Chiral Iron(II) NPPN Complexes: Synthesis and Application in the Asymmetric Strecker Reaction of Azomethine Imines

Article abstract of DOI:10.1021/acs.organomet.5b00357

The open-chain NPPN ligand (1S,1′S)-1,1′-((ethane-1,2-diylbis(phenylphosphanediyl))bis(2,1-phenylene))bis(N-cyclohexylmethanimine) (1) was prepared by condensation of cyclohexylamine with enantiomerically pure (1S,1′S)-2,2′-(ethane-1,2-diylbis(phenylphosphanediyl))dibenzaldehyde ((S,S)-6). Ligand 1 coordinates to [Fe(OH₂)₆](BF₄)₂ or [Fe(MeCN)₆](SbF₆)₂ in acetonitrile to give the dicationic complex [Fe(MeCN)₂(1)](X)₂ (2) (X = BF₄^- or SbF₆^-). The corresponding carbonyl (3), bromocarbonyl (4), and bis(tert-butylisonitrile) (5) derivatives were prepared and fully characterized. Complex 2 reacts with Me₃SiCN to give the corresponding trimethylsilyl isocyanide derivative 18 featuring a Fe-CNSiMe₃ linkage. The X-ray structures of 2, 3, 5, and 18 show that ligand 1 assumes the Λ-cis-α geometry, which allows comparing the trans influence of these ligands. Complexes 2, 3, 5, and 18 were applied in the asymmetric addition of trimethylsilyl cyanide to azomethine imines (Strecker reaction), whose enantioselectivity reached 22% ee. The low enantioselectivity can be explained on the basis of the Me₃SiCN/Me₃SiNC isomerization and of the reaction product partially displacing the NPPN ligand from iron.

Full text of DOI:10.1021/acs.organomet.5b00357

Article
pubs.acs.org/Organometallics
Chiral Iron(II) NPPN Complexes: Synthesis and Application in the
Asymmetric Strecker Reaction of Azomethine Imines
Raffael Huber, Raphael Bigler, and Antonio Mezzetti*
Department of Chemistry and Applied Biosciences, ETH Zurich, CH-8093 Zurich, Switzerland
̈
̈
S
* Supporting Information
ABSTRACT: The open-chain NPPN ligand (1S,1′S)-1,1′-((ethane-1,2-
diylbis(phenylphosphanediyl))bis(2,1-phenylene))bis(N-cyclohexyl-
methanimine) (1) was prepared by condensation of cyclohexylamine with
enantiomerically pure (1S,1′S)-2,2′-(ethane-1,2-diylbis(phenyl-
phosphanediyl))dibenzaldehyde ((S,S)-6). Ligand 1 coordinates to
[Fe(OH2)6](BF4)2 or [Fe(MeCN)6](SbF6)2 in acetonitrile to give the
−
dicationic complex [Fe(MeCN)2(1)](X)2 (2) (X = BF4− or SbF6 ). The
corresponding carbonyl (3), bromocarbonyl (4), and bis(tert-butyliso-
nitrile) (5) derivatives were prepared and fully characterized. Complex 2
reacts with Me3SiCN to give the corresponding trimethylsilyl isocyanide
derivative 18 featuring a Fe−CNSiMe3 linkage. The X-ray structures of 2,
3, 5, and 18 show that ligand 1 assumes the Λ-cis-α geometry, which allows comparing the trans influence of these ligands.
Complexes 2, 3, 5, and 18 were applied in the asymmetric addition of trimethylsilyl cyanide to azomethine imines (Strecker
reaction), whose enantioselectivity reached 22% ee. The low enantioselectivity can be explained on the basis of the Me3SiCN/
Me3SiNC isomerization and of the reaction product partially displacing the NPPN ligand from iron.
new, P-stereogenic synthon that gives access also to diastereo-
and enantiomerically pure open-chain NPPN ligands such as 1
(Scheme 1).
An advantage of Fe/NPPN over Fe/PNNP systems is that
the terminal donor is an N atom, which better matches the
hardness8 of Fe2+ as compared to the soft PPh2 terminus of the
INTRODUCTION
■
Being able to replace precious metals with iron provides several
advantages: iron is cheap, nontoxic, and environmentally
benign. Catalysts based on iron are employed for a broad
range of transformations,1 and nature uses it for some of the
most challenging reactions, such as the direct C−H oxidation of
methane.2 There are, however, serious obstacles that have to be
overcome in order to successfully use ironand base metals, in
generalin catalysis. In particular, 3d metals give weaker
metal−ligand bonds, which often results in high-spin electron
configurations. The resulting paramagnetic species not only
elude characterization by traditional NMR experiments but also
are less stable than their low-spin counterparts because of the
partial occupation of σ*-antibonding orbitals (eg orbitals in the
case of octahedral complexes).
Scheme 1. Synthesis of Ligand 1 and Complexes 2
A common approach to tackle these problems is to exploit
the chelate effect by using pincer3 or tetradentate4 ligands.
Strong-field ligands such as CO, isonitriles, or hydrides are also
often applied to enforce the low-spin electron configuration.
Furthermore, the macrocyclic effect5 uses ligand preorganiza-
tion and rigidity to provide a pocket of appropriate size for the
metal and to increase the kinetic inertness of the ligand
framework, which can be exploited to build robust systems.
Accordingly, iron−porphyrin systems are abundant in nature as
the active site of enzymes and are widely used.2
We recently capitalized on the macrocyclic effect with iron
complexes of tetradentate N2P2-macrocycles,6 whereas Gao has
successfully applied N4P2 macrocycles in asymmetric hydro-
genation of ketones with iron.7 Notably, the diphosphine (S,S)-
6 that we developed to prepare C2-symmetric macrocycles is a
Received: April 28, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.organomet.5b00357
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
PNNP ligands. Also, the bulkiness at the terminal donor is
lesser in the NPPN ligand 1 than in PNNP. Both factors are
expected to increase the stability of NPPN complexes as
compared to their PNNP analogues, which tend to decompose
to iron nanoparticles under harsh reaction conditions.4t The
present paper describes the synthesis of stable, diamagnetic
iron(II) complexes of the type [Fe(L)2(NPPN)]2+ (L = RCN
(Scheme 1), CO, or RNC) and a first application in asymmetric
catalysis.
Following our experience with Ru/PNNP systems,9 we
initially investigated the potential of Fe/NPPN complexes as
Lewis acids. Surprisingly, though, substrate coordination
experiments revealed that these iron(II) complexes are highly
azaphilic rather than oxophilic. This finding prompted us to
study the reactions of azomethine imines, in particular N,N′-
cyclic ones, which are mainly known for their rich cycloaddition
chemistry.10 Overall, very few examples of nucleophilic addition
to azomethine imines are known,11 of which only two11d,f are
metal-catalyzed. Inspired by a recent report of a Strecker-type
addition of cyanide onto azomethine imines catalyzed by a
cinchona-derived hydrogen-bonding catalyst,12 we here report
preliminary results concerning Fe/NPPN-catalyzed cyanide
addition onto 2-benzylidene-5-oxopyrazolidin-2-ium-1-ide.
Figure 1. ORTEP drawing of the complex cation of 2 (with ellipsoids
at 30% probability).
ly),4n whereas the imine−iron bond is longer (2.028(2) vs
2.007(6)/2.010(6) Å). We suggest that this reflects the effect of
the different chelate in the NPPN and PNNP complexes, which
strengthens the Fe−P bond in the former and the Fe−imino
ligation in the latter. The Fe−NCMe bond in 2 (1.976(2) Å) is
about 6 pm longer than in Fe/PNNP (1.915(7)/1.904(4) Å),
which originates from the higher trans influence of phosphine
as compared to acetonitrile. This may lead to a more labile
acetonitrile ligand, which should facilitate ancillary ligand
exchange.
The helical cis-α configuration of 2 renders the complex C2-
symmetric and reduces the number of possible diastereo-
isomers of the catalyst−substrate adduct. This makes 2 a
promising candidate for binding bidentate substrates (vide infra,
Table 4). In fact, a strikingly high number of C2-symmetric
ligands and their metal complexes have proven to be excellent
catalysts for enantioselective transformations in the past.17
Λ-cis-α-[Fe(CO)2(1)](BF4)2 (3). The bis(acetonitrile) com-
plex 2(BF4)2 reacts with CO (2.25 bar) in acetone to give the
dicarbonyl analogue [Fe(CO)2(1)](BF4)2 (3) (Scheme 2). The
reaction is accompanied by a color change from red to yellow.
However, acetonitrile must be removed to drive the reaction to
completion by evaporating the solution to dryness several
times. If not, a mixture of 3 and [Fe(CO)(MeCN)(1)]2+ is
obtained. A 31P{1H} NMR singlet at δ 79.6 confirms that the
C2-symmetric structure is retained in solution. Complex 3
shows ν(CO) bands at 2072 and 2028 cm−1, in the range
observed for dicationic dicarbonyl complexes18 (2094 and 2038
cm−1 for [Fe(CO)2(N2P2)]2+;18a 2059 and 2022 cm−1 for
[Fe(CO)2(P4)]2+).18b Compared with dicationic monocarbonyl
iron(II) complexes such as [Fe(CO)(MeCN)(PNNP)]2+ (ν =
2002 cm−1),19 the wavenumbers are higher and point to a
weaker dFe→π*CO backdonation, as expected for dicarbonyl
derivatives, in which the strong π-accepting CO ligands
compete for the electron density of the iron(II) ion.
RESULTS AND DISCUSSION
■
Λ-cis-α-[Fe(MeCN)2(1)]2+ (2). Condensation of enantio-
merically pure (1S,1′S)-2,2′-(ethane-1,2-diylbis(phenyl-
phosphanediyl))dibenzaldehyde ((S,S)-6)6a with cyclohexyl-
amine gave the new open-chain NPPN ligand 1 (Scheme 1).
Hexaaquairon(II) tetrafluoroborate or hexakis(acetonitrile)-
iron(II) hexafluoroantimonate13 react with 1 in acetonitrile at
room temperature to give the corresponding stable, diamag-
netic iron(II) complexes [Fe(MeCN)2(1)](BF4)2 (2(BF4)2)
and [Fe(MeCN)2(1)](SbF6)2 (2(SbF6)2), respectively. The
hexafluoroantimonate salt gives better elemental analyses as
compared to the tetrafluoroborate derivative, whose carbon
values are generally too low, probably due to combustion
problems.14
The complexation of 1 to Fe(II), which is accompanied by a
color change of the reaction solution from colorless to red,
reaches completion within 3 h, as indicated by the
disappearance of the 31P{1H} NMR signal of free 1 at δ
−25.2. The new singlet at δ 88.3 is assigned to cis-α-2 (98%)
and the low-intensity (2%) singlet at δ 93.0 to a different
isomer of 2 (possibly the trans one). The high chemical shift of
2 as compared to Fe/PNNP complexes,4n which resonate
around δ 50, is a first indication of a strong chelate effect.15 The
1H NMR signal of the equivalent imine H atoms (HCN) is a
singlet, and a P,H correlation experiment showed no cross
peaks between them and the phosphines, which further
supports a cis-α structure for the main isomer.16
Complex 2(BF4)2 was studied by X-ray diffraction. Crystals
of 2(BF4)2 were grown from CH2Cl2/Et2O. The Fe atom lies
on a two-fold axis, and the complex adopts the Λ-cis-α structure
(Figure 1), in contrast to Ru/PNNP9 or Fe/PNNP systems,4n
which usually adopt the trans structure. The pseudo-octahedral
geometry is slightly distorted, as the N(1)−Fe−N(2) angle is
acute (84.85(9)°) due to steric repulsion between the
cyclohexyl group on N(1A) and the MeCN ligand containing
N(2) (Table 1). The Fe−P bond distance is shorter in the
NPPN derivative 2 than in the PNNP complex trans-
[Fe(MeCN)2((R,R)-{PPh2(o-C6H4)CHNC6H10NCH(o-
C6H4)PPh2})] (2.2161(6) vs 2.276(2)/2.272(2) Å, respective-
X-ray-quality crystals of 3 were grown by layering a CH2Cl2
solution with Et2O under CO atmosphere. Similarly to the
bis(acetonitrile) complex 2(BF4)2, the dicarbonyl derivative 3
crystallizes as the Λ-cis-α isomer with the Fe atom on a two-fold
axis (Figure 2). The slight distortions from the octahedral
geometry found in the bis(acetonitrile) analogue 2(BF4)2 are
retained in 3 (Table 2).
The Fe−CO bond length in 3 (1.831(4) Å) is in the range
observed for trans P−Fe−CO moieties in dicarbonyl iron(II)
B
DOI: 10.1021/acs.organomet.5b00357
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Table 1. Selected Bond Lengths (Å) and Angles (deg) for 2(BF4)2
Fe−P(1)
Fe−N(2)
2.2161(6)
Fe−N(1)
N(2)−C(21)
2.028(2)
1.142(3)
1.9757(18)
Fe−N(2)−C(21)
P(1)−Fe−N(1)
P(1)−Fe−N(2)
N(1)−Fe−N(2)
P(1)−Fe−P(1A)
N(2)−Fe−N(2A)
176.8(3)
94.45(5)
93.15(6)
84.85(9)
85.04(3)
89.46(11)
N(2)−C(21)−C(22)
P(1)−Fe−N(1A)
P(1)−Fe−N(2A)
N(1)−Fe−N(2A)
N(1)−Fe−N(1A)
176.8(4)
88.45(5)
172.91(7)
92.35(8)
176.07(11)
[FeBr(CO)(1)](BPh4) (4). The bis(acetonitrile) complex
2(BF4)2 reacts with KBr (2 equiv) under CO (2.25 bar) to
give the bromocarbonyl complex 4, which was isolated as the
tetraphenylborate salt as an orange powder (Scheme 3).
Scheme 2. Synthesis of Dicarbonyl Complex 3
Scheme 3. Synthesis of Bromocarbonyl Complex 4
The 31P{1H} NMR spectrum (Figure 3) indicates that 4 is
formed as a mixture of three isomers A, B, and C in an
approximately 3:4:1 ratio (Figure 4). The amount of A
increases over time (Supporting Information, Figure S11).
However, heating the complex to accelerate the thermodynamic
equilibration led to the formation of dicarbonyl complex 3
(identified by its 31P{1H} NMR chemical shift) and other
unidentified species.
The 31P{1H} NMR signals were assigned on the basis of the
trans influence of the ligands. The AX system at δ 93.7 and 68.5
Figure 2. ORTEP drawing of the complex cation of 3 (with ellipsoids
at 30% probability).
2
(trans to bromo and carbonyl, respectively; JP,P′ = 49.8 Hz) is
assigned to the cis-α isomer 4A (OC-6-24). The 31P{1H}NMR
resonances at δ 85.0 (trans to imine) and 73.7 (trans to
carbonyl; 2JP,P′ = 38.0 Hz) are diagnostic of structure B (OC-6-
34-A). Finally, the signals at δ 91.7 (trans to bromo) and 86.9
complexes,18 such as in cis-[Fe(CO)2(P4)]2+ (1.804(2) and
1.824(3) Å; P4 = tetradentate phosphine ligand).18b Interest-
ingly, all cis-[Fe(CO)2(P−N)2]2+ (P−N = chelating P,N
ligand) exhibits a long Fe−CO distance trans to P (1.837(2)
Å) and a short one (1.788(3) Å) trans to N.18a Unsurprisingly,
phosphine has here a larger trans influence than nitrogen, which
may be relevant to catalysis as the weaker Fe−CO bond may
lead to easy dissociation to a reactive 16-electron complex. The
weak coordination of CO is in agreement with the fact that
MeCN has to be removed several times during the synthesis,
which implies that it competes effectively with CO.
2
(trans to imine; JP,P′ = 35.0 Hz) were assigned to the cis-β
isomer 4C (OC-6-23-A). The IR spectrum of 4 displays two
carbonyl absorptions at ν = 1986 and 1961 cm−1. Thus, the C−
O bond is weaker than in dicarbonyl complex 3, which reflects
the stronger π-backbonding in the monocationic monocarbonyl
4 than in the dicationic dicarbonyl species 3.
Λ-cis-α-[Fe(CNtBu)2(1)](BF4)2 (5). The bis(acetonitrile)
complex 2(BF4)2 reacts with tert-butylisonitrile in dichloro-
methane at room temperature to give Λ-cis-α-[Fe(CNtBu)2-
Table 2. Selected Bond Lengths (Å) and Angles (deg) for 3
Fe−P(1)
Fe−C(21)
2.266(1)
1.831(4)
Fe−N(1)
C(21)−O(1)
2.020(3)
1.139(4)
P(1)−Fe−N(1)
P(1)−Fe−C(21)
N(1)−Fe−C(21)
P(1)−Fe−P(1A)
C(21)−Fe−C(21A)
94.36(8)
91.03(11)
84.15(14)
84.54(5)
94.8(2)
P(1)−Fe−N(1A)
P(1)−Fe−C(21A)
N(1)−Fe−C(21A)
N(1)−Fe−N(1A)
Fe−C(21)−O(1)
86.84(8)
169.63(12)
94.75(14)
178.39(17)
172.4(3)
C
DOI: 10.1021/acs.organomet.5b00357
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Figure 3. Section of the 31P{1H} NMR spectrum of bromocarbonyl complex 4.
Figure 4. Depiction of the isomers observed by 31P{1H} NMR
spectroscopy for 4. Ligand 1 is drawn as N−P−P−N.
(1)](BF4)2 (5) (Scheme 4). The reaction is accompanied by a
color change from red to orange. A singlet at δ 82.4 in the
Scheme 4. Synthesis of Bis(isonitrile) Complex 5
Figure 5. ORTEP drawing of the complex cation of 5 (with ellipsoids
at 30% probability).
the Fe−C distance trans to imino in the macrocycle derivative
is significantly shorter (1.8433(17) Å), which reflects a lower
trans influence of imine as compared to phosphine. A similar
trend is observed in the Fe−C bond length in [Fe(PNP)-
(CNtBu)3](BF4)2 (1.854(2) Å (trans to N); 1.889(2) and
1.881(2) Å (trans to isonitrile)).20
As complexes 2, 3, and 5 exhibit very similar geometries
(Supporting Information, Table S1), the comparison of Fe−P
bond lengths gives insight into the relative trans influence21 of
acetonitrile, carbon monoxide, and tert-butylisonitrile. The
shortest Fe−P bond is found in the bis(acetonitrile) adduct 2
(2.2159(5) Å) and the longest one in the dicarbonyl complex 3
(2.2663(11) Å), with the ones in isonitrile complex 5 lying in
between (2.2458(7) Å). Hence, for such systems, the carbonyl
ligand has the strongest trans influence, followed by tert-
butylisonitrile and eventually acetonitrile. Little is known about
the trans influence of isonitriles22 and, to the best of our
knowledge, this is its first assessment in iron complexes
containing at least one phosphine donor.
31P{1H} NMR spectrum indicates that the complex is C2-
symmetric. On the basis of its electronic similarity to its
dicarbonyl analogue 3, a cis-α structure can be assumed. This
hypothesis is further supported by the presence of two bands
for the isonitrile NC stretching at 2162 and 2143 cm−1 in the
IR spectrum.
An X-ray study of crystals grown from CH2Cl2/Et2O
confirmed the structural assignment. The NPPN ligand adopts
the Λ-cis-α configuration with the Fe atom on a crystallographic
two-fold axis (Figure 5). The Fe−CNtBu bond distance
(1.896(3) Å, Table 3) is just ca. 7 pm longer than the Fe−
CO separation in 3 (1.831(4) Å), which confirms the similarity
of the CO and NCR ligands and the possible use of isonitriles
as “tunable carbonyls”.6b,c The P(1)−Fe−C(21) angle is
slightly enlarged from 91.03(11)° in 3 to 94.80(9)° in 5, in
agreement with tert-butylisonitrile being considerably more
bulky than acetonitrile or carbon monoxide.
Reactivity of [Fe(MeCN)2(1)]2+ (2). Preliminary tests of
the Fe/NPPN complexes in the asymmetric transfer hydro-
genation of ketones were not promising. The bis(acetonitrile)
derivative 2 was inactive and decomposed under reaction
conditions as already observed for analogous Fe(MeCN)2
complexes,4t,6b,c whereas the CNtBu analogue 5 gave low
enantioselectivity (21% ee, see Supporting Information).
As an alternative, we investigated the potential of Fe/NPPN
complexes as Lewis acids. Several substrates containing oxygen
and/or nitrogen functionalities (such as carbonyl or imine)
X-ray structures of dicationic Fe(II) isonitrile complexes
containing both P and N donors are extremely rare.6,20 We
have recently reported N2P2 macrocyclic complexes of iron(II)
of the type [Fe(CNR)2(N2P2)]2+.6 The Fe−CNtBu bond
distances trans to P are indistinguishable (1.896(3) and
1.8856(17) Å in 5 and in the latter complex). In contrast,
D
DOI: 10.1021/acs.organomet.5b00357
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Table 3. Selected Bond Lengths (Å) and Angles (deg) for 5
Fe−P(1)
Fe−C(21)
2.2458(7)
1.896(3)
Fe−N(1)
C(21)−N(2)
2.054(2)
1.157(4)
Fe−C(21)−N(2)
P(1)−Fe−N(1)
P(1)−Fe−C(21)
N(1)−Fe−C(21)
P(1)−Fe−P(1A)
C(21)−Fe−C(21A)
173.5(3)
92.09(6)
94.80(9)
86.70(12)
85.64(4)
85.13(17)
C(21)−N(2)−C(22)
P(1)−Fe−N(1A)
P(1)−Fe−C(21A)
N(1)−Fe−C(21A)
N(1)−Fe−N(1A)
177.2(3)
88.73(7)
175.4(1)
92.47(12)
178.88(13)
were studied to explore the reactivity of the bis(acetonitrile)
complex 2 (Chart 1). In each case, the appropriate substrate
between 2 and 16a. This observation prompted us to test
complexes 2−5 as Lewis acids in catalytic reactions of
azomethine imines.
a
Asymmetric Strecker-Type Addition to Azomethine
Imines. In preliminary experiments with 2-benzylidene-5-
oxopyrazolidin-2-ium-1-ide (16a), acrylonitrile was tested as
1,3-dipolarophile, and trifluoromethyltrimethylsilane, allyltrime-
thylsilane, and trimethylsilyl cyanide as nucleophiles. In these
experiments, only Me3SiCN reacted with azomethine imine 16a
in the presence of the chiral Lewis acid 2(BF4)2 under the
conditions given in Scheme 5. As 2-(3-oxopyrazolidin-1-yl)-2-
Chart 1. Coordination Experiments with 2(BF4)2
Scheme 5. Asymmetric Strecker Reaction
a
Complex 2(BF4)2 (5 mg, 5.4 μmol) was dissolved in CD2Cl2 (0.5
mL). The substrate (54 μmol, 10 equiv) was added, the mixture was
shaken for 30 min, and the 31P{1H} NMR spectrum of the solution
was recorded.
phenylacetonitrile (17a) was formed with low, but significant
enantioselectivity (20% ee), we focused our attention on
cyanide addition onto 16a with the dicationic complexes 2, 3,
and 5 as catalysts under different reaction conditions. An
intriguing feature of this reaction is that both the substrate and
Me3SiCN contain a nitrogen donor and can hence potentially
bind to the iron(II) ion (vide infra).
and 2(BF4)2 were mixed in a 10:1 molar ratio in CD2Cl2 at
room temperature, and the reaction was monitored by 31P{1H}
NMR spectroscopy. Based on our experience with dicationic
Ru/PNNP complexes, which give stable O,O-chelate complexes
with 1,3-dicarbonyl compounds,9a−d we tested ethyl 2-
oxocyclopentane-1-carboxylate (7) and tert-butyl 2-oxoindo-
line-1-carboxylate (8) first. However, neither 7 nor 8 reacted
with 2(BF4)2 (Chart 1). In contrast, chelating diimines such as
N-alkyl-1-(pyridin-2-yl)methanimine (9) or N1,N2-dicyclo-
hexylethane-1,2-diimine (10) displaced the NPPN ligand 1
from 2(BF4)2. Their reaction with 2(BF4)2 was instantaneous,
as indicated by the color change of the reaction solution, and
was confirmed by its 31P{1H} NMR spectra, in which the only
signal was that of the free ligand 1 at δ −25.2. Although
diimines are good ligands,23 the outcome of the reaction was
nonetheless surprising, as the multiple chelate effect of the
tetradentate ligand 1 should give robust complexes.
Therefore, we tried to fine-tune the donor properties of the
substrate by using α-iminoester 11, o-formylpyridine 12,
acylhydrazone 13, nitrostyrene 14, phosphinoylimine 15, and
azomethine imine 16a. All these substrates either did not
coordinate or displaced the NPPN ligand 1 from the iron(II)
ion. Interestingly though, azomethine imine 16a gave an
instantaneous color change from light to deep red upon
addition. As no change was observed in the 31P{1H} NMR
spectrum, we speculate that a weak interaction takes place
The reactions were run with a catalyst loading of 10 mol%
and an excess of Me3SiCN (5 equiv) (Table 4). The
Table 4. Enantioselective Catalytic Addition of Me3SiCN to
a
Azomethine Imines (Strecker Reaction)
b
c
entry substrate
Ar
catalyst
t (h) yield (%)
ee (%)
1
2
3
4
5
6
7
8
9
16a
16a
16a
16a
16a
16b
16c
16a
16a
16a
16a
Ph
Ph
Ph
Ph
Ph
2(BF4)2
2(SbF6)2
2(SbF6)2
3
1.5
1.5
7.5
1.5
4.5
1.5
1.5
1.5
1.5
1.5
1.5
99
99
91
95
89
82
87
49
99
71
91
20
20
22
11
0
d
5
4-MeOPh
4-CF3Ph
Ph
2(SbF6)2
2(SbF6)2
2(SbF6)2
2(SbF6)2
2(SbF6)2
18
21
17
9
e
f
Ph
15
8
g
10
11
Ph
Ph
0
a
Reaction conditions: substrate (0.17 mmol), catalyst (0.017 mmol),
Me3SiCN (0.86 mmol), CH2Cl2 (5 mL), T = −25 °C (unless
b
c
d
otherwise stated). Isolated yields. Determined by chiral HPLC. T =
−78 °C. Ethyl cyanoformate was used as the cyanide source. 1.1
equiv of Me3SiCN was used. Slow addition of Me3SiCN (2 μL/min).
e
f
g
E
DOI: 10.1021/acs.organomet.5b00357
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
bis(acetonitrile) derivative 2(BF4)2 was tested first (entry 1).
At −25 °C, the background reaction was much slower than the
catalyzed reaction at the same temperature, which typically
reaches completion within 1.5 h. Lowering the temperature to
−78 °C slowed the reaction down without substantially
improving the enantioselectivity (22% ee, entry 3). The
counterion (BF4 , entry 1; SbF6 , entry 2) had no influence
on the reaction outcome. The dicarbonyl complex 3 gave lower
enantioselectivity (entry 4), whereas the tert-butylisonitrile
derivative 5 reacted sluggishly and without asymmetric
induction (entry 5).
The substrate scope was briefly studied with the best
performing catalyst 2(SbF6)2. An electron-donating group
(OMe) on the aryl substituent slightly lowered the yield, but
not the enantioselectivity (entry 6), whereas the electron-
withdrawing trifluoromethyl group decreased the enantiomeric
excess to 17% (entry 7). Alternatively to Me3SiCN, ethyl
cyanoformate was used as cyanide source. Although this has
proven to be beneficial in other reports,24 ethyl cyanoformate
lowered both yield and enantioselectivity in combination with
2(SbF6)2 (entry 8). Also, lowering the excess of TMSCN to 1.1
equiv resulted in lower ee values (entry 9). Slow addition of the
cyanide was detrimental to enantioselectivity, too (entry 10).
In an attempt to improve the enantiodiscrimination, we tried
to gather insight into the mechanistic features of the reaction.
The observation that the addition of trimethylsilyl cyanide to
CH2Cl2 solutions of azomethine imine 16a and the catalyst
always caused a color change from dark red to light red-orange
suggested that Me3SiCN coordinates to the metal. Therefore,
we prepared the corresponding complex by treating the
bis(acetonitrile) complex 2(SbF6)2 with trimethylsilyl cyanide
(5 equiv) in dichloromethane (Scheme 6).
0.2%)27 of the isonitrile analogue Me3SiNC (νCN ≈ 2096
cm−1). Calculations suggest that the isonitrile form (Me3SiNC)
is 3−5 kcal/mol less stable than Me3SiCN with an activation
barrier of 28−30 kcal mol−1 for the noncatalyzed interconver-
sion.28 However, the thermodynamic preference for the cyano
form can be reversed by coordination to a transition metal, as
documented for titanium29 and rhenium30 complexes. The
formation of trimethylsilyl isocyanide complexes upon treat-
ment with Me3SiCN, which has been reported for rhenium(I),
is thought to be driven thermodynamically by the stronger net
π-acceptor/σ-donor properties of isocyanides as compared to
nitriles.29 The silyl isonitrile complexes are so stable that their
formation acts as thermodynamic sinks in the cleavage of the
C−C bond of RCN in combination with a silyl ligand on the
metal, as reported for rhodium31 and iron32 (or of the O−CN
bond of cyanates with molybdenum).33
−
−
X-ray of Λ-cis-α-[Fe(CNSiMe3)2(1)](SbF6)2 (18). Orange
crystals of 18 were obtained by layering a CH2Cl2 solution with
hexamethyldisiloxane. The Fe atom lies on a crystallographic
two-fold axis and the NPPN ligand adopts the usual Λ-cis-α
configuration with essentially linear CNSiMe3 ligands (Figure
6). The Fe−CNSiMe3 bond distance in 18 (1.886(7) Å) is very
Scheme 6. Synthesis of Bis(CNSiMe3) Complex 18
Figure 6. ORTEP drawing of the complex dication of 18 (with
ellipsoids at 30% probability).
close to the Fe−CNtBu separation (1.896(3) Å) in 5, whereas
the P(1)−Fe−C(21) angle (93.8(2)°) suggests that CNSiMe3
is slightly less bulky than CNtBu (94.80(9)°).
The reaction of 2 with Me3SiCN, which was accompanied by
a similar color change as observed in catalysis, gave an orange
solid (18) upon precipitation with hexane. The IR spectrum
showed two strong bands at 2112 (CNSiMe)3) and 2083 cm−1,
which are diagnostic of a isonitrile complex (Fe−CNSiMe3).25
The assignment based on IR data was confirmed by a broad
13C{31P,1H} NMR signal at δ 188.8 for Fe−CNSiMe3.25a An X-
ray analysis identified complex 18 as Λ-cis-α-[Fe(CNSiMe3)2-
(1)](SbF6)2 (see below). Accordingly, the 31P{1H} NMR
spectrum of 18 shows a singlet at δ 80.7, which indicates that
the same configuration is retained in solution. The signal is
rather broad (fwhm = 70 Hz), which may hint to facile
dissociation of the bulky Me3SiNC ligand or to a metal-
catalyzed isomerization process (see below). The HRMS trace
shows a peak at m/z = 698.2507 that corresponds to
[Fe(CN)(1)]+, which is not surprising given the relative
instability of the highly polarized C−Si bond in coordinated
trimethylsilyl isocyanide.
Although N and C atoms can be hardly distinguished by X-
ray diffraction, the isonitrile formulation gives a physically more
realistic pattern for the thermal parameters of the atoms
involved than the alternative Fe−NCSiMe3 description (see
Supporting Information).34 The Fe−C and C−N distances in
18 (Table 5) are similar to those found in [Fe(CNSiMe3)-
(CO)(Cp)(PPh3)]+ (1.836(3) and 1.162(4) Å)25a and [Fe-
(CNSiMe3)(CO)2(Cp)]+ (1.862(4) and 1.157(6) Å).25b The
comparison with CNSiR3 complexes of W,35 Re,30 and Rh31
shows that the C−N distance varies significantly with both the
electron density at the metal and the charge of the complex.
Mechanistic Issues. Despite its modest enantioselectivity,
the Strecker reaction catalyzed by 2(SbF6)2 is interesting, as it
is the first transition metal-catalyzed cyanide addition onto an
azomethine imine. Recently, Wang has reported an organo-
catalytic, enantioselective version of this reaction that uses a
cinchona derivative as hydrogen-bonding catalyst to activate the
azomethine imine.12 In the case of Fe(II)/NPPN, the catalyst
may activate either the substrate or Me3SiCN (or even both).
The IR data indicate that Me3SiCN (νNC ≈ 2191 cm−1)
exists in solution in equilibrium with a small amount26 (ca.
F
DOI: 10.1021/acs.organomet.5b00357
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Table 5. Selected Bond Lengths (Å) and Angles (deg) for 18
Fe−P(1)
Fe−C(21)
C(21)−Si(1)
2.241(2)
1.886(7)
1.811(7)
Fe−N(1)
C(21)−N(1)
2.046(6)
1.168(9)
Fe−C(21)−N(2)
P (1)−Fe−N(1)
P(1)−Fe−C(21)
N(1)−Fe−C(21)
P(1)−Fe−P(1A)
C(21)−Fe−C(21A)
171.3(6)
92.78(16)
93.8(2)
C(21)−N(2)−Si(1)
P(1)−Fe−N(1A)
P(1)−Fe−C(21A)
N(1)−Fe−C(21A)
N(1)−Fe−N(1A)
174.8(6)
87.71(16)
172.7(2)
94.4(3)
85.1(3)
85.17(11)
88.2(4)
179.3(3)
This is a central issue in view of the importance of metal-
catalyzed enantioselective trimethylsilyl cyanide additions in
organic synthesis,36 but also because a thorough understanding
of the catalysis intermediates might give insight into the reason
for the low enantioselectivity. To that goal, a number of
experiments were performed.
When the workup of the reaction mixture was performed
under exclusion of water, or when the reaction crude was
directly analyzed, the Me3Si-iminolate derivative 19 (Scheme 7)
The low enantioselectivity of the Strecker reaction may have
a further reason, though. Upon mixing [Fe(MeCN)2(1)]SbF6
(2(SbF6)2), azomethine imine 16a, and Me3SiCN (1:1:1 ratio)
in CD2Cl2 at room temperature, new 31P{1H} NMR resonances
at δ 88.3, 81.6, and −4.2 in a 0.36:0.55:0.07 ratio appeared
within 15 min after mixing the reagents. The chemical shift of
the singlet at δ −4.2 is diagnostic of a species with
noncoordinated P donors, whereas the signals at δ 88.3 and
81.6 are assigned to the bis(acetonitrile) complex 2(SbF6)2 and
to [Fe(CNSiMe3)2(1)]2+ (18), respectively.39 The relative
intensity of the signal at δ −4.2 increased with time, and, after
12 h, the final ratio of integrals was 0.32:0.44:0.24 for the signal
of 2(SbF6)2, 18, and the new species at δ −4.2, respectively.
A control experiment showed that the latter species derives
from the displacement of ligand 1 from iron by the reaction
product. When an excess of independently synthesized, racemic
product 17a (10 equiv) was added to the bis(acetonitrile)
derivative 2(SbF6)2 in CD2Cl2, the signal of 2(SbF6)2 at δ 88.3
disappeared instantaneously, and the singlet at δ −4.2 was the
only signal observed. As the free tetradentate ligand 1 gives a
31P{1H} NMR singlet at δ −25.2, the structure of the species
resonating at δ −4.2 remains elusive, but structures involving
dangling phosphines appear probable. As azomethine imine
does not displace the NPPN ligand (see above), the product of
the Strecker reaction 17a possibly binds to Fe(II) via the nitrile
group. As the phosphine is the stereogenic moiety, its
dissociation from iron may account for the low enantio-
selectivity.
Scheme 7. Possible Reaction Pathway for the Strecker
Reaction
Further experiments were performed with the dicarbonyl
complex 3 and with the bis(tert-butylisonitrile) derivative 5,
whose reaction with Me3SiCN (4 equiv) was monitored by
31P{1H} NMR spectroscopy. The dicarbonyl complex 3 slowly
reacted with Me3SiCN to give a supposedly Fe−NCSiMe3
intermediate. This then isomerized over time to several species
with 31P{1H} NMR signals in the δ range 81.5−85.5, which
supports the Fe−CNSiMe3 formulation (see Supporting
Information, Figure S27). As the MeCN analogue 2 reacts
instantaneously to give 18, and the enantioselective pathway
requires the activation of Me3SiCN by the chiral catalyst, the
different substitution lability of these two catalysts may affect
the rate of formation of the (yet unknown) active species and
explain the lower enantioselectivity of 3 (11% ee, Table 4, entry
4) as compared with 2 (20% ee, entry 1). A further option is
the competition of the non-enantioselective reaction, in which
the Lewis acidic iron catalyst only serves as a Me3Si+ transfer
agent, but does not deliver CN− to the activated substrate.
Finally, the CNtBu complex 5 failed to react either with
Me3SiCN or with the product of the Strecker reaction 17a,
which suggests that there is no involvement of a metal complex
in catalysis, and a slow and non-enantioselective background
was observed instead of 17. To check whether the Me3SiCN/
Me3SiNC isomerization affects the outcome of the Strecker
reaction, the bis(CNSiMe3) complex 18 was tested as catalyst
under the same conditions used with the bis(MeCN) analogue
2, which gave product 17a in 91% yield as racemate (Table 4,
entry 11).
The observation of 19 suggests that the first reaction step is
the transfer of Me3Si+ to the oxygen atom the free azomethine
imine 16a, in line with other transition metal-catalyzed
additions to azomethine imines, which generally do not involve
their coordination to the metal.37 Me3Si+ can arise either from
Fe−NCSiMe3 (A) or from Fe−CNSiMe3 (B), which gives
either an isocyano (Fe−NC, A′) or a cyano complex (Fe−CN,
B′) (Scheme 7).38 The cyano complex B′ is expected to be less
reactive (because more stable) than its isocyano analogue B.
Also, it contains the CN− fragment in the wrong orientation for
the nucleophilic attack onto the silyl-activated azomethine
imine. In turn, the Me3Si-activated azomethine imine may react
with a non-metal-bound trimethylsilyl cyanide molecule in a
non-enantioselective fashion. Albeit essentially speculative, such
a working hypothesis would accommodate the lack of
enantioselectivity of the trimethylsilyl isonitrile complex 18.
G
DOI: 10.1021/acs.organomet.5b00357
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Laboratorium fur Organische Chemie (ETH Zurich). The enantio-
̈
̈
reaction may account for the slow formation of racemic 17a
observed in catalysis (Table 4, entry 5).
meric excess was determined by chiral high-performance liquid
chromatography (HPLC) using an IA 3 μm column or an AM 5
μm column.
(1S,1′S)-N,N′-(((Ethane-1,2-diylbis(phenylphosphinediyl))-
bis(2,1-phenylene))bis(methanylylidene))dicyclohexanamine,
1. Cyclohexylamine (0.178 mL, 1.55 mmol, 2.1 equiv) was added to a
solution of (1S,1′S)-2,2′-(ethane-1,2-diylbis(phenylphosphinediyl))-
dibenzaldehyde6a (336 mg, 0.739 mmol, 1 equiv) in EtOH (7.5
mL). After the mixture was stirred for 24 h at room temperature, the
The above experiments suggest at least two reasons for the
low enantioselectivity observed in catalysis with the Fe/NPPN
catalysts, that is, the isomerization of Me3SiCN to Me3SiNC on
the iron(II) catalyst and the displacement of the chiral NPPN
ligand from the metal by the product of the Strecker reaction
17a. The facile isomerization of Me3SiCN to Me3SiNC, which
is triggered by the stability of the resulting isonitrile complex,
may hamper cyanide transfer to the substrate. Due to its general
nature, this phenomenon is potentially a serious obstacle to the
application of late transition metal catalysts to the enantio-
selective cyanide transfer from silyl cyanides, in contrast to early
transition complexes, which successfully catalyze higly enantio-
selective trimethylsilyl- and hydrocyanation reactions.40
Finally, the lability of the NPPN ligand 1 contrasts with the
sturdy nature of the closely related macrocycle N2P2 ligands
that our group reported recently.6 Albeit disappointing, this is
an instructive example of how subtle changes such as moving
from an open-chain NPPN to a macrocylic ligand can
dramatically affect the stability and catalytic performance of
iron(II) catalysts.
solvent was evaporated, and excess amine was azeotropically removed
20
with toluene to give a white solid. Yield: 460 mg (100%). [α]D
=
1
−72.0° (c = 0.1, CH2Cl2). H NMR (300 MHz, CD2Cl2): δ 9.02 (s,
2H, NCH), 7.95−7.89 (m, 2H, Ar−H), 7.39−7.11 (m, 16H, Ar−
H), 3.21−3.07 (m, 2H, N−CH), 2.14−2.08 (m, 4H, PCHH), 1.84−
1.10 (m, 20H, Cy−H). 31P{1H} NMR (122 MHz, CD2Cl2): δ −25.2.
13C{1H} NMR (75 MHz, CD2Cl2): δ 157.3 (m, NC), 141.0 (CAr),
138.5 (br m, CAr), 134.9 (m, CAr), 133.0 (t, 2JP,C = 9.5 Hz, CAr), 131.7
4
(CAr), 130.2 (CAr), 129.1 (CAr), 128.9 (CAr), 128.8 (t, JP,C = 3.2 Hz,
CAr), 127.7 (CAr), 70.0 (N−C), 34.7 (d, JP,C = 4.0 Hz, PCH2), 26.1
2
(Cy−CH2), 25.09 (Cy−CH2), 25.06 (Cy−CH2), 24.0 (Cy−CH2),
23.9 (Cy−CH2). HRMS (MALDI): m/z calcd for [M]+: 616.3136;
found 616.3134. Anal. Calcd for C40H46N2P2: C, 77.90; H, 7.52; N,
4.54; Found: C, 77.62; H, 7.58; N, 4.53.
Synthesis of (OC-6-22)-[Fe(MeCN)2(1)](X)2, 2. [Fe(OH2)6]-
(BF4)2 (319 mg, 0.945 mmol) was added to a solution of 1 (670
mg, 1.09 mmol, 1.15 equiv) in acetonitrile (15 mL), whose color
immediately changed from colorless to red. After the mixture was
stirred for 3 h, the solvent was removed under reduced pressure. Red
crystals were obtained by dissolving the crude in acetonitrile and
CONCLUSIONS
■
A number of stable, diamagnetic iron(II) complexes have been
prepared with a new, P-stereogenic open-chain NPPN ligand.
The complexes are mostly C2-symmetric and feature the Λ-cis-α
geometry. Reactivity studies revealed that these dicationic
complexes are highly azaphilic. In view of their readiness to
bind nitrogen donors, the new complexes were applied in the
first example of transition metal-catalyzed enantioselective
nucleophilic addition of trimethylsilyl cyanide to N,N′-cyclic
azomethine imines. However, the enantioselectivity is low, for
which at least two possible reasons can be envisaged. First, the
isomerization of Me3SiCN to Me3SiNC, which is thermody-
namically driven by the stability of the resulting isonitrile
complexes, produces a Fe−CN complex that can be expected to
be an unsuitable CN-transfer reagent for electronic and
stereochemical reasons. Second, in the specific case of iron,
the coordination of the nitrile-containing product to iron(II) is
so strong that it displaces the chiral tetradentate NPPN ligand.
This result shows once more that the coordination ability of the
chiral ligand is a particularly delicate issue in asymmetric
catalysis with iron(II) complexes, as even multidentate ligands
are readily displaced despite their substantial chelate effect.
−
layering with diethyl ether. The SbF6 salt, obtained from [Fe-
(MeCN)6](SbF6)2 instead of [Fe(OH2)6](BF4)2, showed essentially
the same spectroscopic data, but improved elemental analytic data,
possibly due to better combustion behavior. Yield: 660 mg (75%). 1H
NMR (300 MHz, CD2Cl2): δ 8.17 (s, 2H, NCH), 7.80−7.58 (m,
6H, Ar-H), 7.56−7.43 (m, 8H, Ar-H), 7.33−7.21 (m, 4H, Ar−H),
3.56−3.47 (m, 2H, N−CH), 2.86 (dd, 2JP,H = 29.0 Hz, 2JH,H′ = 9.0 Hz,
2H, PCHH), 2.45 (s, 6H, NCCH3), 2.28−2.11 (m, 4H, PCHH and
Cy−H), 1.89−1.79 (m, 2H, Cy−H), 1.69−1.51 (m, 4H, Cy−H),
1.38−1.02 (m, 10H, Cy−H), 0.87−0.66 (m, 2H, Cy−H). 31P{1H}
NMR (122 MHz, CD2Cl2): δ 88.3. 13C{1H} NMR (101 MHz,
CD2Cl2): δ 175.4 (NC), 137.7 (m, CAr), 137.1 (m, CAr), 134.9 (m,
3
CAr), 133.6 (CAr), 132.6 (CAr), 132.2 (t, JP,C = 4.3 Hz, CAr), 132.0
3
(CAr), 131.3 (br m, CAr), 130.2 (t, JP,C = 4.9 Hz, CAr), 74.0 (N−C),
36.1 (PCH2), 34.0 (Cy−CH2), 26.5 (Cy−CH2), 26.3 (Cy−CH2), 26.0
(Cy−CH2), 25.8 (Cy−CH2), 4.6 (CH3). The quaternary carbon atom
of the acetonitrile and one quaternary aromatic carbon atom are not
resolved. HRMS (MALDI): m/z calcd for [M − 2MeCN]+: 672.2486,
found 672.2480; calcd for [M − 2MeCN + F]+): 691.2470, found
691.2467. IR (KBr pellet, cm−1): 2315 (NCCH3), 2279 (NCCH3).
−
Anal. Calcd for C44H52FeN4P2B2F8 (BF4 salt): C, 56.93; H, 5.65; N,
6.04. Found: C, 56.76; H, 5.70; N, 7.03. Anal. Calcd for
C44H52FeN4P2Sb2F12 (SbF6 salt): C, 43.10; H, 4.27; N, 4.57.
−
EXPERIMENTAL SECTION
■
Found: C, 43.11; H, 4.28; N, 4.51.
General. All reactions were performed under an argon atmosphere
using Schlenk techniques unless otherwise stated. Solvents were of
puriss p.a. quality and were distilled under argon atmosphere with
standard drying agents (CH2Cl2, MeOH, MeCN: CaH2; Et2O,
toluene: Na/benzophenone; hexane: Na/benzophenone/tetraglyme;
EtOH: Na/diethyl phthalate). All solvents were freshly distilled prior
to use. NMR spectra were measured on Bruker Avance DPX 300 (1H,
300.1; 13C{1H}, 75.5; 31P{1H}, 121.5), Bruker Avance DPX 400 (1H,
400.1; 13C{1H}, 100.6; 31P{1H}, 162.0) or Bruker Avance DPX 500
(1H, 500.2; 13C{1H,31P}, 125.8; 31P{1H}, 202.5) spectrometers
(frequencies in MHz). The internal standard standard was the residual
1H or 13C peak of the deuterated solvent (1H NMR: CDCl3 δ = 7.26;
CD2Cl2 δ = 5.32; 13C{1H} NMR: CDCl3 δ = 77.16; CD2Cl2 δ =
53.84). 31P{1H} NMR spectra are referenced to external 85% H3PO4.
Thin-layer chromatography (TLC) was performed on Merck silica gel
60 F254 TLC plates; UV light (366 or 254 nm) or KMnO4 was used
for detection. Mass spectra were measured by the MS service of the
Synthesis of (OC-6-33)-[Fe(CO)2(1)](BF4)2, 3. Complex 2(BF4)2
(300 mg, 0.323 mmol) was dissolved in acetone (16 mL) in a 50 mL
Young Schlenk flask. The argon atmosphere was then removed and
replaced by CO (2.25 bar), the solution was stirred for 20 min, and the
volatiles were removed under reduced pressure. The procedure was
repeated until no further color change (red to yellow) was observed
(3−4 times). Then, the crude was redissolved in dichloromethane (10
mL) and layered with diethyl ether (70 mL) under CO atmosphere
(1.2 bar). After 2 days, the mother liquor was decanted, and the
product was dried to give yellow-greenish crystals. Yield: 260 mg
(89%). 1H NMR (500 MHz, CD2Cl2): δ 8.23 (s, 2H, NCH), 8.00−
7.91 (m, 4H, Ar−H), 7.88−7.83 (m, 2H, Ar−H), 7.82−7.69 (m, 10H,
Ar−H), 7.63−7.55 (m, 2H, Ar−H), 3.11 (dd, JP,H = 43.9 Hz, 2JH,H′
=
2
3
9.2 Hz, 2H, PCHH), 2.77 (d, JH,H′ = 11.2 Hz, 2H, N−CH), 2.68 (d,
2JH,H′ = 9.2 Hz, 2H, PCHH), 2.43 (d, JH,H′ = 10.8 Hz, 2H, Cy−Heq),
3
3
1.96 (d, JH,H′ = 9.2 Hz, 2H, Cy−Heq), 1.57−1.45 (m, 6H, Cy−H),
H
DOI: 10.1021/acs.organomet.5b00357
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
1.30−1.19 (m, 4H, Cy−H), 1.12−1.02 (m, 2H, Cy−H), 0.74 (d, 3JH,H′
= 11.3 Hz, 2H, Cy−H), 0.29 (q, 3JH,H′ = 13.2 Hz, 2H, Cy−H). 31P{1H}
NMR (122 MHz, CD2Cl2): δ 79.6. 13C{31P,1H} NMR (126 MHz,
CD2Cl2): δ 207.8 (CO), 176.7 (CN), 141.9 (CAr), 137.1 (CAr),
135.9 (CAr), 135.5 (CAr), 134.5 (CAr), 134.0 (CAr), 133.5 (CAr), 131.5
(CAr), 126.6 (CAr), 120.0 (CAr), 84.5 (N−C), 35.6 (PCH2), 34.9 (Cy−
CH2), 26.8 (Cy−CH2), 26.7 (Cy−CH2), 26.5 (Cy−CH2), 25.4 (Cy−
CH2). HRMS (MALDI): m/z calcd for [M + H]+: 729.2462, found
729.2453. IR (KBr pellet, cm−1): 2072 (CO), 2028 (CO). Anal. Calcd
for C42H46B2F8FeN2P2O2·CH2Cl2: C, 52.32; H, 4.90; N, 2.84. Found:
C, 52.34; H, 4.96; N, 2.80. One equivalent of dichloromethane is
incorporated in the crystals obtained by the method described above.
For a powdered, dried sample, the following analytical data were
collected. Anal. Calcd for C42H46B2F8FeN2P2O2: C, 55.91; H, 5.14; N,
3.10. Found: C, 56.14; H, 5.37; N, 2.88.
(CH3)3), 174.5 (CN), 139.5 (CAr), 136.8 (CAr), 136.0 (CAr), 134.4
(CAr), 133.8 (CAr), 133.2 (CAr), 133.0 (CAr), 130.6 (CAr), 130.0 (CAr),
122.1 (CAr), 81.5 (N−CCy), 36.0 (PCH2), 34.1 (Cy−CH2), 27.2 (Cy−
CH2), 26.8 (Cy−CH2), 26.6 (Cy−CH2), 25.5 (Cy−CH2), 0.1
(Si(CH3)3). HRMS (MALDI): m/z calcd for [Fe(CN)(1)]+:
698.2516, found 698.2507. IR (ATR, cm−1): 2112 (CNSi(CH3)3),
2083 (CNSi(CH3)3). Anal. Calcd for C48H64F12FeN4P2Si2Sb2: C,
42.94; H, 4.80; N, 4.17. Found: C, 43.15; H, 4.91; N, 4.23.
General Procedure for the Synthesis of Enantioenriched 2-
(3-Oxopyrazolidin-1-yl)-2-arylacetonitriles. Catalyst (0.017
mmol, 10 mol%) was dissolved in dichloromethane (5 mL), and the
appropriate 2-arylidene-5-oxopyrazolidin-2-ium-1-ide (16a−16c, 0.17
mmol) was added to the reaction solution, which was cooled to −25
°C. Me3SiCN (115 μL, 0.86 mmol, 5 equiv) was added in one portion
and the mixture was stirred at −25 °C. Reaction progress was checked
by TLC (10% MeOH in dichloromethane). After the reaction was
complete, the reaction mixture was filtered through a pad of silica gel,
which was washed thoroughly with ethyl acetate. Evaporation of the
solvent under reduced pressure gave products 17a−17c as off-white
solids. For characterization details, see Supporting Information.
Synthesis of [FeBr(CO)(1)](BPh4), 4. KBr (64 mg, 0.54 mmol, 2
equiv) and 2(BF4)2 (249.5 mg, 0.27 mmol) were dried under vacuum
for 15 min. Acetone (15 mL) was added, and the mixture was gently
heated until all solids dissolved. The argon atmosphere was then
removed and replaced with CO (2.25 bar). The mixture was stirred for
2 days (whereby the CO atmosphere has been replaced 4 times),
followed by removal of volatiles under reduced pressure. Then, the
crude was dissolved in MeOH (7 mL), filtered, and sodium
tetraphenylborate (92 mg, 0.27 mmol, 1 equiv) in MeOH (1 mL)
was added to the filtered solution. The resulting precipitate was filtered
ASSOCIATED CONTENT
■
S
* Supporting Information
1H, 31P, and 13C NMR spectra, details of X-ray structures,
substrate synthesis, and product characterization data. The
Supporting Information is available free of charge on the ACS
Publications website at DOI: 10.1021/acs.organomet.5b00357.
1
off to give an orange powder. Yield: 167 mg (57%). H NMR (400
MHz, CD2Cl2): δ 9.07−9.04 (m, 0.2H, NCH), 8.83−8.77 (m, 0.4H,
NCH), 7.95−7.91 (m, 1.5H, NCH), 7.86−7.35 (m, 18H, Ar−H),
7.34−7.28 (m, 8H, BPh4), 7.00 (t, 3JH,H′ = 7.4 Hz, 8H, BPh4), 6.87 (t,
3JH,H′ = 7.2 Hz, 4H, BPh4), 4.23 (t, 3JH,H′ = 10.7 Hz, 2H, NCH), 3.6−
0.6 (m, 24H, Haliphatic). 31P{1H} NMR (121 MHz, CD2Cl2): δ 93.7 (d,
AUTHOR INFORMATION
Corresponding Author
*E-mail: mezzetti@inorg.chem.ethz.ch.
■
2
2
2JP,P′ = 49.6 Hz), 91.7 (d, JP,P′ = 35.1 Hz), 86.8 (d, JP,P = 34.9 Hz),
85.0 (d, 2JP,P′ = 38.0 Hz), 73.8 (d, 2JP,P′ = 38.1 Hz), 68.6 (d, 2JP,P′ = 49.6
Hz). See Results and Discussion for signal attribution. IR (ATR, in
cm−1): υ = 1986 (CO), 1961 (CO). HRMS (MALDI): m/z calcd for
[M−CO]+: 751.1669, found 751.1663. Anal. Calcd for
C65H66BBrFeN2OP2: C, 70.99; H, 6.05; N, 2.55. Found: C, 70.73;
H, 6.02; N, 2.48.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This work was supported by the Swiss National Foundation
(SNF), Grant No. 200020_146881.
■
Synthesis of (OC-6-33)-[Fe(CNtBu)2(1)](BF4)2, 5. tert-Butyliso-
cyanide (20 μL, 0.181 mmol, 4 equiv) was added to a solution of
2(BF4)2 (42 mg, 0.045 mmol) in dichloromethane (3 mL). After being
stirred for 2 h, the solution was concentrated to 1 mL and layered with
hexane (11 mL) to give orange crystals. Yield: 40 mg (87%). 1H NMR
(300 MHz, CD2Cl2): δ 8.07 (s, 2H, NCH), 7.90−7.80 (m, 2H, Ar−
H), 7.75−7.65 (m, 6H, Ar−H), 7.63−7.47 (m, 10H, Ar−H), 2.96 (dd,
2JP,H = 36.1 Hz, 2JH,H′ = 8.7 Hz, 2H, PCHH), 2.70 (t, 3JH,H′ = 11.0 Hz,
REFERENCES
■
(1) For selected reviews, see: (a) Sherry, B. D.; Furstner, A. Acc.
̈
Chem. Res. 2008, 41, 1500−1511. (b) Enthaler, S.; Junge, K.; Beller, M.
Angew. Chem., Int. Ed. 2008, 47, 3317−3321. (c) Morris, R. H. Chem.
Soc. Rev. 2009, 38, 2282−2291. (d) Gopalaiah, K. Chem. Rev. 2013,
113, 3248−3296. (e) Bauer, I.; Knolker, H.-J. Chem. Rev. 2015, 115,
3170−3387.
2H, N−CH), 2.60 (d, 3JH,H′ = 11.4 Hz, 2H, Cy−Heq), 2.36 (d, 2JH,H′
=
̈
3
8.7 Hz, 2H, PCHH), 1.94 (d, JH,H′ = 12.4 Hz, 2H, Cy−Heq), 1.51 (s,
18H, CNC(CH3)3), 1.46−1.00 (m, 12H, Cy−H), 0.41−0.21 (m, 4H,
Cy−H). 31P{1H} NMR (122 MHz, CD2Cl2): δ 82.4. 13C{1H} NMR
(101 MHz, CD2Cl2): δ 175.4 (NC), 139.7 (m, CAr), 137.0 (m, CAr),
135.8 (m, CAr), 134.4 (CAr), 133.6 (CAr), 133.1 (CAr), 133.0 (CAr),
130.5 (m, CAr), 122.1 (br m, CAr), 81.1 (N−C), 60.9 (C(CH3)3), 35.7
(PCH2), 34.0 (2C, Cy−CH2), 30.7 (C(CH3)3), 26.8 (2C, Cy−CH2),
25.4 (Cy−CH2). The quaternary carbon atom of the isonitrile and one
quaternary aromatic carbon are not resolved. HRMS (MALDI): m/z
calcd for [M]+: 838.3956, found 838.3949. IR (ATR, cm−1): 2162
(tBuNC), 2143 (tBuNC). Anal. Calcd for C50H64B2F8FeN4P2: C,
59.31; H, 6.37; N, 5.53. Found: C, 59.14; H, 6.27; N, 5.32.
(2) Bertini, I.; Gray, H. B.; Lippard, S. J.; Valentine, J. S. Bioinorganic
Chemistry; University Science Books: Mill Valley, CA, 1994.
(3) (a) Bart, S.; Lobkovsky, E.; Chirik, P. J. J. Am. Chem. Soc. 2004,
126, 13794−13807. (b) Tondreau, A. M.; Atienza, C. C. H.; Weller, K.
J.; Nye, S. A.; Lewis, K. M.; Delis, J. G. P.; Chirik, P. J. Science 2012,
335, 567−570. (c) Langer, R.; Leitus, G.; Ben-David, Y.; Milstein, D.
Angew. Chem., Int. Ed. 2011, 50, 2120−2124. (d) Langer, R.; Iron, M.
A.; Konstantinovski, L.; Diskin-Posner, Y.; Leitus, G.; Ben-David, Y.;
Milstein, D. Chem.Eur. J. 2012, 18, 7196−7209. (e) Langer, R.;
Diskin-Posner, Y.; Leitus, G.; Shimon, L.; Ben-David, Y.; Milstein, D.
Angew. Chem., Int. Ed. 2011, 50, 9948−9952. (f) Zell, T.; Butschke, B.;
Ben-David, Y.; Milstein, D. Chem.Eur. J. 2013, 19, 8068−8072.
(g) Lagaditis, P. O.; Sues, P. E.; Sonnenberg, J. F.; Wan, K. Y.; Lough,
A. J.; Morris, R. H. J. Am. Chem. Soc. 2014, 136, 1367−1380.
(4) (a) Chen, M. S.; White, M. C. Science 2007, 318, 783−787.
(b) Chen, M. S.; White, M. C. Science 2010, 327, 566−571. (c) Bigi,
M. A.; Reed, S. A.; White, M. C. Nat. Chem. 2011, 3, 216−222.
(d) Bigi, M. A.; Liu, P.; Zou, L.; Houk, K. N.; White, M. C. Synlett
2012, 23, 2768−2772. (e) Bigi, M. A.; Reed, S. A.; White, M. C. J. Am.
Chem. Soc. 2012, 134, 9721−9726. (f) Gormisky, P. E.; White, M. C. J.
Am. Chem. Soc. 2013, 135, 14052−14055. (g) Zhou, S.; Fleischer, S.;
Junge, K.; Das, S.; Addis, D.; Beller, M. Angew. Chem., Int. Ed. 2010,
Synthesis of (OC-6-22)-[Fe(CNSiMe3)2(1)](SbF6)2, 18.
Trimethylsilyl cyanide (53 μL, 0.4 mmol, 5 equiv) was added to a
solution of 2(SbF6)2 (97 mg, 0.079 mmol) in dichloromethane (4
mL), and stirred for 5 h, after which hexane was added to precipitate
1
the product. Yield: 76 mg (71%). H NMR (300 MHz, CD2Cl2): δ
8.00 (s, 2H, NCH), 7.87−7.33 (m, 18H, Ar−H), 3.20−2.64 (m, 6H,
PCH2 and N−CH), 2.36−2.18 (m, 2H, Cy−Heq), 1.95−1.75 (m, 2H,
Cy−Heq), 1.60−1.34 (m, 4H, Cy−Heq), 1.33−0.82 (m, 1H, Cy−H),
0.37 (s, 18H, Si(CH3)3), 0.26−0.08 (m, 2H, Cy−H). 31P{1H} NMR
−
(122 MHz, CD2Cl2): δ 80.7. For the BF4 salt, a shift of δ 82.9 is
observed. 13C{31P,1H} NMR (126 MHz, CD2Cl2): δ 188.8 (CNSi-
I
DOI: 10.1021/acs.organomet.5b00357
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
49, 8121−8125. (h) Wienhofer, G.; Sorribes, I.; Boddien, A.;
(12) (a) Li, N.; Liu, Z.; Huang, X.; Zhang, J.; Chen, X.; Wang, Y.;
Wang, X. RSC Adv. 2013, 3, 9154−9157. (b) For an organocatalytic
Strecker reaction with Et2AlCN as nucleophile, see: Kaur, P.; Pindi, S.;
Wever, W.; Rajale, T.; Li, G. G. J. Org. Chem. 2010, 75, 5144−5150.
(13) [Fe(MeCN)6](SbF6)2 was prepared by double chloride
abstraction from anhydrous FeCl2 using AgSbF6 in acetonitrile and
stored in a glovebox.
̈
Westerhaus, F.; Junge, K.; Junge, H.; Llusar, R.; Beller, M. J. Am.
Chem. Soc. 2011, 133, 12875−12879. (i) Wienhofer, G.; Westerhaus,
̈
F.; Jagadeesh, R.; Junge, K.; Junge, H.; Beller, M. Chem. Commun.
2012, 48, 4827−4829. (j) Ziebart, C.; Federsel, C.; Pazhamalai, A.;
Jackstell, R.; Baumann, W.; Spannenberg, A.; Beller, M. J. Am. Chem.
Soc. 2012, 134, 20701−20704. (k) Wienhofer, G.; Westerhaus, F.;
̈
Junge, K.; Ludwig, R.; Beller, M. Chem.Eur. J. 2013, 19, 7701−7707.
(14) Marco,
́
A.; Compano, R.; Rubio, R.; Casals, I. Microchim. Acta
́
̃
(l) Wienhofer, G.; Westerhaus, F.; Junge, K.; Beller, M. J. Organomet.
2003, 142, 13−19.
̈
(15) Garrou, P. E. Inorg. Chem. 1975, 14, 1435−1439.
(16) (a) In the case of a trans P−M−NCH arrangement, the 4JP,H
is large enough to split the imine proton into a doublet and to generate
cross peaks in the P,H 2D NMR spectrum.. (b) Schotes, C.;
Ranocchiari, M. M.; Mezzetti, A. Organometallics 2011, 30, 3596−
3602.
Chem. 2013, 744, 156−159. (m) Buchard, A.; Heuclin, H.; Auffrant,
A.; Le Goff, X. F.; Le Floch, P. Dalton Trans. 2009, 1659−1667.
(n) Sui-Seng, C.; Freutel, F.; Lough, A. J.; Morris, R. H. Angew. Chem.,
Int. Ed. 2008, 47, 940−943. (o) Mikhailine, A. A.; Lough, A. J.; Morris,
R. H. J. Am. Chem. Soc. 2009, 131, 1394−1395. (p) Meyer, N.; Lough,
A. L.; Morris, R. H. Chem.Eur. J. 2009, 15, 5605−5610. (q) Lagaditis,
P. O.; Mikhailine, A. A.; Lough, A. J.; Morris, R. H. Inorg. Chem. 2010,
49, 1094−1102. (r) Lagaditis, P. O.; Lough, A. J.; Morris, R. H. Inorg.
Chem. 2010, 49, 10057−10066. (s) Mikhailine, A. A.; Morris, R. H.
Inorg. Chem. 2010, 49, 11039−11044. (t) Sonnenberg, J. F.; Coombs,
N.; Dube, P. A.; Morris, R. H. J. Am. Chem. Soc. 2012, 134, 5893−
5899. (u) Mikhailine, A. A.; Maishan, M. I.; Lough, A. J.; Morris, R. H.
J. Am. Chem. Soc. 2012, 134, 12266−12280. (v) Prokopchuk, D. E.;
Morris, R. H. Organometallics 2012, 31, 7375−7385. (w) Zuo, W.;
Lough, A. J.; Li, Y. F.; Morris, R. H. Science 2013, 342, 1080−1083.
(x) Zuo, W.; Tauer, S.; Prokopchuk, D. E.; Morris, R. H.
Organometallics 2014, 33, 5791−5801. (y) Hoyt, J. M.; Shevlin, M.;
Margulieux, G. M.; Krska, S. W.; Tudge, M. T.; Chirik, P. J.
Organometallics 2014, 33, 5781−5790.
(17) Yoon, T. P.; Jacobsen, E. N. Science 2003, 299, 1691−1693.
(18) (a) Li, S. L.; Mak, T. C. W.; Zhang, Z. Z. J. Chem. Soc., Dalton
Trans. 1996, 3475−3483. (b) Burrows, A. D.; Dodds, D.; Kirk, A. S.;
Lowe, J. P.; Mahon, M. F.; Warren, J. E.; Whittlesey, M. K. Dalton
Trans. 2007, 570−580.
(19) Typical values are around 2000 cm−1; see ref 4p and the
following: (a) Henry, R. M.; Shoemaker, R. K.; Newell, R. H.;
Jacobsen, G. M.; DuBois, D. L.; Rakowski DuBois, M. Organometallics
2005, 24, 2481−2491. (b) Kohl, S. W.; Heinemann, F. W.; Hummert,
M.; Bauer, W.; Grohmann, A. Chem.Eur. J. 2006, 12, 4313−4320.
(c) Benito-Garagorri, D.; Wiedermann, J.; Pollak, M.; Mereiter, K.;
Kirchner, K. Organometallics 2007, 26, 217−222.
(20) Benito-Garagorri, D.; Alves, L. G.; Veiros, L. F.; Standfest-
Hauser, C. M.; Tanaka, S.; Mereiter, K.; Kirchner, K. Organometallics
2010, 29, 4932−4942.
(21) Pidcock, A.; Richards, R. E.; Venanzi, L. M. J. Chem. Soc. A
1966, 1707−1710.
(22) For two reports on the trans-influence of isonitriles in platinum
(5) Cabbiness, D. K.; Margerum, D. W. J. Am. Chem. Soc. 1969, 91,
6540−6541.
(6) (a) Bigler, R.; Otth, E.; Mezzetti, A. Organometallics 2014, 33,
4086−4099. (b) Bigler, R.; Mezzetti, A. Org. Lett. 2014, 16, 6460−
6463. (c) Bigler, R.; Huber, R.; Mezzetti, A. Angew. Chem., Int. Ed.
2015, 17, 5171−5174.
́ ́
complexes, see: (a) Jovanovic, B.; Manojlovic, M. J. J. Chem. Soc.,
(7) (a) Chen, J. S.; Chen, L. L.; Xing, Y.; Chen, G.; Shen, W. Y.;
Dong, Z. R.; Li, Y. Y.; Gao, J. X. Acta Chim. Sin. 2004, 62, 1745−1750.
(b) Yu, S. L.; Shen, W. Y.; Li, Y. Y.; Dong, Z. R.; Xu, Y. Q.; Li, Q.;
Zhang, J. N.; Gao, J. X. Adv. Synth. Catal. 2012, 354, 818−822. (c) Li,
Y.; Yu, S.; Wu, X.; Xiao, J.; Shen, W.; Dong, Z.; Gao, J. J. Am. Chem.
Soc. 2014, 136, 4031−4039.
Dalton Trans: Inorg. Chem. 1972, 11, 1176−1178. (b) Ros, R.; Renaud,
J.; Roulet, R. J. Organomet. Chem. 1976, 104, 271−279.
(23) Johnson, L. K.; Killian, C. M.; Brookhart, M. J. Am. Chem. Soc.
1995, 117, 6414−6415.
(24) (a) Saravasan, S.; Khan, N.; Bera, P. K.; Kureshi, R. I.; Abdi, S.
H. R.; Kumar, P.; Bajaj, H. C. ChemCatChem 2013, 6, 1374−1385.
(b) Kumar, P.; Saravanan, S.; Khan, N.; Hussain, F.; Singh, S. Eur. J.
Inorg. Chem. 2014, 29, 5077−5083.
(25) (a) Nakazawa, H.; Itazaki, M.; Owaribe, M. Acta Crystallogr.
2005, E61, m1172−m1173. (b) Nakazawa, H.; Itazaki, M.; Owaribe,
M. Acta Crystallogr. 2005, E61, m1173−m1174.
(26) Bither, T. A.; Knoth, W. H.; Lindsey, R. V.; Sharkey, W. H. J.
Am. Chem. Soc. 1958, 80, 4151−4153.
(27) (a) Booth, M. R.; Frankiss, S. G. Chem. Commun. 1968, 1347−
1348. (b) Seckar, J. A.; Thayer, J. S. Inorg. Chem. 1976, 15, 501−504.
(28) (a) Tao, J. C.; Guo, Y.; Li, S. H. J. Mol. Struct.: THEOCHEM
2009, 899, 61−70. (b) Wang, J. Y.; Cai, D. L.; Zhang, M.; Wang, M. J.
Organomet. Chem. 2013, 724, 117−128.
(29) Bochmann, M.; Wilson, L. M.; Hursthouse, M. B.; Motevalli, M.
Organometallics 1988, 7, 1148−1154.
(30) Guedes da Silva, M. F.; Lemos, M. A. N. D. A.; Frausto da Silva,
J. J. R.; Pombeiro, A. J. L.; Pellinghelli, M. A.; Tiripicchio, A. Dalton
Trans. 2000, 373−380.
(31) Taw, F. L.; White, P. S.; Bergman, R. G.; Brookhart, M. J. Am.
Chem. Soc. 2002, 124, 4192−4193.
(32) Nakazawa, H.; Kawasaki, T.; Miyoshi, K.; Suresh, C. H.; Koga,
N. Organometallics 2004, 23, 117−126.
(33) Fukumoto, K.; Dahy, A. A.; Oya, T.; Hayasaka, K.; Itazaki, M.;
Koga, N.; Nakazawa, H. Organometallics 2012, 31, 787−790.
(34) (a) On the basis of the same criterion and the IR data, the
structure of [Ni(NCSiMe3)(PPh3)3], which was claimed to be a Ni−
NCSiMe3 complex, may deserve reexamination. (b) Wilhelm, D.;
Hoffmann, T.; Wenschuh, E.; Reck, G.; Winter, G. Z. Anorg. Allg.
Chem. 1993, 619, 1801−1805.
(8) Parr, R. G.; Pearson, R. G. J. Am. Chem. Soc. 1983, 105, 7512−
7516.
(9) (a) Mezzetti, A. Dalton Trans. 2010, 39, 7851−7869. (b) Schotes,
C.; Mezzetti, A. J. Am. Chem. Soc. 2010, 132, 3652−3653. (c) Schotes,
C.; Mezzetti, A. Angew. Chem., Int. Ed. 2011, 50, 3072−3074.
(d) Schotes, C.; Althaus, M.; Aardoom, R.; Mezzetti, A. J. Am. Chem.
Soc. 2012, 134, 1331−1343. (e) Egloff, J.; Ranocchiari, M.; Schira, A.;
Schotes, C.; Mezzetti, A. Organometallics 2013, 32, 4690−4701.
(10) (a) Stanley, L. M.; Sibi, M. P. Chem. Rev. 2008, 108, 2887−
2902. (b) Milosevic, S.; Togni, A. J. Org. Chem. 2013, 78, 9638−9646.
(c) Guo, H.; Liu, H.; Zhu, F. L.; Na, R.; Jiang, H.; Wu, Y.; Zhang, L.;
Li, Z.; Yu, H.; Wang, B.; Xiao, Y.; Hu, X. P.; Wang, M. Angew. Chem.,
Int. Ed. 2013, 52, 12641−12645. (d) Tong, M. C.; Chen, X.; Tao, H.
Y.; Wang, C. J. Angew. Chem., Int. Ed. 2013, 52, 12377−12380.
(e) Yamashita, Y.; Kobayashi, S. Chem.Eur. J. 2013, 19, 9420−9427.
(f) Li, J.; Lian, X.; Liu, X.; Lin, L.; Feng, X. Chem.Eur. J. 2013, 19,
5134−5140. (g) Tsu
̌
pova, S.; Maeorg, U. Heterocycles 2014, 88, 129−
̈
173.
(11) (a) Burger, K.; Thenn, W.; Schickaneder, H. Chem. Ber. 1975,
108, 1468−1474. (b) Kawai, H.; Kusuda, A.; Nakamura, S.; Shiro, M.;
Shibata, N. Angew. Chem., Int. Ed. 2009, 48, 6324−6327. (c) Okusu, S.;
Kawai, H.; Xu, X. H.; Tokunaga, E.; Shibata, N. J. Fluorine Chem. 2012,
143, 216−219. (d) Shintani, R.; Soh, Y. T.; Hayashi, T. Org. Lett. 2010,
12, 4106−4109. (e) Hashimoto, T.; Kimura, H.; Kawamata, Y.;
Maruoka, K. Nat. Chem. 2011, 3, 642−646. (f) Hashimoto, T.; Omote,
M.; Maruoka, K. Angew. Chem., Int. Ed. 2011, 50, 8952−8955.
(g) Hashimoto, T.; Kimura, H.; Kawamata, Y.; Maruoka, K. Angew.
Chem., Int. Ed. 2012, 51, 7279−7281.
J
DOI: 10.1021/acs.organomet.5b00357
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(35) Suzuki, E-; Komuro, T.; Kanno, Y.; Okazaki, Y.; Tobita, H.
Organometallics 2018, 29, 1838−1848.
(36) Weber, W. P. Silicon Reagents for Organic Synthesis; Springer
Verlag: Berlin, 1983; pp 6−20.
(37) For instance, the rhodium(I)-catalyzed phenylation of azo-
methine imines with NaBPh4 is thought to involve the formation of a
rhodium−phenyl intermediate.11d Analogously, a Cu(box)−acetylide
complex transfers RCC− to C,N-cyclic azomethine imines in the
presence of a Brønsted acid cocatalyst in excellent yield and with high
enantioselectivity.11f
(38) A similar ambiguity has been discussed for the oxidative addition
of silyl cyanides to rhodium(II) porphyrins: Chan, K. S.; Zhang, L.;
Fung, C. W. Organometallics 2004, 23, 6097−6098.
(39) In this experiment, the 31P{1H} NMR signal of 18 (δ 81.6) is
slightly shifted toward that of 2(SbF6)2 as compared to the chemical
shift of isolated 18 (δ 80.7), which we tentatively attribute to chemical
exchange between 18 and 2(SbF6)2. Accordingly, when Me3SiCN (1
equiv) is added to 2(SbF6)2 in CD2Cl2 at −78 °C, the 31P{1H} NMR
spectrum of the reaction mixture shows singlets at δ 88.2 and 80.8,
which are assigned to 2(SbF6)2 and 18, respectively.
(40) Brunel, J. M.; Holmes, I. P. Angew. Chem., Int. Ed. 2004, 43,
2752−2778.
K
DOI: 10.1021/acs.organomet.5b00357
Organometallics XXXX, XXX, XXX−XXX