Regioselective ortho Amination of Coordinated 2-(Arylazo)pyridine. Isolation of Monoradical Palladium Complexes of a New Series of Azo-Aromatic Pincer Ligands

Article abstract of DOI:10.1021/acs.inorgchem.5b02110

In an unusual reaction of [Pd(L¹)Cl₂] (L¹ = 2-(arylazo)pyridine) with amines, a new series of palladium complexes [Pd(L^2?-)Cl] (L² = 2-((2-amino)arylazo)pyridine) (1a-1h) were isolated. The complexes

Full text of DOI:10.1021/acs.inorgchem.5b02110

Article
pubs.acs.org/IC
Regioselective ortho Amination of Coordinated 2‑(Arylazo)pyridine.
Isolation of Monoradical Palladium Complexes of a New Series of
Azo-Aromatic Pincer Ligands
Debabrata Sengupta,† Nabanita Saha Chowdhury,† Subhas Samanta,†,◊,# Pradip Ghosh,†
Saikat Kumar Seth,§ Serhiy Demeshko,# Franc Meyer,# and Sreebrata Goswami*,†
†Department of Inorganic Chemistry, Indian Association for the Cultivation of Science, Jadavpur, Kolkata 700 032, India
◊Department of Chemical Sciences, Indian Institute of Science Education and Research Kolkata, Mohanpur 741246, Nadia, India
#Institut fur Anorganische Chemie, Georg-August-Universitat Gottingen, Tammannstraße 4, 37077 Gottingen, Germany
̈
̈
̈
̈
§Department of Physics, Mugberia G. Mahavidyalaya, Bhupatinagar, Purba Medinipur 721425, India
S
* Supporting Information
ABSTRACT: In an unusual reaction of [Pd(L1)Cl2] (L1 = 2-(arylazo)-
pyridine) with amines, a new series of palladium complexes [Pd(L2•−)Cl] (L2
= 2-((2-amino)arylazo)pyridine) (1a−1h) were isolated. The complexes were
formed via N−H and N−C bond cleavage reactions of 1°/2° and 3° amines,
respectively, followed by regioselective aromatic ortho-C−N bond formation
reaction and are associated with ortho-C−H/ortho-C−Cl bond activation. A
large variety of amines including both aromatic and aliphatic were found to be
effective in producing air-stable complexes. Identity of the resultant complexes
was confirmed by their X-ray structure determination. Efforts were also made
to understand the mechanism of the reaction. A series of experiments were
performed, which point toward initial ligand reduction followed by intraligand
electron transfer. Examination of the structural parameters of these complexes
(1) indicates that the in situ generated ligand coordinated to the PdII center
serves as the backbone of these air-stable monoradical complexes. Molecular and electronic structures of the isolated complexes
were further scrutinized by various spectroscopic techniques including cyclic voltammetry, variable temperature magnetic
susceptibility measurements, electron paramagnetic resonance, and UV−vis spectroscopy. Finally the electronic structure was
confirmed by density functional theory calculations. The isolated monoradical complexes adopt an unusual π-stacked array, which
leads to a relatively strong antiferromagnetic interaction (J = −40 cm−1 for the representative complex 1c).
(IL-SET) between two nonconjugative counterparts of a
coordinated azoaromatic ligand (L1) is operative behind a
regioselective ortho-C−N bond formation reaction. Ligand-
radical mediated SET reactivity has been recently identified and
is a rarely observed phenomenon in transition metal complex
mediated organic transformations.6c,7,8
While working on redox-active azo-aromatic ligands,
previously we observed9 that primary aromatic amines react
with metal coordinated 2-(arylazo)pyridine ligands freely to
give complexes of monoanionic tridentate NNN donor ligands
(Scheme 1). However, these reactions were limited only to
primary aromatic amines, and our attempts to achieve similar
reactivity with 2° and 3° aromatic as well as aliphatic amines
were unsuccessful.
INTRODUCTION
■
Events associated with transition metal complexes containing
redox active ligands have attracted considerable interest in
recent times.1 Research in this area is encouraged by, inter alia,
various metalloenzymatic transformations where ligand-radical
containing intermediates were found to be active species that
enable unusual reactivity.2 Moreover such complexes can also
show interesting magnetic properties3 as well as memory device
applications.4 Very recently, it has been argued that transition
metal coordinated redox active ligands can also act as source or
sink of electrons in metal complex mediated/catalyzed
transformations.5 For example, pioneer works of Wieghardt
and Chirik on redox active ligand-containing complexes
catalyzed reactions are prominent examples where ligand
redox is used in bond-breaking and bond-making processes.1a,6
Recently van der Vlugt et al. reported7 an example of a
palladium complex where ligand-to-substrate single electron
transfer (SET) is involved in chemical reactions. Herein we
report the reactivity of a series of redox-active azo-aromatic
palladium complexes where intraligand single electron transfer
Here, in an attempt to isolate a homoleptic palladium
complex, we observe an unusual aromatic C−N bond
formation reaction with a large variety of both aromatic and
Received: September 12, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.5b02110
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
trialkylamine. A similar reaction with primary or secondary
amines, however, failed even though it is known that activation
of N−H bonds is easier than N−C bond activation in alkaline
conditions. These results led us to examine the redox feature of
the [Pd(L1)Cl2] complex vis-a-vis the chemical transformation.
It is noted in the literature11 that the starting palladium
complex [Pd(L1)Cl2] undergoes a single electron ligand
reduction at a low potential (ca. −0.25 V; see Supporting
Information Figure S1 and Table S1). Accordingly, we planned
to examine the chemical reaction between coulometrically
generated one-electron reduced complex [Pd(L1a)Cl2]− with 1°
as well as 2° amines. The strategy worked as anticipated with
aniline and N-methylaniline and resulted in the isolation of the
nonradical complex 1i in case of aniline and the complex 1c in
the case of N-methylaniline, in 76% and 72% yields, respectively
(Table 1). Thus, it is understood that single-electron reduction
of the starting complex [Pd(L1)Cl2] is a prerequisite for the
aforenoted reactions. Trialkyl amines being stronger reductants
than both 1° and 2° amines, our initial attempt was successful
only in the former case, but no reactions were observed in the
later two cases. The above results allowed us to design a general
chemical protocol of the above fusion reactions by the use of an
equimolar quantity of cobaltocene as an added reducing agent.
As anticipated, the reactions proceeded cleanly in the presence
of cobaltocene, producing the desired palladium complexes
with greater ease and in higher yields (70−80%; Scheme 2).
Large varieties of aromatic and aliphatic amines were
attempted, and the isolated complexes are collected in Table
1. Their characterizations were made by elemental analysis and
usual spectroscopic studies. All characterization data are
provided in the Experimental Section (vide infra).
Scheme 1
aliphatic 1ο, 2ο and 3ο amines via N−H/N-C bond activation
(Scheme 2, see later). Detailed investigations reveal that initial
π*-azo reduction associated with π*-azo to π*-aryl ring
electron transfer initiated the C−N bond formation reaction.
The reaction not only leads to the formation of a new series of
tridentate azoaromatic ligand supported stable and uncoupled
palladium monoradical complexes, but also generates the
possibility of isolation of new azoaromatic pincer ligands with
fine-tuning of the coordination environment for further
exploration in redox chemistry. In this context, we may have
to mention that stable, uncoupled radical complexes are rare in
general10 because in most cases the radical character of the
coordinated ligand is obscured through the coupling of the
radical spin with unpaired electrons on metal or on another
radical ligand.
The isolated radical complexes (1a−h) displayed nearly
isotropic ligand-based electron paramagnetic resonance (EPR)
signals at g ≈ 1.99 (vide infra; Figure S2). One-electron
paramagnetism of these complexes was further verified by
solution-state magnetic susceptibility measurements by Evans
method of the representative samples (1c, 1e, 1f, and 1g) at
room temperature (300 K). Their magnetic moments (μeff) lie
between 1.70 and 1.73 μB. The complex 1i is a nonradical
complex diamagnetic and is identical in all respects to the
previously reported one, which was synthesized from the
corresponding preformed ligand.12
RESULTS AND DISCUSSIONS
■
Chemical Reactions. The work began with the typical
chemical reaction of [Pd(L1)Cl2] with trialkylamines (NR3, R =
Me, Et) in neat condition and at 50 °C. The reaction
unexpectedly produced an air-stable square planar PdII-
monoradical complex [Pd(L2a,b•−)Cl] (1a, 1b) (L2 = 2-((2-
amino)arylazo)pyridine) in ca. 15% yield. In the resultant
complex PdII is coordinated to a newly formed NNN donor,
which serves as a backbone of the π-radical (vide infra).
Formation of the new ligand is a result of ortho-Carom-amination
of a coordinated azo-ligand L1 via simultaneous ortho-C−H
bond activation and C−N bond fusion of 3° alkyl amines and is
also associated with aliphatic C−N bond cleavage of the
Crystal Structures Analysis. Final authentication of the
products came from X-ray crystallographic structure determi-
nation of the four representative complexes, specifically, 1c, 1e,
Scheme 2
B
DOI: 10.1021/acs.inorgchem.5b02110
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 1. Products of the Regiospecific ortho-C−N Coupling Reaction on Coordinated Azo-Aromatic Ligand with Amines
NR1NR2X
ligand(L1)
R1
R2
X
AX/BX
ligand(L2)
compound
A = H
B = H
C = H
A = H
B = H
C = Cl
L1a
Me
Et
Me
Et
H, Me
H, Et
H
H2/CH4
H2/C2H6
H2
L2a
L2b
L2c
L2d
L2e
L2f
1a
1b
1c
1d
1e
1f
Me
Me
Et
Ph
Me
Et
L1b
H, Me
H, Et
H
H2/CH4
H2/C2H6
H2
Me
Ph
Me
Ph
Ph
Me
H, Ph
H, Me
H2/C6H6
H2/CH4
L2g
L2h
1g
1h
A = Cl
B = H
L1c
L1d
L1e
L1f
L1a
C = H
A = Cl
B = Cl
C = H
A = H
B = H
Me
Me
Me
H
Me
Me
Me
Ph
H, Me
H, Me
H, Me
H
HCl/CH3Cl
L2h
1h
a
NR
C = CH3
A = CH3
B = CH3
C = H
A = H
B = H
a
NR
H2
L3a
1i
C = H
a
NR: No reaction at 50°C; 1i is a nonradical complex, see later.
Figure 1. View of the molecular structures of complexes (A) 1c, (B) 1e, (C) 1f, and (D) 1h (Hydrogen atoms are omitted for clarity).
1f, and 1h. Their molecular structures indeed support the
chemical reaction depicted in Scheme 2. Views of these
complexes are shown in Figure 1. Structural features of all these
are more or less similar, and we will use the structure of 1c
(Table 2) for our discussion, while details of the other three are
submitted as Supporting Information materials (Figures S3−S6
and Tables S5−S6). Here, two major issues emerged: first, the
structural analysis confirms the formation of the compounds
from the reference chemical reaction; second, there is a
considerable elongation of the N2−N3 bond (dN2−N3
=
1.329(3) Å), which indicates azo-anion radical character of
the coordinated ligand and is consistent with the magnetic
C
DOI: 10.1021/acs.inorgchem.5b02110
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
C(6) and N(3) atoms have the largest deviations in opposite
directions [C(6): +0.025(2) Å, N(3): −0.023(2) Å] from the
least-squares mean plane through the atoms N(1)−C(5)−
N(2)−N(3)−C(6)−C(11)−N(4)−Pd(1). In the solid state,
the structure is stabilized through intermolecular C−H···Cl, C−
H···π, and π−π stacking interactions (Tables S7 and S8). The
methyl carbon atom C(18) in the molecule at (x, y, z) acts as
donor to the Cl(1) in the molecule at (1/2 − x, −1/2 + y, 1/2
− z). The aryl ring carbon atom C(14) is in contact with the
ring centroid of another aryl ring C(6)−C(11) through C−
H···π stacking interactions (Figure S7), thus generating a
dimeric unit through C−H···π interactions.14 Because of their
self-complementary nature, the molecules are juxtaposed via
face-to-face π-stacking interactions.15 The pyridine rings N(1)/
C(1)−C(5) of the molecules at (x, y, z) and (1 − x, 1 − y, 1 −
z) are strictly parallel, with an interplanar spacing of 4.018(2) Å
and a ring centroid separation of 4.128(2) Å, corresponding to
a ring offset of 0.948 Å. The mutual influence of the C−H···π
bonds and face-to-face π-stacking interactions between the
pyridine rings led the molecules to build a two-dimensional
supramolecular assembly in the (1 0 1) plane (Figure S8). In
another substructure, the molecular packing is such that the
π−π stacking interactions between the pyridine rings of
adjacent partner molecules are optimized. These pyridine
rings are also in contact with the centroids of the aryl ring
C(6)−C(11) with an intercentroid distance of 3.840(2) Å and
interplanar spacing of 3.306(2) Å. Thus, the mutual influence
of the two types of π−π stacking interactions leads the
molecules to propagate into a supramolecular layered assembly
in the (1 1 0) plane (Figure 2).
These intermolecular interactions in solid state prompted us
to investigate the variable-temperature magnetic behavior, again
using 1c as a representative example. Magnetic measurements
were performed using polycrystalline material of 1c in a
SQUID magnetometer in the temperature range of 2−295 K
(Figure 3). The χMT value per dimeric unit at room
temperature is ∼0.64 cm3 mol−1 K, corresponding to an
effective magnetic moment of 2.26 μB. This value is lower than
the expected spin-only value of 0.75 cm3 mol−1 K (or 2.45 μB)
for two uncoupled ions with S = 1/2. On lowering the
temperature to 2 K, χMT tends to zero, indicating a singlet
ground state (S = 0). The experimental data were fitted using
the appropriate Heisenberg−Dirac−van Vleck spin Hamilto-
nian that includes isotropic exchange coupling and Zeeman
splitting, (eq 1).
Table 2. Selected Experimental and Calculated Bond
Distances (Å) for the Complex 1c
calculated
bond
parameters
experimental
1c
[1c]+
(S = 0)
[1c]−
(S = 0)
(1c)
(S = 1/2)
Pd1−Cl1
Pd1−N1
Pd1−N3
Pd1−N4
N2−N3
N3−C6
2.3219(9)
2.004(2)
1.9197(18)
2.100(2)
1.329(3)
1.378(3)
1.481(3)
2.362
2.033
1.956
2.165
1.324
1.381
1.489
2.308
2.054
1.960
2.134
1.269
1.405
1.487
2.430
2.018
1.958
2.211
1.379
1.352
1.488
C11−N4
properties as well as EPR spectra of the complexes (see below).
The effects of dN−N bond elongation of the coordinated ligands
in the present complexes are also reflected by the lowering of
vibrational frequencies νNN as compared to uncoordinated
L3a. The νNN band in free L3a appears at ∼1380 cm−1, whereas
those in the present complexes (1c) are considerably lower
appearing near 1303 cm−1.
It is worth comparing the structure of 1c with the previously
reported palladium complex 1i.10 The protonated ligand HL3
was obtained by a cobalt mediated ortho-Carom-amine bond
fusion reaction between coordinated L1 and ArNH2. The
complex is nonradical with S = 0 ground state and its N−N
bond is considerably shorter than that in the present radical
complexes. Relevant bond lengths of 1c and 1i are compared in
Scheme 3. It reveals that dN−N in the former is considerably
Scheme 3. Comparison of Bond Lengths of the Two Related
a
Complexes 1c and 1i
a
1c: neutral NNN donor tridentate ligand complex, 1i: monoanionic
NNN donor tridentate ligand complex.
̂
̂
̂
⃗
⃗
⃗
H = −2JS1S2 + gμBB(S1 + S2)
elongated compared to the later. Moreover C−N lengths on
either side of the central o-phenylene ring in these two
complexes are also quite different. Bond length contraction of
these two bonds in the complexes of the anionic ligand has
been attributed to delocalization of negative charge along the
ligand backbone. However, such delocalization is expected to
be absent9a in the azo-anion radical complexes. Thus, the above
two examples of PdII complexes have provided a direct
experimental evidence that neutral azoaromatic donors13 are
far more promising candidates for sequential redox events than
the corresponding anionic ligands.
Intermolecular Interaction and Magnetic Properties.
Detailed examination of the crystal structure of complex 1c
shows that the metal center is displaced by 0.028 Å out of the
least-squares planes defined by the donor atoms. The ligand
forms two five-membered chelate rings, and the dihedral angle
between the two related mean ring planes is 0.92°. In 1c, the
(eq 1)
The best fit parameters are g = 1.99 and J = −40.2 cm−1.
Better agreement with experimental data was obtained when
additional intermolecular interactions were considered in a
mean field approach by using a Weiss temperature Θ. This
parameter Θ = −15.2 K relates to the intermolecular
interactions of zJinter = −42.3 cm−1, where J is the interaction
parameter between two nearest neighbor magnetic centers and
z is the number of nearest neighbors. On the basis of the
structural information, three pathways for the exchange
interactions can be considered. Two of them may result from
the two different kinds of π−π stacking interactions with
corresponding distances of 3.840 and 4.128 Å (Figure 2); the
third pathway may arise through the C−H···π hydrogen bond
interactions (Figure S7). According to the spin density
distribution from density functional theory (DFT) calculations
D
DOI: 10.1021/acs.inorgchem.5b02110
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 2. Supramolecular layer architecture in the complex 1c assembled through cooperative π−π stacking interactions.
oxidation near 0.25 V and another one-electron reversible
reduction near −0.7 V. Cyclic voltammetry data of all
complexes are collected in the Supporting Information (Table
S2 and Figures S9−S15), and that of a representative complex
1c is shown in Figure 4. The corresponding nonradical
Figure 3. χMT vs T plot for complex 1c, calculated for a dimeric motif.
The solid line represents the best fit curve.
(see below), the unpaired electron is delocalized over the
backbone of the ligand, that is, into the pyridine and aryl rings
that form the five-membered chelate rings, whereas no spin
density is present on the phenyl rings involved in C−H···π
hydrogen bond interactions. Therefore, the coupling J = −40.2
cm−1 can be attributed to the shorter π−π stacking interactions
Figure 4. Cyclic voltammogram of complex 1c (in CH2Cl2/0.1 M
Bu4NClO4, scan rate 50 mV/s; potentials vs SCE). (inset) EPR
spectrum (X-band) of the complex 1c at 120 K in dichloromethane−
toluene glass.
of 3.840 Å, while weaker intermolecular interactions zJinter
=
−42.3 cm−1 (or Jinter = −21.2 cm−1 for the case z = 2) should
arise from π−π stacking interactions with longer distances of
4.128 Å. In general, the strength of the magnetic exchange
interaction in the π−π stacking systems depends on the spin
density on the interacting atoms as well as on the contact
distances and can vary from fractions of one wavenumber up to
extremely strong (larger than −1000 cm−1) coupling.16 The
observed coupling constant of −40.2 cm−1 in 1c with the
vertical distances from ring centroids to rings of 3.31 to 3.39 Å
(Table S8) is comparable with literature data for systems with
similar metric parameters. For example, coupling constants of
−58 cm−1 or −52 cm−1 were observed for the verdazyl radical
with interplanar spacing of 3.37 and 3.44 Å, or for phenylene-
based radicals with the closest C···C distances between
phenalenyl units of 3.54 to 3.66 Å, respectively.17
complex, [Pd(L3a)Cl] (1i), showed12 one reversible azo-based
reduction at −0.2 V and an ill-defined oxidative wave at higher
anodic potentials. The nature of the redox processes in the
representative complex 1c was studied by DFT calculations.
To have a closer look into the electronic structure of the
complexes, we performed a series of calculations using DFT
(B3LYP; see Supporting Information for details). The
computed metrical parameters of the complex 1c are in
reasonable agreement with the experimentally observed
parameters. A spin density plot for 1c, shown in Figure 5,
indicates almost one-electron spin delocalization over the
ligand backbone (96% on L2 and 4% on Pd) signifying that the
unpaired spin resides almost exclusively on the coordinated
ligand, which is well-consistent with the aforenoted exper-
Electrochemistry and Density Functional Theory. The
Pd complexes (1a−h) showed a reversible one-electron
E
DOI: 10.1021/acs.inorgchem.5b02110
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Thus, this absorption is assigned to an intraligand transition.
The computed one-electron oxidized complex showed a
HOMO to LUMO transition at 596 nm that appeared at 580
nm in the experimentally observed spectrum. The HOMO is
localized over the N-phenyl ring, whereas the LUMO is
concentrated on the NN π* array. The broad transition at
506 nm for the one-electron reduced species corresponds well
to the computed HOMO to LUMO+1 transition at 459 nm.
The experimentally observed spectra are shown in Figure 6, and
associated orbitals are tabulated in Table S3.
Figure 5. Spin-density plot for 1c derived from DFT calculations.
imental findings. Comparison of calculated structural parame-
ters between 1c and its corresponding one-electron oxidized
complex reveals a contraction of the azo bond length (dN2−N3
≈
1.269 Å in [1c]+). Our attempts to isolate the oxidized complex
[1c]+ were unsuccessful. However, the complex [1f]+ shows an
S = 0 ground state, as evidenced by its 1H NMR spectrum with
resonances appearing in the normal range of diamagnetic
compounds (see Supporting Information Figure S16). The
complex 1f is identical in all respects to 1c with the only
difference that the ligand here is chloro-substituted. It is
therefore concluded that the oxidation of 1c is also ligand-
centered. The reduction of 1c is identified to impart a single-
bond character to the N(2)−N(3) bond (ca. 1.379 Å). A
comparison of calculated bond lengths in the Pd complexes is
tabulated in Table 2. Notably, the lowest unoccupied molecular
orbital (LUMO) in 1c+ is localized over the π* azo orbital of
the azo-function, while the highest occupied molecular orbital
(HOMO) in the reduced complex, 1c‑ is distributed over the
ligand containing azo-function, which indicates that the redox
processes occurred exclusively at the ligand (Figure S22).
With all these results taken together, it is concluded that
redox events in the Pd complexes occur at the ligand without
affecting the bivalent oxidation state of metal. Thus, the redox
processes in these complexes involve the following three redox
states (Scheme 4).
Figure 6. Electronic spectra of 1c (violet) and coloumetrically
generated [1c]+ (dark yellow) and [1c]− (red) in dichloromethane
solution.
Mechanistic Investigations. To gain insight into the
aromatic ring amination reaction the following experiments
were planned. First we checked the reactivity of the complex
with different quantities of cobaltocene, CoCp2, with respect to
the starting palladium complex. It is found that the reaction
occurs even in the presence of 0.25 equiv of CoCp2. However,
maximum yields of the products were obtained when equimolar
quantities of [Pd(L1)Cl2] and CoCp2 were used. We thus
propose that single-electron reduction of the palladium
complex is needed for this reaction to occur. Moreover, it
was noted before that with less reducing amines such as 1° and
2° amines the reactions do not occur at all without an external
reductant. It is thus believed that formation of the one-electron
reduced ligand radical complex11a (I) is the initial step for the
reaction as shown in Scheme 5.
Scheme 4. Electron-Transfer Series for Complexes 1a−h
The formation of the one-electron reduced complex was
further supported by EPR spectroscopy and mass spectrometry
(MS). For example, the isolated one-electron reduced species
[Pd(L1b)Cl2]− gives a single-line EPR signal at g = 1.99
characteristic for a ligand-centered radical, and a molecular ion
peak at m/z = 392.72 amu in negative ion mode electrospray
ionization (ESI) MS, which matches well with the formulation
of the intermediate I (Supporting Information Figures S18 and
S19). Our attempts to characterize the reduced complex by IR
spectroscopy shows that νNN shifts from 1413 cm−1 [in
Pd(L1a)Cl2] to 1340 cm−1 in the reduced complex, indicating
primarily azo-reduction.
The idea that the reaction goes through a radical
intermediate was further corroborated by the fact that it is
completely ceased in the presence of a radical scavenger like
2,2,6,6-tetramethylpiperidinoxyl radical or 2,6-di-tert-butyl-4-
methylphenol. Notably, it is known that upon ligand reduction
of a transition metal complex bearing redox noninnocent
The radical complexes are intensely blue-violet colored and
showed multiple transitions in the optical spectral range, from
250 to 800 nm (Supporting Information Figure S17). For
example, the complex 1c has two broad bands at 575 and 348
nm. Upon oxidation the broad band at 575 nm slightly shifted
to 580 nm. However, upon reduction this band shifts to a
shorter wavelength 506 nm. To assign the electronic transition,
TD-DFT calculations were performed on three states of the
complex, specifically, 1c, [1c]+, and [1c]‑. Good agreement
between computed absorption spectra and the experimentally
observed ones was noted. For example, in the case of 1c the
computed absorption at 511 nm (HOMO(β) → LUMO(β)) is
in reasonable agreement with the experimentally observed band
at 575 nm, where both HOMO(β) and LUMO(β) orbitals are
localized on the ligand.
F
DOI: 10.1021/acs.inorgchem.5b02110
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 5. Plausible Reaction Mechanism
ligands, the lability of the ligand in the reduced complex
increases substantially.18 Thus, we anticipate that one of the
chloride ligands becomes labile in [Pd(L1)Cl2]− and is
subsequently replaced by the incoming neutral amine to
form19 the intermediate radical complex II as shown in Scheme
5. We have not yet been successful to isolate such an
intermediate complex with amines; however, a similar reaction
in the presence of triphenylphosphine (PPh3) leads to the
isolation of a new complex, Pd−I, confirming the lability of one
of the Cl− ligands. (Scheme 6 and Figure 7).
Scheme 6
Figure 7. ORTEP representation of the isolated palladium complex
(Pd−I) in the presence of PPh3 at 50% probability ellipsoid
(Hydrogen atoms are omitted for clarity).
the incoming amine substrate generating an intermediate
radical complex associated with C−N bond formation, (iii)
intraligand electron transfer from the π*-azo to the π*-phenyl
ring in the ligand associated with C−N bond formation
reaction.
Notably, a similar reaction of PPh3 with the parent complex
[Pd(L1)Cl2] does not show any substitution reaction. Now the
intermediate ligand radical complex is assumed to potentially
follow three different types of electron transfer processes for
the said chemical reactions: (i) delocalization of the unpaired
electron throughout the ligand backbone followed by C−N
bond formation, (ii) electron transfer from the radical ligand to
The phenomenon of delocalization of electronic spin along a
ligand backbone is implicated in the most frequently observed
electron transfer reactivity.20 However, for such a phenomenon
planarity of the ligand backbone is essential.21 Better
delocalization of the electron increases unpaired spin in the
reaction center and thus leads to enhanced reactivity. In our
case it is evident that the pendant phenyl ring of the ligand L1 is
G
DOI: 10.1021/acs.inorgchem.5b02110
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 7. Dihedral Angle between the Ligand Planes and Spin-Density Plots of the Optimized Structures of [Pd(L1)Cl2]−
significantly out of plane from the azo pyridine moiety of the
ligand L1. For example, the dihedral angle between the planes
of the two nonconjugative parts of the ligands is as follows: L1b
(28.22 °), L1a (32.71°), L1e (32.74 °), L1d (64.09 °), L1f
(73.54°) (Scheme 7). Thus, introduction of bulky substituents
at the ortho position increases the dihedral angle. Accordingly
the localization of the unpaired spin on the pendent aryl ring
decreases as shown in Scheme 7. Thus, it is expected that
[Pd(L1d)Cl2] should be the least reactive system in this series.
However, in contrast, [Pd(L1d)Cl2] was shown experimentally
to have the highest tendency for C−N bond formation
reaction. Thus, the argument of C−N bond formation
occurring as a result of electron delocalization along the ligand
backbone is not valid, and an alternative mechanistic scenario
must be operative.
more facile than Cl• elimination. Given this result, we planned
to examine a similar reaction with a dichloro-substituted
complex, [Pd(L1d)Cl2]. We anticipated that the reaction rate in
this case will be slower than the previous reactions of
[Pd(L1b)Cl2]/[Pd(L1c)Cl2] as C−Cl activation is more difficult
than C−H activation reaction. However, it is found that in case
of [Pd(L1d)Cl2], the reaction rate is faster than the reaction of
[Pd(L1b)Cl2]/[Pd(L1c)Cl2], and the product is 1h. The
decreasing order of the reaction rates with the series of
different ligands is as follows: L1d > L1c ≈ L1b > L1a > L1e > L1f.
Thus, a comparison of reaction rates in a series of substituted
ligands (L1) did not follow the anticipated trend, as per the
previous arguments. Thus, the scenario of initial electron
transfer from the reduced azo ligand to the amine substrate
associated with C−N bond formation is unlikely.
The second possibility involves electron transfer from the
reduced ligand to the incoming substrate. Radical ligand-to-
substrate electron transfer reactivity is well-documented1a,e in
literature in a series of transformations catalyzed by redox-
active transition metal complexes. Thus, in our case one may
expect electron transfer from a reduced azo-aromatic ligand in
the palladium complex to the incoming amine, generating a
radical intermediate complex followed by the aforementioned
C−N bond formation reaction. Notably, examination of the
rates of reactions indicates that the reaction rate is greatly
influenced by the substitution on the pendant aryl ring of the
ligand L1 in the complex [Pd(L1)Cl2]. Such observation may
suggest that the elimination step is associated with the rate-
limiting step of the reaction. These reactions will involve either
formal H2 or HCl elimination in case of L1 ligands with at least
one H atom in ortho position (i.e., one ortho position
unsubstituted in ligands L1a, L1b, or L1c) or with chloro
substituents in both ortho positions (ligand L1d), respectively,
with 1°/2° amines. Corresponding reactions with 3° amines
will result in alkane and alkyl/aryl halides, respectively. GC-MS
characterization of the crude reaction mixture indeed reveals
the elimination of chlorobenzene from the reaction of
[Pd(L1d)Cl2] and triphenylamine (supporting Figure S20).
Now at this stage we thought of comparing the reaction rates
between the two monochloro-substituted complexes, namely,
[Pd(L1b)Cl2], [Pd(L1c)Cl2] and the secondary amine Me2NH.
The reaction rates are similar and yielded the products 1d and
1h, respectively. This result indicates that H• elimination is
Then we examine the third possibility of an intraligand
electron transfer22 in L1 for the above reaction. Here, we
assume that the reduced azo chromophore in the ligand
transfers electron from π*-azo (singly occupied molecular
orbital (SOMO)) to the π*-orbital (LUMO) of the pendant
aryl ring of the ligand. Such electron transfer possibility has also
been noted previously for azoaromatic systems, by us23 as well
as by others.7 Examination of the MOs of the reduced complex
[Pd(L1a)Cl2] shows that the LUMO and LUMO+1 of the
complex are close lying and are significantly localized on the
pendant phenyl ring of the ligand L1a. Localization of these
MOs on the pendent aryl ring are expected to be controlled by
the substitution on this ring and will be enhanced in case of
electron-withdrawing substituent(s). Thus, in the complexes of
chloro-substituted ligand LUMO and LUMO+1 are localized
primarily on the pendent aryl ring. In contrast, in the case of
electron-donating substituted ligand like 4-methyl/2,6-dimethyl
groups on the pendent aryl ring, localization becomes more
prominent on the azo function as well as on the pyridine ring.
Time-dependent (TD) DFT calculation of each of the
complexes shows that the SOMO to LUMO and LUMO+1
transition occurs at ∼550 to 600 nm, which matches well with
the experimental data. For example, experimentally this appears
at ∼500−650 nm (Table S4 and Figure S21). It may be noted
here that this value gradually shifts to higher wavelengths with
the introduction of electron-withdrawing substituent and shifts
to shorter wavelength by the introduction of electron-donating
groups on the pendent aryl ring. Such a change indicates
H
DOI: 10.1021/acs.inorgchem.5b02110
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 8. Plot of (a) Experimentally observed SOMO → LUMO/LUMO+1 transition energy with different substituents. (b) Reaction completion
time for different ligands L1.
Figure 9. Associated orbitals for SOMO → LUMO transitions.
smaller and larger energy differences between the SOMO and
LUMO/LUMO+1 in case of electron-withdrawing and
electron-donating substituents on the pendent aryl ring,
respectively (Figure 8). Notably, localization of LUMO and
LUMO+1 on the pendent aryl ring suggests the distinct
possibility of reduction of this aryl ring. The smaller the
difference in energy between SOMO and LUMO/LUMO+1,
the better the reducibility of the aryl ring is. Note here that such
reduction of the pendent aryl ring is essential for the
aforementioned C−N bond formation reaction. Though it
was not possible to isolate the one-electron reduced pendent
aryl ring containing intermediate complex, we, however, were
successful in isolating another compound from a similar
reduction reaction in the presence of triphenylphosphine.
After its isolation and characterization by X-ray structure
determination, it was found that the pendant aryl ring is
metalated in the complex (Scheme 6 and Figure 7). Such
metalation upon azo-reduction goes parallel to our idea of
electron transfer from π*-azo to π*-orbital of pendant aryl ring
of the ligand L1. Thus, we propose that electron transfer to the
aryl ring followed by H• (Cl•) elimination reaction from the
reduced complexes is the key step for this unusual reaction. It
may be noted that without ligand reduction no such ortho-
metalation reaction of the pendent aryl ring of ligands L1 is
possible. Moreover, a comparison of the rate of reaction with
differently substituted ligands shows a good correlation with the
energy difference between the SOMO and LUMO/LUMO+1,
reflecting the effect of ligand substituents on the pendent aryl
ring of the ligand L1 (Figure 8). Associated molecular orbitals
are also shown in Figure 9.
Moreover a prerequisite that for any bond formation reaction
is that the two bond-forming atoms should be close. Thus, the
scenario of an ortho-metalation intermediate followed by
reductive elimination of H2/HCl/alkane/alkyl halide depend-
ing upon the ligand substitution and a C−N bond formation
reaction resulted the one-electron reduced azo-anion radical
ligand containing product as shown in Scheme 5. Given the
above experimental finding we strongly believe that this
reaction pathway is a reasonable explanation of the observed
chemical reaction.24 Moreover, ortho-metalation of the phenyl
ring in Pd-compexes is well-documented25 in the literature.
Notably, reaction with primary amine loses a proton (NH) and
undergoes oxidation to result in the nonradical mono anionic
ligand coordinated to a palladium(II) center.
I
DOI: 10.1021/acs.inorgchem.5b02110
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
N, 15.20%. UV−vis (CH2Cl2): λ [nm] (ε, M−1 cm−1) = 272(9455),
350(13 290), 369sh(12 730), 387sh(12 040), 533sh(4950) 552(5470),
583(5420), 602sh(4050), 643(4520).
CONCLUSIONS
■
To conclude, in this report we have disclosed an unusual ortho-
C−N bond-forming reactivity of amines at the pendent aryl
ring of coordinated azo-aromatic ligands L1 in their palladium
complexes. The reaction is successful for a large number of
aliphatic and aromatic amines including 1°, 2°, and 3° amines.
Detailed investigation shows that initial azo reduction
associated with π*-azo to π*-aryl electron transfer initiates
the C−N bond fusion reaction. The reaction has resulted in a
series of palladium(II) complexes with a new redox non-
innocent azo-anion radical ligands. The complexes have been
characterized thoroughly by X-ray crystal structure determi-
nation, EPR, cyclic voltammetry, solution-state magnetic
properties studies, and finally by DFT calculations. Variable-
temperature magnetic studies of the representative complex 1c
show a relatively strong (J = −40 cm−1) magnetic interaction
between ligand radical centers. A comparison of the redox
features of the palladium complex containing a neutral NNN
donor ligand to that of the mono-anionic ligand shows that the
neutral NNN donor is a far more redox-active. Thus, the
isolated Pd-complexes (1a−h) display two voltammetric waves
due to redox events at the coordinated ligand. As an outcome
of the interplay of redox dynamics, this genre of systems may
turn out to be unique in Pd-catalyzed bond formation reactions
without using any metal-based redox events. Our investigations
in this area are continuing.
[Pd(L2b)Cl], 1b. Blue colored solid. Yield: 73%. Anal. Calcd for
C15H18ClN4Pd: C, 45.47; H, 4.58; N, 14.14. Found C, 45.44; H, 4.56;
N, 14.12%. UV−vis (CH2Cl2): λ [nm] (ε, M−1 cm−1) = 327(13 320),
396sh(2500), 585(960), 606sh(740), 640(880).
[Pd(L2c)Cl], 1c. Blue colored solid. Yield: 74%. Anal. Calcd for
C18H16ClN4Pd: C, 50.25; H, 3.75; N, 13.02. Found C, 50.15; H, 3.70;
N, 13.00%. UV−vis (CH2Cl2): λ [nm] (ε, M−1 cm−1) = 293sh(6212),
409(4457), 562(4074), 641(3841), 390(3096). IR (KBr, cm−1): 1303
[ν(NN)].
[Pd(L2d)Cl], 1d. Blue colored solid. Yield: 74%. Anal. Calcd for
C13H13Cl2N4Pd: C, 38.78; H, 3.25; N, 13.92. Found C, 38.73; H, 3.23;
N, 13.88. UV−vis (CH2Cl2): λ [nm] (ε, M−1 cm−1) = 273(7280),
289sh(7140), 338(9980), 390sh(5600), 555(2210), 583(1990),
603sh(1520), 641(1600).
[Pd(L2e)Cl], 1e. Blue colored solid. Yield: 75%. Anal. Calcd for
C15H17Cl2N4Pd: C, 41.83; H, 3.98; N, 13.01. Found C, 41.80; H, 3.95;
N, 12.97%. UV−vis (CH2Cl2): λ [nm] (ε, M−1 cm−1) = 273(10 320),
355(13 150), 390sh(10 620), 528sh(4470) 548sh(4780), 576(5250),
595sh(8300), 632(4140).
[Pd(L2f)Cl], 1f. Blue colored solid. Yield: 76%. Anal. Calcd for
C18H15Cl2N4Pd: C, 46.53; H, 3.25; N, 12.06. Found C, 46.50; H, 3.24;
N, 12.04%. UV−vis (CH2Cl2): λ [nm] (ε, M−1 cm−1) = 248(1700),
346(8780), 368sh(8090), 393(8000), 532sh(4140), 577(4780),
601sh(4190), 637(3880), 690(1440).
[Pd(L2g)Cl], 1g. Blue colored solid. Yield: 78%. Anal. Calcd for
C23H17Cl2N4Pd: C, 52.45; H, 3.25; N, 10.64. Found C, 52.40; H, 3.23;
N, 10.60%. UV−vis (CH2Cl2): λ [nm] (ε, M−1 cm−1) = 346(1770),
379sh(14 180), 517sh(5810), 537sh(6250), 566(7280), 584sh(5560),
623(5470).
EXPERIMENTAL SECTION
■
(i). Materials. PdCl2 was purchased from Arora-Matthey Limited.
All other reagents and chemicals were purchased from commercial
sources and used without further purifications. Solvents are dried and
deoxygenated prior to use. Tetrabutylammonium perchlorate was
prepared and recrystallized as reported earlier.26 Caution! Perchlorates
have to be handled with care and appropriate safety precautions.
(ii). Physical Measurements. A PerkinElmer Lambda 950
spectrophotometer was used to record UV−vis spectra. Infrared
spectra were obtained using a PerkinElmer 783 spectrophotometer. 1H
NMR spectra were recorded on a Bruker Avance 300, 400, or 500
MHz spectrometer, and SiMe4 was used as the internal standard. A
PerkinElmer 240C elemental analyzer was used to collect micro-
analytical data (C, H, N). ESI mass spectra were recorded on a micro
mass Q-TOF mass spectrometer (serial no. YA 263). All electro-
chemical measurements were performed using a PC-controlled PAR
model 273A electrochemistry system. Cyclic voltammetric experi-
ments were performed under nitrogen atmosphere using a Ag/AgCl
reference electrode, with a Pt disk working electrode and a Pt wire
auxiliary electrode either in dichloromethane (1a−h) or in acetonitrile
[Pd(L1a−1f)Cl2] solution containing supporting electrolyte, 0.1 M
Bu4NClO4. A Pt wire gauge working electrode was used for exhaustive
electrolyses. E1/2 for the ferrocenium-ferrocene couple under our
experimental conditions was 0.39 V. X-band EPR spectra were
recorded with a JEOL JES-FA200 spectrometer.
[Pd(L2h)Cl], 1h. Blue colored solid. Yield: 72%. Anal. Calcd for
C13H13Cl2N4Pd: C, 38.78; H, 3.25; N, 13.92. Found C, 38.69; H, 3.20;
N, 13.88%. UV−vis (CH2Cl2): λ [nm] (ε, M−1 cm−1)= 277(6030),
351(7610), 583(2770), 640(2080).
Nuclear Magnetic Resonance Spectrum of Oxidized Complex
[1f]+. In an NMR tube, 20 mg (0.05 mmol) of 1f was mixed with a
slight excess (15 mg, 0.06 mmol) of iodine in 2 mL of CDCl3 solution.
The solution turned dark blue in color. The 1H NMR spectrum of the
1
crude oxidized product was recorded instantaneously. The H NMR
spectrum is given below:
1
[1f]+. H NMR (500 MHz, CDCl3) δ 7.702 (d, J = 6 Hz), δ 7.537
(t, J = 8 Hz), δ 7.453−7.406 (m), δ 7.353−7.311 (m), δ 7.250, δ 7.200
(d, J = 7.5 Hz), δ 6.996 (t, J = 8 Hz), δ 6.810 (d, J = 8 Hz), δ 6.588 (d,
J = 7 Hz).
(iv). X-ray Crystallography. Crystallographic data for complexes
1c, 1e, 1f, 1h, and Pd−I are collected in Table S5. Suitable X-ray
quality crystals of these complexes are obtained either by the slow
evaporation of a dichloromethane−hexane solution of the complex or
slow diffusion of a dichloromethane solution of the complex into
hexane. All data were collected on a Bruker SMART APEX-II
diffractometer, equipped with graphite-monochromated Mo Kα
radiation (λ = 0.710 73 Å) and were corrected for Lorentz polarization
effects. 1c: A total of 21 535 reflections were collected, of which 3760
were unique (Rint = 0.025) 1e: A total of 14 342 reflections were
collected, of which 3715 were unique (Rint = 0.024), satisfying the I >
2σ(I) criterion, and were used in subsequent analysis. 1f: A total of
(iii). Synthesis. The ligands L1a−L1f were prepared by following
the reported procedure.27 Also complexes [Pd(L1a‑1f)Cl2] were
prepared and characterized by following the reported procedure.11a
Synthesis of [Pd(L2)Cl], 1. All the reactions were performed
following a general procedure. A mixture of [Pd(L1)Cl2] (1.0 equiv),
the respective primary/secondary/tertiary amine (1.0 equiv), and
cobaltocene (1.0 equiv) in methanol (10 mL) was heated at 300−323
K (depending on the ligand substitution) in argon-filled Schlenk line
for 0.5−1.5 h. The crude product, thus obtained, was purified on a
preparative alumina thin-layer chromatography plate using dichloro-
methane−hexane (1:50) mixture as eluent. The isolated yields and
characterization of the products were as follows:
100 094 reflections were collected, of which 2962 were unique (Rint
=
0.089). 1h: A total of 18 681 reflections were collected, of which 5784
were unique (Rint = 0.039). Pd−I: A total of 16 587 reflections were
collected, of which 5724 were unique (Rint = 0.083). The structures
were solved by employing the SHELXS-2013 program package and
were refined by full-matrix least-squares based on F2 (SHELXL-
2013).28 All hydrogen atoms were added in calculated positions.
(v). Solution-State Magnetic Susceptibility Measurement by
Evans Method. Magnetic susceptibility measurements for the
isolated species 1c, 1e, 1f, and 1g were made by Evans method29
with a Bruker Advance 500 MHz spectrometer at 300 K. The solution
[Pd(L2a)Cl], 1a. Blue colored solid. Yield: 72%. Anal. Calcd for
C13H14ClN4Pd: C, 42.41; H, 3.83; N, 15.22. Found C, 42.35; H, 3.81;
J
DOI: 10.1021/acs.inorgchem.5b02110
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
magnetic susceptibility was determined by using the following
equation:
2011 and 01(274)/13/EMR-II, respectively. S.G. sincerely
thanks DST for a J. C. Bose fellowship. D.S. and P.G. are
thankful to the CSIR for their fellowship support. S.S. is
thankful to the DST inspire fellowship support. Crystallography
was performed at the DST-funded National Single Crystal
Diffractometer facility at the Department of Inorganic
Chemistry, IACS. We thank Prof. S. Bhattacharya of Jadavpur
Univ. for GC-MS analysis.
−3Δf
4πFm
χg =
+ χD
where χg is the gram susceptibility of the solvent in cm3 g−1, Δf is the
chemical shift difference (in Hz) between a reference proton in the
sample and that in a solution lacking the paramagnetic compound, F is
the fixed probe frequency of the spectrometer, m is the mass of the
complex in grams in 1 cm3 of solution, and χD is the diamagnetic
susceptibility. The corrections for the susceptibility of the solvent and
the difference in densities of the solvent and the solution are ignored.30
The diamagnetic correction to the molar susceptibility χm was
calculated from Pascal’s constants,31 and the effective magnetic
moment μeff was calculated by using the standard equation, where
χm is the corrected molar susceptibility and T is the temperature in K.
DEDICATION
■
Dedicated to Professor Animesh Chakravorty on his 80th birth
anniversary.
REFERENCES
■
μ
= 2.828(χmT)1/2(μB)
(1) (a) Chirik, P. J.; Wieghardt, K. Science 2010, 327, 794−795.
(b) Dzik, W. I.; van der Vlugt, J. I.; Reek, J. N. H.; de Bruin, B. Angew.
Chem., Int. Ed. 2011, 50, 3356−3358. (c) Lyaskovskyy, V.; de Bruin, B.
ACS Catal. 2012, 2, 270−279. (d) Gruetzmacher, H. Angew. Chem.,
Int. Ed. 2008, 47, 1814−1818. (e) Chirik, P. J. Inorg. Chem. 2011, 50,
9737−9740. (f) Kaim, W. Eur. J. Inorg. Chem. 2012, 2012, 343−348.
(2) (a) Kaim, W.; Schwederski, B. Coord. Chem. Rev. 2010, 254,
1580−1588. (b) Kaim, W. Dalton Trans. 2003, 761−768.
(3) (a) Vickers, E. B.; Giles, I. D.; Miller, J. S. Chem. Mater. 2005, 17,
1667−1672. (b) Wang, Z.-X.; Zhang, X.; Zhang, Y.-Z.; Li, M.-X.; Zhao,
H.; Andruh, M.; Dunbar, K. R. Angew. Chem., Int. Ed. 2014, 53,
11567−11570.
eff
(vi). SQUID Magnetic Susceptibility Measurements of Solid
Material. Temperature-dependent magnetic susceptibility measure-
ments for 1c were performed with a Quantum-Design MPMS-XL-5
SQUID magnetometer equipped with a 5 T magnet in the range from
295 to 2.0 K at a magnetic field of 0.5 T. The powdered sample was
contained in a gelatin capsule and fixed in a nonmagnetic sample
holder. Each raw data file for the measured magnetic moment was
corrected for the diamagnetic contribution of the gelatin capsule
according to Mdia(capsule) = χg·m·H, with an experimentally obtained
gram susceptibility of the gelatin capsule. The molar susceptibility data
were corrected for the diamagnetic contribution using the Pascal
constants and the increment method according to Haberditzl.32 The
experimental data were fitted using the appropriate Heisenberg−
Dirac−van Vleck spin Hamiltonian that includes the isotropic
exchange coupling constant and Zeeman splitting (see eq 1 in the
main text).33 Temperature-independent paramagnetism (TIP) and
paramagnetic impurities (PI) were included according to χcalc = (1 −
PI)·χ + TIP (TIP = 20 × 10−6 cm3 mol−1, PI = 5% (per one molecule
of 1c, fixed value)). Intermolecular interactions were considered in a
mean field approach by using a Weiss temperature Θ.34 The Weiss
temperature Θ = −15.2 K (defined as Θ = zJS(S + 1)/3k) relates to
intermolecular interactions zJ of −42.3 cm−1 for 1c, where J is the
interaction parameter between two nearest neighbor magnetic centers,
k is the Boltzmann constant (0.695 cm−1·K−1), and z is the number of
nearest neighbors.
(4) (a) Goswami, S.; Sengupta, D.; Paul, N. D.; Mondal, T. K.;
Goswami, S. Chem. - Eur. J. 2014, 20, 6103−6111. (b) Paul, N. D.;
Rana, U.; Goswami, S.; Mondal, T. K.; Goswami, S. J. Am. Chem. Soc.
2012, 134, 6520−6523.
(5) (a) Praneeth, V. K. K.; Ringenberg, M. R.; Ward, T. R. Angew.
Chem., Int. Ed. 2012, 51, 10228−10234. (b) Luca, O. R.; Crabtree, R.
H. Chem. Soc. Rev. 2013, 42, 1440−1459.
(6) (a) Russell, S. K.; Lobkovsky, E.; Chirik, P. J. J. Am. Chem. Soc.
2011, 133, 8858−8861. (b) Monfette, S.; Turner, Z. R.; Semproni, S.
P.; Chirik, P. J. J. Am. Chem. Soc. 2012, 134, 4561−4564. (c) Darmon,
J. M.; Stieber, S. C. E.; Sylvester, K. T.; Fernandez, I.; Lobkovsky, E.;
Semproni, S. P.; Bill, E.; Wieghardt, K.; DeBeer, S.; Chirik, P. J. J. Am.
Chem. Soc. 2012, 134, 17125−17137.
(7) Broere, D. L. J.; de Bruin, B.; Reek, J. N. H.; Lutz, M.; Dechert,
S.; van der Vlugt, J. I. J. Am. Chem. Soc. 2014, 136, 11574−11577.
(8) Hall, G. B.; Kottani, R.; Felton, G. A. N.; Yamamoto, T.; Evans,
D. H.; Glass, R. S.; Lichtenberger, D. L. J. Am. Chem. Soc. 2014, 136,
4012−4018.
(9) (a) Sanyal, A.; Chatterjee, S.; Castineiras, A.; Sarkar, B.; Singh, P.;
Fiedler, J.; Zalis, S.; Kaim, W.; Goswami, S. Inorg. Chem. 2007, 46,
8584−8593. (b) Saha, A.; Ghosh, A. K.; Majumdar, P.; Mitra, K. N.;
Mondal, S.; Rajak, K. K.; Falvello, L. R.; Goswami, S. Organometallics
1999, 18, 3772−3774.
(10) (a) Johnston, C. W.; McKinnon, S. D. J.; Patrick, B. O.; Hicks,
R. G. Dalton Trans. 2013, 42, 16829−16836. (b) Sanz, C. A.;
Ferguson, M. J.; McDonald, R.; Patrick, B. O.; Hicks, R. G. Chem.
Commun. 2014, 50, 11676−11678.
(11) (a) Roy, S.; Hartenbach, I.; Sarkar, B. Eur. J. Inorg. Chem. 2009,
2009, 2553−2558. (b) Sinha, C. Ph. D. Thesis, Jadavpur University,
1990.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.5b02110.
■
S
Figures and tables giving crystallographic details and
selected bond parameters of the complexes 1c, 1e, 1f, 1h,
and Pd−I, relevant cyclic voltammograms, spectral data,
ORTEP representations, contour plots and Cartesian
coordinates of the optimized structures. (PDF)
Additional crystallographic details. (CIF)
AUTHOR INFORMATION
Corresponding Author
*E-mail: icsg@iacs.res.in.
■
(12) Kamar, K. K.; Das, S.; Hung, C.-H.; Castineiras, A.; Kuz’min, M.
D.; Rillo, C.; Bartolome, J.; Goswami, S. Inorg. Chem. 2003, 42, 5367−
5375.
(13) Samanta, S.; Ghosh, P.; Goswami, S. Dalton Trans. 2012, 41,
2213−2226.
Notes
The authors declare no competing financial interest.
(14) (a) Desiraju, G. R.; Steiner, T. The Weak Hydrogen Bond In
Structural Chemistry and Biology; Oxford University Press, 1999.
(b) Bhogala, B. R.; Nangia, A. Cryst. Growth Des. 2003, 3, 547−554.
(c) Janiak, C. Dalton Trans. 2000, 3885−3896. (d) Nishio, M.
CrystEngComm 2004, 6, 130−158. (e) Seth, S. K. J. Mol. Struct. 2014,
ACKNOWLEDGMENTS
■
The research was supported by the Department of Science and
Technology (DST), India, and Council of Scientific and
Industrial Research (CSIR) funded projects, SR/S2/JCB-09/
K
DOI: 10.1021/acs.inorgchem.5b02110
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
1070, 65−74. (f) Seth, S. K.; Sarkar, D.; Roy, A.; Kar, T.
CrystEngComm 2011, 13, 6728−6741.
(15) (a) Seth, S. K.; Sarkar, D.; Kar, T. CrystEngComm 2011, 13,
4528−4535. (b) Seth, S. K. CrystEngComm 2013, 15, 1772−1781.
(16) (a) Koivisto, B. D.; Ichimura, A. S.; McDonald, R.; Lemaire, M.
T.; Thompson, L. K.; Hicks, R. G. J. Am. Chem. Soc. 2006, 128, 690−
691. (b) Chi, Y.-H.; Shi, J.-M.; Li, H.-N.; Wei, W.; Cottrill, E.; Pan, N.;
Chen, H.; Liang, Y.; Yu, L.; Zhang, Y.-Q.; Hou, C. Dalton Trans. 2013,
42, 15559−15569.
(17) (a) Hicks, R. G.; Lemaire, M. T.; Oehrstroem, L.; Richardson, J.
F.; Thompson, L. K.; Xu, Z. J. Am. Chem. Soc. 2001, 123, 7154−7159.
(b) Pal, S. K.; Itkis, M. E.; Reed, R. W.; Oakley, R. T.; Cordes, A. W.;
Tham, F. S.; Siegrist, T.; Haddon, R. C. J. Am. Chem. Soc. 2004, 126,
1478−1484.
(18) Chakraborty, I.; Panda, B. K.; Gangopadhyay, J.; Chakravorty, A.
Inorg. Chem. 2005, 44, 1054−1060.
(19) Cope, A. C.; Siekman, R. W. J. Am. Chem. Soc. 1965, 87, 3272−
3.
(20) Widger, L. R.; Jiang, Y.; Siegler, M. A.; Kumar, D.; Latifi, R.; de
Visser, S. P.; Jameson, G. N. L.; Goldberg, D. P. Inorg. Chem. 2013, 52,
10467−10480.
(21) Blonski, C.; Myers, A. W.; Palmer, M.; Harris, S.; Jones, W. D.
Organometallics 1997, 16, 3819−3827.
(22) Pfeffer, M. G.; Schaefer, B.; Smolentsev, G.; Uhlig, J.;
Nazarenko, E.; Guthmuller, J.; Kuhnt, C.; Waechtler, M.; Dietzek,
B.; Sundstroem, V.; Rau, S. Angew. Chem., Int. Ed. 2015, 54, 5044−
5048.
(23) Ghosh, P.; Samanta, S.; Roy, S. K.; Demeshko, S.; Meyer, F.;
Goswami, S. Inorg. Chem. 2014, 53, 4678−4686.
(24) (a) Dimmer, J.-A.; Hornung, M.; Wuetz, T.; Wesemann, L.
Organometallics 2012, 31, 7044−7051. (b) Kumar, S.; Mani, G.;
Mondal, S.; Chattaraj, P. K. Inorg. Chem. 2012, 51, 12527−12539.
(25) Khare, G. P.; Little, R. G.; Veal, J. T.; Doedens, R. J. Inorg. Chem.
1975, 14, 2475−9.
(26) Goswami, S.; Mukherjee, R.; Chakravorty, A. Inorg. Chem. 1983,
22, 2825−32.
(27) Ghosh, P.; Samanta, S.; Roy, S. K.; Joy, S.; Kramer, T.;
McGrady, J. E.; Goswami, S. Inorg. Chem. 2013, 52, 14040−14049.
(28) Sheldrick, G. M. Acta Crystallogr., Sect. A: Found. Crystallogr.
2008, 64, 112−122.
(29) (a) Evans, D. F. J. Chem. Soc. 1959, 2003−5. (b) Naklicki, M. L.;
White, C. A.; Plante, L. L.; Evans, C. E. B.; Crutchley, R. J. Inorg. Chem.
1998, 37, 1880−1885. (c) Casabianca, L. B.; An, D.; Natarajan, J. K.;
Alumasa, J. N.; Roepe, P. D.; Wolf, C.; de Dios, A. C. Inorg. Chem.
2008, 47, 6077−6081.
(30) Weast, R. C. CRC Handbook of Chemistry and Physics; CRC
Press, 1980; Vol. 60.
(31) Drago, R. S. Physical methods for chemists, 2nd ed.; Saunders
College Publishing: Philadelphia, PA, 1992.
(32) Haberditzl, W. Angew. Chem., Int. Ed. Engl. 1966, 5, 288−98.
(33) Simulation of the experimental magnetic data was performed
with the julX program (Bill, E. Max-Planck Institute for Chemical
Energy Conversion: Mulheim/Ruhr, Germany).
̈
(34) Kahn, O. Molecular Magnetism; VCH, Weinheim, 1993.
L
DOI: 10.1021/acs.inorgchem.5b02110
Inorg. Chem. XXXX, XXX, XXX−XXX