Structural and spectroscopic characterization of iron(II), cobalt(II), and nickel(II) ortho-dihalophenolate complexes: Insights into metal-halogen secondary bonding

Article abstract of DOI:10.1021/ic501424e

Metal complexes incorporating the tris(3,5-diphenylpyrazolyl)borate ligand (Tp^Ph2) and ortho-dihalophenolates were synthesized and characterized in order to explore metal-halogen secondary bonding in biorelevant model complexes. The complexes Tp^Ph2ML were synthesized and structurally characterized, where M was Fe(II), Co(II), or Ni(II) and L was either 2,6-dichloro- or 2,6-dibromophenolate. All six complexes exhibited metal-halogen secondary bonds in the solid state, with distances ranging from 2.56 ? for the Tp^Ph2Ni(2,6- dichlorophenolate) complex to 2.88 ? for the Tp^Ph2Fe(2,6-dibromophenolate) complex. Variable temperature NMR spectra of the Tp^Ph2Co(2,6-dichlorophenolate) and Tp^Ph2Ni(2,6-dichlorophenolate) complexes showed that rotation of the phenolate, which requires loss of the secondary bond, has an activation barrier of ～30 and ～37 kJ/mol, respectively. Density functional theory calculations support the presence of a barrier for disruption of the metal- halogen interaction during rotation of the phenolate. On the other hand, calculations using the spectroscopically calibrated angular overlap method suggest essentially no contribution of the halogen to the ligand-field splitting. Overall, these results provide the first quantitative measure of the strength of a metal-halogen secondary bond and demonstrate that it is a weak noncovalent interaction comparable in strength to a hydrogen bond. These results provide insight into the origin of the specificity of the enzyme 2,6-dichlorohydroquinone 1,2-dioxygenase (PcpA), which is specific for ortho-dihalohydroquinone substrates and phenol inhibitors.

Full text of DOI:10.1021/ic501424e

Article
pubs.acs.org/IC
Structural and Spectroscopic Characterization of Iron(II), Cobalt(II),
and Nickel(II) ortho-Dihalophenolate Complexes: Insights into
Metal−Halogen Secondary Bonding
Timothy E. Machonkin,*,† Monica D. Boshart,† Jeremy A. Schofield,† Meghan M. Rodriguez,‡
Katarzyna Grubel,‡ Dalia Rokhsana,† William W. Brennessel,‡ and Patrick L. Holland*,‡,§
†Department of Chemistry, Whitman College, Walla Walla, Washington 99362, United States
‡Department of Chemistry, University of Rochester, Rochester, New York 14627, United States
S
* Supporting Information
ABSTRACT: Metal complexes incorporating the tris(3,5-diphenylpyrazolyl)borate ligand
(TpPh2) and ortho-dihalophenolates were synthesized and characterized in order to explore
metal−halogen secondary bonding in biorelevant model complexes. The complexes TpPh2ML
were synthesized and structurally characterized, where M was Fe(II), Co(II), or Ni(II) and L was
either 2,6-dichloro- or 2,6-dibromophenolate. All six complexes exhibited metal−halogen
secondary bonds in the solid state, with distances ranging from 2.56 Å for the TpPh2Ni(2,6-
dichlorophenolate) complex to 2.88 Å for the TpPh2Fe(2,6-dibromophenolate) complex. Variable
temperature NMR spectra of the TpPh2Co(2,6-dichlorophenolate) and TpPh2Ni(2,6-dichlor-
ophenolate) complexes showed that rotation of the phenolate, which requires loss of the
secondary bond, has an activation barrier of ∼30 and ∼37 kJ/mol, respectively. Density
functional theory calculations support the presence of a barrier for disruption of the metal−
halogen interaction during rotation of the phenolate. On the other hand, calculations using the
spectroscopically calibrated angular overlap method suggest essentially no contribution of the
halogen to the ligand-field splitting. Overall, these results provide the first quantitative measure of the strength of a metal−
halogen secondary bond and demonstrate that it is a weak noncovalent interaction comparable in strength to a hydrogen bond.
These results provide insight into the origin of the specificity of the enzyme 2,6-dichlorohydroquinone 1,2-dioxygenase (PcpA),
which is specific for ortho-dihalohydroquinone substrates and phenol inhibitors.
contrast, the native substrate of PcpA is a chlorinated
compound, 2,6-dichlorohydroquinone.
INTRODUCTION
■
Pentachlorophenol is a toxic xenobiotic compound that was
introduced in the early twentieth century as a pesticide and is
still used today as a wood preservative.1 Although the use of
pentachlorophenol has been restricted, it is still commonly
found in the environment. Interestingly, in the short time
pentachlorophenol has been present in the environment, the
bacterium Sphingobium chlorophenolicum has evolved to fully
degrade it.2 The enzyme 2,6-dichlorohydroquinone 1,2-
dioxygenase (PcpA) provides the key step in this degradation
process, using O2 to cleave the aromatic ring of 2,6-
dichlorohydroquinone.3−5
The structure of PcpA is similar to a well-studied class of
enzymes,6 the catechol extradiol dioxygenases (EDOs),7,8 as
confirmed by a recent crystal structure.9 Both are mononuclear
nonheme iron(II)-containing dioxygenases with similar active
site structures. They are also functionally comparable: while
PcpA cleaves the aromatic ring of hydroquinones, EDOs cleave
the aromatic ring of catechols. Despite of the similarities
between PcpA and the EDOs, they differ greatly in their
behavior toward halogenated substrates. In general, EDOs are
not active toward chlorinated substrates, and instead such
compounds are potent mechanism-based inactivators.10−13 In
Many aspects of the mechanism of PcpA and the basis for its
substrate specificity remain unknown. Kinetic studies have
shown that PcpA exhibits a strong selectivity toward ortho-
dihalohydroquinone substrates and phenol inhibitors over the
corresponding methyl-substituted versions.14 The origin of this
unique specificity is currently unknown, and there are not yet
crystallographic data for PcpA interacting with substrates or
inhibitors. In order to identify the most important factors,
model complexes would be valuable tools to explore the
possible binding modes of substrates and inhibitors to metal
ions like the one in the PcpA active site. In particular, we
anticipated that iron(II) complexes with ortho-halophenols
could demonstrate bonding factors that explain the preference
of PcpA for halogenated substrates and inhibitors.
In our previous work, the supporting ligand was cis,cis-1,3,5,-
tris(E)-(tolylideneimino)-cyclohexane (TACH-o-tolyl), a tri-
dentate facially coordinating ligand. The resulting [(TACH-o-
tolyl)FeL]+ complexes, where L was 2-chloro-, 2,6-dichloro-, or
2,6-dibromophenolate, each exhibited short interactions
Received: June 17, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ic501424e | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
37,38
1
between the halogen substituent and the iron(II) center,
ranging from 2.84 to 3.12 Å.15 These short interactions are
classified as “metal−halogen secondary bonds,” which are
formally defined as halocarbon complexes having a halogen-
metal distance considerably below the sum of their van der
Waals radii but longer than the metal-halide bond length.16−20
On the basis of nuclear quadrupole resonance (NQR)
spectroscopy results, Wulfsberg proposed the criterion that
the metal−halogen distance must be shorter than the average
metal-halide distance plus 1 Å.19 Using the average four-
coordinate Fe(II)-chloride bond length of 2.30 Å,21 all of the
[(TACH-o-tolyl)FeL]+ complexes meet this criterion for
metal−halogen secondary bonding. This shows that an ortho-
halophenolate can form an additional interaction with the
iron(II) center. This direct interaction between the halogen and
the metal could increase the binding affinity of halogenated
substrates and inhibitors to PcpA and thus contribute to the
unusual substrate specificity of this enzyme.
However, the postulation of a metal−halogen secondary
bond based on distance alone is necessarily tentative. Metal−
halogen secondary bonds are relatively rare,15−19,22−27 and
there are few techniques capable of detecting these bonds other
than X-ray crystallography and NQR.17−20 Consequently, the
characteristics of these secondary bonds are in need of greater
understanding. In particular, to our knowledge there are no
quantitative studies of the strength of this interaction, nor is it
clear to what extent the halogen can be classified as a true
“ligand” to the metal center. More detailed studies of TACH-o-
tolyl complexes were not possible because the supporting
ligand was unstable with respect to isomerization and
dissociation.15 Therefore, we chose a tris(pyrazolyl)borate
(Tp) ligand28−30 as a more stable tridentate supporting group
for the synthesis of metal ortho-halophenolate complexes. To
provide a similar level of steric bulk to the TACH-o-tolyl ligand,
the ligand tris(3,5-diphenylpyrazolyl)borate (TpPh2) was
selected for the present study.31 A series of Fe(II)-phenolate
complexes have been reported previously with the ligand
tris(3,5-diisopropylpyrazolyl)borate (TpiPr2), including a 2,6-
dichlorophenolate complex (although it was not structurally
characterized).32 Ni(II)-2,6-diisopropylphenolate complexes
have been reported with the ligands tris(3,5-
dimethylpyrazolyl)borate (Tp*) and tris(3-phenyl-5-
methylpyrazolyl)borate.33 In addition, Fiedler and co-workers
have reported a series of TpPh2FeL complexes where L is a
either one of several hydroquinonates or 2-amino-4,6-di-tert-
butylphenolate.34−36 Here we report the synthesis and
characterization of a series of metal ortho-dihalophenolate
complexes, and examine the nature of the metal−halogen
interaction using crystallography, spectroscopy, and computa-
tional analysis.
procedures and characterized by H NMR spectroscopy.
To each of these complexes was added the sodium salt of either
2,6-dichlorophenol (2,6-dcp) or 2,6-dibromophenol (2,6-dbp)
(see Scheme 1). In each case, the reactions were performed
Scheme 1. Synthesis of Compounds 1 (M = Fe(II), X = Cl, Y
= Br), 2 (M = Fe(II), X = Br, Y = Br), 3 (M = Co(II), X = Cl,
Y = Cl), 4 (M = Co(II), X = Br, Y = Cl), 5 (M = Ni(II), X =
Cl, Y = Br), and 6 (M = Ni(II), X = Br, Y = Br)
anaerobically in THF. Reaction of TpPh2FeBr with these
sodium salts yielded colorless TpPh2Fe(2,6-dcp) (1) and
TpPh2Fe(2,6-dbp) (2) in 71% and 76% yield, respectively.
Reactions of TpPh2CoCl with 2,6-dcp and 2,6-dbp gave purple
TpPh2Co(2,6-dcp) (3) and TpPh2Co(2,6-dbp) (4) in 74% and
80% yield, respectively. Reactions of TpPh2NiBr with 2,6-dcp
and 2,6-dbp gave orange TpPh2Ni(2,6-dcp) (5) and TpPh2Ni-
(2,6-dbp) (6) in 58% and 71% yield, respectively. Complexes
3−6 were air-stable, while compounds 1 and 2 were handled
under anaerobic conditions.
Crystallographic analysis of 1−6 gave structures that are
shown in Figure 1. Complexes 1, 2, 5, and 6 cocrystallized with
disordered solvent. Selected crystallographic data for 1−6 are
summarized in Table 1. For all six complexes, the bond angles
reveal a distorted five-coordinate geometry. In a trigonal
bipyramidal description, N3 and X1 are the axial positions, and
O1, N1, and N5 are the equatorial positions. The axial M−N3
distance is slightly longer than the equatorial M−N distances,
and all the M−N distances decrease in the order Fe > Co > Ni,
in accordance with the trend in metal ion radii.21 For each of
the six complexes, the sum of the angles of the equatorial
ligands is nearly 360° (354−360°), and the axial-to-equatorial
bond angles are ∼90° for both N3−M−Neq (86°−93°) and
X1−M−Neq (92°−101°), supporting the description of a
trigonal bipyramidal geometry. However, the complexes exhibit
varying levels of distortion toward a square pyramidal
geometry, with one O1−M−Neq bond angle significantly larger
than the other (116°−126° versus 132°−150°), which in turn
leads to τ values39 that range from 0.32 and 0.33 for complexes
1 and 5 up to a value of 0.72 for 3. Distortions from an ideal
trigonal bipyramidal geometry are found for the N3−M−O1
bond angle (94°−109°) and the X1−M−O1 bond angle (73°−
81°). In a square-pyramidal description, N1 (N5 in 2 and 3) is
the axial position. The Nax−M−Neq angles are 90°−96°, the
Nax−M−X1 angles are 86°−100°, and the Neq−M−X1 angles
are 166°−176°. The most notable distortion away from a
square pyramidal description again occurs for the N−M−O1
angles, in which the Nax−M−O1 angle is much greater than 90°
and the Neq−M−O1 angle is much less than 180°.
RESULTS
■
Synthesis and Structural Characterization. The bulky
tridentate ligand TpPh2 was synthesized according to literature
procedures.31 While the active site in PcpA contains iron(II),
the ortho-dihalophenolate complexes reported herein contain
iron(II), cobalt(II), and nickel(II) in order to explore trends
with respect to the metal ion and gain deeper insight into
metal−halogen bonding.
All six complexes exhibited metal−halogen secondary
bonding, as judged using Wulfsberg’s geometric criterion.19
The metal−halogen distances decreased in the order Fe > Co >
Ni, ranging from 2.89 Å for 2 to 2.56 Å for 5, in accordance
with ionic radii;21 however, there was no consistent pattern for
the M−Cl versus M−Br distance compared to halogen size.40,41
The M−O1−C46phen bond angle decreased in the same trend,
The starting complex TpPh2FeBr was prepared in a similar
manner to the previously reported TpiPr2FeCl complex and
characterized by 1H NMR spectroscopy.32 Likewise, TpPh2CoCl
and TpPh2NiBr were prepared according to the literature
B
dx.doi.org/10.1021/ic501424e | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 1. Solid state structures of 1−6, with thermal ellipsoids shown at 50% probability. Hydrogen atoms are omitted for clarity. The metal−
halogen secondary bond in each complex is indicated by a dashed line. A, TpPh2Fe(2,6-dcp) (1); B, TpPh2Fe(2,6-dbp) (2); C, TpPh2Co(2,6-dcp) (3);
D, TpPh2Co(2,6-dbp) (4); E, TpPh2Ni(2,6-dcp) (5); F, TpPh2Ni(2,6-dbp) (6).
Table 1. Selected Crystallographic Data for TpPh2Fe(2,6-dcp) (1), TpPh2Fe(2,6-dbp) (2), TpPh2Co(2,6-dcp) (3), TpPh2Co(2,6-
dbp) (4), TpPh2Ni(2,6-dcp) (5), and TpPh2Ni(2,6-dbp) (6)
1
2
3
4
5
6
Bond Lengths (Å)
2.038(3)
M−N1
M−N3
M−N5
M−O1
M−X1
2.1135(17)
2.1250(16)
2.0973(16)
1.8995(14)
2.8618(8)
2.089(2)
2.136(2)
2.103(2)
2.045(4)
2.092(5)
2.062(5)
1.882(4)
2.8105(10)
2.0319(16)
2.0387(15)
2.0348(15)
1.9035(13)
2.5601(6)
2.024(6)
2.043(6)
2.027(6)
1.905(5)
2.6846(12)
2.088(3)
2.036(3)
1.8997(17)
2.8853(5)
1.885(3)
2.8325(12)
Bond Angles (deg)
O1−M−N1
O1−M−N5
N1−M−N5
N3−M−O1
N3−M−N1
N3−M−N5
X1−M−O1
X1−M−N1
X1−M−N5
N3−M−X1
M−O1−C46phen
C47phen−X1−M
119.13(6)
147.67(7)
92.34(6)
98.34(6)
92.21(6)
86.85(6)
72.95(4)
100.76(4)
95.34(4)
166.73(5)
133.66(13)
92.45(7)
143.31(8)
121.48(8)
93.37(8)
102.25(8)
86.32(8)
91.93(8)
76.22(6)
91.86(6)
93.4(6)
132.45(12)
125.66(12)
95.63(12)
108.90(12)
91.47(12)
90.15(12)
75.03(9)
119.52(18)
144.10(18)
95.34(18)
98.14(17)
91.75(18)
88.37(18)
78.39(12)
93.24(12)
92.28(13)
174.88(13)
130.8(4)
116.67(6)
149.52(6)
93.50(6)
93.95(6)
93.48(6)
88.05(6)
80.14(4)
96.82(4)
92.74(4)
169.60(5)
125.79(13)
94.92(7)
118.5(2)
146.7(2)
94.3(2)
95.9(2)
93.4(2)
87.2(2)
80.67(15)
94.06(16)
92.02(16)
172.58(17)
129.1(5)
89.9(2)
86.80(9)
86.18(9)
174.53(6)
133.01(17)
88.35(9)
175.77(9)
130.1(3)
91.35(14)
88.64(19)
Torsion Angles (deg)
M−O1−C46phen−C47phen
N3−M−O1−C46phen
−13.7(3)
−177.32(17)
14.0(4)
15.4(5)
−14.3(8)
−169.0(5)
−5.3(3)
176.33(15)
−8.1(10)
−179.8(6)
172.2(2)
163.5(3)
τ values
0.72
0.32
0.52
0.51
0.33
0.43
C
dx.doi.org/10.1021/ic501424e | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
1H NMR spectra of all six complexes the three arms of the
TpPh2 ligand are equivalent, as are the meta protons of the
phenolate ring. Thus, rapid rotation of the phenolate ring about
the M−O1 and/or O1−C46phen bonds is occurring on the
NMR time scale. This, in turn, indicates that the metal−
halogen secondary bond is not maintained in solution at room
temperature. The same result was observed in the previously
reported TACH-o-tolyl complexes.15 In the case of [(TACH-o-
tolyl)Fe(2,6-dcp)]+, 1H NMR spectra recorded at temperatures
down to −40 °C did not yield decoalescence of the phenolate
meta proton signals.
to accommodate the shorter M−X1 distance. The C47phen
−
X1−M bond angles were all ∼90° (87°−94°), consistent with
what has been seen before for metal−halogen secondary
bonding15,18,19 and with what is known about the anisotropic
charge density of the halogen.42 In all complexes, the phenyl
ring is nearly coplanar with the N3−M−O1 plane, as shown by
the N3−M−O1−C46phen and M−O1−C46phen−C47phen tor-
sion angles, thus allowing for the maximum interaction between
the halogen substituent and the metal center.
1H NMR Spectroscopic Studies. Complexes 1−6 were
analyzed by 1H NMR spectroscopy at room temperature
(Figure S1). These spectra were consistent with the presence of
the desired TpPh2 M(2,6-dihalophenolate) complexes in
solution, and the large shifts indicate that the complexes are
paramagnetic. Based upon integrations and expected patterns of
hyperfine shifts, the pyrazole H4 and the meta and para protons
of the phenolate could be readily identified and assigned (Table
2). Some of the resonances, particularly the TpPh2 phenyl
protons in the nickel(II) complexes, were not resolved and
therefore could not be assigned.
The greater solubility of the present complexes in a wider
range of solvents allowed the acquisition of NMR spectra at
lower temperatures. Complexes 3 and 5 were selected for
variable temperature NMR (VT-NMR) studies. Figure 2A
1
shows the H NMR spectra of 3 in CD2Cl2 at temperatures
from +20 to −89.5 °C. The line width (full width at half-
maximum) of the phenolate para proton signal increased by
roughly 4-fold, whereas the meta proton signal increased by
more than 100-fold (Figure 2B). The increased line width of
the phenolate para proton signal is likely due to greater
hyperfine broadening at lower temperature caused by some
temperature dependence of the electronic relaxation time.43
Although complete decoalesence of the meta protons was not
observed, the steep increase in the line width at low
temperature is most reasonably explained by exchange
broadening from slower rotation of the phenolate ring.
Assuming a two-state model for this exchange broadening,
the temperature-dependence can be fit with two adjustable
parameters: the activation barrier, ΔG‡, and the difference in
the chemical shift of the two meta proton signals, Δδ.43 We
accounted for the change in line width of the meta proton
signal from hyperfine broadening by assuming it to be equal to
that of the para proton signal. Since an estimation of the
contact contribution (from the DFT calculation described
below) and dipolar contribution to the paramagnetic relaxation
1
Table 2. Selected Room Temperature H NMR Data for
Complexes 1−6
complex
pyrazole H4
phenolate meta
phenolate para
Chemical Shifts (ppm)
1
2
3
4
5
6
57.9
58.4
68.1
67.4
72.4
71.3
54.2
54.3
35.2
37.3
29.0
28.0
−23.7
−26.2
−34.1
−34.7
−25.2
−25.3
In the solid state structures, these complexes show
approximate Cs symmetry, and the meta protons of the
phenolate are inequivalent. However, in the room-temperature
1
Figure 2. (A) Variable temperature H NMR spectra of 3 in CD2Cl2. The phenolate meta and para protons are highlighted. (B) The temperature
◊
■
○
▲
dependence of line width of the pyrazole H4 ( ), phenolate meta ( ), Tp-ortho-phenyl ( ), and phenolate para ( ) proton signals. The fit to the
temperature dependence of the line width of the phenolate meta protons due to exchange broadening as described in the text is shown as a solid line.
D
dx.doi.org/10.1021/ic501424e | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 3. UV−vis−NIR absorption spectra of 3 (A) and 5 (B) in CHCl3 at room temperature. The thin black bars indicate the calculated energies of
the LF transitions from an AOM calculation, and the thick gray bars indicates the calculated energies of the phenolate oxygen pπ to metal CT
transitions with high oscillator strengths from a TD-DFT calculation.
yields an average line width for the meta protons that is ∼1.8
that of the para proton, this assumption is justified. With these
assumptions, the fit to the temperature dependence of the meta
proton line width (Figure 2B, solid line) yielded ΔG‡ = 30.3
0.7 kJ/mol and Δδ = 3.4 0.7 ppm. When the meta proton
line width was fit assuming that exchange broadening was the
sole contribution, the values of ΔG‡ and Δδ were the same
within uncertainty. The pyrazole H4 and ortho-phenyl proton
signals of the TpPh2 ligand also showed evidence of exchange
broadening at low temperature, but were not analyzed
quantitatively.
phenolate) complex (28000 cm−1, ε = 1300 M−1 cm−1).44
The other peaks were assigned as ligand field (LF) transitions
of the cobalt(II). The absorption peaks for 5 appeared at
22600, 12700, 11400, and 6800 cm−1 (443, 788, 877, and 1470
nm). Likewise, the intense transition at 22600 cm−1 (ε = 2300
M−1 cm−1) was assigned as a phenolate-O pπ to nickel(II) CT
transition, analogous to the previously reported thiophenolate-S
pπ to nickel(II) CT transition (23200 cm−1, ε = 3060 M−1
cm−1),44 while the other peaks were assigned as LF transitions.
Mull absorption spectra of 3 and 5 were obtained in the
UV−visible region, and the features of these spectra appeared
at the same wavelengths as in the solution spectra. This
indicates that the electronic structures of these complexes in
solution are essentially the same as that in the solid state. This
strongly argues that, while these complexes are fluxional in
Similar results were obtained using the VT-NMR spectra of 5
(Figure S2). Applying the same data analysis described above,
the fit to the temperature dependence of the meta proton line
width (Figure S2B, solid line) yielded ΔG‡ = 37 1 kJ/mol
1
and Δδ = 0.7
0.5 ppm. However, the reliability of these
solution (as indicated by the H NMR spectra), the geometric
values is reduced compared with the results for complex 3
because of two factors. First, NMR spectra of this sample could
only be obtained down to −70 °C. Second, the line width of
the para proton signal increased with decreasing temperature to
a greater extent (about 7-fold), presumably due to a larger
temperature dependence to the hyperfine broadening. Since a
larger proportion of the line width of the meta protons is also
likely due to hyperfine broadening and not just exchange
broadening, the calculated values of ΔG‡ and Δδ will be much
more sensitive to this correction. The assumption made here,
that this paramagnetic contribution to the line broadening is
the same for both the meta and para protons, may not be
accurate, which in turn reduces the accuracy of the fitted values
for ΔG‡ and Δδ.
Electronic Absorption Spectroscopic Studies. UV−
vis−NIR absorption spectra were obtained for complexes 3−6
in CHCl3. The spectra of 3 and 5 are given in Figure 3 (the
spectra of 4 and 6 were nearly identical and are given in Figure
S3). For 3, prominent peaks were seen at 26800, 17480, 15270,
12330, and 6810 cm−1 (373, 572, 655, 811, and 1469 nm). The
transition at 26800 cm−1 was the most intense (ε = 1150 M−1
cm−1) and was assigned as a charge transfer (CT) transition
from the phenolate-O pπ to the cobalt(II), analogous to the
thiophenolate-S pπ to cobalt(II) CT transition reported
previously for a closely related TpiPr2Co(penta¯uorothio-
structure of the energy minimum in solution resembles the
structure observed in the solid state. This, in turn, suggests that
the metal−halogen secondary bond is present in the minimum
energy solution structure.
The LF transitions of 3 were analyzed by the angular overlap
method (AOM) using the AOMX program.45 The AOM
parameters of the analogous (Tpt‑Bu,R)CoL complexes (R = H,
−
methyl, or 2′-thienyl; L = NCS−, NCO−, N3 , or Cl−) of
Krzystek et al.46 were used as a starting point. The geometry
from the crystal structure was used. The pyrazole nitrogens (N-
pz) were treated as σ-donors and the phenolate oxygen (O-
phen) was treated as a σ-donor and an out-of-plane π-donor.47
An excellent fit was obtained with values of εσ(N-pz) = 4120,
εσ(O-phen) = 2050, επ(O-phen) = 1660, and B = 698 cm−1
(C/B was kept fixed at 4.3). The calculated energies of the LF
transitions are shown as thin black bars in Figure 3A. A Cs
symmetrized structure of 3 was also used in an AOM analysis of
the LF transitions. In this structure, the O1−Co−Neq, N3−
Co−Neq, and Cl−Co−Neq bond angles were averaged, and the
phenolate was taken to be exactly coplanar with the N3−Co−O
plane. Again, an excellent fit to the LF transitions could be
obtained with very similar AOM parameters to those above
(the only significant change was that επ(O-phen) increased to
1930 cm−1). When the halogen was included as a σ-donor in
the fit, optimization led to an unphysical (negative) value for
E
dx.doi.org/10.1021/ic501424e | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 3. Energies and Major Orbital Contributions of the β-Spin MOs with Predominantly Metal d and Phenolate O Orbital
Character, from the PBE1PBE Calculations
Tp*Co(2,6-dcp)
energy (eV)
Tp*Ni(2,6-dcp)
energy (eV)
orbital
label
contributions
label
contributions
2
dz
LUMO+2
LUMO+1
LUMO
−0.30
−0.53
−1.23
−5.75
−6.64
−6.77
−7.80
−8.29
50% M, 2% Nax, 9% Neq, 2% Cl
56% M, 3% Nax, 5% Neq
LUMO+1
LUMO
−1.08
−1.30
−8.67
−5.82
−9.22
−8.78
−7.39
81% M, 5% Nax, 3% Cl
80% M, 6% Neq
dxy
2
2
dx −y
86% M, 3% O, 5% Neq
HOMO−10
HOMO
67% M, 7% O
phenolate-O pπ
dxz
HOMO
7% M, 22% O, 60% phenyl C, 9% Cl
47% M, 3% O, 16% Neq
3% M, 23% O, 64% phenyl C, 9% Cl
62% M, 15% O, 4% Cl
HOMO−1
HOMO−4
HOMO−8
HOMO−10
HOMO−13
HOMO−11
HOMO−8
a
dyz
28% M, 12% Nax, 5% Neq, 3% Cl
29% M, 14% O, 17% Neq
20% M, 44% O, 16% Cl
47% M, 14% phenyl C, 26% Cl
phenolate-O pσ
phenolate-O pσ
28% M, 44% O, 3% Neq
a
12% from proximal (coordinating) Cl, 14% from distal Cl.
εσ(Cl). Constraining εσ(Cl) to be positive led to an optimum fit
with εσ(Cl) ≈ 0. If the εσ(N-pz) parameters were allowed to
differ for the axial versus equatorial positions, then fits could be
obtained with any combination of εσ(Nax-pz) and εσ(Cl) such
that the sum was ∼3100 cm−1. Keeping the angular overlap
parameters fixed while altering the angle ψ, which corresponds
to rotation of the phenolate, altered the energies of the LF
transitions by as much as 1700 cm−1. These results suggest that
(1) the phenolate remains approximately coplanar with the
N3−Co−O plane in the minimum energy solution geometry,
and (2) the metal−halogen secondary bond makes a negligible
contribution to the LF splitting of the cobalt(II).
A similar analysis of the LF transitions of 5 was performed,
using the AOM parameters of Tp*NiL (Tp* = tris(3,5-
dimethylpyrazolyl)borate; L = Cl−, Br−, I−) of Derochers et
al.48 as well as of complex 3 as a starting point. Since only three
LF transitions could be clearly resolved (three more were
assumed to be under the shoulder observed at 17000−19000
cm−1), επ(O-phen) was constrained to be equal to 0.8 of εσ(O-
phen). Again, using the geometry from the crystal structure, an
excellent fit was obtained with εσ(N-pz) = 4660, εσ(O-phen) =
1830, επ(O-phen) = 1470, and B = 539 cm−1 (C/B was kept
fixed at 4.7), and the calculated LF transition energies are
shown in Figure 3B. As in the case of 3, a Cs symmetrized
structure of 5 was used, and again a reasonable fit to the LF
transitions could be obtained with very similar AOM
parameters. Likewise, inclusion of the metal−halogen secon-
dary bond was not necessary to yield reasonable results for the
energies of the LF transitions.
Density Functional Theory Calculations. Density func-
tional theory calculations were performed on models of 3 and 5
with Gaussian49 using both the B3LYP50 and PBE1PBE51
functionals. The phenyl groups of the TpPh2 ligand were
replaced with methyl groups (Tp*). Geometry optimization
with the 6-311++G(d) basis set for Cl and 6-311+G(d) basis
set for all other atoms yielded excellent agreement with the
experimentally determined structures with both functionals,
with the exception of the metal−halogen distance (Table S1).
Both DFT methods systematically calculated the metal−
halogen distance to be longer than the experimentally
determined values; however, M−X distances calculated with
PBE1PBE for Tp*Co(2,6-dcp) and Tp*Ni(2,6-dcp) were 0.07
and 0.14 Å longer than the experimental values, respectively,
while B3LYP gave values that were 0.20 and 0.29 Å longer.
Constraining the metal-chlorine distance in Tp*Co(2,6-dcp) to
the experimental value altered the total energy by only 1.5 kJ/
mol in the B3LYP calculation. This is consistent with what has
been previously postulated regarding metal−halogen bonding,
namely, that the potential energy well is much shallower for a
secondary bond than for a covalent bond, and thus the energy
varies only slightly as the distance changes.16
The ground state electronic structures of Tp*Co(2,6-dcp)
and Tp*Ni(2,6-dcp) are typical for high-spin cobalt(II) and
nickel(II) complexes, and strongly resemble the previously
reported pentafluorothio-phenolate complexes.44 There are
2
three unoccupied β-spin orbitals in the Co(II) complex (dz ,
2
2
dxy, and dx −y , defining the z and x axes to be approximately
along the M−Nax and M−O bond vectors, respectively) and
2
two in the Ni(II) complex (dz and dxy). In both cases, the β-
spin HOMO is the phenolate-O pπ orbital. Phenolate-O pσ
orbital character was found distributed in two closely spaced β-
spin MOs at much lower energies in Tp*Co(2,6-dcp) and in
one in Tp*Ni(2,6-dcp). The energies and major orbital
contributions, determined from the Mulliken orbital contribu-
tions52−55 from the PBE1PBE calculations with a smaller basis
set (6-311G(d)) analyzed with AOMix,56,57 are given in Table
3, and contour surface plots of the orbitals of Tp*Co(2,6-dcp)
are shown in Figure S4 (very similar results were obtained with
B3LYP). Small Cl contributions are seen in two of the β-spin d
2
orbitals in both complexes: dz has a small σ interaction, and dyz
has a small π interaction. In Tp*Ni(2,6-dcp), dxz also shows
some interaction with the Cl.
Mayer bond orders58−61 were calculated in order to quantify
the extent of the metal−halogen interaction. Since calculations
with the large, diffuse basis set used for the geometry
optimization gave unphysical values for the bond orders,
single-point calculations with the 6-311G(d) basis set were
used. These results yielded a M−Cl bond order that is about
15−38% of the bond orders for the M−O and M−N bonds
(Table 4). Similar results were obtained from the Wiberg and
Lowdin bond indices (Table S2).62 Thus, the calculations show
̈
that while the chlorine has some bonding interaction with the
metal, it is considerably smaller than that of the “normal”
metal−ligand bonds.
In order to obtain a quantitative estimate of the strength of
the metal-chlorine interaction, the barrier to rotation of the
Table 4. Mayer Bond Orders for Tp*M(2,6-dcp)
bond
M−O
M−N3(ax)
M−N1(eq)
M−N5(eq)
M−Cl
M = Co
M = Ni
0.398
0.244
0.255
0.255
0.059
0.351
0.271
0.265
0.287
0.100
F
dx.doi.org/10.1021/ic501424e | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
phenolate about the O−C bond was examined in Tp*Co(2,6-
dcp). For Tp*Co(2,6-dcp), the Co−O1−C46phen-C47phen
torsion angle in the optimized geometry was 0°. This angle
was then was fixed at 15°, and the geometry was reoptimized.
This was repeated at 15° increments up to 90°. From this, the
calculated barrier to rotation of the phenolate was found to be
10.3 kJ/mol with B3LYP and 13.4 kJ/mol with PBE1PBE
(Figure 4 and Table S3). Likewise for Tp*Ni(2,6-dcp), the
bonding of the phenolate-O does not contribute to the
energetics of phenolate rotation.
While the actual mode of geometric fluxionality of these
complexes in solution is likely more complex, the calculated low
barrier to rotation of the phenolate is qualitatively consistent
with the low barrier to rotation that was determined
experimentally from VT-NMR. However, the DFT calculation
underestimated this barrier by a factor of ∼2.3 for the
cobalt(II) and by even more for the nickel(II) complex. The
lack of quantitative agreement could be due to two factors: (1)
additional steric hindrance from the phenyl substituents on the
Tp ligand could raise this barrier in the actual complex (as
noted above) and (2) DFT methods, while highly successful at
describing open-shell transition metal complexes, are often
poor at accurately calculating weak noncovalent interactions.63
While the PBE1PBE functional performed better than B3LYP
with regards to both the metal−halogen distance and the
barrier to rotation, in agreement with previous claims regarding
the utility of these functionals for the calculation of noncovalent
interactions,63 the energy difference in the latter was small.
Note that the best density functional for noncovalent
interactions, M06X, failed to give a reasonable electronic
structure description for the metal. Nonetheless, the fact that
DFT calculations yield a barrier to rotation for Tp*Co(2,6-
dcp) but not for Tp*Co(2,6-dmp) provides the strongest
evidence that the experimentally observed barrier comes mainly
from the metal−halogen interaction and not steric hindrance or
anisotropic π-bonding of the phenolate.
Figure 4. Barrier to rotation of the phenolate in Tp*Co(2,6-
dichlorophenolate) (circles) and Tp*Co(2,6-dimethylphenolate)
(squares), calculated with B3YLP (open symbols) and PBE1PBE
(filled symbols). The Co−O1−C46phen−C47phen torsion angle was
fixed at 15° increments from 0° to 90° and the geometry was
reoptimized.
Time-dependent DFT (TD-DFT), calculated with B3LYP
using the PBE1PBE-optimized structure, was used to calculate
the energies of the phenolate-O to metal CT transitions in the
cobalt(II) and nickel(II) 2,6-dcp complexes. In the fully
geometry-optimized Tp*Co(2,6-dcp), three phenolate O pπ
2
2
to cobalt(II) CT transitions are observed: pπ → dx −y (22600
energy difference with PBE1PBE between the fully optimized
geometry, where the Ni−O1−C46phen-C47phen torsion angle
was −4.8°, and one in which the torsion angle was frozen at
105° was found to be 9.7 kJ/mol (the torsion angle of 90°
created a steric clash with the methyl group of the Tp* ligand
and had difficulty converging).
−1
−1
cm−1), pπ → dxy (27100 cm ), and pπ → dz (30300 cm ).
Only the pπ → dxy has an oscillator strength >0.0005 (f =
0.0296), in excellent agreement with the experimentally
observed intense transition at 26800 cm−1 in the solution
spectrum. In the TD-DFT calculation of Tp*Co(2,6-dcp) with
the Co−O1−C46phen−C47phen torsion angle frozen at 90°, the
2
Two alternative factors must be considered that could
contribute to a barrier to rotation of the phenolate ring besides
the metal−halogen secondary bond. The first is steric
hindrance from the phenyl substituents of the TpPh2 ligand.
Because the size of the substituents on the Tp ligand was
reduced from phenyl to methyl in the DFT calculation, this
should have reduced the effect of any steric hindrance on
rotation of the phenolate in the calculations. Additionally, we
note that when the Co−O1−C46phen−C47phen dihedral angle in
3 was frozen at 90°, the steric clash between the chloro
substituent and the phenyl ring of the TpPh2 ligand could be
readily relieved by rotation of the phenyl ring. This suggests
three pπ CT transitions occur at very similar energies: pπ →
dx −y (22200 cm−1), pπ → dxy (27700 cm ), and pπ → dz
−1
2
2
2
(28000 cm−1); however, the oscillator strength of the pπ → dxy
transition has dropped to 0.0001 while that of the lower energy
2
2
pπ → dx −y transition has increased to 0.0444. Similar results
are obtained with the TD-DFT calculation of Tp*Ni(2,6-dcp).
Two closely spaced phenolate-O pπ to nickel(II) CT
transitions are observed, pπ → dxy (21700 cm−1) and pπ →
−1
2
dz (23500 cm ), with similar oscillator strengths (f = 0.0501
and 0.0216, respectively). This again agrees well with the
experimentally observed intense transition at 22600 cm−1 in the
solution spectrum.
that while there may be greater steric hindrance from the TpPh2
it is probably not severe.
,
Overall, the results indicate that the minimum energy
solution geometry in each complex is one in which the torsion
angle of the phenolate is ∼0°, since this geometry is the one
predicted to give an intense CT transition at the energy where
it is observed in the solution absorption spectrum. This, in turn,
agrees with crystal structures and with the mull UV−visible
absorption spectra. So, while complexes 3 and 5 are fluxional in
solution, there is a discrete energy minimum in the geometry.
The second factor is the anisotropic π-bonding of the
phenolate-O to the metal center. To address how this might
influence the energy barrier to rotation of the phenolate,
calculations were performed on a complex in which the 2,6-dcp
ligand was replaced with 2,6-dimethylphenolate (2,6-dmp),
with the Co−O1−C46phen−C47phen torsion angle again stepped
in 15° increments from 0° to 90°. This yielded an extremely
low barrier to rotation (Figure 4, Table S4), indicating that π-
G
dx.doi.org/10.1021/ic501424e | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 5. Comparison of M−X Distances for Complexes 1−6 and Other Iron(II), Cobalt(II), and Nickel(II) ortho-
Halophenolate Complexes, the Average Metal−Halide Bond Lengths, and the Difference (Δ M−X)
a
b
c
complex
experimental M−X distance (Å)
average bond length (Å)
ΔM−X (Å)
ref
Fe−Cl
1
2.862
2.301
2.301
2.301
2.35
0.561
0.589−0.821
0.628, 0.709
0.54
(TACH-o-tolyl)Fe(2,6-dcp)
2.890−3.122
2.929, 3.010
2.885
15
15
(TACH-o-tolyl)Fe(2-cp)
Fe−Br
Co−Cl
2
(TACH-o-tolyl)Fe(2,6-dbp)
2.841
2.35
0.49
15
3
2.833
2.254
2.254
0.579
d
Co(NMIm)2(2,4,6-tcp)2
[Co(OMe)(2,4,6-tcp)(MeOH)]4
Co(TMED)(2,4,6-tcp)2
4
3.100
0.846
22
24
25
e
d
2.62, 2.64
2.873, 2.894
2.811
2.414
0.21, 0.23
e
2.414
0.459, 0.480
0.427
Co−Br
Ni−Cl
Ni−Br
2.384
ef
,
Co(py)2(2,4,6-tbp)2
Co(NMIm)2(2,4,6-tbp)2
5
3.116, 3.120
3.027, 3.091
2.560
2.573
2.573
0.543, 0.547
0.454, 0.518
0.343
27
26
ef
,
2.217
e
d
[Ni(OMe)(2,4,6-tcp)(MeOH)]4
Ni(TMED)(2,4,6-tcp)2
6
2.483−2.530
2.586, 2.637
2.685
2.441
2.441
0.042−0.089
0.145, 0.196
23
19
e
d
2.341
0.344
a
Other abbreviations: 2-chlorophenolate (2-cp), N-methylimidazole (NMIm), 2,4,6-trichlorophenolate (2,4,6-tcp), N,N,N′,N′-tetramethyl-1,2-
b
ethanediamine (TMED), pyridine (py), 2,4,6-tribromophenolate (2,4,6-tbp). Different M−X distances are either due to multiple coordinated ortho-
c
halophenolates, or from different crystal forms and/or multiple conformers in the same unit cell. For four-coordinate divalent metal complexes,
unless otherwise noted. NQR results indicate the presence of a metal−halogen secondary bond.17,19 For six-coordinate complexes. This value was
not available in ref 21; therefore, it was determined from 31 structures in the Cambridge Structural Database of six-coordinate cobalt complexes with
one or more terminal bromide ligands where the oxidation state was unambiguously Co(II).
d
e
f
complexes were six-coordinate when including the secondary
bonds, so using a six-coordinate average Ni−Cl (2.441 Å),
ΔM−X was much smaller.
DISCUSSION
■
The model complexes reported here serve to further illuminate
the nature of the metal−halogen secondary bonding observed
with ortho-dihalophenolates, and thus the potential binding
modes of substrates and inhibitors at the active site of PcpA.14
To our knowledge, this is the only case where metal−halogen
bonding has been reported for a series of structurally
homologous complexes where the identity of both the metal
and the halogen is varied. Thus, these complexes can provide
valuable insights into the nature and influence of this
interaction. We use a combination of crystallography, NMR
spectroscopy, and spectroscopically calibrated computational
methods to quantify and elucidate this important interaction.
In the crystal structures of complexes 1−6, the halogen
occupies an axial position of a distorted trigonal bipyramidal
geometry. According to the criterion established by Wulfsberg,
a metal−halogen secondary bond is present if the metal−
halogen distance is less than the average metal-halide distance
plus 1 Å.19 By this definition, all six of these complexes exhibit
metal−halogen secondary bonding. The metal−halogen
distances of these complexes as well as those of other iron(II),
cobalt(II), and nickel(II) ortho-halophenolate complexes
exhibiting metal−halogen secondary bonding are reported in
Table 5, along with the average metal-halide bond lengths for
divalent metal complexes,21 and the difference between these
(ΔM-X). The previously reported Co(N-methylimid-
zole)2(2,4,6-trichlorophenolate)2 complex is a particularly
useful comparison, since it is also five-coordinate (τ = 0.43)
when counting the chlorine.22 The trend in metal−halogen
distance among complexes 1−6 is Fe > Co > Ni, which is in
accordance with the trend in ionic radii of these divalent metal
ions, as well as the expected increase in Zeff. The metal−
halogen distances are particularly short in the nickel(II)
complexes, although within the range of the remarkably short
distances seen in some other nickel(II)-ortho-chlorophenolate
complexes (2.483−2.53023 and 2.586 Å19). However, these
The Tp complexes reported here exhibit no clear trend in
metal−halogen distance or in ΔM−X between the 2,6-dichloro-
and 2,6-dibromophenolates. Since the van der Waals radius of
bromine is larger than that of chlorine,40,41 one might predict
that the metal−bromine distances would be longer than the
corresponding metal-chlorine distances. Conversely, if the
secondary bonding interaction were stronger for the more
polarizable halogen,64 one might predict the opposite trend.
However, the strength of secondary bonding has been
postulated to be only weakly dependent on distance,16 and
the results of our DFT calculations on these complexes suggest
only weak covalent bonding between the metal and the
halogen. Thus, the very shallow potential energy well for this
interaction may make it too difficult to discern a trend in M−Cl
versus M−Br distances even in these structurally related
complexes.
These Tp-supported complexes are all fluxional on the NMR
time scale. Therefore, one must consider whether the metal−
halogen secondary bond observed in the crystal structures is
also present in solution. A comparison of the mull and solution
spectra show that the transitions in the UV−visible region are
at the same energies, arguing that the geometry is the same in
both the solid and the minimum energy solution state
structures. Likewise, the TD-DFT calculation of the energy of
the intense phenolate-O to metal CT transition is only
consistent with a structure where the Co−O1−C46phen
−
C47phen torsion angle is ∼0°. Therefore, we conclude that the
secondary bond is not an artifact of crystallization and is
present in the minimum energy solution structure.
The ultimate question is whether the halogen is truly a
“ligand.” A simple AOM analysis of the LF transitions in the
UV−vis-NIR absorption spectra was entirely consistent with a
trigonal pyramidal geometry, without including the halogen.
H
dx.doi.org/10.1021/ic501424e | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
N2 atmosphere using standard Schlenk techniques or in a glovebox
maintained at or below 1.0 ppm of O2, except crystallization of the
cobalt(II) and nickel(II) complexes.
Although the DFT calculations do show a small percentage of
chlorine character mixed into the metal d-orbitals, calculations
give a bond order that is only 15−38% of the values for the
M−O and M−N bonds. From VT-NMR, the experimentally
determined value for the phenolate barrier to rotation was 30.3
kJ/mol in 3 and 37 kJ/mol in 5 (with the caveat noted above).
The DFT calculations with PBE1PBE gave a barrier to rotation
of the phenolate in Tp*Co(2,6-dcp) of just 13.4 kJ/mol, and
the barrier in the nickel(II) complex was comparable. The
evidence herein suggests that the metal−halogen interaction in
these compounds is primarily electrostatic in nature and that
there is no significant covalent interaction between the metal
and halogen.
For comparison, another type of noncovalent interaction
involving polarizable halogens (especially attached to arenes) is
halogen bonding, where a halogen interacts with the partial
negative charge on an oxygen or nitrogen atom42,65,66 or with
the π electrons from an aromatic ring.67,68 The strength of this
interaction has been estimated as 10−30 kJ/mol for haloarenes,
which is similar to that for a hydrogen bond between neutral
species.69 Assuming that the barrier to rotation of the phenolate
is a reasonable measure of the strength of the metal−halogen
secondary bond, it puts the strength of the metal−halogen
secondary interaction in the same range as halogen bonding.
Note that the directionality of these two types of noncovalent
interactions is very different, due to the strongly anisotropic
charge distribution of polarizable halogens, in which electron
density is pulled away from the crown of the halogen and
drawn toward the equator.42 Thus, while the ideal C-X···O/N
angle in halogen bonding is ∼180°, the ideal C-X···M angle in
metal−halogen secondary bonding is ∼90°, as observed in the
present complexes.
These results further illuminate to what extent metal−
halogen bonding could be responsible for the unusual
specificity of the enzyme PcpA for ortho-dihalophenols and
hydroquinones. Since this interaction gives a stabilization that is
on par with other noncovalent interactions that govern binding
of substrates and inhibitors to the active site pocket, metal−
halogen secondary bonding is but one of several factors that
must be considered in governing the substrate specificity of this
enzyme, and further work is required to elucidate its potential
role. Nonetheless, if one makes the rough assumption that in
PcpA the KmA is primarily determined by substrate binding
(∼KD), then ΔΔG for binding of dihalohydroquinone versus
dimethylhydroquinone is 7−9 kJ/mol, and thus secondary
bonding could certainly be a large contributor to this difference.
Metal−halogen secondary bonding could also play an
important role in other ways; for example, it could help to
orient the ortho-dihalohydroquinone substrates in a geometry
that is optimal for attack by O2, thus leading to the higher kcat
that is observed with ortho-dihalohydroquinone substrates
relative to the ortho-dimethyl version.
Spectra and Data Analysis. NMR spectra were collected on a
Bruker Avance III 400 MHz spectrometer with a broadband probe.
Variable temperature NMR spectra were acquired with an inverse
broadband probe, and a liquid nitrogen heat exchanger was used to
control the temperature. The temperature sensor in the probe was
calibrated with neat methanol.70 All peaks in NMR spectra of the
complexes appeared as singlets and are reported in ppm. The
temperature dependences of the linewidths in the variable temperature
NMR spectra were fit in Excel (Microsoft) using Solver. The
uncertainty in ΔG‡ and Δδ was estimated with a jackknife
procedure.71 FT-IR measurements were carried out on KBr pellets
and acquired on an PerkinElmer Avatar 360 or a Shimadzu Prestige 21
spectrometer. UV−visible absorption spectra were acquired on a Cary
50 Bio UV−visible spectrophotometer, and NIR absorption spectra
were acquired on a StellarNet InGaAs spectrophotometer. Solution
spectra were acquired in quartz cuvettes of 1 cm optical path length,
while mull spectra were acquired in quartz cuvettes of 0.2 mm path
length. Elemental analyses were determined by the CENTC Elemental
Analysis Facility at the University of Rochester. X-ray crystallography
was performed at the X-ray Crystallographic Facility of the
Department of Chemistry at the University of Rochester.
TpPh2MX. TpPh2CoCl and TpPh2NiBr were prepared according to
literature methods.37,38 TpPh2FeBr was prepared in a similar manner to
TpiPr2FeCl and characterized by 1H NMR spectroscopy (Figure S5).32
Sodium 2,6-Dibromophenolate. To a solution of 2,6-
dibromophenol (120.5 mg, 0.4784 mmol) in THF (4 mL), excess
sodium hydride was added slowly until bubbling ceased. The solution
was filtered through Celite. Solvent was removed under vacuum and a
white powder was collected. Crystallization of the solid from THF (3
1
mL) gave clear, colorless crystals (0.0968 g, 73%). H NMR (400
MHz, 25 °C, CD3CN): δ 7.29 (d, 2, meta-H, J = 7.6 Hz), 6.03 (t, 1,
para-H, J = 7.8 Hz) ppm.
Sodium 2,6-Dichlorophenolate. This was prepared by the
method above using 2,6-dichlorophenol (146.7 mg, 0.9000 mmol).
Crystallization yielded clear, colorless crystals (45.1 mg, 27% yield).
1H NMR (400 MHz, 25 °C, CD3CN): δ 7.12 (d, 2, meta-H, J = 8.0
Hz), 6.23 (t, 1, para-H, J = 7.8 Hz) ppm.
TpPh2Fe(2,6-Dichlorophenolate) (1). TpPh2FeBr (56.3 mg,
0.0699 mmol) was dissolved in THF (15 mL). The solution was
transferred to solid sodium 2,6-dichlorophenolate (14.8 mg, 0.0800
mmol), and was stirred overnight. The solvent was removed under
reduced pressure, and the product was suspended in acetonitrile (5
mL). The suspension was filtered through a Celite plug and washed
with additional THF (5 mL). The THF was then removed, leaving an
off-white powder. The product was redissolved in THF (2 mL) and
crystallized via vapor diffusion with pentane to give colorless crystals
(44.1 mg, 71% yield).72 1H NMR (400 MHz, 25 °C, THF-d8): δ 57.9
(3H, pz-H4), 54.2 (2H, m-phenolate), 39.2 (1H, BH), 16.4 (6H, o/m-
Ph), 12.8 (6H, o/m-Ph), 10.0 (3H, p-Ph), 7.0 (∼10H, o/m/p-Ph),
−21.8 (∼4H, o/m/p-Ph), −23.7 (1H, p-phenolate) ppm. IR (KBr):
3061 (w), 2634 (w), 1575 (s), 1544 (s), 1479 (m), 1456 (w), 1415
(m), 1301 (w), 1168 (w), 1066 (w), 759 (w), 696 (s) cm−1. Elem.
Anal.: Calculated for C51H37BCl2FeN6O•THF: C, 68.84; H, 4.73; N,
8.76. Found: C, 69.02; H, 4.85; N, 8.79.
TpPh2Fe(2,6-Dibromophenolate) (2). This was prepared as for 1,
but using TpPh2FeBr (50.5 mg, 0.0627 mmol) and sodium 2,6-
dibromophenolate (19.0 mg, 0.0694 mmol). Crystallization via vapor
diffusion with pentane to give colorless crystals (46.8 mg, 76% yield).
1H NMR (400 MHz, 25 °C, THF-d8): δ 58.4 (3H, pz-H4), 54.3 (2H,
m-phenolate), 35.7 (1H, BH), 15.5 (6H, o/m-Ph), 12.4 (6H, o/m-Ph),
9.6 (3H, p-Ph), 6.7 (∼9H, o/m/p-Ph), −19.22 (∼4H, o/m/p-Ph),
−26.18 (1H, p-phenolate) ppm. IR (KBr): 3061 (w), 2625 (w), 1564
(s), 1544 (s), 1479 (s), 1452 (m), 1415 (s), 1305 (m), 1168 (w), 1066
(w), 763 (m), 696 (s) cm−1. UV−vis−NIR (in THF) λmax (ε/M−1
EXPERIMENTAL SECTION
■
General Considerations. All reagents were reagent grade and
used without purification except where indicated. For the reactions
with iron(II), the phenols were purified by sublimation before use, and
THF was dried by distillation from sodium and benzophenone. For
reactions with cobalt(II) and nickel(II), reagent grade anhydrous THF
was used without further purification and CH2Cl2 was dried over
molecular sieves. THF-d8 was dried over CaH2, then over sodium
metal, and then vacuum distilled into a storage container before use.
The tris(3,5-diphenylpyrazolyl)borate ligand, KTpPh2, was prepared by
literature procedures.31 All manipulations were carried out under an
cm− 1
) 360 (sh) nm. Elem. Anal.: Calculated for
C51H37BBr2FeN6O•THF: C, 63.01; H, 4.33; N, 8.02. Found: C,
62.76; H, 4.26; N, 7.95.
I
dx.doi.org/10.1021/ic501424e | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 6. Crystal and Data Parameters for Structures 1−6
1
2
3
4
5
6
empirical formula
formula weight
temperature (K)
wavelength (Å)
crystal system
space group
a (Å)
C51H37BCl2FeN6O
887.43
C53H41BBr2FeN6O1.5
1012.40
C51H37BCl2CoN6O
890.51
C51H37BBr2CoN6O
979.43
C51H37BCl2NiN6O
890.29
C51H37Br2NiN6O
979.21
100.0(5)
0.71073
100.0(5)
100.0(5)
0.71073
100.0(5)
0.71073
100.0(5)
0.71073
100.0(5)
0.71073
0.71073
triclinic
triclinic
monoclinic
P21/n
monoclinic
P21/n
triclinic
triclinic
P-1
P-1
P-1
P-1
12.305(2)
14.638(3)
15.082(3)
64.193(3)
84.138(4)
71.675(4)
2319.8(7)
2
12.4836(11)
12.8222(11)
15.9833(14)
89.687(2)
79.109(2)
62.179(2)
2211.8(3)
2
10.2718(15)
30.796(4)
13.4527(19)
90
10.2332(11)
30.514(3)
13.6761(15)
90
12.3898(11)
14.2664(13)
15.4042(13)
63.969(2)
84.152(2)
71.450(2)
2317.2(4)
2
12.4504(15)
12.8726(16)
15.8829(19)
90.729(2)
100.184(2)
117.451(2)°
2210.4(5)
2
b (Å)
c (Å)
α (deg)
β (deg)
91.876(3)
90
91.362(2)
90
γ (deg)
volume (Å3)
4253.2(10)
4
4269.2(8)
4
Z
crystal color,
morphology
colorless, block
colorless, needle
blue-violet, plate
blue-violet, plate
yellow-orange, block yellow-orange, plate
crystal size (mm)
0.24 × 0.20 × 0.08
0.22 × 0.18 × 0.12
0.24 × 0.18 × 0.06
0.22 × 0.18 × 0.05
0.20 × 0.20 × 0.16
0.18 × 0.14 × 0.05
data/restraints/
parameters
GOF on F2
18368/0/559
16767/5/592
8696/0/559
7550/0/559
17565/0/559
7809/0/559
0.873
0.998
1.086
1.035
0.909
0.870
final R indices [I > R1 = 0.0563, wR2 = R1 = 0.0535, wR2 =
2σ(I)] 0.1170 0.1021
R1 = 0.0662, wR2 = R1 = 0.0594, wR2 = R1 = 0.0582, wR2 = R1 = 0.0634, wR2 =
0.1163 0.1015 0.1138 0.1057
R indices (all data) R1 = 0.1107, wR2 = R1 = 0.1159, wR2 =
0.1356 0.1224
R1 = 0.1150, wR2 = R1 = 0.1189, wR2 = R1 = 0.0970, wR2 = R1 = 0.1447, wR2 =
0.1322 0.1181 0.1292 0.1276
TpPh2Co(2,6-Dichlorophenolate) (3). A solution of TpPh2CoCl
(15 mg, 0.020 mmol) in THF (4 mL) was added to sodium 2,6-
dichlorophenolate (80 mg, 0.043 mmol). Upon addition, the solution
changed from aqua blue to purple. The reaction was stirred for 3 h, the
solution was filtered, and solvent was removed under reduced
pressure. Crystallization from CH2Cl2 (2 mL) layered with hexanes
(8 mL) yielded purple crystals (12 mg, 74%). 1H NMR (400 MHz, 25
°C, CDCl3): δ 68.1 (3H, pz-H4), 35.2 (2H, m-phenolate), 15.1 (6H,
o-Ph), 9.08 (6H, o/m-Ph), 8.85 (6H, o/m-Ph), 8.67 (3H, p-Ph), 8.196
(6H, o/m-Ph), 7.45 (3H, p-Ph), −34.1 (1H, p-phenolate) ppm. IR
(KBr): 3123 (w), 2626 (w), 1546 (m), 1475 (s), 1169 (s), 760 (s),
695 (s) cm−1. UV−vis−NIR (in CHCl3) λmax (ε/M−1 cm−1) 373
(1190), 572 (283), 655 (620), 811 (19), 1469 (81) nm. Elem. Anal.
Calculated for C51H37BCl2CoN6O: C, 68.78; H, 4.19; N, 9.44. Found:
C, 68.31; H, 4.18; N, 9.38.
TpPh2Ni(2,6-Dibromophenolate) (6). This was prepared as for 3,
but using TpPh2NiBr (13 mg, 0.016 mmol) and sodium 2,6-
dibromophenolate (5.1 mg, 0.019 mmol). Crystallization from
CH2Cl2 (2 mL) layered with hexanes (5 mL) yielded orange crystals
1
(15 mg, 71%). H NMR (400 MHz, 25 °C, CDCl3): δ 71.3 (3H pz-
H4), 28.0 (2H, m-phenolate), 10.4 (6H, o-Ph), 8.38 (3H, p-Ph), 6.78
(6H, o/m-Ph), 6.67 (6H, o/m-Ph), −16.1 (1H, BH), −25.3 (1H, p-
phenolate) ppm. IR (KBr): 3131 (m), 2619 (w), 1546 (w), 1455 (s),
1173 (s), 760 (s), 696 (s) cm−1. UV−vis−NIR (in CHCl3) λmax (ε/
M−1 cm−1) 447 (1790), 789 (55), 880 (63), ∼1500 (46) nm. Elem.
Anal. Calculated for C51H37BBr2NiN6O: C, 62.56; H, 3.81; N, 8.58.
Found: C, 62.29; H, 3.85; N, 8.50.
X-ray Crystallography. Each crystal was placed onto the tip of a
glass fiber and mounted on a Bruker SMART Platform diffractometer
equipped with an APEX II CCD area detector. All data were collected
at 100.0(5) K using Mo Kα radiation (graphite monochromator). For
each sample a preliminary set of cell constants and an orientation
matrix were determined from reflections harvested from three
orthogonal wedges of reciprocal space. Full data collections were
carried out with frame exposure times of 45−120 s at detector
distances of 4 cm. Randomly oriented regions of reciprocal space were
surveyed for each sample: four or five major sections of frames were
collected with 0.50° steps in ω at different φ settings and detector
positions of −38° in 2θ. The intensity data were corrected for
absorption,73 and final cell constants were calculated from the xyz
centroids of 3790−4087 (1804 in the case of 6) strong reflections
from the actual data collection after integration.74 Structures were
solved using SIR9775 and refined using SHELXL-97.76 Direct-methods
solutions were calculated which provided most non-hydrogen atoms
from the difference Fourier map. Least squares (on F2) difference
Fourier cycles located the remaining non-hydrogen atoms. Non-
hydrogen atoms were refined with anisotropic displacement
parameters and hydrogen atoms were placed in ideal positions and
refined as riding atoms with relative isotropic displacement parameters.
Details of each structure are given in Table 6.
TpPh2Co(2,6-Dibromophenolate) (4). This was prepared as for 3,
but using TpPh2CoCl (12 mg, 0.016 mmol) and sodium 2,6-
dibromophenolate (6.0 mg, 0.022 mmol). Crystallization from
CH2Cl2 (2 mL) layered with hexanes (10 mL) yielded purple crystals
1
(13 mg, 80%). H NMR (400 MHz, 25 °C, CDCl3): δ 67.4 (3H, pz-
H4), 37.3 (2H, m-phenolate), 10.5 (6H, o-Ph), 9.90 (6H, o/m-Ph),
8.92 (6H, o/m-Ph), 8.30 (3H, p-Ph), 7.78 (9H, o/m/p-Ph), −34.7
(1H, p-phenolate) ppm. IR (KBr): 3129 (m), 2620 (w), 1451 (s),
1168 (s), 760 (s), 697 (s) cm−1. UV−vis−NIR (in CHCl3) λmax (ε/
M
−1 cm−1) 372 (1230), 570 (273), 660 (539), ∼800 (∼20), 1475 (71)
nm. Elem. Anal. Calculated for C51H37BBr2CoN6O: C, 62.54; H, 3.81;
N, 8.58. Found: C, 62.88; H, 3.90; N, 8.53.
TpPh2Ni(2,6-Dichlorophenolate) (5). This was prepared as for 3,
but using TpPh2NiBr (19 mg, 0.024 mmol) and sodium 2,6-
dichlorophenolate (5.4 mg, 0.029 mmol). Crystallization from
CH2Cl2 (3 mL) layered with hexanes (5 mL) yielded orange crystals
1
(12 mg, 58%). H NMR (400 MHz, 25 °C, CDCl3): δ 72.4 (3H, pz-
H4), 29.0 (2H, m-phenolate), 11.5 (6H, o-Ph), 8.52 (3H, p-Ph), 7.20
(9H, o/m/p-Ph), 6.73 (6H, o/m-Ph), 6.63 (6H, o/m-Ph), −16.8 (1H,
BH), −25.2 (1H, p-phenolate) ppm. IR (KBr): 3060 (w), 2640 (w),
1653 (m), 1459 (s), 1174 (s), 1012 (s), 694 (s) cm−1. UV−vis−NIR
(in CHCl3) λmax (ε/M−1 cm−1) 443 (2300), 788 (80), 877 (91), 1470
(53) nm. Elem. Anal. Calculated for C51H37BCl2NiN6O• 1/2CH2Cl2:
C, 66.31; H, 4.11; N, 9.01. Found: C, 66.20; H, 4.56; N, 8.61.
Structures 1, 5, and 6 (1 and 5 are isomorphous) were found to
have cocrystallized solvent that was highly disordered and unable to be
modeled satisfactorily. Reflection contributions from the solvent were
removed using the SQUEEZE routine of program Platon,77 which
determined there to be 87, 115, and 59 electrons in 353, 375, and 169
J
dx.doi.org/10.1021/ic501424e | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Å3 removed, respectively, per unit cell. Because the exact identity and
amount of solvent were not known, no solvent was included in the
atom list nor the molecular formulas for these structures; thus all
calculated quantities that derive from the molecular formulas (i.e.,
F(000), density, molecular weight, etc.) are known to be incorrect.
The asymmetric unit of structure 2 contained the iron molecule in a
general position and one cocrystallized THF solvent molecule in a
crystallographic inversion center, the latter which was modeled as
disordered over the center. Structures 3 and 4 are also isomorphous,
but without cocrystallized solvent.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This research was supported by grants from the National
Science Foundation (CHE-0951999 to T.E.M. and CHE-
0911314 to P.L.H.). The purchase of the 400 MHz Bruker
Avance III NMR spectrometer was supported by a grant from
the National Science Foundation (MRI-0922775). Analytical
data from the University of Rochester used the CENTC
Elemental Analysis Facility, funded by NSF CHE-0650456.
Calculations. AOM calculations were performed using the AOMX
program.45 DFT calculations were performed using the Gaussian 0949
package installed on an AMD Opteron G34 12 core server. Two
different unrestricted hybrid density functionals, B3LYP (Becke, 3-
parameter, Lee−Yang−Parr)50 and PBE1PBE (Perdew, Burke, and
Ernzerhof with 25% exchange weighting),51 were employed for all
calculations. A third hybrid density functional M062X,63 was explored
initially; however, it gave unreasonable results for the electronic
structure of the metal center and was not pursued further. In all
models, the phenyl groups of TpPh2 were replaced with methyl groups
(Tp*), since these substituents do not affect the overall geometry and
hence the metal−halogen interactions. Initial coordinates were
obtained from the crystal structures, and the phenyl groups on Tp
ligands were modified to methyl groups using DS Viewer Pro
(Accelrys). Two-step calculations were conducted to obtain the lowest
energy optimized structures. First, single point energies (SPE) were
calculated using the valence triple-ζ basis set (6-311G(d))78 for all
atoms. Second, the wave functions generated from SPE calculations
were used as a starting point for the geometry optimizations. All
models were optimized with added polarization and diffuse functions:
6-311+G(d) for all atoms, and 6-311++G(d) for halogens to obtain
improved electronic structures. Tight convergence criteria were used
during all optimizations. To estimate the metal−halogen interactions
in complexes Tp*Co(2,6-dcp) and Tp*Ni(2,6-dcp), the M−O1−
C46phen−C47phen dihedral angle in the optimized structure was rotated
by 15° increments up to 90°, and these modified models were
reoptimized by fixing the dihedral angles at that particular value. Khon-
Sham orbitals were analyzed using Gaussview 5 containing Gaussian
utilities programs. Mulliken orbital contributions52−55 from the
PBE1PBE calculations with a smaller basis set (6-311G(d)) were
analyzed with AOMix.56,57 Mayer bond orders58−61 and Wiberg and
REFERENCES
■
(1) Reregistration Eligibility Decision for Pentachlorophenol; U.S.
Environmental Protection Agency, Office of Pesticide Programs:
Arlington, VA, 2008; EPA 739-R-008.
(2) Copley, S. D. Trends Biochem. Sci. 2000, 25, 261−265.
(3) Ohtsubo, Y.; Miyauchi, K.; Kanda, K.; Hatta, T.; Kiyohara, H.;
Senda, T.; Nagata, Y.; Mitsui, Y.; Takagi, M. FEBS Lett. 1999, 459,
395−398.
(4) Xu, L.; Resing, K.; Lawson, S. L.; Babbitt, P. C.; Copley, S. D.
Biochemistry 1999, 38, 7659−7669.
(5) Xun, L.; Bohuslavek, J.; Cai, M. Biochem. Biophys. Res. Commun.
1999, 266, 322−325.
(6) Machonkin, T. E.; Holland, P. L.; Smith, K. N.; Liberman, J. S.;
Dinescu, A.; Cundari, T. R.; Rocks, S. S. J. Biol. Inorg. Chem. 2010, 15,
291−301.
(7) Lipscomb, J. D. Curr. Opin. Struct. Biol. 2008, 18, 644−649.
(8) Vaillancourt, F. H.; Bolin, J. T.; Eltis, L. D. Crit. Rev. Biochem.
Mol. Biol. 2006, 41, 241−267.
(9) Hayes, R. P.; Green, A. R.; Nissen, M. S.; Lewis, K. M.; Xun, L.;
Kang, C. Mol. Microbiol. 2013, 88, 523−536.
(10) Bartels, I.; Knackmuss, H.-J.; Reineke, W. Appl. Environ.
Microbiol. 1984, 47, 500−505.
(11) Cerdan, P.; Wasserfallen, A.; Rekik, M.; Timmis, K. N.;
Harayama, S. J. Bacteriol. 1994, 176, 6074−6081.
(12) Dai, S.; Vaillancourt, F. H.; Maaroufi, H.; Drouin, N. M.; Neau,
D. B.; Snieckus, V.; Bolin, J. T.; Eltis, L. D. Nat. Struct. Biol. 2002, 9,
934−939.
Lowdin bond indices62 were determined with AOMix from the
̈
Gaussian calculations with the 6-311G basis set. Larger basis sets were
found to give unphysical values with all three bond order methods. To
examine the phenolate-O to metal charge-transfer transitions, time
dependent DFT (TD-DFT) calculations79 were performed with
B3LYP using the geometries obtained from the PBE1PBE calculations.
60 excited states were calculated in each case. Very similar results were
also obtained using the B3LYP geometry-optimized structures.
́
(13) Vaillancourt, F. H.; Labbe, G.; Drouin, N. M.; Fortin, P. D.;
Eltis, L. D. J. Biol. Chem. 2002, 277, 2019−2027.
(14) Machonkin, T. E.; Doerner, A. E. Biochemistry 2011, 50, 8899−
8913.
(15) Rocks, S. S.; Brennessel, W. W.; Machonkin, T. E.; Holland, P.
L. Inorg. Chem. 2010, 49, 10914−10929.
(16) Kuz’mina, L. G. Russ. J. Coord. Chem. 1999, 25, 599−617.
(17) Meyer, R.; Gagliardi, J.; Wulfsberg, G. J. Mol. Struct. 1983, 111,
311−316.
(18) Wulfsberg, G.; Yanisch, J.; Meyer, R.; Bowers, J.; Essig, M. Inorg.
Chem. 1984, 23, 715−719.
(19) Richardson, M. F.; Wulfsberg, G.; Marlow, R.; Zaghonni, S.;
McCorkle, D.; Shadid, K.; Gagliardi, J., Jr.; Farris, B. Inorg. Chem.
1993, 32, 1913−19.
(20) Wulfsberg, G. P., Nuclear Quadrupole Resonance (NQR)
Spectroscopy. In Applications of Physical Methods to Inorganic and
Bioinorganic Chemistry, Scott, R. A., Lukehart, C. M., Eds.; John Wiley
& Sons: Chichester, U.K., 2007; pp 385−399.
(21) Orpen, A. G.; Brammer, L.; Allen, F. H.; Kennard, O.; Watson,
D. G.; Taylor, R. J. Chem. Soc., Dalton Trans. 1989, S1−S83.
(22) Cingi, M. B.; Lanfredi, A. M. M.; Tiripicchio, A.; Reedijk, J.; van
Landschoot, R. Inorg. Chim. Acta 1980, 39, 181−186.
(23) Simonov, Y. A.; Dvorkin, A. A.; Matuzenko, G. S.; Yampol’skaya,
M. A.; Gifeisman, T. S.; Gerbeleu, N. V.; Malinovskii, T. I. Russ. J.
Coord. Chem. 1984, 10, 1247−52.
(24) Simonov, Y. A.; Matuzenko, G. S.; Botoshanskii, M. M.;
Yampol’skaya, M. A.; Gerbeleu, N. V.; Malinovskii, T. I. Russ. J. Inorg.
Chem. 1982, 27, 231−234.
ASSOCIATED CONTENT
■
S
* Supporting Information
Crystallographic data are given in CIF format. Selected
geometric data, Wiberg and Lowdin bond orders, optimized
̈
energies, and optimized coordinates for the Tp*Co(2,6-dcp)
and Tp*Ni(2,6-dcp) models from DFT calculations are
provided in Tables S1−S10. Other experimental and computa-
tional results are given in Figures S1−S5. This material is
available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Authors
*E-mail: machonte@whitman.edu.
*E-mail: patrick.holland@yale.edu.
Present Address
§Department of Chemistry, Yale University, New Haven, CT
06520.
K
dx.doi.org/10.1021/ic501424e | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(25) Basturkmen, M.; Kisakurek, D.; Driessen, W.; Gorter, S.;
Reedijk, J. Inorg. Chim. Acta 1998, 271, 19−23.
(53) Mulliken, R. S. J. Chem. Phys. 1955, 23, 1841−1846.
(54) Mulliken, R. S. J. Chem. Phys. 1955, 23, 2338−2342.
(55) Mulliken, R. S. J. Chem. Phys. 1955, 23, 2343−2346.
(26) Camurlu, P.; Yilmaz, A.; Tatar, L.; Kisakurek, D.; Ulku, D. Cryst.
Res. Technol. 2005, 40, 271−276.
(56) Gorelsky, S. I. AOMix, 6.81; University of Ottawa: Ottawa,
Ontario, 2013.
(27) Ulku, D.; Tahir, M. N.; Kesici, N.; Isci, H.; Kisakurek, D. Z.
Kristallogr.-New Cryst. Struct. 1997, 212, 493−494.
(28) Trofimenko, S. Chem. Rev. 1993, 93, 943−980.
(29) Kitajima, N.; Tolman, W. B. Prog. Inorg. Chem. 1995, 43, 419−
531.
(30) Pettinari, C.; Santini, C., Polypyrazolylborate and Scorpionate
Ligands. In Comprehensive Coordination Chemistry II; McCleverty, J. A.,
Meyer, T. J., Eds.; Elsevier Pergamon: Amsterdam, 2004; Vol. 1, pp
159−210.
(57) Gorelsky, S. I.; Lever, A. B. P. J. Organomet. Chem. 2001, 635,
187−196.
(58) Mayer, I. Chem. Phys. Lett. 1983, 97, 270−274.
(59) Mayer, I. Theor. Chim. Acta 1985, 67, 315−322.
(60) Mayer, I. Int. J. Quantum Chem. 1986, 29, 73−84.
(61) Mayer, I. Int. J. Quantum Chem. 1986, 29, 477−483.
(62) Natiello, M. A.; Medrano, J. A. Chem. Phys. Lett. 1984, 105,
180−182.
(63) Zhao, Y.; Truhlar, D. G. J. Chem. Theory Comput. 2007, 3, 289−
300.
(64) CRC Handbook of Chemistry and Physics, 90th ed.; CRC Press:
(31) Kitajima, N.; Fujisawa, K.; Fujimoto, C.; Moro-oka, Y.;
Hashimoto, S.; Kitagawa, T.; Toriumi, K.; Tatsumi, K.; Nakamura,
A. J. Am. Chem. Soc. 1992, 114, 1277−1291.
Boca Raton, FL, 2009.
(32) Ito, M.; Amagai, H.; Fukui, H.; Kitajima, N.; Moro-Oka, Y. Bull.
Chem. Soc. Jpn. 1996, 69, 1937−1945.
(65) Lommerse, J. P. M.; Stone, A. J.; Taylor, R.; Allen, F. H. J. Am.
Chem. Soc. 1996, 118, 3108−3116.
(33) Deb, T.; Rohde, G. T.; Young, V. G.; Jensen, M. P. Inorg. Chem.
2012, 51, 7257−7270.
(66) Metrangolo, P.; Meyer, F.; Pilati, T.; Resnati, G.; Terraneo, G.
Chem. Commun. 2008, 1635−1637.
(34) Baum, A. E.; Park, H.; Wang, D. N.; Lindeman, S. V.; Fiedler, A.
T. Dalton Trans. 2012, 41, 12244−12253.
(67) Imai, Y. N.; Inoue, Y.; Nakanishi, I.; Kitaura, K. Protein Sci. 2008,
17, 1129−1137.
(35) Bittner, M. M.; Kraus, D.; Lindeman, S. V.; Popescu, C. V.;
Fiedler, A. T. Chem.Eur. J. 2013, 19, 9686−9698.
(36) Bittner, M. M.; Lindeman, S. V.; Fiedler, A. T. J. Am. Chem. Soc.
2012, 134, 5460−5463.
(68) Matter, H.; Nazare, M.; Gussregen, S.; Will, D. W.; Schreuder,
H.; Bauer, A.; Urmann, M.; Ritter, K.; Wagner, M.; Wehner, V. Angew.
Chem., Int. Ed. Engl. 2009, 48, 2911−2916.
(69) Perrin, C. L.; Nielson, J. B. Annu. Rev. Phys. Chem. 1997, 48,
511−544.
(70) Ammann, C.; Meier, P.; Merbach, A. E. J. Magn. Reson. 1982, 46,
319−321.
(37) Harding, D. J.; Harding, P.; Adams, H.; Tuntulani, T. Inorg.
Chim. Acta 2007, 360, 3335−3340.
(38) Harding, D. J.; Harding, P.; Daengngern, R.; Yimklan, S.;
Adams, H. Dalton Trans. 2009, 1314−1320.
(71) Harris, D. C. J. Chem. Educ. 1998, 75, 119−121.
(72) Since complexes 1 and 2 were both colorless, and the UV−vis-
NIR spectrum of 2 showed only a shoulder at ∼360 nm (see below),
the UV−vis-NIR spectrum of 1 was not obtained.
(39) Addison, A. W.; Rao, T. N.; Reedijk, J.; Vanrijn, J.; Verschoor,
G. C. J. Chem. Soc., Dalton Trans. 1984, 1349−1356.
(40) Bondi, A. J. Phys. Chem. 1964, 68, 441−451.
(41) Mantina, M.; Chamberlin, A. C.; Valero, R.; Cramer, C. J.;
Truhlar, D. G. J. Phys. Chem. A 2009, 113, 5806−5812.
(42) Auffinger, P.; Hays, F. A.; Westhof, E.; Ho, P. S. Proc. Natl. Acad.
Sci. U.S.A. 2004, 101, 16789−16794.
(73) Sheldrick, G. M. SADABS, 2008/1; University of Gottingen:
̈
Gottingen, Germany, 2008.
̈
(74) SAINT, 7.60A; Bruker AXS: Madison, WI, 2008.
(75) Altomare, A.; Burla, M.; Cascarano, G. L.; Giacovazzo, C.;
Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.; Spagna, R. SIR97;
Istituto di Cristallografia: CNR Bari, Italy, 1999.
(43) Bertini, I.; Luchinat, C.; Parigi, G. Solution NMR of Paramagnetic
Molecules; Elsevier: Amsterdam, 2001.
(44) Gorelsky, S. I.; Basumallick, L.; Vura-Weis, J.; Sarangi, R.;
Hodgson, K. O.; Hedman, B.; Fujisawa, K.; Solomon, E. I. Inorg. Chem.
2005, 44, 4947−4960.
(76) Sheldrick, G. M. Acta Crystallogr. A 2008, 64, 112−122.
(77) Spek, A. L. Acta Crystallogr. D 2009, 65, 148−155.
(78) Schafer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97,
̈
(45) Adamsky, H. AOMX; Institut fur Theoretische Chemie,
̈
2571−2577.
Heinrich-Heine-Universitat: Dusseldorf, Germany, 1996.
̈
̈
(79) Stratmann, R. E.; Scuseria, G. E.; Frisch, M. J. J. Chem. Phys.
1998, 109, 8218−8224.
(46) Krzystek, J.; Swenson, D. C.; Zvyagin, S. A.; Smirnov, D.;
Ozarowski, A.; Telser, J. J. Am. Chem. Soc. 2010, 132, 5241−5253.
(47) Holm, R. H.; Kennepohl, P.; Solomon, E. I. Chem. Rev. 1996,
96, 2239−2314.
(48) Desrochers, P. J.; Telser, J.; Zvyagin, S. A.; Ozarowski, A.;
Krzystek, J.; Vicic, D. A. Inorg. Chem. 2006, 45, 8930−8941.
(49) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.
P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; J. A. Montgomery, J.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega,
N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.;
Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.;
̈
Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.;
Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09,
revision C 01; Gaussian, Inc.: Wallingford, CT, 2010.
(50) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652.
(51) Adamo, C.; Barone, V. J. Chem. Phys. 1999, 110, 6158−6170.
(52) Mulliken, R. S. J. Chem. Phys. 1955, 23, 1833−1840.
L
dx.doi.org/10.1021/ic501424e | Inorg. Chem. XXXX, XXX, XXX−XXX