Reactivity of Me-pma RhI and IrI Complexes upon Deprotonation and Their Application in Catalytic Carbene Carbonylation Reactions

Article abstract of DOI:10.1002/ejic.201501302

Dehydrogenative oxidation of amines is a relevant process in metal-mediated catalysis, with the amines being either substrates or ligands. Transformation of amine- into imine-type ligands in the coordination sphere of a transition metal can be an important catalyst activation process. The behaviour of secondary pyridin-2-ylmethanamine (pma) ligands in the corresponding rhodium and iridium complexes upon NH deprotonation varies, depending on a number of factors. In this paper the behaviour of the Me-pma ligand [Me-pma = N-methyl-1-(pyridin-2-yl)methanamine] bound to [Rh(cod)]⁺ and [Ir(cod)]⁺ was studied. Whereas the iridium amido complex could be obtained upon NH deprotonation, the rhodium complex instantaneously disproportionated into a free pma ligand and an unusual dinuclear complex, adopting a structure with two Rh^I metal centres hosted by a dianionic (pma-2H)^2- ligand, and with the ligand coordinating to Rh2 as an "aza-allyl" fragment. The study gives further proof for the effect of pyridine ligation on the previously observed charge-transfer from the ligand to the metal. Furthermore, the catalytic activity of both the Ir and the Rh species with Me-pma in carbene carbonylation reactions to generate ketenes was studied.

Full text of DOI:10.1002/ejic.201501302

DOI: 10.1002/ejic.201501302
Full Paper
Reactivity on Deprotonation
Reactivity of Me-pma RhI and IrI Complexes upon
Deprotonation and Their Application in Catalytic Carbene
Carbonylation Reactions
Zhou Tang,[a] Cristina Tejel,[b] Marc Martinez de Sarasa Buchaca,[a,b] Martin Lutz,[c]
Jarl Ivar van der Vlugt,[a] and Bas de Bruin*[a]
Dedicated to Luis Oro on the occasion of his 70th birthday
Abstract: Dehydrogenative oxidation of amines is a relevant tained upon NH deprotonation, the rhodium complex instanta-
process in metal-mediated catalysis, with the amines being ei- neously disproportionated into a free pma ligand and an un-
ther substrates or ligands. Transformation of amine- into imine- usual dinuclear complex, adopting a structure with two RhI
type ligands in the coordination sphere of a transition metal metal centres hosted by a dianionic (pma-2H)2– ligand, and with
can be an important catalyst activation process. The behaviour the ligand coordinating to Rh2 as an “aza-allyl” fragment. The
of secondary pyridin-2-ylmethanamine (pma) ligands in the cor- study gives further proof for the effect of pyridine ligation on
responding rhodium and iridium complexes upon NH deproto- the previously observed charge-transfer from the ligand to the
nation varies, depending on a number of factors. In this paper metal. Furthermore, the catalytic activity of both the Ir and the
the behaviour of the Me-pma ligand [Me-pma = N-methyl-1- Rh species with Me-pma in carbene carbonylation reactions to
(pyridin-2-yl)methanamine] bound to [Rh(cod)]+ and [Ir(cod)]+ generate ketenes was studied.
was studied. Whereas the iridium amido complex could be ob-
Introduction
oxidation of a metal-bound amido group into the correspond-
ing imine, two distinct pathways are generally proposed to be
operational (Scheme 1). The first involves a metal–hydride inter-
mediate arising from ꢀ-H elimination, which can subsequently
be transferred to a hydrogen acceptor[7] or released in the form
of dihydrogen (path i);[8] the second does not involve a metal–
hydride intermediate but proceeds through (stepwise or con-
certed) deprotonation and oxidation steps (one proton and two
electrons), reducing either the metal or an exogenous “oxi-
dant”[9] (path ii). Other mechanisms involving either direct li-
gand-to-(exogenous) substrate hydride-transfer[10] or H-atom
transfer[11] are also reported.
Dehydrogenative oxidation of amines is an important process
in both biological and chemical systems.[1] Amine oxidases are
widely found in bacteria, plants and animals, and are involved
in a number of basic biological processes, such as lysyl oxid-
ation in the cross-linking of collagen.[2] Transition-metal-pro-
moted dehydrogenative amine oxidation reactions are fre-
quently used in synthetic chemistry to generate imines,[3] ni-
triles[4] and diazo compounds.[5] Transformation of amine- into
imine-type ligands in the coordination sphere of transition met-
als can also be an important catalyst activation process.[6] These
reactions involve multiple proton- and electron-transfer steps,
but the mechanisms are not fully understood at the molecular
level. Metal-amido complexes are usually key intermediates in
these transformations, irrespective of the pathway involved. For
[a] Homogeneous, Bioinspired & Supramolecular Catalysis, van 't Hoff Institute
for Molecular Sciences, Faculty of Sciences, University of Amsterdam,
Science Park 904, 1098 XH Amsterdam, The Netherlands
E-mail: b.debruin@uva.nl
www.homkat.nl
[b] Departamento de Química Inorgánica, Instituto de Síntesis Química y
Catálisis Homogénea – ISQCH, CSIC – Universidad de Zaragoza,
Pedro Cerbuna 12, 50009 Zaragoza, Spain
[c] Crystal & Structural Chemistry, Bijvoet Center for Biomolecular Research,
Utrecht University,
Padualaan 8, 3584 CH Utrecht, The Netherlands
Supporting information and ORCID(s) from the author(s) for this article are
available on the WWW under http://dx.doi.org/10.1002/ejic.201501302.
Scheme 1. Two generalized proposed pathways for imine formation via
metal-amido complexes.
Eur. J. Inorg. Chem. 2016, 963–974
963
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Ligand-based redox activity[12] adds to the complexity of the
overall system and the mechanistic interpretation of amine to
imine conversion, because the ligand can participate directly in
the redox process.[13] For example, one-electron oxidation of an
amido-rhodium complex [RhI(bipy)(trop2N)] (trop = 5-H-di-
benzo[a,d]cycloheptene-5-yl, bipy = 2,2′-bipyridyl) led to forma-
tion of a rhodium aminyl radical [RhI(bipy)(trop2N·)]OTf (OTf– =
trifluorosulfonate) instead of the expected rhodium(II) complex,
because the oxidation occurs mainly at the amido ligand.[14a,14b]
Mono-deprotonation of a metal-bound amino group can also
trigger formation of a ruthenium aminyl radical [RuII(Bu2sq)-
(LNH·)] [LNH2
=
bis(2-pyridylmethyl)-2-aminoethylamine,
Bu2sq = 3,5-di-tert-butylsemiquinonate] from its amido electro-
isomer, which is able to catalyse the oxidation of alcohols into
aldehydes.[14c,14d] Two-electron oxidation of a metal-bound am-
ido ligand is usually accompanied by proton loss from the α-
position to the nitrogen, producing an imine complex (which is
often more stable, in particular for late transition metals). Previ-
ous studies investigated the dehydrogenative oxidation of rho-
dium and iridium amino complexes derived from cationic
[RhI(bpa)(cod)]+ and [IrI(bpa)(cod)]+ [bpa = Py-CH2-NH-CH2-Py =
bis(pyridin-2-ylmethyl)amine, cod = 1,5-cyclooctadiene].[15] This
has led to the mixed-valent dirhodium complex [(cod)Rh–I{(bpa-
2H)0}RhI(cod)] (A, Scheme 2), which was obtained by deproto-
nation of the mononuclear cationic amino-precursor
[RhI(bpa)(cod)]+ with one molar equivalent of base. The reaction
likely involves an unstable amido complex as intermediate. The
overall reaction at the ditopic ligand of complex A consists of
two deprotonations and a two-electron transfer process, lead-
ing to formation of the neutral Py-CH2-N=CH-Py imine ligand
(bpa-2H)0 with formal reduction of one of the metal centres
from rhodium(I) to rhodium(–I) (Scheme 2).[15b] The analogous
dinuclear rhodium(norbornadiene) complex B,[15a] and a heter-
odinuclear rhodium-iridium complex C1,[15b,15c] with similar
electronic structures, were also fully characterised. Interestingly,
complex C1 [(cod)Rh–I{(bpa-2H)0}IrI(cod)] was observed in equi-
librium with isomer C2, which holds the formulation of
[(cod)RhI{(bpa-2H)–2}IrI(cod)]. Key structural changes on going
from C1 to C2 (Scheme 2) were decoordination of one of the
pyridyl ligands and a slight slippage of Rh in order to interact
with the “iridaazabutenyl” moiety (Scheme 2). These changes
are associated with two-electron transfer from the rhodium to
the “imine” moiety of the ligand. Moreover, (de)coordination of
the pendant pyridyl ring may be directly related to reversible
metal-to-ligand charge-transfer, because pyridine coordination
is likely required to stabilise the reduced tetrahedral Rh–I centre.
Scheme 2. Top: previous work on mixed-valent dinuclear rhodium complex
A obtained through cooperative double deprotonation and ligand-to-metal
charge transfer.[15b] Bottom: isomerisation of complex C to the two extreme
electronic forms C1 and C2, affected by temperature.[15c]
cently also demonstrated that those amido complexes are ac-
tive catalysts in catalytic carbene carbonylation reactions to
produce ketenes (Scheme 3).[18]
Scheme 3. Top: H2 loss from a Rh-PNN amido complex. Bottom: A Rh-PNN
amido complex as the catalyst for carbene carbonylation reaction to produce
ketenes.[15e]
To elucidate the subtleties involved in the coordination
chemistry and (catalytic) reactivity of pyridyl-amido complexes
of Rh and Ir, as well as pyridyl-imine ligands involving two metal
centres, we herein report on rhodium and iridium(cod) com-
plexes based on the Me-pma ligand [Me-pma = N-methyl-1-
(pyridin-2-yl)methanamine], an analogue to the bpa ligand that
lacks the pendant pyridine. The aim of these studies was to
investigate the effect of the Me-pma ligand on the dehydrogen-
ative ligand oxidation process and to shed more light on the
effect of pyridine ligation on the previously observed charge-
transfer from the ligand to the metal (Scheme 2, bottom). We
further evaluated their activity in catalytic carbene carbonyl-
ation.
Besides the reactions described above (Scheme 1), spontane-
ous H2-loss from a protic and hydridic ligand-arm of PNN-pin-
cer-type ligands has also been observed to occur in apolar sol-
vents such as benzene,[15e] which triggered interesting ligand-
centred reactivity (Scheme 3) that goes beyond chemistry with,
for example, PNP or bipyridine-based PNN ligands.[16] Interest-
ingly, the ligand substituents and/or the solvent seem to play
a crucial role in this process, because the amido form of PNN-
ligands of the type shown in Scheme 3 proved to be much
more stable in tetrahydrofuran (THF), with no indications for
spontaneous H2-loss from these related ligands.[6a,17] We re-
Results and Discussion
The mononuclear complexes 1a (Ir) and 1b (Rh) were prepared
in good yields (87 and 81 % yield, respectively) according to a
Eur. J. Inorg. Chem. 2016, 963–974
www.eurjic.org
964
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
previously reported procedure (Scheme 4),[15b] reacting the Me- Synthesis of Iridium–Amido Complex 2a Involving
pma ligand with the [{M(cod)(μ-Cl)}2] precursors (M = IrI or RhI), Monodeprotonation of 1a
followed by precipitation of the complexes as PF6– salts by add-
Addition of one equivalent of potassium tert-butoxide (KOtBu)
to a THF solution of 1a led to a colour change from yellow to
deep-red. Disappearance of the -NH signal, appearance of the
methylene group as a singlet at δ = 4.55 ppm and chemical
1
ing NH4PF6. In the H NMR spectrum, the –NH signals of com-
plexes 1a and 1b appear at δ = 4.32 and 3.76 ppm, respectively.
The methylene group in both complexes is observed as a well-
defined doublet of doublet (AB spin system), implying inequiva-
lence of the two protons. Similarly, asymmetry is observed for
the cyclooctadiene ligands. Single crystals of both complexes
suitable for X-ray diffraction studies were obtained by top-layer-
ing pentane onto concentrated dichloromethane solutions.
shifts for the cod signals (Δδ = 0.39–0.27 ppm upfield relative
1
to 1a) in the H NMR spectrum support the formation of the
neutral iridium-amido complex 2a (Scheme 5).
Scheme 5. Synthesis of complex 2a by monodeprotonation with KOtBu.
Single crystals suitable for X-ray structure determination
were obtained by storing a concentrated diethyl ether solution
of 2a at –20 °C. The molecular structure of deprotonated com-
plex 2a reveals a square-planar coordination geometry around
iridium (Figure 2). In contrast to the neutral ligands in 1a and
1b, the anionic ligand of 2a coordinates in the coordination
plane without remarkable distortion (∠Ir–N1–C5–C6: 5.26°;
∠N1–C5–C6–N2: 1.18°). The Ir–N2 bond in 2a is shortened from
Scheme 4. Synthesis of complexes 1a and 1b.
2.1043(19) Å in 1a to 1.9730(13) Å, resulting from a stabilising
push-pull π-interaction. The nitrogen lone pair donates to the
metal, which, in turn, back-donates to the trans-C=C bond of
the cod-ligand. Hence, the C12–C13 bond length of the cod
ligand trans to the amido N atom is elongated to 1.427(2) Å
compared with the corresponding bond in 1a [1.412(4) Å], and
is also longer than C8–C9 [1.412(2) Å], in agreement with en-
hanced π-back-donation from the metal upon deprotonation
The crystal structures of 1a and 1b (Figure 1) are isomor-
phous and the two metal centres consequently exhibit almost
identical square planar geometries with the plane defined by
the two nitrogen atoms of the Me-pma ligand and the two
centroids of the C=C bonds coordinated to the metal. The ge-
ometry around atom N2 is tetrahedral in both cases, with the
lone pair coordinated to the metal and the methyl group point-
ing away from and being almost perpendicular to the coordina-
of the secondary amine nitrogen.
tion planes of the metals [∠ Ct1–Ct2–N2–C7 98.10(16) and
99.46(11)° for 1a and 1b, respectively]. The M–N and M–Ct
bonds are slightly more contracted for the iridium complex 1a.
Figure 2. Molecular structure (ORTEP at 50 % level) of 2a. Hydrogen atoms
Figure 1. Molecular structure (ORTEP at 50 % level) of the cations of com-
are removed for clarity, except on C6. Selected bond lengths [Å] and angles
plexes 1a (left) and 1b (right). Hydrogen atoms are removed for clarity, except
[°]: Ir–N1, 2.0901(12); Ir–N2, 1.9730(13); Ir–Ct1, 1.992; Ir–Ct2, 1.994; N1–C1,
for those on C6 and N2; red dots represent the centroids of the coordinated
1.349(2); N1–C5, 1.3493(19); C5–C6, 1.490(2); C6–N2, 1.436(2); N2–C7, 1.449(2);
C=C bonds of the cod ligand. Selected bond lengths [Å] and angles [°], for
C8–C9, 1.412(4); C12–C13, 1.427(2); N1–Ir–Ct1, 172.62; N2–Ir–Ct2, 175.91; N1–
1a: Ir–N1, 2.0814(19); Ir–N2, 2.1044(19); Ir–Ct1, 2.0012(17); Ir–Ct2, 2.0083(15);
Ir–N2, 79.78(5); Ct1–Ir–Ct2, 87.43; Ir–N1–C5–C6, 5.26; N1–C5–C6–N2, –1.18.
N1–C1, 1.353(3); N1–C5, 1.353(3); C5–C6, 1.502(3); C6–N2, 1.495(3); C8–C9,
1.410(3); C12–C13, 1.412(3); N1–Ir–Ct1, 170.02(7); N2–Ir–Ct2, 176.57(7); N1–Ir–
N2, 79.21(7); Ct1–Ir–Ct2, 87.57(7); Ir–N2–C6, 107.08(14); Ct1–Ct2–N2–C7,
Synthesis of the Bimetallic Complex 3b Involving
Cooperative Ligand Deprotonation
98.10(16). For 1b: Rh–N1, 2.0965(12); Rh–N2, 2.1147(13); Rh–Ct1, 2.0143(11);
Rh–Ct2, 2.0210(10); N1–C1, 1.3497(19); N1–C5, 1.3542(18); C5–C6, 1.503(2);
C6–N2, 1.485(2); C8–C9, 1.397(2); C12–C13, 1.392(2); N1–Rh–Ct1, 169.86(4);
N2–Rh–Ct2, 176.36(5); N1–Rh–N2, 79.29(5); Ct1–Rh–Ct2, 87.89(4); Rh–N2–C6,
106.40(9); Ct1–Ct2–N2–C7, 99.46(11).
In contrast to the mono-deprotonation of iridium complex 1a,
treatment of the rhodium complex 1b in [D8]THF with one mo-
Eur. J. Inorg. Chem. 2016, 963–974
www.eurjic.org
965
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
lar equivalent of KOtBu did not produce the rhodium-amido nals of 3b in C6D6 appear in the range between δ = 6.57–
complex 2b (Scheme 6). Instead, spontaneous double deproto- 5.66 ppm in the 1H NMR spectrum. The triplet signal at δ =
nation was observed, despite using only one equivalent of base. 5.16 ppm (J = 0.8 Hz) correlates with a doublet at δ =
1
1
In the H NMR spectrum (see the Supporting Information, Fig- 89.80 ppm (JC,H = 5.5 Hz) in the H,13C-hsqc spectrum, and it is
ure S1), half an equivalent of free Me-pma ligand was detected, assigned to the “imine” HC=N proton. Both signals of the imine
and the pyridine signals of the remaining Me-pma ligand were moiety are markedly shifted downfield compared with those of
remarkably shifted upfield to the region of δ = 6.71–5.84 ppm. complexes A and B (1H: δ = 3.54 and 4.33 ppm, 13C: δ = 76.9
The same chemical shift trend was observed for the cod signals and 81.88 ppm, respectively), pointing to a stronger “imine”
(especially the vinylic protons), compared with those in com- character of the C=N bond in 3b and therefore a higher degree
plex 1b, indicating the electron-rich nature of the new rhodium of electron-transfer from the deprotonated N2–C6 moiety (see
complex 3b. Integration of the peaks in the H NMR spectrum Table 1). However, oxidation of the dianionic “(CH-NHMe)2–”
1
points to the presence of two cod ligands (which is suggestive moiety to form a neutral C(H)=NMe imino moiety could occur
of the presence of two rhodium centres) and one ligand derived either by electron-transfer from the ligand to the metal, with
from Me-pma. However, signals corresponding to the NH or the RhI centre as the formal oxidant, as observed for A and B
CH2 moieties of the original Me-pma ligand were absent, and (Scheme 2), or by charge-delocalisation into the pyridine moi-
instead one sharp pseudo-triplet[19] at δ = 5.52 ppm (1 H) was ety, with the aromatic ring acting as the electron acceptor. The
found that did not seem to correlate with other hydrogen nu- latter is more likely for 3b, because the pyridine proton signals
clei in the 1H,1H-cosy spectrum. The observed reactivity pattern of 3b are markedly shifted upfield relative to complexes A and
resembles that of the previously reported reaction of complex B (see shifts of HPy in Table 1). Furthermore, the absence of a
[Rh(bpa)(cod)]PF6 with one equivalent of KOtBu, resulting in for- pendant coordinating pyridine group in 3b should destabilise
mation of the dinuclear mixed-valent {RhI, Rh–I} complex A a hypothetical Rh–I centre, because a tetrahedral geometry is
[(cod)RhI{(bpa-2H)–2}-RhI(cod)],[15b] which overall shows an in- no longer accessible. Three sharp signals in the region of δ =
teresting pathway for the dehydrogenative oxidation of a 4.23–3.42 ppm (four H atoms) are assigned to the olefinic pro-
metal-bound amine into an imine (Scheme 2).
tons of [cod(1)] coordinated to Rh1 on the pma ligand plane.
Signals at δ = 2.06–1.94 and 1.85–1.75 ppm are assigned to the
allylic protons of this cod(1)-ligand, based on 1H,13C-hsqc,
1
1H,1H-cosy and H,1H-noesy spectra, which all indicate the ab-
sence of any symmetry in cod(1). In contrast, the olefinic pro-
tons of cod(2) above the Me-pma-2H ligand plane give rise to
a very broad signal in the range of 4.98–3.76 ppm, whereas the
corresponding carbon signals are not visible in the 13C NMR
spectrum and the allylic proton signals show up as three sets
of multiplets in a more narrow region of δ = 2.26–1.79 ppm in
the 1H NMR spectrum. The observed coalescence of the olefinic
cod signals of cod(2) points to rotation of this ligand around
the Rh2 centre with an intermediate rate on the NMR timescale
at room temperature. In fact, cod(2) rotation around Rh2 is
slower than observed for complex A, but still faster than cod(1)
rotation around Rh1. This may indicate some “tetrahedral char-
acter” for Rh2,[15b,20] but the unsaturated nature of Rh2 (lacking
an additional donor from a pendant pyridine such as in com-
plexes A and C1) might also contribute to the rotational fluxion-
ality of cod(2) being faster than cod(1).
Scheme 6. Synthesis of complex 3b by treating 1b with one equivalent of
KOtBu.
The new dinuclear species 3b, formulated as [(cod)RhI{(Me-
pma-2H)–2}RhI(cod)], was assigned to contain the “doubly de-
protonated” dianionic {(Me-pma-2H)–2} ligand. Complex 3b was
also efficiently synthesised by reacting [{Rh(cod)(μ-OMe)}2] with
Me-pma in a 1:1 molar ratio. In this case, the bridging methoxy
ligand in the precursor complex functions as an internal base.
Further recrystallisation afforded 3b as a dark brownish-red
solid in 62 % yield.
Double deprotonation of the Me-pma ligand converts the
“CH2-NHMe” moiety into a dianionic “(CH-NHMe)2–” moiety.
Subsequent two-electron oxidation of the (Me-pma-2H)2–
would transform the ligand into a neutral (Me-pma-2H)0 ligand
containing an imine “C(H)=NMe” moiety. In case of complexes A
and B, based on the bpa ligand (see Scheme 2 and surrounding
discussion in the introduction), as well as for their iridium ana-
logues, deprotonation leads to imine formation. This conversion
was observed in an intramolecular redox process wherein one
of the metal ions acts as a two-electron oxidant, thus producing
species that are best described as having a π-bound imine li-
gand coordinated to a tetrahedral rhodium(–I) or iridium(–I) ion,
respectively. To determine whether the same situation could
Table 1. Selected NMR parameters [δ (ppm)] of known complexes A,[15b]
B,[15a] C1[15c] and C2[15c] (see Scheme 2) compared with those of new com-
plex 3b.
δ
A
B
C1
C2[a]
3b
HHC=N
CHC=N (JRh,C
HPy1
HPy3
HPy4
4.33
82.2 (7)
6.80
6.58
6.20
3.54
76.9 (9)
6.95
6.56
6.41
4.84
5.77
5.16
89.8 (5)
6.56
6.47
6.16
)
91.8 (br)[b] 93.3 (5)
6.94[c]
6.90[c]
7.35[c]
6.48[c]
6.95
6.52
6.38
5.81
HPy2
5.78
5.65
5.68
[a] At 80 °C. [b] At –70 °C. [c] At –100 °C.
Deep-red crystals of 3b suitable for X-ray diffraction studies
arise in complex 3b, the spectral and structural parameters of were obtained from diethyl ether at –20 °C (Figure 3, left). The
this complex were examined in detail. The pyridine proton sig- structure contains two Rh(cod) fragments and a ditopic doubly
Eur. J. Inorg. Chem. 2016, 963–974
www.eurjic.org
966
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
deprotonated Me-pma ligand. Atom Rh1 is clearly coordinated
to the two nitrogen atoms (N1 and N2) and to the cod ligand,
resulting in a square-planar environment as expected for
rhodium(+I). However, the geometry around Rh2 is more diffi-
cult to define because, in addition to the cod ligand, it seems
to have an interaction with the N2–C6–C5 moiety,[21] and also
interacts weakly with Rh1 as indicated by the relatively short
Rh1–Rh2 bond length of 2.91468(19) Å (Table 2 and Table 3,
Figure 3). Nonetheless, key parameters to define the oxidation
state for this rhodium atom were the intraligand bond lengths
in the α-pyridylamine moiety (Table 2), which clearly indicates
it to be closer to C2 than to complexes A/B, both having an
oxidised (bpa-2H)0 ligand; for example, the N1–C5, C5–C4 and
C4–C3 bond lengths of 3b are 1.385(2), 1.421(5) and 1.356(7) Å,
respectively, indicating alternating double and single bond
characters, similar to 1.394(5), 1.421(5) and 1.356(7) Å for C2.
Table 3. Selected bond lengths [Å] and angles [°] of X-ray crystal structures
of complexes A,[15b] C2[15c] and 3b.
A
C2
3b
M1–N1
M1–N2
M1–Ct1
M1–Ct2
M1–Rh2
C8–C9
C12–C13
N1–C5
N1–C1
C1–C2
C2–C3
C3–C4
C4–C5
C5–C6
C6–N2
Rh2–C5
Rh2–C6
Rh2–N2
Rh2–Ct3
Rh2–Ct4
C16–C17
C20–C21
N1–M1–Ct1
N2–M1–Ct2
N1–M1–N2
Ct1–M1–Ct2
Ct4–Rh2–M1
Ct3–Rh2–Ct4
M1–C6–Rh2
2.094(3)
2.032(2)
2.010(3)
1.997(3)
3.4093(4)
1.394(4)
1.401(5)
1.373(3)
1.355(4)
1.369(4)
1.401(5)
1.374(5)
1.407(4)
1.435(4)
1.415(4)
2.925(3)
2.130(3)
2.254(2)
2.043(3)
1.964(3)
1.396(4)
1.427(4)
173.6(1)
175.7(1)
80.4(1)
2.054(3)
2.022(3)
1.987(8)
1.953(9)
2.8773(3)
1.419(5)
1.421(5)
1.394(5)
1.372(6)
1.355(7)
1.412(6)
1.356(7)
1.421(5)
1.417(6)
1.396(5)
2.381(4)
2.136(4)
2.384(3)
2.027
2.0711(14)
2.0340(14)
2.0031(13)
2.0016(12)
2.91468(19)
1.403(3)
1.403(3)
1.385(2)
1.368(2)
1.363(2)
1.412(3)
1.359(2)
1.427(2)
1.418(2)
1.387(2)
2.4696(16)
2.1356(16)
2.2246(14)
2.0259(14)
2.0004(13)
1.391(3)
2.001
1.388(5)
1.408(5)
174.25
175.86
79.60
86.93
171.17
88.32
68.01
1.397(3)
170.61(6)
175.65(6)
79.73(5)
87.77(5)
159.52(4)
87.87(6)
87.7(1)
108.72
87.1(1)
83.71
68.87(4)
ter in the former complexes; this is in agreement with the
chemical shift differences of the HC=N signals in the 1H and 13
C
NMR spectra. The C5–C6 bond is also slightly shorter in 3b/C2
than in A/B, perhaps indicating pseudo-double-bond character.
The distance between N1 and C5 is slightly longer in 3b/C2
than in A/B, and the pyridine shows a more distinctive C–C and
C=C bond alteration (see Table 2). It is difficult to draw solid
conclusions from the C6–N2 and C5–C6 bonds concerning the
ligand oxidation states in the complexes because π-coordina-
tion of the ligand to the Rh2 centre has an intrinsic elongating
effect on these bonds. Previous studies showed that (in contrast
to the C6–N2 and C5–C6 bonds) the intraligand bond lengths
within the pyridine ring are diagnostic for the ligand oxidation
state of dinuclear π-coordinated (reduced) α-imine-pyridine li-
gand types such as Me-pma-2H.[15c] As such, these bond
lengths are quite similar to those in complex C2, in between
those expected for the fully neutral and dianionic ligands, but
closer to the dianionic form than to the neutral form. Hence,
the pyridine bond lengths in the molecular structure of 3b are
diagnostic for the presence of a (Me-pma-2H)2– ligand with a
“dearomatised” pyridine ring, rather than a (Me-pma-2H)0 li-
gand. The Rh1–N1 bond length is shorter in 3b [2.0711(14) Å]
than in A/B [2.094(3)/2.082(2) Å], which is consistent with a
higher degree of charge transfer (delocalisation) into the pyr-
idine moiety.
Figure 3. Molecular structure (ORTEP at 50 % level) of 3b and C2. Hydrogen
atoms are removed for clarity, except on C6 and C7. Selected bond lengths
[Å] and angles [°], for 3b: Rh1–N1, 2.0711(14); Rh1–N2, 2.0340(14); Rh1–Ct1,
2.0031(13); Rh1–Ct2, 2.0016(12); C8–C9, 1.403(3); C12–C13, 1.403(3); Rh2–Ct3,
2.0259(14); Rh2–Ct4, 2.0004(13); C16–C17, 1.391(3); C20–C21, 1.397(3); N1–
Rh1–Ct1, 170.61(6); N2–Rh1–Ct2, 175.65(6); N1–Rh1–N2, 79.73(5); Ct1–Rh1–
Ct2, 87.77(5); Ct4–Rh2–Rh1, 159.52(4); Ct3–Rh2–Ct4, 87.87(6).
Table 2. Selected interligand bond lengths [Å] of complexes A,[15b] B,[15a]
C2[15c] and 3b.
A
B
3b
C2
N1–C5
C5–C6
C6–N2
N1–C1
C1–C2
C2–C3
C3–C4
C4–C5
1.373(3)
1.435(4)
1.415(4)
1.355(4)
1.369(4)
1.401(5)
1.374(5)
1.407(4)
1.362(4)
1.437(5)
1.415(4)
1.349(4)
1.379(5)
1.393(5)
1.373(5)
1.406(4)
1.385(2)
1.418(2)
1.387(2)
1.368(2)
1.363(2)
1.412(3)
1.359(2)
1.427(2)
1.394(5)
1.417(6)
1.396(5)
1.372(6)
1.355(7)
1.412(6)
1.356(7)
1.421(5)
The coordination mode of 3b is somewhat similar but not
identical to that observed for rhodium in [(cod)RhI{(bpa-
Thus, the C6–N2 bond lengths in C2/3b are shorter than in 2H)–2}IrI(cod)] complex C2. In that case the Rh2 is also assigned
complexes A/B, which is suggestive of stronger “imine” charac- to be in the +I oxidation state and the bpa-2H ligand bears the
Eur. J. Inorg. Chem. 2016, 963–974
www.eurjic.org
967
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
negative charge. However, the Rh2 centre in C2 is best de- variety of electron-transfer and charge-delocalisation processes
scribed to coordinate to the “iridaazabutenyl” moiety, and Rh2 from the doubly deprotonated C6–N2 moiety, which results in
has a somewhat stronger interaction with the C5=C6 bond than a continuum for the diamide-to-picolylimine redox chemistry.
with the N2=C6 bond. The Rh2 centre in complex 3b, on the
other hand, seems to interact primarily with the N2–C6–C5
Synthesis of the Bimetallic Complex 3a
“aza-allyl” moiety, and has a stronger interaction with the
N2=C6 bond than with the C5=C6 bond. Furthermore, the Rh2–
Rh1 distance in 3b [2.91468(19) Å] is significantly longer than
the Rh2–Ir1 distance in C2 [2.8773(3) Å]. Hence, the N2–C6–C5
moiety in 3b is best viewed as an aza-allyl ligand bound to Rh2,
with the distance between the aza-allyl centroid Ct and Rh2
being 2.034(1) Å.[22] In this description, the geometry of the
Rh2 in 3b is, in fact, somewhere between a “tetrahedral” and
“square planar” coordination geometry. This distorted geome-
try may well contribute to a lower cod rotation barrier, in agree-
The iridium analogue of 3b, [(cod)IrI{(Me-pma-2H)–2}IrI(cod)]
(3a), was also synthesised and was isolated as a deep brownish-
red solid. The complex was prepared by reacting [{Ir(cod)(μ-
OMe)}2] with two equivalents of the Me-pma ligand in Et2O.
Suitable crystals of 3a could not be obtained, and we therefore
turned to DFT calculations to elucidate the preferred geometry.
We first validated our computational approach by comparing
the DFT optimised geometry of [(cod)RhI{(Me-pma-2H)–2}RhI-
(cod)] (3b′) (bp86/def2-TZVPP level employing Grimme's D3 dis-
persion corrections) with the structure of 3b determined by X-
ray diffraction (see Table S1 in the Supporting Information). The
structure of [(cod)IrI{(Me-pma-2H)–2}IrI(cod)] (3a′) was com-
puted at the same level of theory to serve as a model for 3a.
The geometry of 3a′ is analogous to that of 3b, with the
Ir1(cod) fragment coordinated to the Me-pma-2H ligand plane
and the Ir2(cod) fragment seemingly interacting with the N2–
C6–C5 “aza-allyl” moiety. Ir2 is again best described as an
iridium(+I) centre and there is once more substantial charge
delocalisation over the pyridine ring, as can be seen from the
partial dearomatisation of the ring indicated by the alternating
bond C–C lengths. Complexes 3a′ and 3b/3b′ are very similar,
except for the substantially longer N2=C6 distance in 3a′, which
indicates that the N2=C6 moiety has a somewhat smaller “im-
ine” character in 3a′. As observed for Rh2 in 3b/3b′. Atom Ir2
likely interacts slightly stronger with N2=C6 in 3a′ than in 3b′,
given the slightly longer bond length for M2–C5 (Δ = 0.078 Å)
and M2–N2 (Δ = 0.019 Å), and the larger C5–C6–M2 angle (Δ =
2.99°) and N2–C6–M2 (Δ = 1.27°). Overall, the geometric and
electronic structure of 3a′ is similar to those of 3b/3b′.
1
ment with the H NMR spectroscopic data showing that cod(2)
shows fluxional behaviour due to rotation around Rh2, whereas
cod(1) is frozen on the NMR timescale. This distorted geometry
also corresponds to partial charge transfer from the ligand to
Rh2. Nevertheless, taking these observations together, we con-
clude that Rh2 is best described as being in the +I oxidation
state, adopting a distorted square-planar coordination geome-
try in its interactions with Rh1, cod(2) and the N2–C6–C5 aza-
allyl moiety of the (Me-pma-2H)2– ligand.[23]
Hence, in contrast to the observations for complex A, the
structural parameters and NMR spectroscopic observations for
3b suggest limited electron-transfer from the doubly deproto-
nated (Me-pma-2H)2– ligand to Rh2. Instead, electron delocal-
ization into the pyridine moiety occurs with extensive “dearo-
matisation” of the pyridine moiety, which still leads to a sub-
stantial “imine” character of the N2=C6 moiety. The reluctance
of Rh2 to undergo formal reduction from rhodium(+I) to rho-
dium(–I) in complex 3b is presumably due to the absence of an
additional pyridine donor in close proximity [which stabilises
the respective rhodium(–I) centre, because a tetrahedral geom-
etry similar to that observed in complex A is no longer accessi-
ble]. It should be noted that exogenous donors, such as a free
pma ligand, do not seem to readily coordinate to the Rh2 cen-
tre to induce ligand-to-metal charge transfer (see Scheme 6).
The coordination mode observed for 3b is complementary to
the coordination modes I previously observed for complexes A,
B and C1 and coordination mode II observed for complex C2
(Figure 4).
Catalytic Activity of 1a and 1b in Ketene Formation by
Carbonylation of Carbenes
Ketenes are highly active and very useful intermediates for the
synthesis of a wide range of chemicals.[24] Although carbonyl-
ation of transition-metal carbenes is a promising synthetic
method to produce ketenes,[25] providing a valuable alternative
to conventional methods that have clear limitations,[26] most
systems developed to date involve only stoichiometric reac-
tions.[27] Catalytic ketene formation by using cobalt and palla-
dium[28] catalysts was reported recently. A variety of esters, am-
ides and ꢀ-lactams were accessible via the ketene intermediates
that were generated in situ.
However, whereas chiral induction in these coupling reac-
tions is highly desirable,[29] none of the reported catalysts is
capable of both catalysing ketene formation from carbene and
promoting the following coupling reactions in an enantioselect-
ive manner [e.g., when several chiral Co(porphyrin) complexes
were used as catalysts, no ee was detected[28e]]. This is attrib-
Figure 4. Three geometric and electronic structures of the bimetallic Rh or Ir
complexes featuring a doubly deprotonated ditopic picolyl-amine ligand.
The versatility of the Rh2 coordination mode reflects the rich uted to the relatively low affinity of these metal centres to bind
interplay between the metal and ligand, and demonstrates a the ketene intermediates. We reasoned that complexes 1a and
Eur. J. Inorg. Chem. 2016, 963–974
www.eurjic.org
968
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
1b might serve as suitable catalyst candidates for this purpose. For substrate 4a, the amine side product 7a, arising from direct
Rhodium complexes are known to promote stoichiometric coupling between metallocarbene and 4-nitroaniline 5, was al-
ketene formation by carbonylation of carbenes,[30] and rhodium ways formed in a significant amount, along with the alkene side
complexes should display a relatively strong affinity to bind the product 9a formed from dimerisation of the carbene. Addition
resulting ketenes. With suitable metal complexes, this might of K2CO3 reduced the formation of 7a and 9a (Table 4, en-
eventually lead to the development of enantioselective catalytic tries 1–4). For the iridium precatalyst 1a, monodeprotonation
protocols. We recently reported on the activity of rhodium com- of the Me-pma ligand would lead to a more electron-rich amido
plexes with PNN pincer-type ligands in catalytic carbene carb- complex with concomitantly enhanced π-back donation from
onylation reactions,[18] showing that rhodium is indeed a good the metal to the carbene p-orbital. This would reduce the elec-
candidate for further developments in this field. Therefore, we trophilicity of the carbene, thus suppressing its propensity for
wondered whether rhodium(I) complexes with bidentate N-do- coupling with 4-nitroaniline and carbene dimerisation through
nor ligands could also facilitate this reaction, because these li- coupling with the carbene precursor. For rhodium precatalyst
gands may be more easily modified to tune (enantio)selectivity 1b, the exact form of the active metal complex is less clear,
in future studies. Iridium might also be a good candidate, be- because deprotonation of the Me-pma ligand of 1b can lead to
cause it can be expected to bind ketenes even more strongly, formation of dinuclear complex 3b (see above). Whether or not
even though Ir complexes are thus far less explored as media- such a disproportionation reaction also occurs under (diluted)
tors for ketene formation reactions.
catalytic conditions is not clear.
The reactions were performed by using previously reported
Iridium complex 1a showed poor activity and selectivity with
conditions of 20 bar CO at 60 °C (Table 4).[18,28e,28f] Ethyl diazo- respect to product 6 with both carbene precursors 4a and 4b
acetate (EDA) 4a and sodium 2-benzylidene-1-tosylhydrazin-1- (Table 4, entries 1, 2 and 5). The rhodium precatalyst 1b showed
ide (4b) were tested as carbene precursors, and 4-nitroaniline good activity and selectivity to produce 6 from both 4a and 4b
was used as the nucleophile to trap the ketene intermediates under basic conditions (entries 4 and 6; note that 4b is also a
to form amide product 6. By using 1b as catalyst, the corre- moderately strong base, so both 1a and 1b should be in their
sponding amide products were obtained in 78 and 62 % yield deprotonated forms when using an excess of this substrate).
with 4a and 4b, respectively, with base being employed for 4a. Surprisingly, when 4b was used as the carbene precursor, imine
The yields are similar to those obtained with Co,[28e] Pd[28f] and side product 8b was observed and the corresponding amine
Rh[18] systems. The reactions generally yielded a mixture of side product 7b was not detected.[31] The iridium complex 1a
products that was analysed by NMR spectroscopy and GC–MS. clearly has a lower activity than rhodium complex 1b in cata-
lytic ketene formation.
ꢀ-Lactams are medicinally important structures and versatile
Table 4. Catalytic carbene carbonylation reactions studied by using catalyst
precursors 1a and 1b, carbene precursors 4a and 4b, and 4-nitroaniline 5 as
the ketene trapping agent. Besides the desired products 6a and 6b, side
products 7–9 are also formed.
building blocks for the construction of other valuable com-
pounds.[32] A widely applied ꢀ-lactam synthetic method is the
[2+2] Staudinger reaction between a ketene and an imine.[33]
Interestingly, complex 1b can also be used to catalyse the for-
mation of ꢀ-lactams from an in situ generated ketene (using 4b)
and imine 10 in a one-pot synthetic protocol by using similar
conditions to those reported previously (Scheme 7).[18,28e,28f]
Complex 1a shows good activity in this reaction, producing ꢀ-
lactam 11 in 83 % yield, with the trans-configured product be-
ing detected exclusively.[34] The yield is similar to that of the
reported Pd[28f] system but higher than those of the Co[28e] and
Rh systems.[18] Furthermore, the obtained trans-stereoselectivity
is higher than observed with either the Co[28e] or Pd[28f] system.
Entry
4
Catalyst
K2CO3
6
7
8
9
Scheme 7. ꢀ-Lactam 11 synthesis through ketene formation from 4b cata-
lysed by catalyst precursors 1a and 1c.
1
2
3
4
4a
4a
4a
4a
4b
4b
1a
1a
1b
1b
1a
1b
no
yes
no
yes
no
no
3
0
11
78
0
32
14
31
20
0
0
0
0
0
0
9
0
18
0
0
0
Given that methods for enantioselective ꢀ-lactam synthesis
are highly desirable,[35] we were prompted to synthesise a chiral
analogue of 1b for enantioselective catalysis. For this purpose,
we synthesised chiral complex 1c (Figure 5). Complex 1c was
prepared by using Bn*-pma ligand [(R)-1-phenyl-N-(pyridin-2-
5[a]
6
62
0
19
[a] Unidentified side products formed.
Eur. J. Inorg. Chem. 2016, 963–974
www.eurjic.org
969
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ylmethyl)ethanamine], bearing a chiral benzylethyl auxiliary in- the additional ligand donor is proposed to stabilise the corre-
stead of a methyl group in pma. The cation in complex 1c sponding reduced d10 M(–I) ion in a tetrahedral geometry. The
adopts an almost identical square-planar geometry around X-ray crystal structure of 3b and the DFT optimised geometry
rhodium as 1b (Figure 5), with only slightly longer Rh–N2 of 3a further disclose an “aza-allyl” η3-coordination mode of
[Δ = 0.0301(16) Å] and similar Rh–N1 [Δ = –0.0043(17) Å] bonds. the (Me-pma-2H)2– ligand bound to M2. The observations de-
The decreased orbital overlap due to the steric bulk of the scribed in this paper, combined with the previously reported
benzylethyl group at N2 results in a larger (∠N2–Rh–Ct1 angle results, demonstrate a series of intermediate states of metal-
of 98.25(4)° in 1c (vs. 95.66(5)° in 1b). Complex 1c also showed bound picolyl-amines after double deprotonation. The electron-
good catalytic activity to produce ꢀ-lactam 11 with a yield of transfer and charge-delocalisation processes are sensitive to a
63 %. Unfortunately, no ee was detected for this reaction. This number of factors, such as the nature of the sigma-bound metal
could have several reasons: (i) Premature dissociation of the (M1), the nature of the π-coordinated metal M2 acting as a
ketene fragment from the catalyst could be the problem, potential “oxidant”, and the presence or absence of an addi-
(ii) (partial) Bn*-pma ligand dissociation could occur, or (iii) the tional pendant coordinating ligand donor which can stabilise a
chiral moiety is simply too remote from the coordinated ketene tetrahedral M(–I) ion.
for efficient chirality transfer. Further studies on this process are
in progress to unravel the detailed mechanism and prospects
for chirality transfer with related catalysts.
The activity of Ir complex 1a and Rh complex 1b in catalytic
carbene carbonylation reactions was evaluated. Rhodium com-
plex 1c, a chiral analogue of 1b, was also used for this purpose.
Rhodium complex 1b showed much better catalytic activity for
ketene formation from diazo-derived carbenes than Ir complex
1a, with amides and ꢀ-lactams synthesised in good yields via
the intermediacy of metal-bound carbenes. Chiral complex 1c
has an activity similar to that of 1b, but no chirality transfer was
observed in initial attempts to achieve enantioselective one-pot
synthesis of ꢀ-lactams from a carbene precursor, CO and an
imine. Mechanistic studies to elucidate the catalytic process and
to progress catalyst development towards enantioselective ca-
talysis in ketene chemistry are ongoing in our group.
Figure 5. Left: Line drawing of the cation of chiral complex 1c. Right: Molec-
ular structure (ORTEP at 50 % level) of the cation of complex 1c. Hydrogen
atoms are removed for clarity, except for those on C6, N2 and C7. Selected
bond lengths [Å] and angles [°]: Rh–N1, 2.0922(12); Rh–N2, 2.1448(10); Rh–
Ct1, 2.0285(11); Rh–Ct2, 2.0145(9); N1–C5, 1.3586(17); C5–C6, 1.505(2); C6–N2,
1.500(2); N2–C7, 1.5073(17); C15–C16, 1.3879(19); C19–C20, 1.395(2); N1–Rh–
Ct1, 172.12(4); N2–Rh–Ct2, 174.21(4); N1–Rh–N2, 77.60(4); Ct1–Rh–Ct2,
87.47(4); Rh–N2–C6, 101.51(9); Ct1–Ct2–N2–C7, 110.82(9)°.
Experimental Section
General: All the reactions were carried out under argon using
standard Schlenk techniques. Metal precursors [{M(cod)(μ-Cl)}2][36]
and [{M(cod)(μ-OMe)}2],[37] as well as complexes 1,[38] 2b[15b] and
3[15b] were synthesised according to known procedures. The Me-
pma ligand was obtained from Aldrich and used as received. NMR
spectroscopic measurements were recorded with a Bruker AV 400
spectrometer. Chemical shifts are reported and referenced to SiMe4,
using the internal signal of deuterated solvent as reference. HRMS
were obtained with a time-of-flight JEOL AccuTOF LC-plus mass
spectrometer (JMS-T100LP). Elemental analyses were carried out by
Kolbe Mikroanalytisches Labor (Mülheim/Ruhr, Germany).
Conclusions
Rhodium and iridium(cod) cationic complexes 1a and 1b based
on Me-pma ligand were synthesised. Upon reaction with one
equivalent KOtBu, 1a gives rise to quantitative formation of the
neutral amido [Ir(cod)(Me-pma-H)] complex 2a, whereas 1b
leads to the formation of a dinuclear rhodium complex 3b fea-
turing a doubly deprotonated Me-pma-2H ditopic ligand. NMR
spectroscopy and X-ray crystal structure determination confirm
that there is negligible ligand-to-metal charge-transfer from the
In situ Ketene Synthesis from Carbene Carbonylation for Cou-
pling: In a typical carbonylation experiment a stainless steel auto-
clave (150 mL) suitable for seven reaction vessels (equipped with
Teflon mini stirring bars) was used to perform parallel reactions.
Each vial was charged with carbene precursor (EDA 4a or N-tosyl-
hydrazone sodium salt 4b, 0.15 mmol), amine nucleophile 5 or im-
doubly deprotonated Me-pma ligand to Rh2. The negative ine 10 (0.30 mmol), catalyst (0.0075 mmol), internal standard 1,3,5-
charge remains on the (pma-2H)2– fragment, but substantial
charge delocalization into the pyridine ring occurs leading to
trimethoxybenzene (0.015 mmol) and solvent (3 mL). Before start-
ing the catalytic reactions, the charged autoclave was purged (5 ×)
with 20 bar CO and then pressurised to the desired pressure. The
partial pyridine “dearomatisation”. The geometry of the corre-
autoclave was heated in an oil-bath with the vials being stirred
sponding dinuclear iridium complex 3a is similar to that of 3b,
according to DFT studies. The collected data support the previ-
overnight. After the reaction time, the autoclave was cooled to 0 °C
and vented to 1 atm. Around 1.5 mL of each reaction mixture was
ous assumption that coordination of an additional ligand donor
taken out, the solvent was removed by using rotary evaporation,
to M2 is necessary to induce formal reduction of M(+I) to M(–I)
the residue was dissolved in CDCl3 or CH2Cl2 (and filtered when
necessary) for 1H NMR and GC–MS analysis, respectively. The con-
through electron transfer of the dianionic π-coordinated “1,2-
diyldiamide” (N–Py–C=C–N-amido) moiety of the ligand to M2
version and yield for each reaction were calculated based on the
1
to form a neutral imine (NPy–C-C=N–) ligand moiety, wherein integration of H NMR signals of corresponding species relative to
Eur. J. Inorg. Chem. 2016, 963–974
www.eurjic.org
970
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
that of the internal standard (1,3,5-trimethoxybenzene). Product as- (278 mg, 0.418 mmol, 1 equiv.) in diethyl ether (14 m), giving a
signment followed reported assignments.[28e,28f]
deep-red colour. After stirring for 20 min, the solvent was concen-
trated to around 8 mL by evaporation under vacuum, layered with
15 mL of pentane and stored at –4 °C overnight. The mother liquor
was decanted and the solid was washed with pentane and dried.
1H NMR (400 MHz, C6D6): δ = 7.72 (dt, J = 5.9, 1.1 Hz, 1 H, Py), 6.66
(dd, J = 7.6, 1.5 Hz, 1 H, Py), 6.61 (d, J = 7.4 Hz, 1 H, py), 6.14 (tm,
J = 7.5 Hz, 1 H, Py), 4.34 (s, 2 H, cod-vinyl), 3.58–3.52 (m, 2 H, cod-
vinyl), 3.19 (s, 3 H, CH3), 3.02 (s, 1 H, PyCH=N), 2.95 (br., 2 H, cod-
allyl), 2.53 (br., 4 H, cod-allyl), 2.21–2.08 (m, 4 H, cod-allyl), 1.99 (d,
J = 8.0 Hz, 4 H, cod-allyl), 1.41–1.28 (m, 4 H, cod-allyl) ppm. 13C{1H}
NMR (100 MHz, C6D6): δ = 170.95 (CPy), 145.63 (CPy), 135.16 (CPy),
121.52 (CPy), 119.78 (CPy), 89.82 (Py-CH=N), 64.52 (Ccod–vinyl), 59.09
(Ccod–vinyl), 58.59 (Ccod–vinyl), 57.18 (Ccod–vinyl), 54.14 (Ccod–vinyl), 42.11
(CH3), 36.23 (Ccod–allyl), 34.58 (Ccod–allyl), 33.97 (Ccod–allyl), 32.83
(Ccod–allyl), 32.69 (Ccod–allyl), 31.87 (Ccod–allyl) ppm. Correct mass data
could not be obtained possibly due to the instability of the complex
under the applied conditions.
Complex 1a: The ligand Me-pma (93.8 μL, 0.763 mmol, 2.05 equiv.)
was added to
a solution of [{Ir(cod)(μ-OMe)}2] (250.0 mg,
0.372 mmol, 1 equiv.) in methanol (10 mL), followed by addition of
NH4PF6 (121.3 mg, 0.744 mmol, 2 equiv.). The solution was stirred
for 1 h and then concentrated to 5 mL under vacuum. The yellow
precipitate was collected by filtration and washed with a small
amount of cold methanol and dried in vacuo, yield 367.8 mg (87 %).
1H NMR (400 MHz, CD2Cl2): δ = 8.09–8.01 (m, 2 H, Py), 7.74 (d, J =
7.9 Hz, 1 H, Py), 7.49 (ddd, J = 7.4, 5.7, 1.4 Hz, 1 H, Py), 4.71–4.57
(m, 2 H, Py-CH2N, NH), 4.25–4.12 (m, 3 H, Py-CH2N, cod-vinyl), 4.11–
4.02 (m, 2 H, cod-vinyl), 2.67 (d, J = 5.9 Hz, 3 H, CH3), 2.46–2.19 (m,
4 H, cod-allyl), 2.05–1.87 (m, 2 H, cod-allyl), 1.87–1.59 (m, 2 H, cod-
allyl) ppm. 13C{1H} NMR (100 MHz, CD2Cl2): δ = 163.39 (CPy), 148.15
(CPy), 141.46 (CPy), 125.33 (CPy), 123.23 (CPy), 70.69 (Ccod–vinyl), 69.05
(Ccod–vinyl), 68.36 (Ccod–vinyl), 67.72 (Ccod–vinyl), 62.10 (Py-CH2N), 40.05
(CH3), 32.40 (Ccod–allyl), 31.22 (Ccod–allyl), 31.09 (Ccod-allyl), 30.09
(Ccod–allyl) ppm. HRMS (CSI, 243 K, CH2Cl2): m/z calcd. for [M]+
423.1407; found 423.1498. C15H22F6IrN2P (567.5): calcd. C 31.74, H
3.91, N 4.94; found C 31.66, H 3.94, N 4.96.
Complex 3b: The ligand Me-pma (63.4 μL, 0.516 mmol, 1 equiv.)
was added to a suspension of [{Rh(cod)(μ-OMe)}2] (250 mg,
0.615 mmol, 1 equiv.) in diethyl ether (15 mL). The colour changed
to orange and then to deep red. After stirring for 20 min, the sol-
vent was concentrated to around 5 mL by evaporation under vac-
uum, layered with 10 mL of pentane and stored in the fridge
(–4 °C) overnight. The mother liquor was decanted and the solid
was washed with pentane and dried under vacuum, yield 172.8 mg
1
Complex 1b: Synthesised as described for 1a, yield 81 %. H NMR
(400 MHz, CD2Cl2): δ = 7.94 (td, J = 7.7, 1.5 Hz, 1 H, Py), 7.72 (dd,
J = 5.7, 1.3 Hz, 1 H, Py), 7.56 (d, J = 7.9, 1.2 Hz, 1 H, Py), 7.41 (ddd,
J = 7.2, 5.6, 1.3 Hz, 1 H, Py), 4.59–4.42 (m, 2 H, Py-CH2N, cod-vinyl),
4.39–4.18 (m, 3 H, cod-vinyl), 3.99 (dd, J = 15.9, 5.9 Hz, 1 H, Py-
CH2N), 3.82–3.70 (br., 1 H, NH), 2.66–2.37 (m, 4 H, cod-allyl), 2.51 (d,
J = 6.2 Hz, 3 H, CH3), 2.17–1.87 (m, 4 H, cod-allyl) ppm. 13C{1H} NMR
(100 MHz, CD2Cl2): δ = 161.60 (CPy), 147.88 (CPy), 140.59 (CPy), 124.78
(CPy), 123.35 (CPy), 85.03 (d, J = 12.4 Hz, Ccod–vinyl), 83.79 (d, J =
12.7 Hz, Ccod–vinyl), 83.69 (d, J = 6.2 Hz, Ccod–vinyl), 83.57 (d, J = 5.8 Hz,
1
(62 %). H NMR (400 MHz, C6D6): δ = 6.56 (dd, J = 6.5, 1.2 Hz, 1 H,
py), 6.47 (ddd, J = 8.9, 6.4, 1.4 Hz, 1 H, py), 6.16 (dt, J = 8.8, 1.2 Hz,
1 H, py), 5.68 (td, J = 6.4, 1.4 Hz, 1 H, py), 5.16 (dd, J = 0.8, 0.8 Hz,
1 H, PyCH=N), 4.98–3.76 (br., 5 H, cod-vinyl), 4.27–4.19 (m, 1 H, cod-
vinyl), 3.74–3.63 (m, 2 H, cod-vinyl), 3.47–3.38 (m, 1 H, cod-vinyl),
2.78–2.66 (m, 1 H, cod-allyl), 2.65–2.54 (m, 1 H, cod-allyl), 2.58 (t, J =
1.1 Hz, 3 H, CH3), 2.38–2.19 (m, 5 H, cod-allyl), 2.19–2.10 (m, 1 H,
cod-allyl), 2.05–1.92 (m, 4 H, cod-allyl), 1.85–1.71 (m, 4 H, cod-
allyl) ppm. 13C{1H} NMR (100 MHz, C6D6): δ = 141.99 (CPy), 130.93
(d, J = 1.6 Hz, CPy), 128.24 (CPy), 114.54 (CPy), 108.03 (CPy), 89.78 (d,
J = 5.5 Hz, Py-CH=N), 77.78 (d, J = 13.4 Hz, Ccod–vinyl), 75.80 (d, J =
12.8 Hz, Ccod–vinyl), 74.26 (d, J = 12.2 Hz, Ccod–vinyl), 72.06 (d, J =
12.5 Hz, Ccod–vinyl), 43.12 (d, J = 1.7 Hz, CH3), 33.58 (Ccod–allyl), 32.06
(Ccod–allyl), 31.30 (Ccod–allyl), 30.14 (Ccod–allyl), 29.15 (Ccod–allyl) ppm.
HRMS (CSI, 243 K, C6D6): m/z calcd. for [M–H]+ 541.0592; found
541.0626.
Ccod–vinyl), 60.83 (Py-CH2N), 39.16 (CH3), 31.45 (Ccod–allyl), 30.57
(Ccod–allyl), 31.46 (Ccod–allyl), 29.79 (Ccod–allyl) ppm. HRMS (CSI, 243 K,
CH2Cl2): m/z calcd. for [M]+ 333.0833; found 333.0865.
C15H22F6N2PRh (478.2): calcd. C 37.67, H 4.64, N 5.86; found C 37.08,
H 4.63, N 5.72.
Complex 2a. Method I: Potassium tert-butoxide (KOtBu, 3.3 mg,
29.1 μmol, 1.1 equiv.) was added to a yellow solution of 1a
(15.0 mg, 26.4 μmol, 1 equiv.) in [D8]THF (0.6 mL) in a Schlenk vessel
whilst stirring, which resulted in an immediately colour change to
deep-red. The solution was transferred to a sealed NMR tube and
directly analysed by NMR spectroscopy.
Complex 1c: Synthesised as described for 1a, yield 56 %. Surpris-
ingly, complex 1c is in equilibrium with another unidentified spe-
cies with a ratio of ca. 1:0.6 in CD2Cl2 (Figure S20,21), whereas addi-
tion of half the volume of CD3CN resulted in the sole existence of
1c. 1H NMR (400 MHz, CD2Cl2, 1/2 CD3CN): δ = 7.84 (td, J = 7.8,
1.5 Hz, 1 H, HPy), 7.57 (d, J = 5.6 Hz, 1 H, HPy), 7.47–7.38 (m, 1 H, 2
H, HPy, HPh), 7.34–7.22 (m, 1 H, 3 H, HPy, HPh), 4.25 (br., 4 H,
Hcod–vinyl), 4.16–3.89 (br., 1 H, 2 H, NH, Py–CHHN), 3.67 [q, J = 7.0 Hz,
1 H, NCH(Me)Ph], 2.43 (ddt, J = 14.0, 9.8, 6.9 Hz, 2 H, Hcod–allyl), 2.27
(br., 2 H, Hcod–allyl), 1.94 (br., 2 H, Hcod–allyl), 1.812.27 (br., 2 H,
Hcod–allyl), 1.61 (d, J = 6.9 Hz, 3 H, Me) ppm. 13C{1H} NMR (100 MHz,
CD2Cl2, 1/2 CD3CN): δ = 162.01 (CPy), 147.75 (CPy), 139.91 (CPy),
139.71 (CPh), 129.12 (CPh), 128.99 (CPh), 127.91 (CPh), 124.44 (CPy),
123.13 (CPy), 83.42–82.44 (m, Ccod–vinyl), 60.98 (s, PhCHMe), 54.90–
54.67 (m, PyCH2N), 30.48 (Ccod–allyl), 29.81 (Ccod–allyl), 19.72
(CH3) ppm. HRMS (ESI, 243 K, CH2Cl2): m/z calcd. for [M]+ 423.1308;
found 423.1321.
Method II: The ligand Me-pma (51.4 μL, 0.418 mmol, 1 equiv.) was
added to a yellow suspension of [{Ir(cod)(μ-OMe)}2] (278 mg,
0.418 mmol, 1 equiv.) in diethyl ether (14 mL), which resulted in a
colour change to deep-red. After stirring for 20 min, the solvent
was concentrated to around 8 mL by evaporation under vacuum,
layered with 15 mL of pentane and stored at –4 °C overnight. The
mother liquor was decanted and the solid was washed with pent-
ane and dried in vacuo. 1H NMR (400 MHz, [D8]THF): δ = 8.01 (d,
J = 5.9 Hz, 1 H, Py), 7.78 (dd, J = 7.8, 1.5 Hz, 1 H, Py), 7.65 (d, J =
7.6 Hz, 1 H, Py), 7.19 (ddd, J = 7.6, 6.2, 1.3 Hz, 1 H, Py), 4.57 (s, 2 H,
Py-CH2N), 3.74–3.69 (m, 2 H, cod-vinyl), 2.93 (s, 3 H, CH3), 2.71–2.65
(m, 2 H, cod-vinyl), 2.26–2.12 (m, 4 H, cod-allyl), 1.75–1.61 (m, 4 H,
cod-allyl) ppm. 13C{1H} NMR (100 MHz, CD2Cl2): δ = 172.04 (CPy),
146.45 (CPy), 136.78 (CPy), 122.67 (CPy), 120.73 (CPy), 73.49 (Py-CH2N),
63.01 (Ccod–vinyl), 53.31 (Ccod–vinyl), 42.01 (CH3), 33.26 (Ccod–allyl), 32.52
(Ccod–allyl) ppm. HRMS (CSI, 243 K, THF): m/z calcd. for [M]+ 423.1407;
found 423.1406.
X-ray Crystal Structure Determination of 1a: [C15H22IrN2](PF6);
567.52; dark-yellow block; 0.49 × 0.46 × 0.33 mm3; monoclinic;
P21/c (no. 14); a = 6.4624(2), b = 15.9677(4), c = 17.1755(4) Å, ꢀ =
Complex 3a: The ligand Me-pma ligand (51.4 μL, 0.418 mmol,
1 equiv.) was added to a yellow suspension of [{(cod)Ir(μ-OMe)}2]
Eur. J. Inorg. Chem. 2016, 963–974
www.eurjic.org
971
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
101.2505(10)°; V = 1738.28(8) Å3; Z = 4; Dx = 2.169 g/cm3; μ = atoms were refined with a riding model. 180 Parameters were re-
7.83 mm–1. 25066 Reflections were measured with a Bruker Kappa
ApexII diffractometer with sealed tube and Triumph monochroma-
tor (λ = 0.71073 Å) at a temperature of 150(2) K up to a resolution
of (sin θ/λ)max = 0.65 Å–1. The intensities were integrated with the
Saint software.[39] Multiscan absorption correction and scaling was
performed with SADABS[40] (correction range 0.26–0.43). 4020 Re-
flections were unique (Rint = 0.020), of which 3895 were observed
[I > 2σ(I)]. The structure was solved with Direct Methods using
SHELXS-97.[41] Least-squares refinement was performed with
SHELXL-97[41] against F2 of all reflections. Non-hydrogen atoms
were refined freely with anisotropic displacement parameters. All
hydrogen atoms were located in difference Fourier maps. The N–H
hydrogen atom and the four C–H hydrogen atoms at the coordi-
nated double bonds were refined freely with isotropic displacement
parameters. All other hydrogen atoms were refined with a riding
model. 247 Parameters were refined with no restraints. R1/wR2
[I > 2σ(I)]: 0.0146/0.0338. R1/wR2 [all reflections]: 0.0154/0.0340, S =
1.160. Residual electron density between –0.74 and 0.88 e/Å3. Ge-
ometry calculations and checking for higher symmetry was per-
formed with the PLATON program.[42]
fined with no restraints. R1/wR2 [I > 2σ(I)]: 0.0128/0.0260. R1/wR2 [all
reflections]: 0.0157/0.0265, S = 1.057. Residual electron density be-
tween –0.88 and 0.73 e/Å3. Geometry calculations and checking for
higher symmetry was performed with the PLATON program.[42]
X-ray Crystal Structure Determination of 3b: C23H32N2Rh2;
542.33; red needle; 0.32 × 0.12 × 0.08 mm3; triclinic; P1¯ (no. 2); a =
8.5096(3), b = 10.4024(3), c = 12.9938(4) Å, α = 78.8723(11), ꢀ =
79.4202(11), γ = 66.3169(10)°; V = 1026.31(6) Å3; Z = 2; Dx = 1.755 g/
cm3; μ = 1.62 mm–1. 17837 Reflections were measured with a Bruker
Kappa ApexII diffractometer with sealed tube and Triumph mono-
chromator (λ = 0.71073 Å) at a temperature of 150(2) K up to a
resolution of (sin θ/λ)max = 0.65 Å–1. The intensities were integrated
with the SAINT software.[39] Multiscan absorption correction and
scaling was performed with SADABS[40] (correction range 0.67–
0.75). 4683 Reflections were unique (Rint = 0.015), of which 4343
were observed [I > 2σ(I)]. The structure was solved with Direct
Methods using SIR-97.[45] Least-squares refinement was performed
with SHELXL-97[41] against F2 of all reflections. Non-hydrogen atoms
were refined freely with anisotropic displacement parameters. All
hydrogen atoms were located in difference Fourier maps. The four
C–H hydrogen atoms at the coordinated double bonds were refined
freely with isotropic displacement parameters. All other hydrogen
atoms were refined with a riding model. 180 Parameters were re-
fined with no restraints. R1/wR2 [I > 2σ(I)]: 0.0168/0.0419. R1/wR2 [all
reflections]: 0.0188/0.0430, S = 1.026. Residual electron density be-
tween –0.51 and 0.79 e/Å3. Geometry calculations and checking for
higher symmetry was performed with the PLATON program.[42]
X-ray Crystal Structure Determination of 1b: [C15H22N2Rh](PF6);
478.23; yellow-green block; 0.30 × 0.09 × 0.09 mm3; monoclinic;
P21/c (no. 14); a = 6.5163(2), b = 16.0417(4), c = 17.1102(5) Å, ꢀ =
101.2918(10)°; V = 1753.95(9) Å3; Z = 4; Dx = 1.811 g/cm3; μ =
1.13 mm–1. 33645 Reflections were measured with a Bruker Kappa
ApexII diffractometer with sealed tube and Triumph monochroma-
tor (λ = 0.71073 Å) at a temperature of 150(2) K up to a resolution
of (sin θ/λ)max = 0.65 Å–1. The intensities were integrated with the
Saint software.[40] Multiscan absorption correction and scaling was
performed with SADABS[40] (correction range 0.67–0.75). 4050 Re-
flections were unique (Rint = 0.018), of which 3772 were observed
[I > 2σ(I)]. The structure was solved with Direct Methods using
SHELXS-97.[41] Least-squares refinement was performed with
SHELXL-97[41] against F2 of all reflections. Non-hydrogen atoms
were refined freely with anisotropic displacement parameters. All
hydrogen atoms were located in difference Fourier maps. The N–H
hydrogen atom and the four C–H hydrogen atoms at the coordi-
nated double bonds were refined freely with isotropic displacement
parameters. All other hydrogen atoms were refined with a riding
model. 247 Parameters were refined with no restraints. R1/wR2
[I > 2σ(I)]: 0.0174/0.0436. R1/wR2 [all reflections]: 0.0194/0.0446, S =
1.041. Residual electron density between –0.34 and 0.66 e/Å3. Ge-
ometry calculations and checking for higher symmetry was per-
formed with the PLATON program.[42]
X-ray Crystal Structure Determination of 1c: C22H28N2Rh·CH3OH;
600.39; orange needle; 0.57 × 0.31 × 0.13 mm3; triclinic; P1 (no. 1);
a = 7.48444(17), b = 9.5487(2), c = 9.6905(2) Å, α = 63.452(1), ꢀ =
78.364(1), γ = 76.874(1)°; V = 599.34(2) Å3; Z = 1; Dx = 1.663 g/cm3;
μ = 0.84 mm–1. 20838 Reflections were measured with a Bruker
Kappa ApexII diffractometer with sealed tube and Triumph mono-
chromator (λ = 0.71073 Å) at a temperature of 150(2) K up to a
resolution of (sin θ/λ)max = 0.65 Å–1. The intensities were integrated
with the Eval15 software.[43] Multiscan absorption correction and
scaling was performed with SADABS[40] (correction range 0.67–
0.75). 5430 Reflections were unique (Rint = 0.013), of which 5429
were observed [I > 2σ(I)]. The structure was solved with Direct
Methods using SHELXS-97.[41] Least-squares refinement was per-
formed with SHELXL-97[41] against F2 of all reflections. Non-
hydrogen atoms were refined freely with anisotropic displacement
parameters. All hydrogen atoms were located in difference Fourier
maps. The N–H hydrogen and the hydrogen atoms of the cycloocta-
diene double bonds were refined freely with isotropic displacement
parameters. All other hydrogen atoms were refined with a riding
model. 334 Parameters were refined with three restraints (for fixing
the origin). R1/wR2 [I > 2σ(I)]: 0.0104/0.0275. R1/wR2 [all reflections]:
0.0104/0.0276, S = 1.061. Flack parameter[46] x = –0.001(8). Residual
electron density between –0.37 and 0.19 e/Å3. Geometry calcula-
tions and checking for higher symmetry was performed with the
PLATON program.[42]
X-ray Crystal Structure Determination of 2a: C15H21IrN2; 421.54;
red block; 0.21 × 0.17 × 0.16 mm3; monoclinic; P21/c (no. 14); a =
9.2085(3), b = 11.6486(3), c = 13.1155(4) Å, ꢀ = 109.281(1)°; V =
1327.93(7) Å3; Z = 4; Dx = 2.108 g/cm3; μ = 10.04 mm–1. 53739
Reflections were measured with a Bruker Kappa ApexII diffractome-
ter with sealed tube and Triumph monochromator (λ = 0.71073 Å)
at a temperature of 150(2) K up to a resolution of (sin θ/λ)max
=
0.65 Å–1. The intensities were integrated with the Eval15 soft-
ware.[43] Multiscan absorption correction and scaling was per-
formed with SADABS[40] (correction range 0.29–0.44). 5809 Reflec-
tions were unique (Rint = 0.025), of which 5267 were observed
[I > 2σ(I)]. The structure was solved with automated Patterson meth-
ods using DIRDIF-08.[44] Least-squares refinement was performed
with SHELXL-97[41] against F2 of all reflections. Non-hydrogen atoms
were refined freely with anisotropic displacement parameters. All
hydrogen atoms were located in difference Fourier maps. The four
C–H hydrogen atoms at the coordinated double bonds were refined
CCDC Crystallographic Data:
CCDC 1429983 (for 1a), 1429984 (for 1b), 1429985 (for 2a), 1429986
(for 3b), and 1429987 (for 1c) contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre.
DFT Calculations: The gas-phase geometries of complexes 3 and
4 were optimised with the Turbomole program package[47] coupled
freely with isotropic displacement parameters. All other hydrogen to the PQS Baker optimiser[48] via the BOpt package[49] at the ri-
Eur. J. Inorg. Chem. 2016, 963–974
www.eurjic.org
972
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
DFT[50]/BP86[51] level. We used the def2-TZVPP basis set[52] for all
atoms and a small grid (m4), with Grimme's dispersion corrections
(disp3 version).[53] The minima (no imaginary frequencies) were
characterised by calculating the Hessian matrix.
[9]
a) F. R. Keene, M. J. Ridd, M. R. Snow, J. Am. Chem. Soc. 1983, 105, 7075–
7081; b) J. P. Saucedo-Vázquez, V. M. Ugalde-Saldívar, A. R. Toscano,
P. M. H. Kroneck, M. E. Sosa-Torres, Inorg. Chem. 2009, 48, 1214–1222; c)
G. J. Christian, A. Arbuse, X. Fontrodona, M. A. Martinez, A. Llobet, F.
Maseras, Dalton Trans. 2009, 6013–6020.
Supporting Information (see footnote on the first page of this
[10]
J. Gómez, G. García-Herbosa, J. V. Cuevas, A. Arnáiz, A. Carbayo, A. Mu-
ñoz, L. Falvello, P. E. Fanwick, Inorg. Chem. 2006, 45, 2483–2493.
article): NMR spectra, crystallographic details, computational details.
[11]
[12]
G. J. Christian, A. Llobet, F. Maseras, Inorg. Chem. 2010, 49, 5977–5985.
a) W. Kaim, Inorg. Chem. 2011, 50, 9752–9765; b) V. Lyaskovskyy, B.
de Bruin, ACS Catal. 2012, 2, 270–279; c) J. I. van der Vlugt, Eur. J. Inorg.
Chem. 2012, 363–375; d) V. K. K. Praneeth, M. R. Ringenberg, T. R. Ward,
Angew. Chem. Int. Ed. 2012, 51, 10228–10234; Angew. Chem. 2012, 124,
10374; e) O. R. Luca, R. H. Crabtree, Chem. Soc. Rev. 2013, 42, 1440–1459;
f) D. L. J. Broere, R. Plessius, J. I. van der Vlugt, Chem. Soc. Rev. 2015, 44,
6886–6915.
Acknowledgments
This research was funded by the Netherlands Research Council
– Chemical Sciences (NWO-CW) through an ECHO research
grant. The Spanish Ministerio de Economía y Competitividad
(MINECO) and Fondos Europeos para el Desarrollo Regional
(FEDER) (CTQ2014-53033-P) is also acknowledged.
[13]
a) B. de Bruin, T. P. J. Peters, N. N. F. A. Suos, D. Gelder, J. M. M. Smits,
A. W. Gal, Inorg. Chim. Acta 2002, 337, 154–162; b) D. G. H. Hetterscheid,
B. de Bruin, J. M. M. Smits, A. W. Gal, Organometallics 2003, 22, 3022–
3024; c) D. G. H. Hetterscheid, M. Klop, R. J. N. A. M. Kicken, J. M. M.
Smits, E. J. Reijerse, B. de Bruin, Chem. Eur. J. 2007, 13, 3386–3405; d) C.
Tejel, M. A. Ciriano, M. P. del Río, D. G. H. Hetterscheid, N. Tsichlis i Spithas,
J. M. M. Smits, B. de Bruin, Chem. Eur. J. 2008, 14, 10932–10936; e) C.
Tejel, M. P. del Río, M. A. Ciriano, E. J. Reijerse, F. Hartl, S. Záliš, D. G. H.
Hetterscheid, N. Tsichlis i Spithas, B. de Bruin, Chem. Eur. J. 2009, 15,
11878–11889.
Keywords: Carbenes · Iridium · Ketene · Ligand design ·
Redox chemistry · Rhodium
[1] a) J. Choi, A. H. R. MacArthur, M. Brookhart, A. S. Goldman, Chem. Rev.
2011, 111, 1761–1779; b) G. E. Dobereiner, R. H. Crabtree, Chem. Rev.
2010, 110, 681–703; c) I. W. C. E. Arends, R. A. Sheldon, in: Modern Oxid-
ation Methods (Ed.: J.-E. Bäckvall), Wiley-VCH, Weinheim, Germany, 2004,
p. 83–118; d) M. Beller, C. Bolm, in: Transition Metals for Organic Synthesis,
2nd ed., (Ed.: S.-I. Murahashi, Y. Imada), Wiley-VCH, Weinheim, Germany,
2004, 2, p. 497–507; e) J. Finney, H.-J. Moon, T. Ronnebaum, M. Lantz, M.
Mure, Arch. Biochem. Biophys. 2014, 546, 19–32; f) D. Rokhsana, E. M.
Shepard, D. E. Brown, D. M. Dooley in: Copper-Oxygen Chemistry (Eds.: S.
Rokita, K. D. Karlin, S. Itoh), Wiley, Hoboken, NJ, USA, 2011, chapter 3, p.
53–106.
[14]
[15]
a) T. Büttner, J. Geier, G. Frison, J. Harmer, C. Calle, A. Schweiger, H.
Schönberg, H. Grützmacher, Science 2005, 307, 235–238; b) P. Maire, M.
Konigsmann, A. Sreekanth, J. Harmer, A. Schweiger, H. Grützmacher, J.
Am. Chem. Soc. 2006, 128, 6578–6580; c) Y. Miyazato, T. Wada, K. Tanaka,
Bull. Chem. Soc. Jpn. 2006, 79, 745–747; d) Y. Miyazato, T. Wada, J. T.
Muckerman, E. Fujita, K. Tanaka, Angew. Chem. Int. Ed. 2007, 46, 5728–
5730; Angew. Chem. 2007, 119, 5830.
a) C. Tejel, M. A. Ciriano, M. P. del Río, F. J. van den Bruele, D. G. H.
Hetterscheid, N. Tsichlis i Spithas, B. de Bruin, J. Am. Chem. Soc. 2008,
130, 5844–5845; b) C. Tejel, M. P. del Río, L. Asensio, F. J. van den Bruele,
M. A. Ciriano, N. Tsichlis i Spithas, D. G. H. Hetterscheid, B. de Bruin, Inorg.
Chem. 2011, 50, 7524; c) C. Tejel, L. Asensio, M. P. del Río, M. A. Ciriano,
B. de Bruin, J. A. López, M. A. Ciriano, Eur. J. Inorg. Chem. 2012, 512; d)
C. Tejel, L. Asensio, M. A. Ciriano, J. A. López, B. de Bruin, M. P. del Río,
Angew. Chem. Int. Ed. 2011, 50, 8839; Angew. Chem. 2011, 123, 9001; e)
Z. Tang, E. Otten, J. N. H. Reek, J. I. van der Vlugt, B. de Bruin, Chem. Eur.
J. 2015, 21, 12683–12693.
a) J. I. van der Vlugt, J. N. H. Reek, Angew. Chem. Int. Ed. 2009, 48, 8832–
8846; Angew. Chem. 2009, 121, 8990; b) J. R. Khusnutdinova, D. Milstein,
Angew. Chem. Int. Ed. 2015, 54, 12236–12273. For specific examples with
Rh and Ir, see: c) M. Feller, Y. Diskin-Posner, L. J. W. Shimon, E. Ben-Ari,
D. Milstein, Organometallics 2012, 31, 4083–4101; d) M. Feller, A. Karton,
G. Leitus, J. M. L. Martin, D. Milstein, J. Am. Chem. Soc. 2006, 128, 12400–
12401; e) M. Feller, E. Ben-Ari, Y. Diskin-Posner, R. Carmieli, L. Weiner, D.
Milstein, J. Am. Chem. Soc. 2015, 137, 4634–4637. See also: f) Y. Gloa-
guen, C. Rebreyend, M. Lutz, P. Kumar, M. Huber, J. I. van der Vlugt, S.
Schneider, B. de Bruin, Angew. Chem. Int. Ed. 2014, 53, 6814–6818;
Angew. Chem. 2014, 126, 6932.
[2] a) C. Binda, A. Mattevi, D. E. Edmondson, J. Biol. Chem. 2002, 277, 23973–
23976; b) N. S. Scrutton, Nat. Prod. Rep. 2004, 21, 722–730; c) D. E. Ed-
mondson, C. Binda, A. Mattevi, Arch. Biochem. Biophys. 2007, 464, 269–
276.
[3] a) G. J. Christian, A. Llobet, F. Maseras, Inorg. Chem. 2010, 49, 5977–5985;
b) J. P. Saucedo-Vázquez, V. M. Ugalde-Saldívar, A. R. Toscano, P. M. H.
Kroneck, M. E. Sosa-Torres, Inorg. Chem. 2009, 48, 1214–1222; c) G. J.
Christian, A. Arbuse, X. Fontrodona, M. A. Martinez, A. Llobet, F. Maseras,
Dalton Trans. 2009, 6013–6020; d) B. Zhu, M. Lazar, B. G. Trewyna, R. J.
Angelici, J. Catal. 2008, 260, 1–6; e) A. Machkour, D. Mandon, M. Lachkar,
R. Welter, Eur. J. Inorg. Chem. 2005, 158–161; f) V. Amendola, L. Fabbrizzi,
E. Mundum, P. Pallavicini, Dalton Trans. 2003, 773–774; g) X. Gu, W. Chen,
D. Morales-Morales, C. M. Jensen, J. Mol. Catal. A 2002, 189, 119–124; h)
F. R. Keene, Coord. Chem. Rev. 1999, 187, 121–149.
[16]
[4] C. Zhang, N. Jiao, Angew. Chem. Int. Ed. 2010, 49, 6174–6177; Angew.
Chem. 2010, 122, 6310.
[5] a) S. Aiki, A. Taketoshi, J. Kuwabara, T. Koizumi, T. Kanbara, J. Organomet.
Chem. 2011, 696, 1301–1304; b) A. Taketoshi, T. Koizumi, T. Kanbara,
Tetrahedron Lett. 2010, 51, 6457–6459; c) W. H. Bernskoetter, M. Broo-
khart, Organometallics 2008, 27, 2036–2045; d) K. Yamaguchi, N. Mizuno,
Angew. Chem. Int. Ed. 2003, 42, 1480–1483; Angew. Chem. 2003, 115,
1518.
[6] a) R. E. Rodríguez-Lugo, M. Trincado, M. Vogt, F. Tewes, G. Santiso-Quino-
nes, H. Grützmacher, Nat. Chem. 2013, 5, 342–347; for a reversed exam-
ple of active catalyst formation involving imine-to-amine-type ligand
transformation, see: b) A. Mikhailine, M. I. Maishan, A. J. Lough, R. H.
Morris, J. Am. Chem. Soc. 2012, 134, 12266–12280.
[17]
For oxidative loss of H2 in PNP ligands with aliphatic backbones, see: a)
M. Käß, A. Friedrich, M. Drees, S. Schneider, Angew. Chem. Int. Ed. 2009,
48, 905–907; Angew. Chem. 2009, 121, 922; b) A. Friedrich, M. Drees, M.
Käss, E. Herdtweck, S. Schneider, Inorg. Chem. 2010, 49, 5482–5494.
Z. Tang, S. Mandal, N. D. Paul, M. Lutz, P. Li, J. I. van der Vlugt, B. de Bruin,
Org. Chem. Front. 2015, 2, 1561–1577.
A triplet-like doublet of doublet with two small J constants close in value
(1.3–1.5 Hz), presumably one is 3JRh,H and the second is 3JH,H correspond-
ing to coupling with the C–H of the methyl group.
[18]
[19]
[7] See refs.[5c,5d] and: a) C. Jun, K. Chung, J. Hong, Org. Lett. 2001, 3, 301–
303; b) O. Saidi, A. J. Blacker, M. M. Farah, S. P. Marsden, J. M. J. Williams,
Angew. Chem. Int. Ed. 2009, 48, 7375–7378; Angew. Chem. 2009, 121,
7511; c) W. P. Gri, B. Reddy, A. G. F. Shoair, M. Suriaatmaja, A. J. P. White,
D. J. Williams, J. Chem. Soc., Dalton Trans. 1998, 2, 2819–2825.
[8] a) K. T. Tseng, A. M. Rizzi, N. K. Szymczak, J. Am. Chem. Soc. 2013, 135,
16352–16355; b) D. Talwar, A. Gonzalez-de-Castro, H. Y. Li, J. Xiao, Angew.
Chem. Int. Ed. 2015, 54, 5223–5227; Angew. Chem. 2015, 127, 5312; c) S.
Muthaiah, S. H. Hong, Adv. Synth. Catal. 2012, 354, 3045–3053.
[20]
[21]
C. Tejel, J. M. Villoro, M. A. Ciriano, J. A. López, E. Eguizabal, F. J. Lahoz,
V. I. Bakhmutov, L. A. Oro, Organometallics 1996, 15, 2967–2978.
The short distance of Rh2–C6 may indicate a σ-bond between the two
atoms and thus an sp3 carbon, but an almost square-planar bonding
geometry was found around C6 with N2, H6 and C5 [∠N2–C6–C5
116.36(14)°, ∠C5–C6–H6 119.7(12)°, ∠H6–C6–N2 123.1(12)°,
Σ =
359.2(17)°], pointing to an sp2-hybridized carbon atom.
Eur. J. Inorg. Chem. 2016, 963–974
www.eurjic.org
973
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[22] A more localised charge distribution is also found on the ligand in 3b
than in C2, and the Rh2 in 3b coordinates to an N2–C6–C5 “azaallyl”
moiety (Figure 3, III), which is different from that in complex C2. The
rhodium (Rh1 in 3b) is less capable to induce and stabilise a charge-
delocalised M–N2–C6–C5–N1 metallocycle than iridium (Ir1 in C2). The
longer N1=C5 and N2=C6 bond lengths in 3b compared with those in
C2 reflect this difference (N1=C5: 1.385(2) vs. 1.394(5), N2=C6: 1.387(2)
vs. 1.396(5) Å, for 1b and C2, respectively).
[23] a) S. Takemoto, S. Otsuki, Y. Hashimoto, K. Kamikawa, H. Matsuzaka, J.
Am. Chem. Soc. 2008, 130, 8904–905; b) S.-Y. Liu, I. D. Hills, G. C. Fu,
Organometallics 2002, 21, 4323–4325; c) A. S. Batsanov, V. G. Andrianov,
Y. T. Struchkov, V. S. Khandkarova, A. Z. Rubezhov, Metalloorg. Khim.
1988, 1, 68.
[24] a) H. Staudinger, Ber. Dtsch. Chem. Ges. 1905, 38, 1735–1739; b) T. T.
Tidwell, Eur. J. Org. Chem. 2006, 563–576; c) D. H. Paull, A. Weatherwax,
T. Lectka, Tetrahedron 2009, 65, 6771–6803; d) A. D. Allen, T. T. Tidwell,
Eur. J. Org. Chem. 2012, 1081–1096; e) A. D. Allen, T. T. Tidwell, Chem.
Rev. 2013, 113, 7287–7342.
[25] a) N. Ungvári, F. Ungváry, Carbonylation of Diazoalkanes, in: Modern Carb-
onylation Methods (Ed.: L. Kollár), Wiley-VCH, Weinheim, Germany, 2008,
chapter 8, p. 199–221; b) K. H. Dötz, J. Stendel, Chem. Rev. 2009, 109,
3227–3274.
[26] Conventional synthetic methods generally have limitations such as nar-
row substrate scope, strictly controlled conditions and generation of det-
rimental side products, such as thermolysis of ketene dimers, see: a)
H. W. Moore, D. S. Wilbur, J. Org. Chem. 1980, 45, 4483–4491; for Wolff
rearrangement, see: b) Y. Chiang, A. J. Kresge, V. V. Popik, J. Am. Chem.
Soc. 1999, 121, 5930–5932; c) B. D. Wagner, B. R. Arnold, G. S. Brown, J.
Lusztyk, J. Am. Chem. Soc. 1998, 120, 1827–1834; for dehydrohalogena-
tion of acyl halide precursors, see: d) L. S. Hegedus, J. Montgomery, Y.
Narukawa, D. C. Snustad, J. Am. Chem. Soc. 1991, 113, 5784–5791.
[27] a) K. H. Dötz, J. Mühlemeier, Angew. Chem. Int. Ed. Engl. 1982, 21, 929–
933; Angew. Chem. 1982, 94, 936; b) P. T. Barger, B. D. Santarsiero, J.
Armantrout, J. E. Bercaw, J. Am. Chem. Soc. 1984, 106, 5178–5186; c) N.
Ungvári, E. Fördo˝s, J. Balogh, T. Kégl, L. Párkányi, F. Ungváry, Organome-
tallics 2010, 29, 3837–3851.
Aragoncillo, Chem. Rev. 2007, 107, 4437–4492; c) T. T. Tidwell, Angew.
Chem. Int. Ed. 2008, 47, 1016–1020; Angew. Chem. 2008, 120, 1032; d)
N. A. A. El-Kanzi, Heterocycl. Lett. 2013, 3, 81–109; e) A. Gupta, A. K. Halve,
Int. J. Pharm. Sci. Res. 2015, 6, 978–987.
[33] a) M. Staudinger, Justus Liebigs Ann. Chem. 1907, 356, 51–123; b) G. I.
Georg, V. T. Ravikumar, in: The Organic Chemistry of ꢀ-Lactams (Ed.: G. I.
Georg), Verlag Chemie, New York, 1993, p. 295–398; c) Palomo, J. M.
Aizpurua, I. Ganboa, M. Oiarbide, Curr. Med. Chem. 2004, 11, 1837–1872;
d) R. Tuba, Org. Biomol. Chem. 2013, 11, 5976–5988.
[34] The cis-trans stereoselectivity has been a less well understood and con-
troversial topic in the Staudinger reaction, and many factors have been
observed to affect the selectivity, see: a) L. S. Hegedus, J. Montgomery,
Y. Narukawa, D. C. Snustad, J. Am. Chem. Soc. 1991, 113, 5784–5791; b)
S. France, A. Weatherwax, A. E. Taggi, T. Lectka, Acc. Chem. Res. 2004, 37,
592–600; c) Y. Wang, Y. Liang, L. Jiao, D. M. Du, J. Xu, J. Org. Chem. 2006,
71, 6983–6990; d) B. K. Banik, B. Lecea, A. Arrieta, A. De Cózar, F. P. Cossío,
Angew. Chem. Int. Ed. 2007, 46, 3028–3032; Angew. Chem. 2007, 119,
3088; e) F. P. Cossío, A. Arrieta, M. a. Sierra, Acc. Chem. Res. 2008, 41,
925–936; f) H. Qi, X. Li, J. Xu, Org. Biomol. Chem. 2011, 9, 2702–2714.
[35] C. R. Pitts, T. Lectka, Chem. Rev. 2014, 114, 7930–7953.
[36] G. Giordano, R. H. Crabtree, Inorg. Synth. 1990, 28, 88–90.
[37] R. Uson, L. A. Oro, J. A. Cabeza, Inorg. Synth. 1985, 23, 126–129.
[38] B. de Bruin, J. A. Brands, J. J. J. M. Donners, M. P. J. Donners, R. de Gelder,
J. M. M. Smits, A. W. Gal, A. L. Spek, Chem. Eur. J. 1999, 5, 2921–2936.
[39] SAINT, Bruker AXS Inc., Madison, Wisconsin, USA, 2007.
[40] G. M. Sheldrick, SADABS, University of Göttingen, Germany, 2008.
[41] G. M. Sheldrick, Acta Crystallogr., Sect. A 2008, 64, 112–122.
[42] A. L. Spek, Acta Crystallogr., Sect. D 2009, 65, 148–155.
[43] A. M. M. Schreurs, X. Xian, L. M. J. Kroon-Batenburg, J. Appl. Crystallogr.
2010, 43, 70–82.
[44] P. T. Beurskens, G. Beurskens, R. de Gelder, S. Garcia-Granda, R. O. Gould,
J. M. M. Smits, DIRDIF-08, Crystallography Laboratory, University of Nij-
megen, The Netherlands, 2008.
[45] A. Altomare, M. C. Burla, M. Camalli, G. L. Cascarano, C. Giacovazzo, A.
Guagliardi, A. G. G. Moliterni, G. Polidori, R. Spagna, J. Appl. Crystallogr.
1999, 32, 115–119.
[28] a) T. Ye, M. A. Mckervey, Chem. Rev. 1994, 94, 1091–1160; b) R. Tuba, F.
Ungváry, J. Mol. Catal. A 2003, 203, 59–67; c) Z. Zhang, J. Wang, Tetrahe-
dron 2008, 64, 6577–6605; d) E. Fördos, R. Tuba, L. Párkányl, T. Kégl, F.
Ungváry, Eur. J. Org. Chem. 2009, 1994–2002; e) N. D. Paul, A. Chirila, H.
Lu, X. P. Zhang, B. de Bruin, Chem. Eur. J. 2013, 19, 12953–12958; f) Z.
Zhang, Y. Liu, L. Ling, Y. Li, Y. Dong, M. Gong, X. Zhao, Y. Zhang, J. Wang,
J. Am. Chem. Soc. 2011, 133, 4330–4341; g) Q. Xiao, Y. Zhang, J. Wang,
Acc. Chem. Res. 2013, 46, 236–247.
[46] H. D. Flack, Acta Crystallogr., Sect. A 1983, 39, 876–881.
[47] R. Ahlrichs, Turbomole, version 6.5, Theoretical Chemistry Group, Univer-
sity of Karlsruhe, Germany.
[48] PQS, version 2.4, Parallel Quantum Solutions, Fayettevile, Arkansas, USA,
2001; the Baker optimizer is available separately from PQS upon request,
see: I. Baker, J. Comput. Chem. 1986, 7, 385–395.
[49] P. H. M. Budzelaar, J. Comput. Chem. 2007, 28, 2226–2236.
[50] M. Sierka, A. Hogekamp, R. Ahlrichs, J. Chem. Phys. 2003, 118, 9136–
9148.
[51] a) A. D. Becke, Phys. Rev. A 1988, 38, 3098–3100; b) J. P. Perdew, Phys.
Rev. B 1986, 33, 8822–8824.
[52] a) F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 2005, 7, 3297–3305;
b) F. Weigend, M. Häser, H. Patzelt, R. Ahlrichs, Chem. Phys. Lett. 1998,
294, 143–152.
[53] S. Grimme, J. Antony, S. Ehrlich, H. Krieg, J. Chem. Phys. 2010, 132,
154104.
[29] C. R. Pitts, T. Lectka, Chem. Rev. 2014, 114, 7930–7953.
[30] a) P. Schwab, N. Mahr, J. Wolf, H. Werner, Angew. Chem. Int. Ed. Engl.
1993, 32, 1480–1482; Angew. Chem. 1993, 105, 1498; b) H. Werner, P.
Schwab, E. Bleuel, N. Mahr, B. Windmüller, J. Wolf, Chem. Eur. J. 2000, 6,
4461–4470.
[31] However, complex 1b does not show dehydrogenation activity for the
related amine substrates under similar conditions, indicating fast de-
hydrogenation of the intermediate amino/amido species 8b might also
derive from other reactions, e.g. transimination between hydrolyzed 4b
and 5.
[32] a) R. B. Morin, M. Gorman, Chemistry and Biology of ꢀ-Lactam Antibiotics,
6th Ed., Academic Press, New York, 1982; b) B. Alcaide, P. Almendros, C.
Received: November 9, 2015
Published Online: January 25, 2016
Eur. J. Inorg. Chem. 2016, 963–974
www.eurjic.org
974
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim