Synthesis of a 14-electron iridium(III) complex with a xanthene-based bis(silyl) chelate ligand (xantsil): A distorted seesaw-shaped fourcoordinate geometry and reactions leading to 16-electron complexes

Article abstract of DOI:10.1016/j.jorganchem.2013.09.009

Synthesis, structure determination, and reactions of a 14-electron four-coordinate iridium(III) complex bearing a xanthene-based bis(silyl) chelate ligand, i.e., Ir{κ²(Si,Si)-xantsil}(PCy3)Cl (1a, xantsil = (9,9-dimethylxanthene-4,5-diyl)bis(dimethylsilyl)), are reported. A precursor of 1a, the 16-electron (dihydrido)iridium(V) complex Ir{κ² (Si,Si)-xantsil}(H)₂(PCy₃)Cl (2), was prepared by the reaction of [IrCl-(coe)₂]₂ (coe = cyclooctene) with 4,5-bis(dimethylsilyl)-9,9-dimethylxanthene (xantsilH₂) and PCy ₃. Dehydrogenation reaction of 2 with 3,3-dimethylbut-1-ene, a hydrogen acceptor, afforded a mixture of 1a and its isomer 1b, abbreviated as 1a + 1b, together with 2,2-dimethylbutane. Single crystals obtained from a CH ₂Cl₂ solution of 1a + 1b contained only isomer 1a. X-ray crystal structure analysis of one of the crystals revealed that 1a adopts a distorted seesaw-shaped four-coordinate geometry where the coordinatively unsaturated metal center is stabilized by weak agostic interaction of two γ-C-H bonds of the PCy₃ ligand. On the other hand, NMR spectroscopic analysis of 1a + 1b demonstrated that 1a is in fast equilibrium with a minor amount of 1b in solution. Replacement of the chloro ligand in complexes 1a + 1b by a triflato ligand with AgOTf (OTf = OSO₂CF ₃) afforded quantitatively a single product, i.e., the 16-electron iridiumetriflato complex Ir{κ³(Si,O,Si)-xantsil}(PCy ₃)(OTf) (3). 1a + 1b and 16-electron complex 2 are interconvertible via oxidative addition of dihydrogen (from 1a + 1b to 2) and alkene hydrogenation (from 2 to 1a + 1b). Complexes 1a + 1b and 2 were found to catalyze hydrogenation of 3,3-dimethylbut-1-ene.

Full text of DOI:10.1016/j.jorganchem.2013.09.009

Journal of Organometallic Chemistry 751 (2014) 686e694
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Synthesis of a 14-electron iridium(III) complex with a xanthene-based
bis(silyl) chelate ligand (xantsil): A distorted seesaw-shaped four-
coordinate geometry and reactions leading to 16-electron complexes
*
Takashi Komuro, Keisuke Furuyama, Takeo Kitano, Hiromi Tobita
Department of Chemistry, Graduate School of Science, Tohoku University, Sendai 980-8578, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Synthesis, structure determination, and reactions of a 14-electron four-coordinate iridium(III) com-
plex bearing
xantsil ¼ (9,9-dimethylxanthene-4,5-diyl)bis(dimethylsilyl)), are reported. A precursor of 1a, the 16-
electron (dihydrido)iridium(V) complex Ir{
2(Si,Si)-xantsil}(H)2(PCy3)Cl (2), was prepared by the re-
a xanthene-based bis(silyl) chelate ligand, i.e., Ir{k
2(Si,Si)-xantsil}(PCy3)Cl (1a,
Received 11 August 2013
Received in revised form
5 September 2013
k
Accepted 6 September 2013
action of [IrCl-(coe)2]2 (coe ¼ cyclooctene) with 4,5-bis(dimethylsilyl)-9,9-dimethylxanthene (xant-
silH2) and PCy3. Dehydrogenation reaction of 2 with 3,3-dimethylbut-1-ene, a hydrogen acceptor,
afforded a mixture of 1a and its isomer 1b, abbreviated as 1a þ 1b, together with 2,2-dimethylbutane.
Single crystals obtained from a CH2Cl2 solution of 1a þ 1b contained only isomer 1a. X-ray crystal
structure analysis of one of the crystals revealed that 1a adopts a distorted seesaw-shaped four-co-
ordinate geometry where the coordinatively unsaturated metal center is stabilized by weak agostic
Keywords:
Iridium
14-Electron complex
Bis(silyl) chelate ligand
Xanthene-based ligand
Seesaw-shaped complex
Alkene hydrogenation
interaction of two
1a þ 1b demonstrated that 1a is in fast equilibrium with a minor amount of 1b in solution.
Replacement of the chloro ligand in complexes 1a 1b by triflato ligand with AgOTf
g-CeH bonds of the PCy3 ligand. On the other hand, NMR spectroscopic analysis of
þ
a
(OTf ¼ OSO2CF3) afforded quantitatively a single product, i.e., the 16-electron iridiumetriflato com-
plex Ir{
k
3(Si,O,Si)-xantsil}(PCy3)(OTf) (3). 1a þ 1b and 16-electron complex 2 are interconvertible via
oxidative addition of dihydrogen (from 1a þ 1b to 2) and alkene hydrogenation (from 2 to 1a þ 1b).
Complexes 1a þ 1b and 2 were found to catalyze hydrogenation of 3,3-dimethylbut-1-ene.
Ó 2013 Elsevier B.V. All rights reserved.
1. Introduction
influence of the silyl ligand that weakens the coordinate bonds
trans to the silyl coordination sites [3]. For instance, Turculet et al.
Transition-metal complexes bearing silyl-containing multi-
dentate chelate ligands are increasingly attracting considerable
attention due to their unique properties (molecular structures and
reactivity) and potential applications as catalysts [1]. These com-
plexes containing normally labile silyl ligands are stabilized by the
chelate-effect of the multidentate ligand. One of the characteristic
features of these complexes is the ability to generate and stabilize
reactive species with a coordinatively unsaturated metal center in
which the vacant coordination sites are in most cases located trans
to the silyl coordination sites. This feature is attributable to the
following properties of the silyl coordination sites in the multi-
have recently succeeded in isolation and structural determination
of four-coordinate 14-electron Ru complexes having a tridentate
PSiP ligand [4]. These complexes adopt an unusual trigonal pyra-
midal geometry where the silyl moiety of the PSiP ligand occupies
the apical position. It is proposed that this geometry is enforced by
the strong s-donating silyl coordination site in the PSiP ligand.
In this context, we have developed a xanthene-based bis(silyl)
chelate ligand “xantsil” ((9,9-dimethylxanthene-4,5-diyl)bis(di-
methylsilyl)) as a supporting ligand effective for stabilizing coor-
dinatively unsaturated complexes [5]. The xantsil ligand having
double silyl coordination sites is considered to stabilize and
generate reactive, coordinatively unsaturated species more effi-
ciently than the multidentate ligands having only a single silyl
coordination site. Moreover, the xantsil ligand functions as a
hemilabile ligand because the ether-oxygen atom in the xanthene
backbone can weakly coordinate to a metal center. Thus, both
dentate ligands: (1) Strong s-donating ability of the silyl ligand that
makes the metal center electron rich and compensates the elec-
tronic unsaturation of the metal center [2], and (2) strong trans
* Corresponding author. Tel.: þ81 22 795 6539; fax: þ81 22 795 6543.
E-mail address: tobita@m.tohoku.ac.jp (H. Tobita).
k
2(Si,Si) and 3(Si,O,Si) coordination modes are possible (Scheme 1),
k
0022-328X/$ e see front matter Ó 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jorganchem.2013.09.009

T. Komuro et al. / Journal of Organometallic Chemistry 751 (2014) 686e694
687
Scheme 1. Hemilability of the xantsil ligand.
and facile dissociation of the coordinated oxygen in
xantsil complexes generates coordinatively unsaturated
xantsil complexes under mild conditions [5].
k
3(Si,O,Si)-
2(Si,Si)-
k
In our research on the chemistry of metalexantsil complexes,
we have focused on attempting to synthesize the iridiumexantsil
complexes. This is because iridium is known to form stable metale
silyl complexes in various oxidation states, which are potentially
highly active as catalysts [1,6,7]. In this study, we succeeded in the
synthesis of a 14-electron (k
2(Si,Si)-xantsil)iridium complex with a
Scheme 2. Synthesis of iridiumexantsil complexes 1a þ 1b and 2 and interconversion
pathways between them. Reaction conditions: (a) simultaneous mixing of the re-
agents, toluene, r.t., 1 h; (b) toluene, 60 ꢀC, 30 h; (c) C6D6, r.t., 24 h.
unique distorted seesaw-shaped four-coordinate geometry. We
also found that the 14-electron complex exists as an equilibrium
mixture with an isomeric 16-electron k
3(Si,O,Si)-xantsil complex (a
putative structure) and that the equilibrium mixture is an active
catalyst for alkene hydrogenation.
and 2173 cme1 attributable to the two stretching modes of the He
IreH linkage.
The positions of the hydrido ligands can be estimated by
comparing the bond angles around the metal center in 2. The
CleIreSi1 (125.49(7)ꢀ) and PeIreSi1 (122.22(7)ꢀ) angles are
considerably wider than other bond angles, while the CleIreP
(91.74(6)ꢀ) angle is the narrowest. Thus, the two hydrido ligands
are considered to be located in the area surrounded by the P, Si1,
and Si2 atoms and that surrounded by the Cl, Si1, and Si2 atoms.
This indicates that complex 2 adopts a bicapped-tetrahedral
2. Results and discussion
2.1. Synthesis of 16-electron iridiumexantsil complex 2
(Dihydrido)iridium(V) complex Ir{
(2), a precursor of the synthetic target Ir{
(1a), was prepared in 68% yield by the reaction of a dinuclear iri-
dium(I) complex [IrCl(coe)2]2 (coe cyclooctene) with 4,5-
k
2(Si,Si)-xantsil}(H)2(PCy3)Cl
2(Si,Si)-xantsil}(PCy3)Cl
k
¼
geometry in which the trans arrangements of strongly
s-
bis(dimethylsilyl)-9,9-dimethylxanthene (xantsilH2) [5a,b] and
PCy3 (1:2.1:2.7 M ratio, simultaneous mixing) in toluene at room
temperature (Scheme 2). Oxidative addition of two SieH bonds of
xantsilH2 and substitution of PCy3 for coe took place at the iridium
center. Importantly, the main product(s) of this reaction changed by
changing the order of the addition of xantsilH2 and PCy3. Thus, in
contrast with the above procedure producing 2, addition of xant-
silH2 and then PCy3 to [IrCl(coe)2]2 in toluene resulted in direct
formation of a mixture of 1a and its isomer 1b in 48% isolated yield
together with cyclooctane (Scheme 3, also see 2.3). This result
implies that, in the former reaction producing 2, [IrCl(coe)2]2 reacts
first with PCy3 and then with xantsilH2. In fact, the NMR-tube re-
action of [IrCl(coe)2]2 with PCy3 (3 equiv) and then xantsilH2 (2
equiv) in C6D6 at room temperature predominantly afforded com-
plex 2.
donating ligands, i.e., silylehydrido and hydridoehydrido, are
avoided. Similar bicapped-tetrahedral geometries have been re-
ported by Tilley et al. for bis-(silyl)(dihydrido)iridium(V) com-
plexes possessing
chelate ligand [9].
a pyridylindolide or a pyridyl-pyrrolide
The observation of two broadened 1H NMR signals for the
hydrido ligands in 2 implies the existence of a dynamic behavior
exchanging the two hydrido ligands in solution. This was confirmed
by variable-temperature 1H NMR spectroscopy (see Fig. B1 in the
Supplementary material): When a solution of 2 in toluene-d8 was
cooled to 243 K, the hydrido signals appeared as two sharp doublets
2.2. Characterization of complex 2
Single crystal X-ray analysis revealed that 2 has a chloro, PCy3,
and xantsil ligands and they adopt a distorted pseudo-tetrahedral
geometry (Fig. 1). The bond angles around the metal center range
widely from 91.74(6)ꢀ to 125.49(7)ꢀ. The IreSi bond distances (Ire
ꢀ
Si1 ¼ 2.372(2), IreSi2 ¼ 2.344(2) A) are rather normal for typical
ꢀ
(silyl)iridium complexes (2.235(5)e2.479(2) A) [8]. Although the
hydrido ligands of 2 could not be located in the final difference
Fourier map, the following spectroscopic features support the ex-
istence of two hydrido ligands on iridium: The 1H NMR spectrum of
2 at room temperature in C6D6 shows two broad signals assigned to
the hydrido ligands at ꢁ14.33 and ꢁ5.34 ppm in the same intensity.
The IR spectrum of 2 also exhibits two absorption bands at 2297
Scheme 3. One-pot synthesis of 1a þ 1b from [IrCl(coe)2]2. Reaction conditions: (a)
toluene, r.t., 30 min, (b) toluene, r.t., 30 min.

688
T. Komuro et al. / Journal of Organometallic Chemistry 751 (2014) 686e694
Fig. 2. Molecular structure of Ir{
k
2(Si,Si)-xantsil}(PCy3)Cl (1a) (50% probability ellip-
ꢀ
soids). Hydrogen atoms are omitted for clarity. Selected interatomic distances (A) and
angles (ꢀ) for 1a: IreCl 2.3157(11), IreP 2.2593(11), IreSi1 2.3316(13), IreSi2
2.3229(12), Ir$$$C25 3.390(5), Ir$$$C27 3.192(5); CleIreP 150.64(4), CleIreSi1 95.37(5),
CleIreSi2 94.01(4), PeIreSi1 104.08(4), PeIreSi2 103.64(4), Si1eIreSi2 100.47(4), Ire
PeC20 103.56(15), IrePeC26 101.89(16), IrePeC32 136.76(15).
Fig. 1. Molecular structure of Ir{
k
2(Si,Si)-xantsil}(H)2(PCy3)Cl (2) (50% probability el-
ꢀ
lipsoids). Hydrogen atoms are omitted for clarity. Selected bond distances (A) and
angles (ꢀ) for 2: IreCl 2.3771(18), IreP 2.4014(18), IreSi1 2.372(2), IreSi2 2.344(2); Cle
IreP 91.74(6), CleIreSi1 125.49(7), CleIreSi2 105.88(7), PeIreSi1 122.22(7), PeIreSi2
108.94(7), Si1eIreSi2 101.44(7).
dehydrogenation product Ir(xantsil)(PCy3)Cl as a mixture of isomers
1a and 1b, together with 2,2-dimethylbutane (Scheme 2). The
product was isolated from the reaction mixture as a reddish orange
powder in 45% yield. As will be described in sections 2.4e2.6, the
product exists as an equilibrium mixture of two isomers 1a (major)
and 1b (minor) in solution, but exists as only 1a in the solid state. The
same reaction was carried out in C6D6 and was monitored by 1H NMR
spectroscopy, which revealed that 2,2-dimethylbutane was formed in
parallel with the formation of 1a þ 1b. 2,2-Dimethylbutane in the
reaction mixture was identified by comparison of its 1H and 13C{1H}
NMR spectra and GC retention time with those of its authentic
sample.
The same product 1a þ 1b was also formed mainly and was isolated
in 48% yield by the reaction of [IrCl(coe)2]2 with xantsilH2 (2.3 equiv) in
toluene at room temperature followed by the reaction with PCy3 (2.3
equiv) in one-pot (Scheme 3). The same reaction conducted in an NMR
tube using C6D6 as a solvent ([IrCl(coe)2]2:xantsilH2:PCy3 ¼ 1:2:3 in
molar ratio) revealed that cyclooctane was formed as a main product
together with 1a þ 1b,whichwereidentifiedby 1H,13C{1H}, and 31P{1H}
NMR spectroscopy in the reaction mixture. This observation clearly
shows that some of the cyclooctene ligands in [IrCl(coe)2]2 were hy-
drogenated in this reaction.
at ꢁ14.16 (2JPH ¼ 22 Hz) and ꢁ5.07 (2JPH ¼ 112 Hz) ppm coupled
with the phosphorus of the PCy3 ligand. These signals broadened at
room temperature and became much broader at higher tempera-
tures. Existence of the process exchanging the two hydrido ligands
was further supported by the measurement of the 2D EXSY NMR
spectrum [10] for 2 at 243 K where the EXSY cross peaks were
observed between the hydrido signals (see Fig. B2 in the
Supplementary material).
The coupling constants between the hydrogen and phosphorus
nuclei observed in the 1H NMR signals of the hydrido ligands at
243 K are consistent with the positions of the hydrido ligands
estimated by the single crystal X-ray analysis (vide supra). Thus, the
large coupling constant (2JPH ¼ 112 Hz) for the signal at ꢁ5.07 ppm
indicates that the hydrogen atom is approximately trans to phos-
phorus, whereas the small coupling constant (2JPH ¼ 22 Hz) for the
signal at ꢁ14.16 ppm indicates that the hydrogen atom is located at
2
a position nearly cis to phosphorus. Similar JPH values, i.e., 117
(trans to phosphorus) and 18 (cis to phosphorus) Hz, have been
reported for one of the hydrido signals of bis(silyl)(trihydrido)iri-
dium(V) complex Ir{o-(Me2Si)2C6H4}(H)3(PPh3)2 [7b].
The spin-lattice relaxation times (T1) of the two hydrido ligands of 2
were determined using the inversion-recovery method by 1H NMR
spectroscopy. The values at 183 K for the signals at ꢁ13.98
and ꢁ4.69 ppm are 2.3 s and 1.8 s, respectively. These T1 values are
typical of dihydrido complexes (T1 > 150 ms) [11]. Thus, complex 2 is
2.4. Crystal structure analysis of 14-electron complex 1a
Recrystallization of 1a þ 1b from CH2Cl2 afforded reddish or-
ange crystals, from which one crystal was selected and analyzed by
X-ray diffraction method. As a result, the crystal was found to be a
single crystal of 1a and the molecular structure was determined
unequivocally (Fig. 2). Complex 1a has a four-coordinate structure
composed of a bidentate xantsil, PCy3, and chloro ligands, and their
arrangement is considerably distorted from the ideal tetrahedral
geometry. The CleIreP angle (150.64(4)ꢀ) is much wider than other
bond angles around the metal center in 1a (94.01(4)e104.08(4)ꢀ)
and also the CleIreP angle in 2 (91.74(6)ꢀ). The molecular structure
of 1a can be best described as a distorted seesaw-shaped geometry
where the CleIreP moiety corresponds to the long board of the
seesaw [13]. The IreCl, IreP, and IreSi bond distances (2.3157(11),
regarded not as an h
2-dihydrogen complex butas a dihydridocomplex.
The 29Si{1H} NMR spectrum of 2 shows a singlet at ꢁ1.6 ppm,
indicating that the two silicon nuclei in the xantsil ligand are
equivalent in solution. This chemical shift is close to that (1.05 ppm)
for
a bis(silyl)(dihydrido)iridium(V) complex (h
5-C5Me5)Ir{o-
(Me2Si)2C6H4}(H)2 [7c]. The observation of the 29Si NMR signal of 2
as a singlet implies that the coupling with the phosphorus nucleus
is negligible.
2.3. Synthesis of 14-electron iridiumexantsil complex 1a and its
isomer 1b
Ir{k
2(Si,Si)-xantsil}(H)2(PCy3)Cl (2) reacted with 3,3-dimethylbut-
2.2593(11), and av. 2.3273(18) A, respectively) are all slightly
ꢀ
1-ene, a hydrogen acceptor [12], at 60 ꢀC in toluene to give a
shortened in comparison with the corresponding distances of 2

T. Komuro et al. / Journal of Organometallic Chemistry 751 (2014) 686e694
689
ꢀ
(2.3771(18), 2.4014(18), and av. 2.358(3) A, respectively), which is
attributable to the decrease of the coordination number of 1a in
comparison with that of 2.
The coordinatively unsaturated iridium center in 1a is consid-
ered to be stabilized by weak agostic interaction of two CeH bonds
directed toward the vacant coordination sites located in the
widened CleIreP bond angle. Two methylene moieties (C25 and
C27) in two cyclohexyl groups are located trans to the silyl ligands
and are somewhat close to the metal center: The Ir$$$Cg inter-
ꢀ
ꢀ
atomic distances (Ir$$$C25: 3.390(5) A, Ir$$$C27: 3.192(5) A) are
ꢀ
nearly identical with the sum (3.22 A) of the van der Waals radius of
a methylene group (2.0 A) [14] and a single-bond covalent radius of
iridium (1.22 A) [15]. The existence of weak agostic interaction is
ꢀ
ꢀ
also supported by the narrowing of IrePeC20 and IrePeC26 angles
(103.56(15) and 101.89(16)ꢀ, respectively) compared with the Ire
PeC32 angles (136.76(15)ꢀ) where the former two carbons C20 and
C26 belong to the cyclohexyl groups with the weak agostic inter-
action [16]. Similar weak CeH agostic interaction between a metal
and two g-carbons in a cyclohexyl group of PCy3 has been found in
the 16-electron Ruexantsil complex Ru{k
3(Si,O,Si)-xantsil}(P-
Cy3)(CO) [5f].
The two g-CeH agostic interactions in 1a are substantially weak
as demonstrated by the Ir$$$C25 and Ir$$$C27 interatomic distances
that are much longer than the Ir$$$Cg(cyclohexyl) distance (2.923(10)
Fig. 3. 2D EXSY NMR spectrum (in the range of 0e1 ppm) of 1a þ 1b (700 MHz, r.t.,
0
ꢀ
A) in the tris(phosphine)iridium complex [Ir(H)2(PCy2Ph)3]½BAr4ꢂ
CD2Cl2).
(Ar0 ¼ 3,5-(CF3)2C6H3) having a normal agostic interaction between
2.6. Solid state NMR spectroscopic analysis of 1a þ 1b
the iridium(III) center and a g-CeH bond of a cyclohexyl group on
phosphorus [16a]. This weakening of the agostic interactions is
attributable to the strong trans-influence of the silyl ligands Si1 and
Si2 located trans to C27 and C25, respectively. If we neglect these very
weak agostic interactions, 1a can be regarded as a 14-electron bis(-
silyl)iridium(III) complex. To the best of our knowledge,1a is the first
crystallographically characterized 14-electron bis(silyl)iridium com-
plex and is also a rare example of isolated 14-electron iridium(III)
complexes [16,17].
To clarify the composition of 1a and 1b in the solid state, an
NMR spectrum of a solid sample of 1a þ 1b was measured, which
revealed that 1a existed exclusively in the solid state in accordance
with the result of the X-ray crystal structure analysis [18]. The 31P
{1H} CPMAS NMR spectrum shows two signals at ꢁ2.1 and 1.1 ppm
with nearly the same intensity (see Fig. B7 in the Supplementary
material). These chemical shifts are close to that of the 31P NMR
signal of 1a in solution (0.5 ppm in CD2Cl2, 1.8 ppm in C6D6), and
there were no signals in the region where 1b showed its signal in
solution (9.1 ppm in CD2Cl2, 9.6 ppm in C6D6). Therefore, the two
31P signals of the solid sample can be both attributed to 1a. The
observation of two 31P signals implies that the solid sample con-
tains two crystallographically independent molecules of 1a [19].
This sample was obtained as orange crystals by recrystallization of
crude 1a þ 1b from a THF/hexane solution, but its crystal structure
was not determined. The molecular packing in this sample is
considered to be different from that in the single crystal of 1a used
for the X-ray analysis (see 2.4) where the asymmetric unit contains
only one independent molecule of 1a together with one molecule
of CH2Cl2 as the solvent of crystallization.
2.5. Spectroscopic characterization of 1a þ 1b
Although elemental analysis and mass spectroscopic data
confirm that the molecular formula of 1a þ 1b is “Ir(xantsil)(PCy3)
Cl”, the NMR spectroscopic analysis of 1a þ 1b in solution revealed
that this compound exists as an equilibrium mixture of two
isomeric Irexantsil complexes 1a (major) and 1b (minor) (see
Fig. B3e6 in the Supplementary material). The 1H NMR spectrum of
1a þ 1b in CD2Cl2 shows four sharp SiMe signals at 0.72, 0.85 ppm
(1a) and 0.36, 0.77 ppm (1b) with a ca. 5:5:1:1 intensity ratio. In the
29Si{1H} NMR spectrum of 1a þ 1b in CD2Cl2, two doublets were
2
2
observed at ꢁ17.5 (1a, JSiP ¼ 8.4 Hz) and ꢁ6.2 (1b, JSiP ¼ 9.9 Hz)
ppm. These signals are shifted upfield in comparison with that of 2
(ꢁ1.6 ppm). The 31P{1H} NMR spectrum of 1a þ 1b in CD2Cl2 ex-
hibits two broad signals for 1a and 1b at 0.5 ppm (Dn1/2 ¼ 12 Hz)
and 9.1 ppm (Dn1/2 ¼ 8 Hz), respectively. The broadening of these
signals is possibly caused by the fast interconversion between 1a
and 1b. These signals are also substantially shifted upfield in
comparison with that for 2 (41.5 ppm).
2.7. Conversion of chloro complexes 1a þ 1b into triflato complex 3
Replacement of the chloro ligand on iridium of 1a and 1b by a
triflato ligand was examined by treatment of 1a þ 1b with silver
trifluoromethanesulfonate (AgOTf) in C6D6 at room temperature.
This reaction gave a single product Ir{k
3(Si,O,Si)-xantsil}(PCy3)(OTf)
(3) quantitatively (ca. 100% NMR yield) (Scheme 4). Both 1a and 1b
were converted into 3, and this fact further supports that 1b is an
isomer of 1a. A preparative scale reaction of 1a þ 1b with AgOTf in
toluene resulted in isolation of 3 in 65% yield after removal of
precipitated AgCl.
The molar ratio of 1a and 1b in solution changes depending on
the solvent, i.e., 1a:1b ¼ 5:1 in CD2Cl2 and 8:1 in C6D6. This also
implies that the fast equilibrium between 1a and 1b exists. The 2D
EXSY measurement of 1a þ 1b in CD2Cl2 at room temperature
clearly demonstrated it (Fig. 3). Two EXSY cross peaks were
observed between the SiMe signal of 1a at 0.72 ppm and that of 1b
at 0.36 ppm and between the SiMe signal of 1a at 0.85 ppm and that
of 1b at 0.77 ppm. These observations indicate that two sets of
exchange between SiMe groups of 1a and 1b exist and are caused
by the fast equilibrium.
2.8. Characterization of triflato complex 3
The molecular structure of 3 was confirmed by single crystal X-
ray analysis (Fig. 4): Complex 3 adopts a distorted square pyramidal

690
T. Komuro et al. / Journal of Organometallic Chemistry 751 (2014) 686e694
pyridyloxy)methylsilyl (NSiN) ligand Ir(H)(coe)(OTf)(NSiN) where
ꢀ
the IreO(triflato) distance is 2.36 A (the averaged value for two
crystallographically independent molecules) [1l].
The Ir$$$C22(
g
-position of PCy3) interatomic distance (3.123(6)
ꢀ
A) implies the existence of one -CeH agostic interaction at the
g
position trans to apical Si1 on iridium. This distance is close to the
sum of the van der Waals radius of a methylene group [14] and a
ꢀ
single-bond covalent radius of iridium (1.22 A) [15], and is slightly
shorter than the corresponding Ir$$$Cg(agostic) distances for 1a
ꢀ
(3.192(5) and 3.390(5) A). The IR spectrum of 3 exhibits a weak
Scheme 4. Reaction of iridiumexantsil complexes 1a þ 1b with AgOTf that gives a
single complex 3.
absorption band attributable to the CeH stretching vibration of the
agostic interaction at 2721 cme1, which lies at nearly the highest
wavenumber region for nCH in various CeH agostic interactions
(2250e2800 cme1) [21]. Analogous to the strength of the agostic
interactions in 1a (vide supra), the long Ir$$$C22 interatomic dis-
tance for 3 also indicates that this interaction is substantially weak.
Thus, complex 3 can be regarded as a 16-electron complex.
The 1H NMR spectrum of 3 in CD2Cl2 shows two singlets
assigned to the SiMe protons at 0.34 and 0.76 ppm. The 29Si{1H}
NMR spectrum of 3 in CD2Cl2 exhibits a doublet at ꢁ9.4 ppm
(2JSiP ¼ 9.6 Hz). These data clearly shows that two silicon atoms in 3
are equivalent and that two methyl groups on the same silicon
atom are inequivalent. Nevertheless, in the crystal structure shown
in Fig. 4, two silicon atoms are inequivalent. Therefore, there must
be a dynamic process involving exchange of the two silicon moi-
eties in solution. A plausible process is the site change of the triflato
ligand.
geometry composed of a k
3(Si,O,Si)-xantsil ligand, a PCy3 ligand,
and a triflato ligand. The tridentate xantsil ligand is facially coor-
ꢀ
dinated to iridium. The IreO1 distance (2.245(4) A) is close to the
ꢀ
IreO(xanthene) distance (2.220(3) A) in the diiridium complex
having a k
3(P,O,P)-Xantphos ligands on each metal center with a
facial coordination geometry, i.e., [Ir(k m-
3(P,O,P)-Xantphos)(H)(
H)]2½BAr0 ꢂ [20]. The IreSi1 and IreSi2 distances (2.3160(15) and
4 2ꢀ
2.3171(16) A, respectively) are typical of IreSi single bonds. The two
largest bond angles around the iridium center are the PeIreO1 and
Si2eIreO2 angles (176.09(10) and 154.65(11)ꢀ, respectively), indi-
cating that Si1 is located at the apical position of the square
pyramid.
ꢀ
The IreO2 bond (2.324(4) A) is relatively long among common
ꢀ
IreO(triflato) bonds (2.039(6)e2.401(12) A) [8]. This is attributable
The IR spectrum of 3 shows an absorption band at 1317 cme1
assignable to the nS]O stretching vibration of the coordinated tri-
to the large trans influence of a silyl ligand (Si2) located trans to the
triflato ligand. A similar trend for a triflato ligand located trans to a
silyl ligand has been observed in the iridium complex with a bis(2-
flato ligand. This wavenumber is close to that for (h
5-C5Me5)Ir(P-
Me3)(OTf)2 (1324 cme1) in which the OTf groups are coordinated to
the iridium atom [22].
2.9. Estimation of the molecular structure of 1b by comparison of
NMR spectroscopic data for 1a, 1b, and 3
2
The NMR data (1H and 29Si NMR chemical shifts, JSiP coupling
constant) of the SiMe signals for triflato complex 3 are similar to
those for chloro complex 1b but are different from those for 1a: The
1H NMR chemical shifts (0.34 and 0.76 ppm) of the two SiMe signals
for 3 in CD2Cl2 are almost identical with those for 1b (0.36 and
0.77 ppm) but are fairly different from those for 1a (0.72 and
0.85 ppm). Both the 29Si NMR chemical shift and the 2JSiP coupling
constant (ꢁ9.4 ppm, 2JSiP ¼ 9.6 Hz) for 3 are also much closer to the
corresponding data for 1b (ꢁ6.2 ppm, 2JSiP ¼ 9.9 Hz) than those for
1a (ꢁ17.5 ppm, 2JSiP ¼ 8.4 Hz). Accordingly, it is highly probable that
the molecular structure of 1b is similar to that of 3, i.e., a five-
coordinate complex with a tridentate
(Schemes 2 and 3).
k
3(Si,O,Si)-xantsil ligand
On the basis of this estimation and the discussions in 2.5 and 2.6,
14-electron
2(Si,Si)-xantsil complex 1a is expected to be thermo-
dynamically more stable than 16-electron
3(Si,O,Si)-xantsil com-
k
k
plex 1b both in the solid state and in solution. This feature of the
iridiumexantsil complexes 1a and 1b contrasts with those of the
previously known ruthenium, manganese, molybdenum, and
tungsten k k
3(Si,O,Si)-xantsil complexes where the 3(Si,O,Si)-xantsil
geometry exists exclusively both in the solid-state and in solution
[5].
Fig. 4. Molecular structure of Ir{k
3(Si,O,Si)-xantsil}(PCy3)(OTf) (3) (50% probability
ellipsoids). Hydrogen atoms are omitted for clarity. Each of the three fluorine atoms of
the triflato ligand is disordered over two sites with the occupancy factors 0.51:0.49.
Each one of the disordered atoms with the higher occupancy factor (F1A, F2A, and F3A)
In contrast with the above feature of chloro complexes 1a and
1b, the exclusive formation of 16-electron (triflato)(k
3(Si,O,Si)-
ꢀ
ꢀ
is depicted. Selected interatomic distances (A) and angles ( ) for 3: IreP 2.2534(13), Ire
Si1 2.3160(15), IreSi2 2.3171(16), IreO1 2.245(4), IreO2 2.324(4), Ir$$$C22 3.123(6); Pe
IreSi1 101.41(5), PeIreSi2 93.97(5), PeIreO1 176.09(10), PeIreO2 102.72(10), Si1eIre
Si2 100.77(6), Si1eIreO1 80.65(10), Si1eIreO2 94.53(11), Si2eIreO1 82.35(10), Si2e
IreO2 154.65(11), O1eIreO2 80.33(13), IrePeC21104.08(18), IrePeC27 116.97(19), Ire
PeC33 113.27(18), IreO2eS 127.8(2).
xantsil) complex 3 in the reaction of 1a þ 1b with AgOTf (Scheme 4)
indicates that 3 is thermodynamically far more stable than the 14-
electron triflato analog of 1a, i.e., Ir{k
2(Si,Si)-xantsil}(PCy3)(OTf).
This is mainly because the triflato ligand in 3, which is less electron-
donating than the chloro ligand in 1a and 1b, makes the metal

T. Komuro et al. / Journal of Organometallic Chemistry 751 (2014) 686e694
691
Scheme 6. Catalytic hydrogenation of 3,3-dimethylbut-1-ene promoted by 1a þ 1b or
2.
formation of dihydrido complex 2 by the oxidative addition of H2.
High yields of the alkane in these reactions indicate that the xantsil
complexes serve as active catalysts for this hydrogenation.
3. Conclusions
In conclusion, we newly synthesized and characterized a 14-
electron four-coordinate iridium(III) complex 1a bearing a biden-
k
2(Si,Si)-xantsil ligand. Complex 1a adopts a distorted seesaw-
Scheme 5. A possible formation mechanism of 1a þ 1b by the one-pot reaction of
tate
shaped geometry due to the existence of two
teractions that are considerably weakened by strong trans-influ-
ence of two silyl ligands located at the positions trans to the -CeH
bonds. Complex 1a was also found to be in equilibrium with a
smaller amount of its isomer, 16-electron
3(Si,O,Si)-xantsil com-
[IrCl(coe)2]2 with xantsilH2 and then PCy3.
g-CeH agostic in-
g
center of 3 more electron deficient than those of 1a and 1b and,
hence, strengthens the coordination of the xanthene oxygen to
iridium in 3.
k
plex 1b, in solution, whereas 1a exists exclusively in the crystals
formed from the solution. Chloride substitution reaction of 1a þ 1b
with AgOTf gave quantitatively a single 16-electron triflato complex
2.10. A possible formation mechanism of 1a þ 1b by the one-pot
k
3(Si,O,Si)-xantsil ligand. The molecular
synthesis from [IrCl(coe)2]2, xantsilH2, and PCy3
3 having a tridentate
structure of 1b was estimated by comparing the NMR spectroscopic
data with those for 3.
A possible formation mechanism of iridiumexantsil complexes
1a and 1b by the reaction of [IrCl(coe)2]2 with xantsilH2 and then
PCy3 in one-pot (Scheme 3, also see 2.3) is illustrated in Scheme 5.
This mechanism involves (1) oxidative addition of one SieH bond of
xantsilH2 to a coordinatively unsaturated IrCl(coe)2 generated from
the dimer [IrCl(coe)2]2, (2) migratory insertion of cyclooctene
ligand into an IreH bond, (3) oxidative addition of the remaining
SieH bond of the xantsil(H) ligand, (4) CeH reductive elimination
of cyclooctane to give an Irexantsil complex Ir(xantsil)(coe)Cl (A),
and (5) replacement of the cyclooctene ligand in A by PCy3.
1a þ 1b and 16-electron (dihydrido)iridium complex 2 are
interconvertible: Oxidative addition of dihydrogen to 1a þ 1b
afforded 2, while treatment of 2 with 3,3-dimethylbut-1-ene, a
hydrogen acceptor, regenerated 1a þ 1b. Applying these findings,
we found that both 1a þ 1b and 2 function as catalysts for hydro-
genation of 3,3-dimethylbut-1-ene.
This study demonstrates that the xantsil ligand is a useful sup-
porting ligand that effectively stabilizes a coordinatively unsatu-
rated iridium center. We are expecting that iridiumexantsil
complexes 1e3 can serve as active catalysts for various trans-
formation reactions other than olefin hydrogenation. Further
studies on reactions of 1e3 with organic and inorganic molecules
and their application to catalytic reactions are in progress.
2.11. Reaction of complexes 1a þ 1b with H2
When 1a þ 1b was exposed to dihydrogen gas (1 atm) in C6D6 at
room temperature, dihydrido complex 2 was formed in 51% NMR
yield (95% conversion of 1a þ 1b, conversion yield of 2: 54%)
together with some minor products (Scheme 2). This result and the
previously mentioned result in the reaction of 2 with 3,3-
dimethylbut-1-ene demonstrate that complexes 1a þ 1b and 2
are interconvertible via oxidative addition of dihydrogen (from
1a þ 1b to 2) and alkene hydrogenation (from 2 to 1a þ 1b).
4. Experimental section
4.1. General procedures
All manipulations were carried out under dry argon in a glove-
box or using a standard high-vacuum line and Schlenk techniques.
2.12. Hydrogenation of 3,3-dimethylbut-1-ene catalyzed by Ire
xantsil complexes 1a þ 1b and 2
4.1.1. Materials
Benzene-d6, toluene, toluene-d8, hexane, and CH2Cl2 were dried
over CaH2 and vacuum transferred. All solvents were stored under
argon over 4 A molecular sieves in a glovebox. Tricyclohex-
ylphosphine (PCy3) (Strem Chemicals) was used as received.
[IrCl(coe)2]2 [23] and 4,5-bis(dimethylsilyl)-9,9-dimethylxanthene
(xantsilH2) [5a,b] were prepared according to the literature
methods.
ꢀ
According to the above-mentioned finding, it is expected that
complexes 1a þ 1b and 2 are catalytically active for alkene hy-
drogenation. Thus, by employing 1a þ 1b and 2 as catalysts, we
examined the hydrogenation of 3,3-dimethylbut-1-ene (Scheme 6).
In the presence of 5 mol% of 1a þ 1b, 3,3-dimethylbut-1-ene was
exposed to dihydrogen (1 atm) in C6D6. The reaction completed in
32 h at 40 ꢀC to give 2,2-dimethylbutane in 95% NMR yield (100%
conversion of the alkene, TON ¼ 19). Complex 2 (5 mol%) also
promoted the same reaction under similar conditions to produce
2,2-dimethylbutane in 87% NMR yield. Thus, in the hydrogenation
reaction catalyzed by 1a þ 1b, the initial step is probably the
4.1.2. Spectroscopic measurements
1H, 13C{1H}, 29Si{1H}, and 31P{1H} NMR spectra in solution were
recorded on a Bruker AVANCE III 400, a JEOL JNM-ECA 600, or a
JEOL JNM-ECA 700 Fourier transform spectrometer. 29Si{1H} NMR
measurements were performed using the DEPT pulse sequence or

692
T. Komuro et al. / Journal of Organometallic Chemistry 751 (2014) 686e694
using an inverse gate decoupling (IG) pulse sequence. The re-
sidual proton (C6D5H, 7.15 ppm; C6D5CD2H, 2.09 ppm; CDHCl2,
5.32 ppm) and the carbon resonances (C6D6, 128.0 ppm; CD2Cl2,
53.8 ppm) of deuterated solvents were used as internal references
for 1H and 13C NMR chemical shifts, respectively. Aromatic pro-
tons or carbons are abbreviated as ArH and ArC, respectively. 29Si
{1H} NMR chemical shifts were referenced to SiMe4 (0 ppm) as an
external standard. The NMR data were collected at room tem-
perature unless indicated otherwise. 2D exchange spectroscopy
(EXSY) measurements were carried out using a standard pulse
sequence for phase-sensitive 1He1H NOESY spectra [10]. The
solid-state 31P{1H} cross-polarization magic angle spinning
(CPMAS) NMR spectrum of 1a þ 1b was recorded on a JEOL JNM-
ECA 800 Fourier transform spectrometer where the chemical
shifts were referenced to (NH4)H2PO4 (1.0 ppm) as an external
standard. Infrared spectra were measured on a KBr pellet sample
using a Horiba FT-720 spectrometer. Mass spectra and high-
resolution mass spectra (HRMS) were recorded on a Bruker Dal-
tonics solariX 9.4T spectrometer operating in the electrospray
ionization (ESI) mode or on a JEOL JMS-T100GCV spectrometer
operating in the field desorption (FD) mode. Elemental analyses
were carried out using a J-Science Lab JM11 microanalyzer.
Measurements of mass spectra and elemental analyses were
performed at the Research and Analytical Center for Giant Mol-
ecules, Tohoku University.
4.3. Synthesis of Ir(xantsil)(PCy3)Cl (1a þ 1b)
Method A: A toluene (5 mL) solution of 3,3-dimethylbut-1-ene
(0.024 g, 0.29 mmol) was added to complex
2 (0.105 g,
0.126 mmol). The solution was stirred at 60 ꢀC for 30 h. The reaction
mixture was filtered through a membrane filter (Nihon Millipore,
Omnipore membrane, pore size 0.2 mm, 25 mm in diameter). After
the filtrate was concentrated under vacuum, hexane was added to
the concentrated solution, and then the solution was stored over-
night at ꢁ35 ꢀC in a freezer. A reddish orange solid was precipitated
in the solution. The mother liquor was removed, and the solid was
washed with hexane and was dried under vacuum. Ir{k
2(Si,Si)-
xantsil}(PCy3)Cl (1a) (0.047 g, 0.056 mmol) was obtained as a
reddish orange powder in 45% yield. The product exists as a 5:1
equilibrium mixture of isomeric complexes 1a and 1b in CD2Cl2 but
exists as only 1a in the solid state.
Method B (One-pot synthesis of 1a þ 1b from [IrCl(coe)2]2): In a
Schlenk tube, [IrCl(coe)2]2 (0.144 g, 0.161 mmol) and xantsilH2
(0.123 g, 0.377 mmol) were dissolved in toluene (5 mL). The solu-
tion was stirred at room temperature for 30 min. To the resulting
mixture was added PCy3 (0.106 g, 0.378 mmol), and then the
mixture was stirred at room temperature for further 30 min. The
brown reaction mixture was filtered through a Whatman PTFE
syringe filter (pore size: 0.45 mm), and the filtrate was concentrated
under vacuum until most of volatiles were evaporated. When
hexane was added to the concentrated solution, a reddish orange
solid was precipitated. The solution containing the solid was cooled
to ꢁ35 ꢀC overnight. Removing the mother liquor, washing the solid
with hexane, and drying under vacuum gave 1a(þ1b) (0.128 g,
0.154 mmol, 48%) as a reddish orange powder. Data for 1a: 1H NMR
4.2. Synthesis of Ir{k
2(Si,Si)-xantsil}(H)2(PCy3)Cl (2)
[IrCl(coe)2]2 (0.143 g, 0.160 mmol), xantsilH2 (0.112 g,
0.343 mmol), and PCy3 (0.123 g, 0.439 mmol) were dissolved in
toluene (5 mL), and then the solution was stirred at room tem-
perature for 1 h. The resulting red solution was evaporated to
dryness under vacuum. The residue was washed with hexane (ca.
1 mL ꢃ 3) and dried under vacuum. Complex 2 was obtained as a
pale-yellow powder in 68% yield (0.181 g, 0.217 mmol). 1H NMR
(400 MHz, C6D6):
d ca. 0.6e1.8 (m, 33H, PCy3), 0.93 (s, 6H, SiMe),
1.40 (s, 3H, 9-CMe), 1.45 (s, 6H, SiMe), 1.56 (s, 3H, 9-CMe), 6.96 (t,
3
2H, JHH ¼ 7.3 Hz, ArH), 7.13e7.18 (m, 2H, ArH), 7.25 (dd, 2H,
3JHH ¼ 7.3 Hz, 4JHH ¼ 1.4 Hz, ArH). The chemical shifts of the signals
of PCy3 could not be determined exactly due to the overlap with
(400 MHz, C6D6):
d
ꢁ14.33 (br s, Dn1/2 ¼ 117 Hz, 1H, IreH), e5.34
those of 1b. An ArH signal at
with the residual proton signal of C6D6. 1H NMR (400 MHz, CD2Cl2):
0.72 (s, 6H, SiMe), ca. 0.8e2.0 (m, 33H, PCy3), 0.85 (s, 6H, SiMe),
d 7.13e7.18 was partially overlapped
(br d, 2JHP ¼ 99 Hz, 1H, IreH), 0.52e0.70 (m, 3H, PCy3), 0.70e0.89
(m, 6H, PCy3), 0.86 (s, 6H, SiMe), 0.97e1.14 (m, 6H, PCy3), 1.18 (d,
d
3
3
6H, JHP ¼ 1.2 Hz, SiMe), 1.33e1.50 (m, 9H, PCy3), 1.37 (s, 3H, 9-
1.46 (s, 3H, 9-CMe), 1.80 (s, 3H, 9-CMe), 7.05 (t, 2H, JHH ¼ 7.4 Hz,
3 4
CMe), 1.55 (s, 3H, 9-CMe), 1.65e1.81 (m, 6H, PCy3), 1.92e2.06
ArH), 7.21 (dd, 2H, JHH ¼ 7.4 Hz, JHH ¼ 1.4 Hz, ArH), 7.37 (dd, 2H,
3JHH ¼ 7.4 Hz, 4JHH ¼ 1.4 Hz, ArH). The chemical shifts of the signals
of PCy3 could not be determined exactly due to the overlap with
3
(m, 3H, PCy3), 6.95 (t, 2H, JHH ¼ 7.3 Hz, ArH), 7.13e7.17 (m, 2H,
ArH), 7.19 (dd, 2H, 3JHH ¼ 7.3 Hz, 4JHH ¼ 1.4 Hz, ArH). An ArH signal
at
d
7.13e7.17 was partially overlapped with the residual proton
those of 1b. 13C{1H} NMR (101 MHz, CD2Cl2):
d
2.1 (SiMe), 8.2
signal of C6D6. 1H NMR (600 MHz, 243 K, C6D5CD3):
d
ꢁ14.16 (d,
(SiMe), 22.6 (9-CMe), 26.3 (PCy3), 28.1 (d, JCP ¼ 10.2 Hz, PCy3),
29.8e31.8 (br, PCy3), 30.7 (9-CMe), 37.1 (9-CMe), 37.8e39.1 (br,
PCy3), 123.7, 125.6, 129.9, 132.5 (d, 3JCP ¼ 2.5 Hz), 135.2, 159.4 (ArC).
Assignments of the 1H and 13C signals were confirmed by 1He13C
2
2JHP ¼ 22 Hz, 1H, IreH), e5.07 (d, JHP ¼ 112 Hz, 1H, IreH), 0.52e
0.66 (m, 3H, PCy3), 0.66e0.81 (m, 6H, PCy3), 0.85 (s, 6H, SiMe),
1.01e1.16 (m, 6H, PCy3), 1.20 (br s, 6H, Dn1/2 ¼ 3.2 Hz, SiMe), 1.38e
1.53 (m, 9H, PCy3), 1.36 (s, 3H, 9-CMe), 1.53 (s, 3H, 9-CMe), 1.60e
1.80 (m, 6H, PCy3), 1.91e2.04 (m, 3H, PCy3), 6.95 (t, 2H,
3JHH ¼ 7.2 Hz, ArH), 7.10e7.14 (m, 2H, ArH), 7.16 (dd, 2H,
2
HSQC and 1He13C HMBC NMR spectra. 29Si{1H} NMR (79.5 MHz, IG,
2
CD2Cl2):
d
d
ꢁ17.5 (d, JSiP ¼ 8.4 Hz). 31P{1H} NMR (162 MHz, C6D6):
1.8 (br s, Dn1/2 ¼ 12 Hz). 31P{1H} NMR (162 MHz, CD2Cl2):
d 0.5 (br
3JHH ¼ 7.2 Hz, 4JHH ¼ 1.4 Hz, ArH). An ArH signal at
d
7.10e7.14 was
s, Dn1/2 ¼ 12 Hz). 31P{1H} CPMAS NMR (324 MHz, spinning rate:
partially overlapped with a residual ArH signal of C6D5CD3. 13C
22 kHz):
d
ꢁ2.1, 1.1. Data for 1b: 1H NMR (400 MHz, C6D6):
d 0.53 (s,
{1H} NMR (101 MHz, C6D6):
d
2.7 (s with satellites, 1JSiC ¼ 47 Hz,
6H, SiMe),1.00 (s, 6H, SiMe), 2.12e2.28 (m, 6H, PCy3), 2.35e2.53 (m,
3H, PCy3), 7.05e7.09 (m, 4H, ArH), 7.36e7.41 (m, 2H, ArH). The 9-
CMe signals of the xanthene backbone and other 1H signals of
PCy3 could not be assigned because of the overlap with 1H signals of
SiMe), 12.1 (d, 3JCP ¼ 7.5 Hz, SiMe), 22.7 (9-CMe), 26.0 (PCy3), 27.4
2
(d, JCP ¼ 10 Hz, PCy3), 29.3 (PCy3), 30.8 (9-CMe), 33.2 (d,
1JCP ¼ 20 Hz, PCy3), 36.7 (9-CMe), 123.2, 125.2, 129.8, 135.2, 135.3,
158.6 (ArC). 29Si{1H} NMR (79.5 MHz, DEPT, C6D6):
d
ꢁ1.6. 31P{1H}
1a. 1H NMR (400 MHz, CD2Cl2):
d 0.36 (s, 6H, SiMe), 0.77 (s, 6H,
NMR (162 MHz, C6D6):
d
41.5. IR (KBr-pellet): 3053 (w, nCH), 2925
SiMe), ca. 0.8e2.0 (m, 24H, PCy3), 1.60 (s, 3H, 9-CMe), 1.82 (s, 3H, 9-
CMe), 2.08e2.26 (m, 6H, PCy3), 2.32e2.49 (m, 3H, PCy3), 7.15 (t, 2H,
3JHH ¼ 7.3 Hz, ArH), 7.29e7.35 (m, 4H, ArH). The chemical shifts of
(s, nCH), 2850 (m, nCH), 2297 (w, nIrH), 2173 (w, nIrH), 1603 (w), 1444
(w), 1392 (s), 1244 (w), 1232 (w), 1215 (s), 1194 (w), 1124 (w), 1005
(w), 876 (w), 841 (m), 823 (s), 816 (s), 798 (m), 775 (s), 754 (s), 673
(w), 638 (m), 511 (w), 445 (m) cmꢁ1. HRMS (ESI, acetone solu-
tion): m/z calcd for [C37H59OSi2PClIr þ Na]þ 857.3052, found
857.3048. NaI was added to the sample. Anal. Calcd for
the signals of PCy3 at d ca. 0.8e2.0 could not be determined exactly
due to the overlap with those of 1a. 13C{1H} NMR (101 MHz,
CD2Cl2): 5.5 (SiMe), 8.0 (SiMe), 22.7 (9-CMe), 30.2 (9-CMe), 37.6
d
(9-CMe), 123.4, 124.6, 129.7, 133.1, 136.1, 160.2 (ArC). The signals of
PCy3 could not be assigned due to the overlap with those of 1a and
C
37H59OSi2PClIr: C, 53.24; H, 7.12. Found: C, 53.59; H, 7.00.

T. Komuro et al. / Journal of Organometallic Chemistry 751 (2014) 686e694
693
unidentified impurities. Assignments of the 1H and 13C signals were
confirmed by 1He13C HSQC and 1He13C HMBC NMR spectra. 29Si
applied to the tube. This procedure was repeated three times, and
then the tube was shaken well. Conversion of 1a and 1b at room
temperature was monitored by 1H NMR spectroscopy. Dihydrido
complex 2 was formed as the main product. The maximum NMR
yield of 2 (51%) was attained after 24 h (95% conversion of 1a þ 1b,
conversion yield of 2: 54%). The NMR yield was determined by
comparison of the intensities of the SiMe signals for 1a, 1b, and 2
with that of the Cp2Fe signal. The product 2 was identified by
comparing the 1H and 31P{1H} NMR spectra with those of the
sample synthesized by the procedure described in section 4.2.
Prolonged exposure of the reaction mixture to H2 resulted in
decrease of 2 and formation of unidentified products.
{1H} NMR (79.5 MHz, IG, CD2Cl2):
d
ꢁ6.2 (d, 2JSiP ¼ 9.9 Hz). 31P{1H}
NMR (162 MHz, C6D6):
(162 MHz, CD2Cl2):
d
9.6 (br s, Dn1/2 ¼ 7 Hz). 31P{1H} NMR
d
9.1 (br s, Dn1/2 ¼ 8 Hz). Data for a solid sample
of 1a(þ1b): IR (KBr-pellet): 3053 (w, nCH), 2929 (s, nCH), 2852 (s,
nCH), 2819 (m, nCH), 1604 (w), 1446 (m), 1427 (w), 1392 (s), 1267 (w),
1234 (m), 1219 (s), 1192 (m), 1174 (w), 1122 (m), 1003 (w), 877 (w),
835 (s), 816 (s), 793 (s), 773 (s), 750 (s), 735 (s), 673 (w), 638 (m),
513 (w), 447 (m), 424 (w), 407 (m) cmꢁ1. HRMS (FD): m/z calcd for
[C37H57OSi2PClIr]þ 832.2998, found 832.3013. Anal. Calcd for
C
37H57OSi2PClIr: C, 53.37; H, 6.90. Found: C, 53.34; H, 6.93.
4.4. Reaction of Ir(xantsil)(PCy3)Cl (1a þ 1b) with AgOTf: synthesis
of Ir{
3(Si,O,Si)-xantsil}(PCy3)(OTf) (3)
4.6. NMR monitoring of a reaction of 3,3-dimethylbut-1-ene with
dihydrogen catalyzed by Ir(xantsil)(PCy3)Cl (1a þ 1b)
k
To 1a þ 1b (0.047 g, 0.056 mmol) in a Schlenk tube was added a
suspension of AgOTf (0.014 g, 0.054 mmol) in toluene (2 mL). The
tube was covered with a piece of aluminum foil to shield the sample
against the light, and the mixture was stirred at room temperature
for 20 min. The reaction mixture was filtered through a membrane
1a þ 1b (5 mg, 6
mmol), 3,3-dimethylbut-1-ene (11 mg,
0.13 mmol), and Cp2Fe (internal standard, less than 1 mg) were
dissolved in C6D6 (0.6 mL). The solution was transferred to an NMR
tube with a J-Young Teflon valve (5 mm o.d.). A 1H NMR spectrum of
the mixture was measured to determine the ratio of the intensities
of the signals of 3,3-dimethylbut-1-ene and the signal of Cp2Fe. The
NMR tube was evacuated briefly to degas, and then charged with
1 atm of H2 gas. This procedure was repeated three times. The tube
was shaken well and then heated at 40 ꢀC, and the conversion of
3,3-dimethylbut-1-ene was monitored by 1H NMR spectroscopy.
After heating for 32 h, the alkene was completely consumed and
2,2-dimethylbutane was formed in 95% NMR yield (100% conver-
sion of the alkene, TON ¼ 19). During the reaction, the tube was
shaken well after 1, 5, 7, and 24 h to dissolve dihydrogen gas and
volatile 3,3-dimethylbut-1-ene (b.p. 41 ꢀC) in the solution, and after
20 h of the reaction the tube was charged again with 1 atm of H2
gas. The product (2,2-dimethylbutane) was identified by comparing
the 1H and 13C{1H} NMR spectra and the retention time of gas
chromatography (column: SE-30 packed column, carrier gas: N2,
detector: FID) with those of the authentic sample.
filter (Nihon Millipore, Omnipore membrane, pore size 0.2
mm,
25 mm in diameter). The yellow filtrate was concentrated under
vacuum. A yellow solid was precipitated in the solution. Hexane was
added to the solution containing the solid, and then the suspension
was stored at ꢁ35 ꢀC in a freezer for 1 h. The solid was collected by
filtration through a membrane filter and was dried under vacuum. Ir
{k
3(Si,O,Si)-xantsil}(PCy3)(OTf) (3)$0.5toluene (0.035 g, 0.035 mmol)
was obtained as a yellow powder in 65% yield based on AgOTf. 1H
NMR (400 MHz, C6D6): 0.31 (s, 6H, SiMe), 0.50e2.90 (m, 33H, PCy3),
d
0.83 (s, 6H, SiMe), 1.43 (s, 3H, 9-CMe), 1.92 (s, 3H, 9-CMe), 7.05 (t, 2H,
4
3JHH ¼ 7.3 Hz, ArH), 7.11 (dd, 2H, 3JHH ¼ 7.3 Hz, JHH ¼ 1.2 Hz, ArH),
3
7.27 (dd, 2H, JHH ¼ 7.3 Hz, 4JHH ¼ 1.2 Hz, ArH). 1H NMR (400 MHz,
CD2Cl2):
d 0.34 (s, 6H, SiMe), 0.76 (s, 6H, SiMe), 1.15e1.45 (m, 9H,
PCy3), 1.67 (s, 3H, 9-CMe), 1.71e2.05 (br, 15H, PCy3), 1.82 (s, 3H, 9-
CMe), 2.06e2.29 (br, 6H, PCy3), 2.38e2.72 (br, 3H, PCy3), 7.21 (t,
2H, 3JHH ¼ 7.3 Hz, ArH), 7.32 (dd, 2H, 3JHH ¼ 7.3 Hz, 4JHH ¼ 1.1 Hz, ArH),
3
4
7.42 (dd, 2H, JHH ¼ 7.3 Hz, JHH ¼ 1.1 Hz, ArH). 13C{1H} NMR
4.7. NMR monitoring of a reaction of 3,3-dimethylbut-1-ene with
(101 MHz, CD2Cl2):
d
2.7 (s with satellites, 1JSiC ¼ 46 Hz, SiMe), 6.0 (s
dihydrogen catalyzed by Ir{k
2(Si,Si)-xantsil}(H)2(PCy3)Cl (2)
with satellites, 1JSiC ¼ 49 Hz, SiMe), 24.0 (9-CMe), 26.6 (PCy3), 27.9 (d,
2JCP ¼ 10.7 Hz, PCy3), 28.7e34.5 (br, Dn1/2 ¼ 134 Hz, PCy3), 30.9 (9-
CMe), 37.7 (9-CMe), 38.5e44.6 (br, Dn1/2 ¼ 202 Hz, PCy3), 120.7 (q,
1JCF ¼ 320.6 Hz, CF3), 125.9, 126.1, 129.3, 129.5, 136.5, 160.7 (ArC).
Assignments of the 1H and 13C signals were confirmed by 1He13C
HSQC and 1He13C HMBC NMR spectra. 29Si{1H} NMR (79.5 MHz,
The title reaction was monitored by a procedure similar to that
for the above-mentioned hydrogenation catalyzed by 1a þ 1b using
2 (5 mg, 6 mmol), 3,3-dimethylbut-1-ene (11 mg, 0.13 mmol), and
Cp2Fe (internal standard, less than 1 mg) in C6D6 (0.6 mL) at 40 ꢀC.
After 34 h, 2,2-dimethylbutane was formed in 87% NMR yield.
DEPT, C6D6):
CD2Cl2):
d
ꢁ9.5 (d, 2JSiP ¼ 9.7 Hz). 29Si{1H} NMR (79.5 MHz, DEPT,
d
ꢁ9.4 (d, 2JSiP ¼ 9.6 Hz). 31P{1H} NMR (162 MHz, C6D6):
d
1.0.
4.8. X-ray crystal structure determination
31P{1H} NMR (162 MHz, CD2Cl2):
d 1.6. IR (KBr-pellet): 3043 (w, nCH),
2931 (m, nCH), 2852 (m, nCH), 2721 (w, nCH), 1448 (w), 1421 (w), 1389
(s), 1317 (s, nS]O), 1230 (s), 1207 (s), 1201 (s), 1159 (m), 1144 (m), 1115
(w), 1086 (w), 1024 (m), 1011 (s), 839 (m), 823 (m), 802 (m), 785 (m),
764 (m), 731 (w), 675 (w), 648 (w), 633 (s), 540 (w), 517 (w), 444 (w),
409 (w) cmꢁ1. FD-MS: m/z 946 (Mþ, 10), 797 (Mþ ꢁ OTf, 100). Anal.
Calcd for C38H57O4F3Si2PSIr þ 0.5 C7H8: C, 50.23; H, 6.20. Found: C,
49.89; H, 6.07.
An X-ray quality single crystal of 2$toluene (a pale-yellow plate)
was obtained from a concentrated toluene solution at ꢁ35 ꢀC in a
freezer. That of 3$0.2CH2Cl2 (a pale-yellow plate) was obtained
from a hexane layered solution in CH2Cl2 at ꢁ35 ꢀC. For 1a$CH2Cl2, a
reddish orange blocklike crystal for the X-ray analysis was selected
from crystals obtained from a concentrated CH2Cl2 solution of
1a þ 1b at ꢁ35 ꢀC. Intensity data for the analysis were collected on a
Rigaku RAXIS-RAPID imaging plate diffractometer with graphite-
ꢀ
4.5. NMR monitoring of a reaction of Ir(xantsil)(PCy3)Cl (1a þ 1b)
monochromated Mo K
a
radiation (
l
¼ 0.71069 A) under a cold
with dihydrogen
nitrogen stream (T ¼ 150 K). Numerical absorption corrections
were applied to the data. The structures were solved by the Pat-
terson method using the DIRDIF-2008 program [24] and refined by
full matrix least-squares techniques on all F2 data with SHELXL-97
[25]. Each of the three fluorine atoms of the triflato ligand in 3 was
disordered over two sites with the occupancy factors 0.51:0.49.
These atoms were refined isotropically. In the case of 1a$CH2Cl2,
each of the two chlorine atoms of a CH2Cl2 molecule, the solvent of
1a þ 1b (6 mg, 7
mmol) and Cp2Fe (internal standard, less than
1 mg; Cp ¼
h
5-C5H5) were dissolved in C6D6 (0.6 mL). The solution
was transferred to an NMR tube with a J-Young Teflon valve (5 mm
o.d.). A 1H NMR spectrum of the mixture was measured to deter-
mine the ratio of the intensities of signals of 1a, 1b, and Cp2Fe. The
NMR tube was evacuated briefly to degas, and 1 atm of H2 was then

694
T. Komuro et al. / Journal of Organometallic Chemistry 751 (2014) 686e694
(2009) 2507;
crystallization, was disordered over two sites with the occupancy
factors 0.67:0.33. For 3$0.2CH2Cl2, the structure refinement was
carried out by assuming partial occupancy (occupancy
factor ¼ 0.20) of the solvent of crystallization (CH2Cl2) because the
determined electron density of the CH2Cl2 was far less than ex-
pected for one molecule. The carbon and chlorine atoms of the
solvent molecules in 1a$CH2Cl2 and 3$0.2CH2Cl2 were refined iso-
tropically, and no hydrogen atoms on the carbon atoms were
included. Except for these solvent molecules and the disordered
atoms, anisotropic refinements were applied to all non-hydrogen
atoms. Two hydrido hydrogen atoms of 2 were not located in the
difference Fourier map and were not included in the refinement.
Other hydrogen atoms were placed at calculated positions. The
correct absolute structure of 1a was confirmed by refinement of the
Flack parameter (0.000(4)) [26]. All calculations were carried out
using Yadokari-XG 2009 [27]. Selected crystallographic data for 1a,
2, and 3 are listed in Table B1 (see the Supplementary material).
(j) H. Fang, Y.-K. Choe, Y. Li, S. Shimada, Chem. Asian J. 6 (2011) 2512;
(k) Y.-H. Li, Y. Zhang, X.-H. Ding, Inorg. Chem. Commun. 14 (2011) 1306;
(l) R. Lalrempuia, M. Iglesias, V. Polo, P.J.S. Miguel, F.J. Fernández-Alvarez,
J.J. Pérez-Torrente, L.A. Oro, Angew. Chem. Int. Ed. 51 (2012) 12824;
(m) C.-I. Lee, J. Zhou, O.V. Ozerov, J. Am. Chem. Soc. 135 (2013) 3560.
[2] D.L. Lichtenberger, A. Rai-Chaudhuri, J. Am. Chem. Soc. 113 (1991) 2923.
[3] (a) J. Chatt, C. Eaborn, S. Ibekwe, Chem. Commun. (Lond.) (1966) 700;
(b) J. Chatt, C. Eaborn, S.D. Ibekwe, P.N. Kapoor, J. Chem. Soc. A 125 (1970) 1343;
(c) V.K. Dioumaev, L.J. Procopio, P.J. Carroll, D.H. Berry, J. Am. Chem. Soc. 125
(2003) 8043.
[4] M.C. MacInnis, R. McDonald, M.J. Ferguson, S. Tobisch, L. Turculet, J. Am. Chem.
Soc. 133 (2011) 13622.
[5] (a) H. Tobita, K. Hasegawa, J.J.G. Minglana, L.-S. Luh, M. Okazaki, H. Ogino,
Organometallics 18 (1999) 2058;
(b) J.J.G. Minglana, M. Okazaki, K. Hasegawa, L.-S. Luh, N. Yamahira, T. Komuro,
H. Ogino, H. Tobita, Organometallics 26 (2007) 5859;
(c) J.J.G. Minglana, M. Okazaki, H. Tobita, H. Ogino, Chem. Lett. (2002) 406;
(d) M. Okazaki, J.J.G. Minglana, N. Yamahira, H. Tobita, H. Ogino, Can. J. Chem.
81 (2003) 1350;
(e) M. Okazaki, N. Yamahira, J.J.G. Minglana, H. Tobita, Organometallics 23
(2004) 4531;
(f) M. Okazaki, N. Yamahira, J.J.G. Minglana, T. Komuro, H. Ogino, H. Tobita,
Organometallics 27 (2008) 918;
Acknowledgments
(g) H. Tobita, N. Yamahira, K. Ohta, T. Komuro, M. Okazaki, Pure Appl. Chem.
80 (2008) 1155;
(h) R. Begum, T. Komuro, H. Tobita, Chem. Commun. (2006) 432;
(i) R. Begum, T. Komuro, H. Tobita, Chem. Lett. 36 (2007) 650;
(j) T. Komuro, R. Begum, R. Ono, H. Tobita, Dalton Trans. 40 (2011) 2348;
(k) T. Komuro, S. Okawara, K. Furuyama, H. Tobita, Chem. Lett. 41 (2012) 774.
[6] (a) J.Y. Corey, Chem. Rev. 111 (2011) 863;
(b) J.Y. Corey, J. Braddock-Wilking, Chem. Rev. 99 (1999) 175.
[7] For selected examples of iridium complexes having a bis(silyl) chelate ligand,
see: (a) J. Greene, M.D. Curtis, J. Am. Chem. Soc. 99 (1977) 5176;
(b) M. Loza, J.W. Faller, R.H. Crabtree, Inorg. Chem. 34 (1995) 2937;
(c) Y.-J. Lee, J.-D. Lee, S.-J. Kim, J. Ko, I.-H. Suh, M. Cheong, S.O. Kang, Organ-
ometallics 23 (2004) 135;
This work was supported by Grants-in-Aid for Scientific
Research (Grant Numbers 25410058, 22350024, and 23750053)
from the Japan Society for the Promotion of Science (JSPS) and also
partly supported by “Nanotechnology Platform” of the Ministry of
Education, Culture, Sports, Science and Technology (MEXT), Japan,
at the Center for Integrated Nanotechnology Support, Tohoku
University. We are grateful to Mr. Shinichiro Yoshida (Tohoku
University) for his help with the 2D EXSY measurements of 1a þ 1b
and 2 and with the solid-state NMR spectroscopic measurements of
1a þ 1b. We also acknowledge the Research and Analytical Center
for Giant Molecules, Tohoku University, for spectroscopic mea-
surements and elemental analyses. This work was also performed
under the Cooperative Research Program of “Network Joint
Research Center for Materials and Devices”. We acknowledge Prof.
Hideo Nagashima (Kyusyu University) for his warm support in this
program. We are also grateful to Ms. Keiko Ideta for her help with
the low-temperature NMR spectroscopic measurements of 2 and
with the determination of spin-lattice relaxation times (T1) of 2 in
the Evaluation Center of Materials Properties and Function, Insti-
tute for Material Chemistry and Engineering, Kyusyu University.
(d) H. Hashimoto, T. Suzuki, H. Tobita, Dalton Trans. 39 (2010) 9386.
[8] Based on a survey of the Cambridge Structural Database, CSD Version 5.34,
November 2012.
[9] (a) D. Karshtedt, A.T. Bell, T.D. Tilley, Organometallics 25 (2006) 4471;
(b) J.L. McBee, J. Escalada, T.D. Tilley, J. Am. Chem. Soc. 131 (2009) 12703.
[10] (a) C.L. Perrin, T.J. Dwyer, Chem. Rev. 90 (1990) 935;
(b) L. Cronin, C.L. Higgitt, R.N. Perutz, Organometallics 19 (2000) 672.
[11] D.G. Hamilton, R.H. Crabtree, J. Am. Chem. Soc. 110 (1988) 4126.
[12] (a) R.H. Crabtree, C.P. Parnell, Organometallics 4 (1985) 519;
(b) D.W. Lee, W.C. Kaska, C.M. Jensen, Organometallics 17 (1998) 1.
[13] The s4 value, developed by Houser et al. for evaluating the geometry of four-
coordinate complexes, was calculated for 1a as 0.75, which is close to the
reported values (w0.70 ꢄ 0.02) for complexes adopting a seesaw-shaped
geometry. See: L. Yang, D.R. Powell, R.P. Houser Dalton Trans. (2007) 955.
[14] L. Pauling, The Nature of the Chemical Bond, third ed., Cornell University
Press, Ithaca, New York, 1960, pp. 257e264.
[15] P. Pyykkö, M. Atsumi, Chem. Eur. J. 15 (2009) 186.
[16] (a) A.C. Cooper, E. Clot, J.C. Huffman, W.E. Streib, F. Maseras, O. Eisenstein,
K.G. Caulton, J. Am. Chem. Soc. 121 (1999) 97;
Appendix A. Supplementary material
(b) A.C. Cooper, W.E. Streib, O. Eisenstein, K.G. Caulton, J. Am. Chem. Soc. 119
(1997) 9069.
[17] (a) N.M. Scott, R. Dorta, E.D. Stevens, A. Correa, L. Cavallo, S.P. Nolan, J. Am.
Chem. Soc. 127 (2005) 3516;
CCDC 959048, 959049, and 959050 contain the supplementary
crystallographic data for this paper. These data can be obtained free
of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
(b) R. Dorta, R. Goikhman, D. Milstein, Organometallics 22 (2003) 2806;
(c) C.Y. Tang, N. Phillips, M.J. Kelly, S. Aldridge, Chem. Commun. 48 (2012)
11999.
[18] The solid sample used for the measurement of the 31P{1H} CPMAS NMR
spectrum, which involves only 1a, was dissolved in CD2Cl2, and the 1H NMR
spectrum of the solution was measured. The result revealed the formation of a
mixture of 1a and 1b in 5:1 M ratio due to the fast equilibrium between them
in solution.
Appendix B. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.jorganchem.2013.09.009.
[19] As a related phenomenon, Davies et al. reported that the plural signals (ca. 2:1
intensity ratio) were observed in the solid-state 31P CPMAS NMR spectrum of
tricyclohexylphosphine oxide (OPCy3) due to the existence of three crystal-
lographically independent molecules. See: J.A. Davies, S. Dutremez,
A.A. Pinkerton Inorg. Chem. 30 (1991) 2380.
[20] A.J. Pontiggia, A.B. Chaplin, A.S. Weller, J. Organomet. Chem. 696 (2011) 2870.
[21] M. Brookhart, M.L.H. Green, L.-L. Wong, Prog. Inorg. Chem. 36 (1988) 1.
[22] P.J. Stang, Y.-H. Huang, A.M. Arif, Organometallics 11 (1992) 231.
[23] J.L. Herde, J.C. Lambert, C.V. Senoff, Inorg. Synth. 15 (1974) 18.
[24] P.T. Beurskens, G. Beurskens, R. de Gelder, S. Garcia-Granda, R.O. Gould,
J.M.M. Smits, The DIRDIF2008 Program System, Crystallography Laboratory,
University of Nijmegen, Nijmegen, The Netherlands, 2008.
[25] G.M. Sheldrick, Acta Crystallogr. Sect. A 64 (2008) 112.
[26] H.D. Flack, Acta Crystallogr. Sect. A 39 (1983) 876.
References
[1] For selected examples of iridium complexes bearing silyl-containing multi-
dentate ligands, see: (a) M.J. Auburn, S.R. Stobart, Inorg. Chem. 24 (1985) 318;
(b) P.I. Djurovich, A.L. Safir, N.L. Keder, R.J. Watts, Inorg. Chem. 31 (1992) 3195;
(c) M.J. Auburn, R.D. Holmes-Smith, S.R. Stobart, P.K. Bakshi, T.S. Cameron,
Organometallics 15 (1996) 3032;
(d) M. Okazaki, H. Tobita, H. Ogino, J. Chem. Soc. Dalton Trans. (1997) 3531;
(e) M. Stradiotto, K.L. Fujdala, T.D. Tilley, Chem. Commun. (2001) 1200;
(f) P. Sangtrirutnugul, T.D. Tilley, Organometallics 26 (2007) 5557;
(g) D.F. MacLean, R. McDonald, M.J. Ferguson, A.J. Caddell, L. Turculet, Chem.
Commun. (2008) 5146;
[27] (a) K. Wakita, Yadokari-XG, Software for Crystal Structure Analyses, 2001;
(b) Yadokari-XG 2009, Software for Crystal Structure Analyses. See: C. Kabuto,
S. Akine, T. Nemoto, E. Kwon J. Cryst. Soc. Jpn. 51 (2009) 218.
(h) E. Morgan, D.F. MacLean, R. McDonald, L. Turculet, J. Am. Chem. Soc. 131
(2009) 14234;
(i) M.T. Whited, N.P. Mankad, Y. Lee, P.F. Oblad, J.C. Peters, Inorg. Chem. 48