Synthesis and characterization of Fe(II) β-diketonato complexes with relevance to acetylacetone dioxygenase: Insights into the electronic properties of the 3-histidine facial triad

Article abstract of DOI:10.1021/ic201115s

A series of high-spin iron(II) β-diketonato complexes have been prepared and characterized with the intent of modeling the substrate-bound form of the enzyme acetylacetone dioxygenase (Dke1). The Dke1 active site features an Fe(II) center coordinated by three histidine residues in a facial geometry-a departure from the standard 2-histidine-1-carboxylate (2H1C) facial triad dominant among nonheme monoiron enzymes. The deprotonated β-diketone substrate binds to the Fe center in a bidentate fashion. To better understand the implications of subtle changes in coordination environment for the electronic structures of nonheme Fe active sites, synthetic models were prepared with three different supporting ligands (L_N3): the anionic ^Me2Tp and ^Ph2Tp ligands (^R2Tp = hydrotris(pyrazol-1-yl)borate substituted with R-groups at the 3- and 5-pyrazole positions) and the neutral ^PhTIP ligand (^PhTIP = tris(2-phenylimidazol-4-yl)phosphine). The resulting [(L_N3) Fe(acac^X)]^0/+ complexes (acac^X = substituted β-diketonates) were analyzed with a combination of experimental and computational methods, namely, X-ray crystallography, cyclic voltammetry, spectroscopic techniques (UV-vis absorption and ¹H NMR), and density functional theory (DFT). X-ray diffraction results for complexes with the ^Me2Tp ligand revealed six-coordinate Fe(II) centers with a bound MeCN molecule, while structures of the ^Ph2Tp and ^PhTIP complexes generally exhibited five-coordinate geometries. Each [(L _N3)Fe(acac^X)]^0/+ complex displays two broad absorption features in the visible region that arise from Fe(II)→acac ^X charge transfer and acac^X-based transitions, consistent with UV-vis data reported for Dke1. These absorption bands, along with the Fe redox potentials, are highly sensitive to the identity of L_N3 and substitution of the β-diketonates. By interpreting the experimental results in conjunction with DFT calculations, detailed electronic-structure descriptions of the complexes have been obtained, with implications for our understanding of the Dke1 active site.

Full text of DOI:10.1021/ic201115s

ARTICLE
pubs.acs.org/IC
Synthesis and Characterization of Fe(II) β-Diketonato Complexes
with Relevance to Acetylacetone Dioxygenase: Insights into the
Electronic Properties of the 3-Histidine Facial Triad
Heaweon Park, Jacob S. Baus, Sergey V. Lindeman, and Adam T. Fiedler*
Department of Chemistry, Marquette University, Milwaukee, Wisconsin 53201-1881, United States
S Supporting Information
b
ABSTRACT: A series of high-spin iron(II) β-diketonato complexes have
been prepared and characterized with the intent of modeling the substrate-
bound form of the enzyme acetylacetone dioxygenase (Dke1). The Dke1
active site features an Fe(II) center coordinated by three histidine residues in
a facial geometry—a departure from the standard 2-histidine-1-carboxylate
(2H1C) facial triad dominant among nonheme monoiron enzymes. The
deprotonated β-diketone substrate binds to the Fe center in a bidentate fashion.
To better understand the implications of subtle changes in coordination
environment for the electronic structures of nonheme Fe active sites,
synthetic models were prepared with three different supporting ligands
(LN3): the anionic Me2Tp and Ph2Tp ligands (R2Tp = hydrotris(pyrazol-1-
yl)borate substituted with R-groups at the 3- and 5-pyrazole positions) and the neutral PhTIP ligand (PhTIP = tris(2-phenylimidazol-
4-yl)phosphine). The resulting [(LN3)Fe(acacX)]0/+ complexes (acacX = substituted β-diketonates) were analyzed with a
combination of experimental and computational methods, namely, X-ray crystallography, cyclic voltammetry, spectroscopic
techniques (UVÀvis absorption and 1H NMR), and density functional theory (DFT). X-ray diffraction results for complexes with
the Me2Tp ligand revealed six-coordinate Fe(II) centers with a bound MeCN molecule, while structures of the Ph2Tp and PhTIP
complexes generally exhibited five-coordinate geometries. Each [(LN3)Fe(acacX)]0/+ complex displays two broad absorption
features in the visible region that arise from Fe(II)facacX charge transfer and acacX-based transitions, consistent with UVÀvis data
reported for Dke1. These absorption bands, along with the Fe redox potentials, are highly sensitive to the identity of LN3 and
substitution of the β-diketonates. By interpreting the experimental results in conjunction with DFT calculations, detailed electronic-
structure descriptions of the complexes have been obtained, with implications for our understanding of the Dke1 active site.
that employ the three histidine (3His) facial triad instead.9 The
1. INTRODUCTION
first member of this class to be structurally characterized was
The widespread presence of pollutants in the environment
remains a major ecological problem and threat to human health.
One viable approach for the restoration of contaminated soils and
cysteine dioxygenase (CDO), an enzyme that catalyzes the first
step in the catabolism of L-cysteine.10 Other 3His enzymes have
since been discovered in bacteria, where they act to degrade
groundwaters utilizes the power of microorganisms to degrade
xenobiotic compounds.11 The enzyme acetylacetone dioxygen-
and detoxify pollutants, a process known as bioremediation.1 The
ase (Dke1), for instance, is one of the few Fe-dependent
aerobic degradation of pollutants relies heavily on mononuclear
dioxygenases capable of oxidatively cleaving aliphatic CÀC
nonheme iron dioxygenases found in the catabolic pathways of
bonds.12À14 Dke1 allows Acinetobacter johnsonii to convert the
bacteria.2 Examples include the Rieske dioxygenases,3 extradiol
toxic and prevalent pollutant acetylacetone to acetic acid and
catechol dioxygenases,4 and (chloro)hydroquinone dioxygenases.5
2-oxopropanal. X-ray diffraction (XRD) studies confirmed that
These enzymes share a common active-site structure in which the
Fe(II) center is coordinated to one aspartate (or glutamate) and
the metal center in Dke1 is facially coordinated by three His
residues and presumably 2À3 H2O molecules, although these
two histidine residues in a facial array; two or three bound H2O
were not resolved in the structure.15,16 While the active site can
molecules are also found in the resting states.6,7 The advantage of
bind several first-row transition metal ions, only Fe(II) results in
this 2-His-1-carboxylate (2H1C) motif is that it permits the Fe
center to bind both substrate and O2 at adjacent coordination sites
catalytic activity.12 Spectroscopic and computational studies
indicate that the substrate coordinates to Fe as the deprotonated
in an ordered mechanism.8 The endogenous ligands also play an
β-diketone (acac) in a bidentate manner.14,15
important role in modulating the electronic structure of the active
site, particularly the redox potential of the Fe center.
Despite the predominance of the 2H1C motif, a new class of
mononuclear nonheme Fe dioxygenases has recently emerged
Received: May 25, 2011
Published: October 28, 2011
r
2011 American Chemical Society
11978
dx.doi.org/10.1021/ic201115s Inorg. Chem. 2011, 50, 11978–11989
|

Inorganic Chemistry
ARTICLE
Scheme 1
facial triad; pyrazole rings have different electronic properties
than histidines (i.e., imidazoles), and the overall negative charge
of the Tp ligand contrasts with the neutral 3His set of the
enzyme. For such reasons, we have also pursued the tris(2-
phenylimidazol-4-yl)phosphine ligand (PhTIP; see Scheme 1) to
more faithfully replicate the charge and donor strength of the
3His coordination environment. Related TIP ligands were first
developed to model the active sites of Zn-containing carbonic
anhydrase and various Cu systems.22,23 Iron complexes withthese
trisimidazole ligands are relatively rare, though, with all such examp-
les involving homoleptic [Fe(TIP)2]2+/3+ complexes24,25 or car-
boxylate-bridged multinuclear species.25,26
In an effort to better understand the significance of the 3His
triad for Dke1, we have synthesized a series of Fe(II)-acacX
complexes featuring the three supporting ligands (LN3) shown in
Scheme 1: Me2Tp, Ph2Tp, and PhTIP. As noted above, the R2Tp
and PhTIP ligands each reproduce the facial N3 coordination
environment of the Dke1 active site, yet they have important
differences with respect to charge and electronic properties that
resemble those differences between the 2H1C and 3His triads.
Since a previous study by Solomon and co-workers suggested
that the acac-bound Dke1 site partially retains a bound H2O
ligand,27 we employed both Me2Tp and Ph2Tp ligands to generate
six-coordinate (6C) and five-coordinate (5C) complexes, respec-
tively. In addition to the natural Dke1 substrate (acac), our
models were prepared with acacX ligands featuring bulky and/or
electron-withdrawing substituents (Scheme 1) to evaluate the
effect such variations on the structural and spectroscopic features
of the resulting complexes. Each complex was characterized with
X-ray crystallography, cyclic voltammetry, and electronic absorp-
The emergence of the 3His family of Fe dioxygenases raises a
pertinent question: what is the significance of variations in Fe
coordination environment for the electronic structure and
catalytic activity of dioxygenases? Interestingly, a mutant of
Dke1 in which the His104 ligand was replaced with Glu was
able to partially bind Fe (∼30% of wild type) yet exhibited no
catalytic activity.16 Thus, the 2H1C and 3His motifs are not
functionally interchangeable, yet it remains unclear exactly how
these ligand-sets tune the catalytic properties of their respective
enzymes. Another important question concerns the mechanism
of oxidative CÀC bond cleavage in Dke1. On the basis of kinetic
data, Straganz and co-workers have proposed that the mechan-
ism proceeds via concerted addition of O2 to give an alkylper-
oxidate intermediate.14 Such a mechanism would resemble the
one proposed for the intradiol catechol dioxygenases in which
the role of the Fe center is to activate the substrate, not O2.17 Yet
a more conventional O2-activation mechanism has also been
suggested that involves initial formation of an Fe-superoxo
intermediate, followed by reaction with bound substrate.15,18
These fundamental questions concerning the structure and
function of Dke1 can be addressed, in part, through the
development of synthetic model complexes. Two Fe-acacX
complexes related to the Dke1 active site have been previously
reported. Several years prior to the discovery of Dke1,
Kitajima et al. published the synthesis and X-ray structure of
[(iPr2Tp)Fe(acac)(MeCN)], where iPr2Tp = hydrotris(3,5-
diisopropylpyrazol-1-yl)borate(À1).19 In 2008, Siewert and
Limberg prepared [(Me2Tp)Fe(Phmal)] (Phmal = anion of
diethyl phenylmalonate) and demonstrated that reaction with
O2 at room temperature in MeCN resulted in dioxygenolytic
ring cleavage of the bound ligand.18 Thus, both Dke1 models
reported to date utilize R2Tp ligands (Scheme 1), which have been
widely employed to replicate the 2H1C facial triad.7,20,21 However,
these pyrazole-based ligands have limitations as mimics of the 3His
1
tion and H NMR spectroscopies. Density functional theory
(DFT) calculations were also performed to examine the effects of
ligand charge and coordination number on Fe/ligand bond-
ing interactions. This combined experimental and computational
approach has provided detailed insights into the electronic struc-
tures of the synthetic Fe(II)-acacX complexes and, by extension,
the Dke1 active site.
2. EXPERIMENTAL SECTION
Materials. All reagents and solvents were purchased from commercial
sources and used as received unless otherwise noted. Acetonitrile (MeCN),
dichloromethane, and tetrahydrofuran (THF) were purified and dried using
a Vacuum Atmospheres solvent purification system. The LN3 supporting
ligands K(Me2Tp),28 K(Ph2Tp),29 and PhTIP23 were prepared according to
literature procedures. The synthesis and handling of air-sensitive materials
were carried out under inert atmosphere using a Vacuum Atmospheres
Omni-Lab glovebox equipped with a freezer set to À30 °C.
Physical Methods. Elemental analyses were performed at Midwest
Microlab, LLC, in Indianapolis, IN. Infrared (IR) spectra of the Fe-acacX
complexes were measured as KBr pellets using a Nicolet Magna-IR 560
spectrometer. 1H and 19F NMR spectra were collected at room
temperature with a Varian 400 MHz spectrometer; 19F NMR spectra
were referenced using the benzotrifluoride peak at À63.7 ppm.
UVÀvis spectra were obtained with an Agilent 8453 diode array
spectrometer. Electrochemical measurements were performed with
an epsilon EC potentiostat (iBAS) under nitrogen atmosphere at a scan
rate of 100 mV/s with 60 mM (NBu4)PF6. A three-electrode cell
containing a Ag/AgCl reference electrode, a platinum auxiliary electrode,
and a glassy carbon working electrode was employed for cyclic voltammetric
(CV) measurements. Under these conditions, the ferrocene/ferrocenium
11979
dx.doi.org/10.1021/ic201115s |Inorg. Chem. 2011, 50, 11978–11989

Inorganic Chemistry
ARTICLE
(Fc+/0) couple has an E1/2 value of +474 mV in MeCN and 1:1 MeCN:
CH2Cl2 solutions.
IR (KBr, cmÀ1): 3062, 2619 [ν(BH)], 1606 [ν(CO)], 1479, 1362, 1169,
1144 (CF3). 19F NMR (C6D6): À42.1.
General Procedure for Synthesis of (R2Tp)Fe(acacX) Com-
plexes (R = Me, Ph). These complexes were generated using a
modified version of the procedure reported by Siewert et al.18 The
Na(acacX) salts were prepared by mixing the appropriate β-diketone
(1.0 mmol) with NaOMe (1.0 mmol) in THF for 30 min. Afterward,
the solvent was removed under vacuum to yield the Na(acacX) salt as awhite
solid. Anhydrous FeCl2 (1.0 mmol), KR2Tp (1.0 mmol), and 10 mL of
MeCN were then added to the flask, and the reaction mixture was stirred
overnight at ambient temperature. The solution was filtered to remove
inorganic salts (NaCl and KCl) and any Fe(R2Tp)2 byproduct, and
the volume of the resulting solution was reduced by half. For the
Fe(Me2Tp)(acacX) complexes, storage of this solution (with a concen-
tration of ∼0.2 M) at À30 °C yielded crystals suitable for X-ray
diffraction analysis. The crystallization methods used for the Fe(Ph2Tp)-
(acacX) complexes are described below for each complex.
General Procedure for Synthesis of [(PhTIP)Fe(acacX)]OTf
Complexes. The appropriate acacX ligand (1.0 mmol) was stirred with
NaOMe (1.0 mmol) in THF for 30 min, after which the solvent was
removed under vacuum to yield the Na(acacX) salt. The PhTIP ligand
(1.0 mmol) and anhydrous Fe(OTf)2 (1.0 mmol) were each dissolved in
3 mL of MeOH and added dropwise to a flask containing the Na(acacX)
salt in 5 mL of MeOH. The reaction mixture was stirred overnight at
ambient temperature, and the solvent was removed by vacuum. The
method used to obtain crystals of each PhTIP-containing complex is
described below.
[(PhTIP)Fe(acac)]OTf. The solid obtained after removal of the
MeOH solvent was dissolved in 5 mL of MeCN. After filtration to
remove insoluble materials, the 0.2 M solution was cooled to À30 °C,
and yellow crystals formed after several days. Yield: 36%. Anal. Calcd for
C33H28F3FeN6O5PS: C, 51.84; H, 3.69; N, 10.99. Found: C, 51.52; H,
3.65; N, 10.68. UVÀvis [λmax, nm (ε, MÀ1 cmÀ1) in MeCN]: 415 (160),
351 (sh). IR (KBr, cmÀ1): 3207, 1587 [ν(CO)], 1559, 1516, 1478, 1458,
1389. 19F NMR (CD3CN): À79.8 (OTf).
(
Me2Tp)Fe(acac). Yield: 114 mg, 25%. Anal. Calcd for C20H29BFe-
N6O2: C, 53.13; H, 6.46; N, 18.59. Found: C, 53.03; H, 6.28; N, 18.57.
UVÀvis [λmax, nm (ε, MÀ1 cmÀ1) in MeCN]: 438 (620), 348 (630).
IR (KBr, cmÀ1): 2955, 2922, 2520 [ν(BH)], 1590 [ν(CO)], 1450,
1400, 1200.
[(PhTIP)Fe(acacF3)]OTf. The solid obtained after removal of the
MeOH solvent was dissolved in 5 mL of CH2Cl2. After filtration, the
solution was layered with pentane, providing orange crystals. Anal. Calcd
for C33H25F6FeN6O5PS: C, 48.43; H, 3.08; N, 10.27. Found: C, 48.05;
H, 3.24; N, 10.04. UVÀvis [λmax, nm (ε, MÀ1 cmÀ1) in MeCN]: 461
(220), 373 (sh). IR (KBr, cmÀ1): 3221, 1630 [ν(CO)], 1559, 1478,
1458. 19F NMR (CD3CN): À45.1 (acacF3), À79.8 (OTf).
(
Me2Tp)Fe(acacF3). Yield: 46 mg, 11%. Anal. Calcd for C20H26BF3Fe-
N6O2: C, 47.46; H, 5.18; N, 16.61. Found: C, 47.65; H, 5.15; N, 16.39.
UVÀvis [λmax, nm (ε, MÀ1 cmÀ1) in MeCN]: 479 (580), 377 (400). IR
(KBr, cmÀ1): 2962, 2928, 2534 [ν(BH)], 1616 [ν(CO)], 1543, 1450,
1292, 1188, 1138 (CF3). 19F NMR (CD3CN): À56.2.
Me2Tp)Fe(acacPhF3). Yield: 352 mg, 62%. Anal. Calcd for C25H28BF
[(PhTIP)Fe(acacPhF3)]OTf. The solid obtained after removal of the
MeOH solvent was dissolved in 5 mL of CH2Cl2. The solution was filtered
and then layered with pentane, yielding reddish crystals. Anal. Calcd for
C38H27F6FeN6O5PS: C, 51.83; H, 3.09; N, 9.54. Found: C, 50.30; H, 3.18;
N, 9.13 (the slight discrepancy in the carbon value indicates that small
amounts of impurities are present). UVÀvis [λmax, nm (ε, MÀ1 cmÀ1) in
MeCN]: 528 (630), 408 (320). IR (KBr, cmÀ1): 3223, 1609 [ν(CO],
1574, 1478, 1457. 19F NMR (CD3CN): À48.3 (acacPhF3), À79.8 (OTf).
[(PhTIP)Fe(acacF6)]OTf. The solid obtained after removal of the
MeOH solvent was dissolved in 5 mL of CH2Cl2, filtered to remove
insoluble materials, and then layered with pentane to give purple crystals.
Yield: 25%. Anal. Calcd for C33H22F9FeN6O5PS: C, 45.43; H, 2.54; N,
9.63. Found: C, 43.19; H, 2.80; N, 9.32 (the minor discrepancies indi-
cate that small amounts of impurities are present). UVÀvis [λmax, nm
(ε, MÀ1 cmÀ1) in MeCN]: 509 (450), 381 (390). IR (KBr, cmÀ1): 3206,
1632 [ν(CO)], 1560, 1480. 19F NMR (CD3CN): À63.1 (acacF6), À79.8
(OTf).
(
3FeN6O2: C, 52.85; H, 4.97; N, 14.79. Found: C, 53.03; H, 5.03; N
14.72. UVÀvis [λmax, nm (ε, MÀ1 cmÀ1) in MeCN]: 567 (1190), 403
(570). IR (KBr, cmÀ1): 2962, 2931, 2526 [ν(BH)], 1604 [ν(CO)],
1573, 1537, 1450, 1288, 1196, 1142 (CF3). 19F NMR (CD3CN): À64.6.
(
Me2Tp)Fe(acacF6). Yield: 206 mg, 48%. Anal. Calcd for C20H23BF6
FeN6O2: C, 42.89; H, 4.14; N, 15.00. Found: C 42.97, H 4.22, N 15.22.
UVÀvis [λmax, nm (ε, MÀ1 cmÀ1) in MeCN]: 571 (820), 398 (560). IR
(KBr, cmÀ1): 2966, 2928, 2530 [ν(BH)], 1628 [ν(CO)], 1543, 1450,
1254, 1196, 1142 (CF3). 19F NMR (CD3CN): À78.9.
(
Ph2Tp)Fe(acac). Slow evaporation of a MeCN solution provided
yellow-green crystals after several days. Yield: 172 mg, 21%. Anal. Calcd
for C50H41BFeN6O2: C, 72.83; H, 5.01; N, 10.19. Found: C, 71.69; H,
5.03; N, 10.02. UVÀvis [λmax, nm (ε, MÀ1 cmÀ1) in MeCN]: 420 (410),
356 (sh). IR (KBr, cmÀ1): 3059, 2611 [ν(BH)], 1585 [ν(CO)],
1477, 1362.
(
Ph2Tp)Fe(acactBu). All MeCN was removed under vacuum, and
the yellow solid was redissolved in 5 mL of THF and filtered. Layering
the filtrate with MeCN provided yellow crystals suitable for X-ray
diffraction. Yield: 199 mg, 23%. Anal. Calcd for C53H47BFeN6O2: C,
73.25; H, 5.47; N, 9.70. Found: C, 72.85; H, 5.46; N, 9.39. UVÀvis
[λmax, nm (ε, MÀ1 cmÀ1) in MeCN]: 424 (420), 361 (sh). IR
(KBr, cmÀ1): 3062, 2615 [ν(BH)], 1583 [ν(CO)], 1479, 1362, 1169.
Crystallographic Studies. The 12 newly synthesized complexes
reported here were each characterized using X-ray crystallography. The
X-ray diffraction data were collected at 100 K with an Oxford Diffraction
SuperNova kappa-diffractometer equipped with dual microfocus
Cu/Mo X-ray sources, X-ray mirror optics, Atlas CCD detector and a
low-temperature Cryojet device. Crystallographic data for the com-
pounds are provided in Table 1. The data were processed with
CrysAlisPro program package (Oxford Diffraction Ltd., 2010) typically
using a numerical Gaussian absorption correction (based on the real
shape of the crystal) followed by an empirical multiscan correction using
the SCALE3 ABSPACK routine. The structures were solved using the
SHELXS program and refined with the SHELXL program30 within the
Olex2 crystallographic package.31 All computations were performed on
an Intel PC computer with Windows 7 OS. The majority of the
structures contain a certain degree of disorder that was detected in
difference Fourier syntheses of electron density and accounted for using
capabilities of the SHELX package. In most cases, hydrogen atoms were
localized in difference syntheses of electron density but were refined
using appropriate geometric restrictions on the corresponding bond
lengths and bond angles within a riding/rotating model (torsion angles
(
Ph2Tp)Fe(acacF3). All MeCN was removed under vacuum, and the
orange solid was redissolved in 5 mL of CH2Cl2 and filtered. Layering
the filtrate with MeCN provided reddish orange crystals suitable for
X-ray diffraction. Yield: 194 mg, 22%. Anal. Calcd for C50H38BF3Fe-
N6O2: C, 68.36; H, 4.36; N, 9.57. Found: C, 68.63; H, 4.51; N, 9.71.
UVÀvis [λmax, nm (ε, MÀ1 cmÀ1) in MeCN]: 462 (440), 363 (sh). IR
(KBr, cmÀ1): 3060, 2620 [ν(BH)], 1630 [ν(CO)], 1480, 1360, 1170,
1140 (CF3). 19F NMR (C6D6): À46.6.
(
Ph2Tp)Fe(acacPhF3). All MeCN was removed under vacuum, and
the purple solid was redissolved in 5 mL of CH2Cl2 and filtered. Layering
the filtrate with MeCN provided reddish purple crystals suitable
for X-ray diffraction. Yield: 239 mg, 25%. Anal. Calcd for C55H40BF3Fe-
N6O2: C, 70.23; H, 4.29; N, 8.93. Found: C, 69.83; H, 4.29; N, 8.90.
UVÀvis [λmax, nm (ε, MÀ1 cmÀ1) in MeCN]: 540 (980), 415 (460).
11980
dx.doi.org/10.1021/ic201115s |Inorg. Chem. 2011, 50, 11978–11989

Inorganic Chemistry
ARTICLE
Table 1. Summary of X-ray Crystallographic Data Collection and Structure Refinement
1-acac 2MeCN
1-acacF3 MeCN
1-acacPhF3 MeCN
1-acacF6
2-acac
2-acactBu 0.5THF 0.5 MeCN
3
3
3
3
3
empirical formula
formula weight
crystal system
space group
a, Å
C26H38BFeN9O2
575.31
C24H32BF3FeN8O2
588.24
C29H34BF3FeN8O2
650.30
C22H26BF6FeN7O2
601.16
C50H41BFeN6O2
824.55
C56H52.5BFeN6.5O2.5
923.20
triclinic
monoclinic
P21/n
monoclinic
P21/c
monoclinic
P21/n
monoclinic
P21/n
monoclinic
P21/n
P1
7.8894(2)
12.1386(3)
16.3511(4)
91.901(2)
98.880(2)
97.026(2)
1533.33(7)
2
11.32026(9)
19.30082(13)
13.78973(11)
90
16.3762(7)
12.3527(9)
16.0614(6)
90
11.72084(10)
19.56928(16)
11.95953(9)
90
13.88320(12)
18.87572(15)
16.93576(18)
90
9.75151(12)
35.2674(4)
13.67554(17)
90
b, Å
c, Å
α, deg
β, deg
106.1061(8)
90
102.273(5)
90
99.6456(8)
90
112.1370(11)
90
94.0112(11)
90
γ, deg
V, Å3
2894.66(4)
4
3174.8(3)
4
2704.36(4)
4
4110.95(7)
4
4691.63(10)
4
Z
D
calc, g/cm3
1.246
1.350
1.361
1.477
1.332
1.307
λ, Å
0.7107
1.5418
1.5418
0.7107
0.7107
1.5418
μ, mmÀ1
0.53
4.658
4.303
0.633
0.416
2.984
θ-range, deg
3 to 33
4 to 77
5 to 74
3 to 38
3 to 33
3 to 71
reflections collected
independent reflections
71038
27495
13228
81670
59544
22809
10798
6018
6281
14031
14181
8776
[Rint = 0.0537]
10789/3/374
1.010
[ Rint = 0.0263]
6018/24/402
1.010
[ Rint = 0.0229]
6281/0/405
1.034
[ Rint = 0.0247]
14031/6/386
1.039
[ Rint = 0.0268]
14181/0/543
1.085
[ Rint = 0.0406]
8776/13/645
0.837
data/restraints/parameters
GOF (on F2)
R1/wR2 (I > 2σ(I)) a
0.0344/0.0878
0.0474/0.0914
0.0288/0.0710
0.0327/0.0728
0.0331/0.0818
0.0372/0.0853
0.0309/0.0957
0.0406/0.0982
0.0352/0.1023
0.0502/0.1054
0.0335/0.0673
0.0547/0.0710
R1/wR2 (all data)
2-
[3-acac]OTf
[3-acacF3]OTf
[3-acacPhF3]OTf
4CH2Cl2
[3-acacF6(MeCN)]OTf
2-acacF3 CH2Cl2
acacPhF3 CH2Cl2
MeCN
CH2Cl2
2MeCN
b
3
3
3
3
3
3
empirical formula
C50.87H 39.73 BCl1.73
F3FeN6O2
952.17
C56H42BCl2-
F3FeN6O2
1025.52
monoclinic
P21/n
C35H31F3Fe-
N7O5PS
805.55
C34H27Cl2-
F6FeN6O5PS
903.40
C42H35Cl8-
F6FeN6O5PS
C39H31F9FeN9-
O5PS
formula weight
crystal system
space group
a, Å
1220.24
995.61
monoclinic
P21/n
triclinic
triclinic
triclinic
orthorhombic
Pna21
P1
P1
P1
9.9620(3)
33.2844(10)
13.5814(4)
90
9.9045(7)
35.733(2)
13.4769(9)
90
10.8489(3)
12.6264(3)
14.4290(4)
87.443(2)
71.223(2)
89.004(2)
1869.46(8)
2
10.9737(3)
12.6383(3)
14.6113(3)
85.8018(19)
71.277(2)
87.183(2)
1913.32(8)
2
11.4327(3)
12.8330(2)
17.9561(3)
85.0393(15)
78.0672(17)
83.3042(17)
2554.66(9)
2
20.3397(3)
11.83101(16)
17.8160(2)
90
b, Å
c, Å
α, deg
β, deg
94.637(3)
90
92.429(7)
90
90
γ, deg
V, Å3
90
4488.6(2)
4
4765.4(5)
4
4287.23(10)
4
Z
D
calc, g/cm3
1.409
1.429
1.431
1.568
1.586
1.542
λ, Å
1.5418
1.5418
0.7107
0.7107
1.5418
0.7107
μ, mmÀ1
4.153
4.090
0.568
0.709
7.525
0.532
θ -range, deg
reflections collected
independent reflections
3 to 71
4 to 75
2 to 29
2 to 29
2 to 74
3 to 29
25772
26605
42236
86437
23048
41132
8377
9427
9554
10144
10023
10734
[Rint = 0.2268]
8377/0/597
1.063
[Rint = 0.0545]
9427/0/640
1.058
[Rint = 0.0335]
9554/0/481
1.041
[Rint = 0.0498]
10144/26/584
1.081
[Rint = 0.0542]
10023/0/631
1.033
[Rint = 0.0251]
10734/1/590
1.057
data/restraints/parameters
GOF (on F2)
R1/wR2 (I > 2σ(I)) a
0.0804/0.1983
0.0661/0.1452
0.0895/0.1553
0.0356/0.0797
0.0486/0.0868
0.0361/0.0998
0.0477/0.1036
0.0630/0.1744
0.0671/0.1810
0.0247/0.0581
0.0292/0.0589
R1/wR2 (all data)
0.1107/0.2217
2
a R1 = ∑||Fo| À |Fc||/∑|Fo|; wR2 = [∑w(Fo À Fc2)2/∑w(Fo2)2]1/2. b The solvate molecule in 2-acacF3 CH2Cl2 is only partially (87%) populated.
3
of methyl hydrogens were optimized to better fit the residual electron
density).
Density Functional Theory (DFT) Calculations. DFT cal-
culations were performed using the ORCA 2.7 software package
11981
dx.doi.org/10.1021/ic201115s |Inorg. Chem. 2011, 50, 11978–11989

Inorganic Chemistry
ARTICLE
Figure 1. Thermal ellipsoid plots (50% probability) derived from 1-acac 2MeCN (A), 2-acac (B), and 2-acactBu 0.5MeCN 0.5MeCN
3
3
3
(C). Hydrogen atoms and noncoordinating solvent molecules have been omitted for clarity, as well as the Ph-rings at the 3- and 5-positions of the
Ph2Tp ligand in 2-acactBu
.
developed by Dr. F. Neese.32 In each case, the corresponding X-ray
structure provided the starting point for geometry optimizations and
the computational model included the entire complex (excluding
counteranions and uncoordinated solvent molecules). Geometry
optimizations employed the BeckeÀPerdew (BP86) functional33
and Ahlrichs’ valence triple-ζ basis set (TZV) for all atoms, in
conjunction with the TZV/J auxiliary basis set.34 Extra polarization
functions were used on non-hydrogen atoms. Single-point (SP)
calculations involving the optimized models were carried out with
Becke’s three-parameter hybrid functional for exchange along with
the LeeÀYangÀParr correlation functional (B3LYP).35 These SP
calculations also utilized the TZV basis set noted above, but addi-
tional polarization functions were included for all atoms, including
hydrogens. The same enlarged basis set was used for time-dependent
DFT (TD-DFT) calculations,36 which computed absorption energies
and intensities within the TammÀDancoff approximation.37 In each
case, at least 20 excited states were calculated. Finally, the gOpenMol
program38 developed by Laaksonen was used to generate isosurface
plots of molecular orbitals.
Table 2. Selected Metric Parameters for 1-acac 2MeCN,
3
1-acacF3 MeCN, 1-acacPhF3 MeCN, and 1-acacF6 a
3
3
1-acac
1-acacF3
1-acacPhF3
3
3
3
2MeCN
MeCN
MeCN
1-acacF6
FeÀO1
FeÀO2
FeÀN1
FeÀN3
FeÀN5
FeÀN7
2.0882(8)
2.0510(8)
2.1535(9)
2.1851(9)
2.1748(9)
2.2363(10)
2.070
2.0563(10)
2.0843(10)
2.1798(12)
2.1353(11)
2.2141(11)
2.2212(12)
2.070
2.0644(11)
2.0730(11)
2.1436(13)
2.1635(13)
2.1695(13)
2.2550(14)
2.069
2.1116(6)
2.0976(6)
2.1502(7)
2.1154(7)
2.1461(7)
2.2461(7)
2.105
FeÀOacac (ave)
FeÀNTp (ave)
2.171
2.176
2.158
2.137
FeÀN7ÀCXb
159.30(9)
20.0
161.20(12)
10.7
171.49(14)
18.4
169.23(7)
24.4
acacX tiltc
a Bond distances in Å and angles in deg. b N7 and CX are atoms in the
coordinated MeCN ligand. c acacX tilt = average angle between the plane
of the acac ligand and a plane defined by the O1ÀFeÀO2 atoms.
3. RESULTS AND ANALYSIS
A. Synthesis and Solid State Structures. Mononuclear
Fe(II) β-diketonato complexes with the Me2Tp and Ph2Tp
ligands (1-acacX and 2-acacX, respectively; Scheme 1) were
prepared by mixing 1 equiv of the sodium salt of the appro-
priate β-diketone, Na(acacX), with equimolar amounts of
FeCl2 and K(R2Tp) in MeCN. The analogous PhTIP-contain-
ing complexes [3-acacX]OTf were generated via addition of
the ligand to MeOH solutions of Na(acacX) and Fe(OTf)2. All
syntheses were performed under anaerobic conditions. Each
complex was characterized with single-crystal X-ray crystal-
lography; details concerning the data collection and analysis
are summarized in Table 1. While the majority of these
complexes crystallize as solvates, elemental analysis indicates
that no solvent is present in ground vacuum-dried crystals.
Upon cooling to À30 °C in MeCN, concentrated solutions of
1-acacX (acacX = acac, acacF3, acacF6, and acacPhF3) provided
crystals suitable for X-ray diffraction studies. Each of the resulting
structures revealed a 6C monoiron(II) center with a facially
coordinating Me2Tp ligand, bidentate acacX ligand, and bound
MeCN. In some cases, the crystals contained one or two equivalents
of noncoordinating MeCN. Figure 1A displays the molecular
structure of 1-acac, and selected bond distances and angles
for the four 1-acacX complexes are provided in Table 2.39
The iron-pyrazole bond lengths (average FeÀNTp = 2.16 Å)
are similar to those observed for other six-coordinate high-spin
Fe(II) complexes with Tp ligands.19,40 The FeÀOacac bond
distances vary slightly across the series, with the average value
ranging from 2.070 Å for 1-acac to 2.105 Å for 1-acacF6. The acac
ligands are tilted out of the equatorial N2O2 plane by an average
of 18° because of steric interactions with the 3-Me group of the
axial pyrazole moiety. The MeCN ligands exhibit FeÀNMeCN
bond distances near 2.24 Å and exist in bent conformations with
FeÀNÀC angles between 159 and 172°.
Complexes with the Ph2Tp ligand (2-acacX; acacX = acac,
acactBu, acacF3, and acacPhF3) were crystallized using various
methods: slow evaporation of a concentrated MeCN solution
(2-acac) and diffusion of MeCN into solutions of CH2Cl2 (2-
acacF3 and 2-acacPhF3) or THF (2-acactBu). In each case, the
resulting structures revealed 5C Fe(II) centers lacking bound
solvent, even though all crystallizations involved MeCN. As
illustrated in the structure of 2-acac (Figure 1B), two phenyl
rings from the Ph2Tp ligand are positioned over the vacant
coordination site, apparently restricting access of solvent to the
metal center. The Fe-ligand bond lengths (Table 3) are typical
for high-spin Fe(II) complexes, although somewhat shorter than
those found in the Me2Tp structures because of the decrease in
11982
dx.doi.org/10.1021/ic201115s |Inorg. Chem. 2011, 50, 11978–11989

Inorganic Chemistry
ARTICLE
Table 3. Selected Bond Distances (Å) and Bond Angles (deg)
for 2-acac, 2-acactBu 0.5MeCN 0.5THF, 2-acacF3, and
3
3
2-acacPhF3 CH2Cl2
3
2-acactBu 0.5MeCN
2-acacF3
2-acacPhF3
3
3
3
3
2-acac
0.5THF
CH2Cl2
CH2Cl2
FeÀO1
FeÀO2
FeÀN1
FeÀN3
FeÀN5
1.9945(10)
2.0239(9)
2.1625(10)
2.1462(11)
2.0984(9)
2.0492(13)
1.9443(14)
2.0945(14)
2.2603(15)
2.1074(14)
1.997
2.050(3)
1.986(3)
2.096(3)
2.178(3)
2.117(3)
2.018
2.054(3)
1.967(3)
2.092(3)
2.231(3)
2.095(3)
2.011
FeÀOacac (ave) 2.009
FeÀNTp (ave) 2.136
2.154
2.130
2.139
O1ÀFeÀO2
O1ÀFeÀN1
O1ÀFeÀN3
O1ÀFeÀN5
O2ÀFeÀN1
O2ÀFeÀN3
O2ÀFeÀN5
N1ÀFeÀN3
N1ÀFeÀN5
N3ÀFeÀN5
τ-value a
88.82(4)
88.14(6)
94.55(6)
177.39(5)
94.64(5)
140.95(6)
93.35(6)
125.96(6)
82.95(6)
92.70(5)
86.23(5)
0.607
86.40(12) 87.21(11)
94.81(12) 95.95(11)
91.24(4)
158.66(4)
110.44(4)
164.31(4)
92.64(4)
103.37(4)
81.71(4)
91.31(4)
89.94(4)
0.094
172.58(11) 177.69(11)
97.33(12) 95.34(11)
152.94(12) 144.46(12)
92.26(12) 91.01(11)
115.49(13) 122.84(12)
83.07(12) 84.65(11)
91.21(12) 92.19(12)
89.83(11) 86.86(11)
0.327
0.554
a For a definition of the τ-value, see references 41 and 42. A five-
coordinate complex with ideal square-pyramidal geometry would have a
τ-value of 0.0, while those with ideal trigonal bipyramidal geometry
would have a value of 1.0.
Figure 2. Thermal ellipsoid plots (50% probability) derived from
[3-acac]OTf MeCN (A) and [3-acacF6(MeCN)]OTf 2MeCN (B).
Noncoordinating solvent molecules, counteranions, and most hydrogen
atoms have been omitted for clarity.
3
3
coordination number. The overall geometry depends strongly on
the steric properties of the acacX ligand. Structures possessing the
acac and acacF3 ligands are best described as distorted square-
pyramids (τ = 0.09 and 0.33 for 2-acac and 2-acacF3, re-
spectively), whereas 2-acacPhF3 and 2-acactBu exhibit trigonal
bipyramidal geometries (τ = 0.56 and 0.61, respectively) in which
lie roughly parallel to the plane of the acacX ligand, shielding the
vacant coordination site. By contrast, the presence of the
coordinated MeCN ligand in [3-acacF6(MeCN)]OTf forces
two of the Ph rings to adopt orientations perpendicular to the
acacF6 ligand (Figure 2B).
t
the Oacac atom proximal to the bulky substituent (Ph or Bu)
occupies an axial position (Figure 1C).41,42 The axial and equatorial
FeÀOacac bond lengths diverge as the structure shifts toward a
trigonal bipyramidal geometry. This asymmetry is also evident in the
metric parameters of the acacX ligands. For 2-acactBu, the C3ÀC4
bond is 0.043(7) Å shorter than the corresponding C2ÀC3 bond,
while the O2ÀC4 bond is 0.031(5) Å longer than the O1ÀC2
bond (similar values are observed in 2-acacPhF3). These results
indicate that the negative charge of the acac ligand is primarily
localized on the equatorial O-atom in structures with distorted
trigonal bipyramidal geometries.
Comparison of structures with the same coordination number
and acacX ligand indicates that FeÀN bond distances involving
the neutral PhTIP ligands are consistently longer than those
involving the anionic R2Tp ligands. For example, the FeÀNTIP
bond distances in 6C [3-acacF6(MeCN)]OTf are lengthened by
∼0.08 Å (on average) relative to the FeÀNTp distances in 1-acacF6.
Yet this difference is less pronounced when one compares 5C
complexes with the same acacX ligand. In these cases, the FeÀ
NTIP bond distances are only ∼0.02 Å longer (on average) than
the corresponding FeÀNTp bond distances.
Metric data for complexes with the neutral PhTIP ligand
([3-acacX]OTf; acacX = acac, acacF3, acacF6, and acacPhF3) are
provided in Table 4, and two prototypical crystallographic
structures from this series are shown in Figure 2. Crystals of
these triflate salts were obtained by either the slow cooling of
MeCN solutions or the diffusion of pentane into CH2Cl2
solutions. Analysis of the crystal packing reveals that each triflate
counteranion forms hydrogen bonds with three imidazole NÀH
groups in the solid state. In general, the [3-acacX]OTf complexes
exhibit 5C Fe(II) geometries that are intermediate between
square pyramidal and trigonal bipyramidal (τ-values between
0.38 to 0.56). The lone exception is [3-acacF6(MeCN)]OTf,
which features a 6C Fe(II) center with a bound solvent ligand
(Figure 2B). In the 5C structures, the Ph rings of the PhTIP ligand
B. Spectroscopic and Electrochemical Properties. UVÀvis
absorption spectra of the reported Fe-acacX complexes, measured
in MeCN at room temperature, are shown in Figure 3. Along with
intense near-UV peaks (not shown), two broad absorption mani-
folds with ε-values between 0.2 and 1.2 mMÀ1 cmÀ1 are observed
in the visible region, giving the complexes their distinctive colors.
These two features are separated by ∼6000À8000cmÀ1, although
the higher-energy band is often obscured in the 2-acacX and
[3-acacX]OTf spectra because of the onset of Ph-based transitions
in the near-UV. It isapparentin most spectra that the lower-energy
band is composed of two (or more) overlapping peaks. Within
each series, this feature red-shifts as the acacX ligand becomes
more electron-poor, suggesting that it primarily arises from an
Fe(II)facacX MLCT transition—an assignment confirmed by
11983
dx.doi.org/10.1021/ic201115s |Inorg. Chem. 2011, 50, 11978–11989

Inorganic Chemistry
ARTICLE
Table 4. Selected Bond Distances (Å) and Bond Angles (deg)
for [3-acac]OTf MeCN, [3-acacF3]OTf CH2Cl2,
3
3
[3-acacPhF3]OTF 4CH2Cl2, and [3-acacF6]OTf 3MeCN
3
3
[3-acac]
[3-acacF3
]
[3-acacPhF3
]
[3-acacF6(MeCN)]
OTf MeCN OTf CH2Cl2 OTF 4CH2Cl2 OTf 2MeCN
3
3
3
3
FeÀO1
FeÀO2
FeÀN1
FeÀN3
FeÀN5
FeÀN7
1.9619(12) 1.930(11)
2.0586(11) 2.120(11)
2.2122(13) 2.2225(12)
2.1212(13) 2.0985(13)
2.1376(13) 2.1289(12)
1.974(2)
2.070(2)
2.193(3)
2.109(2)
2.120(2)
2.0792(10)
2.0918(11)
2.2639(13)
2.1705(11)
2.2123(13)
2.1854(14)
2.086
FeÀOacac (ave) 2.010
FeÀNTIP (ave) 2.157
2.025
2.151
2.022
2.141
2.216
O1ÀFeÀO2 87.47(5)
O1ÀFeÀN1 94.46(5)
O1ÀFeÀN3 142.76(5) 138.0(8)
O1ÀFeÀN5 123.96(5) 126.7(7)
O2ÀFeÀN1 176.36(5) 176.2(3)
86.5(3)
96.9(4)
85.96(9)
95.77(9)
150.09(10)
114.71(9)
172.64(9)
89.86(9)
95.87(9)
85.08(10)
89.92(9)
95.17(9)
0.376
83.89(4)
93.67(4)
178.87(4)
89.97(4)
175.81(5)
96.03(5)
92.45(5)
86.48(5)
90.95(5)
88.90(5)
À
O2ÀFeÀN3 92.12(5)
O2ÀFeÀN5 90.73(5)
N1ÀFeÀN3 84.47(5)
N1ÀFeÀN5 90.75(5)
N3ÀFeÀN5 93.28(5)
91.0(3)
88.5(4)
85.35(5)
90.62(5)
95.04(5)
0.637
τ-value a
0.560
a For a definition of the τ-value, see references 41 and 42. A five-
coordinate complex with ideal square-pyramidal geometry would have a
τ-value of 0.0, while those with ideal trigonal bipyramidal geometry
would have a value of 1.0.
literature precedents27 and time-dependent DFT (TD-DFT)
studies (vide infra). Our TD-DFT calculations further indicate that
the higher-energy feature corresponds to an acacX-based transition
with some Fe(II)facacX MLCT character. The MLCT intensities
are strongly dependent on the identity of the acacX ligand, following
the order acacPhF3 > acacF6 > acacF3 > acac in each series. A
complete summary of absorption energies and intensities are
provided in Table 5.
Figure 3. Electronic absorption spectra of the Fe(II)-acac complexes
1-acacX, 2-acacX, and [3-acacX]OTf measured at room temperature
in MeCN. For the 2-acacX complexes, a small amount (<5%) of toluene
was present to enhance solubility.
It is instructive to compare absorption data for complexes
with the same acacX ligand but different supporting ligands.
For instance, the MLCT bands of the Ph2Tp-based complexes
(2-acacX) are blue-shifted by 900 ( 100 cmÀ1 relative to their
counterparts in spectra of the Me2Tp-based complexes (1-acacX).
Given that these species share virtually identical ligand en-
vironments, such significant disparities in absorption energies
suggest that the difference in coordination number observed
in the solid-state structures persists in solution. The MLCT
absorption features of the PhTIP complexes ([3-acacX]OTf) are
higher in energy than those of the corresponding Me2Tp and
Ph2Tp complexes by an average of 1400 and 250 cmÀ1, respec-
tively. This result indicates that the Fe(II) d-orbitals are stabilized
in the PhTIP models relative to the Tp complexes, likely because
of the difference in charge of the supporting ligands. In addition,
MLCT absorption features in the [3-acacX]OTf series are much
weaker than analogous bands in the 1-acacX and 2-acacX series
(Table 5), suggesting that the LN3 ligands also modulate Fe-acacX
covalency.
2-acac, and [3-acac]OTf are shown in Figure 4. The wide range
of observed chemical shifts confirms that these Fe(II) complexes
possess high-spin (S = 2) electronic configurations. Assignments
were made on the basis of chemical shifts and peak integrations,
T1-relaxation values, and literature precedents. In all cases, the three
pyrazole (or imidazole) ligands are spectroscopically equivalent in
solution, suggesting that the positions of the acacX ligands are
dynamically averaged on the NMR time scale.43 1H NMR data
for complexes in the 1-acacX and 2-acacX series are summarized
in Supporting Information, Tables S1 and S2, respectively. The
4-pyrazole protons appear between 54 and 62 ppm, consistent
with prior studies of Fe(II) species with R2Tp ligands. For com-
plexes with the Me2Tp ligand, signals arising from the 3-Me and
5-Me pyrazole substituents are easily distinguished by their differ-
ent T1-values of 2.5 ( 0.1 and 38 ( 4 ms, respectively. In the
2-acacX series, the observed resonances largely conform to
the pattern reported by Mehn et al. for five-coordinate
1H and 19F NMR spectra of the Fe-acacX complexes were
measured in MeCN-d3 (1-acacX and [3-acacX]OTf) or benzene-
(
Ph2Tp)Fe(II)(α-keto carboxylate) complexes.21 For both
sets of compounds, the ortho protons of the 3-Ph rings display
1
d6 (2-acacX) at ambient temperature; H spectra for 1-acac,
11984
dx.doi.org/10.1021/ic201115s |Inorg. Chem. 2011, 50, 11978–11989

Inorganic Chemistry
ARTICLE
negative chemical shifts and remarkably short T1-values (<1 ms),
reflecting their close proximity to the paramagnetic Fe(II) center.
Each [3-acacX]OTf complex exhibits a downfield resonance
near 65 ppm with a T1-value of 6 ( 1 ms (Table 6). This peak
gradually disappears upon mixing with MeOH-d4, indicating that
it arises from the exchangeable proton of the imidazole moieties.
The other intense downfield signal (found between 35 and
50 ppm) also integrates to three protons, and it is assigned to
the 5-imidazole protons. On the basis of similarities to NMR
spectra of the Fe(Ph2Tp) complexes, we attribute the fast-relaxing
peak (T1 ∼ 1 ms) in the negative δ-region to ortho protons of the
2-phenyl substituents. The corresponding meta and para protons
appear near 6.5 and 7.8 ppm, respectively. 19F NMR spectra of
[3-acacX]OTf complexes with fluorinated β-diketonates each
display two resonances: a sharp peak at À79.2 ppm from the
triflate counteranion, and a fast-relaxing feature (T1 = 6 ( 3 ms)
between À45 and À65 ppm derived from the acacX ligands. The
chemical shift measured for the triflate counteranions is identical
to that observed for [NBu4]OTf in MeCN-d3.44 Interestingly,
both 1-acacF6 and [3-acacF6]OTf exhibit only one acacF6-
derived resonance, indicating that the two ÀCF3 groups are
equivalent in solution because of dynamic averaging (vide supra).
Regardless of the LN3 supporting ligands, peaks arising from
methyl substituents of the acac and acacF3 ligands exhibit upfield
chemical shifts ranging from À7 to À34 ppm. Interestingly, the T1-
values of the acac-Me resonances fall into two classes: those
measured for 1-acac and 1-acacF3 are near 10 ms, while those
measured for [3-acac]OTf and [3-acacF3]OTf are considerably
shorter (∼3 ms) and close to the value found for 2-acac in benzene-
d6 (3.3 ms). This result suggests that the acacX ligands adopt
different orientations with respect to the Fe(II) center in the two
sets of complexes, and provides further evidence that complexes
with bulky Ph-substituents remain pentacoordinate even in MeCN
solution. In addition, the T1-values of [3-acacF6]OTf—the only
6-coordinate PhTIP complex in the solid state—are significantly
larger than those measured for the other three [3-acacX]OTf
species, indicating differences in solution structures.
Figure 4. 1H NMR spectra of 1-acac in MeCN-d3 (A), 2-acac in
benzene-d6 (B), and [3-acac]OTf in MeCN-d3. Spectra were referenced
to the residual solvent peaks (indicated by *). Note that peak intensities
in the middle portions of the spectra were reduced for the sake of clarity.
both sets of data indicate the Fe(II) oxidation state is stabilized in
the order PhTIP > Ph2Tp > Me2Tp.
C. Density Functional Theory (DFT) Calculations. Energy-
minimized structures of the Fe-acacX complexes were generated
via DFT geometry optimizations. On the basis of the crystal-
lographic results, computational models of the 1-acacX com-
plexes included a bound MeCN ligand, while those of the
2-acacX series were exclusively 5C. For the sake of comparison,
structures of the [3-acacX]+ complexes were computed both with
and without coordinated MeCN. As shown in Supporting
Information, Tables S3ÀS5, the DFT-derived structures agree
quite well with the crystallographic results, generally providing
bond distances within 0.05 Å of the experimental values. Con-
sistent with the XRD data, DFT predicts FeÀNTIP bond
distances to be 0.103 ( 0.002 Å longer (on average) than
FeÀNTp distances for 6C complexes and 0.023 ( 0.002 Å
longer for 5C complexes, assuming the same acacX ligand.
However, DFT uniformly overestimates all FeÀNTp/TIP bond
lengths by approximately 0.03 Å, regardless of LN3 ligand,
while underestimating the FeÀOacac by the same amount. In
addition, the computed 5C models tend to exhibit larger
τ-values (i.e., geometries closer to the trigonal-bipyramidal
limit) than the experimental structures.
The redox properties of the Fe-acacX complexes were examined
with cyclic voltammetry in MeCN solutions with (NBu4)PF6 as
the supporting electrolyte (a 1:1 MeCN:CH2Cl2 mixture was
used for the 2-acacX complexes because of their limited solubility
in MeCN). The electrochemical data are summarized in Table 5
(potentials are reported vs Fc+/0), while Figure 5 displays
representative cyclic voltammograms for complexes with the
acacPhF3 ligand. Each R2Tp complex exhibits a quasi-reversible
one-electron oxidation wave corresponding to the Fe(II/III)
couple. As expected, within each series the redox potentials shift
to more positive values as the acacX ligand becomesmoreelectron-
poor (Eacac < EacacF3 < EacacPhF3 < EacacF6). Potentials measured for the
2-acacX complexes are 210 ( 25 mV more positive than those in
the 1-acacX series. As shown in Figure 5, the PhTIP complex [3-
acacPhF3]OTf displays an anodic wave at +410 mV along with
a much weaker cathodic wave at +190 mV, suggesting that
the oxidation is irreversible. Such behavior is typical of the
[3-acacX]OTf complexes, and thus only the Ep,a values are
provided in Table 5 ([3-acacF6]OTf failed to show any electro-
chemical events in the range examined). Regardless, the data
clearly indicate that the [3-acacX]OTf complexes are harder to
oxidize than the corresponding R2Tp-based complexes, with Ep,a
values shifted positively by 100À200 mV relative to the 2-acacX
series. Thus, the electrochemical results are consistent with the
trend observed for Fe(II)facacX MLCT energies (vide supra);
Spin-unrestricted single-point DFT calculations utilizing the
B3LYP hybrid functional were performed with the optimized models.
Representative molecular orbital (MO) energy-level diagrams are
showninFigure6for1-acacF6 and [3-acacF6(MeCN)]+.45 The lone
spin-down Fe electron lies in the 3dxz-based MO that bisects the
NeqÀFeÀO angles, while the highest-occupied acac-based MO
11985
dx.doi.org/10.1021/ic201115s |Inorg. Chem. 2011, 50, 11978–11989

Inorganic Chemistry
ARTICLE
Table 5. Physical Properties of Fe(II)-acacX Complexes and Comparison to Enzymatic Systems
UVÀvis
electrochemistry
a
complex
1-acac
1-acacF3
1-acacPhF3
1-acacF6
color
yellow
energy, cmÀ1 (ε, MÀ1 cmÀ1
)
E1/2, mV vs Fc+/0 (ΔE, mV) b
22830 (620)
28730 (630)
À303 (72)
À34 (107)
À2 (98)
+225 (127)
À58 (133)
+47 (116)
+158 (91)
+195 (94)
orange
20880 (580)
17640 (1190)
17510 (820)
23810 (410)
23590 (420)
21650 (440)
18520 (980)
24090 (160)
21690 (220)
18940 (630)
19650 (450)
24000 (1000)
23000 (760)
22200 (270)
18500 (350)
26520 (400)
24810 (570)
25130 (560)
28090 (sh)
27700 (sh)
27550 (sh)
24100 (460)
28490 (sh)
26810 (sh)
24510 (320)
26180 (390)
28000 (sh)
27500 (sh)
26200 (280)
reddish purple
purple
2-acac
2-acactBu
2-acacF3
2-acacPhF3
yellow
yellow
orange
purple
[3-acac]OTf
[3-acacF3]OTf
[3-acacPhF3]OTf
[3-acacF6]OTf
Dke1-acac d
HPPD-acac d
Dke1-acacF3 e
Dke1-acacPhF3 e
faint yellow
orange
E
p,a = +120 c
Ep,a = +360 c
Ep,a = +410 c
N/A
reddish purple
purple
a sh = shoulder; no intensity is reported. b ΔE = Ep,a À Ep,c c Only the Ep,a value is provided because of irreversibility. N/A = no electrochemical event was
observed. d Data obtained from reference 27. HPPD = hydroxyphenylpyruvate dioxygenase. e Data obtained from reference 14.
exhibits a large lobe of electron density on the central carbon
atom that reflects the anionic nature of the ligand (Figure 7).
The spin-down lowest unoccupied MO (LUMO) has mainly
acacF6 CdO* character, albeit with non-negligible Fe character
(∼7%). The Fe 3dxy- and 3dyz-based MOs lie at slightly higher
energies. Thus, the acac LUMO orbital is approximately isoener-
getic with the Fe(II) “t2g-set” of orbitals, resulting in significant
π-backbonding interactions. Similar bonding patterns were found
for all 6C complexes in the 1-acacX and [3-acacX(MeCN)]+
series, although the strength of the π-backbonding interac-
tions varied according to the electronic properties of the acacX
substituents.
The DFT results permitted further comparison of the electro-
nic properties of the R2Tp and PhTIP supporting ligands. Because
of the difference in charge between the two sets of complexes, it
was first necessary to normalize the orbital energies. This was
accomplished by assuming that the acac highest occupied MO
Figure 5. Cyclic voltammograms of 1-acacPhF3, 2-acacPhF3, and
[3-acacPhF3]OTf in MeCN (or 1:1 MeCN:CH2Cl2 for 2-acacPhF3
with 60 mM (NBu4)PF6 as the supporting electrolyte and a scan rate of
100 mV/s.
(HOMO), which is essentially nonbonding with respect to the
Fe(II) center and LN3 ligand, has identical energies in com-
plexes with the same acacX ligand (that is, the acacX HOMO
served as an “internal energy standard”). Following this proce-
dure, it is evident in Figure 6 that the Fe d-orbital manifold of
[3-acacF6(MeCN)]+ is uniformly stabilized relative to the
corresponding set of 1-acacF6 orbitals, reflecting the reduced
donor strength of the neutral PhTIP ligand compared to the
anionic Me2Tp ligand. Indeed, an analysis of DFT results for the
four pairs of 6C Fe-acacX species reveals that Fe d-based MOs in
PhTIP complexes are stabilized by an average of 0.9 ( 0.3 eV
relative to their counterparts in Me2Tp models. The contrast
between the Tp and TIP ligands is less dramatic for the 5C
complexes, where the difference in Fe d-orbital energies is only
∼0.3 eV (Supporting Information, Figure S1 provides MO
energy-level diagrams for 5C models of 2-acacF3 and [3-acacF3]+;
MO contour plots for [3-acacF3]+ are shown in Supporting
Information, Figure S2).
)
To aid in assignment of the observed UVÀvis absorption
features, TD-DFT calculations were performed for both 5C
and 6C models of the [3-acacX]+ complexes. Regardless of
coordination number, TD-DFT predicts two intense features
in the visible region: a Fe(xz) f acacX MLCT transition and a
higher-energy acacX-based transition with some MLCT char-
acter (electron density difference maps for the calculated
transitions are provided in Supporting Information, Figure S3).
As shown in Table 7, the computed energies for both types of
transitions agree reasonably well with the experimental data.
While TD-DFT consistently overestimates the MLCT intensi-
ties, it nicely reproduces the trend (observed experimentally)
that these transitions weaken as the acacX ligand becomes more
11986
dx.doi.org/10.1021/ic201115s |Inorg. Chem. 2011, 50, 11978–11989

Inorganic Chemistry
ARTICLE
Table 6. Summary of 1H NMR Parameters for [3-acacX]OTf Complexesa
1H NMR parameters
4-imid
o-2-Ph
m-2-Ph
δ, ppm
p-2-Ph
δ, ppm
NÀH
acacX
complex
δ, ppm
δ, ppm
δ, ppm
δ, ppm
[3-acac]OTf
50 (4.0)
47 (4.3)
45 (4.5)
35 (8.0)
À19 (0.45)
À14 (0.56)
À13 (0.62)
À4.5 (1.2)
6.5 (8.1)
6.4 (8.9)
6.4 (9.3)
7.2 (15)
7.8 (23)
7.9 (24)
7.8 (25)
7.8 (33)
65 (5.2)
66 (6.1)
66 (6.5)
67 (7.0)
À12 (ÀCH3, 2.9), 44 (-H, 1.0)
À21 (ÀCH3, 3.4), 27.2 (-H, 1.6)
[3-acacF3]OTf
[3-acacPhFe]OTf
[3-acacF6]OTf
8.0 (m-Ph, 35), 16 (p-Ph, 77), 19 (o-Ph, 2.9), 26 (-H, 1.4)
5.1 (-H, 3.3)
a All spectra were measured in MeCN-d3. The numbers in parentheses are the relaxation times (T1) in milliseconds.
Figure 7. Isosurface plots of spin-down MOs computed for [3-acacF6
(MeCN)]+ by DFT.
Figure 6. MO energy diagrams for geometry-optimized models of
1-acacF6 and [3-acacF6(MeCN)]+ obtained from DFT calculations.
To account for differences in overall charge, the computed MO energies
for [3-acacF6(MeCN)]+ were uniformly increased by 2.11 eV, such that the
acac HOMOs are isoenergetic for the two models (see text for more details).
Table 7. Comparison of Experimental and TD-DFT
Computed Transition Energies
MLCT
acacX-based
E, cmÀ1 ε, mMÀ1 cmÀ1a E, cmÀ1 ε, mMÀ1 cmÀ1a
electron rich. The MLCT transitions are most intense for the 6C
complexes, since overlap between the donor Fe(xz) orbital and
the acceptor acacX LUMO is maximized when the acacX ligand
lies in the equatorial plane. As the geometry shifts toward trigonal
bipyramidal in the 5C complexes, this orbital overlap is reduced.
In addition, intermediate τ-values facilitate mixing between the
MLCT and acacX-based transitions, thereby increasing the intensity
of the latter at the expense of the former in 5C complexes.
[3-acac]+
[3-acacF3
6C DFT 25363
5C DFT 26130
0.90
0.60
0.16
1.50
0.39
0.22
2.27
1.02
0.63
2.02
1.18
0.45
25363
26130
28490
23683
25842
26810
22139
22873
24510
22093
22406
26180
0.07
0.14
N/A
0.03
0.11
N/A
0.27
0.83
0.32
0.01
0.17
0.39
exp
24090
+
]
6C DFT 22597
5C DFT 23542
exp
21690
[3-acacPhF3
]
6C DFT 20036
5C DFT 20244
+
exp
18940
4. SUMMARY AND IMPLICATIONS FOR DKE1
[3-acacF6
]
6C DFT 19834
5C DFT 18830
+
This Paper has described the synthesis of three series of
Fe(II) β-diketonato complexes designed to model the acac-
bound form of Dke1 and replicate variations in the facial triad
(2H1C vs 3His) found in nonheme Fe dioxygenases. Adjustment
of the steric properties of the Tp ligands resulted in the formation
of both 5C and 6C complexes, and β-diketonato ligands with a
range of steric and electronic properties were employed to aid in
the interpretation of results. Each complex was extensively
characterized with experimental and computational methods,
including X-ray crystallography, UVÀvis and NMR spectro-
scopies, CV, and DFT calculations. Thus, the 12 reported
exp
19650
a Assumed ν1/2-value of 2500 cmÀ1
.
complexes have permitted a systematic examination of the roles of
the LN3 and acacX ligands in determining the structural, spectro-
scopic, electrochemical, and electronic properties of the Fe(II)
models. Comparison of complexes featuring anionic (R2Tp) and
neutral (PhTIP) supporting ligands, but identical acacX ligands,
reveals the following key differences: (i) regardless of coordination
11987
dx.doi.org/10.1021/ic201115s |Inorg. Chem. 2011, 50, 11978–11989

Inorganic Chemistry
ARTICLE
number, FeÀNTIP bond distances are consistently longer than
FeÀNTp distances (Tables 2À4), (ii) the Fe(II)facacX MLCT
features appear at higher energies for the PhTIP complexes
(Figure 3; Table 5), and (iii) redox potentials of the R2Tp
complexes are more negative than those of the corresponding
PhTIP complexes (Figure 5; Table 5). DFT calculations further
confirm that the PhTIP ligand is a significantly weaker donor, as
seen in the relative stabilization of the Fe d-orbitals (Figure 6). We
will now discuss the relevance of these findings for the electronic
structure of the Dke1 active site.
Diebold et al. recently published a study in which the spectro-
scopic features of acac-bound Dke1 were compared with those of
acac-bound hydroxyphenylpyruvate dioxygenase (HPPD), an
enzyme that possesses the conventional 2H1C facial triad.27
Prior to that, Straganz and Nidetzky reported the absorption
spectra of Dke1 coordinated to various β-diketonates.14 Like our
models, substrate-bound Dke1 exhibits an intense near-UV band
and two broad features in the visible region with ε-values
between 0.2 and 1.0 mMÀ1 cmÀ1 (Table 5). Diebold et al. also
used CD and MCD spectroscopies to observe much weaker
ligand-field transitions at lower energies. Analysis of these ligand-
field bands revealed only minor differences between enzymes
with the 3His and 2H1C triads; however, the MLCT feature is
shifted to lower energy by ∼1000 cmÀ1 in the 2H1C system.
Similarly, for our synthetic [(LN3)Fe2+(acacX)]0/+ complexes,
the absorption features of the 1-acacX series are red-shifted by an
average of 1400 cmÀ1 relative to the [3-acacX]OTf series. In
general, the 5C 2-acacX and [3-acacX]OTf spectra exhibit
excellent agreement with the Dke1-acacX absorption data, while
the 6C 1-acac spectrum is nearly identical to the one reported for
HPPD-acac.27 Thus, while our results indicate that the PhTIP
ligand accurately reproduces the enzymatic 3His coordination
environment, they would also seem to corroborate the conclu-
sion of Diebold et al. that variations in the facial triad give rise to
only modest spectral perturbations.
Yet analysis of the electronic transitions may not provide a
complete picture. Our electrochemical results indicate that the
[3-acacX]OTf complexes are harder to oxidize than the corre-
sponding 2-acacX complexes by an average of 145 mV, even
though the two sets exhibit similar absorption energies. Thus, the
charge of the supporting ligands has a significant impact on the
redox potential of the Fe center—a crucial parameter in tuning
the O2 reactivity of the Fe-acacX unit. These experimental results
are consistent with DFT calculations that indicate a sizable stabiliza-
tion of the Fe d-orbital manifold in the PhTIP complexes relative to
the Ph2Tp complexes (vide supra). Of course, the Fe redox
potential is somewhat irrelevant if the catalytic cycle proceeds via
direct reaction of O2 with the bound acac ligand, a possibility
suggested by Straganz and Nidetzky;14 our results, however, cast
some doubts on this scenario. First, the proposed mechanism
would require significant spin delocalization from the Fe center
to the acacX ligand to overcome the spin-forbidden nature of
concerted reaction with O2. While such a scenario has been shown
to occur in Fe3+-containing intradiol catechol dioxygenases,17 our
DFT calculations indicate that only a small amount of unpaired
spin-density resides on the acacX ligands in our models. Second,
the highest-occupied MO of the coordinated acac ligand, which
would play a central role in the electrophilic attack of O2, is at
least 1.0 eV lower in energy than the Fe-based MOs in all DFT
models.46 Even for complexes with electron-rich acac ligands, the
frontier MOs are exclusively Fe-based, suggesting that reaction
with O2 is more likely at Fe than the ligand. This conclusion has
been confirmed experimentally in our laboratory via O2 reactivity
studies, which will be described in a future paper. Regardless,
further biochemical and synthetic studies are required to fully
understand the significance ofthe3His triad for enzymatic function.
’ ASSOCIATED CONTENT
Supporting Information. 1H NMR data for 1-acacX
S
b
and 2-acacX complexes, metric parameters for DFT-opti-
mized models, MO energy-level diagrams for 2-acacF3 and
[3-acacF3]+, electron-density difference maps from TD-DFT
calculations, and crystallographic data in CIF format. This
material is available free of charge via the Internet at http://
pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: adam.fiedler@marquette.edu.
’ ACKNOWLEDGMENT
A.T.F. thanks Marquette University and the National
Science Foundation (CAREER CHE-1056845) for generous
financial support. We also thank Dr. James Gardinier for use of
his electrochemical apparatus and appreciate the assistance of
Dr. Sheng Cai in measuring T1-values
’ REFERENCES
(1) (a) Parales, R. E.; Haddock, J. D. Curr. Opin. Biotechnol. 2004,
15, 374–379.(b) Wackett, L. P.; Hershberger, C. D. Biocatalysis and
Biodegradation: Microbial Transformation of Organic Compounds; ASM
Press: Washington, D.C., 2001. (c) Singh, A.; Ward, O. P. In Biode-
gradation and Bioremediation; Singh, A., Ward, O. P., Eds.; Springer:
Heidelberg, Germany, 2004, pp 1À18.
(2) (a) Gibson, D. T.; Parales, R. E. Curr. Opin. Biotechnol. 2000,
11, 236–243.(b) Parales, R.; Resnick, S. M. In Biodegradation and
Bioremediation; Singh, A., Ward, O. P., Eds.; Springer: Heidelberg,
Germany, 2004, pp 175À196. (c) Costas, M.; Mehn, M. P.; Jensen,
M. P.; Que, L., Jr. Chem. Rev. 2004, 104, 939–986.
(3) (a) Kauppi, B.; Lee, K.; Carredano, E.; Parales, R. E.; Gibson,
D. T.; Eklund, H.; Ramaswamy, S. Structure 1998, 6, 571–586.
(b) Karlsson, A.; Parales, J. V.; Parales, R. E.; Gibson, D. T.; Eklund,
H.; Ramaswamy, S. Science 2003, 299, 1039–1042.
(4) (a) Han, S.; Eltis, L. D.; Timmis, K. N.; Muchmore, S. W.; Bolin,
J. T. Science 1995, 270, 976–980. (b) Vaillancourt, F. H.; Bolin, J. T.;
Eltis, L. D. Crit. Rev. Biochem. Mol. 2006, 41, 241–267. (c) Kovaleva,
E. G.; Lipscomb, J. D. Science 2007, 316, 453–457.
(5) (a) Xun, L. Y.; Bohuslavek, J.; Cai, M. A. Biochem. Biophys. Res.
Commun. 1999, 266, 322–325. (b) Xu, L.; Resing, K.; Lawson, S. L.;
Babbitt, P. C.; Copley, S. D. Biochemistry 1999, 38, 7659–7669.
(c) Nagata, Y.; Endo, R.; Ito, M.; Ohtsubo, Y.; Tsuda, M. Appl. Microbiol.
Biotechnol. 2007, 76, 741–752.
(6) Koehntop, K. D.; Emerson, J. P.; Que, L., Jr. J. Biol. Inorg. Chem.
2005, 10, 87–93.
(7) Bruijnincx, P. C. A.; Lutz, M.; Spek, A. L.; Hagen, W. R.;
Weckhuysen, B. M.; vanKoten, G.; Gebbink, R. J. M. K. J. Am. Chem.
Soc. 2007, 129, 2275–2286.
(8) (a) Kovaleva, E. G.; Lipscomb, J. D. Nature Chem. Biol. 2008,
4, 186–193. (b) Lipscomb, J. D. Curr. Opin. Struct. Biol. 2008,
18, 644–649.
(9) (a) Straganz, G. D.; Nidetzky, B. ChemBioChem 2006,
7, 1536–1548. (b) Leitgeb, S.; Nidetzky, B. Biochem. Soc. Trans. 2008,
36, 1180–1186.
11988
dx.doi.org/10.1021/ic201115s |Inorg. Chem. 2011, 50, 11978–11989

Inorganic Chemistry
ARTICLE
(10) (a) Joseph, C. A.; Maroney, M. J. Chem. Commun. 2007,
3338–3349. (b) Pierce, B. S.; Gardner, J. D.; Bailey, L. J.; Brunold,
T. C.; Fox, B. G. Biochemistry 2007, 46, 8569–8578.
(34) (a) Schafer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992,
97, 2571–2577. (b) Schafer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys.
1994, 100, 5829–5835.
(11) (a) Chen, J.; Li, W.; Wang, M. Z.; Zhu, G. Y.; Liu, D. Q.; Sun, F.;
Hao, N.; Li, X. M.; Rao, Z. H.; Zhang, X. C. Protein Sci. 2008,
17, 1362–1373. (b) Matera, I.; Ferraroni, M.; Burger, S.; Scozzafava,
A.; Stolz, A.; Briganti, F. J. Mol. Biol. 2008, 380, 856–868. (c) Hintner,
J. P.; Remtsma, T.; Stolz, A. J. Biol. Chem. 2004, 279, 37250–37260.
(12) Straganz, G. D.; Glieder, A.; Brecker, L.; Ribbons, D. W.;
Steiner, W. Biochem. J. 2003, 369, 573–581.
(35) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652. (b) Becke,
A. D. J. Chem. Phys. 1993, 98, 1372–1377. (c) Lee, C. T.; Yang, W. T.;
Parr, R. G. Phys. Rev. B 1988, 37, 785–789.
(36) (a) Stratmann, R. E.; Scuseria, G. E.; Frisch, M. J. J. Chem. Phys.
1998, 109, 8218–8224. (b) Casida, M. E.; Jamorski, C.; Casida, K. C.;
Salahub, D. R. J. Chem. Phys. 1998, 108, 4439–4449. (c) Bauernschmitt,
R.; Ahlrichs, R. Chem. Phys. Lett. 1996, 256, 454–464.
(13) Straganz, G. D.; Hofer, H.; Steiner, W.; Nidetzky, B. J. Am. Chem.
Soc. 2004, 126, 12202–12203.
(14) Straganz, G. D.; Nidetzky, B. J. Am. Chem. Soc. 2005, 127,
12306–12314.
(37) (a) Hirata, S.; Head-Gordon, M. Chem. Phys. Lett. 1999,
314, 291–299. (b) Hirata, S.; Head-Gordon, M. Chem. Phys. Lett.
1999, 302, 375–382.
(38) Laaksonen, L. J. Mol. Graphics 1992, 10, 33–34.
(15) (a) Straganz, G. D.; Diebold, A. R.; Egger, S.; Nidetzky, B.;
Solomon, E. I. Biochemistry 2010, 49, 996–1004. (b) Diebold, A. R.;
Straganz, G.; Solomon, E. I. J. Amer. Chem. 2011, 113, 15979–15991.
(16) Leitgeb, S.; Straganz, G. D.; Nidetzky, B. Biochem. J. 2009, 418,
403–411.
(17) (a) Pau, M. Y. M.; Davis, M. I.; Orville, A. M.; Lipscomb, J. D.;
Solomon, E. I. J. Am. Chem. Soc. 2007, 129, 1944–1958. (b) Pau,
M. Y. M.; Lipscomb, J. D.; Solomon, E. I. Proc. Natl. Acad. Sci. U.S.A.
2007, 104, 18355–18362.
(18) Siewert, I.; Limberg, C. Angew. Chem., Int. Ed. 2008, 47, 7953–7956.
(19) Kitajima, N.; Amagai, H.; Tamura, N.; Ito, M.; Morooka, Y.;
Heerwegh, K.; Penicaud, A.; Mathur, R.; Reed, C. A.; Boyd, P. D. W.
Inorg. Chem. 1993, 32, 3583–3584.
(20) (a) Ogihara, T.; Hikichi, S.; Akita, M.; Moro-oka, Y. Inorg.
Chem. 1998, 37, 2614–2615. (b) Mukherjee, A.; Cranswick, M. A.;
Chakrabarti, M.; Paine, T. K.; Fujisawa, K.; Munck, E.; Que, L. Inorg.
Chem. 2010, 49, 3618–3628. (c) Paine, T. K.; Zheng, H.; Que, L., Jr.
Inorg. Chem. 2005, 44, 474–476.
(39) In the crystal structures of 1-acacF3 and [3-acacF3]OTf, the
acacF3 ligand is disordered because of interchange of the -CF3 and -
CH3 groups. This disorder is fairly minor for 1-acacF3, as the two
species exist in a 9:1 ratio; however, the problem is more severe for
[3-acacF3]OTf, where the relative occupancies are ∼2:1. Hence,
uncertainties in bond lengths and angles are greater for the latter
structure. The metric parameters reported in Tables 2 and 4 refer to
the high-occupancy species.
(40) (a) Reger, D. L.; Gardinier, J. R.; Elgin, J. D.; Smith, M. D.;
Hautot, D.; Long, G. J.; Grandjean, F. Inorg. Chem. 2006,
45, 8862–8875. (b) Reger, D. L.; Elgin, J. D.; Smith, M. D.; Grandjean,
F.; Rebbouh, L.; Long, G. J. Polyhedron 2006, 25, 2616–2622.
(c) Calogero, S.; Lobbia, G. G.; Cecchi, P.; Valle, G.; Friedl, J. Polyhedron
1994, 13, 87–97. (d) Oliver, J. D.; Mullica, D. F.; Hutchinson, B. B.;
Milligan, W. O. Inorg. Chem. 1980, 19, 165–169.
(41) The geometric parameter τ is defined as τ = |(α À β)|/60,
where α and β are the two basal angles in pseudo-square pyramidal
geometry (see ref 42). The τ-value is 0.0 in idealized square-planar
geometries and 1.0 in idealized trigonal bipyramidal geometries.
(42) Addison, A. W.; Rao, T. N.; Reedijk, J.; Vanrijn, J.; Verschoor,
G. C. J. Chem. Soc., Dalton Trans. 1984, 1349–1356.
(21) Mehn, M. P.; Fujisawa, K.; Hegg, E. L.; Que, L., Jr. J. Am. Chem.
Soc. 2003, 125, 7828–7842.
(22) (a) Brown, R. S.; Huguet, J. Can. J. Chem. 1980, 58, 889–901.
(b) Breslow, R.; Hunt, J. T.; Smiley, R.; Tarnowski, T. J. Am. Chem. Soc.
1983, 105, 5337–5342. (c) Slebocka-Tilk, H.; Cocho, J. L.; Frakman, Z.;
Brown, R. S. J. Am. Chem. Soc. 1984, 106, 2421–2431. (d) Allen, W. E.;
Sorrell, T. N. Inorg. Chem. 1997, 36, 1732–1734. (e) Kimblin, C.; Allen,
W. E.; Parkin, G. J. Chem. Soc., Chem. Commun. 1995, 1813–1815.
(f) Lynch, W. E.; D. M. Kurtz, J.; Wang, S.; Scott, R. A. J. Am. Chem. Soc.
1994, 116, 11030–11038. (g) Kunz, P. C.; Reiss, G. J.; Frank, W.; Klaui,
W. Eur. J. Inorg. Chem. 2003, 3945–3951.
(43) Equivalency between the pyrazole ligands was also found in 1H
NMR studies of five-coordinate (Ph2Tp)Fe(II)(α-keto carboxylate)
complexes. See ref 21.
(44) T1-values for triflate anions in the [3-acacX]OTf series were
found to lie near 0.5 s, which is shorter than the value of 4 s measured for
[NBu4]OTf. This result suggests that the triflate anions are weakly
associated with the [Fe(PhTIP)(acacX)]+ units in solution.
(45) Spin-polarization lowers the energies of the spin-up (α) Fe d
orbitals relative to their spin-down (β) counterparts, resulting in
substantial mixing of the former with ligand-based orbitals. For this
reason, only the spin-down Fe d-based MOs are shown in Figure 6.
(46) The gap between the acac HOMO and Fe d-orbital energies is
the smallest for the 5C model of [3-acac]+, since the acac ligand is
electron rich and the Fe-d-orbitals are relatively stabilized by the PhTIP
ligand and 5C environment.
(23) Kunz, P. C.; Klaui, W. Collect. Czech. Chem. Commun. 2007,
72, 492–502.
(24) (a) Malkhasian, A. Y. S.; Nikolovski, B.; Kucera, B. E.; Loloee,
R.; Chavez, F. A. Z. Anorg. Allg. Chem. 2007, 633, 1000–1005. (b) Batten,
M. P.; Canty, A. J.; Cavell, K. J.; Ruther, T.; Skelton, B. W.; White, A. H.
Acta. Crystallogr., Sect. C 2004, 60, M311–M313.
(25) Wu, F. J.; Kurtz, D. M. J. Am. Chem. Soc. 1989, 111, 6563–6572.
(26) (a) Wu, F. J.; Kurtz, D. M.; Hagen, K. S.; Nyman, P. D.; Debrunner,
P. G.; Vankai, V. A. Inorg. Chem. 1990, 29, 5174–5183. (b) Vankai, V. A.;
Newton, M. G.; Kurtz, D., Jr. Inorg. Chem. 1992, 31, 342–343.
(27) Diebold, A. R.; Neidig, M. L.; Moran, G. R.; Straganz, G. D.;
Solomon, E. I. Biochemistry 2010, 49, 6945–6952.
(28) Malbosc, F.; Chauby, V.; Serra-Le Berre, C.; Etienne, M.;
Daran, J. C.; Kalck, P. Eur. J. Inorg. Chem. 2001, 2689–2697.
(29) Kitajima, N.; Fujisawa, L.;Fujimoto, C.;Moro-oka, Y.; Hashimoto,
S.; Kitagawa, T.; Toriumi, K.; Tatsumi, K.; Nakamura, A. J. Am. Chem. Soc.
1992, 114, 1277–1291.
(30) Sheldrick, G. M. Acta. Crystallogr., Sect. A 2008, 64, 112–122.
(31) Dolomanov, O. V.; Bourhis, L. J.; Gildea, R. J.; Howard, J. A. K.;
Puschmann, H. J. Appl. Crystallogr. 2009, 42, 339–341.
(32) Neese, F. ORCA - an ab initio, Density Functional and Semi-
empirical Program Package, version 2.7; University of Bonn: Bonn,
Germany, 2009.
(33) (a) Becke, A. D. J. Chem. Phys. 1986, 84, 4524–4529. (b) Perdew,
J. P. Phys. Rev. B 1986, 33, 8822–8824.
11989
dx.doi.org/10.1021/ic201115s |Inorg. Chem. 2011, 50, 11978–11989