Electronic versus steric control in palladium complexes of carboranyl phosphine-iminophosphorane ligands

Article abstract of DOI:10.1039/c8dt04006k

A new family of carboranyl phosphine-iminophosphorane ligands was prepared and characterized. The new ligands present a carboranyl group directly attached to the iminophosphorane nitrogen atom through a cage carbon atom (C-carboranyl derivatives L1-L3) or

Full text of DOI:10.1039/c8dt04006k

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Electronic versus steric control in palladium
complexes of carboranyl phosphine-
iminophosphorane ligands†
Cite this: Dalton Trans., 2019, 48,
486
José Luis Rodríguez-Rey,a David Esteban-Gómez,b Carlos Platas-Iglesiasb and
a
Antonio Sousa-Pedrares
*
A new family of carboranyl phosphine-iminophosphorane ligands was prepared and characterized. The
new ligands present a carboranyl group directly attached to the iminophosphorane nitrogen atom
through
a cage carbon atom (C-carboranyl derivatives L1–L3) or through the B3 boron atom
(B-carboranyl derivatives L4 and L5), and the phosphine group on a side chain derived from the dipho-
sphine dppm, i.e. with a two-atom spacer between the P and N donor atoms. The non-carboranyl ana-
logue L6, with a biphenyl group on the nitrogen atom, was also synthesized for comparison. These
potential (P, N) ligands were used to obtain palladium complexes (Pd1–Pd6) and, thus, study how the
different inductive effect of the carboranyl substituents can modify the coordinating ability of the nitrogen
atom. The structural analysis of the complexes revealed two different coordination modes for the ligands:
the (P, N) chelate coordination and the unexpected P-terminal coordination, which is not observed for
non-carboranyl phosphine-iminophosphoranes. These unexpected structural differences led us to
perform DFT calculations on the ligands and metal complexes. The calculations show that the final
coordination modes depend on the balance between the electronic and steric properties of the particular
carboranyl group.
Received 5th October 2018,
Accepted 29th November 2018
DOI: 10.1039/c8dt04006k
rsc.li/dalton
of the o-carborane cluster is its irregular charge distribution. It
has been established that the CH vertices, with more electron
Introduction
ortho-Carboranes are bulky clusters with ten boron atoms and density involved in the skeletal electrons are holders of +δ
two adjacent carbon atoms in an icosahedral arrangement. charges, and that the electron density on the BH vertices
The B–H hydride bonds confers them a high hydrophobic increases with the distance from the CH vertices, increasing
character, while the close shell structure gives them high the −δ charge on the boron atoms.6 Thus, the electron density
thermal and chemical stability.1 The functionalization of other of the five different positions of the ortho-carborane moiety
materials with carborane substituents has been used to (Fig. 1), follow the order: 1 (2) < 3 (6) < 4 (5, 7, 11) < 8 (10) < 9
produce robust materials2 and hydrophobic compounds that (12) (1, 2: carbon atoms; 3–12: boron atoms). This order of
have found several medical applications.3 The carborane mole- electron density is reflected in the inductive effect that every
cule is a bulky cluster, comparable to adamantane and signifi- position of the cluster exerts on its hydrogen atom or on any
cantly larger than the phenyl ring rotation envelope.4 Thus, it group that may substitute it. Thus, the cluster carbon atoms
has found application as a steric group, for example for the show a clear electron-withdrawing character, which is less pro-
construction of luminescent materials, as the bulky cage can
hinder intermolecular interactions.5 A very interesting property
aDepartamento de Química Inorgánica, Universidade de Santiago de Compostela,
15782 Santiago de Compostela, Spain. E-mail: antonio.sousa.pedrares@usc.es
bCentro de Investigacións Científicas Avanzadas (CICA) and Departamento de
Química, Universidade da Coruña, Campus da Zapateira-Rúa da Fraga 10, 15008 A
Coruña, Spain
†Electronic supplementary information (ESI) available: Crystallographic data for
all crystal structures, NMR spectra and optimized Cartesian coordinates
obtained with DFT calculations. CCDC 1855597–1855602. For ESI and crystallo-
graphic data in CIF or other electronic format see DOI: 10.1039/c8dt04006k
Fig. 1 Numbering scheme used for o-carborane derivatives (1,2:
carbon atoms; 3–12: boron atoms). Equivalent positions: (C1, C2), (B3,
B6), (B4, B5, B7, B11), (B8, B9), (B9, B12).
486 | Dalton Trans., 2019, 48, 486–503
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
nounced for the B3/B6 positions (connected to both carbon
atoms). The boron atoms at the central positions (B4, B5, B7,
B11) are electroneutral, while the further boron atoms B8/B10,
and especially the antipodal positions B9/B12, are electron-
donating. These irregularities determine the reactivity of every
position and, thus, the possibility of preparing substituted
derivatives. The electron-withdrawing character of the carbon
atoms is reflected in acidic CH groups (pKa = 23.3),7 which is
commonly exploited for the preparation of C-functionalized
carboranes (via deprotonation with strong bases and reaction
with electrophiles).1 The positively charged boron atoms, B3
and B6, react with nucleophiles to produce nido derivatives,8
which can be capped to yield closo-B3-functionalized deriva-
tives.1,9 The electron-donating character of the antipodal B9/
B12 positions is reflected in a higher reactivity towards electro-
philic substitution (which proceeds to the B8 and B10 posi-
tions under forcing conditions), which enables the easy prepa-
ration of other B-functionalized carboranes.1
The different inductive effect of C- and B-carboranyl deriva-
tives has been used to modify chemical and physical pro-
perties of other substrates. However, most of these studies
only exploit the electron-withdrawing character of C-carboranyl
derivatives. In fact, this character leads to unique electronic
effects and has been used for the preparation of electronic and
luminescent materials.5 The few studies that compare the
different inductive effect of C- and B-carboranyl groups have
only focused on the more extreme members of the series, i.e.
the electron-withdrawing C-carboranyl derivatives and the elec-
tron-donating B8- and B9-carboranyl derivatives. Thus, carbor-
anyl groups have been used to modify the acidity of carboxylic
acid6b and thiol groups (–SH),6c,d concluding, for example, that
the C-connection increases the acidity of a –SH group (pKa =
3.30) while the B8- (pKa = 9.32) and B9- (pKa = 10.08) connec-
Fig. 2 closo-Carboranyl-iminophosphoranes reported in the literature.
compounds,17,18 and showed that the nitrogen donor atom is
very affected by the C-carboranyl group, although its donating
properties are also influenced by the nature of the second
(more coordinating) donor atom on the other cage carbon
atom.
The combination with a thiolate group (Fig. 2, E = S−) com-
pletely deactivates the nitrogen atom towards the coordination
to a palladium(II) center,12 while the combination with a phos-
phine group (Fig. 2, E = PPh2) reduces its coordination
strength although the ligand can still engage in (P, N) coordi-
nation to a palladium center.13 However, while the phosphine
group was proven to favor the coordination of the nitrogen
atom, the direct connection of both neutral donor groups
(phosphine and iminophosphorane) to the cage carbon atoms
promotes the evolution to the anionic nido derivatives which
results in an increase of the donor strength of the nitrogen
atom,13 as also found for 10 and 12-vertex closo-carborane
anions, which are strong electron donor substituents.19
As a continuation of our research on the modulation of the
coordinating ability of an iminophosphorane nitrogen atom by
effect of a carborane group, we have designed new carboranyl
phosphine-iminophosphorane ligands with the nitrogen atom
directly attached to a cage atom (C- or B3-), and the more coor-
dinating phosphine group on
a side chain, to favor
tions decrease the acidity, compared to a phenyl ring (pKa
=
N-coordination by chelate effect but without promoting the
evolution to nido derivatives (Fig. 3). Although no studies have
been reported so far on the inductive effect of B3-carboranes,
we have chosen the C- and B3-points of attachment to the car-
borane cage as they are both associated with electron-with-
drawing character, although to a different extent. Thus, these
ligands allow us to compare the different deactivating modifi-
cations that the carborane cage can induce on the nitrogen
donor atom, and provide valuable information regarding the
design principles associated with the phosphine-iminopho-
sphorane ligand system.
7.50).6c,d In the field of coordination chemistry, the carboranyl
group has also been used to modify the coordinating strength
of vicinal donor atoms. Once again, these studies have only
taken into account the deactivating C-carboranyl moiety and
the donating B9-carboranyl group. Published results report the
reduction of the coordinating strength of C-carboranyl phos-
phines10 and thioethers,11 and the enhancement of the
strength for B9-thioethers.11 Some of us have also contributed
to this field by studying the deactivation of the coordinating
ability of an iminophosphorane nitrogen donor atom by effect
of a C-carboranyl group.12,13 Iminophosphoranes, R3P = NR′,
are basic compounds that can act as ligands though the sp2-
hybridized nitrogen atom of their highly polarized PvN bond.
They are predominantly σ-donor ligands with no significant
π-accepting capacity. Although the basicity of the nitrogen
atom depends on its substituent (R′),14 neutral iminopho-
sphoranes can be usually exchanged by other ligands.15
However, the combination with stronger donors produces
more stabilizing chelating ligands that enable the preparation
of stable coordination complexes, some of them with interest-
ing catalytic activities.15,16 Our recent studies on carboranyl-
Fig. 3 Carboranyl phosphine-iminophosphorane ligands (L1–L5)
iminophosphoranes increased the limited knowledge on these designed for this study.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 486–503 | 487

View Article Online
Paper
Dalton Transactions
are described in the literature.24,25 The versatile 1-R-3-NH2-o-
caborane derivatives have been previously used as precursors
of other nitrogen functionalities, like amines,26 isocyanes,27
Results and discussion
Synthesis and characterization of the carborane ligands
The designed carboranyl phosphine-iminophosphorane and diazonium salts,22 which, in turn, have been used to
ligands (L1–L5, Fig. 3) were obtained following the Staudinger prepare other B3-substituted carboranes.22,28 However, they
method,20 i.e. by reaction of B- or C-carboranyl-azides and 1,1- have never been used to prepare 3-azide derivatives. The iso-
bis(diphenylphosphino)methane (dppm). The difference lated B-carboranyl amines (Scheme 2) were converted into the
between the synthesis of C-carboranyl and B-carboranyl imino- desired azides under mild conditions, by reaction with tert-
phosphoranes lies in the different method for the in situ butyl nitrite and azidotrimethylsilane.23 The azides, obtained
preparation of the corresponding carboranyl-azide. In the case in situ, were reacted with one equivalent of diphosphine dppm,
of the C-carboranyl derivatives (ligands L1–L3), the yielding the projected B-carboranyl phosphine-iminophos-
preparations of the R-carboranyl-azides have been described in phoranes L4 and L5 (Scheme 2).
the literature, R = H,12 Me,18,21 Ph.18 The reaction of the
Due to the few number of palladium complexes with phos-
different C-carboranyl azides, obtained in situ, with the phine-iminophosphorane ligands described in the literature
equimolar amount of the diphosphine dppm, yields the (see below), we decided to synthesize the non-carboranyl phos-
projected C-carboranyl phosphine-iminophosphoranes L1–L3 phine-iminophosphorane analogue L6, with a bulky biphenyl
(Scheme 1). This method has already been reported for the substituent on the iminophosphorane nitrogen atom, which
preparation of the other known closo-C-carboranyl imino- provides a good comparison with the carboranyl ligands.
phosphoranes (Fig. 2).12,13,18
Ligand L6 was obtained following the same procedure used for
In the case of the B-carboranyl derivatives (ligands L4 and the B-carboranyl derivatives, i.e. by reaction of diphosphine
L5, Scheme 2), although the synthesis of 3-azide-carborane has dppm with 1-azido-2-phenylbenzene, obtained in situ from
been described in the literature using 3-diazonium-o-carbor- 2-phenyl-aniline (Scheme 3).23
ane tetrafluoroborate as starting material,22 we decided to
The analysis of solid samples of the carboranyl ligands
apply the method described for the preparation of organic aro- L1–L5 by IR spectroscopy confirms the presence of the closo-
matic azides, starting from the corresponding amines.23 Thus, carborane cage, as the spectra exhibit a very intense ν(B–H)
this method requires the previous synthesis of the different band in the range 2566–2588 cm−1, typical of closo-carborane
3-amine-carborane precursors (R = H, Me, Scheme 2), which derivatives. The IR spectra of L1–L6 also show an intense band
Scheme 1 Synthesis of the C-carboranyl phosphine-iminophosphoranes L1–L3.
Scheme 2 Synthesis of the B-carboranyl phosphine-iminophosphoranes L4 and L5.
Scheme 3 Synthesis of the organic phosphine-iminophosphorane L6.
488 | Dalton Trans., 2019, 48, 486–503
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
Table 1 Selected spectroscopic data (IR and 31P NMR) for ligands
L1–L6
(10–12 ppm, R = phenyl rings with –F, –CN or –NO2 substitu-
ents).34,35 The latter values, shifted to low field, are more
similar to those found for the carboranyl ligands L1–L5.
However, it is interesting to note that the pairs of C- and
B-carboranyl analogues show almost the same chemical shifts
for the P(V) atom (5.9 and 5.8 for L1 and L4, respectively, and
9.1 and 8.1 ppm for L2 and L5, respectively), which does not
reflect the more electron-withdrawing character of the
C-carboranyl derivatives compared to the B-carboranyl ones.
IR
31P NMRa
ν(PN)/cm−1
–PPh2
PvN
2JP,P
L1
L2
L3
L4
L5
L6
1370
1304
1375
1433
1374
1429
−30.9
−30.1
−30.5
−30.2
−29.9
−29.6
5.9
9.1
7.4
5.8
8.1
54.7
57.0
56.6
50.8
52.7
52.0
1
The H NMR spectra of the ligands show signals due to the
−1.2
diphosphine fragment and to the groups on the nitrogen
atom; the carborane cage (L1–L5) or the biphenyl group (L6).
The more interesting signals of the diphosphine fragment are
those due to the aliphatic protons of the methylene spacer,
a Chemical shifts in ppm and coupling constants in Hz.
attributable
to
the
ν(PvN)
stretching
frequency which appear as one multiplet in the range 2.89–3.32 ppm,
(1304–1433 cm−1) that proves the formation of the iminophos- due to coupling with both phosphorus atoms. The organic
phorane PvN bond (Table 1). These values are similar to ligand L6 presents a very similar chemical shift for these
those found for other free carboranyl-iminophosphorane protons (3.27 ppm). The 1H NMR spectra of the carboranyl
derivatives described in the literature (Fig. 2), in the range ligands also show signals due to the carborane cage. Except in
1332–1423 cm−1 12,13,18
and also similar to those found for the cases of the disubstituted C-carboranyl ligands (L2 and L3)
,
non-coordinated (non-carboranyl) aromatic derivatives.29 the spectra show signals due to the protons on the cage
A detailed analysis of the data shows the different strength of carbon atoms (Cc–H). The C-carboranyl derivative L1 shows
the PvN bond depending on the type of ligands, which is this signal at 3.69 ppm, slightly deshielded compared to the
reflected in different ranges for the C-carboranyl ligands B-carboranyl analog L4, 3.03 ppm. The substitution of one of
L1–L3, 1304–1375 cm−1, and the B-carboranyl ligands L4 the cage carbon atoms of the B-carboranyl derivative with a
(1433 cm−1) and L5 (1374 cm−1) more similar to the organic methyl group further shifts the Cc–H signal to high field (δ =
ligand L6 (1429 cm−1).
2.69 ppm for L5). The methyl group of the C-carboranyl ligand
The 31P spectra of all of L1–L6 consist of two doublets (see L2 (2.01 ppm) appears only slightly downfield shifted with
Table 1). The signal observed at higher field was assigned to respect to the B-carboranyl ligand L5 (1.90 ppm). These chemi-
the P(III) phosphine group (–PPh2) and the other one to the cal shifts are very similar to those reported for 1-triphenylimi-
P(V) iminophosphorane group (PvN). The chemical shift of nophosphorane-2-methyl-carborane (2.02 ppm).23
the signal due to the phosphine group (–PPh2) does not
The 11B NMR spectra show very similar ranges for the
depend on the nature of the substituent on the nitrogen atom C-carboranyl derivatives (L1–L3), with signals between −15.7
(carboranyl or non-carboranyl), as L6 shows this signal at and −5.9 ppm, typical of closo-carborane compounds. The
−29.6 ppm, while for the carboranyl ligands L1–L5 they are B-carboranyl derivatives, L4 and L5, show signals in wider
observed in the range (−30.9)–(−29.9) ppm. Other unbound ranges, (−23.3)–(2.0) ppm, with the resonance due to the sub-
phosphine-iminophosphorane ligands derived from the stituted boron atom at lower field. This signal, which is the
diphosphine dppm found in the literature also show very only one with a positive shift, is very easy to assign by compar-
similar values for the –PPh2 group, near −29 ppm.30–33 ing the 1H decoupled 11B NMR spectrum with the coupled
However, the signal due to the P(V) iminophosphorane group one, as this is the only boron atom not attached to a hydrogen
(PvN) is sensitive to the nature of the group on the nitrogen atom (see spectra in ESI†). The ranges of signals shown by the
atom. Thus, the carboranyl ligands L1–L5 show these signals B-carboranyl iminophosphoranes are also wider than those
in the range 5.8–9.1 ppm, while the signal found for the imi- found for other nitrogen-substituted B-carboranyl compounds
nophosphorane group in the spectrum of L6 (−1.2 ppm) is found in the literature, like 3-amino-carborane [(−19.14)–(0.14)
shifted to higher field with respect to the carboranyl ligands. ppm],28 3-nitro-carborane [(−14.8)–(−3.5) ppm],26 3-azido-car-
The deshielding of the iminophosphorane phosphorus atom borane [(−14.2)–(−3.8) ppm],26 3-amide derivatives [(−15)–(−3)
by effect of the carboranyl cage reflects its electron-withdraw- ppm, aprox.],22,24,36a 3-triazol derivatives [(−12.9)–(−2.8)
ing character. Other unbound phosphine-iminophosphorane ppm],36b or 3-amine derivatives [(−17)–(−1) ppm, aprox.].22,36c
ligands derived from the diphosphine dppm found in the It is interesting to note that none of the literature examples
literature also show different values for the PvN–R group present positive shifts for the substituted boron atom, indicat-
depending on the electron-withdrawing character of the R ing a more deshielding effect for the iminophosphorane
group on the nitrogen atom. Thus, N-aryl (organic) derivatives substituent.
show values for the iminophosphorane phosphorus atom at
highfield, between 0 and 5 ppm,31–33 while derivatives with
Synthesis and characterization of the palladium complexes
electron-withdrawing groups on the nitrogen atom show the The neutral carboranyl phosphine-iminophosphorane ligands
signals of the P(V) phosphorus atom shifted to lower fields L1–L5, which are potentially (P, N) bidentate ligands, were
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 486–503 | 489

View Article Online
Paper
Dalton Transactions
used to synthesize palladium complexes and thus compare
The IR spectra of the palladium complexes are very infor-
their coordinating abilities. The ligands were stirred overnight mative. The integrity of the closo-carborane cage is reflected in
with the palladium precursor cis-[PdCl2(PhCN)2]37 in the values of the strong ν(B–H) bands, 2551–2587 cm−1, which
a
1 : 1 molar ratio in dry tetrahydrofuran at room temperature show almost the same range as the free closo-carboranyl
(see Schemes 4 and 5). The substitution of the labile benzo- ligands (see before). However, the most interesting band of the
nitrile ligands of the precursor leads to the formation of new IR spectra of the metal complexes is the strong band associ-
palladium complexes, which were precipitated from the solu- ated with the iminophosphorane stretching frequency, ν(PvN)
tion by adding hexane (see Experimental section). The organic (see Table 2). This band is sensitive to the coordination of the
ligand L6 was also used to prepare a similar palladium iminophosphorane group to a metal atom, which weakens the
complex under the same conditions, Pd6 (see Scheme 6).
P–N bond and thus, shifts the band to lower wavenumbers.38
All the palladium complexes Pd1–Pd6 were characterized by The extent of the shift can be used to estimate the strength of
standard spectroscopic techniques, elemental analysis and the metal–ligand interaction. The palladium complexes Pd2
X-ray diffraction analysis. Although the spectroscopic tech- (1336 cm−1
) ) with disubstituted
and Pd3 (1396 cm−1
niques evidence the presence of the ligands in the complexes, C-carboranyl ligands, show ν(PvN) bands at almost the same
the X-ray studies were crucial to assign the coordination values as the corresponding free ligands. These bands are even
modes of the phosphine-iminophosphorane ligands. The X-ray shifted to higher wavenumbers (positive shifts of 32 and
studies (see below) revealed two different coordination modes 21 cm−1, respectively), which indicates that the iminopho-
for these ligands: P-terminal ligands (dimeric complexes Pd2 sphorane nitrogen atom is not coordinated to the palladium
and Pd3, see Scheme 4) and (P, N) chelating ligands (mono- atom in these cases. This result is in accordance with the
meric complexes Pd1, Pd4–6 see Schemes 4–6).
P-terminal coordination mode revealed for these complexes by
Scheme 4 Synthesis of palladium complexes with C-carboranyl phosphine-iminophosphorane ligands, Pd1–Pd3.
Scheme 5 Synthesis of palladium complexes with B-carboranyl phosphine-iminophosphorane ligands, Pd4 and Pd5.
Scheme 6 Synthesis of the palladium complex with the organic phosphine-iminophosphorane ligand, Pd6.
490 | Dalton Trans., 2019, 48, 486–503
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
Table 2 Selected spectroscopic data (IR and 31P NMR) for the palladium complexes Pd1–Pd6a
IR/cm−1
31P NMRb
Coord. Mode
ν(PN)
Δ(PN)
–PPh2
Δδ
PvN
Δδ
2JP,P
Δ(2JP,P
)
Pd1
Pd2
Pd3
Pd4
Pd5
Pd6
(P,N)-Chelate
P-Terminal
P-Terminal
(P,N)-Chelate
(P,N)-Chelate
(P,N)-Chelate
1238
1336
1396
1253
1222
1271
−132
32
21
−180
−152
−158
8.2
20.5
20.6
11.4
9.6
39.1
50.6
51.1
41.6
39.5
42.5
38.2
17.5
2.1
40.9
40.8
41.7
32.3
8.4
5.3
35.1
32.7
42.9
32.7
8.6
13.9
27.6
28.4
35.2
−22
−48.4
−42.7
−23.2
−24.3
−16.8
12.9
a Δ values are the differences with respect to the corresponding free ligands. b Chemical shifts in ppm and coupling constants in Hz.
X-ray crystallography. For the rest of the palladium complexes, iminophosphorane ligands, all of them with (P, N) chelating
with ligands with (P, N) chelating mode, the ν(PN) bands are coordination. By comparison with literature data,30–34,41,42 the
shifted to lower wavenumbers respect to the corresponding assignment of the 31P NMR spectra of these complexes was
bands of the free ligands, due to the formation of the Pd–N reversed compared to the complexes with P-terminal ligands,
bond. However, there is a difference between the shift found i.e. the signal at high field (8.2–12.9 ppm) was assigned to the
for the C-carbonyl derivative Pd1, (Δυ = −132 cm−1), and those –PPh2 group and the signal at low field (38.2–41.7 ppm) was
found for the B-carboranyl derivatives Pd4 (Δυ = −180 cm−1
)
assigned to the PN group (see Table 2). In all cases, both
and Pd5 (Δυ = −152 cm−1), more similar to the shift found for signals are very shifted downfield with respect to the signals of
the non-carboranyl complex Pd6 (Δυ = −158 cm−1).
the free ligands by effect of the (P, N) coordination, although
The palladium complexes were also characterized by NMR for each complex this effect is more pronounced for the phos-
spectroscopy (31P, 1H and 11B). The 31P NMR spectra of the phine group than for the PN group (Table 2), as also found for
palladium complexes show two doublets due to the coupling the literature examples.35a For these cases with (P, N) coordi-
of the two phosphorus groups, phosphine (–PH2) and imino- nation, the magnitude of the 2JPP coupling constants decreases
phosphorane (PN) (see data in Table 2). However, the assign- upon complexation (Δ ≈ −20 Hz), as shown in Table 2. In con-
ment of the peaks is not straightforward, as it depends on the clusion, although the 31P data is in agreement with the struc-
coordination mode of the ligand. The assignment was possible tures found in the solid state by X-ray crystallography, the simi-
due to the knowledge of the coordination mode of the ligands larities among the spectra of the different complexes, regard-
in the complexes (provided by X-ray crystallography) and by less the coordination mode of the ligand, make difficult the
comparison with literature data for other palladium complexes assignment of the signals without the previous knowledge of
with similar organic ligands. The dimeric palladium com- the final structures.
1
plexes Pd2 and Pd3, [PdCl2(L)]2, present P-terminal di-
The H NMR spectra of the palladium complexes show the
substituted C-carboranyl ligands (Scheme 4). The signal due to same signals as the corresponding free ligands, proving the
the phosphine group is not affected by the different substi- presence of the ligands in the complexes. The signal due to
tution of the carborane cage, being observed at 20.5 ppm for the Cc–H group of complexes Pd1 (5.15 ppm) Pd4 (5.16 ppm)
Pd2 and 20.6 ppm for Pd3. These signals are shifted downfield and Pd5 (6.05 ppm) appears shifted to lower field compared to
to ∼50 ppm respect to the signals of the free ligands, which the free ligands. The extent of the shift observed upon ligand
confirms the formation of the Pd–P bond. While metal com- coordination appears to reflect the strength of the Pd–N inter-
plexes with P-terminal organic phosphine-iminophosphorane action, as it is lower for the complex with the more withdraw-
ligands are not reported in, the literature, examples of similar ing C-carboranyl ligand, Pd1 (Δ = 1.43 ppm), than for the com-
palladium dimers with terminal phosphines show one singlet plexes with B-carboranyl ligands, Pd4 (Δ = 2.13 ppm) and Pd5
signal in the 31P NMR spectra, usually between 25 and (Δ = 3.36 ppm). The 1H NMR spectra of the palladium com-
40 ppm, that experiences important shifts with respect to the plexes also show the signals due to the aliphatic protons of the
free ligands.39,40 The other signal in the 31P NMR spectra of methylene spacer. These signals are slightly shifted compared
the dimeric complexes Pd2 (17.5 ppm) and Pd3 (2.1 ppm) can to the corresponding signals of the free ligands, but to a
be assigned to the iminophosphorane phosphorus atom. different extent depending on the particular case, without a
These signals at high field are not very shifted respect to the clear influence of the coordination mode of the ligand. The
corresponding signals of the free ligands (see Table 2), which spectra of the complexes Pd2 and Pd5 also show the singlet
is in agreement with the P-terminal coordination found for the signal due to methyl group at higher field (1.27 and 2.11 ppm
ligands in the solid state. The rest of the synthesized palla- respectively).
dium complexes (Pd1, Pd4–Pd6) present phosphine-iminopho-
The 11B NMR spectra of the palladium complexes with car-
sphorane ligands with (P, N) coordination, as shown by X-ray boranyl ligands show broad resonances in the range (−0.4)–
crystallography. The literature reports several examples of (−17.9) ppm, confirming the presence of the closo-carborane
palladium complexes with non-carboranyl phosphine- cage. The complexes with B-carboranyl ligands [(−0.4)–(−17.9)
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 486–503 | 491

View Article Online
Paper
Dalton Transactions
ppm] show wider ranges than the complexes with C-carboranyl
ligands [(−3.9)–(−17.1) ppm], although to a lesser extent than
the free ligands (see before). The spectra of the B-carboranyl
derivatives Pd4 and Pd5, show the signal due to the B3 boron
atom at higher field (−0.4 and −1.2 ppm, respectively) than
the corresponding free ligand (1.4 ppm for L4 and 2.0 ppm for
L5), by effect of the Pd–N coordination.
Crystal structures of the metal complexes Pd1–Pd6
Single crystals suitable for X-ray diffraction analysis of all the
palladium complexes were obtained by slow concentration of
solutions of the complexes in mixtures of dichloromethane
and hexane, except in the case of Pd6, that was recrystallized
from a mixture of chloroform and hexane. The molecular
structures of the complexes are shown in Fig. 4–9. The crystal-
lographic data can be found in Table S1 (ESI†). Selected bond
lengths and angles for these compounds are collected in
Tables 3–5.
Fig. 6 Molecular structure of compound Pd1. Thermal ellipsoids are
shown at the 40% probability level. Hydrogen atoms are omitted for
clarity.
The crystal structures of the palladium complexes can be
grouped in two different classes according to the coordination
mode of the potential (P, N) donating ligands: (i) dimeric com-
Fig. 7 Molecular structure of compound Pd4. Thermal ellipsoids are
shown at the 40% probability level. Hydrogen atoms are omitted for
clarity.
Fig. 4 Molecular structure of compound Pd2. Thermal ellipsoids are
shown at the 40% probability level. Hydrogen atoms are omitted for
clarity.
Fig. 5 Molecular structure of compound Pd3. Thermal ellipsoids are
shown at the 40% probability level. Hydrogen atoms are omitted for
clarity.
Fig. 8 Molecular structure of compound Pd5. Thermal ellipsoids are
shown at the 40% probability level. Hydrogen atoms are omitted for
clarity.
492 | Dalton Trans., 2019, 48, 486–503
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
respectively). The equivalent ligands are related by an inver-
sion center. The ligands that show this coordination mode are
disubstituted C-carboranyl derivatives with a phosphine-imino-
phosphorane group on one cage carbon atom and a methyl
(L2) or phenyl (L3) group on the other cage carbon atom. The
inversion center at the middle point of the line than join the
two palladium atoms reflects the planarity of the central
[(μ-Cl)2Pd2] four-membered rings and the anti-disposition of
the terminal ligands around the central Pd2Cl2 unit. Each
palladium atom presents a [PCl3] coordination environment in
a distorted square planar geometry. The main source of distor-
tion with respect to the ideal square planar geometry is the for-
mation of the central [(μ-Cl)2Pd2] ring [Cl–Pd–Cl#1: 85.72(3)°
for Pd2 and 85.31(3)° for Pd3]. The coordination plane
[PdPCl3] is essentially planar in both cases [rms: 0.0735 Å
(Pd2) and 0.0602 Å (Pd3)]. Both complexes present the same
conformation, with the carborane cages occupying the space
above and below the central [(μ-Cl)2Pd2] ring, and the methyl
(Pd2) and phenyl (Pd3) substituents pointing towards this
ring.
Fig. 9 Molecular structure of compound Pd6. Thermal ellipsoids are
shown at the 40% probability level. Hydrogen atoms are omitted for
clarity.
In the case of the carboranyl P-terminal ligands, the struc-
tural parameters that involve the iminophosphorane nitrogen
atom are not affected by the coordination to the palladium
atom. The non-coordinated nitrogen atom can engage in exo-
π-bonding with the carborane cage, which produces the
elongation of the Cc–Cc distance and shortens the N–Cc
plexes with two P-terminal ligands and (ii) monomeric com-
plexes with one (P, N) chelating ligand.
(i) P-terminal ligands. Compounds Pd2 (Fig. 4) and Pd3
(Fig. 5) are neutral dinuclear palladium(II) complexes with a di-
μ-chloro bridge, two equivalent terminal chlorine ligands and
two equivalent P-terminal carborane ligands (L2 and L3,
Table 3 Selected bond lengths (Å) for Pd1–Pd6
Pd–N
Pd–P
Pd–Cl1
Pd–Cl2
Pd–Cl2#1
PvN
N–Cc/B
Cc–Cc
Pd1a
2.132(3)
2.118(3)
—
2.2248(10)
2.2219(9)
2.2099(7)
2.2120(8)
2.2258(17)
2.2243(6)
2.2135(10)
2.2948(9)
2.2888(9)
2.2853(8)
2.2699(9)
2.3012(16)
2.2908(6)
2.2999(10)
2.3725(9)
2.3717(9)
2.3486(8)
2.3251(8)
2.3878(17)
2.4023(6)
2.3789(9)
—
—
1.628(3)
1.621(3)
1.586(2)
1.568(3)
1.606(4)
1.6075(18)
1.605(3)
1.417(4)
1.430(4)
1.388(3)
1.366(4)
1.437(8)
1.461(3)
1.430(4)
1.680(5)
1.665(5)
1.769(4)
1.786(5)
1.628(7)
1.623(3)
—
Pd2
Pd3
Pd4
Pd5
Pd6
2.4209(7)
2.4210(8)
—
—
—
—
2.103(4)
2.0972(17)
2.059(3)
a Two molecules per asymmetric unit. Symmetry transformation used to generate equivalent atoms #1: −x + 1, −y + 1, −z + 1 (Pd2) and #1: −x + 1,
−y, −z + 1 (Pd3).
Table 4 Selected bond angles (°) for the dimeric complexes Pd2 and Pd3
P2–Pd–Cl1
P2–Pd–Cl2
P2–Pd–Cl2#1
Cl1–Pd–Cl2
Cl1–Pd–Cl2#1
Cl2–Pd–Cl2#1
C1/B3–N1–P1
Pd2
Pd3
92.83(3)
86.96(3)
90.37(3)
95.09(3)
175.52(3)
178.22(3)
170.52(3)
172.54(3)
90.70(3)
92.43(3)
85.72(3)
85.31(3)
127.5(2)
132.7(2)
Table 5 Selected bond angles (°) for the monomeric complexes Pd1, Pd4–Pd6
P–Pd–N
P–Pd–Cl1
P–Pd–Cl2
N–Pd–Cl1
N–Pd–Cl2
Cl1–Pd–Cl2
Cc/B–N–P
Dihedralb
Pd1a
90.51(8)
88.13(8)
89.45(13)
89.95(5)
88.34(9)
86.34(3)
88.36(3)
87.18(6)
86.65(2)
87.22(4)
174.51(4)
177.73(3)
179.47(6)
171.618(18)
173.40(4)
176.57(8)
176.06(8)
176.58(13)
172.74(4)
173.76(9)
92.34(8)
92.69(8)
90.68(13)
91.62(5)
91.80(9)
90.91(3)
90.89(3)
92.69(7)
92.68(2)
93.11(3)
125.4(2)
126.0(2)
125.2(4)
124.30(15)
120.3(2)
4.86(8)
2.75(7)
0.79(16)
10.23(4)
7.75(9)
Pd4
Pd5
Pd6
a Two molecules per asymmetric unit. b Dihedral angle between the PdPN plane and the PdCl2 plane.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 486–503 | 493

View Article Online
Paper
Dalton Transactions
bond.43 Thus, these compounds show long Cc–Cc distances ively],39 or [PdCl2(PPh2CH2Ph)]2, [2.273, 2.412 and 2.315 Å,
[1.769(4) for Pd2 and 1.786(5) Å for Pd3] (the longest presented respectively].40
in this paper), and short N–Cc distances [1.388(3) for Pd2 and
In conclusion, the presence of the non-coordinated carbora-
1.366(4) Å for Pd3] (the shortest of this paper). The non-coordi- nyl-iminophosphorane moiety does not affect the coordinating
nation of the nitrogen atom is also reflected in the P–N dis- abilities of the phosphine group of the ligand, which behaves
tances of the iminophosphorane groups, which present short like a typical terminal phosphine ligand.
values [1.586(2) Å for Pd2 and 1.568(3) Å for Pd3], and also in
(ii) (P, N) chelating ligands. Compounds Pd1 (Fig. 6), Pd4
opened Cc–N–P angles [127.5(2) for Pd2 and 132.7(2) Å for (Fig. 7), Pd5 (Fig. 8), and Pd6 (Fig. 9) are neutral monomeric
Pd3]. All these parameters (Cc–Cc, N–Cc and P–N distances; palladium(II) complexes in which the metal atom is co-
Cc–N–P angles) are very similar to those found for the other ordinated to two cis terminal chlorine ligands and to the phos-
non-coordinated carboranyl iminophosphorane ligands found phorus and nitrogen donor atoms of a chelating phosphine-
in the literature,12,13,18 especially for the very similar methyl- iminophosphorane ligand (L1, L4, L5, and L6, respectively).
and phenyl- carborane derivatives of triphenylimino- The carborane ligands that show this coordination mode are
phosphorane.18
the C-carboranyl derivative L1, with no other substituents on
The Pd–P distances [2.2099(7) for Pd2 and 2.2120(8) Å for the cage carbon atoms, and the B3-carboranyl ligands L4, with
Pd3], trans to a bridging chlorine atom, are in the expected unsubstituted cage carbon atoms, and L5, with a methyl group
range for this type of geometry. Similar values are found in on a Cc atom. The non-carboranyl ligand L6 synthesized for
other palladium dimers with related terminal phosphines comparison purposes also provides the (P, N) chelate coordi-
(Ph2P–CH2–), like for example [PdCl2(PPh2Pr)]2, 2.227 Å,39 or nation mode in the palladium complex Pd6.
[PdCl2(PPh2CH2Ph)]2, 2.221 Å.40 The effect of the coordination
The coordination of the chelating (P, N) ligands gives rise
of the phosphine phosphorus atom to the palladium metal to five-membered chelate rings, which present an envelope dis-
atom is also reflected in the P–C bond distances, but this is position. The N1–Pd1–P2–C fragment is planar [rms: 0.0449
only made clear by comparison with data found in the litera- and 0.1451 Å for Pd1; 0.0673 Å for Pd4; 0.0283 Å for Pd5;
ture for free phosphine-iminophosphorane ligands. The litera- 0.0185 Å for Pd6] and forms an angle of 53.19(16)° and 49.83
ture only presents four examples of similar free ligands (18)° (Pd1), 53.85(23)° (Pd4), 50.07(9)° (Pd5) and 43.96(16)°
derived from the diphosphine dppm.35 These examples show (Pd6) with the N1–P1–C plane. The coordination of the chelate
that the iminophosphorane P(V)–C distances (1.787–1.808 Å)35 ligands only produces a small distortion with respect to the
are shorter than the phosphine P(III)–C distances regular geometry, which is reflected in P–Pd–N chelate angles
(1.826–1.864 Å)35 and that, in both cases, the distances that in the range 88.13(8)–90.51(8)°. The deviation of the trans
involve the phenyl rings are always slightly shorter [P(V)–C(Ph): angles respect to the value of 180° is a reflection of the planar-
1.787–1.806 Å; P(III)–C(Ph): 1.826–1.839 Å] than the corres- ity of the coordination plane, [PdPNCl2]. For example (Table 5),
ponding distances that involve the aliphatic spacer [P(V)–C the more distorted trans angles [171.618(18)° and 172.74(4)°,
(CH2): 1.799–1.808 Å; P(III)–C(CH2): 1.853–1.864 Å].35 In the Pd5] correspond to the structure with a higher dihedral angle
case of the carboranyl complexes Pd2 and Pd3, although the between the PdPN plane and the PdCl2 plane [10.23(4)°], while
P(V)–C distances involving the iminophosphorane P atom are the more regular trans angles [179.47(6)° and 176.58(13)°, Pd4]
almost unaffected by the coordination of the ligand [P(V)–C correspond to the structure with the lower dihedral angle
(Ph): 1.798(3)–1.806(3) Å; P(V)–C(CH2): 1.815(3) for Pd2 and between these planes [0.79(16)°].
1.820(3) Å for Pd3], the P(phosphine)–C distances [P(III)–C(Ph):
In the case of the complex Pd5 with a B3–N–P connection
1.799(3)–1.816(3) Å; P(III)–C(CH2): 1.824(3) Å for Pd2 and 1.815 and a methyl substituent on one cage carbon atom, the methyl
(3) Å for Pd3] are shorter than those found for the literature group points away from the coordination plane, while the Cc–H
free ligands and more similar to P(V)–C distances.
group point to the coordination plane and engage in an intra-
The Pd–Cl bond distances depend on the coordination molecular Cc–H⋯Cl interaction [d(H⋯Cl): 2.373 Å]. In the case
mode of the chlorine atom and on the donor atom located on of complex Pd4, with a B3–B–P connection and no substituents
trans position (trans influence). The shortest Pd–Cl distances on the Cc atoms, both Cc–H groups engage in intramolecular
correspond to the terminal Cl ligand, trans to a bridging chlor- Cc–H⋯Cl interactions [d(H⋯Cl): 2.683 and 2.854 Å].
ine atom [2.2853(8) Å for Pd2 and 2.2699(9) Å for Pd3]. The
The Pd–Cl distances are a reflection of the different trans
other two Pd–Cl distances are different (asymmetric chloro influence of the phosphine phosphorus atom and the imino-
bridge) due to the different trans influence of the phosphine phosphorane nitrogen atom. Thus, the values found for the
group and the terminal chlorine ligand. The Pd–Cl bond dis- Pd–Cl distances trans to the more donating phosphine phos-
tances trans to the phosphine phosphorus atom are longer phorus atom [2.3717(9)–2.4023(6) Å, mean value: 2.3826 Å] are
[2.4209(7) for Pd2 and 2.4210(8) Å for Pd3] and the Pd–Cl longer than the Pd–Cl distances trans to the nitrogen atom
bond distances trans to the terminal chlorine ligand are [2.2888(9)–2.3012(16) Å, mean value: 2.2951 Å]. These values
shorter, [2.3486(8) Å for Pd2 and 2.3251(8) Å for Pd3]. The are similar to the values found for PdCl2 adducts of organic
same pattern of Pd–Cl distances is found in the literature for (P, N) chelating phosphine-iminophosphorane ligands (12 struc-
similar palladium dimers with terminal phosphines, for tures, CSD), with mean values of 2.371 Å for Pd–Cl(trans-P) dis-
example [PdCl2(PPh2Pr)]2 [2.268, 2.444 and 2.321 Å, respect- tances and 2.302 Å for Pd–Cl(trans-N) distances.35–38,46,48,49
494 | Dalton Trans., 2019, 48, 486–503
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
The Pd–P distances that involve the carboranyl ligands atom connected to a boron atom [P–N: 1.606(4) and 1.6075(18)
(complexes Pd1, Pd4 and Pd5) are all very similar [2.2219(9)– Å, respectively]. The slightly shorter values found for the
2.2258(17) Å, mean value: 2.2242 Å], and the corresponding B-carboranyl derivatives are more similar to the value found
distance found in Pd6, with the organic ligand L6, is only for the organic analog Pd6, 1.605(3) Å, which in turn is similar
slightly shorter, 2.2135(10) Å. The comparison of these dis- to the mean value of 1.601 Å found in the literature for palla-
tances with those found in the literature for PdCl2 adducts of dium complexes with organic (P, N) chelating phosphine-
organic phosphine-iminophosphorane ligands, shows
a
iminophosphoranes (CSD, 22 structures).30–34,42,44–47 It is
similar mean value, 2.2266 Å, although in a wider range, interesting to note that complex Pd1 (Cc–N–P connection)
2.203–2.253 Å.31–34,42,44,45 This result indicates that the shows a weaker Pd–N interaction than Pd4, Pd5 and Pd6, as
carboranyl moiety does not affect the Pd–P interaction to a judged by the Pd–N distances (see before), so the higher P–N
great extent.
elongation is an effect of the C-carboranyl group. This result
The Pd–N distances found for Pd1, Pd4 and Pd5 (carboranyl was not observed in the case of the literature nido exo-
ligands), 2.0972(17)–2.132(3) Å (mean value: 2.1125 Å), are phosphine-iminophosphoranes, which show a normal elonga-
clearly longer than the value found for (non-carboranyl) Pd6, tion of the P–N bond upon Pd–N coordination [1.6100(14) and
2.059(3), which is similar to the mean value of 2.073 Å found 1.611(5) Å],17 indicating that the higher P–N elongation is an
in the literature for the PdCl2 adducts of organic phosphine- effect of the closo-C-carboranyl cage.
iminophosphorane ligands.31–34,42,44,45 This result indicates
that the carboranyl moiety reduces the coordinating ability of the Pd–N coordination also affects the Cc–Cc and the Cc–N
the iminophosphorane nitrogen atom. bond distances, as it competes with the exo-π bonding to the
In the case of compound Pd1, with a Cc–N–P connection,
It is also interesting to compare the structural data of the cage carbon atoms. Thus, the Pd–N coordination leads to
chelating carboranyl ligands with the data described in the lit- shorter Cc–Cc distances [1.665(5), 1.680(5) Å] and longer Cc–N
erature for palladium complexes with carboranyl phosphine- distances [1.417(4), 1.430(4) Å] than those found in the litera-
iminophosphoranes. The two examples described recently,13 ture for a non-coordinated carboranyl iminophosphorane with
are palladium complexes with (P, N) nido-carboranyl phos- no other substituents on the other cage carbon atom, 1.688(3)
phine-iminophosphorane ligands. The Pd–N distances found Å and 1.368(2) Å, respectively.12 In the case of the complexes
for these (P, N)-chelating nido ligands 2.0447(13) Å (trans to a Pd4 and Pd5 (B3–N–P connection), with no donor atoms on
bridging chlorine ligand) and 2.072(4) Å (trans to a terminal the cage carbon atoms, the Cc–Cc distances are very short
chlorine ligand), are more similar to the values found for the [1.628(7) and 1.623(3) Å, respectively], even shorter than the
organic derivatives (see before), indicating that only the closo- value found for free o-carborane [Cc–Cc distance: 1.630 Å].48
carboranyl derivatives show the reduction of the donating abil- Unexpectedly, the shortest distance, 1.623(3) Å, is found for
ities of the nitrogen atom by effect of the carborane cage.
Pd5, with a methyl substituent on a cage carbon atom, while
The (P, N) coordination of the phosphine-iminophosphor- the longest distance, 1.628(7) Å, is found for Pd4, with no sub-
ane ligands is also reflected in other structural parameters. As stituents on Cc atoms. The B3–N distances [1.437(8) for Pd4
mentioned before for the complexes with P-terminal ligands and 1.461(3) Å for Pd5] are longer than the mean value of
(Pd2 and Pd3), the Pd–P coordination is reflected in the P–C 1.379
Å found for free organic boryl-iminophosphorane
distances, that follow the same pattern. Thus, the coordination ligands (20 structures, CSD).
reduces the P(phosphine)–C bond distances in comparison to
Finally, the P–N coordination of the chelating carboranyl
the values found in the literature for free phosphine-imino- ligands is also reflected in closed Cc/B3–N–P angles
phosphoranes (see before) [P(III)–C(Ph): 1.791(4)–1.822(3) Å; [124.30(15)–126.0(2)°, mean value: 125.2°], compared to the
P(III)–C(CH2): 1.831(6)–1.849(2) Å], but it does not affect the angles found for the P-terminal ligands in Pd2 [127.5(2)°] and
P(iminophosphorane)–C bond distances [P(V)–C(Ph): 1.788(6)– Pd3 [132.7(2)°]. However, these angles are more opened than
1.800(4) Å; P(V)–C(CH2): 1.786(4)–1.8128(18) Å]. It is interesting the one found for the organic analog Pd6, 120.3(2)°, possibly
that there are no clear differences among the complexes with due to the steric bulk of the carborane cage.
carboranyl ligands and the organic analog Pd6, showing a
similar Pd–P interaction. The Pd–N coordination also affects
some structural parameters that involve the nitrogen atom.
The coordination of the iminophosphorane nitrogen atom to
Discussion
the metal weakens the P–N bond,38 which is reflected in P–N The Cambridge Structural Database (Version 1.22, 2018)
bond distances [1.605(3)–1.628(3) Å] that are longer than those collects 24 crystal structures of palladium complexes with
found for non-coordinated carboranyl-iminophosphoranes, organic (non carboranyl) (P, N) phosphine-iminophosphorane
like the known literature examples [1.5659(17)–1.5870(16) ligands,30–34,42,44–47 of which 14 are PdCl2 adducts.31–34,42,44,45
Å],12,13,18 or the P-terminal complexes Pd2 [1.586(2) Å] and Pd3 In all cases the ligands behave as (P, N) chelating ligands,
[1.568(3) Å] described before. A more detailed analysis reveals regardless the spacer between the donor atoms or the substitu-
minor differences among Pd1, with the nitrogen atom con- ent on the iminophosphorane nitrogen atom. The only excep-
nected to the cage carbon atom [P–N: 1.621(3), 1.628(3) Å], and tions are two PdCl2 adducts with P-terminal ligands, although
Pd4 [1.606(4) Å] and Pd5 [1.6075(18) Å], with the nitrogen in those cases the P-terminal coordination is forced by the
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 486–503 | 495

View Article Online
Paper
Dalton Transactions
stoichiometry of the complexes, [PdCl2(κ1-P)2] or [PdCl2(κ1-P) inductive effect of the methyl group is higher when the nitro-
(L′)] where (κ1-P) is the P-terminal phosphine-iminophosphor- gen atom is connected to the other cage carbon atom
ane ligand and L′ is a neutral coligand.44b Indeed, the same (C-carboranyl derivative L2) than when it is connected to the
ligands show (P, N) chelate coordination when both coordi- B3 boron atom (B-carboranyl derivative L5). The natural
nation sites are available.44b The palladium complex Pd6 pre- charges (Table S2†) confirm the inductive effect of the methyl
sented in this paper, with a (P, N) chelating organic phos- and phenyl substituents, although the inductive effect calcu-
phine-iminophosphorane ligand, follows the same trend.
lated for both groups is less accused. Thus, the calculations
The conclusion that the only known coordination mode for show the donating character of the methyl group when con-
organic phosphine-iminophosphorane ligands is the (P, N) nected to a cage carbon atom, which is in contrast with the
chelation, providing both sites are available, is in line with the acceptor character observed when the methyl group is con-
structures found for the palladium complexes Pd1, Pd4 and nected to cage boron atoms.50
Pd5, but in contrast with the P-terminal coordination found
We next performed DFT calculations on the [PdLCl2] com-
for the palladium dimers Pd2 and Pd3. The (P, N) chelating plexes (L = L1–L5), with the ligand providing a chelating (P,N)
mode found for Pd1, Pd4 and Pd5 confirms that the phos- coordination. The DFT calculations provide calculated Pd–N
phine group can promote the coordination of a carboranyl distances (Fig. 10) that can be used to estimate the coordinat-
iminophosphorane nitrogen atom,13 despite its usual reluc- ing strength of the nitrogen atom. The calculations show that
tance to bind a palladium atom.12 However, it is interesting to the derivatives with the iminophosphorane unit at the cage
note that the carboranyl ligands maintain their closo structure carbon atom (L1–L3) present longer Pd–N distances than the
in the final complexes. Thus, the (P, N) coordination does not B-substituted analogues (L4, L5), which is in line with the
promote the evolution to nido derivatives, as was observed trends in Mulliken charges described above. The calculated
when both donor groups, phosphine and iminophosphorane, Pd–N distances also show the combined electronic and steric
were directly attached to the cage carbon atoms.13 The kind of effects that the methyl or phenyl substituents on the cage
dimeric structure found for Pd2 and Pd3, with a central carbon atom induce on the nitrogen donor atom. The much
[(μ-Cl)2Pd2] ring and a [PCl3] coordination environment, is very longer Pd–N distance calculated for the hypothetical complex
common for (1 : 1) PdCl2 adducts of monophosphines, but [Pd(L3)Cl2], with a phenyl substituent, is easy to rationalize as
rare for potentially bidentate ligands. Thus, although other both electronic and steric factors cooperate to the elongation
potentially (P, N) donating ligands are also known to form this of the Pd–N bond. The case of the methyl group is particularly
kind of dimers through P-terminal coordination,49 the nitro- interesting as it leads to longer Pd–N distances compared to
gen atom of the iminophosphorane group is always donating the unsubstituted ligands. The higher lengthening calculated
enough to promote (P, N) coordination in organic phosphine- for the methyl substituted B-carboranyl ligand L5, compared
iminophosphoranes.
to the substituted C-carboranyl analog L2, is a reflection of the
Aiming to rationalize the different structures observed for different inductive effect of the methyl group, as the steric
the Pd1–Pd5 complexes, we performed a density functional hindrance is exactly the same in both cases. The better trans-
theory (DFT) investigation (see computational details below). mission of the donating effect of the methyl group in the
We started the study by optimizing the structures of the C-carboranyl ligand L2, as suggested by Mulliken charges of
ligands, L1–L5. The Mulliken charges of the N atoms calcu- the unbound ligands (see before), compensates its steric hin-
lated using DFT (Table S2, ESI†) reflect the different (elec- drance and leads only to a small elongation of the Pd–N dis-
tronic) inductive effect of the carboranyl substituents. These tance. In the case of the B-carboranyl analog L5, the steric
charges are clearly more negative for the B-carboranyl ligands
L4 and L5 (−0.83 and −0.90, respectively) than for the
C-carboranyl ones, L1–L3 (−0.05, −0.24 and +0.16, respect-
ively), which show a rather strong electron-withdrawing effect.
The ν(PN) stretching frequencies of L1 and L2 (1349 and
1357 cm−1) are calculated at lower wavenumbers than those of
L4 and L5 (1409–1393 cm−1), in line with the experimental
data (Table S2†). The natural charges obtained with natural
population analysis (NPA) show the same trend, calculating
more negative charges for B-carboranyl derivatives (Table S2†).
This suggests that the higher electron-withdrawing effect of
the carborane cage in C-carboranyl derivatives weakens the
iminophosphorane bond. The Mulliken charges also show the
different inductive effect that the substituents on the cage
carbon atoms exert on the nitrogen atom. The calculations
confirm the electron-withdrawing effect of the phenyl substitu-
Fig. 10 Pd–N bond distances calculated using DFT for the mono-
nuclear complexes [PdLCl2] with chelating (P,N) ligand coordination. The
ent (L3) and the electron-donating character of the methyl
distances become longer as the steric demand of the substituent at the
group (L2 and L5). However, it is important to note that the Cc atom (H, Me or Ph) increases.
496 | Dalton Trans., 2019, 48, 486–503
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
effects determine the final Pd–N distance, as the inductive
effect is less accused.
Conclusions
The nature of the Pd–N bonds was also characterized by the We have synthesized five new carboranyl phosphine-iminopho-
values of the electron density at the bond critical points (ρBCP sphorane ligands with the carboranyl group directly attached
)
and the corresponding Laplacian values (∇2ρBCP).51 Longer to the iminophosphorane nitrogen atom through a cage
Pd–N distances are associated to lower electron densities at carbon atom (C-carboranyl derivatives L1–L3) or through the
the bond critical points (ρBCP) and lower Laplacian bond B3 boron atom (B-carboranyl derivatives L4 and L5), and the
orders (LBOs),52 reflecting weaker bonds (Table S3, ESI†). The phosphine group on a side chain derived from the dipho-
positive ∇2ρBCP values obtained for all complexes (0.31–0.34) sphine dppm, i.e. with a two-atom spacer between the P and N
are typical of Pd–N bonds,53 and indicate a mostly ionic inter- donor atoms (see Fig. 3). The Cc- and B3-carboranyl connec-
action with a minor covalent contribution suggested by the tions are associated with different electron-withdrawing char-
negative values of the total electron energy density (H(r)).54 In acter, although they present the same steric bulk. Thus, the
the case of the B-carboranyl derivatives (L4 and L5) the intro- designed ligands were used to study how the different induc-
duction of a methyl group at a cage carbon atom (L5) results in tive effect of these substituents can modify the coordinating
longer Pd–N bonds and lower ρBCP values, indicating that the ability of the nitrogen atom. The experimental results show
substituent hinders the approximation of the N atom to the Pd that although some complexes present the expected (P, N)
center. In the C-substituded derivatives L1 and L2 the Pd–N coordination mode that is always found for non-carboranyl
bonds are characterized by nearly identical distances, ρBCP analogs, the disubstituted C-carboranyl ligands L2 and L3
values and LBOs, reflecting the compensation of the steric hin- present a P-terminal coordination mode in the dimeric palla-
drance with the electron-donating character, as commented dium complexes Pd2 and Pd3.
before.
The DFT calculations have shown that the (donating) induc-
In practice, the B-carboranyl unit, although slightly elec- tive effect of a methyl substituent on a cage carbon atom is
tron-withdrawing, does not deactivate the coordinating charac- better transmitted to a nitrogen donor atom on the other cage
ter of the iminophosphorane nitrogen atom to a great extent. carbon atom than to a nitrogen donor atom on the adjacent
The complexes with B-carboranyl ligands, Pd4 and Pd5, show B3- position, where it mainly acts as a steric group. The calcu-
the expected (P, N) chelating coordination, which indicates lations also show the great deactivation associated with the
that the destabilization of the Pd–N interaction by the steric C-carboranyl group, which marks the limit for the Pd–N bond
hindrance of the methyl group (Pd5) is not enough to prevent formation. The attempts to further modulate the donor
the formation of the Pd–N bond. However, in the case of the strength of the nitrogen atom by substituting the other cage
more electron-withdrawing C-carboranyl analogs Pd1–Pd3, carbon atom (Pd2 and Pd3) resulted in an unexpected change
although the unsubstituted ligand (Pd1) can still promote the of coordination mode from (P, N) chelating to P-terminal
expected chelate coordination, even the small destabilization coordination. This indicates that further modulation of the
induced by the methyl group (Pd2) can hinder the formation donating strength of the nitrogen atom by functionalization of
of the Pd–N bond, preventing the formation of a very stable the other cage carbon atom is not possible. Thus, the
five-membered chelate ring and forcing a P-terminal coordi- B-carboranyl ligands are possibly more versatile, as they can be
nation of the ligand. The unsubstituted C-carboranyl ligand L1 further functionalized on the cage carbon atoms for a finer
seems to mark the limit for the Pd–N bond formation, as tuning of the coordinating strength of the nitrogen atom. The
further substitution of the other cage carbon atom leads to the modulation of the deactivation of the nitrogen atom is very
total deactivation of the nitrogen coordination capacity, even important as it can lead to hemilabile behavior, which is very
with small (donating) methyl groups.
interesting for catalytic applications.
The relative stability of both coordination modes was also
We have shown that the carborane cage can produce
studied by calculating the relative free energies of the chelating different deactivating effects without changing its steric pro-
(P,N) species with respect to the dinuclear palladium(II) com- perties. Thus, regardless the point of connection, the carbor-
plexes with a di-μ-chloro bridge. The results favor the chelating ane cage always retains its bulky character. The large alkyl
(P,N) coordination by 5.6, 1.7 and 0.1 kcal mol−1 for the com- groups that are generally used to increase the steric bulk
plexes of L1, L2 and L3, respectively. Thus, according to our around a certain atom are associated with electron-donating
calculations the introduction of the methyl group stabilizes character. In contrast, the carborane substituent combines
the dinuclear complex by ca. 4 kcal mol−1. The relative free steric bulk with (tuneable) electron-withdrawing character.
energies of the mononuclear Pd4 and Pd5 complexes with
respect to the dinuclear complexes with P-terminal coordi-
nation favor the mononuclear species by 45.9 and 42.9 kcal Experimental and computational
mol−1, respectively. The comparison of these energy data with
section
Instrumentation
the corresponding values obtained for Pd1 and Pd2 (5.6 and
1.7 kcal mol−1) highlights the dramatic effect that the different
substitution of the carboranyl moiety (B- or C-substitution) has Elemental analysis were performed using a Thermo Finnigan
on the coordinating ability of the N atom.
Flash 1112 microanalyzer. Solid state ATR-IR spectra were
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 486–503 | 497

View Article Online
Paper
Dalton Transactions
recorded on a high-resolution spectrometer FT-IR PerkinElmer BH), 3.23 (d, 2H, P-CH2-P), 3.72 (s, 1H, Ccage-H), 7.27 (m, 10H,
Spectrum Two. The 1H NMR and 31P NMR spectra were PPh2), 7.39 (m, 4H, m-PPh2), 7.50 (m, 2H, p-PPh2), 7.58 (m, 4H,
2
recorded on a Varian Mercury 300 spectrometer. The 11B NMR o-PPh2) ppm. 31P{1H} NMR (CDCl3) δ (ppm): −30.9 (d, JPP
=
2
spectra were recorded on a Varian Inova 500. All NMR spectra 54.7 Hz, –CH2PPh2), 5.9 (d, JPP = 54.7 Hz, PvN) ppm. 11B
were recorded in CDCl3 solutions at 25 °C. Chemical shift NMR (CDCl3) δ (ppm): −15.7, −14.8, −13.3, −12.2, −11.0, −5.9.
1
values for H NMR were referenced to SiMe4 (TMS), those for IR (ATR, ν/cm−1): 3057w, 2613s, 2588vs ν(B–H), 2562vs, 1434vs,
31P were referenced to 85% H3PO4, and those for 11B{1H} NMR 1370s ν(PvN), 1326s, 1307s, 1117s, 1107s, 770s, 692vs. MS (EI,
were referenced to external BF3·OEt2. Chemical shifts are m/z): 541 (27.8%) [L]+, 464 (66.5%) [L-Ph]+, 342 (7.5%)
reported in units of parts per million downfield from the refer- [L-CH2PPh2]+, 262 (25.1%) [PPh3]+, 199 (96.2%) [CH2PPh2]+,
ence, and all coupling constants are reported in Hertz. Mass 185 (56.2%) [PPh2]+, 121 (100.0%) [CH2PPh]+. Elemental ana-
spectra of the ligands L1–L6 were determined on a Micromass lysis calcd (%) for C27H33B10NP2: C 59.8, H 6.1, N 2.6; found: C
Autospec instrument (positive electronic impact). MALDI-TOF 59.2, H 6.1, N 2.5.
mass spectra of the metal complexes Pd1–Pd6 were recorded
in negative-ion mode on a Bruker Autoflex instrument. In the
Synthesis of L2
case of the dimeric complexes Pd2 and Pd3, their mass spectra
Starting materials. ortho-1-Me-carborane (0.5 g, 3.14 mmol);
were also determined in the ESI+ mode, on a Bruker Microtof n-BuLi 1.6 M in hexanes (2.15 mL, 3.45 mmol); tosyl azide
instrument.
(0.55 mL, 3.45 mmol) and dppm (1.39 g, 3.61 mmol). Yield:
0.889 g (51%); white solid. 1H NMR (CDCl3), δ (ppm):
1.12–3.02 (bm, 10H, BH), 2.01 (s, 3H, Ccage-CH3), 3.32 (d, 2H,
Materials
All reactions were performed under atmosphere of dinitrogen P-CH2-P), 7.28 (m, 10H, PPh2), 7.38 (m, 4H, m-PPh2), 7.51 (m,
employing standard Schlenk techniques. Tetrahydrofuran, 2H, p-PPh2), 7.63 (m, 4H, o-PPh2) ppm. 31P{1H} NMR (CDCl3) δ
2
2
acetonitrile and dichloromethane were purchased from Merck (ppm): −30.1 (d, JPP = 57.0 Hz, –CH2PPh2), 9.1 (d, JPP = 57.0
and distilled from sodium benzophenone, P4O10 and calcium Hz, PvN) ppm. 11B NMR (CDCl3) δ (ppm): −14.5, −13.4,
hydride, respectively, previously to use. Commercial grade −12.7, −12.2, −9.1. IR (ATR, ν/cm−1): 3056w, 2566vs ν(B–H),
diethyl ether, ethyl acetate, hexane, chloroform and dichloro- 1481w, 1433m, 1372w, 1304s ν(PvN), 1121m, 1108m, 1087m,
methane
were
used
without
further
purification. 772m, 741m, 687vs. MS (EI, m/z): 555 (12.8%) [L]+, 478 (9.8%)
Demineralized water was used in the aqueous stages of syn- [L-Ph]+, 370 (3.4%) [L-PPh2]+, 356 (4.0%) [L-CH2PPh2]+ 276
thesis. The precursors tosyl azide,55 bis(benzonitrile)palladium (100.0%) [PPh2-CH2Ph]+. Elemental analysis calcd (%) for
(II) chloride, cis-[PdCl2(PhCN)2],37 and 1-R-3-NH2-1,2-ortho-car- C28H35B10NP2: C 60.5, H 6.3, N 2.5; found: C 59.8, H 6.4, N 2.6.
borane, R = H, Me,24,25 were synthesized according to the lit-
Synthesis of L3
erature. o-Carborane precursors, o-1-R-C2B10H11 R = H, Me, Ph,
were supplied from KatChem Ltd (Prague) and used as
Starting materials. ortho-1-Ph-carborane (0.5 g, 2.26 mmol);
received. The precursors terc-butyl nitrite, trimethylsilyl azide, n-BuLi 1.6 M in hexanes (1.55 mL, 2.49 mmol); tosyl azide
2-phenylaniline, bis(diphenylphosphino)methane (dppm) and (0.39 mL, 2.49 mmol) and dppm (1 g, 2.6 mmol). Yield: 0.558 g
1
n-BuLi solution (1.6 M in hexanes) were purchased from (40%); white solid. H NMR (CDCl3), δ (ppm): 1.11–3.44 (bm,
Aldrich and used as received.
10H, BH), 2.89 (d, 2H, P-CH2-P), 7.14 (m, 4H, m-PPh2), 7.25 (m,
12H: 4H, o-PPh2; 4H, m-PPh2; 4H, p-PPh2), 7.33 (m, 2H, m-Ph),
Synthesis of C-carboranyl iminophosphoranes, L1–L3
7.42 (m, 4H, o-PPh2), 7.52 (m, 1H, p-Ph), 7.70 (m, 2H, o-Ph).
The compounds were synthesized in a similar way, following 31P{1H} NMR (CDCl3) δ (ppm): −30.5 (d, JPP = 56.6 Hz,
2
the literature procedure described for other C-carboranyl –CH2PPh2), 7.4 (d, 2JPP = 56.6 Hz, PvN) ppm. 11B NMR (CDCl3)
iminophosphoranes.12,13,18
δ (ppm): −14.8, −12.6, −11.5, −10.4, −6.5. IR (ATR, ν/cm−1):
3436s, 3058w, 2580s ν(B–H), 1703vw, 1435s, 1375vs ν(PvN),
1115m, 1074m, 1028w, 738s, 692vs. MS (EI, m/z): 617 (19.1%)
Synthesis of L1
To a solution of ortho-carborane (0.5 g, 3.47 mmol) in dry THF [L]+, 540 (13.0%) [L-Ph]+, 432 (3.0%) [L-PPh2]+, 418 (5.4%)
(40 mL) at −10 °C, was added dropwise a solution of n-BuLi [L-CH2PPh2]+, 341 (1.1%) [L-(CH2PPh2)-Ph]+, 276 (100.0%)
1.6 M in hexanes (2.28 mL, 3.65 mmol). The mixture was [PPh2-CH2Ph]+, 262 (15.3%) [PPh3]+. Elemental analysis calcd
stirred for 30 minutes and then tosyl azide (0.6 mL, (%) for C33H37B10NP2: C 64.2, H 6.0, N 2.3; found: C 63.6, H
3.68 mmol) was added. The reaction mixture was stirred at 5.8, N 2.4.
−10 °C for 30 minutes and then 30 minutes at room tempera-
ture. A solution of dppm (1.43 g, 3.72 mmol) in dry THF
Synthesis of B-carboranyl iminophosphoranes, L4 and L5
(10 mL) was added dropwise. The reaction mixture was heated
Synthesis of L4. To a solution of 3-NH2-1,2-ortho-carborane
to reflux for 1.5 hours (evolving nitrogen gas). The cold reac- (0.2 g, 1.25 mmol) in dry acetonitrile (40 mL) at 0 °C, was
tion mixture was evaporated to dryness to yield a brown oil. added dropwise terc-butyl nitrite (0.22 mL, 1.84 mmol) and tri-
The crude mixture was purified by silica column chromato- methylsilyl azide (0.24 mL, 1.84 mmol). The yellow solution
graphy (90 : 10, hexane : ethyl acetate). Yield: 0.798 g (43%); was stirred at room temperature for 1.5 hours and then a solu-
white solid. 1H NMR (CDCl3), δ (ppm): 0.90–2.81 (bm, 10H, tion of dppm (1.41 g, 3.67 mmol) in dry THF (10 mL) was
498 | Dalton Trans., 2019, 48, 486–503
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
added dropwise. The reaction mixture was heated to reflux for (67.1%) [L-Ph]+, 352 (60.3%) [L-(CH2PPh2)]+, 199 (40.4%)
2 hours (evolving nitrogen gas). The cold reaction mixture was [CH2PPh2]+. Elemental analysis calcd (%) for C24H19NPBr: C
evaporated to dryness to yield a brown oil which was purified 80.6, H 5.7, N 2.5; found: C 81.1, H 5.9, N 2.3.
by silica column chromatography (90 : 10, hexane : ethyl
acetate). Yield: 0.462 g (67%); white solid. 1H NMR (CDCl3)
δ (ppm): 1.20–2.76 (bm, 10H, BH), 3.03 (bs, 2H, Ccage-H), 3.17
Synthesis of the palladium complexes, Pd1–Pd6
(d, 2H, P-CH2-P), 7.30 (m, 6H, 2H, p-PPh2, 4H, m-PPh2), 7.38
Synthesis of Pd1. To a solution of the metal precursor bis
(m, 4H, o-PPh2), 7.43 (m, 4H, m-PPh2), 7.50 (m, 2H, p-PPh2), (benzonitrile)palladium(II) chloride, [PdCl2(PhCN)2] (106 mg,
7.71 (m, 4H, o-PPh2). 31P{1H} NMR (CDCl3) δ (ppm): −30.2 (d, 0.28 mmol) in dry acetonitrile (15 mL), was added dropwise a
2
2JPP = 50.8 Hz, –CH2-PPh3), 5.8 (d, JPP = 50.8 Hz, PvN). solution of ligand L1 (150 mg, 0.28 mmol) in dry dichloro-
11B NMR (CDCl3) δ (ppm): −23.3, −17.1, −15.3, −13.6, −7.5, methane (10 mL). The reaction mixture was stirred overnight
1.3. IR (ATR, ν/cm−1): 3071w, 3058w, 2586s, ν(B–H) 2560s ν(B– (16 hours) at room temperature. The mixture was concentrated
H), 1586vw, 1480vs, 1433vs ν(PvN), 1133w, 1108m, 1025w, to half of its volume under reduced pressure. Commercial
980w, 773m, 747m, 736m, 721m, 692s, 615w, 531m, 498m. MS hexane (5 mL) was added to the solution, precipitating a yellow
(EI, m/z): 540 (67.6%) [L]+, 462 (72.4%) [L-Ph]+, 356 (15.7%) solid. The solid was filtered, washed with diethyl ether and
[L-PPh2]+, 342 (52.0%) [L-CH2PPh2]+, 262 (67.1%) [PPh3]+, 185 dried under vacuum. Yield: 50 mg (26%), yellow solid. 1H
(21.0%) [PPh2]+. Elemental analysis calcd (%) for NMR (CDCl3) δ (ppm): 0.89–2.79 (bm, 10H, BH), 4.04 (m, 2H,
C27H33B10NP2: C 60.2, H 6.1, N 2.6; found: C 60.2, H 6.3, N 2.4. P-CH2-P), 5.15 (bs, 1H, Ccage-H), 7.50 (m, 20H, HAr) 31P{1H}
2
Synthesis of L5
NMR (CDCl3) δ (ppm): 8.2 (d, JPP 32.7 Hz, CH2-PPh2), 38.2 (d,
Starting materials. 1-Me-3-NH2-1,2-ortho-carborane (0.2 g, 2JPP 32.7 Hz, PvN). 11B NMR (CDCl3) δ (ppm): −13.0, −11.6,
1.15 mmol); tert-butyl nitrite (0.22 mL, 1.84 mmol); trimethyl- −10.7, −10.1, −9.3, −8.2. IR (ATR, ν/cm−1): 3434s, 3055w,
silyl azide (0.24 mL, 1.80 mmol); and dppm (0.75 g, 2958w, 2864w, 2581s ν(B–H), 1706vw, 1627w, 1462s, 1436vs,
1.95 mmol). Yield: 0.421 g (66%); white solid. 1H NMR (CDCl3) 1385m, 1238m ν (PvN), 1159m, 1112s, 1070m, 1021m, 875m,
δ (ppm): 1.12–2.50 (bm, 9H, BH), 1.90 (s, 3H, Ccage-CH3), 2.69 772w, 732m, 690m, 648vw, 600w, 530w, 503w, 482w. MS
(bs, 1H, Ccage-H), 3.22 (t, 1H, P-CH2-P), 3.32 (t, 1H, P-CH2-P), (MALDI, m/z): 682.1 [PdLCl–H], 541.7 [L]. Elemental analysis
7.30 (m, 4H, m-PPh2), 7.38 (m, 2H, p-PPh2), 7.44 (m, 8H, 4H calcd (%) for C27H33B10NP2Cl2Pd: C 45.0, H 4.6, N 1.9, found:
o-PPh2, 4H m-PPh2), 7.52 (m, 2H, p-PPh2), 7.78 (m, 4H, C 44.6, H 4.8, N 1.8.
2
o-PPh2). 31P{1H} NMR (CDCl3) δ (ppm): −29.9 (d, JPP = 52.7
Synthesis of Pd2. Starting materials: bis(benzonitrile)palla-
Hz, CH2-PPh3), 8.1 (d, 2JPP = 52.7 Hz, PvN). 11B NMR (CDCl3) δ dium(II) chloride (104 mg, 0.27 mmol) and ligand L2 (150 mg,
(ppm): −22.2, −17.4, −14.8, −13.8, −12.9, −10.6, −8.0, 2.0.IR 0.27 mmol). Yield: 53 mg (27%); yellow solid. 1H NMR (CDCl3)
(ATR, ν/cm−1): 3055w; 2571s ν(B–H); 2147m; 1480m; 1433s; δ (ppm): 0.78–2.90 (bm, 10H, BH), 1.27 (s, 3H, Ccage-CH3), 2.20
1403s; 1374s ν(PvN); 1113m; 1026m; 997m; 771m; 735vs; (s, 2H, P-CH2-P), 7.34 (m, 6H, 4H m-PPh2 + 2H p-PPh2), 7.52
691vs; 530m; 500s. MS (EI, m/z): 555 (21.6%) [L]+, 478 (20.6%) (m, 8H, 4H o-PPh2 + 4H m-PPh2), 7.65 (m, 6H, 2H p-PPh2 + 4H
[L-Ph]+, 370 (4.3%) [L-PPh2]+, 356 (9.1%) [L-CH2PPh2]+, 276 o-PPh2). 31P{1H} NMR (CDCl3) δ (ppm): 17.5 (d, JPP = 8.6 Hz,
3
(100.0%) [NvPPh3]+, 262 (42.9%) [PPh3]+, 185 (11.0%) [PPh2]+. PvN), 20.5 (d, JPP = 8.6 Hz, –CH2PPh2). 11B NMR (CDCl3) δ
2
Elemental analysis calcd (%) for para C28H35B10NP2: C 60.5, H (ppm): −17.1, −14.8, −12.1, −5.6, −3.9. IR (ATR, ν/cm−1):
6.3, N 2.5; found: C 59.3, H 6.2, N 2.6.
3056w, 2926m, 2854w, 2580s ν(B–H), 1437s, 1336m ν(PvN),
1104m, 1027w, 777w, 740w, 691m. MS (MALDI, m/z): 697.1
[PdLCl–H], 555.5 [L]+. MS (ESI, m/z): 1393.4 [Pd2L2Cl2]+, 607.2
Synthesis of non-carboranyl iminophosphorane L6
The non-carboranyl derivative L6 was obtained following the [PdLCl–H]+.
Elemental
analysis
calcd
(%)
for
same procedure used for the preparation of the B-carboranyl C28H35NP2B10Cl2Pd: C 45.9, H 4.8, N 1.9, found: C 45.0, H 4.7,
derivatives, starting from 2-phenylaninile.
Synthesis of L6. Starting materials: 2-phenylaniline (0.3 g,
N 2.1.
Synthesis of Pd3. Starting materials: bis(benzonitrile)palla-
1.77 mmol); tert-butyl nitrite (0.25 mL, 2.12 mmol); trimethyl- dium(II) chloride (62 mg, 0.16 mmol) and ligand L3 (100 mg,
silyl azide (0.28 mL, 2.12 mmol); and dppm (1.02 g, 0.16 mmol). Yield: 64 mg (31%); yellow solid. 1H NMR (CDCl3)
2.65 mmol). Purification: silica column chromatography δ (ppm): 1.07–3.27 (bm, 10H, BH), 3.68 (t, 2H, P-CH2-P), 7.20
(80 : 20, hexane : ethyl acetate) Yield: 0.52 g (53%); white solid. (m, 4H), 7.31 (m, 8H), 7.47 (m, 6H), 7.54 (m, 4H) 7.62 (m, 1H),
2
1H NMR (CDCl3), δ (ppm): 3.27 (bd, 2H, P-CH2-P), 6.45 (bd, 7.67 (m, 2H). 31P{1H} NMR (CDCl3) δ (ppm): 2.1 (d, JPP 13.9
1H, HAr), 6.75 (t, 1H, HAr), 6.86 (m, 1H, HAr), 7.16 (m, 4H, Hz, PvN), 20.6 (d, 2JPP 13.9 Hz, CH2-PPh2). 11B NMR (CDCl3) δ
m-PPh2), 7.22 (m, 5H, HAr), 7.28 (m, 3H, m-Ph), 7.34 (m, 4H, (ppm): −13.2, −11.6, −8.3. IR (ATR, ν/cm−1): 3057w, 2923w,
HAr), 7.41 (m, 5H, HAr), 7.71 (m, 3H, HAr), 7.79 (m, 2H, HAr). 31
P
2852w, 2586s ν(B–H), 2551s ν(B–H), 1587w, 1483w, 1434m,
2
{1H} NMR (CDCl3) δ (ppm): −29.6 (d, JPP = 52.0 Hz, 1396vs ν(PvN), 1189w, 1160m, 1103m, 1073m, 1000w, 796w,
2
–CH2PPh2), −1.2 (d, JPP = 52.0 Hz, PvN) ppm. IR (ATR, ν/ 771m, 741s, 715m, 687s, 540w, 502m, 481w. MS (MALDI, m/z):
cm−1): 3050w; 2959vw; 1585w; 1473s; 1429s ν(PvN); 1328m; 759.2 [PdLCl–H], 616.3 [L]+. MS (ESI, m/z): 1554.4 [Pd2L2Cl3]+.
1285m; 1117; 1061m; 1023m; 998w; 796w; 771m; 731s; 691vs; Elemental analysis calcd (%) for C66H74B20N2P4Cl4Pd2: C 49.8,
613w; 570m; 521m; 494s. MS (EI, m/z): 551 (100.0%) [L]+, 474 H 4.7, N 1.8, found: C 49.7, H 4.9, N 1.8.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 486–503 | 499

View Article Online
Paper
Dalton Transactions
Synthesis of Pd4. Starting materials: bis(benzonitrile)palla- crystal structures of all compounds were solved by direct
dium(II) chloride (71 mg, 0.18 mmol) and ligand L4 (100 mg, methods. Crystallographic programs used for structure solu-
0.18 mmol). Yield: 48 mg (37%); yellow solid. 1H NMR (CDCl3) tion and refinement were those of SHELXL-201457 installed on
δ (ppm): 1.08–2.58 (bm, 10H, BH), 3.71 (m, 2H, P-CH2-P), 5.16 a PC clone. Scattering factors were those provided with the
(bs, 2H, CcageH), 7.36 (m, 4H, m-PPh2), 7.50 (m, 6H, 4H o-PPh2, SHELX program system. Missing atoms were located in the
2H p-PPh2), 7.67 (m, 6H, 2H p-PPh2, 4H m-PPh2), 7.86 (m, 4H, difference Fourier map and included in subsequent refine-
2
o-PPh2). 31P{1H} NMR (CDCl3) δ (ppm): 11.4 (d, JPP 27.6 Hz, ment cycles. The structures were refined by full-matrix least-
CH2-PPh2), 40.9 (d, JPP 27.6 Hz, PvN). 11B NMR (CDCl3) squares refinement on F2, using anisotropic displacement
2
δ (ppm): −17.9, −15.3, −12.1, −5.5, −1.2. IR (ATR, ν/cm−1): parameters for all non-hydrogen atoms. Hydrogen atoms were
3038m, 2953w, 2900w, 2587s ν(B–H), 2569s ν(B–H), 1484w, placed geometrically and refined using a riding model, includ-
1436s, 1253vs ν(PvN), 1197w, 1118s, 1101s, 1023m, 989m, ing free rotation about C–C bonds for methyl groups, with C–
971m, 878w, 765m, 732vs, 686s, 593w, 525w, 502w, 481w. MS H distances of 0.95–0.99 Å and B–H distances of 1.12 Å. For all
(MALDI, m/z): 1401 [Pd2L2Cl3–H], 717 [PdLCl2], 683 [PdLCl–H]. compounds, hydrogen atoms were refined with Uiso con-
Elemental analysis calcd (%) for C27H33B10NP2Pd Cl2: C 45.1, strained at 1.2 (for non-methyl groups) and 1.5 (for methyl
H 4.6, N 1.9, found: C 45.8, H 4.8, N 1.8.
groups) times Ueq of the carrier C or B atom. The crystal struc-
Synthesis of Pd5. Starting materials: bis(benzonitrile)palla- ture of Pd1 contains two acetonitrile molecules per asymmetric
dium(II) chloride (69 mg, 0.18 mmol) and ligand L5 (100 mg, unit, one of them disordered over two positions. The disorder
0.18 mmol). Yield: 41 mg (31%); yellow solid. 1H NMR (CDCl3) was handled by introducing split positions in the refinement
δ (ppm): 1.07–3.27 (bm, 10H, BH), 2.11 (s, 3H, Ccage–CH3), 3.13 with the respective occupancies (80/20). This structure also
(t, 1H, P-CH2-P), 4.46 (m, 1H, P-CH2-P), 6.05 (bs, 1H, Ccage-H), contains a fraction of another molecule of acetonitrile (20%).
7.12 (m, 2H), 7.37 (m, 4H), 7.51 (m, 3H), 7.57 (m, 3H), 7.66 (m, This crystal structure contains 2 palladium complexes per
2
5H), 8.02 (m, 3H). 31P{1H} NMR (CDCl3) δ (ppm): 9.6 (d, JPP
=
asymmetric unit; one of them presents an 85/15 disordered
28.4 Hz, CH2PPh2), 40.8 (d, JPP = 28.4 Hz, PvN). 11B NMR phenyl ring. The crystal structures of Pd2 and Pd3 contain two
(CDCl3) δ (ppm): −16.6, −12.7, −9.9, −6.8, −4.1, −0.4. IR (ATR, and one dichloromethane molecules, respectively, per asym-
ν/cm−1): 2984m, 2904w, 2571m ν(B–H), 2544m ν(B–H), 1587w, metric unit. The crystal structure of Pd4 contains half mole-
1483w, 1437m, 1383w, 1364w, 1222s ν(PvN), 1121m, 1098m, cule of dichloromethane per asymmetric unit. Three phenyl
1027w, 998m, 955m, 877m, 784m, 747s, 728vs, 686vs, 596m, rings of the PPh2 moieties are disordered over two positions,
551m, 530m, 501s, 481s. MS (MALDI, m/z): 697 [PdLCl–H], 555 all of them with 50/50 occupancies. The crystal structure of
[L]. Elemental analysis calcd (%) for C28H35B10NP2PdCl2: C Pd6 contains two molecules of chloroform per asymmetric
2
45.9, H 4.8, N 1.9, found: C 46.6, H 5.0, N 1.7.
Synthesis of Pd6. Starting materials: bis(benzonitrile)palla-
unit.
Table S1 (ESI†) summarizes pertinent details of the data
dium(II) chloride (70 mg, 0.18 mmol) and ligand L6 (100 mg, collection and the structure refinement of the crystal struc-
0.18 mmol). Yield: 62 mg (47%); yellow solid. 1H NMR (CDCl3) tures. The program ORTEP358 was used to generate the pic-
δ (ppm): 2.79 (t, 1H, P-CH2-P), 3.00 (m, 1H, P-CH2-P), 6.98 (m, tures of all the molecular structures. CCDC reference
3H), 7.07 (m, 2H), 7.13 (m, 2H), 7.21 (m, 5H), 7.31 (m, 2H), numbers: 1855597–1855602.†
7.43 (m, 4H), 7.54 (m, 7H), 7.68 (d, 2H), 8.02 (m, 2H). 31P{1H}
2
Computational details
NMR (CDCl3) δ (ppm): 12.9 (d, JPP = 35.2 Hz, CH2PPh2), 41.7
2
(d, JPP = 35.2 Hz, PvN). IR (ATR, ν/cm−1): 3057w, 2966w, Full geometry optimizations of ligands L1–L5 were performed
2920w, 1588w, 1472w, 1346w, 1271m ν(PvN), 1233w, 1190vw, within the hybrid-meta GGA approximation to DFT employing
1162vw, 1131w, 1112m, 1053w, 1028w, 1007w, 999w, 917vw, the TPSSh functional (10% exchange)59 and the standard 6-
863vw, 816m, 784m, 766m, 746s, 732vs, 686s, 632w, 613w, 31+G(d,p) basis set. The Pd complexes were optimized using
535m, 505s, 494m, 473s. MS (MALDI, m/z): 1421 (7.8%) the same computational approach together with the relativistic
[Pd2L2Cl3–H]+, 694 (100.0%) [PdLCl–H]+. Elemental analysis effective core potential of Dolg (ECP28MDF),60 which includes
calcd (%) for C37H31NP2PdCl2: C 61.0, H 4.3, N 1.9, found: C 28 electrons in the core ([Ar]3d10). The valence space
60.8, H 4.2, N 2.0.
4s4p4d5s5p was described by the cc-pVDZ-PP basis set, which
presents a (25s22p13d1f)/[4s4p3d1f] contraction scheme.60
The ECP28MDF/cc-pVDZ-PP approximation was found to
X-ray crystallography
Intensity data sets for all compounds were collected with the provide good results for different Pd complexes.61 Due to the
use of a Bruker X8 Kappa APEXII diffractometer (Mo Kα radi- large computational effort required for the calculation of
ation, λ = 0.71073 Å) equipped with a graphite monochroma- second derivatives, the calculations performed on the dinuc-
tor. All crystals were measured at 100 K. The omega and lear complexes were carried out using the smaller 6-31G(d)
phi scans technique was employed to measure intensities in basis set for the ligand atoms. No symmetry constraints were
all crystals. No decomposition of the crystals occurred during imposed during the optimizations. All stationary points found
data collection. The intensities of all data sets were corrected on the potential energy surfaces as a result of geometry optimi-
for Lorentz and polarization effects. Absorption effects in all zations were characterized by frequency analysis to confirm
compounds were corrected using the program SADABS.56 The that they correspond to true energy minima. Wave function
500 | Dalton Trans., 2019, 48, 486–503
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
analysis (ρBCO, ∇ρ, H(r) and LBO) was carried out with the com-
8 (a) R. A. Wiesboeck and M. F. Hawthorne, J. Am. Chem.
Soc., 1964, 86, 1642–1643; (b) B. Wrackmeyer,
E. V. Klimkina, W. Milius, T. Bauer and R. Kempe, Chem. –
Eur. J., 2011, 17, 3238–3251; (c) M. A. Fox, W. R. Gill,
P. L. Herbertson, J. A. H. MacBride and K. Wade,
Polyhedron, 1996, 15, 565–571; (d) M. A. Fox and K. Wade,
J. Organomet. Chem., 1999, 573, 279–291; (e) T. D. Getman,
Inorg. Chem., 1998, 37, 3422–3423; (f) J. Yoo, J. W. Hwang
and Y. Do, Inorg. Chem., 2001, 40, 568–570;
(g) A. F. Armstrong and J. F. Valliant, Inorg. Chem., 2007,
46, 2148–2158; (h) H. Lee, T. Onak, J. Jaballas, U. Tran,
T. U. Truong and H. T. To, Inorg. Chim. Acta, 1999, 289, 11–
19.
puter program Multiwfn 3.2.62
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
This work was supported by Xunta de Galicia (Spain) (grant
no. 10PXIB209285PR).
9 (a) J. Li and M. A. Jones Jr., Inorg. Chem., 1990, 29, 4162–
4163; (b) C. Viñas, G. Barberá, J. M. Oliva, F. Teixidor,
A. J. Welch and G. M. Rosair, Inorg. Chem., 2001, 40, 6555–
6562; (c) H. Yamazaki, K. Ohta and Y. Endo, Tetrahedron
Lett., 2005, 46, 3119–3122; (d) S. Robertson, D. Ellis,
T. D. McGrath, G. M. Rosair and A. J. Welch, Polyhedron,
2003, 22, 1293–1301; (e) W. Chen, J. J. Rockwell,
C. B. Knobler, D. E. Harwell and M. F. Hawthorne,
Polyhedron, 1999, 18, 1725–1734; (f) A. V. Safronov,
N. I. Shlyakhtina and M. F. Hawthorne, Organometallics,
2012, 31, 2764–2769; (g) M. F. Hawthorne and
P. A. Wegner, J. Am. Chem. Soc., 1968, 90, 896–901.
References
1 R. N. Grimes, Carboranes, Academic Press, London, UK,
2nd edn, 2011.
2 (a) M. Y. Tsang, S. Rodríguez-Hermida, K. C. Stylianou,
F. Tan, D. Negi, F. Teixidor, C. Viñas, D. Choquesillo-
Lazarte, C. Verdugo-Escamilla, M. Guerrero, J. Sort,
J. Juanhuix, D. Maspoch and J. G. Planas, Cryst. Growth
Des., 2017, 17, 846–857 and references therein;
(b) R. Núñez, I. Romero, F. Teixidor and C. Viñas, Chem.
Soc. Rev., 2016, 45, 5147–5173; (c) M. Patel, A. C. Swain,
A. R. Skinner, L. G. Mallinson and G. F. Hayes, Macromol. 10 (a) A. R. Popescu, F. Teixidor and C. Viñas, Coord. Chem.
Symp., 2003, 202, 47–58; (d) E. N. Peters, Ind. Eng. Chem.
Prod. Res. Dev., 1984, 23, 28–32; (e) A. Gonzalez-Campo,
R. Núñez, F. Teixidor and B. Boury, Chem. Mater., 2006, 18,
4344–4353; (f) A. Ferrer-Ugalde, E. J. Juarez-Perez,
F. Teixidor, C. Viñas and R. Nunez, Chem. – Eur. J., 2013,
19, 17021–17030; (g) O. K. Farha, A. M. Spokoyny,
Rev., 2014, 269, 54–84; (b) A. M. Spokoyny, C. D. Lewis,
G. Teverovskiy and S. L. Buchwald, Organometallics, 2012,
31, 8478–8481 and references therein; (c) A. R. Popescu,
A. Laromaine, F. Teixidor, R. Sillanpää, R. Kivekäs,
J. I. Llambias and C. Viñas, Chem. – Eur. J., 2011, 17, 4429–
4443.
K. L. Mulfort, M. F. Hawthorne, C. A. Mirkin and 11 A. M. Spokoyny, C. W. Machan, D. J. Clingerman,
J. T. Hupp, J. Am. Chem. Soc., 2007, 129, 12680–12681.
3 (a) F. Issa, M. Kassiou and L. M. Rendina, Chem. Rev., 2011,
111, 5701–5722; (b) M. Scholz and E. Hey-Hawkins, Chem.
Rev., 2011, 111, 7035–7062.
4 J. G. Planas, F. Teixidor and C. Viñas, Crystals, 2016, 6, 50–
71.
M. S. Rosen, M. J. Wiester, R. D. Kennedy, C. L. Stern,
A. A. Sarjean and C. A. Mirkin, Nat. Chem., 2011, 3, 590–
596.
12 P. Crujeiras, J. L. Rodríguez-Rey and A. Sousa-Pedrares,
Dalton Trans., 2017, 46, 2572–2593.
13 P. Crujeiras, J. L. Rodríguez-Rey and A. Sousa-Pedrares,
Eur. J. Inorg. Chem., 2017, 4653–4667.
5 (a) W. Li, X. Yan, H. Zhang, R. He, M. Li and W. Shen,
Eur. J. Inorg. Chem., 2018, 99–108; (b) K. O. Kirlikovali, 14 (a) E. I. Matrosov, V. A. Gilyarov, V. Yu. Kovtun and
J. C. Axtell, A. Gonzalez, A. C. Phung, S. I. Khan and
A. M. Spokoyny, Chem. Sci., 2016, 7, 5132–5138; (c) X. Li,
H. Yan and Q. Zhao, Chem. – Eur. J., 2016, 22, 1888–1898;
(d) A. Ferrer-Ugalde, A. González-Campo, C. Viñas,
M. I. Kabachnik, Izv. Akad. Nauk Az. SSR, 1971, 6, 1162–
1168; (b) V. Yu. Kovtun, V. A. Gilyarov, B. A. Korolev,
E. I. Matrosov and M. I. Kabachnik, Zh. Obshch. Khim.,
1971, 41, 772–778.
J. Rodríguez-Romero, R. Santillan, N. Farfán, R. Sillanpää, 15 J. García-Álvarez, S. E. García-Garrido and V. Cadierno,
A. Sousa-Pedrares, R. Núñez and F. Teixidor, Chem. – Eur.
J., 2014, 20, 9940–9951.
J. Organomet. Chem., 2014, 751, 792–808, and references
therein.
6 (a) V. I. Bregadze, Chem. Rev., 1992, 92, 209–223; 16 (a) S. Ramírez-Rave, F. Estudiante-Negrete, R. A. Toscano,
(b) L. I. Zakharkin and V. A. Ol’shevskaya, Zh. Obshch.
Khim., 1987, 57, 368–372; (c) J. Plesek and S. Hermanek,
Collect. Czech. Chem. Commun., 1979, 44, 24–33;
(d) J. Plesek, S. Hermanek and B. Stibr, J. Less-Common
Met., 1979, 67, 225–228.
S. Hernández-Ortega, D. Morales-Morales and J. M. Grévy,
J. Organomet. Chem., 2014, 749, 287–295; (b) M. Alajarin,
C. Lopez-Leonardo, P. Llamas-Lorente, R. Raja, D. Bautista
and R. A. Orenes, Dalton Trans., 2012, 41, 12259–12269.
17 M. G. Davidson, M. A. Fox, T. G. Hibbert, J. A. K. Howard,
A. Mackinnon, I. S. Neretin and K. Wade, Chem. Commun.,
1999, 1649–1650.
7 K. Hermansson, M. Wójcik and S. Sjölberg, Inorg. Chem.,
1999, 38, 6039–6048.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 486–503 | 501

View Article Online
Paper
Dalton Transactions
18 R. D. Kennedy, Chem. Commun., 2010, 46, 4782–
R. McDonald and R. G. Cavell, Organometallics, 1996, 15,
1033–1041.
4784.
19 J. Estrada, C. A. Lugo, S. G. McArthur and V. Lavallo, Chem. 36 (a) Y. Sevryugina, R. L. Julius and M. F. Hawthorne, Inorg.
Commun., 2016, 52, 1824–1826.
Chem., 2010, 49, 10627–10634; (b) R. Cheng, Z. Qiu and
Z. Xie, Nat. Commun., 2017, 8, 14827–14833;
(c) R. M. Dziedzic, L. M. A. Saleh, J. C. Axtell, J. L. Martin,
S. L. Stevens, A. Timothy Royappa, A. L. Rheingold and
A. M. Spokoyny, J. Am. Chem. Soc., 2016, 138, 9081–9084.
37 G. K. Anderson and M. Lin, Inorg. Synth., 1990, 28, 60–63.
38 (a) E. W. Abel and S. A. Mucklejohn, Phosphorus Sulfur
Relat. Elem., 1981, 9, 235–266; (b) R. Bielsa, A. Larrea,
R. Navarro, T. Soler and E. P. Urriolabeitia, Eur. J. Inorg.
Chem., 2005, 1724–1736; (c) C. J. Wallis, I. L. Kraft,
J. N. Murphy, B. O. Patrick and P. Mhrkhodavandi,
Organometallics, 2009, 28, 3889–3895; (d) S. D. J. Brown,
W. Henderson, K. J. Kilpin and B. K. Nicholson, Inorg.
Chim. Acta, 2007, 360, 1310–1315.
20 (a) H. Staudinger and J. Meyer, Helv. Chim. Acta, 1919, 2,
619–635; (b) H. Staudinger and J. Meyer, Helv. Chim. Acta,
1919, 2, 635–646.
21 R. J. Blanch, L. C. Bush and M. Jones Jr., Inorg. Chem.,
1994, 33, 198–199.
22 D. Zhao and Z. Xie, Chem. Sci., 2016, 7, 5635–5639.
23 K. Barral, A. D. Moorhouse and J. E. Moses, Org. Lett.,
2007, 9, 1809–1811.
24 J. F. Valliant and P. Schaffer, J. Inorg. Biochem., 2001, 85,
43–51.
25 (a) L. I. Zakharkin, V. N. Kalinin and V. V. Gedymin,
J. Organomet. Chem., 1969, 16, 371–379; (b) R. A. Kasar,
G. M. Knudsen and S. B. Kahl, Inorg. Chem., 1999, 38,
2936–2940.
26 (a) A. S. Nuraeva, D. S. Vasileva, S. G. Vasilev,
P. S. Zelenovskiy, D. A. Gruzdev, V. P. Krasnov,
39 S. J. Coles, P. Faulds, M. B. Hursthouse, G. C. Ranger,
A. J. Toner and N. M. Walker, J. Organomet. Chem., 1999,
586, 234–240.
V. A. Olshevskaya, V. N. Kalinin and V. Ya. Shur, 40 (a) R. Meijboom, A. Muller and A. Roodt, Acta Crystallogr.,
Ferroelectrics, 2016, 496, 1–9; (b) G. L. Levit, V. P. Krasnov,
D. A. Gruzdev, A. M. Demin, I. V. Bazhov,
L. Sh. Sadretdinova, V. A. Olshevskaya, V. N. Kalinin,
C. S. Cheong, O. N. Chupakhin and V. N. Charushin,
Sect. E: Struct. Rep. Online, 2006, 62, m897–m899;
(b) V. A. Stepanova, L. M. Egan, L. Stahl and
I. P. Smoliakova, J. Organomet. Chem., 2011, 696, 3162–
3168.
Collect. Czech. Chem. Commun., 2007, 72, 1697–1706; 41 M. Alajarín, C. López-Leonardo and P. Llamas-Lorente,
(c) G. L. Levit, V. P. Krasnov, A. M. Demin, M. I. Kodess, Tetrahedron Lett., 2001, 42, 605–607.
L. Sh. Sadretdinova, T. V. Matveeva, V. A. Ol’shevskaya, 42 (a) M. Alajarín, C. López-Leonardo, P. Llamas-Lorente,
V. N. Kalinin, O. N. Chupakhin and V. N. Charushin, D. Bautista and P. G. Jones, Dalton Trans., 2003, 426–434.
Mendeleev Commun., 2004, 14, 293–295; (d) V. P. Krasnov, 43 L. A. Boyd, W. Clegg, R. C. B. Copley, M. G. Davidson,
A. M. Demin, G. L. Levit, A. N. Grishakov,
L. Sh. Sadretdinova, V. A. Olshevskaya, V. N. Kalinin and
M. A. Fox, T. G. Hibbert, J. A. K. Howard, A. McKinnon,
R. J. Peace and K. Wade, Dalton Trans., 2004, 2786–2799.
`
V. N. Charushin, Russ. Chem. Bull., 2008, 57, 2535–2539; 44 (a) K. V. Katti, R. J. Batchelor, F. W. B. Einstein and
(e) V. P. Krasnov, G. L. Levit, V. N. Charushin,
A. N. Grishakov, M. I. Kodess, V. N. Kalinin,
V. A. Ol’shevskaya and O. N. Chupakhin, Tetrahedron:
Asymmetry, 2002, 13, 1833–1835.
R. G. Cavell, Inorg. Chem., 1990, 29, 808–814;
(b) V. Cadierno, J. Diez, J. Garcia-Alvarez, J. Gimeno,
N. Nebra and J. Rubio-Garcia, Dalton Trans., 2006, 5593–
5604.
27 P. Morel, P. Schaffer and J. F. Valliant, J. Organomet. Chem., 45 (a) C.-Y. Liu, D.-Y. Chen, M.-C. Cheng, S.-M. Peng and
2003, 668, 25–30.
S.-T. Liu, Organometallics, 1995, 14, 1983–1991;
(b) A. Arques, P. Molina, D. Aunon, M. J. Vilaplana,
M. D. Velaso, F. Martinez, D. Bautista and F. J. Lahoz,
J. Organomet. Chem., 2000, 598, 329–338; (c) L. Boubekeur,
L. Ricard, N. Meezailles, M. Demange, A. Auffrant and P. Le
Floch, Organometallics, 2006, 25, 3091–3094.
28 (a) D. Zhao and Z. Xie, Angew. Chem., Int. Ed., 2016, 55,
3166–3170; (b) D. Zhao, J. Zhang and Z. Xie, Angew. Chem.,
Int. Ed., 2014, 53, 8488–8491.
29 T. P. Brauna, P. A. Gutsch and H. Zimmer, Z. Naturforsch.,
1999, 54b, 858–862.
30 A. Buchard, B. Komly, A. Auffrant, X. F. Le Goff and P. A. Le 46 (a) J. Vicente, A. Arcas, D. Bautista and M. C. R. de
Floch, Organometallics, 2008, 27, 4380–4385.
31 L. Boubekeur, L. Ricard, N. Mezailles and P. Le Floch,
Organometallics, 2005, 24, 1065–1074.
Arellano,
Organometallics,
1998,
17,
4544–4550;
(b) T. Cheisson and A. Auffrant, Dalton Trans., 2016, 45,
2069–2078.
32 P. Molina, A. Arques, A. Garcia and M. C. Ramirez de 47 D. Bézier, O. Daugulis and M. Brookhart, Organometallics,
Arellano, Tetrahedron Lett., 1997, 38, 7613–7616. 2017, 36, 2947–2951.
33 P. Molina, A. Arques, A. Garcia and M. C. Ramirez de 48 M. G. Davidson, T. G. Hibbert, J. A. K. Howard,
Arellano, Eur. J. Inorg. Chem., 1998, 1359–1368.
34 R. S. Pandurangi, K. V. Katti, L. Stillwell and C. L. Barnes,
J. Am. Chem. Soc., 1998, 120, 11364–11373.
35 (a) K. V. Katti, B. D. Santarsiero, A. A. Pinkerton and
R. G. Cavell, Inorg. Chem., 1993, 32, 5919–5925; (b) J. Li,
A. Mackinnon and K. Wade, Chem. Commun., 1996, 19,
2285–2286.
49 (a) S. C. To and F. Y. Kwong, Chem. Commun., 2011, 47,
5079–5081; (b) J. O. Yu, E. Lam, J. L. Sereda,
N. C. Rampersad, A. J. Lough, C. S. Browning and
502 | Dalton Trans., 2019, 48, 486–503
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
D. H. Farrar, Organometallics, 2005, 24, 37–47; 55 F. Nanteuil and J. Waser, Angew. Chem., Int. Ed., 2011, 50,
(c) G. R. Newkome, D. W. Evans and F. R. Fronczek, Inorg. 12075–12079.
Chem., 1987, 26, 3500–3506; (d) N. M. Vinogradova, 56 G. M. Sheldrick, SADABS: Program for absorption correction using
I. L. Odinets, K. A. Lyssenko, M. P. Pasechnik, area detector data, University of Göttingen, Germany, 1996.
P. V. Petrovskii and T. A. Mastryukova, Mendeleev Commun., 57 G. M. Sheldrick, Acta Crystallogr., Sect. A: Fundam.
2001, 11, 219–220; (e) J. Autio, S. Vuoti, M. Haukka and
J. Pursiainen, Inorg. Chim. Acta, 2008, 361, 1372–1380.
Crystallogr., 2008, 64, 112–122.
58 L. Farrugia, J. Appl. Crystallogr., 1997, 30, 565.
50 F. Teixidor, G. Barbera, A. Vaca, R. Kivekäs, R. Sillampää, 59 J. M. Tao, J. P. Perdew, V. N. Staroverov and G. E. Scuseria,
J. Oliva and C. Viñas, J. Am. Chem. Soc., 2005, 127, 10158–
10159.
51 R. F. W. Bader, Atoms in Molecules: A Quantum Theory,
Oxford, University Press, Oxford, 1990.
52 T. Lu and F. Chen, J. Phys. Chem. A, 2013, 117, 3100–3108.
53 S. Boonseng, G. W. Roffe, J. Spencer and H. Cox, Dalton
Trans., 2015, 44, 7570–7577.
54 P. K. Sajith and C. H. Suresh, Inorg. Chem., 2011, 50, 8085–
8093.
Phys. Rev. Lett., 2003, 91, 146401–146404.
60 K. A. Peterson, D. Figgen, M. Dolg and H. J. Stoll, Chem.
Phys., 2007, 126, 124101–124112.
61 (a) M. Orbach, S. Shankar, O. V. Zenkina, P. Milko,
Y. Diskin-Posner and E. van der Boom, Organometallics,
2015, 34, 1098–1106; (b) J. Turek, I. Panov, M. Semler,
P. Stepnicka, F. De Proft, Z. Padelkova and A. Ruzicka,
Organometallics, 2014, 33, 3108–3118.
62 T. Lu and F. Chen, J. Comput. Chem., 2012, 33, 580–592.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 486–503 | 503