Charge Effects in PCP Pincer Complexes of NiII bearing Phosphinite and Imidazol(i)ophosphine Coordinating Jaws: From Synthesis to Catalysis through Bonding Analysis

Article abstract of DOI:10.1002/chem.201502491

This contribution reports on a new family of Ni^II pincer complexes featuring phosphinite and functional imidazolyl arms. The proligands ^RPIMC^HOP^R′ react at room temperature with Ni^II precursors to giv

Full text of DOI:10.1002/chem.201502491

DOI: 10.1002/chem.201502491
Full Paper
&
Pincer Complexes
Charge Effects in PCP Pincer Complexes of NiII bearing
Phosphinite and Imidazol(i)ophosphine Coordinating Jaws: From
Synthesis to Catalysis through Bonding Analysis
Boris Vabre,[a] Yves Canac,*[b, c] Christine Lepetit,*[b, c] Carine Duhayon,[b, c] Remi Chauvin,*[b, c]
and Davit Zargarian*[a]
This report is dedicated to the memory of our colleague Dr. Guy Lavigne (1947–2015) and to his lifelong passion for organometallic
chemistry.
Abstract: This contribution reports on a new family of NiII
3–5 were then treated with AgOTf in acetonitrile to give the
corresponding cationic species [(RPIMCOPR)Ni(MeCN)][OTf]
[R=Ph, 6a (89%) or iPr, 6b (90%)], [(RPIMIOCOPR)Ni(MeCN)]
[OTf]2 [R=Ph, 7a (79%) or iPr, 7b (88%)], and
[(NHCCOPR)Ni(MeCN)][OTf] [R=Ph, 8a (85%) or iPr, 8b
(84%)]. All new complexes have been characterized by NMR
and IR spectroscopy, whereas 3b, 3c, 5b, 6b, and 8a were
also subjected to X-ray diffraction studies. The acetonitrile
adducts 6–8 were further studied by using various theoreti-
cal analysis tools. In the presence of excess nitrile and
amine, the cationic acetonitrile adducts 6–8 catalyze hydroa-
mination of nitriles to give unsymmetrical amidines with cat-
alytic turnover numbers of up to 95.
pincer complexes featuring phosphinite and functional imi-
R
dazolyl arms. The proligands PIMCHOPR’ react at room tem-
perature with NiII precursors to give the corresponding com-
plexes [(RPIMCOPR’)NiBr], where RPIMCOPR =kP,kC,kP-{2-
(R’2PO),6-(R2PC3H2N2)C6H3}, R=iPr, R’=iPr (3b, 84%) or Ph
(3c, 45%). Selective N-methylation of the imidazole imine
moiety in 3b by MeOTf (OTf=OSO2CF3) gave the corre-
sponding imidazoliophosphine [(iPrPIMIOCOPiPr)NiBr][OTf],
4b, in 89% yield (iPrPIMIOCOPiPr =kP,kC,kP-{2-(iPr2PO),6-
(iPr2PC4H5N2)C6H3}). Treating 4b with NaOEt led to the NHC
derivative [(NHCCOPiPr)NiBr], 5b, in 47% yield (NHCCOPiPr
=
kP,kC,kC-{2-(iPr2PO),6-(C4H5N2)C6H3)}). The bromo derivatives
k3 ligand architecture, conformation, and steric/electronic
properties can be altered to generate a multitude of chemical
entities possessing different physical and chemical properties.
As the central carbyl moiety of ECE pincer ligands is strongly
s-electron donating, particular efforts have been devoted to
the study of “balanced” pincers with electron-withdrawing “co-
ordinating jaws”, like phosph(in)ite moieties. The objective is
to design new, unsymmetrical pincer ligands by judicious com-
binations of charge-neutral donor moieties, represented by E,
and positively charged jaws such as imidazoliophosphine moi-
eties, also regarded as acceptor–donor adducts of phospheni-
ums with N-heterocyclic carbenes (NHCs), [(NHC)R2P:]+. Our in-
terests in the coordination chemistry of bi-[4] or tridentate[5] li-
gands featuring imidazoliophosphine units,[6] and in the orga-
nonickel chemistry of pincer complexes (Figure 1; e.g., NCN,[7]
Introduction
Pincer ligands are considered as privileged platforms for devel-
oping practical applications in homogeneous catalysis and
functional materials, as well as for exploring fundamental
bonding and reactivity patterns.[1] Since their introduction
nearly 40 years ago,[2] research has shown that pincer ligands
featuring varying donor moieties, backbones, and metallacycle
sizes often lead to unusual bonding/structural features and
unique reactivities.[3] An important factor that has accelerated
the growth of pincer chemistry is the ease with which the
[a] Dr. B. Vabre, Prof. D. Zargarian
DØpartement de chimie, UniversitØ de MontrØal
MontrØal (QuØbec), H3C 3J7 (Canada)
E-mail: zargarian.davit@umontreal.ca
PCP,[8] POCOP,[9] POCN[10] [11]
led us to collaborate on the devel-
)
[b] Dr. Y. Canac, Dr. C. Lepetit, Dr. C. Duhayon, Prof. R. Chauvin
Laboratoire de Chimie de Coordination (LCC), CNRS
205 Route de Narbonne, 31077 Toulouse (France)
E-mail: yves.canac@lcc-toulouse.fr
opment of a family of unsymmetrical ECE’-type pincer ligands
incorporating [(NHC)R2P:]+ and related coordination moieties.
In a previous report, we communicated the initial fruits of
our collaborative effort, namely an unprecedented family of NiII
christine.lepetit@lcc-toulouse.fr
chauvin@lcc-toulouse.fr
pincer complexes featuring
a meta-phenylene backbone
[c] Dr. Y. Canac, Dr. C. Lepetit, Dr. C. Duhayon, Prof. R. Chauvin
UniversitØ de Toulouse, UPS, INP, LCC
flanked by donor moieties that were either electron-poor
(phosphinites, imidazolyl- and imidazoliophosphines) or elec-
tron-rich (NHCs), and generated metallacycles of different sizes
(5,5 and 5,6).[12] The present contribution reports on the results
31077 Toulouse (France)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201502491.
Chem. Eur. J. 2015, 21, 17403 – 17414
17403
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Results and Discussion
Syntheses
The title complexes were prepared as outlined in the synthetic
R
sequence shown in Scheme 1. The PIMCHOPR’ proligands 2b
and 2c were prepared similarly to their previously reported Ph
analogue PhPIMCHOPPh, 2a,[12] by adding two equivalents of
chlorophosphines to the doubly deprotonated 3-hydroxyphe-
nylimidazole 1.[14] Thus, iPrPIMCHOPiPr 2b was obtained in 70%
yield by using 2 equivalents of iPr2PCl, whereas sequential ad-
dition of iPr2PCl and Ph2PCl (1 equivalent each) gave the
“mixed” iPrPIMCHOPPh proligand 2c in 72% yield. Formation of
the latter confirmed the anticipated greater nucleophilicity of
the lithiated carbon atom of the imidazole ring, which was
phosphinylated first. The [(PIMCOP)NiBr] complexes 3b and
c were then obtained by direct nickelation of the central CÀH
bond when the PIMCOP proligands 2b and c were treated at
[15]
RT with a slight excess of [Br2Ni(iPrCN)]n in the presence of
NEt3.[16]
Figure 1. Various types of NiII pincer complex.
of our follow-up investigations
on PIMCOP–, PIMIOCOP–, and
NHCCOP–Ni species (Figure 1).
The original synthetic methodol-
ogy has been used to prepare
R
new examples of PIMCOPR’ pro-
ligands (R=iPr; R’=Ph, iPr) and
new [(pincer)NiBr] complexes
bearing PIMCOP, PIMIOCOP, and
NHCCOP ligands. The related
mono- and dicationic acetonitrile
adducts [(RPIMCOPR’)Ni(NCMe)]
[(OTf)], [(RPIMIOCOPR’)Ni(NCMe)]
[(OTf)]2
and
[(NHCCOPR’)Ni(NCMe)][(OTf)]
have also been prepared and
characterized. It was anticipated
that the cationic character of the
latter complexes would enhance
the electrophilicity of the nickel
center. To test this contention,
the CꢀN stretching frequency,
n(CꢀN), and redox potentials, Eopx,
of these cationic species were
measured to evaluate their apti-
tude for promoting nucleophilic
Scheme 1. Synthesis of neutral and cationic PIMCOP-derived NiII pincer complexes.
hydroamination of coordinated
nitriles,[13] allowing the develop-
ment of a new catalytic route to
a variety of amidines. Finally, the experimental studies were
complemented by extensive theoretical analyses to improve
our understanding of the bonding and reactivities in these
complexes.
Similarly to the synthesis of its previously reported Ph ana-
logue 4a,[12] the complex [(iPrPIMIOCOPiPr)NiBr], 4b, was pre-
pared in 89% yield by N-methylation of [(iPrPIMCOPiPr)NiBr], 3b,
with MeOTf (Scheme 1). Subsequent treatment of 4b with
a stoichiometric amount of NaOEt gave [(NHCCOPiPr)NiBr], 5b,
in 47% yield. A similar reactivity of the Ph analogue 4a has
Chem. Eur. J. 2015, 21, 17403 – 17414
www.chemeurj.org
17404
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
been observed with a weaker nucleophile ([Et4N][Cl]), giving
[(NHCCOPPh)NiX] (X=Cl, Br) 5a.[12] Finally, all of the bromo de-
rivatives were converted into their corresponding cationic ace-
tonitrile analogues by reaction with a slight excess of AgOTf in
acetonitrile (Scheme 1).[17]
Table 1. Selected bond lengths (Š) and bond angles (8) for complexes
3a–c, and 5a–b.
Formation of the NHCCOP derivatives 5 is thought to arise
from the attack of a nucleophile (e.g., EtOÀ) on the phospheni-
um center [R2P:]+ in the PIMIOCOP complexes 4 (see the can-
onical forms shown in Scheme 2).[18] The relative reactivities of
4a and 4b were investigated through NMR spectroscopy,
whereby 1 equivalent each of 4a and 4b was allowed to com-
1
NiÀC11 NiÀBr
NiÀC1
NiÀP11
NiÀP1
NiÀP12
NiÀC7
P11-Ni-P12 C11-Ni-Br
pete for 1 equivalent of NaOEt (Scheme 2). The 31P and H NMR
P1-Ni-C7
C1-Ni-Br
spectra (Figures S1 and S2 in Supporting Information) of the
3a[12] 1.945(2) 2.3345(3) 2.1322(5) 2.169(2) 170.30(2)
170.68(5)
172.06(8)
169.54(8)
174.82(6)
173.14(6)
mixture recorded after heating at 508C for one hour indicated
3b
3c
1.939(3) 2.3414(5) 2.1339(8) 2.1682(8) 176.60(3)
1.943(2) 2.3404(5) 2.1383(7) 2.1985(7) 175.90(3)
5a[12] 1.875(2) 2.3521(3) 2.1451(5) 1.946(2)
2
that 4a [dP = +11.1 (d), +147.0 ppm (d); J(P,P)=367 Hz] was
totally converted into 5a [dP = +144.9 ppm (s)] while 4b re-
mained unreacted [dP = +32.6 (d), +184.8 ppm (d); 2J(P,P)=
309 Hz].[19] The greater reactivity of [(PhPIMIOCOPPh)NiBr] 4a rel-
ative to its iPr analogue 4b is in agreement with the higher
electrophilicity of [Ph2P:+] vs. [iPr2P:+].
161.21(6)
162.31(6)
5b
1.870(2) 2.3464(3) 2.1478(5) 1.951(2)
Table 1, and Tables S1 and S2 in Supporting Information);[20]
these structures can be compared to those of their previously
reported all-Ph analogues 3a and 5a, respectively.[12]
The Ni atom in the PIMCOP complexes is located at the
center of a somewhat distorted square-planar environment,
with the two P atoms occupying mutually trans positions
(Figure 2, left and middle). This family of complexes features 5-
Solid-state structural studies
Single crystal X-ray diffraction studies allowed us to establish
the solid-state structures of the [(iPrPIMCOPR’)NiBr] complexes
3b and 3c, and of the [(NHCCOPiPr)NiBr] complex 5b (Figure 2,
Scheme 2. Relative reactivities of [(RPIMIOCOPR’)NiBr][OTf] 4a and 4b towards NaOEt.
Figure 2. Molecular views of the X-ray crystal structures of charge neutral NiII pincer complexes 3b (left), 3c (middle), and 5b (right). Thermal ellipsoids are
drawn at the 50% probability level and H atoms have been omitted for clarity.
Chem. Eur. J. 2015, 21, 17403 – 17414
www.chemeurj.org
17405
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
and 6-membered fused metallacycles and bite angles (P11-Ni-
P12) that are much greater than in their POCOP analogues fea-
turing two fused 5-membered metallacycles (ca. 170–1768 vs.
1648).[21] The bite angles in [(RPIMCOPR’)NiBr] are also influenced
by the P-substituents, being roughly 6–78 wider in the struc-
tures of 3b and 3c bearing the bulkier iPr2P moieties. Steric
factors appear to have less influence over bond lengths, but
the NiÀP distances in 3a and 3b (R=R’: 2.132–2.134 and
2.168–2.169 Š) are significantly shorter than those in 3c (R¼ R’:
2.138 and 2.199 Š).
A distorted square-planar geometry is also adopted by the
[(NHCCOPR’)NiBr] complexes 5a and b, and minimal structural
variations result from the different P-substituents. The small
bite angles (ca. 161–1628) can be attributed to the metallacycle
size (5,5) and correlated with the shorter N1ÀC7 bonds in 5a
and 5b relative to the OÀP bonds in the analogous POCOP
complexes.[22] The NiÀP1 distances in 5a and 5b (av. 2.147 Š)
are longer than those in 3a and 3c (av. 2.133 Š) but remain
comparable to the corresponding distance in a typical POCOP
structure (2.148 Š).[13,22] These observations establish the fol-
lowing order of trans influence: ArOP(iPr)2 ꢁNHC>2-imidazol-
yl-PPh2. The NiÀCN2 distances of approximately 1.95 Š in 5a
and 5b are longer than the corresponding distances of ap-
proximately 1.90–1.92 Š found in previously reported Ni–NHC
complexes.[23] Finally, the central NiÀC1 bond lengths are much
shorter in the 5,5-complexes 5a and 5b than in the 5,6-com-
plexes 3a–c (1.87 vs. 1.94 Š).
Figure 3. Molecular views of 6b (top) and 8a (bottom). Thermal ellipsoids
are drawn at the 50% probability level and the H atoms and the triflate
anion have been omitted for clarity. Selected bond lengths (Š) and bond
angles (8) for 6b: NiÀC1 1.944(2), NiÀN 1.882(2), NiÀP1 2.1572(5), NiÀP2
2.2005(5); P1-Ni-P2 175.70(2), C1-Ni-N 168.92(7), Ni-N-C 172.48(15). Selected
bond lengths (Š) and bond angles (8) for 8a: NiÀC1 1.876(4), NiÀN 1.888(4),
NiÀP1 2.166(2), NiÀC7 1.943(5); P1-Ni-C7 161.0(2), C1-Ni-N 177.5(2), Ni-N-C
177.3(4).
The crystal structures of the cationic complexes 6b and 8a
(Figure 3) also revealed a somewhat distorted square-planar
geometry around the Ni center. The above-noted smaller bite
angle P1-Ni-C7 and shorter NiÀC1 bond in NHCCOP complexes
are confirmed here. Likewise, the NiÀP1 bond lengths in 6b
(ca. 2.157 Š) and 8a (ca. 2.166 Š) reflect the superior trans in-
fluence of the NHC moiety, even though the PR2 moieties
being compared are different. A greater deviation from the
ideally linear NiÀNCMe coordination mode is found in 6b than
in 8a: C1-Ni-Nꢁ1698 vs. 1788; Ni-N-Cꢁ1728 vs. 1778. This de-
viation is likely caused by the steric demand of the inserted
iPr2P moiety in 6b.
features, whereas the extent of d deshielding upon N-meth-
31P
ylation depends on the P-substituents (by ca. 8 ppm for 3a!
4a vs. 21 ppm for 3b!4b). The N+ÀCH3 substituent of the
1
[(PIMIOCOP)NiX] complexes gives a single H NMR resonance
at dH ꢁ +3.2 ppm. Finally, the NHCCOP complexes 5a,b and
8a,b display characteristic 13C NMR carbene signals at dC
ꢁ172–175 ppm [d; 2J(C,P)ꢁ90–110 Hz].
IR spectroscopy, cyclic voltammetry, and theoretical studies
In an effort to estimate the electronic density of the Ni center
in the title complexes, the oxidation potentials Eopx and IR CꢀN
stretching frequencies n(CꢀN) were measured. Wide-scope
computational studies at the density functional theory (DFT)
level were also performed on the cationic acetonitrile adducts
6a,b, 7a,b, 8a,b, and their previously reported POCOP ana-
logues 9a,b (Figure 4).[13,22] The results that are most pertinent
to the present discussion are presented sequentially in the fol-
lowing sections and some of the computational results (includ-
ing geometry optimization studies, Figures S3–S10 and Ta-
bles S3–S8) are provided in the Supporting Information.
Characterization of the proligands and complexes by NMR
spectroscopy
R
The most characteristic NMR features of the PIMCOPR’ proli-
gands 2b and 2c consist of the pair of singlet 31P resonances
for the phosphinite and imidazolophosphine moieties, as well
as the 13C signals for the central C nuclei that are flanked by
3
4
these moieties [dC ꢁ118–119 ppm (dd); J(C,P)ꢁ7–8 Hz, J(C,P’)
ꢁ3–5 Hz].[12] Upon nickelation, a dramatic 31P downfield shift
occurs, giving rise to doublets of doublets with 2J(P,P’)>
300 Hz, as anticipated for the two inequivalent and mutually
trans P nuclei.[24] The 13C resonances for the reacting carbon
nucleus also moved downfield and turned into pseudo-triplets
Measured and calculated n(CꢀN): IR stretching frequencies
n(CꢀN) of coordinated acetonitrile molecules are a priori ex-
pected to reflect the overall donor character (s-donating vs. p-
accepting properties) of coligands. It was indeed reported that
the (a priori nonbonding) lone pair of a nitrile, which is strong-
ly polarized on the nitrogen lone pair, could actually exhibit
some antibonding character. Therefore, n(CꢀN) values are ex-
pected to increase upon s-donation from the coordinated ace-
2
2
[dC ꢁ125 ppm; J(C,P)ꢁ J(C,P’)ꢁ22 Hz].
Conversion of the PIMCOP complexes 3a,b and PIMIOCOP
complexes 4a,b into their respective cationic acetonitrile ad-
ducts 6a,b and 7a,b has a relatively minor impact on the NMR
Chem. Eur. J. 2015, 21, 17403 – 17414
www.chemeurj.org
17406
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
An increase of the n(CꢀN) resulting from s-donation of ace-
tonitrile to the nickel center would normally imply a bond
order increase or a shortening of the CꢀN bond. However, the
calculated CꢀN bond length remains close to 1.158 Š over the
entire series of the pincer complexes depicted in Figure 4 (see
the Supporting Information, Table S6). Similarly, a constant Cꢀ
N bond order of about 2.2 is found over the entire series by
electron localization function (ELF) topological analysis (see
below and Table S7 in the Supporting Information). The n(CꢀN)
values of Table 2 might, therefore, be probing the local electro-
static environment of the CN moiety rather than the electron
density of the nickel center. Indeed, the vibrational frequency
of nitrile groups has been reported to be very sensitive to
their local environment, and nitriles have been used as IR
probes of biomolecular structures and dynamics.[27] Finally, the
n(CꢀN) values of 6b and 7b might also be influenced by the
experimentally observed (Figure 3) and computationally con-
firmed (Table S6) deviation of the Ni-N-CCH3 angle from 1808,
which would decrease the orbital overlap between the NiII
center and NCCH3.
Figure 4. Cationic acetonitrile adducts analyzed by DFT studies.
For the purpose of comparison, a few calculations were per-
formed on theoretical complexes (in the Ph2P series) in which
the MeCN ligand was replaced by a CO ligand, which is cur-
rently used for assessing relative electron densities of metal
centers as a function of the overall electron-donating character
of coligands.[28] The calculated n(CꢀO) values (Table 2) are con-
sistent with the electrochemical trend of Eopx; they decrease
with the electron-richness of the nickel center: PhPIMIOCOPPh >
tonitrile to the nickel center in the pincer complexes under
study.[25,26] Indeed, an increase of 31–46 cmÀ1 was observed for
the n(CꢀN) values within the pincer series (Table 2).
We found that calculated and experimental n(CꢀN) values
are comparable, with a shift of 100–120 cmÀ1 at the PCM-
B3PW91/6-31G** level. In particular, complexes of pincer li-
gands bearing Ph2P moieties (6a, 7a, 8a) exhibit higher n(Cꢀ PhPIMCOPPh
>
PhPOCOPPh >NHCCOPPh.[29]
N) values than their iPr2P analogues (6b, 7b, 8b), an observa-
tion that could be anticipated from the greater electron-donat-
ing character of the iPr substituents, the same trend having
been observed for the previously reported complexes [(PO-
COP)Ni(NCMe)][OTf].[13,17d,24] However, the IR data do not corre-
late with the relative overall electron-donating character of the
pincer ligand, as established previously on the basis of electro-
chemical measurements (NHCCOP>PIMCOPꢁPOCOP>PIMIO-
COP).[12] For example, both DFT calculations and experimental
measurements showed that the n(CꢀN) value is lower in the
PIMCOP adduct 6b than in the NHCCOP adduct 8b (2284 vs.
2296 cmÀ1), whereas the opposite order is a priori expected.
Molecular orbital (MO) analysis: The near-frontier MOs of
[(pincer)Ni(NCMe)]+ complexes (iPr2P series in Figure 4) are
shown in Figures S5–S8 (see the Supporting Information), and
their energy levels are presented in Figure 5. A selection of the
more pertinent frontier MOs are depicted in Figure 6 to facili-
tate the present discussion. Inspection of the MO shapes and
energies indicate that those of the POCOP and PIMCOP com-
plexes are similar in shape and quasi-degenerate, but signifi-
cant differences can be noted in the NHCCOP and PIMIOCOP
complexes. Consistent with the respective electron-richness of
the pincer ligand, the MO levels are shifted up to higher ener-
Table 2. Experimental and calculated IR data for acetonitrile adducts and selected CO adducts of cationic pincer Ni complexes.
[a]
[b]
L=MeCN; n(CN) [cmÀ1
]
L=CO; n(CO) [cmÀ1
]
Calcd
Complexes/L
Exptl
Dn(CN)c
Calcd
Dn(CN)[c]
[(PhPIMIOCOPPh)NiL][OTf]2 7a
[(iPrPIMIOCOPiPr)NiL][OTf]2 7b
[(PhPIMCOPPh)NiL][OTf] 6a
[(iPrPIMCOPiPr)NiL][OTf] 6b
[(PhPOCOPPh)NiL][OTf] 9a[24]
[(iPrPOCOPiPr)NiL][OTf] 9b[17b]
[(NHCCOPPh)NiL][OTf] 8a
[(NHCCOPiPr)NiL][OTf] 8b
MeCN
2291
2284
2287
2284
2297
2292
2299
2296
2253
38
31
34
31
44
39
46
43
0
2408
2393
2407
2395
2403
2397
2400
2397
2378
30
15
29
17
25
19
22
19
0
2079
2071
2065
2062
[a] PCM-B3PW91/6-31G** level of calculation used for MeCN adducts (acetonitrile solvent, e =35.688); [b] PBE/6-31G** level of calculation used for CO ad-
ducts; [c] n(CꢀN) difference between the complexes and free acetonitrile.
Chem. Eur. J. 2015, 21, 17403 – 17414
www.chemeurj.org
17407
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 3. Oxidation potentials (vs. FeCp2) of NiII pincer complexes.[a]
Complexes
Eopx [mV]
Reversibility
[(PhPIMIOCOPPh)Ni(NCMe)][OTf]2 7a
[(iPrPIMIOCOPiPr)Ni(NCMe)][OTf]2 7b
[(PhPIMIOCOPPh)NiBr][OTf] 4a
[(iPrPIMIOCOPiPr)NiBr][OTf] 4b
[(PhPIMCOPPh)Ni(NCMe)][OTf] 6a
[(iPrPIMCOPiPr)Ni(NCMe)][OTf] 6b
[(PhPIMCOPPh)NiBr] 3a
1220
1195
1120
1001
1020
966
820
862
834
530
quasi
no
no
no
no
no
no
no
no
no
no
quasi
quasi
[(iPrPIMCOPiPr)NiBr] 3b
[(iPrPIMCOPPh)NiBr] 3c
[(NHCCOPPh)Ni(NCMe)][OTf] 8a
[(NHCCOPiPr)Ni(NCMe)][OTf] 8b
[(NHCCOPPh)NiBr] 5a
618
500
547
[(NHCCOPiPr)NiBr] 5b
[a] The experiments were carried out at room temperature on solutions
prepared in dry CH2Cl2 containing [nBu4N][PF6] as electrolyte (0.1m). For
complete details of the experimental set-up and conditions, see the Sup-
porting Information.
from a charge-neutral bromo complex to the cationic acetoni-
trile derivative results in a significant increase of the Eopx value
(by ca. 100–200 mV); by comparison, ionization has a more
muted impact on the oxidation potentials of the NHCCOP
complexes, which increased by 30 (5a/8a) and 71 mV (5b/8b).
The influence exerted on the Eopx values by the relative donat-
ing characters of the P-substituents (Ph2P<iPr2P) varied by as
little as 25 mV (7a/7b) and as much as 119 mV (4a/4b). The
highest Eopx value was observed for the dicationic complexes
[(PIMIOCOP)Ni(NCMe)][OTf]2 (7a,b; Eopx ꢁ +1.2 V), thus reflect-
ing the extreme electron deficiency of PIMIOCOP ligands.
Figure 5. MO diagrams for the complexes in Figure 4 (iPr2P series). Main con-
tributions to the MOs are indicated; m-phen and NHC refer to the central
m-phenylene and the N-heterocyclic carbene moieties, respectively.
PCM-B3PW91/6-31G** level of calculation in the acetonitrile continuum
(e=35.688).
The oxidation potentials of the complexes are correlated
with the character of the HOMOs, which involve mainly p orbi-
tals of the m-phenylene moiety, whereas the HOMOÀn (n=1–
6) present a strong nickel d-orbital character. The relative
energy of these HOMOs is found in qualitative agreement with
the relative oxidation potentials of the complexes. Further-
more, the experimental oxidation potentials correlate linearly
with the energy of the particular HOMOÀn with the highest
contribution of a Ni dz2 orbital (see the Supporting Information,
Figure S9), thus corroborating the assumption that the ob-
served redox events are Ni-based, even in the case of irreversi-
ble oxidations.
Figure 6. Representative frontier MOs of the complexes in Figure 4 (R=iPr).
PCM-B3PW91/6-31G** level of calculation in the acetonitrile continuum
(e=35.688).
gies in the NHCCOP complex, whereas the reverse is observed
in the PIMIOCOP complex. Also noteworthy is the changing
character of the LUMOs from p*z(CN) in the POCOP and
PIMCOP systems to p*(NHC) in the NHCCOP complex and to
p*(PÀC) in the PIMIOCOP complex. In contrast, the nature of
the HOMO is fairly invariable along the series (Figures 5 and
6). The relative order of the HOMOÀn is in line with the dis-
torted square-planar geometry of the nickel center. These MO
features will be valuable for the analysis of experimental cyclic
voltammetry (CV) and electrophilic reactivity results (see
below).
Theoretical analysis of the reactivities of pincer-Ni com-
plexes
Although the oxidation properties of the complexes have been
correlated with the main features of their HOMOs, the electro-
philic reactivities should a priori be correlated with their
LUMOs, the features of which vary considerably with the
pincer ligand (Figures 5 and 6). In POCOP and PIMCOP com-
plexes where the major contribution to the LUMO is the p* or-
bital of the nitrile moiety, interaction with nucleophiles is ex-
pected to weaken the CꢀN bond. In PIMIOCOP complexes in
which the main contribution to the LUMO involves the amidi-
Electrochemical Measurements: Given that the oxidation po-
tential (Eopx) of a complex can reflect the electronic endowment
of its ligands, the title NiII pincer complexes were investigated
by CV to establish the relative electron donation from various
pincer ligands. The Eopx values (Table 3) showed much variation
as a function of the pincer ligand: [(NHCCOP)NiBr] (ca.+
0.5 V)<[(PIMCOP)NiBr] (ca.+0.8 V)<[(PIMIOCOP)NiBr] (Eopx ꢁ +
1.0 V). In the POCOP, PIMCOP, and PIMIOCOP series, passing
Chem. Eur. J. 2015, 21, 17403 – 17414
www.chemeurj.org
17408
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
niophosphine p* orbital (Figure 6),30 interaction with nucleo-
philes is expected to weaken the PÀC bond.
Access to the new set of electrophilic PIMCOP- and PIMIO-
COP-based pincer complexes 6 and 7 allowed us to screen
them for the catalytic hydroamination of nitriles. The bench-
mark reaction of piperidine with benzonitrile was thus studied
under the following conditions: equimolar quantities of piperi-
dine and benzonitrile, no solvent, 508C, and 1 mol% of a cat-
ionic pincer complex (Table 4). For the sake of comparison, we
also screened the catalytic competence of the previously re-
ported cationic adduct [(PhPOCOPPh)Ni(NCMe)]+, 9a.24 Control
experiments demonstrated that no hydroamination occurs in
the absence of a Ni precursor.
The chemical reactivity was first investigated by using ELF
and AIM (AIM=atoms in molecules) topological analyses and
Fukui functions. The results of these studies, including average
ELF populations and AIM atomic charges, are provided and de-
scribed in the Supporting Information (Figure S10). Altogether,
these results indicate that regions most susceptible to attack
by nucleophiles should be the acetonitrile Csp center in POCOP,
PIMCOP, and perhaps NHCCOP systems, whereas in PIMIOCOP-
+
based cations nucleophilic attack should occur at the PÀCN2
region.
Table 4. Catalytic addition of piperidine to benzonitrile.[a]
Catalytic conversion of nitriles to amidines
As dinitrogen analogues of carboxylic acids and esters, ami-
dines exhibit particular biological activities and can be used
in the synthesis of functional heterocycles.[31] Among many
possible routes to amidines,[32] the most straightforward
strategy is the addition of an amine to a nitrile.[33] Such addi-
tions can be promoted by stoichiometric amounts of Lewis
acids,[34] but in most cases such additions require harsh con-
ditions.[35] Alternatively, nitriles can be converted into ami-
dines in a catalytic fashion by using lanthanide(III) ions.[36]
For instance, YbIII–amide complexes have been shown to cat-
alyze the formation of amidines along with pyrimidine and
triazine byproducts.[37,38]
Entry
Catalyst
TON
1
2
3
4
5
7a
7b
6a
6b
13
30
0
8
2
[(PhPOCOPPh)Ni(NCMe)]+ 9a
[a] TON values determined by GC-MS analysis of the final mixtures, using n-
In the course of previous studies on Michael-type hydroal-
koxylation and hydroamination of olefins catalyzed by cat-
ionic POCOP-Ni complexes (Scheme 3), path A),[17b,24,39] it was
discovered that addition of morpholine to 4-cyanostyrene or
cinnamonitrile took place at the nitrile moiety, giving the ami-
dine products with catalytic turnover numbers (TONs) of 25 or
35, respectively (Scheme 3, paths B and C).[13a] The same reac-
tivity was also evident with nitriles lacking an olefinic moiety,
as exemplified by the reaction of morpholine with [(POCOP)-
Ni(NCMe)]+, from which an acetamidine adduct was isolated
(Scheme 3, path D). A report by Arnold and co-workers also
showed that the addition of piperidine to acetonitrile is cata-
lyzed, at ambient temperature and with TONꢁ14, by the
C12H26 as internal standard.
The highest TON values obtained with the PIMIOCOP-based
precursors 7a and 7b were at first attributed to the increased
electrophilic character relative to the PIMCOP analogues 6a
and 6b (see above). Nevertheless, the very low TON obtained
with the PhPOCOPPh-based precursor 9a is at odds with our ex-
pectations of the greater electrophilicity of this diphosphinite
complex. Likewise, the higher catalytic activity observed with
complex 7b vs. 7a does not correlate with the greater electro-
philicity of the latter, which is both anticipated in view of its
less electron-donating P-substituents (Ph2P vs. iPr2P) and also
revealed by the more positive oxidation potential of this spe-
cies.
related
dicationic
“pincer-like”
complex
[kP,kN,kP’-
{(iPr2PCH2CH2)2NH}Ni(NCMe)][BF4]2.[40]
Optimal reaction conditions
were then sought for the hydro-
amination of benzonitrile with
the most competent precursor,
complex 7b (Table 5). We first
found that the TON values de-
clined significantly when the re-
action time was shortened from
67 to 3 h, even when the tem-
perature was raised from 50 to
808C (Table 5, entry 1). Maintain-
ing this temperature and run-
ning the reaction over 21 h re-
sulted in much higher TON
Scheme 3. Hydroamination of nitriles promoted by [(POCOP)Ni(NCMe)]+.
Chem. Eur. J. 2015, 21, 17403 – 17414 www.chemeurj.org
values (Table 5, entry 2 vs.
17409
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
z=288; TON=95), whereas cyclohexylamine gave a mixture of
mono- and disubstituted products (m/z=201 and 284, respec-
tively; total TON=41). The disubstituted amidines result from
a second nucleophilic attack on the coordinated amidine and
subsequent elimination of ammonia; it is worth noting that
this step occurs in an uncatalyzed manner.[38] Finally, PhCN re-
acted with neither weakly nucleophilic aromatic amines, such
as aniline and diphenylamine, nor bulky aliphatic amines, such
as iPr2NH.
Table 5. Conditions for the preparation of piperidinobenzamidine from
benzonitrile and piperidine using 7b as catalyst precursor.[a]
Entry Cat. loading T [8C] t [h] PhCN/piperidine Yield [%] TON
[mol%]
Mechanistic issues
1
2
3
4
5
6
1
1
1
1
5
80
80
80
100
80
80
3
21
21
21
21
21
1:1
1:1
1:2
1:2
1:2
1:2
13
49
68
72
80
68
13
49
68
72
16
7
Two different mechanistic postulates have been put forth for
the hydroamination of nitriles catalyzed by lanthanide com-
plexes. Forsberg et al. proposed that hydroaminations cata-
lyzed by LnIII ions proceed via an outer-sphere mechanism in-
volving nucleophilic attack of amines on coordinated nitriles.[36]
This proposal appears to be somewhat incongruent with the
fact that amines are more nucleophilic towards lanthanide
ions, and they are present in equimolar or even excess quanti-
ties in these catalytic settings; these considerations call into
question the coordination–activation of nitriles. The authors ra-
tionalize this apparent contradiction by proposing that labile
binding of amines to LnIII allows competitive nitrile binding
and activation. The hydroamination reactions catalyzed by Yb
amides are believed to proceed via a conceptually different,
inner-sphere mechanism involving insertion of R’CꢀN into the
YbÀNR2 bond, followed by aminolysis of the resulting amidi-
nate YbÀN=C(R’)NR2.[37]
10
[a] Yields and TON values determined by GC-MS analysis of the final mix-
tures, using n-C12H26 as internal standard.
Table 4, entry 2 and Table 5, entry 1). Further improvements in
TON resulted from using 2 equivalents of piperidine (Table 5,
entry 3) and raising the reaction temperature to 1008C
(entry 4). Finally, increasing the precursor loading from 1% to
5% resulted in the anticipated yield enhancement (to 80%;
Table 5, entry 5), but a further increase to a 10% loading led to
a yield decrease (entry 6); this unexpected result will be ad-
dressed later.
The reaction scope in terms of compatible amine and nitrile
substrates was investigated next. As shown in Figure 7, benzo-
nitrile was found to be much more reactive than acetonitrile,
and piperidine was found to be a more effective nucleophile
than morpholine (TON=49 vs. 33).
In the search for a mechanism adapted for the Ni-catalyzed
hydroamination of nitriles under discussion, we selected as
guiding concepts the following known reactivities of NiII spe-
cies with nitriles and amines: i) Nitriles coordinate very readily
to the NiII center in cationic species; ii) cationic adducts of the
type [(pincer)Ni(NCMe)]+ promote Michael-type additions of
amines to N-coordinated acrylonitrile derivatives.[13b,39] There-
fore, an outer-sphere mechanism similar to that proposed by
Forsberg et al. for LnIII catalysts (see above)[36] seems plausible
in the system under discussion. This postulate is also support-
ed by Fukui index values (see the Supporting Information, Fig-
ure S10). We have studied the stoichiometric reactions shown
in Scheme 4 to examine whether the hydroamination of nitriles
proceeds by direct attack by amines on the coordination-acti-
vated nitriles. The results are described below.
Primary aliphatic amines led to various mixtures of N-mono-
substituted or N,N’-disubstituted amidines, depending on the
amine/PhCN ratio. For instance, using a 4:1 amine/PhCN ratio
with n-hexylamine led to the disubstituted product only (m/
Addition of excess piperidine to a CD3CN solution of 7b
(Scheme 4, reaction A) led to disappearance of the 31P doublets
of 7b [d= +194.2 and +31.2 ppm; J(P,P)=250 Hz] and ap-
pearance of a broad singlet at d=187–188 ppm, which could
be attributed to the cationic adduct [(NHCCOPiPr)Ni(amidine)]+
(10-CH3 or 10-CD3). Consistent with the in situ formation of an
NHC species, the 13C NMR spectrum of the reaction A mixture
indeed showed a signal at d= +173.9 ppm [J(C,P)ꢁ100 Hz]
characteristic of a carbene center and comparable to the 13C
signal at d=172.9 ppm [J(C,P)=93 Hz] attributed to the car-
bene carbon in 8b. The proposed formation of NHC species
10 was further supported by the observation that adding an
authentic amidine sample to the NHC complex 8b gave the
Figure 7. Products of nitrile hydroamination reactions carried out at 808C
for 21 h (with PhCN) or 5 d (with acetonitrile), in a solvent-free manner
using 1 mol% of the precursor complex 7b, and an amine/RCN ratio of 1:1
(for piperidine and morpholine) or 4:1 (for cyclohexylamine and n-hexyla-
mine). Yields were determined by GC-MS with n-C12H26 as internal standard.
Chem. Eur. J. 2015, 21, 17403 – 17414
www.chemeurj.org
17410
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
the hydroamination reaction in-
volving the PIMIOCOP precata-
lyst 7b revealed no significant
induction period (see the Sup-
porting Information, Figure S11),
indicating that the initial de-
phosphinylation of 7 to 8 is very
rapid.
An intriguing observation
from the above experiments is
that, although the catalytic hy-
droaminations require high tem-
peratures, the stoichiometric
conversion of the cationic nitrile
adduct 8b into the amidine ad-
ducts 10-R proceeds at room
temperature. This suggests that
the hydroamination catalysis is
susceptible to some degree of
product inhibition, which can be
counteracted by heating to
induce substrate uptake and
turnover. This scenario is sup-
ported by the finding that the
Scheme 4. Formation of amidine complexes 10-R’ from cationic acetonitrile adducts 7b and 8b. Triflate counter-
ions are omitted for clarity.
same broad 31P singlet at d=187–188 ppm (Scheme 4, reac-
tion B).
amidine ligand in 10-Ph (Scheme 4) is resistant to substitution
by PhCN: the 31P NMR spectrum of a mixture of 10-Ph with
excess PhCN and piperidine showed no sign of a new adduct
even after it had been heated at 808C for over 20 min. The
amidine adduct is thus the only observable species during the
Ni-catalyzed hydroamination reaction; in other words, this
adduct represents the resting state of the catalysis, as illustrat-
ed in the proposed mechanism shown in Scheme 5.
The hydroamination reaction was then monitored step-by-
step by conducting reaction C (Scheme 4). The 31P NMR spec-
trum of a RT solution of 7b in benzonitrile[41] showed the ap-
pearance of two AX spin systems. One of these resonated at
virtually the same chemical shifts as 7b, thus prompting us to
assign it to the corresponding PhCN adduct formed by dis-
placement of MeCN. The second AX spin system consisted of
doublets centered at d=191.5 and 14.0 ppm [J(P,P)=244 Hz],
which was tentatively assigned to a cationic adduct arising
from the displacement of the MeCN in 7b by residual water.[41]
Subsequent addition of piperidine to the sample (1 equivalent
with respect to PhCN) led to appearance of the broad singlet
at d= +187–188 ppm (assigned to 10-Ph; see above), along
with a sharp singlet at d= +57 ppm. The latter was attributed
to iPr2P(O)H, based on our previous observations that hydrolyt-
ic oxidation of iPr2PCl gives similarly sharp singlets in the 53–
55 ppm region of the 31P NMR spectrum. We postulate that
iPr2P(O)H arises from a rearrangement of iPr2P(OH) generated
by hydrolysis of either the phosphenium moiety in 7b or that
of the decomposition byproduct iPr2P(c-N(CH2)5) resulting from
the nucleophilic attack of piperidine on 7b.
The proposed mechanism also provides a rationale for the
aforementioned lower activity observed with a 10% loading of
the precatalyst 7b. Recall that the in situ transformation of the
PIMIOCOP adduct 7 to its NHCCOP analogue at the outset of
the catalysis would result in the release into the reaction
medium of one equivalent of R2PNR’’2. The presence of this
free ligand in runs using 1% catalyst loadings would have neg-
ligible impact on the course of the catalysis, but it would be
reasonable to expect an inhibition of catalysis during a catalytic
run using a 10% loading of the precatalyst.
Conclusion
To consider its results in broad strokes, this report illustrates
how deeply interactive experimental and theoretical ap-
proaches can help address challenges in the three traditionally
quite delineated fields of coordination chemistry of new com-
plexes, their spectroscopic and structural characterization, and
their use in catalysis. The title complexes were thus extensively
characterized by complementary experimental techniques
(crystallography, spectroscopy, voltammetry) and DFT model-
ing (structure optimization; orbital, ELF and AIM analyses; reac-
tivity Fukui indices calculation), and their reactivities in the hy-
droamination of nitriles were investigated. Beyond their obvi-
ous potential developments (generalization of structures and
The above observations allow us to conclude that the PI-
MIOCOP complexes 7 are dephosphinylated to the NHCCOP
analogues at the outset of the catalysis. This assertion is con-
sistent with both the calculated Fukui indices (see the Support-
ing Information, Figure S10) and previous reports of nucleo-
phile-induced PÀC bond cleavage in PIMIOCOP species and
other imidazoliophosphanes.[18a] As a final argument in support
of the proposed scenario, we found that benzonitrile hydroa-
mination proceeds with similar catalytic activities with the pre-
catalysts 8b and 7b (TON=55 vs. 49). Kinetic monitoring of
Chem. Eur. J. 2015, 21, 17403 – 17414
www.chemeurj.org
17411
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
over numbers (TON=95 with n-hexylamine and 72 with piperi-
dine), but the range of amine substrates active in this reaction
remains quite limited. Future studies will be aimed at improv-
ing our understanding of how metal–pincer bonding is influ-
enced by ligand features (electrostatic effects and structural ar-
chitecture), and what factors govern the electrophilicity of the
metal center.
Acknowledgements
The authors are grateful to: Centre National de la Recherche
Scientifique (CNRS) for financial support in the form of operat-
ing funds and also for a 50% teaching sabbatical to R.C.;
NSERC of Canada for a Discovery Grant to D.Z.; UniversitØ de
MontrØal, Centre in Green Chemistry and Catalysis (CGCC) and
Fonds de recherche du QuØbec: Nature et technologies
(FQRNT) for graduate fellowships to B.V.; Dr. Michel Simard for
his valuable assistance with crystallography. Computational
studies were performed by using HPC resources from CALMIP
(Grant 2013[0851]) and from GENCI-[CINES/IDRIS] (Grant
2013[085008]). The Direction des Relations Internationales of
UniversitØ de MontrØal and UniversitØ Toulouse 3 Paul Sabatier
are gratefully acknowledged for the travel grants that made
this collaborative project possible.
Scheme 5. Proposed mechanism for catalytic nitrile hydroamination using
PIMIOCOP pincer complexes 7 as precatalysts.
tuning of reactivities), these investigations aim to advance our
understanding of the interplay between electronic and electro-
static effects in organometallic coordination chemistry and cat-
alysis, in particular by reference to electrostatics-governed en-
zymatic recognition and transformation processes.
Keywords: electron localization function · Fukui indices ·
hydroamination · nickel · pincer complexes
In more concrete terms, the results described in the present
study serve to highlight the differences in structural and spec-
troscopic properties of PIMCOP, PIMIOCOP, and NHCCOP com-
plexes of NiII and the reactivities of their cationic derivatives in
the hydroamination of nitriles. IR spectroscopic analysis of the
cations based on experimental and computational n(CꢀN)
values has demonstrated how P-substituents and the
positive charge on the imidazoliophosphine moieties exert in-
fluence on Ni-ligand interactions, in the process shaping the
electrophilic character of the complexes. Whereas the n(CꢀN)
values of various cationic adducts followed the counter-intui-
[1] Topics in Organometallic Chemistry, Vol. 40: Organometallic Pincer
Chemistry (Eds.: G. van Koten, D. Milstein), Springer-Verlag, Berlin, Hei-
delberg, 2013.
[2] a) C. J. Moulton, B. L. Shaw, J. Chem. Soc. Dalton Trans. 1976, 1020–
1024; b) G. van Koten, J. T. B. H. Jastrzebski, J. G. Noltes, A. L. Spek, J. C.
Schoone, J. Organomet. Chem. 1978, 148, 233–245; c) G. van Koten, K.
Timmer, J. G. Noltes, A. L. Spek, J. Chem. Soc. Chem. Commun. 1978,
250–252.
[3] a) J. T. Singleton, Tetrahedron 2003, 59, 1837–1857; b) M. E. van der
Boom, D. Milstein, Chem. Rev. 2003, 103, 1759–1792; c) The Chemistry
of Pincer Compounds (Eds.: D. Morales-Morales, C. G. M. Jensen), Elsevier,
Amsterdam, 2011.
[4] a) N. Debono, Y. Canac, C. Duhayon, R. Chauvin, Eur. J. Inorg. Chem.
2008, 2991–2999; b) Y. Canac, N. Debono, L. Vendier, R. Chauvin,
Inorg. Chem. 2009, 48, 5562–5568; c) Y. Canac, N. Debono, C. Lepetit, C.
Duhayon, R. Chauvin, Inorg. Chem. 2011, 50, 10810–10819; d) Y. Canac,
C. Bijani, C. Duhayon, R. Chauvin, Organometallics 2013, 32, 4054–4057.
[5] a) C. Barthes, C. Lepetit, Y. Canac, C. Duhayon, D. Zargarian, R. Chauvin,
Inorg. Chem. 2013, 52, 48–58; b) C. Lepetit, J. Poater, M. Esmail Alikhani,
B. Silvi, Y. Canac, J. Contreras-García, M. Solà, R. Chauvin, Inorg. Chem.
2015, 54, 2960–2969.
[6] a) N. Kuhn, J. Fahl, D. Bläser, R. Boese, Z. Anorg. Allg. Chem. 1999, 625,
729–734; b) M. Azouri, J. Andrieu, M. Picquet, P. Richard, B. Hanquet, I.
Tkatchenko, Eur. J. Inorg. Chem. 2007, 4877–4883; c) Y. Canac, C. Maali-
ki, I. Abdellah, R. Chauvin, New J. Chem. 2012, 36, 17–27; d) S. Gaillard,
J. L. Renaud, Dalton Trans. 2013, 42, 7255–7270; e) M. Alcarazo, Chem.
Eur. J. 2014, 20, 7868–7877.
[7] a) M. Albrecht, G. van Koten, Angew. Chem. Int. Ed. 2001, 40, 3750–
3781; Angew. Chem. 2001, 113, 3866–3898; b) J. P. Cloutier, B. Vabre, B.
Mougang-SoumØ, D. Zargarian, Organometallics 2015, 34, 133–145.
[8] a) A. R. Kennedy, R. J. Cross, K. W. Muir, Inorg. Chim. Acta 1995, 231,
195–200; b) W. T. S. Huck, B. Snellink-Rul, F. C. J. M. van Veggel, D. N.
Reinhoudt, Organometallics 1997, 16, 4287–4291; c) K. A. Kozhanov,
M. P. Bubnov, V. K. Cherkasov, G. K. Fukin, G. A. Abakumov, Chem.
Commun. 2003, 2610–2611; d) J. Cµmpora, P. Palma, D. del Río, E. lvar-
tive order NHCCOPPh
>
PhPOCOPPh >NHCCOPiPr
>
iPrPOCOPiPr >
PhPIMIOCOPPh PhPIMCOPPh
>
>
iPrPIMIOCOPiPr and iPrPIMCOPiPr, the
oxidation potentials of both charge-neutral bromo derivatives
and cationic adducts followed the intuitively anticipated order
PIMIOCOP>PIMCOP>NHCCOP.
In terms of their catalytic activities in hydroamination of
benzonitrile, the cationic precatalysts [(PIMIOCOP)Ni(NCCH3)]
[OTf]2 proved to be significantly more active than their POCOP
and PIMCOP analogues. Interestingly, the more electrophilic
Ph2P complexes proved to be less active than their iPr2P coun-
terparts. Moreover, it was found that the PIMIOCOP complexes
are readily converted into NHCCOP analogues during the catal-
ysis, implying that such in situ-generated NHCCOP cationic
species have superior catalytic activities in spite of their lower
electrophilicities. This unexpected observation can be attribut-
ed to the greater thermal stability of the reaction intermedi-
ates bearing an NHC moiety. Optimization studies allowed the
formation of benzamidines with relatively high catalytic turn-
Chem. Eur. J. 2015, 21, 17403 – 17414
www.chemeurj.org
17412
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ez, Organometallics 2004, 23, 1652–1655; e) M. E. van der Boom, S.-Y.
Liou, L. J. W. Shimon, Y. Ben-David, D. Milstein, Inorg. Chim. Acta 2004,
357, 4015–4023; f) L. F. Groux, F. BØlanger-GariØpy, D. Zargarian, Can. J.
Chem. 2005, 83, 634–639; g) A. Castonguay, C. Sui-Seng, D. Zargarian,
A. L. Beauchamp, Organometallics 2006, 25, 602–608; h) A. Castonguay,
A. L. Beauchamp, D. Zargarian, Inorg. Chem. 2009, 48, 3177–3184; i) B. J.
Boro, E. N. Duesler, K. I. Goldberg, R. A. Kemp, Inorg. Chem. 2009, 48,
5081–5087; j) T. J. Schmeier, N. Hazari, C. D. Incarvito, J. A. Raskatov,
Chem. Commun. 2011, 47, 1824–1826; k) V. A. Levina, A. Rossin, N. V.
Belkova, M. R. Chierotti, L. M. Epstein, O. A. Filippov, R. Gobetto, L. Gon-
salvi, A. Lledós, E. S. Shubina, F. Zanobini, M. Peruzzini, Angew. Chem.
Int. Ed. 2011, 50, 1367–1370; Angew. Chem. 2011, 123, 1403–1406.
[9] a) V. Gómez-Benítez, O. Baldovino-Pantaleón, C. Herrera-lvarez, R. A.
Toscano, D. Morales-Morales, Tetrahedron Lett. 2006, 47, 5059–5062;
b) S. Chakraborty, J. Zhang, J. A. Krause, H. Guan, J. Am. Chem. Soc.
2010, 132, 8872–8873; c) B. Vabre, F. Lindeperg, D. Zargarian, Green
Chem. 2013, 15, 3188–3194; d) B. Vabre, P. Petiot, R. Declercq, D. Zar-
garian, Organometallics 2014, 33, 5173–5184; e) J. Hao, B. Mougang-
SoumØ, B. Vabre, D. Zargarian, Angew. Chem. Int. Ed. 2014, 53, 3218–
3222; Angew. Chem. 2014, 126, 3282–3286; f) J. Hao, B. Vabre, B. Mou-
gang-SoumØ, D. Zargarian, Chem. Eur. J. 2014, 20, 12544–12552; g) J.
Hao, B. Vabre, D. Zargarian, Organometallics 2014, 33, 6568–6576.
[10] a) D. M. Spasyuk, D. Zargarian, d. E. A. van, Organometallics 2009, 28,
6531–6540; b) D. M. Spasyuk, D. Zargarian, Inorg. Chem. 2010, 49,
6203–6213; c) B. S. Zhang, W. Wang, D. D. Shao, X. Q. Hao, J. F. Gong,
M. P. Song, Organometallics 2010, 29, 2579–2587; d) J. L. Niu, Q. T.
Chen, X. Q. Hao, Q. X. Zhao, J. F. Gong, M. P. Song, Organometallics
2010, 29, 2148–2156; e) D. M. Spasyuk, S. I. Gorelsky, d. E. A. van, D. Zar-
garian, Inorg. Chem. 2011, 50, 2661–2674; f) M. J. Yang, Y. J. Liu, J. F.
Gong, M. P. Song, Organometallics 2011, 30, 3793–3803; g) J. Sanford,
C. Dent, J. D. Masuda, A. Xia, Polyhedron 2011, 30, 1091–1094; h) B.
Mougang-SoumØ, F. Belanger-GariØpy, D. Zargarian, Organometallics
2014, 33, 5990–6002.
[23] a) A. Liu, X. Zhange, W. Chen, Organometallics 2009, 28, 4868–4871;
b) S. Hameury, P. de Fremont, P. A. R. Breuil, H. Olivier-Bourbigou, P.
Braunstein, Inorg. Chem. 2014, 53, 5189–5200.
[24] A. B. Salah, C. Offenstein, D. Zargarian, Organometallics 2011, 30, 5352–
5364.
[25] Since nitriles are primarily s-donor ligands and their lone pairs have
slightly antibonding character with respect to the CÀN bond,[26] the
n(CꢀN) values are expected to reflect the degree of donation to the
metal center, greater values implying greater degree of donation (and
by extension, greater electrophilic character of the metal fragment).
[26] a) W. E. Buschmann, J. S. Miller, Chem. Eur. J. 1998, 4, 1731–1737; b) S.
Liang, H. Wang, T. Deb, J. L. Petersen, G. T. Yee, M. P. Jensen, Inorg.
Chem. 2012, 51, 12707–12719; c) J. H. Choi, K. I. Oh, H. Lee, C. Lee, M.
Lee, J. Chem. Phys. 2008, 128, 134506–134508.
[27] a) M. D. Watson, X. S. Gai, A. T. Gillies, S. H. Brewer, E. E. Fenlon, J. Phys.
Chem. B 2008, 112, 13188–13192; b) B. A. Lindquist, K. E. Furse, S. A.
Corcelli, Phys. Chem. Chem. Phys. 2009, 11, 8119–8132; c) K. H. Park, J.
Jeon, Y. Park, S. Lee, H. J. Kwon, C. Joo, S. Park, H. Han, M. Cho, J. Phys.
Chem. Lett. 2013, 4, 2105–2110.
[28] In contrast to what was stated in the case of the acetonitrile adducts, it
is understood that the NiÀCO bond is dominated by Ni!CO p-back-
bonding interactions. For a representative example of a study discus-
sing the use of calculated n(CO) values for classification of monoden-
tate spectator ligands in terms of net electron-donation character, see:
D. G. Gusev, Organometallics 2009, 28, 763–770.
[29] Our analysis of the IR spectra of acetonitrile and CO adducts discussed
here is based on the commonly accepted notion that significant M!
ligand p-backbonding (usually dominant in CO complexes but weak in
acetonitrile complexes) leads to a lowering of the stretching frequency
of the ligand CÀX multiple bond, whereas an increase in this frequency
is seen with significant ligand!M s-donation (usually dominant in ace-
tonitrile complexes but weak in CO complexes). An alternative analysis
based on the greater importance of electrostatic effects in determining
n(CO) values, especially in cationic species, has been reported: A. S.
Goldman, K. Krogh-Jespersen, J. Am. Chem. Soc. 1996, 118, 12159–
12166.
[30] Previous studies have shown that the amidiniophosphine moiety in-
volves a bonding interaction between the HOMO of the NHC donor
(nonbonding s-NHC) and the 3pp orbital of the phosphenium acceptor
in addition to the antibonding interaction of the HOMO of the phos-
phenium (pp) with the LUMO of the NHC (p*). For a detailed study of
this analysis see references [18a] and [18d].
[31] a) F. Pietra, Chem. Biodiversity 2012, 9, 331–351; b) J. Kotthaus, T. Stein-
metzer, A. van de Locht, B. Clement, J. Enzyme Inhib. Med. Chem. 2011,
26, 115–122; c) H. Oediger, F. Mçller, K. Eiter, Synthesis 1972, 1972, 591–
598; d) J. E. Taylor, S. D. Bull, J. M. J. Williams, Chem. Soc. Rev. 2012, 41,
2109–2121; e) J. Y. Quek, T. P. Davis, A. B. Lowe, Chem. Soc. Rev. 2013,
42, 7326–7334.
[32] a) A. R. Katritzky, C. Cai, S. K. Singh, J. Org. Chem. 2006, 71, 3375–3380;
b) A. A. Aly, A. M. Nour-El-Din, ARKIVOC (Gainesville, FL, U.S.) 2008, 153–
194; c) J. Sävmarker, J. Rydfjord, J. Gising, L. R. Odell, M. Larhed, Org.
Lett. 2012, 14, 2394–2397; d) M. A. McGowan, C. Z. McAvoy, S. L. Buch-
wald, Org. Lett. 2012, 14, 3800–3803.
[33] It has long been known that coordination to various Lewis acids (LA) or
electrophilic transition metal ions (Mn+) can activate nitriles toward nu-
cleophilic attack by amines, a reaction that generates amidines. See for
example: a) P. Oxley, M. W. Partridge, W. F. Short, J. Chem. Soc. 1947,
1110–1116; b) G. Rouschias, G. Wilkinson, J. Chem. Soc. A. 1968, 489–
496; c) P. Uguagliati, U. Belluco, R. A. Michelin, P. Guerriero, Inorg. Chim.
Acta 1984, 81, 61–67; d) L. Maresca, G. Natile, F. P. Intini, F. Gasparrini,
M. Tiripicchio-Camellini, J. Am. Chem. Soc. 1986, 108, 1180–1185; e) J. E.
Bercaw, D. L. Davies, P. T. Wolczanski, Organometallics 1986, 5, 443–450.
[34] G. Rousselet, P. Capdevielle, M. Maumy, Tetrahedron Lett. 1993, 34,
6395–6398.
[11] For a recent review on ECE-type pincer complexes of nickel see: D. Zar-
garian, A. Castonguay, D. M. Spasyuk in Topics in Organometallic Chemis-
try, Vol. 40: Organometallic Pincer Chemistry (Eds.: G. van Koten, D. Mil-
stein), Springer-Verlag, Berlin, Heidelberg, 2013, pp. 131–173.
[12] B. Vabre, Y. Canac, C. Duhayon, R. Chauvin, D. Zargarian, Chem.
Commun. 2012, 48, 10446–10448.
[13] a) X. Lef›vre, G. Durieux, S. Lesturgez, D. Zargarian, J. Mol. Catal. A
2011, 335, 1–7; b) A. Castonguay, D. M. Spasyuk, N. Madern, A. L. Beau-
champ, D. Zargarian, Organometallics 2009, 28, 2134–2141.
[14] D. A. Pratt, R. P. Pesavento, W. A. van der Donk, Org. Lett. 2005, 7, 2735–
2738.
[15] B. Vabre, D. M. Spasyuk, D. Zargarian, Organometallics 2012, 31, 8561–
8570.
[16] This step possibly occurs via an electrophilic mechanism. For a discus-
sion of this issue see: B. Vabre, M. L. Lambert, A. Petit, D. H. Ess, D. Zar-
garian, Organometallics 2012, 31, 6041–6053.
[17] a) V. Pandarus, D. Zargarian, Chem. Commun. 2007, 978–980; b) V. Pan-
darus, D. Zargarian, Organometallics 2007, 26, 4321–4334; c) V. Panda-
rus, A. Castonguay, D. Zargarian, Dalton Trans. 2008, 4756–4761; d) S.
Lapointe, B. Vabre, D. Zargarian, Organometallics 2015, 34, 3520–3531.
[18] a) I. Abdellah, C. Lepetit, Y. Canac, C. Duhayon, R. Chauvin, Chem. Eur. J.
2010, 16, 13095–13108; b) I. Abdellah, M. Boggio-Pasqua, Y. Canac, C.
Lepetit, C. Duhayon, R. Chauvin, Chem. Eur. J. 2011, 17, 5110–5115; c) C.
Maaliki, C. Lepetit, C. Duhayon, Y. Canac, R. Chauvin, Chem. Eur. J. 2012,
18, 16153–16160; d) B. D. Ellis, P. J. Ragogna, C. L. B. Macdonald, Inorg.
Chem. 2004, 43, 7857–7867.
[19] In the 31P NMR spectrum, a new resonance appeared at dp = +124 (s)
ppm which was attributed to the formation of Ph2POEt (See Figure S1).
[20] CCDC 1012838 (3b), 1012839 (3c), 1012840 (5b), 1012837 (6b), and
1013472 (8a) contain the supplementary crystallographic data for this
paper. These data are provided free of charge by The Cambridge Crys-
tallographic Data Centre (See also Tables S1 and S2 in the Supporting
Information).
[35] a) R. S. Garigipati, Tetrahedron Lett. 1990, 31, 1969–1972; b) R. A. Moss,
W. Ma, D. C. Merrer, S. Xue, Tetrahedron Lett. 1995, 36, 8761–8764.
[36] J. H. Forsberg, V. T. Spaziano, T. M. Balasubramanian, G. K. Liu, S. A. Kins-
ley, C. A. Duckworth, J. J. Poteruca, P. S. Brown, J. L. Miller, J. Org. Chem.
1987, 52, 1017–1021.
[21] A. Salah, D. Zargarian, Dalton Trans. 2011, 40, 8977–8985.
[22] X. Lef›vre, D. M. Spasyuk, D. Zargarian, J. Organomet. Chem. 2011, 696,
864–870.
[37] J. Wang, F. Xu, Q. Shen, T. Cai, Org. Lett. 2008, 10, 445–448.
Chem. Eur. J. 2015, 21, 17403 – 17414
www.chemeurj.org
17413
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[38] For a more comprehensive list of reports on hydroamination of nitriles
see reference [37] and the references cited therein.
[41] Note that the reagents used in these experiments are of the same qual-
ity as those used in the catalytic runs and might thus contain some re-
sidual moisture.
[39] a) C. Sui-Seng, A. Castonguay, Y. Chen, D. Gareau, L. F. Groux, D. Zargari-
an, Top. Catal. 2006, 37, 81–90; b) A. Castonguay, A. L. Beauchamp, D.
Zargarian, Organometallics 2008, 27, 5723–5732.
Received: June 26, 2015
Published online on October 9, 2015
[40] S. S. Rozenel, J. B. Kerr, J. Arnold, Dalton Trans. 2011, 40, 10397–10405.
Chem. Eur. J. 2015, 21, 17403 – 17414
www.chemeurj.org
17414
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim