteroscorpionate Ligands Based on the Bis(pyrazol-1-yl)methane Moiety: Catalytic Chemistry. Coord. Chem. Rev. 2013, 257, 1806−1868. (5) Werner, T. W.; Reschke, S.; Bunzen, H.; Nidda, H.-A. K. v.; Deisenhofer, J.; Loidl, A.; Volkmer, D. [Co5Tp*4(Me2bta)6]: A Highly Symmetrical Pentanuclear Kuratowski Complex Featuring Tris(pyrazolyl)borate and Benzotriazolate Ligands. Inorg. Chem. 2016, 55, 1053−1060. (6) Ruman, T.; Ciunik, Z.; Wolowiec, S. Synthesis, X-ray crystallographic and 1H NMR spectroscopic structural studies on cobalt(II) complexes of homoscorpionate, heteroscorpionates and chiral tris(pyrazolyl)borates obtained from 5(3)-Isopropyl-3(5)-phenylpyrazole and 3,5-dimethylpyrazole. Eur. J. Inorg. Chem. 2003, 2003, 2475−2485. (7) Jacobsen, F. E.; Breece, R. M.; Myers, W. K.; Tierney, D. L.; Cohen, S. M. Model Complexes of Cobalt-Substituted Matrix Metalloproteinases: Tools for Inhibitor Design. Inorg. Chem. 2006, 45, 7306−7315. (8) Pellei, M.; Papini, G.; Lobbia, G. G.; Santini, C. Chemistry and Relevant Biomimetic Applications of Group 6 Metals Systems Supported by Scorpionates. Curr. Bioact. Compd. 2009, 5, 321−352. (9) Krzystek, J.; Swenson, D. C.; Zvyagin, S. A.; Smirnov, D.; Ozarowski, A.; Telser, J. Cobalt(II) “Scorpionate” Complexes as Models for Cobalt-Substituted Zinc Enzymes: Electronic Structure Investigation by High-Frequency and -Field Electron Paramagnetic Resonance Spectroscopy. J. Am. Chem. Soc. 2010, 132, 5241−5253. (10) Reger, D. L.; Little, C. A.; Rheingold, A. L.; Lam, M.; LiableSands, L. M.; Rhagitan, B.; Concolino, T.; Mohan, A.; Long, G. J.; Briois, V.; Grandjean, F. A Synthetic, Structural, Magnetic, and H

DOI: 10.1021/acs.inorgchem.6b02520 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry (33) McGarvey, B. R. Theory of the Isotropic NMR Shifts in Trigonal Co(II) Complexes. J. Chem. Phys. 1970, 53, 86−91. (34) Mims, W. B. Pulsed ENDOR Experiments. Proc. R. Soc. London, Ser. A 1965, A283, 452−457. (35) Doan, P. E.; Nelson, M. J.; Jin, H.; Hoffman, B. M. An Implicit TRIPLE Effect in Mims Pulsed ENDOR: A Sensitive New Technique for Determining Signs of Hyperfine Couplings. J. Am. Chem. Soc. 1996, 118, 7014−7015. (36) Hutchison, C. A.; McKay, D. B. The Determination of Hydrogen Coordinates in Lanthanum Nicotinate Dihydrate Crystals by Nd(+3)-Proton Double Resonance. J. Chem. Phys. 1977, 66, 3311− 3330. (37) Walsby, C. J.; Krepkiy, D.; Petering, D. H.; Hoffman, B. M. Cobalt-Substituted Zinc Finger 3 of Transcription Factor IIIA: Interactions with Cognate DNA Detected by 31P ENDOR Spectroscopy. J. Am. Chem. Soc. 2003, 125, 7502−7503. (38) Hoffmann, S. K.; Goslar, J.; Lijewski, S. Electron Paramagnetic Resonance and Electron Spin Echo Studies of Co2+ Coordination by Nicotinamide Adenine Dinucleotide (NAD+) in Water Solution. Appl. Magn. Reson. 2013, 44, 817−826. (39) Kang, P. C.; Eaton, G. R.; Eaton, S. S. Pulsed Electron Paramagnetic Resonance of High-Spin Cobalt(II) Complexes. Inorg. Chem. 1994, 33, 3660−3665. (40) Shimizu, T.; Mims, W. B.; Davis, J. L.; Peisach, J. Biochim. Biophys. Acta, Gen. Subj. 1983, 757, 29−34. (41) Wang, J.; Xu, Y.-L.; Zhou, H.-B.; Wang, H.-S.; Song, X.-J.; Song, Y.; You, X.-Z. Octacyanotungstate(V)-based square W2M2 (M = Co, Mn) complexes: synthesis, structure and magnetic properties. Dalton Trans. 2010, 39, 3489−3494. (42) Eaton, D. R.; Seville, L.; Jesson, J. P. Structure of the Second Coordination Sphere of Transition Metal Complexes. III. Ion Pairing involving Cobalt Complexes. Can. J. Chem. 1971, 49, 2751−2754. (43) Eaton, D. R. Structure of the second coordination sphere of transition metal complexes. 1. Co(II) complexes with aniline and pyridine. Can. J. Chem. 1969, 47, 2645−2654.

I

DOI: 10.1021/acs.inorgchem.6b02520 Inorg. Chem. XXXX, XXX, XXX−XXX