Two mononuclear iron(II) p-semiquinonate (pSQ) complexes have been generated via one-electron reduction of precursor complexes containing a substituted 1,4-naphthoquinone ligand. Detailed spectroscopic and computational analysis confirmed the presence of a coordinated pSQ radical ferromagnetically coupled to the high-spin Fe(II) center. The complexes are intended to model electronic interactions between (semi)quinone and iron cofactors in biology.
Two mononuclear iron(II) p-semiquinonate (pSQ) complexes have been generated via one-electron reduction of precursor complexes containing a substituted 1,4-naphthoquinone ligand. Detailed spectroscopic and computational analysis confirmed the presence of a coordinated pSQ radical ferromagnetically coupled to the high-spin Fe(II) center. The complexes are intended to model electronic interactions between (semi)quinone and iron cofactors in biology.
Substituted p-quinones are common redox-active molecules
that appear in a variety of biological contexts, including respiration,
photosynthesis, and enzymatic transformations.[1] These cofactors participate in reversible one- and two-electron
reactions that are often coupled to proton transfer(s), yielding the
corresponding semiquinone (pSQ•–) and hydroquinone (pHQ) species, respectively.
(Hydro)quinones often function in concert with protein-bound metal
ions, including heme and nonheme iron centers. For instance, cytochrome bc1 (respiratory complex III) catalyzes the two-electron
oxidation of the pHQ cofactor ubiquinol; the first
electron is transferred to an Fe–S (Rieske) cluster and the
second to heme bL.[2] Recent studies with electronic paramagnetic resonance (EPR) spectroscopy
have detected the triplet state arising from ferromagnetic interactions
between the ubisemiquinone radical and reduced Fe–S cluster.[3] Similarly, a nonheme FeII center in
photosystem II (PSII) is associated with two p-quinones
that shuttle electrons away from the photoexcited P680 cofactor,
resulting in short-lived FeII-pSQ•– states than have been detected by EPR.[4] Finally, a superoxo-FeII-pSQ•– intermediate has been proposed in the
catalytic cycle of mononuclear nonheme iron enzymes known as hydroquinone
dioxygenases,[5] which play an important
role in the biodegradation of aromatic pollutants.Despite these
biological precedents, only a handful of synthetic iron complexes
with p-quinone ligands have been crystallographically
characterized, and the majority of these are diiron complexes with
a bridging quinone.[6] The dearth of monoiron
complexes is likely due to the inability of the p-quinone unit to form chelates, combined with the weak donor properties
of carbonyls. To the best of our knowledge, there are no prior examples
of mononuclear iron complexes featuring pSQ•– ligands, although Riordan and co-workers recently generated a five-coordinate
iron(II) complex containing a bidentate o-semiquinonate
ligand.[7]This paper describes the
synthesis and characterization of metastable FeII-pSQ•– complexes with relevance
to the transient intermediates found in proteins. Our approach required
the initial preparation of two monoiron(II) complexes (1 and 2) each containing a juglone-derived ligand (either J or J; Figure 1a). The presence of the phenolatedonor ensures coordination of the 1,4-naphthoquinone moiety via formation
of a favorable six-membered ring chelate. The high-spin FeII center is also bound to the hydrotris(3,5-diphenylpyrazol-1-yl)borate
(Ph2Tp) supporting ligand. Significantly, treatment of 1 and 2 with a suitable one-electron reductant
yielded brown chromophores (1 and 2, respectively) that
are stable at low temperatures. Extensive analysis with spectroscopic
[UV–vis absorption, EPR, and resonance Raman (rR)] and computational
methods, described below, confirmed the formation of mononuclear FeII-pSQ•– species,
the first to be reported in the literature. Our synthetic efforts
thus provide a valuable platform for detailed explorations of iron–(semi)quinone
interactions.
Figure 1
(a) Schematic of complexes 1 and 2. (b) Thermal ellipsoid diagram of 1. H atoms and 5-Ph
substituents of the Ph2Tp ligand have been omitted for
clarity.
(a) Schematic of complexes 1 and 2. (b) Thermal ellipsoid diagram of 1. H atoms and 5-Ph
substituents of the Ph2Tp ligand have been omitted for
clarity.Complexes 1 and 2 were prepared by mixing [Fe2+(Ph2Tp)(OBz)]
with equimolar amounts of NaOMe and the appropriate juglone (JH or JH) in CH2Cl2 under anaerobic
conditions. X-ray-quality crystals were obtained by layering 1,2-dichloroethane
solutions with either MeCN (1) or pentane (2). In each case, the resulting structure revealed a pentacoordinate
FeII center bound to a facially coordinating Ph2Tp ligand and bidentate J anion (Figure 1b and Table S1 in the Supporting Information, SI). The observed metric parameters (Table S2 in the SI) are quite similar for the two complexes.
Both structures feature distorted trigonal-bipyramidal coordination
geometries (τ values[8] of 0.63) with
the phenolate (O3) and carbonyl (O2) donors of J in equatorial
and axial positions, respectively. The J ligand binds in an asymmetric
manner, with a short Fe–O3 distance of 1.93 Å and a longer
Fe–O2 distance near 2.07 Å. The average Fe–NTp bond length of ∼2.13 Å is typical of high-spin
(S = 2) ferrous complexes,[9] consistent with the magnetic moments of 4.85(10) μB. The J ligands exhibit short O1–C1 and O2–C4 distances
(1.22 and 1.24 ± 0.01 Å, respectively) characteristic of
quinone moieties.Complex 1 possesses a deep-violet
color in solution because of the presence of two absorption bands
at 510 and 910 nm (ε = 5.3 and 2.0 mM–1 cm–1, respectively; see Figure 2, left). Time-dependent (TD) density functional theory (DFT) calculations
attribute the lower-energy band to a FeII → J metal-to-ligand charge transfer (MLCT) transition, while the higher-energy
feature arises from an intraligand π → π* transition
localized on J (Figure S1 in the SI). Both transitions
red-shift by approximately 700 cm–1 in the spectrum
of 2 (Figure 2) because of the
presence of electron-withdrawing Br substituents.
Figure 2
Red, solid lines: UV–vis
absorption spectra of 1 and 2 in CH2Cl2 at room temperature. Black, dashed lines: Spectra
measured after the addition of 1 equiv of CoCp*2 (1) or CoCp2 (2) at −30 °C
in CH2Cl2 to give the corresponding X species.
Red, solid lines: UV–vis
absorption spectra of 1 and 2 in CH2Cl2 at room temperature. Black, dashed lines: Spectra
measured after the addition of 1 equiv of CoCp*2 (1) or CoCp2 (2) at −30 °C
in CH2Cl2 to give the corresponding X species.To assess the feasibility of generating iron(II) semiquinonate
species via one-electron reduction, the electrochemical properties
of 1 and 2 were examined in CH2Cl2 solutions. In each case, cyclic voltammetry (CV) exhibits
a semireversible one-electron oxidation near 0.15 V versus Fc+/0 (Figure 3), which is assigned to
the Fe2+/3+ redox couple based on our experience with related
complexes.[9] At lower potentials, complex 1 displays a semireversible couple at −1.33 V and an
irreversible peak at −1.97 V. A comparison to redox potentials
measured for 1,4-naphthoquinone in MeCN (−1.11 and −1.74
V)[10] indicates that these low-potential
events arise from successive one-electron reductions of the J ligand.
The corresponding features in the CV of 2 are shifted
to more positive potentials by 0.23 V (Figure 3), reflecting the electron-poor nature of the J ligand.
Figure 3
Cyclic voltammograms
of 1 and 2 in CH2Cl2 with 0.1 M (NBu4)PF6 as the supporting electrolyte.
Having established the redox properties of the iron(II) p-quinone complexes, we proceeded to monitor their reactivity with
chemical reductants. Guided by the CV data, complex 1 was treated with 1 equiv of CoCp*2 (E° = −1.94 V in CH2Cl2), while the
more-easily reduced 2 was treated with CoCp2 (E° = −1.33 V).[11] The addition of these reductants to solutions of 1 and 2 in CH2Cl2 instantly
generates new chromophores (1 and 2, respectively) with
two intense absorption features in the near-UV region (ε values
of 8–10 mM–1 cm–1; Figure 2). In addition, each complex displays a shoulder
near 600 nm and a broad near-IR feature. These features do not arise
from either (decamethyl)cobaltocene or its oxidized derivative (Figure
S2 in the SI). The X species are stable for several minutes under anaerobic
conditions and low temperatures (<−30 °C) but decay
rapidly at room temperature to the starting material. The sizable
intensities and sharp nature of the near-UV bands suggest that they
arise from ligand-based π–π* transitions, consistent
with transient absorption studies of p-semiquinones
that detected strong absorption features in the same region.[12] Moreover, the X absorption spectra bear close similarities to the one reported
for Riordan’s FeII-oSQ•– complex.[7]Cyclic voltammograms
of 1 and 2 in CH2Cl2 with 0.1 M (NBu4)PF6 as the supporting electrolyte.The X-band EPR spectrum of 2, collected at 20 K in frozen
CH2Cl2, consists of two S = 5/2 signals, as is evident by the low-field peaks
at g = 8.6 and 10.1 in Figure 4. Simulation of the 2 data
at multiple temperatures (Figure S3 in the SI) determined that the dominant S = 5/2 species, accounting for 92% of the total intensity, has a
negative D value and moderate rhombicity (E/D = 0.18). The minor S = 5/2 species is highly rhombic, characteristic
of high-spin FeIII ions in low-symmetry environments. This
species accounts for the sharp feature at g = 4.3.
Likewise, the 1 spectrum (Figure
S4 in the SI) has contributions from major
(88%) and minor (12%) S = 5/2 species with spin-Hamiltonian parameters nearly identical with their 2 counterparts. For both species,
we propose that the major S = 5/2 signal corresponds to X,
while the minor signal is attributed to ferric impurities generated
from trace amounts of O2. The EPR results therefore suggest
that the putative pSQ• ligand
is ferromagnetically coupled to the high-spin FeII center.
Figure 4
X-band
EPR spectrum of 2 in frozen
CH2Cl2 at 20 K. The sharp feature (*) arises
from a radical impurity. Simulation of the data required two S = 5/2 species with the following
parameters: major (92%), D = −4.0 cm–1, E/D = 0.18, g = 2.0, 2.0, and 2.1; minor (8%), D = 1.6 cm–1, E/D = 0.33, giso = 2.0.
X-band
EPR spectrum of 2 in frozen
CH2Cl2 at 20 K. The sharp feature (*) arises
from a radical impurity. Simulation of the data required two S = 5/2 species with the following
parameters: major (92%), D = −4.0 cm–1, E/D = 0.18, g = 2.0, 2.0, and 2.1; minor (8%), D = 1.6 cm–1, E/D = 0.33, giso = 2.0.Because we were unable to grow X-ray-quality crystals of 1 and 2, structural models of the reduced species with S = 5/2 were derived from DFT calculations;
Table S3 in the SI compares the metric
parameters for the 1/1 and 2/2 pairs of geometry-optimized models. One-electron reduction of 1 to 1 causes the
O2–C4 and O1–C1 bonds to elongate by 0.049 and 0.033
Å, respectively. In addition, the quinoidal character of the
ligand is diminished because of lengthening of the C2–C3 bond
(by 0.037 Å) and comparable shortening of the C1–C2 and
C3–C4 bonds. Similar differences in the ligand bond lengths
are observed for the 2/2 pair (Table S3 in the SI). Such
dramatic changes in the O–C and C–C bond distances upon
conversion of X → X are indicative of a change in the ligand oxidation state from
a closed-shell quinone to a semiquinone radical.[13] This conclusion is further supported by the Mulliken spin
populations of the iron center (3.78) and the J ligand (1.20) in the 1 model.In agreement with the EPR data, broken-symmetry
(BS) calculations found that the iron- and ligand-based spins are
ferromagnetically coupled, with computed J values
of ∼65 cm–1 (H = −2JSA·SB). Moreover,
absorption spectra computed using TD-DFT nicely reproduce key features
of the experimental spectra (Figure S5 in the SI); in particular, the calculations predict that each X model will display two pSQ-based π → π* transitions in the near-UV
region, resulting in an intense absorption band with λmax near 410 nm. The computed spectra also exhibit a moderate FeII → pSQ• MLCT band
at 530 nm (1) or 645 nm (2).Direct evidence for the
presence of a semiquinonate ligand in the X complexes was provided by rR studies. The rR spectrum
of 1 (Figure 5) displays an intense peak at 1392 cm–1 that
is not evident in the precursor spectrum; 2 exhibits a similar feature at 1382 cm–1 (Figure S6 in the SI). Both peaks are
strongly enhanced by excitation into the near-UV absorption bands.
Coordinated SQ radicals typically display an intense peak in the 1400–1500
cm–1 region that arises primarily from stretching
motions of the C–O bonds.[6e,14] While the
observed X frequencies fall
slightly below this range, the C–O bonds of the semiquinonate
ring are likely weakened by charge donation from the phenolate moiety
of the J ligand. This conclusion is supported by DFT frequency calculations
with the 1 model, which predict
a mode at 1407 cm–1 with mixed ν(O–C)
and ring-breathing character.
Figure 5
rR spectra obtained with 413.1 nm excitation
(40 mW) of 1 (black) and 1 (red) in frozen CD2Cl2 solutions.
Peaks with an asterisk (*) arise from solvent.
rR spectra obtained with 413.1 nm excitation
(40 mW) of 1 (black) and 1 (red) in frozen CD2Cl2 solutions.
Peaks with an asterisk (*) arise from solvent.In summary, two mononuclear iron(II) complexes (1 and 2) featuring juglone-derived ligands have been
prepared and examined with crystallographic and electrochemical methods.
Treatment with strong one-electron reductants gives rise to species X, with coordinated pSQ radicals ferromagnetically coupled to high-spin ferrous centers.
Current efforts are focused on the preparation of species with greater
distances between the iron- and ligand-based spins, thus permitting
studies of long-range FeII–SQ• interactions like those found in PSII.
Authors: Kil Sik Min; Antonio G DiPasquale; James A Golen; Arnold L Rheingold; Atta M Arif; Joel S Miller Journal: J Am Chem Soc Date: 2007-02-02 Impact factor: 15.419
Authors: Kil Sik Min; Antonio G Dipasquale; Arnold L Rheingold; Henry S White; Joel S Miller Journal: J Am Chem Soc Date: 2009-05-06 Impact factor: 15.419
Authors: Kil Sik Min; Krzysztof Swierczek; Antonio G DiPasquale; Arnold L Rheingold; William M Reiff; Atta M Arif; Joel S Miller Journal: Chem Commun (Camb) Date: 2008-01-21 Impact factor: 6.222
Authors: Peng Wang; Michelle M Killian; Mohamed R Saber; Tian Qiu; Glenn P A Yap; Codrina V Popescu; Joel Rosenthal; Kim R Dunbar; Thomas C Brunold; Charles G Riordan Journal: Inorg Chem Date: 2017-08-15 Impact factor: 5.165