The divalent iron complexes trans-[FeBr2(BINC)2], [Cp*FeCl(BINC)] (Cp* = Me5C5), and [FeBr2(CNAr3NC)2] with the chelating bis(isonitrile) ligands BINC (bis(2-isocyanophenyl)phenylphosphonate) and CNAr3NC (2,2″-diisocyano-3,5,3″,5"tetramethyl-1,1':3',1″-terphenyl) have been prepared and characterized. Their subsequent reduction yields the di- and trinuclear compounds [Fe3(BINC)6], [Cp*Fe(BINC)]2, [Fe(CNAr3NC)2]2, and [K(Et2O)]2[Fe(CNAr3NC)2]2. The molecular structures of all new species were determined by X-ray crystallography and compared to those of related iron carbonyl complexes, demonstrating that the bidentate isonitrile ligands are capable surrogates for two CO ligands with only minimal distortion of the tetrahedral or octahedral geometry of the parent complexes. The complexes were further characterized by NMR and IR spectroscopy, and the electrochemical properties of selected compounds were analyzed by UV-vis-NIR spectroelectrochemistry.
The divalent ironcomplexes trans-[FeBr2(BINC)2], [Cp*FeCl(BINC)] (Cp* = Me5C5), and [FeBr2(CNAr3NC)2] with the chelating bis(isonitrile) ligands BINC (bis(2-isocyanophenyl)phenylphosphonate) and CNAr3NC (2,2″-diisocyano-3,5,3″,5"tetramethyl-1,1':3',1″-terphenyl) have been prepared and characterized. Their subsequent reduction yields the di- and trinuclearcompounds [Fe3(BINC)6], [Cp*Fe(BINC)]2, [Fe(CNAr3NC)2]2, and [K(Et2O)]2[Fe(CNAr3NC)2]2. The molecular structures of all new species were determined by X-ray crystallography and compared to those of related iron carbonyl complexes, demonstrating that the bidentate isonitrile ligands are capable surrogates for two CO ligands with only minimal distortion of the tetrahedral or octahedral geometry of the parent complexes. The complexes were further characterized by NMR and IR spectroscopy, and the electrochemical properties of selected compounds were analyzed by UV-vis-NIR spectroelectrochemistry.
With the discovery
of n class="Chemical">isonitriles by Lieke[1] in 1859 came a
new and emerging field within organometallic chemistry.[2] Isonitriles show stronger σ-donating character
and weaker π-acceptor properties in comparison to carbon monoxide.
However, they feature similar behavior when they coordinate to metals
as both terminal and bridging ligands.[3,4] Isonitriles
can be electronically and sterically tuned through modification of
their organic substituents, which makes them versatile ligands for
metal catalysis.[5] Chelating bis(isonitriles)
confer rigidity to metalcomplexes, which has been proven to be especially
valuable for the design of 3d-metalcomplexes with sufficient excited-state
lifetimes to efficiently initiate single-electron-transfer (SET) processes.
This approach has been successful for copper,[6] molybdenum,[7] and rhenium,[8] providing alternatives to established but scarce ruthenium-
and iridium-based photocatalysts.[9]
Extending this concept to n class="Chemical">iron holds the promise of making this
metalbroadly available for photocatalysis. This is especially challenging,
given that excited-state lifetimes of ironcomplexes are often in
the picosecond range and have been reported at best in the low-nanosecond
range.[10] However, there is only limited
data available on coordination complexes of iron with bidentate isonitriles,[20−23] in contrast with monodentate isonitrileironcomplexes, which have
attracted considerable attention due to the high prevalence of ironcarbonyl complexes.[11] For example, dinuclear
ironcomplexes such as [CpFe(CNPh)2][12] (A, Ph = C6H5) and [Fe2(CNPh)9][13] (B) feature terminal and bridging isonitrile ligands (Figure ).
Figure 1
Selected examples of
di- and trinuclear iron isonitrile complexes.
Selected examples of
di- and trinucleariron isonitrilecomplexes.The trinuclear compound [n class="Chemical">Fe3(CO)10(CNPh)2] (C)[14] possesses
both CO and monodentate isonitrile ligands, having replaced two of
the axial CO ligands in [Fe3(CO)12]. Iron isonitrile
analogues of [Fe(CO)4]2–, [Fe(CNXyl)4]2– (Xyl = 2,6-Me2C6H3)[15] and [Fe(CNArMes2)4]2– (ArMes2 = 2,6-(2,4,6-Me3C6H2)2C6H3),[16] further demonstrate the propensity
of monodentate isonitriles to act as surrogates for CO.
In comparison
to monodentate n class="Chemical">isonitrile ligands, the reactivity
and coordination chemistry of bidentate isonitriles has been generally
unexplored.[5] Given the linear structure
of the isonitrile unit, such ligands need to form much larger chelates
relative to the typical N,N- or P,P-ligands that are commonly found in five- and six-membered rings.
The minimal ring size in a bidentate isonitrile–metalcomplex
was found to be 12.[19] With this design
concept, a number of iron(II)complexes with bidentate isonitrile
ligands were realized; representative examples are D–G (Figure ).[5,20−23]
Some of us have previously
been interested in the synthesis and
reactivity of low-oxidation-state polyarene and n class="Chemical">alkene ferrates.[24] Such complexes have been used for various applications,
including the activation of small molecules (e.g., P4 and
CO2) and as catalysts for hydrogenation and hydroboration
reactions.[25] Taking advantage of the well-known
ability of isonitriles to stabilize low-oxidation-state transition-metalcomplexes,[15−18,26] in combination with imposing
rigidity, we were interested in exploring the potential of the aforementioned
chelating isonitrile ligands in the stabilization of highly reduced
ironcomplexes. We particularly questioned if polynuclear ironcomplexes
with chelating bis(isonitrile) ligands can still be formed, for which
to the best of our knowledge no previous examples have been described
in the literature. Here, we report the synthesis and characterization
of new ironcomplexes with the ligands BINC (bis(2-isocyanophenyl)phenylphosphonate)
and CNAr3NC (2,2″-diisocyano-3,5,3″,5”tetramethyl-1,1′:3′,1″-terphenyl,
demonstrating that indeed di- and trinuclear structures can be realized.
Results
and Discussion
Treating anhydrous Fen class="Chemical">Br2 with 2
equiv of BINC resulted
in the formation of trans-[FeBr2(BINC)2] (1) as a blue solid in 33% isolated yield (Scheme ). Compound 1 displays structural and spectroscopic features similar to
those of compounds D–G (see Figure S34 in the Supporting Information for
a graphical representation of the structure).
Scheme 1
Synthesis of [FeBr2(BINC)2] (1) and [Fe3(BINC)6] (2)
Reduction of 1 with an excess of KC8 in
n class="Chemical">THF afforded an orange-brown solution from which brown crystals of
[Fe3(BINC)6] (2) were isolated
in a yield of 26% by slow diffusion of n-hexane into
the reaction mixture.
Single-crystal X-ray diffraction (XRD, Figure ) revealed that 2 crystallizes
in the space group P1̅ with two molecules in
the unit cell. At first glance, the structure of 2 resembles
that of [n class="Chemical">Fe3(CO)12] (H) with respect
to the Fe1–Fe2–Fe3 core (Table ), showing that the the BINC ligand can displace
the carbonyl groups without imposing constraints due to its chelating
structure. Each atom in the triironcore is coordinated by two BINC
ligands, displaying one shorter and two longer Fe–Fe bonds
(Fe2–Fe3 2.495(6) Å vs Fe1–Fe2 2.682(6) Å
and Fe1–Fe3 2.685(6) Å) resembling an isosceles triangle.
The shorter Fe2–Fe3 bond features two bridging isonitrile donor
moieties, similar to the case for H which possesses ten
terminal and two bridging carbonyl ligands.[27] Nevertheless, due to the chelate ring in the bis(isonitrile) ligand
the triironcore of 2 shows slightly widened bond angles
of 62.37 and 62.23° for Fe1–Fe2–Fe3 and Fe2–Fe3–Fe1,
respectively, and a more acute angle of 55.41° for Fe2–Fe1–Fe3
in comparison with H (61.6, 61.2, and 57.2°, respectively).
Figure 2
Solid-state molecular structure of the complex
[Fe3(BINC)6] (2). Ellipsoids are
drawn at the 40% probability
level. H atoms and disorder in the phenyl groups are omitted for clarity.
Selected bond lengths (Å) and angles (deg): Fe1–C1 1.839(3),
C1–N1 1.176(4), Fe2–C81 2.014(3), Fe3–C81 1.945(3),
C81–N10 1.235(4), Fe2–C41 1.853(3), C41–N5 1.168(4);
Fe2–C81–Fe3 78.10(1), Fe1–Fe2–C81 89.64(8),
C81–Fe3–Fe1 91.04(8). Fe–Fe distances and Fe–Fe–Fe
angles are given in Table .
Table 1
Selected Bond Lengths (Å) and
Angles (deg) of 2 and [Fe3(CO)12] (H)a
2
[Fe3(CO)12]
Fe1–Fe2
2.682(6)
2.677(2)
Fe2–Fe3
2.495(6)
2.558(1)
Fe3–Fe1
2.685(6)
2.683(1)
Fe1–Fe2–Fe3
62.37(2)
61.6(2)
Fe2–Fe3–Fe1
62.23(2)
61.2(2)
Fe3–Fe1–Fe2
55.41(2)
57.2(2)
The data for 2 were
collected at 123 K, while the data for H were collected
at room temperature.[27]
The data for 2 were
collected at 123 K, while the data for H were collected
at room temperature.[27]Solid-state molecular structure of the n class="Chemical">complex
[Fe3(BINC)6] (2). Ellipsoids are
drawn at the 40% probability
level. H atoms and disorder in the phenyl groups are omitted for clarity.
Selected bond lengths (Å) and angles (deg): Fe1–C1 1.839(3),
C1–N1 1.176(4), Fe2–C81 2.014(3), Fe3–C81 1.945(3),
C81–N10 1.235(4), Fe2–C41 1.853(3), C41–N5 1.168(4);
Fe2–C81–Fe3 78.10(1), Fe1–Fe2–C81 89.64(8),
C81–Fe3–Fe1 91.04(8). Fe–Fe distances and Fe–Fe–Fe
angles are given in Table .
The 31P{n class="Chemical">1H} NMR spectrum of 2 shows only one singlet at δ
13.2 ppm in THF-d8, indicating a fluxional
behavior by fast exchange of
the bridging C≡N units in solution. In agreement with this
analysis, only one 13C{1H} NMR signal is observed
for the isonitrilecarbon atoms at δ 218.7 ppm, which is shifted
downfield in comparison to the corresponding signal of the free BINC
ligand (δ 172.4 ppm). The solid-state infrared spectrum of 2 features a broad band centered at 2035 cm–1 assignable to the isonitrile ν[CN] stretch, shifted to lower
wavenumbers in comparison to the isonitrile ν[CN] stretch in 1 (2122 cm–1). The ν[CN] stretching
frequency for the divalent complex 1 is similar to that
of the free BINC (2126 cm–1), as the isonitrile
functions essentially as a pure donor ligand. In contrast, compound 2, with an iron center in the formally zerovalent oxidation
state, shows ν[CN] stretching frequencies shifted to lower energies,
shortened terminal Fe–C bond lengths (av. 1.83(1) Å vs
1.87(4) Å in 1), and slightly more bent Fe–C–N
angles (172.5(3)° vs 176.1(1)° in 1). When
they are taken together, these data indicate substantial M(dπ)
→ π* back-bonding in 2.[28]
The electrochemical behavior of 2 (see Figure S41 in the Supporting Information) was
analyzed with UV–vis–NIR spectroelectrochemistry (SEC)
in an optically transparent thin-layered electrode (OTTLE) cell.[29] The cyclic voltammogram showed a single partially
reversible oxidation at Epa = −0.6
V versus [FeCp2]0/+, which was used as the internal
standard for all potentials reported herein unless otherwise noted.
The neutral n class="Chemical">complex could be regenerated at Epc = −1.35 V. The SEC for the oxidation process showed
that the bands at 350 and 715 nm were consumed and replaced by a band
at 475 nm with a shoulder at 580 nm. The spectrum of 2 could be regenerated after a cathodic potential of <1.5 V was
reached. The electrochemical behavior of [Fe3(CO)12] in THF was reported to show only irreversible processes.[30]
Other prominent motifs for iron(II)n class="Chemical">complexes
are mononuclear and
dinuclear ironcarbonyl species such as [CpFeCl(CO)2] (I), which was first synthesized by Hubbard,[31] and [CpFe(CO)2]2 (J).[32] In the case of J it was demonstrated
that the CO ligands could be substituted with monodentate isonitrile
ligands, resulting in A (Figure ). Again, we questioned if substitution by
the bidentate isonitrile ligand BINC is also tolerated.
With
[Cp*FeCl(tmeda)][33] (Scheme ; K, n class="Chemical">tmeda = tetramethylethylene-1,2-diamine)
as the starting material, the mononuclear [Cp*FeCl(BINC)] (3, Cp* = Me5C5) could be readily synthesized
in 82% yield and fully characterized by NMR/IR spectroscopy and XRD
(see Figure S35 in the Supporting Information).
The iron atom possesses a distorted-tetrahedral geometry, being η5-coordinated by the Cp* ligand, the two isonitrile moieties
of the BINC ligand, and a chloride ligand. The carbon monoxide analogue I has a similar distorted-tetrahedral geometry in which the
chloride and the two CO ligands are sterically repelled by the η5-coordinated cyclopentadienyl ligand and possess a C–Fe–C
angle of 95.6(2)° and a C–Fe–Cl angle of 90.4(2)°
(vs angles in 3: C1–Fe–C20 88.4(7)°,
C1–Fe–Cl 94.3(6)°, C20–Fe–Cl 91.4(6)°).[34] The 31P{1H} NMR spectrum
of 3 gives rise to a sharp singlet at δ 15.2 ppm.
The IR spectrum shows two broad bands at 2084 and 2023 cm–1, which are assigned to the symmetrical and asymmetrical CN stretching
vibrations.
Scheme 2
Synthesis of [Cp*FeCl(BINC)] (3) and
[Cp*Fe(BINC)]2 (4)
The reduction of 3 in THF at low temperatures with
1 equiv of n class="Chemical">KC8 affords green crystals of 4 in a yield of 18% after workup and crystallization from THF/n-hexane (Scheme ). Single-crystal XRD revealed that 4 crystallizes
as the cis isomer, in which the Cp* groups are both
coordinated above the Fe–Fe bond (Figure ). One C≡N unit of the BINC ligands
links terminally to iron, while the other bridges the iron–iron
bond. As a result of the double bridging isonitrile units there is
an Fe–Fe bond of 2.575 Å. The iron-bound bridging isonitrile
groups result in an almost planar butterfly motif (Fe1–C20–Fe2–C40)
with a torsion angle of 7.82°.
Figure 3
Solid-state molecular structure of the
complex [Cp*Fe(BINC)]2 (4). Ellipsoids are
drawn at the 40% probability
level. H atoms, solvent molecules, and disorder in the Cp* groups
are omitted for clarity. Selected bond lengths (Å) and angles
(deg): Fe1–Fe2 2.575(3), Fe1–C20 1.914(2), Fe2–C20
1.956(2), C20–N2 1.242(2), C21–N3 1.178(2), Fe1–C1
1.787(2), Fe2–C21 1.795(2) C1–N1 1.176(2), Fe1–C40
1.936(2), Fe2–C40 1.919(2), C40–N4 1.246(2); C1–Fe1–Fe2
95.48(3), C1–Fe1–C20 89.88(5), C21–Fe2–Fe1
96.85(5), Fe1–C1–N1 173.08(1), Fe2–C21–N3
173.93(1), C20–Fe1–C40 96.26(7), Fe1–C20–Fe2
83.43(7).
Solid-state molecular structure of the
complex [Cp*Fe(BINC)]2 (4). Ellipsoids are
drawn at the 40% probability
level. H atoms, solvent molecules, and disorder in the Cp* groups
are omitted for clarity. Selected bond lengths (Å) and angles
(deg): Fe1–Fe2 2.575(3), Fe1–C20 1.914(2), Fe2–C20
1.956(2), C20–N2 1.242(2), C21–N3 1.178(2), Fe1–C1
1.787(2), Fe2–C21 1.795(2) C1–N1 1.176(2), Fe1–C40
1.936(2), Fe2–C40 1.919(2), C40–N4 1.246(2); C1–Fe1–Fe2
95.48(3), C1–Fe1–C20 89.88(5), C21–Fe2–Fe1
96.85(5), Fe1–C1–N1 173.08(1), Fe2–C21–N3
173.93(1), C20–Fe1–C40 96.26(7), Fe1–C20–Fe2
83.43(7).The structure of 4 is related to that of the phenyl
n class="Chemical">isonitrile complex trans-[CpFe(CNPh)2]2 (A, see Figure ), which was already reported by Fehlhammer in 1980.[12] In contrast to 4, compound A crystallizes in a trans/anti configuration with a strictly planar Fe2C2core. The Fe–Fe bond in A is 2.523(1) Å,
which is shorter than that in complex 4 (2.575(3) Å).
In both complexes the aryl groups of the bridging isonitrile units
show an anti conformation.
The NMR spectroscopic
measurements indicate 4 is in
a cis/trans isomer equilin class="Chemical">brium in
THF-d8 (Scheme ) that is not affected by temperature changes
(see Figure S16 in the Supporting Information).
Scheme 3
cis/trans Isomer Equilibrium of
Compound 4 in Solution
The ratio of the isomers depends on the isolated form of 4; the crystalline product gives rise to the cis isomer as the main component, while for an amorphous powder the trans isomer predominates (see the Supporting Information for spectra).The occurrence of cis and trans isomers observed for 4 is
also found in the dinuclear
carbonyl analogue n class="Chemical">[CpFe(CO)2]2 (J).[35] X-ray crystal structures for both trans-J[32] and cis-J[36,37] were obtained. The
structure of the cis isomer contains a shortened
iron–iron bond (2.531(2) Å vs compound 4,
Fe1–Fe2 2.575(3) Å; Table ) due to the smaller ligands coordinated to the iron
centers.[36] It is noteworthy that the sterically
more hindered permethylated derivative [Cp*Fe(CO)2]2[38] (L) appears to
exist exclusively as the trans isomer, although the cis configuration has been observed as a short-lived, photogenerated
species.[37] In contrast, 4 exists
both as the cis and trans isomers
in solution. Nevertheless, the structural parameters of A, J, L, and 4 are similar.
This shows that the exchange of the ancillary ligands (Cp vs Cp*,
and CO vs CNPh and BINC) influences the solution behavior, but it
does not strongly affect Fe–C and C–N bond lengths and
Fe–C–N angles.
Table 2
Comparison of Selected
Bond Lengths
(Å) and Angles (deg) of Compound 4, cis-[CpFe(CO)2]2 (J),[36]trans-[Cp*Fe(CO)2]2 (L),[39] and
[CpFe(CNPh)2]2 (A)[12]
4
J
L
A
Fe1–Fe2
2.575(3)
2.531(2)
2.560(1)
2.523(1)
Fe1–Cterminal
1.787(2)
1.730(7)
1.753(3)
1.775(6)
Cterminal–N/O
1.176(2)
1.159(9)
1.154(3)
1.161(8)
Fe1–Cbridged
1.914(2)
1.918(7)
1.936(2)
1.934(6)
Fe2–Cbridged
1.956(2)
1.908(7)
1.936(2)
1.905(6)
Cbridged–N/O
1.242(2)
1.178(9)
1.172(3)
1.242(8)
Cterminal–Fe1–Cbridged
89.88(5)
92.3(3)
95.2(1)
Fe1–Cterminal–N/O
173.08(1)
176.5(8)
175(8)
178.4(4)
The electrochemical behavior of 4 in 0.1 M nBu4NPF6/n class="Chemical">THF showed a partially reversible
oxidation
wave at Epa = −0.7 V versus [FeCp2]+/0 and two irreversible processes at Epa = −0.35 and 0.05 V. SEC of 4 showed that the bands at 600 and 390 nm decreased upon oxidation,
and the process was irreversible (Figure S42 in the Supporting Information). In contrast, the CV of compound 4 in CH2Cl2 (Figure S43 in the Supporting Information) showed more resolved redox
transitions at −0.7, – 0.15, 0.3, and 0.65 V. No reduction
signal was observed. These transitions were partially reversible according
to the UV–vis SEC spectrum, which showed that the absorption
bands at 310 and 370 nm were consumed and a band at 290 nm appeared.
Upon further oxidation the band at 600 nm was consumed and the band
at 290 nm coalesced.
In hopes of isolating isonitrile ferraten class="Chemical">complexes with greater
kinetic shielding, BINC was substituted for the bulkier bidentate
chelating isonitrileCNAr3NC.[7] The addition of 2 equiv of CNAr3NC to a solution of anhydrous
FeBr2 in THF affords an orange suspension, from which [FeBr2(CNAr3NC)2] (5) can be
isolated as orange crystals in 55% yield after crystallization from
CH2Cl2/n-hexane (Scheme ). The solid-state molecular
structure of 5 shows a distorted-octahedral geometry
at the iron center with a cis configuration of the
bromido ligands (Br1–Fe1–Br2 92.12(2)°; see the Supporting Information for a graphical representation
of the structure and further characterization details).
Scheme 4
Synthesis
of [FeBr2(CNAr3NC)2] (5), [Fe(CNAr3NC)2]2 (6), and [K(Et2O)]2[Fe(CNAr3NC)2]2 (7)
Reduction of 5 in THF at low temperature with 2 equiv
of n class="Chemical">KC8 (Scheme ) results in an orange-red solution from which dark red crystals
of [Fe(CNAr3NC)2]2 (6) were isolated in a yield of 12%. The 1H NMR spectra
and two-dimensional experiments (1H,1H-NOESY)
show that compound 6 is isolated in an equilibrium with
one main isomer (cis isomer; see the solid-state
molecular structure in Figure ) and two other species (presumably isomers) with minor populations
(see the Supporting Information for spectra).
Figure 4
Solid-state
molecular structures of [Fe(CNAr3NC)2]2 (6, on the left) and [K(Et2O)]2[Fe(CNAr3NC)2]2 (7,
on the right). Ellipsoids are drawn at the 40%
probability level. H atoms and solvent molecules are omitted for clarity.
Selected bond lengths (Å) and angles (deg) of 6:
Fe1–Fe2 2.434(6), Fe1–C1 1.835(3), C1–N1 1.165(4),
Fe1–C25 1.977(3), Fe2–C25 1.900(3), C25–N3 1.244(4),
Fe1–C24 1.860(3), C24–N2 1.176(4), Fe2–C48 1.800(3),
C48–N4 1.186(4); Fe1–C25–Fe2 77.73(1), C1–Fe1–C24
87.39(1), Fe1–C1–N1 178.3(3). Selected bond lengths
(Å) and angles (deg) of 7: Fe1–Fe2 2.552(8),
Fe1–C1 1.747(5), Fe1–C24 1.838(5), Fe1–C25 1.951(4),
Fe2–C25 1.963(4), Fe2–C48 1.789(4), Fe1–K1 3.439(1),
Fe2–K2 3.410(1), C1–N1 1.214(9), C24–N2 1.176(6),
C25–N3 1.266(5), C48–N4 1.206(5), Fe1–C25–Fe2
81.38(2), Fe1–C1–N1 146.0(6), Fe1–C24–N2
170.7(4), Fe1–C25–N3 142.9(3), Fe2–C25–N3
135.2(3).
Solid-state
molecular structures of [Fe(CNAr3NC)2]2 (6, on the left) and [K(Et2O)]2[Fe(CNAr3NC)2]2 (7,
on the right). Ellipsoids are drawn at the 40%
probability level. H atoms and solvent molecules are omitted for clarity.
Selected bond lengths (Å) and angles (deg) of 6:
Fe1–Fe2 2.434(6), Fe1–C1 1.835(3), C1–N1 1.165(4),
Fe1–C25 1.977(3), Fe2–C25 1.900(3), C25–N3 1.244(4),
Fe1–C24 1.860(3), C24–N2 1.176(4), Fe2–C48 1.800(3),
C48–N4 1.186(4); Fe1–C25–Fe2 77.73(1), C1–Fe1–C24
87.39(1), Fe1–C1–N1 178.3(3). Selected bond lengths
(Å) and angles (deg) of 7: Fe1–Fe2 2.552(8),
Fe1–C1 1.747(5), Fe1–C24 1.838(5), Fe1–C25 1.951(4),
Fe2–C25 1.963(4), Fe2–C48 1.789(4), Fe1–K1 3.439(1),
Fe2–K2 3.410(1), C1–N1 1.214(9), C24–N2 1.176(6),
C25–N3 1.266(5), C48–N4 1.206(5), Fe1–C25–Fe2
81.38(2), Fe1–C1–N1 146.0(6), Fe1–C24–N2
170.7(4), Fe1–C25–N3 142.9(3), Fe2–C25–N3
135.2(3).Single-crystal XRD of 6 confirms the formation of
a dimeric n class="Chemical">iron complex with four CNAr3NC isonitrile ligands,
two of which are bridging the iron–iron bond. Compound 6 features a relatively short metal–metal bond (2.434(6)
Å) because of the bridging isonitrile ligands and assumes a butterfly
structural motif. The Fe1–C25–Fe2–C49 torsion
angle is 17.9°, which indicates the Fe–C–Fe–C
moiety is nearly planar.
A structural arrangement n class="Chemical">comparable
to that of 6 is
the structure of the unsaturated dinuclear iron carbonyl complex [Fe2(CO)8] (M), which was first observed
in 1971 by Poliakoff and Turner as one of the main products from the
low-temperature photolysis of [Fe2(CO)9] conducted
in CO matrixes.[40,41] On the basis of infrared spectroscopic
measurements the CO-bridged and unbridged isomers of M were proposed.[40] The calculated iron–iron
bond length (2.482 Å, Table , BP86/TZVP) of the singlet state of M with C symmetry (global
minimum) matches with the symmetry (two bridged CO units) and bond
lengths of the solid-state molecular structure of 6 (Fe1–Fe2
2.434(6) Å).
Table 3
Comparison of Selected Bond Lengths
(Å) and Angles (deg) (Average Values) of 6 and [Fe2(CO)8] (M[40])a
6
M
Fe–Fe
2.434(6)
2.482
Fe–Cbridged
1.939
1.982
Cbridged–N/O
1.240
1.177
Fe–Cterminal
1.848
1.802
Cterminal–N/O
1.171
1.154
Fe–Cterminal–N/O
174.88
178.00
Optimized singlet
geometry in C symmetry
(BP86/TZVP).
Optimized singlet
geometry in C symmetry
(BP86/TZVP).The heightened
stability and analogous nature of 6 in comparison to M allows for the latter’s stn class="Chemical">ructural
realization, something not possible previously due to its inherent
instability. As expected for CN-bridged 6 with C symmetry, an analysis of
the DFT calculated CN stretching frequencies determined six IR-active
modes (for a comparison of the calculated IR spectra of M and 6, see the Supporting Information). The solid-state IR spectrum of 6 features two specific
ν[CN] bands at 2070 and 2022 cm–1 (broad bands)
for the terminally and bridged coordinated isonitrile units, respectively.
These data are in good agreement with the DFT calculated values (see
the Supporting Information for further
details).
It was previously reported that when monodentate isonitrile
ligands
were used it was not possible to isolate a dinuclear n class="Chemical">iron complex
bearing eight ligands. An example is [Fe2(CNPh)9] (B, Fe–Fe 2.458(1) Å; see Figure ), which is synthesized in
a manner comparable to that of 6 except partial decomposition
must occur to provide an extra isonitrile ligand.[13] This result shows that the use of the sterically demanding
bidentate ligand CNAr3NC is key to the successful stabilization
of 6, allowing for the isolation of a dinuclear ironcomplex with a lower coordination number.
When the reduction
of 5 is repeated with 3 equiv of
finely ground KC8 instead of 2 equiv in n class="Chemical">THF at −80
°C, a brownish orange solution forms from which [K(Et2O)]2[Fe(CNAr3NC)2]2 (7) can be isolated as a black solid in 46% yield by crystallization
from Et2O (Scheme ). The solid-state molecular structure of 7 shows
a motif similar to that for complex 6. The iron–iron
bond is bridged by two isonitrile units, and the other ligands coordinate
terminally to the iron centers. In the solid state, 7 is a contact-ion pair with significant interactions between the
potassium and bridging isonitrile units.
The dinuclear dianionic
complex [n class="Chemical">Fe2(CO)8]2– (N), which was first reported
by Hieber and Brendel,[42] is the carbon
monoxide analogue of compound 7. The solid-state structures
of N feature different coordination modes depending on
the counterion used. Single-crystal XRD of [Na(DMF)3]2[Fe2(CO)8] (Na-N) shows
no bridging CO ligands, explaining why the Fe–Fe bond is significantly
longer (2.804(1) Å) than the metal–metal bond in 7 (Fe–Fe 2.552(8) Å).[43] The solid-state structure of [Li(THF)3]2[Fe2(CO)8] (Li-N),[44] in which the lithium atoms are bonded to the oxygen atoms
of bridging CO groups, shows the same symmetry (C2) and coordination motif as for compound 7, which results in similar bond lengths (Table , 2.551(1) Å vs 2.552(8)
Å for 7). Because of the bridging CO/isonitrile
ligands, complexes Li-N and 7 both assume
a butterfly structural motif and show similar Fe–C–Fe–C
torsion angles of 40.6(3)° for Li-N and 38.5(2)°
for 7.
Table 4
Comparison of Selected
Bond Lengths
(Å) and Angles (deg) of 7, [Li(THF)3]2[Fe2(CO)8] (Li-N),[44] and [Na(DMF)3]2[Fe2(CO)8] (Na-N)[43]
7
[Li-N]
[Na-N]
Fe–Fe
2.552(8)
2.551(1)
2.804(1)
Fe–Cbridged
1.951(4)
1.929(8)
Cbridged–N/O
1.266(6)
1.216(9)
Fe–Cterminal
1.838(5)
1.765(8)
1.781(2)
Cterminal–N/O
1.176(6)
1.164(1)
1.158(2)
Fe–Cterminal–N/O
170.7(4)
177.5(7)
176.8(2)
Conclusion
We have reported the synthesis and characterization of n class="Chemical">isonitrile
analogues of the classic tri- and dinuclear iron carbonyl complexes.
The reduction of the iron(II) precursors trans-[FeBr2(BINC)2] (1), [Cp*FeCl(BINC)] (3) and [FeBr2(CNAr3NC)2]
(5) led to oligonuclear ironcompounds with chelating
bidentate isonitrile ligands. [Fe3(BINC)6] (2) possesses a triangular iron arrangement that is comparable
to the structural motif of [Fe3CO12] and has
been shown to have fluxional behavior with the bridging and terminal
coordinated isonitrile units in solution. The reduction of compound 3 resulted in the dimeric heteroleptic complex [Cp*Fe(BINC)]2 (4) as a cis/trans isomer mixture. The reduction of 5 bearing a bulkier
isonitrile ligand results in [Fe(CNAr3NC)2]2 (6) and [K(Et2O)]2[Fe(CNAr3NC)2]2 (7), both possessing
a butterfly arrangement. In contrast to its CO analogue [Fe2(CO)8], compound 6 is stable and isolable
at room temperature. A detailed comparison of the spectroscopic and
structural parameters of complexes 4–7 to mono(isonitrile) and CO analogues shows that replacing these
monodentate ligands with bidentate isonitrile molecules, such as BINC
and CNAr3NC, does not dramatically affect the molecular
or electronic structures. Thus, future investigations will be aimed
at the synthesis of further highly reduced ferratecomplexes and their
utilization for electro- or photocatalysis, where the stable coordination
of bidentate (chelating) isonitrile ligands might be beneficial.
Experimental Section
General Procedures
All experiments were performed under
an atmosphere of dry argon (n class="Chemical">Argon 4.6, Linde) using standard Schlenk
line techniques, a MBraun UniLab Glovebox, or a GS Glovebox (GS117717).
Solvents were dried and degassed with the MBraun SPS800 solvent-purification
system. THF, diethyl ether, and DCM were stored over molecular sieves
(3 Å). n-Hexane was stored over a potassium
mirror. Deuterated tetrahydrofuran was purchased from Sigma-Aldrich
and used as received.
Elemental analyses were determined by
the analytical department of the University of Regensburg with a Micro
Vario Cube (Elementar).
NMR Spectroscopy
NMR spectra were
recorded onn class="Chemical">Bruker
Avance 300 (300 MHz) and Bruker Avance 400 (400 MHz) spectrometers
at 300 K if not stated otherwise. 1H and 13C{1H} spectra were referenced internally to residual solvent
resonances, while 31P{1H} spectra were referenced
externally to 85% H3PO4(aq). The assignment
of 1H and 13C NMR signals was confirmed by two-dimensional
(1H,1H-COSY, 1H,1H-NOESY, 1H,13C-HSQC, 1H,13C-HMBC)
experiments. The following abbreviations have been used for multiplicities:
s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; dd,
doublet of doublets.
IR Spectroscopy
IR spectra were
recorded using a n class="Chemical">Bruker
ALPHA spectrometer equipped with a diamond ATR unit. The following
abbreviations were used for the intensities and characteristics of
important IR absorption bands: vs, very strong; s, strong; m, medium;
w, weak; br, broad.
Electrochemistry and UV–Vis–NIR
Spectroelectrochemistry
UV–vis–NIR absorption
spectra were recorded on a
J&M TIDAS spectrophotometer. Cyclic voltammetry was n class="Gene">carried out
in 0.1 M Bu4NPF6/THF solutions using a three-electrode
configuration (Pt working, Pt-wire counter, and Ag quasi-reference
electrodes) and was performed on a Metrohm Autolab potentiostat. THF
was distilled from Na/K amalgam and degassed directly before use in
the electrochemistry.
X-ray Crystallography
Single-crystal
X-ray diffraction
data were recorded on an Agilent Technologies SuperNova diffractometer
with Cu Kα radiation (λ = 1.54184 Å). Crystals were
selected under n class="Chemical">mineral oil, mounted on MicroMount loops, and quench-cooled
using an Oxford Cryo-system open flow N2cooling device.
Either semiempirical multiscan absorption corrections[45] or analytical corrections[46] were
applied to the data. Using Olex2,[47] the
structures were solved with the SHELXT[48] structure solution program using intrinsic phasing and refined with
the SHELXL[49] refinement package using least-squares
refinements on F2.[50] The hydrogen atoms were located in idealized positions
and refined isotropically with a riding model. Crystallographic data
for the structures in this paper have been deposited with the Cambridge
Crystallographic Data Centre, CCDC, 12 Union Road, Cambridge CB21EZ,
UK. Copies of the data can be obtained free of charge on quoting the
depository numbers: 2056040 (for 1), 2056041 (for 2), 2056042 (for 3), 2056043 (for 4), 2056044 (for 5), 2056046 (for 6), and 2056045 (for 7).
Starting Materials
The chelating bidentate isonitrile
ligands bis(2-isocyanophenyl) phenylphosphonate (BINC)[6] and 2,2″-diisocyano-3,5,3″,5′′-tetramethyl-1,1′:3′,1″-terphenyl
(CNAr3NC)[7] and [Cp*FeCl(tmeda)][33] were prepared according to reported procedures.
Synthesis of trans-[FeBr2(BINC)2] (1)
A solution of BINC (0.350 g, 0.970
mmol) in n class="Chemical">THF (10 mL) was added
to a solution of FeBr2 (0.100 g, 0.460 mmol) in THF (60
mL) at ambient temperature. The brownish orange reaction mixture turned
into a grayish suspension after 30 min. After the mixture was stirred
overnight, product 1 could be isolated as a blue solid
after decanting off the supernatant solution, washing with THF (3
× 10 mL), and drying in vacuo. Recrystallization
by slow diffusion of n-hexane into a DCM solution
resulted in blue crystals of 1. Yield: 154 mg, 33% (the
calculated yield assumes the presence of 8% FeBr2 in the
sample, vide infra). Anal. Calcd for C40H26Br2FeN4O6P2 (MW = 936.27 g/mol): C, 51.31; H 2.80; N, 5.98. Found: C, 50.27;
H, 2.87; N, 5.77. Elemental analyses consistently gave low carbon
and nitrogen, which indicated the presence of a persistent impurity.
The impurity could not be removed by extraction or crystallization.
We presume the impurity to be unreacted FeBr2. A calculated
C, H, N analysis for 1·0.08FeBr2 (C,
50.39; H, 2.75; N, 5.88) fits well to the values found (C, 50.27;
H, 2.87; N, 5.77). 1H NMR (400.13 MHz, 300 K, CD2Cl2): δ (ppm) 8.31–8.37 (m, 4H, H2), 7.71 (d, 3JHH = 8.5 Hz,
4H, H4), 7.65 (d, 3JHH = 7.9 Hz, 4H, H7), 7.6 (dt, 3JHH = 1.4 Hz, 3JHH = 7.5 Hz, 2H, H1), 7.46–7.50 (m, 2H, H3), 7.40 (dt, 3JHH = 1.6 Hz, 3JHH = 7.7 Hz, 4H, H5), 7.28 (t, 3JHH = 7.7 Hz,
4H, H6). 13C{1H}-NMR (100.61 MHz,
300 K, CD2Cl2) δ (ppm) 147.4 (d, J = 7.0 Hz, C9), 134.6 (d, J = 2.6 Hz, C1), 133.9 (d, J = 11.5 Hz,
C2), 130.8 (s, C5), 129.4 (d, J = 16.4 Hz, C3), 127.6 (s, C7), 126.1 (s, C6), 123.8 (d, J = 190 Hz, C10),
121.9 (d, J = 2.7 Hz, C4), 121.2 (d, J = 5.1 Hz, C8), (C11 ≡ N)
signals were not detected. 31P{1H}-NMR (161.98
MHz, 300 K, CD2Cl2): δ (ppm) 15.5 (s,
2P). IR: ν (cm–1) 2122 (ν[C≡N],
vs), 1585 (w), 1486 (s), 1486 (m), 1439 (m), 1288 (m), 1239 (m), 1130
(s), 1100 (s), 1031 (m), 916 (br s), 799 (s), 755 (s), 713 (m), 596
(s), 563 (s), 531 (vs).
Authors: Charles C Mokhtarzadeh; Grant W Margulieux; Alex E Carpenter; Nils Weidemann; Curtis E Moore; Arnold L Rheingold; Joshua S Figueroa Journal: Inorg Chem Date: 2015-05-12 Impact factor: 5.165
Authors: Myles J Drance; Jeffrey D Sears; Anthony M Mrse; Curtis E Moore; Arnold L Rheingold; Michael L Neidig; Joshua S Figueroa Journal: Science Date: 2019-03-14 Impact factor: 47.728
Authors: Julia Leitl; Michael Marquardt; Peter Coburger; Daniel J Scott; Verena Streitferdt; Ruth M Gschwind; Christian Müller; Robert Wolf Journal: Angew Chem Int Ed Engl Date: 2019-09-13 Impact factor: 15.336