Article Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX
pubs.acs.org/IC
Electron and Oxygen Atom Transfer Chemistry of Co(II) in a Proton Responsive, Redox Active Ligand Environment Brian J. Cook, Maren Pink, Kuntal Pal, and Kenneth G. Caulton* Department of Chemistry, Indiana University, Bloomington, Indiana 47405, United States S Supporting Information *
ABSTRACT: The bis-pyrazolato pyridine complex LCo(PEt3)2 serves as a masked form of three-coordinate CoII and shows diverse reactivity in its reaction with several potential outer sphere oxidants and oxygen atom transfer reagents. N-Methylmorpholine N-oxide (NMO) oxidizes coordinated PEt3 from LCo(PEt3)2, but the final cobalt product is still divalent cobalt, in LCo(NMO)2. The thermodynamics of a variety of oxygen atom transfer reagents, including NMO, are calculated by density functional theory, to rank their oxidizing power. Oxidation of LCo(PEt3)2 with AgOTf in the presence of LiCl as a trapping nucleophile forms the unusual aggregate [LCo(PEt3)2Cl(LiOTf)2]2 held together by Li+ binding to very nucleophilic chloride on Co(III) and triflate binding to those Li+. In contrast, Cp2Fe+ effects oxidation to trivalent cobalt, to form (HL)Co(PEt3)2Cl+; proton and the chloride originate from solvent in a rare example of CH2Cl2 dehydrochlorination. An unexpected noncomplementary redox reaction is reported involving attack by 2e reductant PEt3 nucleophile on carbon of the 1e oxidant radical Cp2Fe+, forming a P−C bond and H+; this reaction competes in the reaction of LCo(PEt3)2 with Cp2Fe+.
■
We have shown5 that deprotonation of both pyrazolyl rings in (LH2)CoCl2 (Scheme 1) is accompanied by chloride removal from cobalt to generate a unit of formula LCo, which is trapped with added PEt3 to form five-coordinate, square pyramidal LCo(PEt3)2. This molecule has one unpaired electron, and the long Co−P(axial) bond and the electron paramagnetic resonance (EPR) hyperfine coupling to a single P are all consistent with a (dz2)1 HOMO. We are interested in whether this can serve as a source of the simple LCo moiety under reactive conditions and thus a source of low coordinate Co(II). Of special interest is whether, under oxidizing conditions, cobalt here would serve as a one- or two-electron reductant. Also of interest is what role liberated free PEt3 would have in such behavior; as metal is oxidized, one might anticipate that the soft phosphine donor might dissociate from cobalt. In addition, the pyrazolyl rings are established to participate in electron transfer,6 and thus the envisioned reactivity under oxidizing conditions might involve electron transfer from these pincer ligand rings, not only from cobalt. At the very least, the anionic, amide-bearing pyrazolate rings might serve as electron donating substituents to the metal (Scheme 1), increasing its reducing power and/or accessing CoIV. We report here on the reactivity of LCo(PEt3)2 with three very different oxidants: an amine oxide, Ag+, and the outer sphere oxidant Cp2Fe+.
INTRODUCTION A frequent consequence of providing oxidants to a metal, usually by installing a terminal oxo ligand, is that the redox equivalents are rapidly spent oxidizing nearby ligand functionality, accomplishing ligand degradation (e.g., phosphine oxidation, CH hydroxylation, H atom abstraction, etc.).1 Thus, development of robust ligand architectures ensures that oxidizing equivalents are not squandered on intramolecular oxidation. We felt that a pyrazolate ligand (Scheme 1, showing pincer ligand H2L) had relatively robust imine ligand functionality nearby and other functionality in the ligand might resist oxidation or be remote from the metal reactive site in a geometrically protected location. The work reported here is designed to test some of these hypotheses. There are no known structurally characterized Co(IV)O complexes; however high-valent CoO intermediates have been invoked in Co-catalyzed O atom transfer reactions and water oxidation, and characterized computationally and spectroscopically.2 Those that have been characterized are not best described as Co(IV)O, but instead as some contribution of Co(III) and Co(II) with radical character on the oxide ligand.3 Very recently, Cummins and Nocera illustrated the surprising ability of Co(II) to resist oxidation by O atom sources.4 In this case, a di-Co(II) anion (hence electron rich) was oxidized at the cryptand encapsulating the countercation, not the Co centers. This example nicely sets the stage for the results reported here, where the product and oxidation state formed are highly dependent on the choice of oxidant (O atom, inner-sphere, outer-sphere). © XXXX American Chemical Society
Received: March 28, 2018
A
DOI: 10.1021/acs.inorgchem.8b00816 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
center from CoII to CoIII, which is then followed by some chemical change competitive with the CV scan rate. The +1.19 V oxidation is also irreversible; the oxidation product may contain CoIV or form L•− from L2−. However, the CV shows that, even after traversing this second oxidation, some electroactive species remain; degradation is incomplete on the CV time scale. The CV serves to reinforce the fact that CoII in this system is quite electron-rich, and an ideal starting material for exploring oxidation chemistry. A Two-Electron O Atom Transfer Oxidant, Amine Oxide. Reaction of LCo(PEt3 ) 2 with a slurry of Nmethylmorpholine-N-oxide (NMO) in a 1Co/NMO ratio in Et2O occurs over the course of 12 h at 25 °C with a lor change from dark red to goldenrod yellow. Workup by filtration and removal of volatiles give a yellow solid whose 1H NMR spectrum in C6D6 shows chemical shifts in the range of +100 to −120 ppm, indicating paramagnetism (Figure S9). The number and intensity of NMR signals show 2-fold symmetry, with one t Bu peak at 19.6 ppm, together with other peaks assigned to the NMO ring and methyl hydrogens; pyrazole and pyridyl ring protons are also seen consistent with 2-fold symmetry. Assay of the reaction solution by 31P NMR prior to workup shows only one broad signal centered at +48 ppm, consistent with the value for OPEt3 (Figure S10). Single crystal X-ray diffraction of crystals grown by slow diffusion of pentane vapors into a saturated benzene solution reveals a product of formula LCo(NMO)2; see Figure 2.
Scheme 1. Deprotonation of H2L, Showing Different Resonance Forms (top), Dehydrohalogenation of Corresponding Divalent Metal Halide Complexes (Middle), and Oxidative-Induced FLP-like Activity (Bottom) Presented in This Work
■
RESULTS Cyclic Voltammetry of LCo(PEt3)2. To evaluate the potential metal- and ligand-based redox character of LCo(PEt3)2, cyclic voltammetry (CV) was performed. Scanning anodically (100 mV·s−1) in CH2Cl2, two irreversible oxidations are observed (Figure 1), with an Epa of −0.54 V and +1.19 V vs
Figure 2. ORTEP drawing (50% probability) of the non-hydrogen atoms of molecule A of LCo(NMO)2, showing selected atom labeling. Unlabeled atoms are carbon. Selected structural parameters: Co1A− O1A, 1.954(2) Å; Co1A−O3A, 1.984(2) Å; Co1A−N3A, 2.068(3) Å; Co1A−N5A, 2.161(3) Å; Co1A−N1A, 2.194(3) Å; O1A−Co1A− O3A, 111.81(11)°; O1A−Co1A−N3A, 126.23(11)°; O3A−Co1A− N3A, 121.90(10)°; N5A−Co1A−N1A, 151.98(10)°.
The asymmetric unit contains two full molecules, but there are no significant differences between the two, so only data from molecule A is discussed here. The absence of any PEt3 ligand in the unit cell is consistent with the signal observed in the 31P NMR indicating complete conversion to OPEt3 and no coordinated OPEt3. The Co here is five-coordinate, with a tridentate doubly anionic pincer and two additional NMO ligands. The structure shows idealized C2 symmetry, and the bond lengths are faithful to this symmetry. Unlike square pyramidal LCo(PEt3)2, the structure is closer to trigonal bipyramidal, with an Addison τ5 parameter7 of 0.83. The Co−O bond distances here are consistent with the few previous crystallographically characterized Co−N-oxide complexes.8 The lattice of the crystal studied incorporates stoichiometric amounts of water. In each LCo(NMO)2 molecule, one water
Figure 1. Cyclic voltammogram of 2.0 mM LCo(PEt3)2 in CH2Cl2 with 0.1 M [TBA]PF6 at 100 mV s−1. Arrow indicates scan origin and direction.
Fc/Fc+. The first oxidation wave is less irreversible (more return current observed in Figure 1, see Figure S2) if only that peak is scanned (i.e., no scan to +1.19 V); the return current is increased to 7.50 μA from 3.49 μA, but is still not electrochemically reversible. The product of this oxidation is subsequently reduced at −1.59 V, and we will discuss below a possible identity of this electroactive species (vide infra). Thus, the first oxidative wave corresponds to an oxidation at the metal B
DOI: 10.1021/acs.inorgchem.8b00816 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry Scheme 2. Quantitative Ranking of Oxygen Atom Transfer Agents, from Weakest (Bottom) to Strongest (Top)
how electron rich and how seemingly easy it should be to oxidize it to CoIII, at least. The persistence of NMO as a ligand, as opposed to O atom transfer to one or perhaps two cobalt centers, is somewhat surprising. In particular, if the zwitterionic NMO oxygen is primarily a σ donor to cobalt, then the second NMO might be activated to oxidize (O atom transfer) a hypothesized nucleophile-augmented (electron rich) four coordinate LCo(NMO) intermediate. The thermodynamic potential of NMO for O atom transfer as well as other traditional O atom transfer reagents thus needs to be evaluated. Relevant Thermodynamic Information. The thermodynamic strength of the available various single oxygen atom transfer reagents has not been evaluated quantitatively. To assist in choosing among such reagents, we felt that density functional theory (DFT) calculation of some reaction energies could shed light on atom transfer reagent strength. Moreover, since these involve only light atoms, the results should be computationally reliable, especially for ranking the relative thermodynamic O atom transfer strength. We particularly wanted to compare the thermodynamic potential of the various amine oxide oxidants. DFT calculations on pyridine N-oxide (pyNO), Me3NO, N-methylmorpholine N-oxide (NMO), and N2O were carried out, and Scheme 2 shows these, listed from strongest to weakest O atom transfer reaction energies toward pyridine. Iodosobenzene is clearly the most powerful O atom transfer reagent among those considered, but its insolubility, due to intermolecular interactions in the solid state, limits its utility to heterogeneous conditions. Those intermolecular interactions also diminish its performance in oxygen atom transfer, although soluble variants have been synthesized and are used readily.10
molecule is found between one pyrazolate nitrogen (O/N = 2.83 Å) and one NMO oxygen (O/O = 2.84 Å). This hydrogen bonding lengthens both the Co−O(NMO) and Co−N(pyrazolate) distances (by about 0.03 Å) but does not alter the amine oxide N/O distance. This is further detailed in Supporting Information. An Evans method magnetic susceptibility determination gave a μeff value of 3.39 μB, consistent with three unpaired electrons on CoII. We note that the high spin and trigonal bipyramidal LCo unit here with NMO contrasts to the low spin and square pyramidal structure when the two neutral ligands are PEt3.5 Low spin results from the more strongly binding phosphines raising the energy of a σ*Co−P orbital, thus favoring a low spin ground state. Co−N distances here are lengthened compared to LCo(PEt3)2 due to the spin state change at cobalt. C−N−N bond angles at N2a and N4a, 106.9° and 108.0°, are consistent with the absence of protons on these nitrogens.9 A (cobalt-free) solution of PEt3 and NMO (1:3 mol ratio) in Et2O shows only 1% production of OPEt3 in time comparable to that employed above (Figure S16), which indicates that cobalt in the synthesis of LCo(NMO)2 catalyzes phosphine oxidation; that further reactivity is beyond the scope of this report. However, we note that this result with O atom transfer agents is complementary to our recent report effecting that catalysis with N2O.5 The persistence of CoII in this product is noteworthy; as noted previously,3f,4 CoII can be recalcitrant toward innersphere oxidation and O atom transfer, and apparently no O atom transfer from NMO to Co in LCo(NMO) 2 is thermodynamically favored vs possible Co−O−Co or CoO species. Persistence of CoII remains a surprise, however, due to C
DOI: 10.1021/acs.inorgchem.8b00816 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
chloride. This [LCo(PEt3)2Cl(LiOTf)2] monomer unit is then dimerized, around a center of symmetry, by triflate oxygens, which form a four coordinate NClO2 tetrahedral environment around Li+. Two Li+ are incorporated per Co because they profit from donation by two pyrazolate β nitrogens as well as from two chloride lone pairs. Triflates are bridging between two Li + in the dimer. Cobalt is saturated in the moiety LCo(PEt3)2Cl, which is rich with Lewis basic sites, so incorporation of each Li+ originates from the electron rich pocket formed by pyrazolate β nitrogen and the nearby electron rich12 chloride on a d6 metal; the Cl(CoN)N unit is thus a bidentate ligand to Li+. Usually an “-ate” complex involves charge balance between an anionic complex and an alkali metal cation, but here even the neutral cobalt complex LCo(PEt3)2Cl incorporates two neutral LiOTf, with Li+ and Cl−, apparently a carryover impurity from the synthesis of LCo(PEt3)2 from (H2L)CoCl2 and LiN(SiMe3)2.13 The yield of product improves if equimolar chloride (as LiCl; see below) is added intentionally as a trapping nucleophile. Oxidations in THF without LiCl lead to a mixture of diamagnetic and paramagnetic Co(III) products, demonstrating the high reactivity of the transient five-coordinate Co(III) species (Figures S23− S24).The lengths of corresponding bonds in each pyrazolate ring are identical within one esd, confirming C2v symmetry. The Co−Cl distances are not significantly lengthened by the two lithium electrophiles compared to [(HL)Co(PEt3)2Cl+], below. Angles NNC are both 106.5 ± 0.1°, consistent with no protons being on these pyrazolates. With this composition established, a systematic synthesis was attempted by combining LCo(PEt3)2, AgOTf, and (subsequently, after oxidation) LiCl and in a mole ratio of 1:2:2 in THF. Reaction of LCo(PEt3)2 with AgOTf in THF in the dark occurs in less than 1 h with a color change from original dark red to yellow-orange, and with precipitation of gray solid (Ag). Workup, including removal of volatiles followed by trituration with 3 × 5 mL pentane, resulted in an orange solid. The 1H NMR of this orange crude mixture shows both diamagnetic and paramagnetic features in ∼8:1 ratio (Figure S17). Diamagnetic product signals integrate well for C2v symmetry, with ethyl signals of two equivalent phosphines, all consistent with [LCo(PEt3)2Cl(LiOTf)2]2 (Figure S12). The 11% yield paramagnetic product has 1H NMR chemical shifts between +110 and −60 ppm (see Supporting Information), in a pattern consistent with divalent cobalt in this pincer ligand system seen previously without phosphine ligands.13 This paramagnetic product has two distinct tBu signals, and overall correct number and relative intensity of signals for an L2CoII2X where X (Scheme 3) is either a neutral or an anionic ligand bridging the two divalent metals. Some fraction of cobalt thus escapes oxidation under these reaction conditions. The 19F NMR spectrum of the crude material shows a single sharp singlet at −78.4 ppm, which corresponds to SO3CF3− in [LCo(PEt3)2Cl(LiOTf)2]2 (Figure S13). However, the 31P NMR shows (Figures S11 and S18) two sets of signals, one broad at +20 ppm, which is consistent with equivalent phosphine atoms bound to quadrupolar Co (I = 7/2). Another set of peaks appears as two sharp doublets at +13 and +9 ppm, respectively, consistent with PEt3 bound to Ag with 1J107Ag−P = 628 Hz. Silver has two naturally occurring isotopes, 107Ag and 109 Ag, both with I = 1/2, and thus the observed two doublets are in fact due to the difference in the gyromagnetic ratios of the two isotopomers. As a control experiment, reacting AgOTf with two equivalents of PEt3 in THF in the dark reproduces the
The calculations show N2O to be the next most powerful O atom source, followed by the two tertiary amine oxides, and hence pyridine N-oxide is the weakest. The small difference between Me3NO and N-methylmorpholine N-oxide is probably not significant at our level of calculation, so we consider them as essentially equally potent oxo transfer reagents; the practical difference is that the morpholine ring confers desirably increased solubility vs Me3NO. The energy difference between axial and equatorial O (not shown) in NMO is 4.4 kcal/mol, favoring O axial. For comparison, we also give reaction energies for the weakest of the amine oxide reagents to oxidize representative sulfur and phosphorus species. pyNO + Me2S → py + Me2SO
pyNO + PPh3 → py + OPPh3
ΔE(SCF) = − 12.9
ΔE(SCF) = − 59.1
These confirm that amine oxides are much more potent O atom transfer reagents than Me2SO or OPPh3, and that phosphines (specifically PPh3) are more powerful reductants than thioethers (specifically Me2S). This is consistent with the general experience that phosphines are readily oxidized.11 With reactivity with O atom transfer agents ascertained, we next explored single electron oxidation of LCo(PEt3)2, for comparison to the result with NMO. The oxidation with Epa of −0.54 vs Fc/Fc+ in Figure 1 is accomplished by either Cp2Fe+ or Ag+, so we next attempted these oxidations chemically, to identify products, and find the chemical follow-up origin of the observed voltammetric irreversibility. Oxidation of LCoII(PEt3)2 by AgOTf: Synthesis and Characterization of [LCo(PEt3)2Cl(LiOTf)2]2. Reaction of LCoII(PEt3)2 with two equivalents of AgOTf, in pursuit of CoIV or L−1•CoIII, in THF in the dark occurs in less than 1 h with a color change from original dark red to yellow-orange, and with precipitation of gray solid (Ag). Formation of crystals begins within 1 h of initiating the reaction. Single crystal X-ray diffraction structure determination establishes (Figure 3) formula [LCo(PEt3)2Cl(LiOTf)2]2, hence an adduct of LCoIII(PEt3)2Cl with two LiOTf, all of which dimerizes. Each Li binds to a different pyrazolate nitrogen, and also to the same
Figure 3. ORTEP view (50% probabilities) of the non-hydrogen atoms of [LCo(PEt3)2Cl(LiOTf)2]2. Green is Cl and yellow is S. The unlabeled half of the molecule is related to the first by a center of inversion. Fluorines and ethyl groups have been removed for clarity and unlabeled atoms are carbon. Selected structural parameters: Co1− N3, 1.8708(17) Å; Co1−N5, 1.9057(17); Co1−N1, 1.9150(17); Co1−Cl1, 2.2520(6); Co1−P1, 2.3158(6); Co1−P2, 2.3254(7); Cl1− Li2, 2.516(4); Cl1−Li1, 2.543(4); N5−Co1−N1, 163.63(8); N3− Co1−Cl1, 177.96(6); N5−Co1−Cl1, 98.24(6); P1−Co1−P2, 177.49(2). D
DOI: 10.1021/acs.inorgchem.8b00816 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry Scheme 3. Dimerization of LCo Subunits, Linked by a “X” Bridge Ligand
observed 31P NMR doublets centered at 11 ppm, albeit at a lower resolution of those doublets (Figure S15). Thus, multinuclear NMR techniques suggest the formation of Ag(PEt3)2OTf, which coexists in solution with some unoxidized CoII. Since this unoxidized CoII has lost some of its phosphine ligands to Ag+, this explains the observed low yield paramagnetic 1H NMR species. We suggest that the Ag+ with two attached phosphines lacks the power (Ag+ is a weaker oxidant in coordinating MeCN than in CH2Cl2 and THF14) to oxidize P-deficient divalent cobalt, hence accounting for the paramagnetic side product. Using a 3:1 Ag/Co ratio in the synthesis yields still smaller quantities of the paramagnetic side product. This supports the hypothesis that Ag+ oxidant scavenging by phosphine is indeed the cause of the residual Co(II) and thus explains the 1H NMR observation of some L2Co2X side product. A Product of Outer Sphere Electron Transfer: Oxidation by Cp2Fe+. We next turned out attention to outer-sphere oxidants, as that hypothetically would diminish binding of excess oxidant to Co-containing products as well as diminish PR3 scavenging. Accordingly, oxidation of LCo(PEt3)2 by [Cp2Fe]PF6 (1:1 mol ratio) in CH2Cl2 is complete in less than 5 min and gives a diamagnetic product. The NMR spectra (Figure S6−S8) are consistent with a diamagnetic molecule with only a mirror plane relating the two phosphines; the 31P NMR spectrum shows one chemical shift (trans phosphines) with evidence of quadrupolar broadening of this signal by cobalt; also seen is the multiplet due to PF6−. The 1H NMR shows two tBu signals of equal intensity (hence inequivalent pincer arms) and one unit intensity NH proton at ∼8.7 ppm, together with other pyrazolate and pyridyl resonances, all of intensity one. Single crystal X-ray diffraction reveals (Figure 4) the product to be comprised of [(HL)Co(PEt3)2Cl+] cations and PF6− anions, with octahedral environment around cobalt. The single NH proton explains the NMR inequivalence of the two pincer arms. This proton was evident in the diffraction data, but also because cations are arranged pairwise around a crystallographic center of symmetry, with a head to tail arrangement of N2H··· Cl hydrogen bonding; the N2/Cl distance is 3.31 Å and the H/ Cl distance is 2.41 Å. The NNC angle at N4 is 6° smaller (105.4°) than it is at the protonated nitrogen N2 (111.4°). It has been shown9,15 that this is a general characteristic which can be used to support the presence or absence of a proton in pyrazole/pyrazolate. Also consistent with only one proton on the pincer ligand, the Co/N(pyrazolate) distances are inequivalent in [(HL)Co(PEt3)2Cl+], while they are equivalent in [LCo(PEt3)2Cl(LiOTf)2]2. The single NH proton makes that pyrazole a weaker donor to cobalt, with that CoN distance longer by 0.040(5) Å. Trivalent cobalt here is thus six coordinate since the metal abstracts chloride, whose source was next identified as CH2Cl2.
Figure 4. ORTEP view (50% probabilities) of the non-hydrogen atoms of [(HL)Co(PEt3)2Cl]+ as its PF6− salt. Unlabeled atoms are carbon, and the one pyrazole hydrogen is shown, on N2. Selected structural parameters: Co1−N3, 1.890(5) Å; Co1−N5, 1.906(5); Co1−N1, 1.946(4); Co1−Cl1, 2.2465(16); Co1−P2, 2.2948(16); Co1−P1, 2.3180(16); N5−Co1−N1, 162.0(2)°; N3−Co1−Cl1, 177.71(14); N5−Co1−Cl1, 100.99(15); P2−Co1−P1, 172.51(6).
The characterized complex is the formal product of one electron oxidation of LCo(PEt3)2, but with HCl added; this is confirmed in the mass of the molecular ion detected by ESI(+) mass spectrometry. Repeating the oxidation in CD2Cl2 yields an ESI mass spectrum showing a molecular ion of mass heavier by one unit than when oxidized in CH2Cl2 (Figures S20−S22). This proves that the source of the proton is solvent and provides support for electrophilic attack on solvent as the source of the observed product: Outer sphere single electron oxidation in nondonor solvent produces LCo(PEt3)2+, whose high electrophilicity binds solvent by a chloride lone pair (Scheme 4). The coordinated solvent has enhanced Bronsted acidity, and the nearby pyrazolate nitrogen deprotonates coordinated CH2Cl2 concurrent with heterolysis of the C/Cl bond to liberate transient carbene, CHCl, which would rapidly react with neighboring solvent. Abstraction of chloride from CH2Cl2 has precedence, but not the concurrent abstraction of H and Cl.16 We suggest that the (HL)Co(PEt3)2Cl]+ species is Scheme 4. Concerted FLP-like HCl Abstraction from CH2Cl2 by [LCo(PEt3)2]+
E
DOI: 10.1021/acs.inorgchem.8b00816 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry rapidly formed upon electrochemical oxidation (Figure 1) of LCo(PEt3)2, followed by HCl abstraction from solvent, which is faster than the CV time scale. We next describe the fate of the Cp2Fe+ reagent in this oxidation of LCo(PEt3)2. A Competitive Reaction: P−C Bond Formation from Nucleophilic Attack on Ferrocenium Carbon. The presence, in LCo(PEt3)2, of free PEt3 in solution complicates the above reaction with Cp2Fe+. When the reaction of [Cp2Fe]PF6 with LCo(PEt3)2 was initially run at a 2:1 Fe/Co ratio, a surprising low yield product crystallizes with [(HL)Co(PEt3)2Cl]PF6 as an orange solid, and was identified (see structural study below) as the phosphonium species CpFe(C5H4PEt3)+, isolated as its PF6− salt. It was physically separated from [(HL)Co(PEt3)2Cl+]PF6 in a mixture of crystals. Prior synthetic approaches to ferrocenes with a phosphonium substituent directly on the ring invariably involve an electrophilic attack on a ferrocenylphosphine CpFe(C5H4PR2), so do not involve electron transfer.17 We suspected that this product might arise from reaction of electron rich free phosphine with electrophilic ferrocenium oxidant, by nucleophilic attack on ring carbon, and then this transfers a hydrogen atom to some receptor. We have tested this experimentally and find indeed a reaction occurs in time of mixing of dark blue [Cp2Fe]PF6 (a pale blue solution together with undissolved solid) with PEt3 (1:1 stoichiometry) in THF, with color change to orange. The 1H NMR spectrum of this reaction solution (redissolved in CD3CN, see Figures S3 and S4) shows signals for ethyl groups, along with intensity 2:2:5 C5 ring protons, while the 31P NMR spectrum shows complete consumption of PEt3 at this 1:1 stoichiometry to give a singlet 31P NMR signal, together with a signal for PF6−; the 19F NMR spectrum shows only the doublet of PF6−. Free ferrocene was also observed (1H NMR). The ESI mass spectrum of the reaction solution in THF shows the ion CpFe(C5H4PEt3)+ with proper isotopomer multiplet intensities, all of which is consistent with the product of this reaction being CpFe(C5H4PEt3)+. This cation identity was confirmed by a single crystal X-ray structure determination (Figure 5). The angle between the two C5 ring planes in a given cation is 4.5 and 5.6°. The phosphonium phosphorus lies only 0.16 Å from its respective ring plane; the PC bond makes an angle of 5.3° to the five carbon plane. The 10 Fe/C distances
show no significant variation with the presence or absence of phosphonium substituent and vary by 3° C−H Bond Selectivity and a Novel Bis(μ-alkylperoxo)dicopper Intermediate. Inorg. Chem. 1998, 37 (9), 2102−2103. (b) Lintvedt, R. L.; Ranger, G.; Ceccarelli, C. Reactions of Coordinated βpolyketonate ligands. 2. Ligand Oxidation and Benzilic Acid Type Rearrangement in the Nickel(II) Complex of 2,2-dimethyl-3,5,7octanetrione. Molecular Structure of the Binuclear Nickel(II) Complex of the Resultant 2-tert-butyl-2-hydroxy-3,5-dioxohexanoic acid. Inorg. Chem. 1985, 24 (15), 2359−2363. (c) Sikari, R.; Sinha, S.; Jash, U.; Das, S.; Brandão, P.; de Bruin, B.; Paul, N. D. Deprotonation Induced Ligand Oxidation in a NiII Complex of a Redox Noninnocent N1-(2Aminophenyl)benzene-1,2-diamine and Its Use in Catalytic Alcohol H
DOI: 10.1021/acs.inorgchem.8b00816 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry Oxidation. Inorg. Chem. 2016, 55 (12), 6114−6123. (d) Carson, E. C.; Lippard, S. J. Oxidation of Sulfide, Phosphine, and Benzyl Substrates Tethered to N-Donor Pyridine Ligands in Carboxylate-Bridged Diiron(II) Complexes. J. Am. Chem. Soc. 2004, 126 (11), 3412− 3413. (e) Kumar, M.; Colpas, G. J.; Day, R. O.; Maroney, M. J. Ligand Oxidation in a Nickel Thiolate Complex: A Model for the Deactivation of Hydrogenase by Oxygen. J. Am. Chem. Soc. 1989, 111 (21), 8323− 8325. (f) Reynolds, A. M.; Lewis, E. A.; Aboelella, N. W.; Tolman, W. B. Reactivity of a 1:1 Copper-Oxygen Complex: Isolation of a Cu(ii)o-iminosemiquinonato Species. Chem. Commun. 2005, No. 15, 2014− 2016. (2) (a) Hong, S.; Pfaff, F. F.; Kwon, E.; Wang, Y.; Seo, M.-S.; Bill, E.; Ray, K.; Nam, W. Spectroscopic Capture and Reactivity of a Low-Spin Cobalt(IV)-Oxo Complex Stabilized by Binding Redox-Inactive Metal Ions. Angew. Chem., Int. Ed. 2014, 53 (39), 10403−10407. (b) Wang, B.; Lee, Y.-M.; Tcho, W.-Y.; Tussupbayev, S.; Kim, S.-T.; Kim, Y.; Seo, M. S.; Cho, K.-B.; Dede, Y.; Keegan, B. C.; Ogura, T.; Kim, S. H.; Ohta, T.; Baik, M.-H.; Ray, K.; Shearer, J.; Nam, W. Synthesis and Reactivity of a Mononuclear Non-haem Cobalt(IV)-oxo Complex. Nat. Commun. 2017, 8, 14839. (c) Hadt, R. G.; Hayes, D.; Brodsky, C. N.; Ullman, A. M.; Casa, D. M.; Upton, M. H.; Nocera, D. G.; Chen, L. X. X-ray Spectroscopic Characterization of Co(IV) and Metal− Metal Interactions in Co4O4: Electronic Structure Contributions to the Formation of High-Valent States Relevant to the Oxygen Evolution Reaction. J. Am. Chem. Soc. 2016, 138 (34), 11017−11030. (d) Brodsky, C. N.; Hadt, R. G.; Hayes, D.; Reinhart, B. J.; Li, N.; Chen, L. X.; Nocera, D. G. In situ Characterization of Cofacial Co(IV) Centers in Co4O4 Cubane: Modeling the High-Valent Active Site in Oxygen-Evolving Catalysts. Proc. Natl. Acad. Sci. U. S. A. 2017, 114 (15), 3855−3860. (3) (a) Yoshida, J.; Kondo, S.; Yuge, H. A Synthetic Strategy for a New Series of Oxo-Centered Tricobalt Complexes with Mixed Bridging Ligands of Acetate and Pyrazolate Snions. Dalton Trans. 2013, 42 (7), 2406−2413. (b) Kim, D.; Cho, J.; Lee, Y.-M.; Sarangi, R.; Nam, W. Synthesis, Characterization, and Reactivity of Cobalt(III)−Oxygen Complexes Bearing a Macrocyclic N-Tetramethylated Cyclam Ligand. Chem. - Eur. J. 2013, 19 (42), 14112−14118. (c) Tiné, M. R. Cobalt Complexes in Aqueous Solutions as Dioxygen Carriers. Coord. Chem. Rev. 2012, 256 (1), 316−327. (d) Suzuki, M.; Furutachi, H.; O̅ kawa, H. Bimetallic Dioxygen Complexes Derived from ‘end-off’ Compartmental Ligands. Coord. Chem. Rev. 2000, 200−202, 105−129. (e) Crandell, D. W.; Ghosh, S.; Berlinguette, C. P.; Baik, M.-H. How a [CoIV=O]2+ Fragment Oxidizes Water: Involvement of a Biradicaloid [CoII−(·O·)]2+ Species in Forming the O−O Bond. ChemSusChem 2015, 8 (5), 844−852. (f) Corcos, A. R.; Villanueva, O.; Walroth, R. C.; Sharma, S. K.; Bacsa, J.; Lancaster, K. M.; MacBeth, C. E.; Berry, J. F. Oxygen Activation by Co(II) and a Redox Non-Innocent Ligand: Spectroscopic Characterization of a Radical−Co(II)−Superoxide Complex with Divergent Catalytic Reactivity. J. Am. Chem. Soc. 2016, 138 (6), 1796−1799. (4) Stauber, J. M.; Bloch, E. D.; Vogiatzis, K. D.; Zheng, S.-L.; Hadt, R. G.; Hayes, D.; Chen, L. X.; Gagliardi, L.; Nocera, D. G.; Cummins, C. C. Pushing Single-Oxygen-Atom-Bridged Bimetallic Systems to the Right: A Cryptand-Encapsulated Co−O−Co Unit. J. Am. Chem. Soc. 2015, 137 (49), 15354−15357. (5) Cook, B. J.; Chen, C. H.; Caulton, K. G. A Multifunctional Pincer Ligand for Cobalt-Promoted Oxidation by N2O. Chem. - Eur. J. 2018, 24 (22), 5962−5966. (6) Cook, B. J.; Chen, C.-H.; Pink, M.; Lord, R. L.; Caulton, K. G. Coordination and Electronic Characteristics of a Nitrogen Heterocycle Pincer Ligand. Inorg. Chim. Acta 2016, 451, 82−91. (7) Addison, A. W.; Rao, T. N.; Reedijk, J.; van Rijn, J.; Verschoor, G. C. Synthesis, Structure, and Spectroscopic Properties of Copper(II) Compounds Containing Nitrogen-Sulfur Donor Ligands; The Crystal and Molecular Structure of Aqua[1,7-bis(N-methylbenzimidazol-2′yl)-2,6-dithiaheptane]copper(II) perchlorate. J. Chem. Soc., Dalton Trans. 1984, No. 7, 1349−1356. (8) (a) Jin, S.; Nieuwenhuyzen, M.; Robinson, W. T.; Wilkins, C. J. Structures of Dichlorobis(trimethylamine oxide)cobalt(II), Diiodobis-
(trimethylamine oxide)cobalt(II), Dichlorobis(trimethylamine oxide) zinc and Dibromobis(trimethylamine oxide)zinc. Acta Crystallogr., Sect. C: Cryst. Struct. Commun. 1992, 48 (2), 274−279. (b) Jin, S.; Nieuwenhuyzen, M.; Wilkins, C. J. Ionic Complexes from Cobalt and Zinc Halides with Pyridine N-oxide and Trimethylamine Oxide. J. Chem. Soc., Dalton Trans. 1992, No. 13, 2071−2078. (9) Yoshinari, A.; Tazawa, A.; Kuwata, S.; Ikariya, T. Synthesis, Structures, and Reactivities of Pincer-Type Ruthenium Complexes Bearing Two Proton-Responsive Pyrazole Arms. Chem. - Asian J. 2012, 7 (6), 1417−25. (10) (a) Zhou, A.; Prakash, J.; Rohde, G. T.; Klein, J. E. M. N.; Kleespies, S. T.; Draksharapu, A.; Fan, R.; Guo, Y.; Cramer, C. J.; Que, L. The Two Faces of Tetramethylcyclam in Iron Chemistry: Distinct Fe−O−M Complexes Derived from [FeIV(Oanti/syn)(TMC)]2+ Isomers. Inorg. Chem. 2017, 56 (1), 518−527. (b) Puri, M.; Biswas, A. N.; Fan, R.; Guo, Y.; Que, L. Modeling Non-Heme Iron Halogenases: High-Spin Oxoiron(IV)−Halide Complexes That Halogenate C−H Bonds. J. Am. Chem. Soc. 2016, 138 (8), 2484− 2487. (c) Monte Pérez, I.; Engelmann, X.; Lee, Y.-M.; Yoo, M.; Kumaran, E.; Farquhar, E. R.; Bill, E.; England, J.; Nam, W.; Swart, M.; Ray, K. A Highly Reactive Oxoiron(IV) Complex Supported by a Bioinspired N3O Macrocyclic Ligand. Angew. Chem., Int. Ed. 2017, 56 (46), 14384−14388. (d) Stubbs, A. W.; Braglia, L.; Borfecchia, E.; Meyer, R. J.; Román- Leshkov, Y.; Lamberti, C.; Dincă, M. Selective Catalytic Olefin Epoxidation with MnII-Exchanged MOF-5. ACS Catal. 2018, 8 (1), 596−601. (e) de Ruiter, G.; Carsch, K. M.; Takase, M. K.; Agapie, T. Selectivity of C−H versus C−F Bond Oxygenation by Homo- and Heterometallic Fe4, Fe3Mn, and Mn4 Clusters. Chem. Eur. J. 2017, 23 (45), 10744−10748. (11) (a) Holm, R. H.; Donahue, J. P. A Thermodynamic Scale for Oxygen Atom Transfer Reactions. Polyhedron 1993, 12 (6), 571−589. (b) Holm, R. H. Metal-Centered Oxygen Atom Transfer Reactions. Chem. Rev. 1987, 87 (6), 1401−1449. (12) Cook, B. J.; Polezhaev, A. V.; Chen, C.-H.; Pink, M.; Caulton, K. G. Deprotonation, Chloride Abstraction, and Dehydrohalogenation as Synthetic Routes to Bis-Pyrazolate Pyridyl Iron(II) Complexes. Eur. J. Inorg. Chem. 2017, 2017 (34), 3999−4012. (13) Cook, B. J.; Chen, C.-H.; Pink, M.; Caulton, K. G. Dehydrohalogenation of Proton Responsive Complexes: Versatile Aggregation via Pyrazolate Pincer Ligand Arms. Dalton Trans. 2018, 47, 2052−2060. (14) (a) Connelly, N. G.; Geiger, W. E. Chemical Redox Agents for Organometallic Chemistry. Chem. Rev. 1996, 96 (2), 877−910. (b) Geiger, W. E. Reflections on Future Directions in Organometallic Electrochemistry†. Organometallics 2011, 30 (1), 28−31. (15) (a) Toda, T.; Kuwata, S.; Ikariya, T. Unsymmetrical PincerType Ruthenium Complex Containing β-Protic Pyrazole and NHeterocyclic Carbene Arms: Comparison of Brønsted Acidity of NH Groups in Second Coordination Sphere. Chem. - Eur. J. 2014, 20 (31), 9539−9542. (b) Kuwata, S.; Ikariya, T. Metal-Ligand Bifunctional Reactivity and Catalysis of Protic N-heterocyclic Carbene and Pyrazole Complexes Featuring β-NH units. Chem. Commun. 2014, 50 (92), 14290−14300. (16) (a) Liaw, B.-J.; Lobana, T. S.; Lin, Y.-W.; Wang, J.-C.; Liu, C. W. Versatility of Dithiophosphates in the Syntheses of Copper(I) Complexes with Bis(diphenylphosphino)alkanes: Abstraction of Chloride from Dichloromethane. Inorg. Chem. 2005, 44 (26), 9921− 9929. (b) He, W.; Xie, L.; Xu, Y.; Xiang, J.; Zhang, L. Electrophilicity of α-oxo Gold Carbene Intermediates: Halogen Abstractions from Halogenated Solvents Leading to the Formation of Chloro/ Bromomethyl Ketones. Org. Biomol. Chem. 2012, 10 (16), 3168− 3171. (c) Tellers, D. M.; Bergman, R. G. Binding of Chlorohydrocarbons to Metal Centers: Quantitative Evaluation of Relative Binding Constants and Structural Characterization of the First Isolable Transition Metal−Chloromethane Adduct. J. Am. Chem. Soc. 2001, 123 (46), 11508−11509. (d) Broere, D. L. J.; van Leest, N. P.; de Bruin, B.; Siegler, M. A.; van der Vlugt, J. I. Reversible Redox Chemistry and Catalytic C(sp3)−H Amination Reactivity of a I
DOI: 10.1021/acs.inorgchem.8b00816 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry Paramagnetic Pd Complex Bearing a Redox-Active o-AminophenolDerived NNO Pincer Ligand. Inorg. Chem. 2016, 55 (17), 8603−8611. (17) (a) McEwen, W. E.; Sullivan, C. E.; Day, R. O. Crystal and Molecular Structure of Benzyldiphenylferrocenylphosphonium Chloride: Through-Space Overlap Effects in the Wittig Reaction. Organometallics 1983, 2 (3), 420−425. (b) Ramos, A.; Lough, A. J.; Stephan, D. W. Activation of H2 by Frustrated Lewis Pairs Derived from monoand bis-phosphinoferrocenes and B(C6F5)3. Chem. Commun. 2009, 7 (9), 1118−1120. (c) Imrie, C.; Modro, T. A.; Van Rooyen, P. H. Structural Study of Arylmethyldiphenylferrocenylphosphonium salts. Polyhedron 1994, 13 (11), 1677−1682. (d) Olah, G. A.; McFarland, C. W. Organophosphorus Compounds. XI. 1H and 31P Nuclear Magnetic Resonance Studies of the Protonation of Phosphines. J. Org. Chem. 1969, 34 (6), 1832−1834. (e) Iio, H.; Fujii, A.; Ishii, M.; Tokoroyama, T. Enantioselective Vinylation of Aldehydes with Chiral βsilylphosphorus ylides. J. Chem. Soc., Chem. Commun. 1991, 19, 1390−1392. (18) Hong, S.; Gupta, A. K.; Tolman, W. B. Intermediates in Reactions of Copper(I) Complexes with N-Oxides: From the Formation of Stable Adducts to Oxo Transfer. Inorg. Chem. 2009, 48 (14), 6323−6325. (19) (a) Heyduk, A. F.; Krodel, D. J.; Meyer, E. E.; Nocera, D. G. A Luminescent Heterometallic Dirhodium−Silver Chain. Inorg. Chem. 2002, 41 (4), 634−636. (b) Klüfers, P. Die Kristallstruktur von AgCo(CO)4. In Zeitschrift für Kristallographie - Crystalline Materials, 1984; Vol. 166, p 143. (c) Blundell, T. L.; Powell, H. M. Stereochemistry of Cobalt Complexes with Heteronuclear MetalMetal Bonds: the Crystal and Molecular Structure of Tetracarbonyl(triphenylphosphineaurio)cobalt, and of Tetracarbonyl{[bis-(odimethylarsinophenyl)methylarsine]argentio}cobalt. J. Chem. Soc. A 1971, 0, 1685−1690. (d) Connelly, N. G.; Lucy, A. R.; Galas, A. M. R. Synthesis, reactivity, and X-ray crystal structure of [Ag{Rh(CO)(PPh3)(η5-C5H5)}2]-[PF6].toluene: A Silver-Stabilised Radical Cation. J. Chem. Soc., Chem. Commun. 1981, 2, 43−44. (e) Garcia, M. P.; Jimenez, M. V.; Lahoz, F. J.; Oro, L. A. Synthesis and Reactivity of Mononuclear Anionic Pentafluorophenyl Compounds of Rhodium(I) and Iridium(I). X-ray Structure of [{P(OPh)3}2(C6F5)2RhAg(PPh3)]. Inorg. Chem. 1995, 34 (8), 2153−2159. (f) Stephan, D. W.; Erker, G. Angew. Chem., Int. Ed. 2015, 54, 6400. (20) Zimmerman, R. L. Process for the preparation of nmethylmorpholine oxide. US5216154A, 1993. (21) Jaguar 7.7; Schrödinger, LLC: New York, NY, 2007. (22) (a) Becke, A. D. Density-functional Exchange-Energy Approximation with Correct Asymptotic Behavior. Phys. Rev. A: At., Mol., Opt. Phys. 1988, 38, 3098−3100. (b) Becke, A. D. DensityFunctional Thermochemistry. III. The Role of Exact Exchange. J. Chem. Phys. 1993, 98, 5648−5652. (c) Vosko, S. H.; Wilk, L.; Nusair, M. Accurate Spin-Dependent Electron Liquid Correlation Energies for Local Spin Density Calculations: A Critical Analysis. Can. J. Phys. 1980, 58, 1200−1211. (d) Lee, C.; Yang, W.; Parr, R. G. Development of the Colle-Salvetti Correlation-Energy Formula into a Functional of the Electron Density. Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 37, 785−789. (23) (a) Hay, P. J.; Wadt, W. R. Ab initio Effective Core Potentials for Molecular Calculations. Potentials for the Transition Metal Atoms Sc to Hg. J. Chem. Phys. 1985, 82, 270−283. (b) Wadt, W. R.; Hay, P. J. Ab initio Effective Core Potentials for Molecular Calculations. Potentials for Main Group Elements Na to Bi. J. Chem. Phys. 1985, 82, 284−298. (24) Kendall, R. A.; Dunning, T. H., Jr.; Harrison, R. J. J. J. Chem. Phys. 1992, 96, 6796−6806.
J
DOI: 10.1021/acs.inorgchem.8b00816 Inorg. Chem. XXXX, XXX, XXX−XXX
