Bis(2-pyridylimino)isoindolato iron(ii) and cobalt(ii) complexes: Structural chemistry and paramagnetic NMR spectroscopy

Article abstract of DOI:10.1039/c1dt10617a

Condensation of phthalodinitrile and 2-amino-5,6,7,8-tetrahydroquinoline gave the bis(2-pyridylimino)isoindole protioligand 1 (thqbpiH) in high yield. Deprotonation of thqbpiH (1) using LDA in THF at -78 °C yields the corresponding lithium complex [Li(THF)(thqbpi)] (2) in which the lithium atom enforces almost planar arrangement of the tridentate ligand, with an additional molecule of THF coordinated to Li. Reaction of cobalt(ii) chloride or iron(ii) chloride with one equivalent of the lithium complex 2 in THF led to formation of the metal complexes [CoCl(THF)(thqbpi)] (3a) and [FeCl(THF)(thqbpi)] (3b). The paramagnetic susceptibility of 3a,b in solution was measured by the Evans method (3a: μ_eff = 4.17 μ_B; 3b: μ_eff = 5.57 μ_B). Stirring a solution of 1 and cobalt(ii) acetate tetrahydrate in methanol yielded the cobalt(ii) complex 4 which was also accessible by treatment of 3a with one equivalent of silver or thallium acetate in DMSO. Whereas 3a,b were found to be mononuclear in the solid state, the acetate complex 4 was found to be dinuclear, the two metal centres being linked by an almost symmetrically bridging acetate. For all transition metal complexes paramagnetic ¹H as well as ¹³C NMR spectra were recorded at variable temperatures. The complete assignment of the paramagnetic NMR spectra was achieved by computation of the spin densities within the complexes using DFT. The proton NMR spectra of 3a and 3b displayed dynamic behaviour. This was attributed to the exchange of coordinating solvent molecules by an associative mechanism which was analysed using lineshape analysis (ΔS ^≠= -154 ± 25 J mol^-1 K^-1 for 3a and ΔS^≠ = -168 ± 15 J mol^-1 K^-1 for 3b).

Full text of DOI:10.1039/c1dt10617a

View Article Online / Journal Homepage / Table of Contents for this issue
This article is published as part of the Dalton Transactions themed issue entitled:
Dalton Transactions 40th Anniversary
Guest Editor Professor Chris Orvig, Editorial Board Chair
University of British Columbia, Canada
Published in issue 40, 2011 of Dalton Transactions
Image reproduced with permission of Shinobu Itoh
Welcome to issue 40 of the 40th volume of Dalton Transactions-40/40! Articles in the issue include:
PERSPECTIVE:
Synthesis and coordination chemistry of macrocyclic ligands featuring NHC donor groups
Peter G. Edwards and F. Ekkehardt Hahn
Dalton Trans., 2011, 10.1039/C1DT10864F
FRONTIER:
The future of metal–organic frameworks
Neil R. Champness
Dalton Trans., 2011, DOI: 10.1039/C1DT11184A
ARTICLES:
Redox reactivity of photogenerated osmium(II) complexes
Jillian L. Dempsey, Jay R. Winkler and Harry B. Gray
Dalton Trans., 2011, DOI: 10.1039/C1DT11138H
Molecular squares, cubes and chains from self-assembly of bis-bidentate bridging ligands with
transition metal dications
Andrew Stephenson and Michael D. Ward
Dalton Trans., 2011, DOI: 10.1039/C1DT10263J
Visit the Dalton Transactions website for more cutting-edge inorganic and organometallic research
www.rsc.org/dalton

Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2011, 40, 10406
www.rsc.org/dalton
PAPER
Bis(2-pyridylimino)isoindolato iron(II) and cobalt(II) complexes: Structural
chemistry and paramagnetic NMR spectroscopy†
Matthias Kruck, D e´ sir e´ e C. Sauer, Markus Enders, Hubert Wadepohl and Lutz H. Gade*
Received 11th April 2011, Accepted 13th May 2011
DOI: 10.1039/c1dt10617a
Condensation of phthalodinitrile and 2-amino-5,6,7,8-tetrahydroquinoline gave the
bis(2-pyridylimino)isoindole protioligand 1 (thqbpiH) in high yield. Deprotonation of thqbpiH (1)
◦
using LDA in THF at -78 C yields the corresponding lithium complex [Li(THF)(thqbpi)] (2) in which
the lithium atom enforces almost planar arrangement of the tridentate ligand, with an additional
molecule of THF coordinated to Li. Reaction of cobalt(II) chloride or iron(II) chloride with one
equivalent of the lithium complex 2 in THF led to formation of the metal complexes
[
CoCl(THF)(thqbpi)] (3a) and [FeCl(THF)(thqbpi)] (3b). The paramagnetic susceptibility of 3a,b in
solution was measured by the Evans method (3a: meff = 4.17 m ; 3b: meff = 5.57 m ). Stirring a solution of
and cobalt(II) acetate tetrahydrate in methanol yielded the cobalt(II) complex 4 which was also
B
B
1
accessible by treatment of 3a with one equivalent of silver or thallium acetate in DMSO. Whereas 3a,b
were found to be mononuclear in the solid state, the acetate complex 4 was found to be dinuclear, the
two metal centres being linked by an almost symmetrically bridging acetate. For all transition metal
1
13
complexes paramagnetic H as well as C NMR spectra were recorded at variable temperatures. The
complete assignment of the paramagnetic NMR spectra was achieved by computation of the spin
densities within the complexes using DFT. The proton NMR spectra of 3a and 3b displayed dynamic
behaviour. This was attributed to the exchange of coordinating solvent molecules by an associative
π
-1
-1
π
mechanism which was analysed using lineshape analysis (DS = -154 ± 25 J mol
K
for 3a and DS =
-
1
-1
-
168 ± 15 J mol K for 3b).
4
Introduction
ligands of this type. They were first reported in the 1950s and their
5
coordination chemistry has been investigated since the 1970s.
During the past decade there has been a renaissance of molecular
catalysis using late 3d transition metals instead of the precious
platinum metals, which has given rise to a range of novel classes
Several examples of applications of these ligands in catalysis have
been reported, including the cobalt-catalysed aerobic oxidation of
6
hydrocarbons, the palladium-catalysed hydrogenation of C
C
1
of catalysts. This has entailed the development of appropriate
7
8
bonds as well as iridium-catalysed epoxidations of alkenes.
Chiral derivatives have been successfully used as stereodirecting
ligands in the iron-catalysed hydrosilylation and the cobalt-
ancillary ligand systems which on the one hand stabilise the
molecular catalyst and on the other hand create a well defined
active space around the metal centre. The key elements of ligand
design for this purpose have been the combination of ligand
architecture (denticity and construction of the ligand backbone)
and the choice of ligating atoms. For the 3d transition metals
nitrogen donor ligands have proved to meet the requirements for
9
catalysed cyclopropanation.
Although bpi complexes of many transition metals are known,
the derivatives of the 3d metals containing only a single bpi
ligand and additional labile ligands, which thus are suitable for
catalysis, are often only poorly characterised. This may be due
to their substitutional lability and paramagnetism which renders
2
the development of active and stable catalysts.
Monoanionic meridionally coordinating tridentate ligands,
frequently referred to as “pincers”, have been extensively stud-
10
in situ characterisation difficult. There have been only a few
II
II
5a–c,6a,b,9
reports on Co mono(bpi) and Fe mono(bpi) complexes,
3
ied in development of novel molecular catalysts. Bis(2-pyridy-
and their characterisation by paramagnetic NMR in solution is
virtually unexplored. The structural features of, particularly, the
iron(II) compounds, recently found to be active hydrosilylation
catalysts, remain to be established. Herein we report the synthesis
as well as structural and detailed NMR spectroscopic charac-
terisation of iron(II) and cobalt(II) complexes containing bis(2-
pyridylimino)isoindolate ligands.
limino)isoindoles (bpi) are highly modular, readily accessible
Anorganisch-Chemisches Institut, Universit a¨ t Heidelberg, Im Neuenheimer
Feld 270, 69120, Heidelberg, (Germany). E-mail: lutz.gade@uni-hd.de;
Fax: +49-6221-545609
†
CCDC reference numbers 821499–821503. For crystallographic data in
CIF or other electronic format see DOI: 10.1039/c1dt10617a
1
0406 | Dalton Trans., 2011, 40, 10406–10415
This journal is © The Royal Society of Chemistry 2011

Results and discussion
Synthesis of the ligand and its lithium salt
Since the syntheses of the desired mono(bpi) complexes using
established bpiH derivatives have proved to be cumbersome due
to the formation of [M(bpi) ] complexes, the sterically more
2
demanding ligand 1 (thqbpiH) was synthesised, for which the
formation of “homoleptic” complexes is precluded (Scheme 1).
The C2v symmetry of the ligand facilitates the spectroscopic
assignments for the in situ characterisation in solution. Moreover,
it may be viewed as an achiral analogue to the known chiral ligands
9
featuring similar steric and electronic properties.
ThqbpiH (1) was prepared by a modified procedure originally
Fig. 1 Molecular structure of 1. Hydrogen atoms, except N(1)H, C(16)H2
and C(25)H
, have been omitted for clarity. Selected bond lengths ( A˚ )
and angles ( ): C(1)–N(1) 1.3877(12), C(1)–N(2) 1.2840(12), C(8)–N(4)
1.2912(12), C(8)–N(1) 1.3867(12), C(9)–N(2) 1.4004(12), C(18)–N(4)
1.4024(12), N(2)–C(1)–N(1) 130.94(8), N(4)–C(8)–N(1) 130.45(8).
5
a,b
published by Siegl starting from phthalodinitrile and 2-amino-
,6,7,8-tetrahydroquinoline which were refluxed in n-hexanol with
2
◦
5
11
catalytic amounts of calcium chloride (Scheme 1). The analytical,
1
13
H and C NMR and the high resolution mass spectrometric data
are consistent with the proposed molecular structure, which was
further confirmed by X-ray diffraction.
ThqbpiH (1) displays approximate molecular C symmetry in
2
Common complexation procedures, based on the direct reac-
tion of the protioligand with different metal salts, which were
the solid state, with the rotational axis passing through the N(1)H
bond (Fig. 1). The isoindole backbone is essentially planar, while
5a–c,6a–b
previously reported for bpi ligands,
did not readily lead
to the formation of the desired mono(bpi) iron(II) or cobalt(II)
complexes. Therefore a two step route was used, involving the
◦
◦
the tetrahydroquinoline units are tilted by 12.1 and -20.3 out
of the molecular plane defined by the isoindole unit. This effect
is presumably induced by steric repulsion between the methylene
◦
deprotonation of thqbpiH (1) using LDA in THF at -78 C to yield
the corresponding lithium complex 2. Formation of the lithium
complex was monitored by disappearance of the NH signal using
NMR spectroscopy. Its formulation and structure were confirmed
by high resolution mass spectrometry, elemental analysis and
X-ray structure analysis.
groups C(16)H
2
and C(25)H . Evidently the protonated ligand
2
is already highly pre-organised, which is a favourable feature
for metal complexation. The pre-organisation is caused by the
delocalized p-electron system, which favours a planar arrangement
of the rings, and is further supported by hydrogen bonds formed
between the isoindole proton and the pyridine-nitrogen atoms.
Crystals suitable for X-ray diffraction analysis were obtained
from a saturated solution of the lithium complex 2 in THF.
Scheme 1 Preparation of the compounds 1–4.
This journal is © The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 10406–10415 | 10407

Substitution of the NH proton by a lithium atom forces both
spectra as well as elemental analysis. The structural details of both
complexes were established by X-ray diffraction.
Both complexes 3a and 3b were crystallised from a saturated
tetrahydroquinoline rings into the isoindole plane resulting in
◦
◦
reduced interplanar angles (6.1 and 13.2 ) and coordination
of the metal centre by three nitrogen atoms. Additionally one
molecule of THF is found in the coordination sphere, which is
best described as distorted tetrahedral. The molecule possesses
solution in DMSO. Their C
s
-symmetric molecular structures
are almost superimposable with regard to their key features.
12
6
According to Hoffmann et al. the preferred geometry for d
7
approximate C
flexible cyclohexene subunits of the tetrahydroquinoline moieties
Fig. 2).
s
symmetry neglecting the deviation caused by the
and d high spin pentacoordinate metal complexes is a trigonal
bipyramid. The observed distortion of the complex structures is
presumably dominated by the steric demand of the rigid bpi ligand,
which also disfavours the coordination of an additional ligand to
form stable octahedral complexes.
(
In both cases the aromatic parts of the tetrahydroquinoline units
are almost coplanar to the isoindole plane. The angles between the
◦
◦
isoindole plane and the pyridine rings are 6.0 and 10.3 in case
◦
◦
of the iron complex as well as 3.1 and 11.4 with respect to
the cobalt complex and thus within the same order of magnitude
as the ones observed in the lithium complex 2. As expected, the
bond between the metal and the anionic nitrogen atom is about
˚
0
.17 A shorter than to the neutral ligating nitrogen atoms. The
two remaining coordination sites are occupied by a chloro ligand
and a solvent molecule, in this case DMSO. Corresponding bond
lengths and angles of both derivatives are similar, keeping the
slightly smaller ionic radius of cobalt(II) in mind. The bpi ligand
itself is virtually invariant upon substituting cobalt(II) for iron(II)
regarding its structural features. For this reason the larger iron
Fig. 2 Molecular structure of 2. Hydrogen atoms have been omit-
◦
ted for clarity. Selected bond lengths ( A˚ ) and angles ( ): N(1)–Li(1)
˚
atom is forced out of the isoindole plane by 0.2 A. This also leads
1
.896(2), N(3)–Li(1) 2.125(2), N(5)–Li(1) 2.137(2), O(1)–Li(1) 2.049(2),
to increased distortion of the coordination geometry. The angle
between the anionic nitrogen atom and the coordinating solvent
N(1)–C(1) 1.3699(14), N(2)–C(1) 1.3034(14), N(2)–C(9) 1.3920(14),
N(1)–Li(1)–O(1) 112.58(11), O(1)–Li(1)–N(3) 97.36(9), O(1)–Li(1)–N(5)
◦
molecule in 3b N(1)–Fe(1)–O(1) of 97.7 is significantly smaller
9
8.97(9), N(3)–Li(1)–N(5) 162.37(12).
◦
than the corresponding angle N(1)–Co(1)–O(1) in 3a (106.3 )
(
Fig. 3).
Since most reactions catalysed by cobalt(II) bpi systems depend
Synthesis and structural characterisation of the cobalt(II) and
iron(II) complexes
on the application of acetato complexes their structures both in
the solid state and in solution are of great interest for a mechanistic
understanding of the catalytic reaction. Stirring a solution of 1 and
cobalt(II) acetate tetrahydrate in methanol yielded the cobalt(II)
complex 4. The acetato complex 4 is alternatively accessible by
treatment of 3a with one equivalent of silver or thallium acetate
in DMSO.
Reaction of cobalt(II) chloride or iron(II) chloride with one
equivalent of the lithium complex 2 in THF led to formation
of the metal complexes 3a and 3b (Scheme 1). Their formulation
as depicted in Scheme 1 was confirmed by the observation of
+
+
the molecular ion peaks [M+H] and [M-Cl] in their FAB mass
Fig. 3 Molecular structures of 3a (left) and 3b (right). Hydrogen atoms, non-coordinating solvent molecules and disordered atoms have been
◦
omitted for clarity. Selected bond lengths ( A˚ ) and angles ( ) for 3a: Co(1)–N(1) 1.941(2), Co(1)–O(1) 2.1015(18), Co(1)–N(5) 2.148(2), Co(1)–N(3)
2
1
2
1
.173(2), Co(1)–Cl(1) 2.3327(11), N(1)–C(1) 1.386(3), N(2)–C(1) 1.292(3), N(1)–Co(1)–O(1) 106.28(8), O(1)–Co(1)–N(3) 87.39(7), N(5)–Co(1)–N(3)
72.98(7), N(1)–Co(1)–Cl(1) 111.54(6), O(1)–Co(1)–Cl(1) 142.13(5); 3b: Fe(1)–N(1) 1.9955(13), Fe(1)–O(1) 2.1405(12), Fe(1)–N(3) 2.1667(13), Fe(1)–N(5)
.1718(13), Fe(1)–Cl(1) 2.3407(9), N(1)–C(1) 1.3752(17), N(2)–C(1) 1.2994(18), N(1)–Fe(1)–O(1) 97.65(5), O(1)–Fe(1)–N(3) 83.09(5), N(3)–Fe(1)–N(5)
68.37(4), N(1)–Fe(1)–Cl(1), 112.75(4), O(1)–Fe(1)–Cl(1) 149.51(3).
1
0408 | Dalton Trans., 2011, 40, 10406–10415
This journal is © The Royal Society of Chemistry 2011

Whereas the NMR spectroscopic data supported the formu-
Paramagnetic NMR spectroscopy of the iron and cobalt complexes
lation of 4a as given in Scheme 1, the FAB mass spectrum only
displayed the fragment peak of [M-OAc] as highest m/z ion peak.
1
13
+
For all complexes paramagnetic H as well as C NMR spectra
were recorded at variable temperatures. The cobalt(II) complexes
give rise to well resolved paramagnetic NMR spectra. Due to
X-ray diffraction analysis revealed dinuclear molecules in the solid
state. The two cobalt atoms each bear a bpi ligand and are linked by
a bridging acetate. Furthermore, one cobalt atom is coordinated
by an additional acetato unit, while the second bears a molecule
of methanol, which was employed as the solvent for synthesis
as well as crystallisation. The observed coordination geometry
is approximately distorted trigonal bipyramidal with the angles
7
6
the rapid electron relaxation of the high spin d or d systems
relatively sharp resonances are expected. At 295 K line widths
in the H NMR in the range of 200 to 490 Hz for the
iron system and 100 to 600 Hz for the cobalt complexes are
observed.
13
1
◦
The theory of the unpaired electron-nucleus interaction and its
consequences for NMR spectroscopy has been developed over
between the bridging acetato anion and methanol [143.39(7) ] or
◦
the terminal acetato ligand [139.19(7) ] being opened up. Each
13
the past decades and has been summarised comprehensively. In
diamagnetic molecules the orbital shift (dorb) provides the principal
contribution to the observed chemical shifts. In paramagnetic
samples the temperature dependent hyperfine shift (dhf ) adds to
the orbital shift, leading to the observed chemical shift:
metal atom is located within a plane defined by the particular
isoindole backbone (Fig. 4).
As anticipated, the bond between the negatively charged
˚
˚
isoindolato nitrogen and the cobalt atom (1.94 A) is 0.21 A
˚
shorter than the average of the pyridine nitrogen atoms (2.15 A)
for both cobalt atoms. The bridging acetato ligand is almost
symmetrically bonded to both cobalt atoms (bond lengths Co(1)–
d
obs = dorb + dhf
(1)
˚
O(4) 2.02 and Co(2)–O(3) 2.04 A), which is also reflected in the
The hyperfine shift itself may be expressed as stated in eqn
2), where S is the total electron spin, b the Bohr magneton, g
the nuclear gyromagnetic ratio, finally g and A are the g- and
equal distance between the carboxyl carbon and both oxygen
atoms. In contrast to the only other cobalt-acetato-bpi structure
the complex reported in the literature does not crystallise in a
polymeric form. This is probably caused by the steric demand
of this pincer ligand. Nevertheless this complex is only sparingly
soluble in organic solvents with the exception of methanol and
DMSO.
(
e
N
A-hyperfine tensors.
5f
S(S +1)be
dhf
=
gA
(2)
3kTg N
Fig. 4 Molecular structure of 4. Hydrogen atoms, non-coordinating solvent molecules and disordered atoms have been omitted for clarity.
◦
Selected bond lengths ( A˚ ) and angles ( ): Co(1)–N(1) 1.9405(19), Co(1)–O(3) 2.0378(18), Co(1)–O(1) 2.0412(19), Co(1)–N(3) 2.141(2), Co(1)–N(5)
2
.1628(19), C(77)–O(3) 1.256(3), C(77)–O(4) 1.262(3), Co(2)–N(51) 1.939(2), Co(2)–O(4) 2.0188(19), Co(2)–O(5) 2.0774(18), Co(2)–N(53) 2.119(2),
Co(2)–N(55) 2.171(2), N(1)–Co(1)–O(3) 109.64(7), O(3)–Co(1)–O(1) 139.19(7), N(1)–Co(1)–N(3) 90.22(8), O(3)–Co(1)–N(3) 89.57(7), N(1)–Co(1)–N(5)
9.84(8), O(1)–Co(1)–N(5) 90.64(7), N(3)–Co(1)–N(5) 178.77(7), N(51)–Co(2)–O(4) 112.37(7), O(4)–Co(2)–O(5) 143.39(7), N(51)–Co(2)–N(53) 91.37(8),
O(4)–Co(2)–N(53) 92.16(7), N(51)–Co(2)–N(55) 89.27(8), O(5)–Co(2)–N(55) 83.06(7), N(53)–Co(2)–N(55) 175.78(7).
8
This journal is © The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 10406–10415 | 10409

For an arbitrary number of unpaired electrons a general
expression of the NMR shielding tensor has been developed only
14a
recently. Although there are several isotropic and anisotropic
contributions to dhf , it is possible to analyse paramagnetic NMR
spectra in solution by considering only three contributions, namely
the orbital shift (dorb), the Fermi contact shift (dcon) and the
pseudocontact shift (dpc). Consequently eqn (1) may be simplified
to eqn (3):
~
d
obs = dorb + dhf = dorb + dcon + dpc
(3)
Fig. 5 Labelling of the atoms used in discussion of the paramagnetic
NMR spectra.
The Fermi contact shift is caused by coupling of the unpaired
electrons with the atomic nuclei and is transmitted through
chemical bonds. dcon is proportional to the residual spin density
at the atom centre (rab). The latter can be extracted from DFT
calculations so that dcon can be obtained by using eqn (4):
2
2
m m g (S +1)
0
B
e
dcon
=
⋅
rab
(4)
9
k
T
The pseudocontact shift arises from a dipolar through-space
interaction between the magnetic moment of unpaired electrons
and the magnetic moments of the nuclei. It depends strongly on the
distance of the examined nucleus to the paramagnetic centre and
may be of little influence to the chemical shift of nuclei distant
from the metal core. Thus the Fermi contact is presumed to
dominate the chemical shifts of covalently bound ligands in 3d
15,16
metal complexes.
In order to allow the assignment of the paramagnetic NMR
1
3
Fig. 6 Plot of the measured paramagnetic C NMR chemical shift
against the corresponding sum of the calculated hyperfine shift and the
orbital contribution for all carbon atoms of 3a; the approximate linear
relation was used to assign the resonances. The expanded sections of
17
spectra, DFT calculations were performed using the B3LYP
18
functional and employing a TZVP basis set for the metal atoms as
19
well as a 6-311G(d,p) basis set for all other atoms. This approach
the spectrum illustrate the multiplicity of some of the signals. (295 K,
14–16
reliably predicts hyperfine shifts as has been shown recently.
13
6
00 MHz, DMSO-d
6
). For the numbering of the C-nuclei, see Fig. 5.
The computed spin densities were employed to calculate the
hyperfine shifts caused by the Fermi contact interaction using
eqn (4). The chemical shifts of the protio ligand 1 were used to
determine the experimental orbital terms. These values deviate
only slightly from the chemical shifts of the lithium complex 2,
which may also serve as diamagnetic model.
As expected for the C
observed in the C NMR spectrum in the range of -250 to
1000 ppm. The aromatic carbons can be identified by their
CH-coupling to one proton giving rise to doublets in the H-
s
-symmetric complex 13 signals are
1
3
1
1
J
coupled spectra. Analogously, the CH carbon atoms appear as
2
1
3
The use of a cryogenically cooled C detection probe and
optimised repetition times allowed recording of paramagnetic
triplets. Secondary and tertiary carbon atoms were assigned to
the corresponding protons by correlation experiments, whilst the
1
3
13
1
C NMR spectra within a few minutes. The assignment of the
remaining quarternary C nuclei give rise to signals without JCH-
1
3
carbon resonances is based upon a series of C NMR experiments
with selective H-decoupling, correlation experiments and DFT
coupling to protons.
1
Unpaired electrons of the cobalt core are delocalised into
molecular orbitals also located on the ligand. This is reflected
in large calculated values for the spin density at certain atoms
calculations.
The cobalt(II) chloro complex 3a displays eleven proton res-
onances at 295 K which arise from eleven of the twelve non-
20
of the bpi system. The highest values are predicted for the C3
and C4 atoms, which are part of the isoindole-subunit (Fig. 6).
Likewise the pyridine carbon atoms also bear a high amount of
spin density. The atoms C5, C7 and C13, which are connected to
the cobalt core by an even number of bonds, show negative Fermi
contact contributions, while C6 and C8 exhibit positive values and
are linked to the metal centre by an odd number of bonds. The
aromatic carbon atoms C1 and C2 are only slightly influenced by
the unpaired electrons. Within the aliphatic carbon atoms (C9–
C12) C12 is separated from the Co atom by the smallest number
of bonds and consequently bears the highest spin density.
equivalent H atoms of a C
. One signal was not detected, presumably due to its close vicinity
to the paramagnetic metal centre and a resulting substantial
s
-symmetric molecule as depicted in Fig.
5
dipolar relaxation (r- dependence).
6
13
Four resonances can be easily assigned to aromatic protons,
1
3
1
as a C- H correlation experiment shows their direct connection
1
to carbon atoms, which give rise to doublets in the H-coupled
1
3
C NMR spectra (Fig. 6). These signals display Curie behaviour
with increased temperature, while the seven resonances of the
aliphatic protons start to broaden. Coalescence is reached between
3
30 and 340 K. Above that temperature four averaged signals
The calculated carbon shifts are in good agreement with the
observed chemical shifts and enable a reasonable assignment of the
resonances. However, some of the theoretically determined proton
were detected. No signals of coordinating solvent molecules were
detectable.
1
0410 | Dalton Trans., 2011, 40, 10406–10415
This journal is © The Royal Society of Chemistry 2011

1
shifts deviate from the measured values. This discrepancy may be
caused by a fast ring inversion in the cyclohexene unit. Whereas
the calculated spin-densities are based on a single energy minimum
of the fluxional molecule, the experimental NMR values are a
result of several conformational contributions. Coordination or
decoordination of the coordinating solvent molecule dramatically
influences the calculated proton shifts and leads to a much worse
As anticipated the cobalt acetato complex 4a shows a H
NMR spectrum similar to the one of the chloro complex 3a with
almost identical chemical shifts. The coalescence temperature of
the dynamic behaviour for this compound is observed at about
310 K. An additional signal can be found at 39.27 ppm which may
be assigned to the acetato group.
In the C NMR spectra all of these metal complexes display
Curie behaviour in solution over the studied temperature range
of 290 to 340 K. The Curie plot of the proton chemical shifts
illustrates the observed dynamic behaviour and shows the expected
linear correlation between chemical shifts and inverse temperature
(Fig. 7). Spectrometer and solvent limitations prohibited the
examination of a broader temperature interval.
1
3
1
correlation with the observed H NMR spectrum. Although
the lability of coordinating solvent molecules is already known
for similar cobalt(II) bpi complexes the NMR spectra are not
in accordance with a significant contribution of a fourfold
5f
coordinated species in solution.
The isostructural iron(II) complex 3b displays similar behaviour
to 3a. For four resonances of the aromatic protons a normal,
Curie-type temperature dependence was observed. These signals
are assigned according to their computed spin densities. Compared
to the protons of the pyridine rings, the isoindole protons carry
only small spin densities and thus give rise to signals with only
small paramagnetic shifts. The two protons of the pyridine units
are shifted more strongly, as they are closer to the paramagnetic
centre. Of the eight aliphatic protons only six are detectable at
Finally, the paramagnetic susceptibility of the chloro complexes
in solution was measured by the Evans method. The magnetic
moment of 3a is comparable to that of previously reported high
21
10c
spin cobalt(II) bpi complexes (meff = 4.17 m
B
) and the one of 3b is
4a
in the range of high spin iron(II) bpi complexes (meff = 5.57 m
B
).
The rather large value of the iron complex may be attributed to
pronounced spin orbit coupling, which is commonly observed for
22
distorted iron(II) complexes.
2
7
95 K. All of these signals show coalescence at about 320 K (Fig.
). Again these signals are assigned according to their computed
Analysis of the dynamic behaviour
spin densities, which are decreasing with the number of bonds to
the paramagnetic centre. The proton signals of the CH -group 12
are not detectable, presumably due to a strong dipolar interaction
with the spatial close iron core. Additionally there is one equivalent
of uncoordinated THF present, which was substituted by the
1
2
The H NMR spectra of 3a and 3b display dynamic behaviour
of some resonances of the complexes at temperatures above
3
00 K. The corresponding pairs of signals of the diastereotopic
CH proton atoms show broadening, coalescence and collapse to
2
solvent DMSO-d .
6
singlets. The other NMR signals are not affected by varying the
13
1
Fig. 7 Curie plot showing the linear dependence of C resonance shifts on the inverse temperature for 3a (left). The H spectra of 3b (right) display
Curie behaviour in the fast and slow exchange region and a deviation close to coalescence temperature caused by “pseudorotation” of the Cl-M-DMSO
unit. (The signal of H10¢ is superposed by the solvent signal at lower temperatures.).
This journal is © The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 10406–10415 | 10411

temperature (apart from the Curie shift and typical line shape
variations due to temperature dependent relaxation behaviour).
sixfold coordinated transition state or high energy intermediate
(Scheme 2).
Considering a fast cyclohexene ring inversion, each CH group has
2
one H atom pointing to the same side of the molecular plane as the
chloro ligand and the other H atom to the side of the coordinated
solvent molecule. Interconversion of the diastereotopic positions
has to occur along with a pseudo-rotation of the Cl-M-solvent
moiety. An exchange of this type can be realised either by
dissociation of the solvent or by an associative mechanism in which
a second dimethylsulfoxide molecule enters the coordination
sphere to form a sixfold coordinated transition state (or high
energy intermediate). We have performed line shape analysis of the
Conclusions
A reliable and convenient synthetic route to mono-bpi cobalt(II)
and iron(II) complexes has been developed, and for the first
time iron complexes of this type, having gained considerable
9
importance in catalysis, have been structurally characterised.
Moreover the structural features of all metal complexes in solution
have been determined using paramagnetic NMR spectroscopy in
combination with computational methods. These data support
some of the previous proposals about the structure of mono-
bpi metal complexes and provide new insight into their geometry
and their dynamic behaviour. The analysis of the paramagnetic
and dynamic NMR spectra shows that associative exchange
mechanisms are operative which is a valuable information for the
understanding of catalytic processes with such systems. Based on
these results the methods employed in this work appear to be
suitable for the in situ characterisation of the structurally more
1
dynamic H NMR spectra (Fig. 8) in order to extract activation
entropies from the temperature dependence of the activation Gibbs
‡
energies (DG ). For modelling the line shape behaviour of a two-site
exchange only three parameters are necessary: the chemical shifts
of the exchanging signal pairs at various temperatures, the line
width in the absence of exchange and the exchange rate itself. The
temperature dependent shifts of the signal pairs in and above the
coalescence region were calculated by linear extrapolation against
-
1
T
(Curie behaviour).
The line shape was derived from a similar non-exchanging
9
complicated chiral derivatives.
signal so that the exchange rates as the third parameters could
be extracted. Extrapolation within an Eyring plot leads to
Experimental
π
-1
strongly negative activation entropies (DS = -154 ± 25 J mol
-
1
for 3a and DSπ = -168 ± 15 J mol-1
-1
General experimental procedures
K
K
for 3b). These
values are in accordance with an associative exchange mechanism
for the rearrangement of the Cl-M-DMSO unit through a
All manipulations of air and moisture-sensitive materials were
performed under an inert atmosphere of dry argon using standard
2
Fig. 8 Results of the line shape analysis of 3b. Section corresponding to CH (11) of the experimental spectrum (right) compared to the simulated one
(
left).
Scheme 2 Inversion on the metal centre via associative mechanism. On the NMR-timescale rapid coordination and elimination of a second DMSO
molecule leads to averaged signals of the diastereotopic protons.
1
0412 | Dalton Trans., 2011, 40, 10406–10415
This journal is © The Royal Society of Chemistry 2011

Schlenk techniques. Solvents were purified and dried by standard
and the reaction mixture was stirred for 45 min. THF was removed
in vacuo. The remaining solid was washed with hexane to yield
[Li(thqbpi)] (2) as a yellow solid (230 mg, 98%). Crystals suitable
for an X-ray diffraction study were obtained by slow diffusion of
23
methods. 2-(2-Cyanoethyl)cyclohexanone oxime was synthesised
according to reported procedures. All other reagents were obtained
1
13
commercially and used as received. H and C spectra were
recorded on Bruker Avance II 400 and Bruker Avance III
hexane into a saturated solution of 2 in THF. Found: C, 75.6; H,
1
6
00 NMR spectrometers and were referenced using the residual
6.2; N, 16.9. Calc. for C26
NMR (600 MHz, THF-d
H
24LiN
5
: C, 75.5; H, 5.85; N, 16.9. H
1
13
protio solvent ( H) or solvent ( C) resonances. The paramagnetic
8
, 295 K) d 7.88–7.87 (m, 2H, H1), 7.45–
7.44 (m, 2H, H2), 7.44 (d, J = 7.8 Hz, 2H, H6), 7.22 (d, J = 7.8
1
3
3
3
C NMR spectra were measured with a Bruker QNP Cryo Probe.
3
3
Mass spectrometry and elemental analyses were obtained by the
analytical service of the Heidelberg Chemistry Department.
Hz, 2H, H5), 3.12 (t, J = 6.2 Hz, 4H, H10), 2.82 (t, J = 6.2 Hz, 4H,
13 1
H9), 2.06–2.02 (m, 4H, H11), 1.93–1.89 (m, 4H, H10). C { H}
NMR (150 MHz, THF-d , 295 K) d 167.9 (C5), 161.8 (C13), 154.8
(C4), 143.6 (C7), 140.0 (C3), 130.2 (C1), 126.7 (C8), 124.4 (C2),
8
Preparation of the compounds
7
1
1
22.4 (C6), 35.3 (C12), 29.7 (C9), 24.8 (C11), 24.4 (C10). Li { H}
2
-Amino-5,6,7,8-tetrahydroquinoline. In a round bottom flask
equipped with a condenser 2-(2-cyanoethyl)cyclohexanone oxime
56.1 g, 338 mmol) was dissolved in acetic anhydride (57 ml,
06 mmol) while being cooled using an ice bath. Acetyl chlo-
NMR (155 MHz, THF-d , 295 K) d 2.13. HR-MS (FAB+) m/z
8
+
+
(
(
%): Calc. for C26
100).
H
25LiN
5
([M+H] ) 414.2270; Found 414.2228
(
5
[
CoCl(THF)(thqbpi)] (3a). A solution of LDA (28.9 mg,
ride (59 ml, 676 mmol) was added and the red solution was
refluxed for 16 h. The resulting mixture was neutralised with
saturated, aqueous sodium bicarbonate and extracted four times
with dichloromethane (300 ml). The combined organic layers
were concentrated in a rotary evaporator to yield crude 2-
0
.27 mmol) in THF (3 ml) was added to a solution of the ligand
◦
(
100 mg, 0.25 mmol) in THF (5 ml) at -78 C. After the addition
was completed the solution was warmed to room temperature
and stirred for another 30 min. Then a solution of CoCl
·THF
58.1 mg, 0.27 mmol) in THF (4 ml) was added dropwise
via cannula. The solution was stirred at room temperature for
8 h. Then the solvent was filtered off and dried in vacuo.
2
(
(
N-acetylamino)-5,6,7,8-tetrahydroquinoline as an orange solid,
which was used without further purification in the next step.
After addition of NaOH (85.0 g, 2125 mmol) in water (350 ml),
the resulting mixture was refluxed for three hours. The obtained
solution was extracted three times with dichloromethane (500 ml).
1
After subsequent washing with THF the desired product was
obtained as a dark green powder (99 mg, 80%). Crystals suitable
for an X-ray diffraction study were obtained from a saturated
(
Phase separation occurs only very slowly.) The combined organic
phases were dried over Na SO and concentrated in a rotary
evaporator. The crude product was purified using flash column
chromatography (SiO , Et O/NEt = 97 : 3, R 0.2) to yield 2-
amino-5,6,7,8-tetrahydroquinoline as an orange, crystalline solid
9.4 g, 20% over two steps). The recorded spectroscopic data are
solution of 3a in DMSO. Found: C, 57.3; H, 5.1; N, 11.8. Calc.
2
4
1
for C28
H
30ClCoN
5
OS: C, 57.1; H, 5.2; N, 12.1. H NMR (600
MHz, DMSO-d
6
, 295 K, paramagnetic) d 12.1 (2H, H6), 11.0
2
2
3
f
(
2H, H1), 9.8 (2H, H9), 8.1 (2H, H9¢), 7.9 (2H, H2), -3.2 (2H,
H7), -4.4 (2H, H10), -5.4 (2H, H11), -5.6 (2H, H10¢), -8.2
(2H, H11¢), -42.9 (2H, H12); The second resonance of H12 is
(
23
in good agreement with the reported data. Found: C, 72.4; H,
1
3
1
not detected. C { H} NMR (150 MHz, DMSO-d
paramagnetic) d 994.6 (C3), 627.0 (C4), 578.8 (C6), 526.1 (C8),
60.7 (C12), 213.3 (C1), 208.3 (C2), 90.8 (C11), 84.0 (C9), 45.0
6
, 295 K,
8
.2; N, 18.4. Calc. for C
9
H
12
2
N : C, 72.9; H, 8.2; N 18.9.
ThqbpiH (1). A suspension of phthalodinitrile (3.3 g,
4
2
6 mmol), 2-amino-5,6,7,8-tetrahydroquinoline (9.0 g, 60 mmol)
(
C10), 14.7 (C7), -39.0 (C13), -245.9 (C5). HR-MS (FAB+)
+
+
and CaCl
2
(0.3 g, 3 mmol) in n-hexanol (100 ml) was refluxed
m/z (%): Calc. for C26
5
4
H
25ClCoN
5
([M+H] ) 501.1131; Found
([M-Cl] ) 465.1364; Found
◦
+
+
at 160 C for 40 h. After cooling to room temperature, hexanol
was removed in vacuo. The resulting black solid was dissolved
in dichloromethane, filtered over celite and concentrated in a
rotary evaporator. The crude product was recrystallised from
EtOH/MeOH = 1 : 1 to yield 1 as yellow crystals (8.6 g, 81%).
Crystals suitable for an X-ray diffraction study were obtained
01.1135 (32); Calc. for C26
H
24CoN
5
65.1340 (100). Magnetic susceptibility in DMSO-d (295 K):
6
m = 4.17 m
B
.
[
FeCl(THF)(thqbpi)] (3b). A solution of LDA (290 mg,
2
.75 mmol) in THF (20 ml) was added to a solution of thqbpiH
◦
(1020 mg, 2.50 mmol) in THF (75 ml) at -78 C. The cooling
from a saturated solution of 1 in CH
N, 16.95; Calc. for C26 : C, 76.6; H, 6.2; N, 17.2. H NMR
600 MHz, CD Cl , 295 K) d 12.37 (s, 1H, NH), 7.99–7.06 (m, 2H,
H1), 7.64–7.61 (m, 2H, H2), 7.41 (d, J = 8.3 Hz, 2H, H6), 7.05
2
Cl . Found: C, 76.1; H, 6.2;
2
1
bath was removed and the yellow solution was stirred for 60 min.
Subsequently the yellow reaction mixture was transferred to a
suspension of anhydrous FeCl (350 mg, 2.75 mmol) in THF (75
ml). The reaction mixture was stirred at room temperature for
18 h. The precipitating brown solid was filtered off, washed with
THF (10 ml) and dried in vacuo to yield [FeCl(THF)(thqbpi)] (3b)
as a brown solid (1060 mg, 74%). Crystals suitable for an X-ray
diffraction study were obtained from a saturated solution of 3b in
DMSO. Found: C, 63.2; H, 5.6; N, 12.4. Calc. for C30
C, 63.2; H, 5.7; N, 12.3. H NMR (600 MHz, DMSO-d
paramagnetic) d 41.7 (2H, H6), 14.7 (2H, H9), 11.4 (2H, H9¢), 6.2
(2H, H1), 5.4 (2H, H2), 4.1 (2H, H10), 0.1 (2H, H11), -0.35 (2H,
H7), -3.5 (2H, H11¢); The second resonance of H10 is superposed
by the solvent signal, while H12 are not detected. C { H} NMR
H
25
N
5
(
2
2
3
2
3
3
(
d, J = 8.3 Hz, 2H, H5), 2.81 (t, J = 6.4 Hz, 4H, H12), 2.76 (t,
J = 6.4 Hz, 4H, H9), 1.89–1.85 (m, 4H, H11), 1.83–1.79 (m, 4H,
3
1
3
1
H10). C { H} NMR (150 MHz, CD
56.1 (C13), 152.7 (C4), 139.4 (C7), 136.3 (C3), 132.0 (C1), 129.3
C8), 122.7 (C2), 119.4 (C6), 33.6 (C12), 28.9 (C9), 23.6 (C11), 23.4
2
2
Cl , 295 K) d 158.5 (C5),
1
(
(
4
+
+
H
32ClFeN
5
O:
C10). HR-MS (ESI+) m/z (%): Calc. for C26
08.2183; Found: 408.2184 (100).
H
26
N
5
([M+H] ):
1
6
, 295 K,
[
Li(THF)(thqbpi)] (2). A solution of LDA (58 mg, 0.55 mmol)
in THF (4 ml) was added to a solution of thqbpiH (204 mg, 0.50
mmol) in THF (10 ml) at -78 C. The cooling bath was removed
◦
13
1
This journal is © The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 10406–10415 | 10413

Table 1 Details of the crystal structure determinations of 1, 2, 3a, 3b and 4
Compound
Formula
1
2
3a
3b·DMSO
36ClFeN
654.06
Monoclinic
4·1.54 MeOH
C
26
H
25
N
5
C
30
H
32LiN
5
O
C
28
H
30ClCoN
579.01
Orthorhombic
P c a 2
5
OS
C
30
H
5
O
2
S
2
58.53 2 10 6.53
C H64.15Co N O
1129.91
Monoclinic
M
r
407.51
Monoclinic
P 2 /n
11.864(5)
15.282(6)
12.121(5)
111.42(1)
2046(2)
4
864
1.323
0.081
485.55
Monoclinic
P 2 /n
11.194(5)
18.995(9)
11.738(6)
90.69(1)
2496(2)
4
1032
1.292
0.080
0.8623, 0.7964
2.0 to 32.3
-16 . . . 16, 0 . . . 28,
0 . . . 17
Crystal system
Space group
1
1
1
P 2
1
/c
P 2 /c
1
a/A˚
16.954(9)
8.904(5)
17.198(8)
8.838(4)
16.324(7)
20.602(9)
90.21(2)
2972(2)
4
13.578(7)
14.502(7)
26.573(15)
100.720(7)
5141(5)
4
b/A˚
c/A˚
◦
b ( )
V/A˚
3
2596(2)
4
1204
1.481
0.877
0.7464, 0.6923
2.3 to 30.5
0 . . . 24, 0 . . . 12,
-24 . . . 24
63380
Z
F
000
1368
1.462
0.776
0.9791, 0.8147
1.99 to 32.3
-13 . . . 13, 0 . . . 24,
0 . . . 30
75048
2366
1.460
0.711
0.7464, 0.6560
1.97 to 31.50
-19 . . . 19, 0 . . . 21,
0 . . . 39
127398
d
c
/Mg m-
3
-1
m(Mo-Ka)/mm
max., min. transmission factors
0.8623, 0.8196
2.1 to 31.5
-17 . . . 16, 0 . . . 22,
◦
q range/
Index ranges (indep. set) h,k,l
0
. . . 17
Reflections measured
51391
63099
unique [Rint
]
6741 [0.0329]
5732
356
1.058
0.0433, 0.1130
0.0529, 0.1246
8437 [0.0428]
6541
430
1.059
0.0492, 0.1314
0.0675, 0.1445
7944 [0.0864]
6622
336
10075 [0.0426]
8279
412
1.041
0.0359, 0.0862
0.0495, 0.0930
16947 [0.0735]
11952
790
1.034
0.0547, 0.1136
0.0918, 0.1280
observed [I≥2s(I)]
Parameters refined
GooF on F
2
1.037
2
R indices [F>4s(F)] R(F), wR(F )
0.0388, 0.0672
0.0571, 0.0725
0.009(11)
0.495, -0.418
2
R indices (all data) R(F), wR(F )
absolute structure parameter
largest residual peaks/e·A˚
-3
0.511, -0.262
0.540, -0.248
0.844, -0.514
0.980, -0.925
(
150 MHz, DMSO-d
6
, 295 K, paramagnetic) d 848.6 (C3), 686.9
were characterised as minima by the absence of imaginary
frequencies. During the calculations the spin contamination was
(
C12), 514.5 (C4), 418.2 (C6), 382.7 (C8), 204.2 (C1), 200.8 (C2),
2
1
79.2 (C7), 117.2 (C9), 94.7 (C11), 51.9 (C10), -72.4 (C5/13).
controlled. Values of ꢀS ꢁ only slightly deviated from the expected
+
+
HR-MS (FAB+) m/z (%): Calc. for C26
4
H
25ClFeN
5
([M+H] )
value.
+
+
98.1142; Found 498.1120 (29); Calc. for C26
H
24FeN
5
([M-Cl] )
4
d
62.1376; Found 462.1338 (23). Magnetic susceptibility in DMSO-
X-ray crystal structure determinations
6
(295 K): m = 5.57 m
B
.
Crystal data and details of the structure determinations are listed
in Table 1. Intensity data were collected at low temperature
[
{Co(OAc)(thqbpi)}
2
(MeOH)] (4). ThqbpiH (100 mg,
0
.25 mmol) was added to a solution of Co(OAc)
in methanol and stirred for 18 h at room temperature. The
precipitate was filtered off, washed with THF, and dried in vacuo
2
·4 H
2
O
(100 K) with a Bruker AXS Smart 1000 CCD diffractome-
ter (Mo-Ka radiation, graphite monochromator, l = 0.71073
˚
A). Data were corrected for air and detector absorption,
to yield [{Co(OAc)(thqbpi)}
2
(MeOH)] as an orange powder (105
25
Lorentz and polarisation effects; absorption by the crystal
was treated numerically or with a semiempirical multiscan
method.
mg, 80%). Crystals suitable for an X-ray diffraction study were
obtained from a saturated solution of 4 in methanol. Found: C,
9.9; H, 5.1; N, 11.9. Calc. for C30
N, 11.6. H NMR (600 MHz, DMSO-d
9.3 (OAc), 14.1 (H6), 11.9 (H1), 10.9 (H9), 10.1 (H2), 8.7 (H9¢),
26,27
5
H
33CoN
5
O
3
S: C, 59.8; H, 5.5;
1
The structures were solved by conventional direct methods
6
, 295 K, paramagnetic) d
2
8,29
30,31
(
compound 1)
or by the charge flip procedure (all others)
3
-
2
and refined by full-matrix least squares methods based on F
against all unique reflections.
1.7 (H7), -3.0 (H10), -4.8 (H11), -35.8 (H12); The second
28,29
13 1
All non-hydrogen atoms were
resonances of H10, H11, and H12 are not detected. C { H}
NMR (150 MHz, DMSO-d , 295 K, paramagnetic): d 1010.8
OAc), 963.3 (C3), 581.9 (C4), 576.1 (C6), 518.8 (C8), 471.2 (C12),
12.4 (C1), 208.2 (C2), 91.0 (C11), 66.6 (C9), 41.8 (C10), 15.1
C7), -52.1 (C13), -217.7 (C5). HR-MS (FAB+) m/z (%): Calc.
given anisotropic displacement parameters. For the Fe and Co
complexes hydrogen atoms were placed at calculated positions
and refined with a riding model. For compounds 1 and 2 the
positions of all hydrogen atoms were taken from difference Fourier
syntheses and refined. Disorder of the methylene groups of the
tetrahydroquinoline moieties was found in the complexes 3b and
4. The crystal solvent DMSO molecule in 3b·DMSO and one
of the acetate ligands in 4 were also disordered. All disorder was
treated with split-atom models, applying geometry restraints when
necessary.
6
(
2
(
+ +
for C26
H
24CoN
5
([M-OAc] ) 465.1364; Found 465.1376 (100).
Computational details. NMR shift calculations were per-
formed using the spin unrestricted B3LYP density functional
as implemented in the Gaussian 09 program package. Larger
triple-zeta-valence plus polarisation basis sets were used for
the metal atoms and 6-311g(d,p) basis sets for the non-
metal atoms. The molecular systems studied were optimised
starting from X-ray diffraction data without any symmetry
restrictions. Stationary points on the potential energy surface
17
24
18
19
CCDC 821499–821503 contains the supplementary crystallo-
graphic data for this paper.† These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
1
0414 | Dalton Trans., 2011, 40, 10406–10415
This journal is © The Royal Society of Chemistry 2011

Acknowledgements
T. Solvi, H. Stoeckli-Evans and M. Palaniandavar, J. Inorg. Biochem.,
005, 99, 2110.
1 M. B. Meder and L. H. Gade, Eur. J. Inorg. Chem., 2004, 2716.
2 A. R. Rossi and R. Hoffmann, Inorg. Chem., 1975, 14, 365.
3 (a) G. N. La Mar, W. DeW. Horrocks Jr., R. H. Holm, (ed.); NMR
of Paramagnetic Molecules, Academic Press, New York, 1973; (b) I.
Bertini and C. Luccinat, Coord. Chem. Rev., 1996, 150, 1; (c) F. H.
K o¨ hler, Magnetism: Molecules to Materials, ed. J. S. Miller, M. Drillon,
Wiley-VCH, Weinheim, 2001, p. 379–430; (d) I. Bertini, C. Luchinat
and G. Parigi, Prog. Nucl. Magn. Reson. Spectrosc., 2002, 40, 249; (e) I.
Bertini, C. Luchinat, G. Parigi and R. Pierattelli, ChemBioChem, 2005,
6, 1536; (f) F. Rastrelli and A. Bagno, Chem.–Eur. J., 2009, 15, 7990;
(g) M. Kaupp and F. H. K o¨ hler, Coord. Chem. Rev., 2009, 253, 2376.
14 (a) T. O. Pennanen and J. Vaara, Phys. Rev. Lett., 2008, 100, 133002;
(b) H. Liimatainen, T. O. Pennanen and J. Vaara, Can. J. Chem., 2009,
87, 954.
15 (a) J. Mao, Y. Zhang and E. Oldfield, J. Am. Chem. Soc., 2002, 124,
13911; (b) R. Knorr, H. Hauer, A. Weiss, H. Polzer, F. Ruf, P. L o¨ w, P.
Dvorts a´ k and P. B o¨ hrer, Inorg. Chem., 2007, 46, 8379.
16 (a) P. Fern a´ ndez, H. Pritzkow, J. Carb o´ , P. Hofmann and M. Enders,
Organometallics, 2007, 26, 4402; (b) P. Roquette, A. Maronna, M.
Reinmuth, E. Kaifer, M. Enders and H.-J. Himmel, Inorg. Chem., 2011,
50, 1942.
17 (a) A. D. Becke, J. Chem. Phys., 1993, 98, 5648; (b) C. Lee, W. Yang
and R. G. Parr, Phys. Rev. B, 1988, 37, 785.
18 (a) A. Sch a¨ fer, C. Huber and R. Ahlrichs, J. Chem. Phys., 1994, 100,
5829; (b) F. Weigend and R. Ahlrichs, Phys. Chem. Chem. Phys., 2005,
7, 3297.
2
1
1
1
We thank the Deutsche Forschungsgemeinschaft for funding and
the Studienstiftung des Deutschen Volkes for a doctoral fellowship
(
to D. C. S.).
References
1
See for example: (a) R. M. Bullock, Angew. Chem., Int. Ed., 2007,
6, 7360; (b) S. Enthaler, K. Junge and M. Beller, Angew. Chem.,
4
Int. Ed., 2008, 47, 3317; (c) C. Bolm, J. Legros, J. Le Paih and L.
Zani, Chem. Rev., 2004, 104, 6217; (d) M. Carril, A. Correa and
C. Bolm, Angew. Chem., Int. Ed., 2008, 47, 4862; (e) S. C. Bart, E.
Lobkovsky and P. J. Chirik, J. Am. Chem. Soc., 2004, 126, 13794;
(
f) A. M. Archer, M. W. Bouwkamp, M.-P. Cortez, E. Lobkovsky and
P. J. Chirik, Organometallics, 2006, 25, 4269; (g) R. J. Trovitch, E.
Lobkovsky and P. J. Chirik, Inorg. Chem., 2006, 45, 7252; (h) S. C.
Bart, E. J. Hawrelak, E. Lobkovsky and P. J. Chirik, Organometallics,
2
1
5
005, 24, 5518; (i) H. Egami and T. Katsuki, J. Am. Chem. Soc., 2007,
29, 8940; (j) K. P. Bryliakov and E. P. Talsi, Angew. Chem., 2004, 116,
340; (k) M. Bullock, Catalysis without precious metals, Wiley-VCH
Weinheim, 2010.
2
See for example: (a) P. Di Bernardo, A. Melchior, R. Portanova, M.
Tolazzi and P. L. Zanonato, Coord. Chem. Rev., 2008, 252, 1270; (b) C.
A. Caputo and N. D. Jones, Dalton Trans., 2007, 4627; (c) T. P. Yoon
and E. N. Jacobsen, Science, 2003, 299, 1691; G. Desimoni, G. Faita
and P. Quadrelli, Chem. Rev., 2003, 103, 3119; (d) M. G o´ mez, G. Muller
and M. Rocamora, Coord. Chem. Rev., 1999, 193–195, 769; (e) N. C.
Fletcher, J. Chem. Soc., Perkin Trans. 1, 2002, 1831; (f) S. D. Ittel, L.
K. Johnson and M. Brookhart, Chem. Rev., 2000, 100, 1169; (g) L.
M. Rendina and R. J. Puddephatt, Chem. Rev., 1997, 97, 1735; (h) T.
Katsuki, Coord. Chem. Rev., 1995, 140, 189.
Reviews covering “pincer” complex chemistry: (a) M. Albrecht and G.
van Koten, Angew. Chem., Int. Ed., 2001, 40, 3750; (b) M. von der
Boom and D. Milstein, Chem. Rev., 2003, 103, 1759; (c) H. Nishiyama,
Chem. Soc. Rev., 2007, 36, 1133.
19 (a) M. M. Francl, W. J. Pietro, W. J. Hehre, J. S. Binkley, M. S. Gordon,
D. J. Defrees and J. A. Pople, J. Chem. Phys., 1982, 77, 3654; (b) W. J.
Hehre, R. Ditchfield and J. A. Pople, J. Chem. Phys., 1972, 56, 2257;
(c) P. C. Hariharan and J. A. Pople, Theor. Chim. Acta, 1973, 28, 213.
20 The calculated Mulliken spin densities reflect the total spin density
within the space around the corresponding atom, whereas the Fermi-
contact shift is a consequence of the spin density at the atom nucleus
only (i.e. MO’s with s-atomic orbital contribution). Due to the different
pathways of spin delocalisation the Mulliken spin densities and the
Fermi-contact shifts are not correlated in a linear manner and may
even show opposite signs.
3
4
5
(a) J. A. Elvidge and R. P. Linstead, J. Chem. Soc., 1952, 5000; (b) J. A.
Elvidge and R. P. Linstead, J. Chem. Soc., 1952, 5008; (c) P. F. Clark,
J. A. Elvidge and R. P. Linstead, J. Chem. Soc., 1953, 3593.
(a) R. R. Gagn e´ , W. A. Marritt, D. N. Marks and W. O. Siegl, Inorg.
Chem., 1981, 20, 3260; W. O. Siegl, Inorg. Chim. Acta, 1977, 25, L65;
21 D. F. Evans, J. Chem. Soc., 1959, 2003.
22 (a) T. J. J. Sciarone, A. Meetsma and B. Hessen, Inorg. Chim. Acta,
2006, 359, 1815; (b) J. Scott, S. Gambrotta, I. Korobkov and P. H. M.
Budzelaar, J. Am. Chem. Soc., 2005, 127, 13019; (c) A. M. Archer,
M. W. Bouwkamp, M.-P. Cortez, E. Lobkovsky and P. J. Chirik,
Organometallics, 2006, 25, 4269.
(
b) R. R. Gagn e´ , R. S. Gall, G. S. Lisensky, R. E. Marsh and L. M.
Speltz, Inorg. Chem., 1979, 18, 771; (c) D. N. Marks, W. O. Siegl and R.
R. Gagn e´ , Inorg. Chem., 1982, 21, 3140; (d) M. B. Meder, C. H. Galka
and L. H. Gade, Monatsh. Chem., 2005, 136, 1693; (e) M. Br o¨ ring,
C. Kleeberg and S. K o¨ hler, Inorg. Chem., 2008, 47, 6404; (f) B. K.
Langlotz, J. Lloret Fillol, J. H. Gross, H. Wadepohl and L. H. Gade,
Chem.–Eur. J., 2008, 14, 10267.
23 R. J. Vijn, H. J. Arts, P. J. Maas and A. M. Castelijns, J. Org. Chem.,
1993, 58, 887.
24 Gaussian 09, Revision A.02, M. J. Frisch, G. W. Trucks, H. B. Schlegel,
G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,
B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P.
Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M.
Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T.
Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery,
Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N.
Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari,
A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J.
M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J.
Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R.
Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V.
G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich,
A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and
D. J. Fox, Gaussian Inc., Wallingford CT, 2009.
25 SAINT, Bruker AXS, 1997–2008.
26 R. H. Blessing, Acta Crystallogr., Sect. A: Found. Crystallogr., 1995,
51, 33.
27 G. M. Sheldrick, SADABS, Bruker AXS, 2004–2008.
28 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr., 2007,
64, 112.
29 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr., 1990,
46, 467.
6
7
(a) W. O. Siegl, Inorg. Nucl. Chem. Lett., 1974, 10, 825; (b) W. O. Siegl,
J. Organomet. Chem., 1976, 107, C27; (c) W. O. Siegl, J. Org. Chem.,
1
977, 42, 1872; (d) L. Saussine, E. Brazi, A. Robine, H. Mimoun, J.
Fischer and R. Weiss, J. Am. Chem. Soc., 1985, 107, 3534; (e) C. A.
Tolman, J. D. Druliner, P. J. Krusic, M. J. Nappa, W. C. Seidel, I. D.
Williams and S. D. Ittel, J. Mol. Catal., 1988, 48, 129.
(a) B. Siggelkow, M. B. Meder, C. H. Galka and L. H. Gade, Eur.
J. Inorg. Chem., 2004, 3424; (b) B. A. Siggelkow and L. H. Gade, Z.
Anorg. Allg. Chem., 2005, 631, 2575.
8
9
J. A. Camerano, C. S a¨ mann, H. Wadepohl and L. H. Gade,
Organometallics, 2011, 30, 379.
B. K. Langlotz, H. Wadepohl and L. H. Gade, Angew. Chem., Int. Ed.,
2
008, 47, 4670.
1
0 In contrast, most of the “homoleptic” [M(bpi)
2
] complexes are well-
known, including their structural features. These systems are usually
substitutionally inert and therefore unsuitable as catalysts: (a) M. A.
Robinson, S. I. Trotz and T. H. Hurley, Inorg. Chem., 1967, 6, 392; (b) P.
J. Domaille, R. L. Harlow, S. D. Ittel and W. G. William, Inorg. Chem.,
1
983, 22, 3944; (c) M. B. Meder, B. A. Siggelkow and L. H. Gade,
´
Z. Anorg. Allg. Chem., 2004, 630, 1962; (d) E. Balogh-Hergovich, G.
Speier, M. R e´ glier, M. Giorgi, E. Kuzmann and A. V e´ rtes, Inorg. Chem.
Commun., 2005, 8, 457; (e) M. Wicholas, A. D. Garrett, M. Gleaves, A.
M. Morris, M. Rehm, O. P. Anders and A. la Cour, Inorg. Chem., 2006,
30 (a) G. Oszl a´ nyi and A. S u¨ t o¨ , Acta. Cryst., 2004, A60, 134; (b) G.
Oszl a´ nyi and A. S u¨ t o¨ , Acta Crystallogr., Sect. A: Found. Crystallogr.,
2004, 61, 147.
31 L. Palatinus, SUPERFLIP, EPF Lausanne, Switzerland, 2007; L.
Palatinus and G. Chapuis, J. Appl. Crystallogr., 2007, 40, 786.
4
5, 5804; (f) J. Kaizer, G. Bar a´ th, R. Csonka, G. Speier, L. Korecz, A.
Rockenbauer and L. P a´ rk a´ nyi, J. Inorg. Biochem., 2008, 102, 773; (g) P.
This journal is © The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 10406–10415 | 10415