Reactivity and electronic properties of a ferrocene molecule bearing an N,C-chelated BMes2 unit

Article abstract of DOI:10.1021/om500138f

A dimesitylboron-functionalized benzimidazolylferrocene, B(2-ferrocenyl-N-Me-benzimidazolyl)Mes₂ (1), has been synthesized and fully characterized. The B-N bond in 1 was found to undergo a dynamic dissociation/association process in solution, leading to a dynamic exchange of the two mesityls bound to the boron atom and slow hydrolysis of 1 under ambient conditions. The hydrolyzed product 2-BMes(OH)-1-(N-methylbenzimidazol-2-yl) ferrocene (2) was isolated and characterized, in which the boron center has a trigonal-planar geometry with one of the mesityls being replaced by an OH ^- group. The photoisomerization process of the boron unit in 1 was fully inhibited by the low lying d-d/MLCT states of the ferrocene unit. Compound 1 can be oxidized readily by I₂, forming the ferrocenium species [1⁺]I₃^- (3). NMR and EPR data for 3 indicated a notable spin delocalization through space from the Fe(III) center to a flanking mesityl group.

Full text of DOI:10.1021/om500138f

Article
pubs.acs.org/Organometallics
Reactivity and Electronic Properties of a Ferrocene Molecule Bearing
an N,C-Chelated BMes2 Unit
Ying-Li Rao,† Tetsuro Kusamoto,‡ Ryota Sakamoto,‡ Hiroshi Nishihara,‡ and Suning Wang*,†
†Department of Chemistry, Queen’s University, Kingston, Ontario K7M 3N6, Canada
‡Department of Chemistry, School of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
S
* Supporting Information
ABSTRACT: A dimesitylboron-functionalized benzimidazolylferro-
cene, B(2-ferrocenyl-N-Me-benzimidazolyl)Mes2 (1), has been
synthesized and fully characterized. The B−N bond in 1 was found
to undergo a dynamic dissociation/association process in solution,
leading to a dynamic exchange of the two mesityls bound to the
boron atom and slow hydrolysis of 1 under ambient conditions. The
hydrolyzed product 2-BMes(OH)-1-(N-methylbenzimidazol-2-yl)-
ferrocene (2) was isolated and characterized, in which the boron
center has a trigonal-planar geometry with one of the mesityls being
replaced by an OH− group. The photoisomerization process of the
boron unit in 1 was fully inhibited by the low lying d−d/MLCT
states of the ferrocene unit. Compound 1 can be oxidized readily by
‑
I2, forming the ferrocenium species [1+]I3 (3). NMR and EPR data for 3 indicated a notable spin delocalization through space
from the Fe(III) center to a flanking mesityl group.
To examine the mutual influence of a ferrocene unit and a
N,C-chelated BMes2 unit on the electronic, chemical, and
photochemical properties, we designed and synthesized the
new molecule B(2-ferrocenyl-N-Me-benzimidazolyl)Mes2 (1),
in which a four-coordinate boryl moiety is connected to the
ferrocene. Compound 1 was found to display an unusual
dynamic exchange and ring-opening hydrolysis. Furthermore, 1
can be oxidized readily by I2, generating the new ferrocenium
species [1+]. Details of our investigation on the reactivity and
the electronic properties of 1 and [1+] are presented herein.
INTRODUCTION
■
Organoboron compounds are an emerging class of photo-
chromic materials.1 In particular, N,C-chelated dimesitylbor-
anes exhibit tunable colors and rich photochemical and
chemical reactivity.1c−g The charge transfer transition from
the mesityl to the chelated backbone is a key driving force for
the photoisomerization of N,C-chelated BMes2 compounds. To
further tune the photochromic properties of N,C-chelated
BMes2 compounds, we incorporated transition-metal ions such
as Au(I), Pt(II), and Re(I) to the chelated backbone, because
this strategy was used successfully in tuning other photo-
chromic systems such as dithienylethenes (DTE).2 However,
we observed that, in most instances, the introduction of a metal
ion unit greatly inhibited the photoreactivity of the boron unit
because of the low-lying metal-to-ligand charge transfer
(MLCT) state.3 On the other hand, redox-active transition
metal ions have been extensively explored as a control to
modulate photochromic systems by influencing the electronic
structure of the photoswitch through the oxidation states of the
metal ion.4 One of the most commonly used metal units for
regulating photoswitching systems is ferrocene.5 In addition to
influencing the properties of photochromic systems, the
ferrocene unit was also incorporated into a variety of main-
group systems to create new materials that possess interesting
electronic, chemical, and photophysical properties and
applications.6,7 In particular, three-coordinate boryl units were
found to be very effective in promoting electronic communi-
cation with or between the ferrocene units,7 leading to
important applications such as electrochemical sensing of
anions.7f−h
RESULTS AND DISCUSSION
■
Synthesis and Structure of 1. Compound 1 was
synthesized according to Scheme 1. The key starting material,
(N-methylbenzimidazol-2-yl)ferrocene (1a), was obtained by a
procedure modified from that used for a related compound
reported previously.8 The benzimidazolyl-directed ortho
lithiation of 1a, followed by addition of 1 equiv of
Scheme 1. Synthesis of Compound 1
Received: February 7, 2014
Published: March 21, 2014
© 2014 American Chemical Society
1787
dx.doi.org/10.1021/om500138f | Organometallics 2014, 33, 1787−1793

Organometallics
Article
dimesitylboron fluoride, led to the formation of compound 1,
which was isolated in ∼60% yield as a reddish crystalline solid.
The 11B NMR spectrum of 1 has a peak at 0.60 ppm that is
characteristic of a four-coordinate boron. The four aryl protons
and the six methyls on the two mesityl groups display well-
1
resolved and distinct peaks in the H NMR spectrum of 1, in
agreement with the asymmetric nature of the molecule and the
highly congested environment around the boron center.
However, variable-temperature 1H NMR and 2D NOESY
NMR spectra indicated the existence of a dynamic exchange
between the two mesityls and between the two aryl protons on
the same mesityl ring. From NOESY/EXSY experiments (see
the Supporting Information), the Gibbs free energy for the
exchange between two mesityl groups was determined to be
67.0 kJ/mol at 273 K, while that for the exchange between two
aryl protons of the same mesityl is 70.0 kJ/mol. The proton
chemical shift exchange within the same mesityl ring can be
explained by the rotation around the B−C bond.9a The
dynamic exchange between the two mesityl groups is most
likely caused by the dissociation of the B−N bond, leading to
the formation of the intermediate A, followed by the rotation of
B−CCp bond and the re-formation of the B−N bond, as shown
in Scheme 2. The activation energy for the B−N bond
Figure 1. Crystal structure of one of the independent molecules in the
asymmetric unit of 1 with 50% thermal ellipsoids. H atoms are omitted
for clarity.
Table 1. Selected Bond Lengths (Å) and Angles (deg)
compd
1
2
3
Lengths
Fe−C (Cp-B)
2.024(2)
2.027(2)
2.033(2)
2.054(2)
2.125(2)
2.029(2)
2.035(3)
2.035(3)
2.039(3)
2.047(3)
1.665(3)
1.646(3)
1.659(3)
1.658(3)
2.030(3)
2.035(3)
2.035(3)
2.035(3)
2.036(3)
2.027(3)
2.044(3)
2.044(3)
2.046(2)
2.052(3)
1.566(4)
1.590(4)
2.033(8)
2.067(8)
2.080(8)
2.148(7)
2.200(7)
2.052(8)
2.059(8)
2.109(8)
2.116(8)
2.129(9)
1.635(11)
1.637(11)
1.653(11)
1.647(10)
Scheme 2. Proposed Mechanism for the Interconversion of
the Two Mesityl Groups in 1
Fe−C (Cp)
B−CCp
B−CMes
B−N
B−O
1.351(4)
dissociation in 1 is similar to that reported previously for a class
of o-(aminomethyl)arylboronate compounds9b which undergo
a reversible B−N dative bond dissociation/association process
similar to that observed for 1. The B−N dissociation in 1 is
quite unusual, since the same phenomenon was not observed in
the previously reported N,C-chelated BMes2 compounds
involving the phenyl−benzimidazolyl and the related phenyl−
azolyl chelate ligands.1f The greater steric congestion imposed
by the ferrocene unit in 1, in comparison to the nonferrocene-
containing N,C-chelated BMes2 compounds, may be respon-
sible for the unusual B−N bond dissociation/reassociation
process of 1.
The crystal structure of 1 was examined by a single-crystal X-
ray analysis, which confirmed the congested nature of 1, as
shown in Figure 1. There are two independent molecules in the
asymmetric unit of 1. Selected bond lengths and angles for one
of them are shown in Table 1. The B−N bond length
(1.658(3), 1.644(3) Å) in 1 is indeed much longer than those
observed previously in the phenyl−azolyl chelated BMes2
compounds (1.620−1.635 Å),1f while the B−CCp and B−CMes
bonds are longer (by ∼0.01 Å) than the typical B−C bond
lengths observed in previously reported N,C-chelated BMes2
compounds.1c−f The Fe−C bonds on the Cp ring that is bound
to the boron atom vary significantly in bond lengths, as shown
in Table 1. The Fe−C(2) bond (2.125(2) Å) is ∼0.1 Å longer
than other Fe−C bonds, which can be attributed to the binding
of C(2) to the bulky BMes2 unit. The two Cp rings in 1 are
nearly parallel to each other with tilt angles of 8.7 and 4.8°,
Angles
CCp−B−N
CMes−B−CMes
CMes−B−N
94.14(17)
93.7(5)
113.36(19)
117.22(18)
105.82(18)
106.33(18)
119.00(19)
115.8(6)
114.1(6)
105.5(6)
118.1(6)
107.7(6)
CMes−B−CCp
118.2(3)
CCp−B−O
CMes−B−O
123.4(3)
118.4(3)
2.9
tilt angle between the two Cp rings 8.7/4.8
13.7
respectively, for the two independent molecules in the
asymmetric unit.
Chemical and Photochemical Reactivity of 1. Under
ambient conditions, we observed that the solution color of 1
slowly changed from red to yellow-orange. NMR spectra
indicated that 1 was converted to the new species 2
quantitatively over a period of a few weeks if the solvent
used was not dried rigorously. NMR and single-crystal X-ray
diffraction analyses established that 2 is a ring-opened,
hydrolyzed product of 1 (Scheme 3).
As shown in Figure 2, the boron atom in 2 is no longer
bound to the benzimidazolyl nitrogen atom and one of the
mesityls on the boron is replaced by a hydroxyl group. The
geometry of the boron atom is a trigonal plane. The B−O bond
length (1.351(4) Å) is characteristic of a BO bond due to the
formation of a π bond between the O and the B atoms.10 The
1788
dx.doi.org/10.1021/om500138f | Organometallics 2014, 33, 1787−1793

Organometallics
Article
would lead to the product 2. The fact that the addition of
excess H2O to the solution of 1 did not accelerate the
hydrolysis process supports that the formation of the
intermediate A is likely the rate-determining step. All attempts
to trap the intermediate A by adding other Lewis donors such
as NBu4F and NBu4Cl to a rigorously dried solution of 1 led to
the formation of 2, in which the OH− group came from the
H2O molecules associated with the NBu4F and NBu4Cl
compounds. This could be attributed to the high stability of
compound 2. The facile hydrolysis of 1 is in sharp contrast with
the previously reported non-ferrocene-containing N,C-chelated
BMes2 compounds, which did not exhibit the same behavior at
all. In addition, similar hydrolysis was not observed for a related
Fe(Cp)[(Cp-3,5-Me2py)BMes2] compound in which, instead
of a benzimidazolyl ring, a pyridyl ring is bound to the BMes2
unit.11 Since N-methylbenzimidazole has a donor strength
which is comparable to that of 3,5-Me2pyridine and the steric
environment around the boron atom in 1 is similar to that in
Fe(Cp)[(Cp-3,5-Me2py)BMes2], the greater hydrolysis ten-
dency by 1 is likely caused by the more strained chelation to the
boron atom by the two five-membered rings in 1, in
comparison to that by a five-membered and a six-membered
ring in Fe(Cp)[(Cp-3,5-Me2py)BMes2], which may also be a
key factor in the dynamic exchange of the two mesityl groups in
1.
Scheme 3. Proposed Mechanism for the Formation of 2
Compounds 1 and 2 have distinct electronic properties. As
shown in Figure 3, the characteristic ferrocene low-energy
Figure 2. Crystal structure of 2 with 50% thermal ellipsoids. H atoms
are omitted for clarity.
B−CCp and B−CMes bonds are much shorter than those in 1,
due to the much reduced congestion. The Fe−C bond lengths
in 2 show much less variation in comparison those in 1. The
Fe−C(6) bond length (2.052(3) Å) in 2 is also much shorter
than the corresponding length (Fe−C(2)) in 1, further
supporting that steric congestion is the main cause for the
long Fe−C and B−N bonds in 1. As a consequence of the
reduced steric congestion, the tilt angle between the two Cp
rings in 2 is smaller (2.9°). The H atom of the OH group in 2
was located in the difference Fourier map and refined
successfully. It forms a fairly strong intramolecular hydrogen
bond with the benzimidazolyl N atom, as indicated by the
O(1)···N(1) distance (2.686(7) Å). The chemical shift of this
Figure 3. (left) UV−vis absorption spectra of 1 and 2 in CH2Cl2.
(right) CV diagrams of 1 and 2 showing the oxidation wave (vs Fc/
Fc+) at a 75 mV/s scan rate with [Bu4N][PF6] (0.1 M) as the
electrolyte in CH3CN.
absorption band appears at ∼490 nm for 1 and at ∼450 nm for
2 with ε ≈ 1200 M−1 cm−1, which accounts for the distinct
colors of these two compounds. The 450 nm absorption band
of 2, in fact, is similar to that of ferrocene. Both compounds
also display the characteristic reversible 1e oxidation peaks at
E1/2 = −67 and 205 mV, respectively (relative to Fc/Fc+), in
the CV diagram shown in Figure 3. The 270 mV increase of the
oxidation potential from 1 to 2 indicates a considerable
decrease of the electron density on the Fe(II) center, which
may be attributed to the BMes(OH) unit, since the same trend
was observed in the previously reported ferrocene compounds
functionalized by an electron-deficient three-coordinate boron
unit.7a The loss of chelation between the Cp and the
benzimidazolyl ring is believed to be another key factor in
the higher oxidation potential of 2, because the oxidation
potential of the nonborylated precursor compound 1a was
found to be more positive than that of ferrocene (E1/2 = 128
mV, relative to Fc/Fc+), albeit to a lesser degree, in comparison
1
H atom appears at ∼14 ppm in the H NMR spectrum of 2,
which is in agreement with the H bond formation. The boron
atom in 2 displays a broad peak at 48.6 ppm in the 11B NMR
spectrum, consistent with a three-coordinated boron center.
The formation of the hydrolyzed product 2 is consistent with
the proposed B−N bond dissociation process of 1 shown in
Scheme 2. It is conceivable that, after the formation of the
intermediate A, external donor molecules such as H2O can
compete with the internal nitrogen donor atom for binding to
the boron center, forming the intermediate B shown in Scheme
3. An intramolecular H bond between the H2O and the N atom
would stabilize the intermediate B, making the hydrolysis a
highly competitive pathway, in comparison to the re-formation
of the B−N bond. A similar internal H bond facilitated
hydrolysis phenomenon involving a BMes2 group was observed
previously.10b Elimination of a mesitylene molecule from B
1789
dx.doi.org/10.1021/om500138f | Organometallics 2014, 33, 1787−1793

Organometallics
Article
to that of 2. The attachment of the relatively electron negative
benzimidazolyl ring to the Cp ring would certainly decrease the
π electron density on the Cp ring, which in turn would reduce
the electron density on the Fe(II) center, leading to the more
positive oxidation potential of 1a, relative to ferrocene. Adding
the BMes2 chelating unit enhances the π conjugation of the Cp
and the benzimidazolyl ring. However, because the mesityls are
electron-donating groups, the net consequence of the BMes2
chelation is to put more electron density on the Fe(II) center;
thus, the oxidation potential of 1 is more negative, relative to
1a. TD-DFT computational studies in fact confirmed that one
of the mesityls has a large contribution to the HOMO of 1 (see
the Supporting Information). TD-DFT data further established
that the weak absorption band of 1 at ∼490 nm is mainly from
d → π*Cp‑benzimidazolyl and d → π*Cp transitions (MLCT) with a
small contribution from a Mes→ π*Cp transition (see the
Supporting Information). The BMes2 unit in 1 lowers the π*
energy by enhancing the π conjugation of the Cp−
benizimidazolyl unit via chelation and raises the filled d orbital
energy via σ donation (as evidenced by the oxidation potential
of 1), which leads to the relatively low energy MLCT band and
the distinct red color of 1. Well-defined MLCT transitions in
borylated ferrocenes have been observed previously.12 For 2,
the key contributions to the absorption band at ∼450 nm were
found to involve mainly d → π*Cp‑B and d → π*benzimidazolyl‑B
transitions (MLCT). The boron center in 2 has a significant
contribution to the π* level, which should lower the π* energy.
This is, however, offset by the loss of the chelation between the
Cp and the benzimidazolyl ring in 2. Thus, the higher MLCT
energy of 2 relative to that of 1 is mainly caused by the
decreased electron density (or the stabilization of the filled d
orbitals) on the Fe(II) center induced by the unsaturated boron
unit.
Figure 4. Crystal structure of 3 with 50% thermal ellipsoids. The H
atoms are omitted for clarity.
In comparison to those of 1, the Fe−C bonds in 3 are much
longer, which is in agreement with the trend reported in the
literature for ferrocenium cations and can be attributed to the
decreased π back-bonding between the Fe(III) ion and the Cp
rings.13 As observed in the structure of 1, one of the mesityls
(the C(28) ring) is on the same side as the Fe(III) center, thus
being very close to the Fe atom (Fe(1)···C(28) = 3.52(1) Å).
Significantly, the Fe−C bonds that are close to the C(28)
mesityl ring are all longer than other Fe−C bonds (Fe(1)−
C(1) 2.200(7) Å, Fe(1)−C(5) 2.148(7) Å, Fe(1)−C(6)
2.129(9) Å, Fe(1)−C(7) 2.116(8) Å). Furthermore, the two
Cp rings in 3 have a much greater tilt angle (13.7°) and are
more open toward the C(28) mesityl ring in comparison to
those in 1.
Despite the high degree of steric congestion, compound 1 is
very stable toward irradiation by light. No change was observed
at all when the dry C6D6 solution of 1 was irradiated under
nitrogen at 365 nm or using white light for several hours.
Although the HOMO level of 1 has appreciable contributions
from a mesityl and the LUMO has a large contribution from
the π* orbital of the benzimidazolyl−Cp chelate, the HOMO
→ LUMO transition contributes very little to the first five
excited states, which are dominated by d−d and MLCT
transitions (see the Supporting Information). Thus, the
inactivity of the boron unit in 1 toward photoisomerization
can be explained by the presence of the low-lying electronic
states involving the ferrocene unit that effectively quench the
photoisomerization pathway by trapping the excitons.
Electronic and Chemical Properties of 3. To examine
the electronic structures of 3, we recorded the UV−vis−near-IR
absorption spectra of [1+] generated in situ using a
stoichiometric amount of tris(4-bromophenyl)ammoniumyl
hexachloroantimonate and iodine, respectively. Upon the
addition of the oxidant, a broad band covering the range
600−1300 nm appeared with λmax ∼950 nm (ε = 400 M−1
cm−1), which matches well with that observed for the isolated
pure compound 3 (Figure 5). This low-energy band may be
assigned to a ligand-to-metal CT transition, according to the
well-documented work for ferrocenium ions in the literature
(∼617 nm, ε = 390 M−1 cm−1 for the nonsubstituted
ferrocenium at 300 K).14 This band is known for its sensitivity
to substituent groups on the Cp ring.14b
Compound 1 can be oxidized readily by a mild oxidant such
as I2, generating the dark brown species [1+]I3 (3)
quantitatively (Scheme 4), which was fully characterized by
NMR, elemental, and single-crystal X-ray diffraction analysis.
The crystal structure of 3 is shown in Figure 4.
Scheme 4. Formation of 3 via the Oxidation of 1 by I2
Figure 5. UV−vis−near-IR absorption spectra change of 1 upon the
addition of an oxidant: [N(4-Br-C6H4)3][SbCl6] (left) and I2 (right)
in CH2Cl2.
1790
dx.doi.org/10.1021/om500138f | Organometallics 2014, 33, 1787−1793

Organometallics
Article
The EPR spectra of 3 was recorded at 21 and 9 K,
respectively. At both temperatures, a highly anisotropic EPR
signal with an axial character was observed at g∥ = 3.11 and g⊥ =
1.94 (Figure 6), which is characteristic of ferrocenium cations.15
1
Figure 7. H NMR spectra of 3 in CD3CN at 298 K; (bottom) full
spectrum (the asterisks indicate all the methyl groups); (top) enlarged
region from +13.0 to −1.0 ppm showing the assigned peaks (color
coded).
Figure 6. EPR spectrum of 3 recorded at 21 K. Inset: spin density
distribution diagram of [1+] obtained from DFT data (plotted with an
isospin value of 0.004).
delocalization/electronic interaction between the Fe(III) center
and the flanking C(28) mesityl ring.
2D NOESY NMR data indicated that there is a dynamic
exchange between the two mesityl groups in 3 similar to that
observed in compound 1, which is likely through the same B−
N bond dissociation process proposed for 1. Compound 3 is
stable in the solid state under ambient conditions for months.
However, it can be reduced back to compound 1 in solution by
weak donor solvent molecules such as DMSO. A similar
phenomenon was observed for some of the previously reported
ferrocenium compounds.16,17 The presence of the low-lying
LMCT/d−d states in 3 could retard the photoisomerization of
the boron unit in 3. In fact, 3 was found to be sensitive toward
light in solution and rapidly decomposes to unidentified species
upon exposure to light, and no evidence of photoisomerization
of the boron unit was observed.
Nonetheless, the much smaller g∥ value in comparison to that
(∼3.8) of a nonsubstituted ferrocenium indicates that the spin
density extends partially to the organic substituent.5a,15 The
spin density distribution obtained from the DFT calculation
data of [1+] corroborated the EPR results. Single-point
calculations on the radical cation [1+] without the counterion
were performed at the theory level of the BP86 function, using
the basis set of def2SVP for Fe and 6-31G(d) for other atoms.
The tilt angle of the two Cp rings from the optimized geometry
of [1+] is 13°, similar to that observed in the crystal structure.
As shown in Figure 6, the calculated spin density in [1+] is
mainly localized on the ferrocenium center with a notable
delocalization to the flanking mesityl group (the C(28) ring).
The NMR spectrum of 3 in CD3CN was recorded to further
examine the influence of the paramagnetic Fe(III) center on the
surrounding organic ligands.16 The 11B NMR signal of 3
appears at −30 ppm, which is shifted ∼30 ppm upfield relative
to that of 1. As shown in Figure 7, the 1H chemical shifts of the
aromatic protons on the benzimidazole moiety in 3 are similar
to those in 1. The chemical shifts of all seven methyls are
distinct and are spread over the range of +26 to −8 ppm.
Assignments for some of the chemical shifts of 3 were made
possible with the aid of 2D COSY and NOESY data. The
chemical shifts for two of the methyls and the two aryl H atoms
on the C(19) mesityl ring that is far away from the Fe(III)
center were assigned without ambiguity. Because the difference
in the chemical shifts of these protons in comparison to those
in 1 is relatively small (<2 ppm), the influence of the Fe(III)
center on this mesityl ring is considered to be small. In contrast,
the peaks of the methyls and the aryl H atoms on the C(28)
mesityl ring experience a much greater shift. For example, the
chemical shifts of the two aryl H atoms on the C(28) mesityl
ring appear at 12.4 and 2.3 ppm, respectively, which are shifted
by more than 3 ppm in comparison to those of 1. These data
further support the greater influence of the unpaired electron of
the Fe(III) center and the partial through-space spin
delocalization to the C(28) mesityl ring. The greater tilting
of the two Cp rings in 3 is likely a consequence of the spin
CONCLUSIONS
■
Using the N,C-chelated BMes2-functionalized ferrocene com-
pound 1, we have illustrated the great mutual influence by the
boron unit and the ferrocene unit on the electronic,
photochemical, and chemical properties of the molecule. The
low-lying electronic states of the ferrocene unit prohibit the
photoisomerization of the boron unit. In addition, the steric
congestion imposed by the ferrocene unit and the chelate strain
by Cp and the N-methylbenzimidazolyl groups are likely
responsible for the dissociation of the B−N bond, which leads
to an unusual dynamic exchange of the two mesityl rings and
the ring-opening hydrolysis of the boron unit. On the other
hand, the BMes2 chelate unit reinforces the MLCT process on
the ferrocene unit, providing molecule 1 a low-energy MLCT
state and a distinct red color. Furthermore, one of the mesityl
rings on the boron center has been found to have a notable
interaction with the unpaired electron of the Fe(III) ion in 3,
which leads to a significant delocalization of the spin density
through space.
EXPERIMENTAL SECTION
General Procedures. All starting materials were purchased from
■
Aldrich Chemical Co. All reactions were carried out under a nitrogen
atmosphere using Schlenk and vacuum line techniques. The H, 13C,
1
1791
dx.doi.org/10.1021/om500138f | Organometallics 2014, 33, 1787−1793

Organometallics
Article
11B, and 2D HH COSY and NOESY NMR spectra were recorded on a
Bruker Avance 400 or 600 MHz spectrometer. UV−vis spectra were
recorded on a Cary50 UV−visible spectrometer. High-resolution mass
spectra (HRMS) were recorded on an Applied Biosystems Qstar XL
spectrometer. Elemental analyses were conducted at the Laboratoire
at the Cambridge Crystallographic Data Center (CCDC 983808,
983809, and 983810 for 1−3, respectively).
ASSOCIATED CONTENT
* Supporting Information
■
S
́
́ ́ ́
d’Analyze Elementaire de l’Universite de Montreal.
Figures, tables, and CIF files giving 1D and 2D NMR data,
computational data and diagrams, and crystal structure data.
This material is available free of charge via the Internet at
http://pubs.acs.org.
Synthesis of B(2-ferrocenyl-N-Me-benzimidazolyl)Mes2 (1).
tert-Butyllithium (1.76 mL, 1.70 M in hexanes, 3.0 mmol) was added
dropwise to a solution of 1a (N-methylbenzimidazol-2-yl-ferrocene,
0.79 g, 2.0 mmol) at −78 °C in THF. After the solution was stirred for
1 h, BMes2F (0.86 g, 3.0 mmol) in 5 mL of THF was added dropwise.
The reaction mixture was then warmed to room temperature and
stirred overnight. Compound 1 was isolated as a reddish solid (0.70 g,
∼60% yield) by column chromatography on silica gel using hexanes/
AUTHOR INFORMATION
Corresponding Author
*E-mail for S.W.: wangs@chem.queensu.ca.
■
1
CH2Cl2 (50/50 v/v) as the eluent. H NMR (400 MHz, CD2Cl2, 25
Notes
°C, δ, ppm): 7.43 (d, 1H, J = 8.5 Hz), 7.38 (d, 1H, J = 8.0 Hz), 7.29 (t,
1H, J = 8.0 Hz, J = 1.0 Hz), 7.21 (t, 1H, J = 8.5 Hz, J = 1.5 Hz), 6.84
(s, 1H), 6.70 (s, 1H), 6.61 (s, 1H), 6.43 (s, 1H), 4.81 (d, 1H, J = 2.5
Hz), 4.75 (d, 1H, J = 2.5 Hz), 4.58 (t, 1H, J = 2.5 Hz), 3.95 (s, 3H),
3.72 (s, 5H), 2.25 (s, 3H), 2.21 (s, 3H), 2.13 (s, 3H), 2.11 (s, 3H),
1.92 (s, 3H), 1.26 (s, 3H). 13C NMR (100 MHz, CD2Cl2, 25 °C, δ,
ppm): 161.1, 142.9, 141.6, 141.0, 137.2, 137.0, 136.5, 133.8, 132.3,
129.7, 129.4, 128.9, 128.7, 123.9, 122.5, 115.7, 110.3, 74.5, 73.3, 72.3,
69.0, 59.2, 30.9, 26.4, 24.8, 24.3, 23.7, 20.3. 11B NMR (160 MHz,
CD2Cl2, 25 °C, δ, ppm): 0.6. Anal. Calcd for C36H37N2BFe·
0.5CH2Cl2: C, 72.19; H, 6.26; N, 4.61. Found: C, 72.28; H, 6.33;
N, 4.62. HRMS (TOF MS EI+): calcd, 564.2421; obsd, 564.2399.
Synthesis of 2-[BMes(OH)]-1-[N-methylbenzimidazol-2-yl]-
ferrocene (2). Compound 2 was isolated as yellow crystals from a
solution of 1 in CH2Cl2 after standing under ambient conditions for
several weeks. 1H NMR (400 MHz, CD2Cl2, 25 °C, δ, ppm): 13.8 (s,
1H), 7.79 (m, 1H), 7.46 (m, 1H), 7.36 (m, 2H), 6.86 (s, 2H), 5.23
(dd, 1H, J = 2.4 Hz, J = 1.2 Hz), 4.67 (t, 1H, J = 2.4 Hz), 4.21 (m,
6H), 4.03 (s, 3H), 2.33 (s, broad, 9H). 13C NMR (100 MHz, CD2Cl2,
25 °C, δ, ppm): 139.3, 136.8, 127.0, 122.7, 122.6, 118.1, 109.4, 79.8,
72.8, 72.4, 70.6, 32.3, 22.3, 20.8. 11B NMR (160 MHz, CD2Cl2, 25 °C,
δ, ppm): 48.6. Anal. Calcd for C27H27N2BOFe·0.5H2O: C, 68.77; H,
5.94; N, 5.94. Found: C, 68.47; H, 5.76; N, 6.08. HRMS (TOF MS
EI+): calcd, 462.1566; obsd, 462.1583.
Synthesis of [2-BMes2-1-(N-methylbenzimidazol-2-yl)-
ferrocenium]I3 (3). A 5 mL portion of a CH2Cl2 solution of I2
(128 mg, 0.5 mmol) was added to a stirred solution of 1 (200 mg, 0.34
mmol) in 5 mL of CH2Cl2 at ambient temperature. The dark brown
precipitates of 3 were collected and washed with hexanes after the
solution was kept standing overnight at ambient temperature. 1H
NMR (600 MHz, CD3CN, 25 °C, δ, ppm): 28.1 (s, broad), 26.0 (s,
broad), 16.7 (s), 12.4 (s), 9.22 (s), 8.46 (s, broad), 8.2 (s), 7.22 (d, J =
5.4 Hz), 6.93 (s), 6.60 (s), 5.52 (s), 3.48 (s), 2.35 (s), −0.56 (s),
−5.53 (s), −7.47 (s). Anal. Calcd for C36H37BFeN2I3: C, 45.71; H,
3.91; N, 2.96. Found: C, 45.19; H, 3.82; N, 2.87.
Computational Studies. All calculations were performed using
the Gaussian 09 software.18a For compounds 1 and 2, DFT and TD-
DFT calculations were performed at the B3LYP level of theory,18b,c
using the basis set of LanL2DZ for Fe18d,e and 6-31G(d) for other
atoms. For the cation [1+], the DFT calculations were performed at
the theory level of the BP86 function, using the basis set of def2SVP
for Fe18f and 6-31G(d) for other atoms. The geometric parameters
obtained from crystal structure data were used as the initial points for
geometry optimization calculations.
X-ray Crystallographic Analysis. Single crystals of 1−3 were
mounted on glass fibers and were collected on a Bruker Apex II single-
crystal X-ray diffractometer with graphite-monochromated Mo Kα
radiation, operating at 50 kV and 30 mA at 180 K. Data were
processed on a PC with the aid of the Bruker SHELXTL software
package (version 6.10)19 and corrected for absorption effects. All non-
hydrogen atoms were refined anisotropically. The crystal of 1 contains
two independent molecules and one CH2Cl2 solvent molecule in the
asymmetric unit. Complete crystal structure data can be found in the
Supporting Information. The crystal data of 1−3 have been deposited
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the Natural Sciences and Engineering Research
Council for financial support. Y-L.R. acknowledges the 2013
JSPS Summer Program Fellowship and NSERC Michael Smith
Foreign Study Supplement. The authors are also grateful to
Hiroaki Maeda, Kenji Takada, and Dr. Francoise Sauriol for
their assistance in recording vis−near-IR absorption spectra and
NMR spectra, respectively.
REFERENCES
■
(1) (a) Iida, A.; Saito, S.; Sasamori, T.; Yamaguchi, S. Angew. Chem.,
Int. Ed. 2013, 52, 376. (b) Ansorg, K.; Braunschweig, H.; Chiu, C.-W.;
Engels, B.; Gamon, D.; Hugel, M.; Kupfer, T.; Radacki, K. Angew.
̈
Chem., Int. Ed. 2011, 50, 2833. (c) Rao, Y. L.; Amarne, H.; Zhao, S. B.;
McCormick, T. M.; Martic, S.; Sun, Y.; Wang, R. Y.; Wang, S. J. Am.
Chem. Soc. 2008, 130, 12898. (d) Rao, Y.-L.; Wang, S. Inorg. Chem.
2011, 50, 12263. (e) Rao, Y.-L.; Chen, L. D.; Mosey, N. J.; Wang, S. J.
Am. Chem. Soc. 2012, 134, 11026. (f) Amarne, H.; Baik, C.; Murphy, S.
K.; Wang, S. Chem. Eur. J. 2010, 16, 4750. (g) Rao, Y.-L.; Amarne, H.;
Chen, L. D.; Brown, M. L.; Mosey, N. J.; Wang, S. J. Am. Chem. Soc.
2013, 135, 3407.
(2) (a) Vos, J. G.; Pryce, M. T. Coord. Chem. Rev. 2010, 254, 2519.
(b) Guerchais, V.; Ordronneau, L.; Le Bozec, H. Coord. Chem. Rev.
2010, 254, 2533 and references therein.
(3) (a) Rao, Y.-L.; Wang, S. Organometallics 2011, 30, 4453.
(b) Wang, N.; Ko, S.-B.; Lu, J.-S.; Chen, L. D.; Wang, S. Chem. Eur. J.
2013, 19, 5314.
(4) (a) Kume, S.; Kurihara, M.; Nishihara, H. Chem. Commun. 2001,
1656. (b) Meng, F.; Hervault, Y.-M.; Shao, Q.; Hu, B.; Norel, L.;
Rigaut, S.; Chen, X. Nat. Commun. 2014, DOI: 10.1038/ncomms4023.
(c) Wen, H.-W.; Wang, J.-Y.; Hu, M.-Q.; Li, B.; Chen, Z. N.; Chen, C.-
N. Dalton Trans. 2012, 41, 11813. (d) Tanaka, Y.; Inagaki, A.; Akita,
M. Chem. Commun. 2007, 1169. (e) Motoyama, K.; Koike, T. K.;
Akita, M. Chem. Commun. 2008, 5812.
(5) (a) Kurihara, M.; Matsuda, T.; Hirooka, A.; Yutaka, T.; Nishihara,
H. J. Am. Chem. Soc. 2000, 122, 12373. (b) Sakamoto, R.; Murata, M.;
Nishihara, H. Angew. Chem., Int. Ed. 2006, 45, 4793. (c) Sakamoto, R.;
Kume, S.; Nishihara, H. Chem. Eur. J. 2008, 14, 6978.
(6) (a) Miesel, D.; Hildebrandt, A.; Korb, M.; Low, P. J.; Lang, H.
Organometallics 2013, 32, 2993. (b) Kusamoto, T.; Takada, K.;
Sakamoto, R.; Kume, S.; Nishihara, H. Inorg. Chem. 2012, 51, 12102.
(c) Kusamoto, T.; Nishihara, H.; Kato, R. Inorg. Chem. 2013, 52,
13809. (d) Herbert, D. E.; Mayer, U. F. J.; Manners, I. Angew. Chem.,
Int. Ed. 2007, 46, 5060.
(7) (a) Heilmann, J. B.; Scheibitz, M.; Qin, Y.; Sundararaman, A.;
Jakle, F.; Kretz, T.; Bolte, M.; Lerner, H.-W.; Holthausen, M. C.;
̈
Wagner, M. Angew. Chem., Int. Ed. 2006, 45, 920. (b) Venkatasubbaiah,
K.; Zakharov, L. N.; Kassel, W. S.; Rheingold, A. L.; Jakle, F. Angew.
̈
Chem., Int. Ed. 2005, 44, 5428. (c) Heilmann, J. B.; Qin, Y.; Jakle, F.;
̈
Lerner, H.-W.; Wagner, M. Inorg. Chim. Acta 2006, 359, 4802.
1792
dx.doi.org/10.1021/om500138f | Organometallics 2014, 33, 1787−1793

Organometallics
Article
(d) Kaufmann, L.; Breunig, J.-M.; Vitze, H.; Schodel, F.; Nowik, I.;
̈
Pichlmaier, M.; Bolte, M.; Lerner, H.-W.; Winter, R. F.; Herber, R. H.;
Wagner, M. Dalton Trans. 2009, 2940. (e) Braunschweig, H.; Chiu, C.-
W.; Gamon, D.; Kaupp, M.; Krummenacher, I.; Kupfer, T.; Muller, R.;
̈
Radacki, K. Chem.Eur. J. 2012, 18, 11732. (f) Day, J. K.; Bresner, C.;
Coombs, N. D.; Fallis, I. A.; Ooi, L.-L.; Aldridge, S. Inorg. Chem. 2008,
47, 793. (g) Siewert, I.; Fitzpatrick, P.; Broomsgrove, A. E. J.; Kelly,
M.; Vidovic, D.; Aldridge, S. Dalton Trans. 2011, 40, 10345. (h) Remi
Tirfoin, R.; Aldridge, S. Dalton Trans. 2013, 42, 12836.
́
(8) Herault, D.; Aelvoet, K.; Blatch, A. J.; Al-Majid, A.; Smethurst, C.
A.; Whiting, A. J. Org. Chem. 2007, 72, 71.
(9) (a) Li, Y.-F.; Kang, Y.; Ko, S.-B.; Rao, Y.; Sauriol, F.; Wang, S.
Organometallics 2013, 32, 3063. (b) Hoang, C. T.; Prokes, I.; Clarkson,
G. J.; Rowland, M. J.; Tucker, J. H. R.; Shipman, M.; Walsh, T. R.
Chem. Commun. 2013, 49, 2509.
(10) (a) Shuvalov, R. R.; Burns, P. C. Acta Crystallogr. 2003, C59, i47.
(b) Rao, Y. L.; Wang, S. Inorg. Chem. 2009, 48, 7698.
(11) Jakle, F. Personal communication.
̈
(12) (a) Thomson, M. D.; Novosel, M.; Roskos, H. G.; Muller, T.;
̈
Scheibitz, M.; Wagner, M.; de Biani, F. F.; Zanello, P. J. Phys. Chem. A
2004, 108, 3281. (b) de Biani, F. F.; Gmeinwieser, T.; Herdtweck, E.;
Jakle, F.; Laschi, F.; Wagner, M.; Zanello, P. Organometallics 1997, 16,
̈
4776.
(13) Martinez, R.; Tiripicchio, A. Acta Crystallogr., Sect. C 1990, C46,
202.
(14) (a) Sohn, Y. S.; Hendrickson, D. N.; Gray, H. B. J. Am. Chem.
Soc. 1970, 92, 3233. (b) Prins, R. Chem. Commun. 1970, 280.
(15) Prins, R. Mol. Phys. 1970, 19, 603.
(16) Bliimel, J.; Hebendanz, N.; Hudeczek, P.; Kohler, F. H.; Strauss,
W. J. Am. Chem. Soc. 1992, 114, 4223.
(17) Khobragade, D. A.; Mahamulkar, S. G.; Pospisil, L.; Cisarova, I.;
Rulisek, L.; Jahn, U. Chem. Eur. J. 2012, 18, 12267.
(18) (a) Frisch, M. J. et al. Gaussian 09, Revision C.01; Gaussian, Inc.,
Wallingford, CT, 2010. (b) Petersson, G. A.; Bennett, A.; Tensfeldt, T.
G.; Al-Laham, M. A.; Shirley, W. A.; Mantzaris, J. J. Chem. Phys. 1988,
89, 2193. (c) Petersson, G. A.; Al-Laham, M. A. J. Chem. Phys. 1991,
94, 6081. (d) Dupuis, M.; Rys, J.; King, H. F. J. Chem. Phys. 1976, 65,
111. (e) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299.
(f) Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7, 3297.
(19) SHELXTL Version 6.14; Bruker AXS, Madison, WI, 2000−2003.
NOTE ADDED AFTER ASAP PUBLICATION
■
This paper was published ASAP on March 21, 2014, with an
error to ref 11 and the Table of Contents graphic. The
corrected version was reposted on March 26, 2014.
1793
dx.doi.org/10.1021/om500138f | Organometallics 2014, 33, 1787−1793