Imidazoliophosphines are true N-heterocyclic carbene (NHC)-phosphenium adducts

Article abstract of DOI:10.1002/chem.201001721

Whereas the external nucleophilic reactivity of α-amidiniophosphines has been previously illustrated by their complexation to transition-metal centers, their internal electrophilic reactivity is herein investigated by using BIMIONAP (BIMIONAP=N-methylated BIMINAP cation, BIMINAP=formal contraction of the acronyms BIMIP=2,2′-bis(diphenylphosphino)-1,1′-bibenzimidazole and BINAP=2,2′-bis(diphenylphosphino)-1,1′-binaphthyl). Reaction of tetraethylammonium chloride with free BIMIONAP is found to induce heterolytic cleavage of the N₂C-P bond to give chlorodiphenylphosphine and a transient phosphine-N-heterocyclic carbene (NHC) species that is trapped in situ by protonation to the corresponding phosphine-benzimidazolium cation. When the chloride anion reacts with the cationic [Pd(η²-BIMIONAP)Cl ₂] complex, the same cleavage occurs and the phosphine-NHC moiety is trapped in the corresponding [PdCl₂(η²-phosphine-NHC)] complex. When the chloride anion reacts with the dicationic [Pd(π-allyl)(η²-BIMIONAP)]⁺ complex, allyldiphenylphosphine is produced, and the [PdCl(η²-phosphine- NHC)(PPh₂CH₂CH=CH₂)]⁺ complex is obtained. Reaction of free BIMIONAP with the harder n-butyllithium nucleophile also induces heterolytic cleavage of the N₂C-P bond, from which the phosphine-NHC moiety is trapped by hydrolysis of the benzimidazole ring or by P,C-sulfurization. Cleavage of a C-P bond with the weak Cl^- nucleophile to release the reactive NHC moiety (according to the unusual scheme C-P+Cl^-→C:+Cl-P) is a definite experimental indication of the dative nature of the N₂C-P bond of amidiniophosphines, which are, therefore, better described as NHC→phosphenium adducts. This interpretation is supported by the calculation, at the DFT level, of a heterolytic dissociation mode of the N₂C-P bond lower in energy than the homolytic one. A mesomeric description of the NHC→phosphenium entity is also proposed on the basis of electron localization function (ELF) and atoms in molecules (AIM) analyses. Finally ELF and AIM-based Fukui indices, molecular orbitals, and MESP analyses show that the initial attack of Cl^- takes place at the carbenic atom of BIMIONAP. Addition of anionic nucleophiles (Cl^-, nBu^-) to free BIMIONAP and BIMIONAP-containing palladium complexes results in selective cleavage of the N₂C-P bond, from which the released N-heterocyclic carbene (NHC) fragment can be trapped by protonation, hydrolysis, sulfurization, or coordination to Pd^II centers (see scheme).

Full text of DOI:10.1002/chem.201001721

FULL PAPER
DOI: 10.1002/chem.201001721
Imidazoliophosphines are True N-Heterocyclic Carbene (NHC)–
Phosphenium Adducts
Ibrahim Abdellah,[a, b] Christine Lepetit,[a, b] Yves Canac,*[a, b] Carine Duhayon,[a, b] and
Remi Chauvin*[a, b]
Abstract: Whereas the external nucleo-
philic reactivity of a-amidiniophos-
phines has been previously illustrated
by their complexation to transition-
metal centers, their internal electro-
philic reactivity is herein investigated
by using BIMIONAP (BIMIONAP=
N-methylated BIMINAP cation, BIMI-
NAP =formal contraction of the acro-
nyms BIMIP=2,2’-bis(diphenylphos-
complex, the same cleavage occurs and
the phosphine–NHC moiety is trapped
in the corresponding [PdCl2(h2-phos-
phine–NHC)] complex. When the chlo-
ride anion reacts with the dicationic
Clꢀ nucleophile to release the reactive
NHC moiety (according to the unusual
ꢀ
ꢀ
ꢀ
scheme C P+Cl !C:+Cl P) is
a
definite experimental indication of the
ꢀ
dative nature of the N2C P bond of
[Pd
N
com-
amidiniophosphines, which are, there-
fore, better described as NHC!phos-
phenium adducts. This interpretation is
supported by the calculation, at the
DFT level, of a heterolytic dissociation
plex, allyldiphenylphosphine is pro-
duced, and the [PdCl(h2-phosphine–
NHC)(PPh2CH2CH=CH2)]+ complex
is obtained. Reaction of free BIMIO-
NAP with the harder n-butyllithium
nucleophile also induces heterolytic
ꢀ
phino)-1,1’-bibenzimidazole
and
mode of the N2C P bond lower in
BINAP=2,2’-bis(diphenylphosphino)-
1,1’-binaphthyl). Reaction of tetraethyl-
ammonium chloride with free BIMIO-
NAP is found to induce heterolytic
energy than the homolytic one.
A
ꢀ
cleavage of the N2C P bond, from
mesomeric description of the NHC!
phosphenium entity is also proposed
on the basis of electron localization
function (ELF) and atoms in molecules
(AIM) analyses. Finally ELF and AIM-
based Fukui indices, molecular orbitals,
and MESP analyses show that the ini-
tial attack of Clꢀ takes place at the car-
benic atom of BIMIONAP.
which the phosphine–NHC moiety is
trapped by hydrolysis of the benzimi-
dazole ring or by P,C-sulfurization.
ꢀ
cleavage of the N2C P bond to give
ꢀ
chlorodiphenylphosphine and a transi-
ent phosphine–N-heterocyclic carbene
(NHC) species that is trapped in situ
by protonation to the corresponding
Cleavage of a C P bond with the weak
Keywords: dative bonding · elec-
trostatic interactions
· N-hetero-
phosphine–benzimidazolium
cation.
cyclic carbenes · palladium · phos-
phenium
When the chloride anion reacts with
the cationic [Pd(h2-BIMIONAP)Cl2]
Introduction
In current chemical drawing, the dative symbol D!A is for-
mally equivalent to the a-zwitterionic Lewis symbol D+–
Aꢀ, for which the dash “–” denotes a shared electron pair.
Nevertheless, the D!A symbol features more than topolog-
ical information, and should be reserved for demonstrated
relevant cases. Indeed, the dative nature of a covalent bond
[a] I. Abdellah, Dr. C. Lepetit, Dr. Y. Canac, C. Duhayon,
Prof. R. Chauvin
CNRS, LCC (Laboratoire de Chimie de Coordination)
205, route de Narbonne, 31077 Toulouse (France)
Fax : (+33)561-553-003
E-mail: yves.canac@lcc-toulouse.fr
chauvin@lcc-toulouse.fr
ꢀ
A B is determined by an energetical criterion, namely, by
+
ꢀ
its preference for a heterolytic dissociation mode A B!A
+jBꢀ versus the corresponding homolytic mode AꢀB!
[b] I. Abdellah, Dr. C. Lepetit, Dr. Y. Canac, C. Duhayon,
Prof. R. Chauvin
C
C [1]
A +B . The difference in energy between the two modes
Universitꢀ de Toulouse, UPS, INP, LCC
31077 Toulouse (France)
C
C
of dissociation is DD=IP(A )ꢀEA(B ), in which IP and EA
denote the ionization potential and the electron affinity, re-
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201001721.
C
C
spectively. If A and B are neutral, the homolytic rupture is
Chem. Eur. J. 2010, 16, 13095 – 13108
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
13095

generally favored because of the destabilizing electrostatic
term of the heterolytic rupture (for neutral species indeed,
In the cationic series, phospheniums R2P:+, which are iso-
lobal and isovalent of singlet carbenes, formally contain a
sp2-hybridized PIII atom with a lone pair of electrons in a s-
type orbital and an empty pp-type orbital. Their Lewis am-
photeric character make them ideal candidates to act simul-
taneously as acceptors and as donors (Scheme 1a). Beyond
the stable free diaminophospheniums,[14] phosphenium cat-
ions with alkyl and aryl substituents have been a priori re-
garded as “D!phosphenium adducts” with different donor
groups D, such as phosphines, amines, or carbenes.[15] In the
phosphine–phosphenium series, the dative nature of the in-
teraction has been demonstrated by the reactivity criterion,
namely, by facile ligand exchange at the phosphenium
center.[16] In the NHC–phosphenium series, few theoretical
studies in the gas phase[17] and X-ray diffraction analysis in
the crystal state[18] supported the relevance of this appella-
tion. Any convincing reactivity result is, however, missing,
and the same cations were more prudently designated by
C
C
IP@EA in order of magnitude). If A or B is charged (e.g.
C
if B is a cation), the heterolytic dissociation leads to at least
one neutral closed-shell species, and is thus a priori lower in
energy due to the vanishing of the electrostatic term. Al-
though the minimum-energy dissociation mode in the gas
phase can be investigated by DFT or ab initio calculations,
experimental indications are provided by intrinsic structural
features of the bond: high polarity, great length, and low
ꢀ
strength. In particular, the low strength of an A B dative
bond and the accepting character of the A moiety are exper-
imentally revealed by the displacement of an intrinsically re-
active B entity with a relatively stable B’ counterpart.[2] This
kind of reactivity defines the field of coordination chemistry,
in which the acceptor component A (a Lewis acid) is a met-
allic center M, and the donor component (a Lewis base) is a
neutral ligand L. Assignment of the donor–acceptor charac-
ter may, however, require refining when M and L are am-
other authors as “amidiniophosphines” with the formula
+
photeric and the M L interaction is not a pure simple cova-
Im+ R2P instead of N2C!PR2 (in which Im+ denotes a
ꢀ
ꢀ
lent bond, that is, it involves more than two electrons. The
coordination of carbonyl and corresponding Fisher carbenes
are paradigms of this situation. In contrast, diaminals there-
of, namely cyclic diamino-carbenes (N-heterocyclic carbenes
(NHCs)), are currently recognized as pure two-electron do-
nating neutral ligands,[3] but selective displacement of coor-
dinated NHCs from transition-metal centers by other nucle-
ophiles has not been exemplified to the best of our knowl-
edge: experimental evidence of the dative nature of the
NHC!M bond is thus missing. However, NHCs may also
be regarded as “stabilizing ligands” of highly reactive non-
metallic electrodeficient species with low coordinence. Sta-
bilization of p-block elements in the zero-oxidation state by
N-aryl,N’-alkyl-imidazolio substituent, Scheme 1a). The
same questions apply in the metal–P-coordinated series
(Scheme 1b).[19]
Following pioneering reports of monodentate amidinio-
phosphine ligands,[20] we recently described the BIMIONAP
(BIMIONAP=N-methylated BIMINAP cation, BIMI-
NAP =formal contraction of the acronyms BIMIP=2,2’-
bis(diphenylphosphino)-1,1’-bibenzimidazole and BINAP=
2,2’-bis(diphenylphosphino)-1,1’-binaphthyl) chelate version
1, combining an amidiniophosphine end with a neutral phos-
phine end.[18a] In the palladium series, BIMIONAP was
shown to act as a cis-chelating ligand in a seven-membered
metallacycle. It was observed that this cationic ligand is
much less electron-donating than the BIMINAP neutral an-
alogue and exhibits specific catalytic properties.[18a] In the
NHCs was recently reported for the :Si=Si:,[4] :Ge=Ge:,[5]
[6]
:P P:, Pn,[7] and C1 fragments.[8] Additionally, various oxi-
ꢀ
dation states encountered in germylenes GeII,[9] silylenes
SiII,[10] borenium BIII,[11] or tellerium TeII[12] cations have also
been reported to be efficiently stabilized by NHC ligands.
By extension, even tetraaminoethenes could be interpreted
as the ultimate products of primary donor–acceptor adducts
between the filled sp2 orbital of one diaminocarbene and
the formally empty p orbital of the other.[13]
following report, the nature of the N2C PR2 bond of BI-
ꢀ
MIONAP is investigated through both the reactivity criteri-
on by using the chloride ion as a relatively “stable ” nucleo-
phile (relative to NHC) and theoretical analyses by using
the electron localization function (ELF).
Results and Discussion
Displacement of a NHC unit from the free phosphenium
end of BIMIONAP (1) by a chloride ion is expected to
afford chlorodiphenylphosphine (Ph2PCl), which can be
easily identified by 31P NMR spectroscopy. The relevance of
such a transformation in the field of coordination chemistry
is a priori supported by evidence for the dative character of
P–halogen bonds.[2d–e] The concommitent generation of the
highly reactive free NHC unit could be indicated by the
final presence of the corresponding iminium resulting from
hydrolysis (Scheme 2a). Hydrolytic trapping is, however, dif-
ficult to control in two steps and evidence of the putative
transient NHC is thus also sought for by an intramolecular
trapping strategy from palladium(II) complexes of BIMIO-
Scheme 1. Lewis representation of amidiniophosphines (left) versus
dative representation of the corresponding NHC–phospheniums (right):
a) in the free state; b) in the metal–P-coordinated state.
13096
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13095 – 13108

N-Heterocyclic Carbene (NHC)–Phosphenium Adducts
FULL PAPER
the 13C NMR signal at dC =+
170.5 ppm characteristic of the
metal–carbenic NCN carbon
atom of 3, which is strongly de-
shielded with respect to the cor-
responding NCN carbon atom
of
29.4 Hz). The exact structure of
complex was finally con-
2 (dC =+146.1 ppm, JPC =
3
firmed by X-ray diffraction
analysis of yellow crystals de-
posited from a CH2Cl2/MeOH/
Et2O mixture at room tempera-
ture (Figure 1).[22]
The palladium(II) center
adopts a classical square-planar
geometry within a pseudo-boat-
Scheme 2. Characterization of the possibly displaced NHC donor upon addition of an anionic nucleophile to
the BIMIONAP ligand 1: a) intermolecular trapping by hydrolysis of free BIMIONAP; b) intramolecular trap-
ping strategy of the P,P-coordinated BIMIONAP (M=Pd).
NAP, aiming at obtaining the corresponding NHC–phos-
phine complex (Scheme 2b).[21]
The results are then confronted with theoretical analyses.
In particular, the relative weights of the imidazoliophos-
phine and NHC–phosphenium resonance forms depicted in
Scheme 1 are targeted through ELF analysis.
ꢀ
Experimental reactivity of the N2C P bond of BIMIONAP:
The reactivity of a chloride ion is first envisaged from palla-
dium complexes of BIMIONAP. The case of free BIMIO-
NAP is then addressed and generalized by using the more
nucleophilic n-butyl anion.
Reactivity of palladium(II) complexes of BIMIONAP with a
chloride ion: The BIMIONAP–PdCl2 complex 2 was pre-
pared by addition of a stoichiometric amount of [Pd-
Figure 1. ORTEP view of the X-ray crystal structure of the NHC–phos-
phine palladium complex 3 (Scheme 3), with thermal ellipsoids drawn at
the 30% probability level (for clarity, H atoms are omitted). Selected
ACHTUNGTRENNUNG(CH3CN)2Cl2] to BIMIONAP (1) according to the previous-
ly described procedure.[18a] Upon addition of tetraethylam-
monium chloride (TEACl) to an acetonitrile solution of 2 at
room temperature, no reaction took place. At 508C, howev-
er, a slow degradation of 2 was observed. Monitoring the re-
action by 31P NMR spectroscopy indicated the complete dis-
appearance of the signals of 2 (dp =+18.8, +24.0 ppm) after
3 h and the appearance of two new signals at dp =+23.2 and
+80.0 ppm. The latter chemical shift is characteristic of
ClPPh2, whereas the former could be consistently attributed
to the NHC–phosphine palladium complex 3 (Scheme 3).
Complex 3 was finally isolated in 62% yield from 2 and
fully characterized. The essential information is provided by
ꢀ
ꢀ
bond distances [ꢂ] and angles [8]: C1 Pd1: 1.9811(17), P1 Pd1:
ꢀ
ꢀ
ꢀ
ꢀ
2.2405(4), N1 C1: 1.344(2), N2 C1: 1.370(2), Pd1 Cl1: 2.3497(5), Pd1
Cl2: 2.3324(5); C1-Pd1-Cl1: 92.39(5), P1-Pd1-Cl2: 97.011(18), N1-C1-N2:
106.72(15).
shaped six-membered palladacycle. The metal–ligand bond
lengths are comparable to those of related NHC–phos-
phine–PdCl2 complexes.[23] The value of the dihedral angle
between the planes of the naphthyl and benzimidazolyl
rings (C18-C9-N2-C1: 45.338) is smaller than in the recently
reported NHC–ylide palladium complex homologue
(64.108), for which the ligand backbone is involved in a less
strained seven-membered palladacycle.[24]
The formation of the NHC–phosphine–PdCl2 complex 3
from 2 formally results from the nucleophilic displacement
of the NHC fragment by the chloride anion at the phosphe-
nium center. This transformation may a priori proceed
through three possible mechanisms, depending on the site of
initial nucleophilic attack of complex 2, that is, the Pd, P1,
or C1 atom (Scheme 4). Initial attack at Pd could be facili-
Scheme 3. Formation of the NHC–phosphine palladium complex 3 from
the BIMIONAP complex 2.
Chem. Eur. J. 2010, 16, 13095 – 13108
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13097

Y. Canac, R. Chauvin et al.
tated by the accessible planar geometry and by the vacancy
of the 16-electron valence shell of 2. The palladate inter-
mediate could then undergo migration of one of the chlo-
ride ligands to the P1 atom with subsequent or simultaneous
Scheme 9). Alternatively, direct attack at P1 could thus be
driven by the strong P+ character of this center to give a
MacGregor-type pentacoordinate intermediate (Scheme 4a-
P).[25a] However, initial attack at P1 remains a priori disfa-
vored because the atom formally satisfies an octet electron
count (with no empty low-lying valence orbital) and because
of the steric demand at this tetrahedral center. Moreover,
both attacks at P1 or Pd lead to N+/Pdꢀ g-zwitterionic palla-
date species, which are exemplified but still quite energe-
tic.[25b]
ꢀ
cleavage of the P1 C1 bond, as suggested by MacGegor
(Scheme 4a-Pd).[25a] Nevertheless, attack at Pd would not be
guided by any positive formal charge at the Pd center of 2:
The Cl ligands are covalently bonded to Pd, and the P1 li-
gands even bring some formal negative charge onto the Pd
center as the P!Pd dative bond is equivalent to the a-zwit-
ꢀ
terionic Lewis structure P+ Pd . The overall partial charge
In contrast, initial attack at the electrodeficient C1 center
(see the section entitled: ELF and AIM-based mesomeric
description of BIMIONAP and Scheme 9) instantly reduces
the +/ꢀ charge separation and can be followed by a classi-
cal a-elimination process at a
ꢀ
at Pd is indeed calculated to be significantly smaller than
the partial charges at P1 (see the section entitled: ELF and
AIM-based mesomeric description of BIMIONAP and
chlorinated sp3 carbon center.
The corresponding likely mech-
anism is shown in Scheme 4a-C
(further argumentation is given
in the section entitled: ELF and
AIM-based mesomeric descrip-
tion of BIMIONAP).
Whatever the mechanism is,
ꢀ
however, cleavage of a P C
bond by a chloride ion selec-
tively bonding to the P atom is
rather unusual and is definitive-
ly indicative of a dative car-
bene!phosphenium bond in
the amidiniophosphine complex
2.
Treatment with free chloride
ions was then envisaged from
the homologous BIMIONAP–
PdACTHNUTRGENUG(N p-allyl) complex. This com-
plex was first targeted by reac-
tion of BIMIONAP (1) with a
stoichiometric amount of [Pd-
AHCTUNGRTEN(NGUN p-allyl)Cl]2, but all attempted
experimental conditions (sol-
vent, temperature) lead to non-
selective decomposition prod-
ucts of BIMIONAP. By follow-
ing the preceeding observations,
this result could be interpreted
by the dissociation of free chlo-
ride from the PdCl center,
which would instantaneously
result in the decomposition of
free BIMIONAP (see the sec-
tion entitled: Reactivity of the
free BIMIONAP cation with
chloride and n-butyllithium
anions) or of the putative [Pd-
Scheme 4. Possible mechanisms for the formation of NHC–phosphine palladium complexes: a) formation of
the PdCl2 complex 3 from the cationic complex 2. A priori possible mechanisms: a-P) and a-Pd) starting with
attack of Clꢀ at the P1 or Pd atoms and ending with a MacGregor-type rearrangement, and likely mechanism
a-C initiated by attack of Clꢀ at the C1 atom resulting in an instant cancellation of the formal charge separa-
tion. b) Formation of the [PdCl(Ph2Pallyl)]+ complex 5 from the dicationic complex 4; likely mechanism ini-
tially driven by the charge deficiency at the Pd center and passing through a formal resonance-stabilized PdIV
intermediate.
A
ACHTUNGERTN(NUNG BIMIONAP)Cl] primary
13098
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13095 – 13108

N-Heterocyclic Carbene (NHC)–Phosphenium Adducts
FULL PAPER
A nonchlorinated [Pd
aged. Treatment of 1 with the ꢃchloride freeꢄ [Pd
(MeCN)2]
[TfO] cationic complex[26] in CH2Cl2 afforded the
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
A
ACHTUNGTRENNUNG
targeted complex as a mixture (60:40) of two diastereoiso-
mers 4a,b in 84% yield (Scheme 5). The presence of two
diastereoisomers is attributed to a slow 1808 rotary flip of
the p-allyl ligand and to the restricted rotation of the N2C-
P-Pd-P-Napht unit within the seven-membered metallacycle.
The 31P NMR spectroscopic chemical shifts of 4a and 4b
(dP =17.5, 21.1 ppm, JPP =40.8 Hz; 18.0, 24.1 ppm, JPP
=
41.5 Hz) are close to those of complex (dP =18.8,
2
24.0 ppm), whereas the dicationic character of 4 was defini-
tively confirmed by mass spectroscopy (electrospray MS: m/
z: 923 [M2+ +OTfꢀ]+). In the 13C NMR spectrum, the two
NCN carbon atoms of 4a,b resonate at dc =++148.5 (d, JCP
15.7 Hz) and +148.1 ppm (d, JCP =18.0 Hz), respectively.
=
Figure 2. ORTEP view of the X-ray crystal structure of the NHC–phos-
phine palladium complex 5 with thermal ellipsoids drawn at the 30%
probability level (for clarity, triflate anion and H atoms are omitted). Se-
The dicationic complex 4 (4a+4b) was then reacted with
TEACl in MeCN, and by contrast to complex 2, the reaction
proceeded at room temperature, due to the enhanced elec-
trophilicity of 4. Monitoring the reaction by 31P NMR spec-
troscopy indicated the disappearance of 4 and the appear-
ance of two new signals at dP =+17.76 (d) and +15.13 ppm
ꢀ
ꢀ
lected bond distances [ꢂ] and angles [8]: C1 Pd1: 1.983(5), P1 Pd1:
ꢀ
ꢀ
ꢀ
ꢀ
2.3183(14), P2 Pd1: 2.3145(15), N1 C1: 1.323(7), N2 C1: 1.356(6), Pd1
Cl1: 2.3144(14); C1-Pd1-Cl1: 176.74(16), P1-Pd1-P2: 169.62(5), N1-C1-
N2: 107.8(4).
ate, stabilized by resonance
with a phosphoniacarbene–PdII
form (Scheme 4b). A related
complex resulting from oxida-
ꢀ
tive addition of a P Br bond to
a Pd0 center was recently de-
scribed by a divalent phosphe-
nium=Pd structure.[2d] In the
Scheme 5. Synthesis of the cationic P,P- and P,C-chelated palladium complexes 4 and 5, respectively.
present case, the PdIV/PdII spe-
(d) with a large P–P coupling constant (JPP =457.9 Hz),
along with the signals of the above described complex 3.
After column chromatography, the complexes 3 and 5 were
isolated in 29 and 60% yield, respectively. In the 13C NMR
spectrum of 5, a deshielded signal at dc =170.5 ppm (dd,
cies would then undergo a converse process, namely reduc-
tive elimination of the phosphide and allyl ligands producing
the allyldiphenylphosphine complex 5. Beyond mechanistic
issues, the two BIMIONAP complexes 2 and 4 behave glob-
ally in a similar manner towards the chloride anion, which
induces the cleavage of the assumed dative NHC!phosphe-
nium bond to give a transient NHC unit. The latter is imme-
diately trapped by the vicinal palladium center affording the
NHC–phosphine-chelated complexes 3 and 5.
J
CP =6.8, 11.5 Hz) is the signature of a Pd-coordinated car-
benic carbon atom in the vicinity to two nonequivalent
phosphorus atoms, themselves positioned in a trans position
(JPP =457.9 Hz).
The exact structure of 5 was determined by X-ray diffrac-
tion analysis of yellow crystals deposited from a CH2Cl2/
pentane mixture at ꢀ208C (Figure 2).[22] The square-planar
palladium center is thus coordinated to the bidentate NHC–
phosphine ligand, to an allyldiphenylphosphine ligand, and
to a chloride ligand. As previously mentioned for complex
3, the metal–ligand bond lengths in 5 are comparable to
those reported for similar NHC–phosphine–Pd com-
plexes.[23]
Reactivity of the free BIMIONAP cation with chloride and
n-butyllithium anions: By following the procedure used for
the BIMIONAP complex 2 (see the section entitled “Reac-
tivity of palladium(II) complexes of BIMIONAP with a
chloride ion”), the free BIMIONAP cation 1 was treated
with one equivalent of TEACl in MeCN at 508C. After 3 h,
the 31P NMR spectrum indicated the complete disappear-
ance of the signals of 1 (dP =ꢀ18.6, ꢀ16.3 ppm, JPP
=
As in the case of 3, the formation of 5 results from the
35.5 Hz) giving way to a slightly more deshielded signal at
1
cleavage of the N2C P bond of BIMIONAP. In the present
dP =ꢀ15.6 ppm.[27] The H NMR spectrum showed a signal
ꢀ
case, however, initial attack at the cationic Pd center, which
is likely to be more electronegative than the C1 atom, also
instantly reduces the +/ꢀ charge separation. The process
would now go through a formal phosphide–PdIV intermedi-
at dH =9.4 ppm in the classical range of iminium CH protons
(Scheme 6). The imidazoliophosphine 6 was finally isolated
in 66% yield and its structure was confirmed by multinu-
clear NMR spectroscopy and mass spectrometry technics.
Chem. Eur. J. 2010, 16, 13095 – 13108
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13099

Y. Canac, R. Chauvin et al.
The iminium 6 is the signature of the formation of a transi-
ent free NHC, which would be protonated by the MeCN
solvent.
The displacement of a NHC moiety from a P atom by
such a weak nucleophile as Clꢀ is indicative of the NHC!
ꢀ
phosphenium dative character of the N2C P bond in non-
coordinated BIMIONAP.
A similar procedure was conducted at ꢀ788C with nBuLi
as a stronger and likely harder nucleophile (instead of Clꢀ),
and in THF as a nonprotic solvent (instead of MeCN). To
trap the expected but elusive free NHC product, the reac-
tion was carried out in the presence of elemental sulfur (S8),
a classical trapping agent of carbenes (Scheme 6). Monitor-
ing the reaction by 31P NMR spectroscopy indicated the im-
mediate disappearance of the starting BIMIONAP (1) and
the presence of two new signals at dp=+42.7 and
+43.5 ppm. After purification, these two signals were as-
Figure 3. ORTEP view of the X-ray crystal structure of the o-phenylene-
diamine derivative 8 with thermal ellipsoids drawn at the 30% probabili-
ty level (for clarity, H atoms are omitted). Selected bond distances [ꢂ]
ꢀ
ꢀ
ꢀ
ꢀ
and angles [8]: C1 N1: 1.340(2), C2 N1: 1.455(3), C1 O1: 1.218(2), C10
signed to thioACHTUNGTRENNUNG(butyl)diphenylphosphine and thiophosphine-
P1: 1.8371(17); N1-C1-O1: 124.38(18).
thiourea 7 which was isolated in 71% yield, respectively.[28]
AHCTUNGTRENNUNG
a
Scheme 4a-C), the mechanism
operating in the degradation of
free BIMIONAP could also be
initiated by nucleophilic attack
of Clꢀ at the C1 atom, followed
by a-elimination of the diami-
nocarbene
and
Ph2PCl
(Scheme 7a). Alternatively, and
especially for the nBuLi nucleo-
phile from which the a-elimina-
tion process is less likely, the in-
itial attack could occur at the
phosphorus center to give a
neutral phosphonium C-diami-
Scheme 6. Reactions of free BIMIONAP with chloride and alkyl anions, followed by treatment with protic
sources (MeCN, H2O) or elemental sulfur (S8).
no-ylide
amino-ylide being highly destabi-
lized,[31] it could then undergo dissociation to the neutral
phosphine and the neutral diaminocarbene, which would be
intermediate
The formation of 7 is a signature of the generation of the
transient free carbene. Isolation or detection of the latter
was finally envisaged.
(Scheme 7b). The C-diACHTUNGTRENNUNG
AHCTUNGTRENNUNG
If the same reaction (BIMIONAP+nBuLi in THF) is car-
ried out in the absence of sulfur, the expected butyl-
stabilized by protonation, sulfurization, or hydrolysis
(Scheme 7b).
A
is
produced
(31P NMR:
d=
ꢀ16.0 ppm)[29] along with the phosphine–formamide
8
Theoretical analysis of the nature of the N2C P bond of BI-
ꢀ
(31P NMR: d=ꢀ16.3 ppm), which was isolated in 87% yield
and the structure of which was confirmed by an X-ray dif-
fraction analysis (Figure 3).[22] The formation of the o-phe-
nylenediamine derivative 8 results from a hydrolytic opening
of a benzimidazole ring, which may be explained by the
presence of residual water molecules in the starting BIMIO-
NAP (1).[30] All attempts to characterize the free NHC in
situ by VTP NMR experiments after drying beforehand the
hygroscopic BIMIONAP crystals were, however, unsuccess-
ful.
MIONAP: In a complementary approach, the nature of the
ꢀ
N2C P bond of BIMIONAP and its reactivity towards nu-
cleophiles were investigated with suitable theoretical tools,
in particular, ELF[32] and atoms in molecules (AIM) analy-
ses.[33]
ꢀ
Dative nature of the N2C P bond of BIMIONAP in the gas
ꢀ
phase: The preferred dissociation mode of the N2C P bond
for a model of BIMIONAP (1) has been calculated in the
gas phase at the B3PW91/6-31G** level of theory
(Scheme 8). The heterolytic mode (Dhetero =76.9 kcalmolꢀ1,
ZPE-corrected value) is found to be favored by about
15 kcalmolꢀ1 over the homolytic mode (Dhomo =91.8 kcal
By analogy with the mechanism proposed for the reaction
of a Clꢀ anion with the P-coordinated BIMIONAP ligand in
complex 2 (see the section entitled: Reactivity of palladi-
13100
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13095 – 13108

N-Heterocyclic Carbene (NHC)–Phosphenium Adducts
FULL PAPER
cally more favorable heterolytic
cleavage. According to the
mesomeric
description
of
Scheme 10b, BIMIONAP is
thus expected to retain signifi-
cant s-donating properties as
the lone pair of P1 is not in-
volved in the electron delocali-
zation. Moreover, P1 and P2
atoms are hardly distinguisha-
ble from the AIM charges and
ELF populations of the mono-
synaptic V(P) basin correspond-
ing to their lone pair. Only the
shift in population of the P–C
disynaptic basins
Scheme 7. Possible mechanisms for the nucleophilic degradation of free BIMIONAP: a) initial attack at the
C1 center by a Clꢀ ion, followed by a-elimination of Ph2PCl (see analogy with the proposed mechanism for
the Cl2Pd–P-coordinated BIMIONAP in complex 2 (Scheme 4a-C); b) intitial attack at the P1 atom to give an
unstable phosphonium C-diamino-ylide. In both mechanisms, the released NHC would be trapped in situ by
protonation, hydrolysis, or sulfurization (see Scheme 6).
(V
A
VACTHNGUTERN(UNG P2,C10):
ꢀ
ꢀ
2.31) allows us to distinguish between the P2 C2 and P1 C1
bonds (Scheme 9). It is noteworthy that the excess of elec-
ꢀ
trons in the P1 C1 bond cannot be attributed to a P1=C1
Scheme 8. Homolytic (top) and heterolytic (bottom) dissociation schemes
and corresponding energies (B3PW91/6-31G** level of calculation) of a
model of BIMIONAP (1).
molꢀ1, ZPE-corrected value). This result is in complete ac-
cordance with the experimental results.
ELF- and AIM-based mesomeric description of BIMIO-
NAP: From a general standpoint, the ELF partition of the
molecular space provides basins that correspond to the
cores, lone or shared-pairs of the Lewis model.[32] Their
mean populations and (co)variances can be interpreted in
terms of weighted combinations of mesomeric structures.[34]
ꢀ
The method is here applied to the N2C P bond in both the
free cation 1 and the BIMIONAP palladium dichloride com-
plex 2 (Scheme 9).
AIM atomic charges and scaled populations[35] of the ELF
valence basins of the carbene phosphenium moieties of 1
and 2 are very similar (Scheme 9). The same Lewis forms
can, therefore, be invoked for their mesomeric description.
In the ELF-based description of Scheme 10b, the C!P+
symbol is not strictly equivalent to the C+–P Lewis symbol
of Scheme 10a because it has the additional meaning that
ꢀ
the N2C P bond is energetically labile (an indirect criterion
for a donor–acceptor interaction) and dissociates in a heter-
olytic manner. This description is in agreement with litera-
ture data on related compounds[17,18] that concluded to
“dative C!P bonds” according to the definition of Mullike-
n[1a–b] and Haaland,[1c] that is, polar bonds with an energeti-
Scheme 9. Scaled populations of selected ELF valence basins of free BI-
MIONAP (1) (top; B3PW91/6-31G** level of calculation) and BIMIO-
NAP–Pd complex 2 (bottom; B3PW91/6-31G**/DGDZVP(Pd) level of
calculation). AIM charges are given in square brackets.
Chem. Eur. J. 2010, 16, 13095 – 13108
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13101

Y. Canac, R. Chauvin et al.
in which QN+e(A) denotes the
atomic charge (Mulliken, NPA,
AIM) of the atom A.
Recently, alternative conden-
sation schemes of frontier mo-
lecular orbital Fukui functions
within an AIM[39] or ELF[40]
topological partition have been
disclosed [Eq. (3)]:
Z
2
fXa
¼
jꢀFKSðrÞj dr
ð3Þ
Scheme 10. Mesomeric descriptions of the imidazoliophosphine unit of BIMIONAP (the remainder of the
structure is omitted for clarity). a) Lewis topological mesomerism. b) Topological-energetical mesomerism by
X
ꢀ
using a dative bond symbol reflecting the lability and the heterolytic dissociation preference of the N2C P
bond: the weights are determined from the ELF valence basins populations and covariances (Scheme 9).
Equation (3) expresses the
contribution of the FMO F to
the atomic AIM basin or to the
double character because the bond length is typical of a
pure single bond (1.847 ꢂ vs. an even shorter length of
1.808 ꢂ in the neutral nonmethylated BIMINAP precursor
depicted in Scheme 11).[18a]
core or valence ELF basin X (a=+: F=LUMO; a=ꢀ: F=
HOMO). These indices are more attractive than the above
atomic indices because they are confined in the 0–1 range
and they sum to one: 0ꢃfaX ꢃ1 and ꢀxfXa =1. The larger the
value of faX, the more reactive the corresponding site X. The
dual reactivity descriptor Df=f +(r)ꢀfꢀ(r) can also be used
for the simultaneous detection of nucleophilic or electrophil-
ic behavior of a given site and is particularly useful for am-
biphilic reactants.[41]
ELF- and AIM-based electrophilic Fukui indices of BIMIO-
ꢀ
NAP: Insights into the mechanism of the observed N2C P
bond cleavage (see the section entitled: Experimental reac-
ꢀ
tivity of the N2C P bond of BIMIONAP and Scheme 4)
have been gained within the framewok of a “pre-reacting”
approach, thus focusing on the early stage of the process.
The electrophilicity of BIMIONAP was thus investigated by
using two complementary analytical tools,[36] namely, the
molecular electrostatic potential (MESP) and the Fukui
function. The former index is suitable for the description of
hard–hard or charge-controlled interactions, whereas the
latter will allow for probing soft sites of the molecules that
are involved in orbital-controlled interactions with electro-
philes, nucleophiles, or radicals. The Fukui function was in-
troduced by Parr and Yang as a generalization of Fukuiꢄs
frontier molecular orbitals (FMO) concept as the response
of the electron density of the molecular system to a change
in the global number of electrons.[37] It can be expressed as
the derivative of the electron density 1(r) with respect to
the number of electrons N, calculated at constant external
potential v(r). Because of the discontinuity induced by the
discrete intrinsic nature of N, left and right derivatives have
to be considered. The local Fukui functions f +(r)=
Fukui functions f +, characterizing the reactivity towards
nucleophilic attack, condensed on ELF or AIM basins, were
calculated for BIMIONAP and its neutral parent BIMINA-
P[18a] at the B3PW91/6-31G** level (Scheme 11, Table 1).
ꢀ
The largest f +
value (0.16) is thus obtained for the P1
ELF
Table 1. Selected values of Fukui functions f + condensed on ELF or
AIM basins for BIMINAP and BIMIONAP. See atom labelling in
Schemes 9 and 11.
BIMINAP
ELF basin
BIMIONAP (1)
ELF basin
f +
DfELF
f +
DfELF
ELF
ELF
V
N
0.00
0.09
0.07
ꢀ0.01
0.04
0.06
V
V
V(P2)
U
0.16
0.05
0.05
0.16
0.05
0.04
V
V
N
E
AIM basin
f +
DfAIM
AIM basin
f +
DfAIM
AIM
AIM
C9
C11
P1
0.16
0.10
0.11
0.00
ꢀ0.27
P1
C1
N2
P2
0.06
0.05
0.24
0.10
ꢀ0.34
0.13
0.00
0.02
0.24
0.10
0.01
P2
(@1/@N)+ and fꢀ(r)=(@1/@N)ꢀ are the respective response
for adding or removing electrons from the system and allow-
ing, therefore, for an investigation of nucleophilic and elec-
trophilic attacks, respectively (radical attack is assumed to
C1 bond of BIMIONAP, in agreement with the experimen-
tal observations that this bond is the most sensitive to nucle-
ophilic attack (see the section entitled: Reactivity of the
free BIMIONAP cation with chloride and n-butyllithium
ACTHNUTRGNEUNG
be correlated with f0(r)=1/2[f+(r)+fꢀ(r)]). The finite differ-
ence approach (DN=ꢁ1) yielding the following expressions
[Eq. (1) and (2)] of the local Fukui function f(r) has been
extensively used for the estimation of atomic Fukui indices
f(A) from atomic charges:[38]
anions). The largest f +
value calculated for the C1 atom
AIM
(0.24) points to the latter as the most reactive site towards
sufficiently soft nucleophiles. In contrast, the lower f +
AIM
value calculated for the P1 atom (0.06) suggests that the
latter is more prone to react with hard nucleophiles. The
free chloride Clꢀ is bigger and thus softer than the carba-
nionic atom of the nBuꢀ anion. Moreover, the latter is made
f
þðrÞ ꢂ 1Nþ1ðrÞꢀ1NðrÞ ! f þðAÞ ¼ QNðAÞꢀQNþ1ðAÞ
ð1Þ
ð2Þ
fꢀðrÞ ꢂ 1NðrÞꢀ1Nꢀ1ðrÞ ! fꢀðAÞ ¼ QNꢀ1ðAÞꢀQNðAÞ
13102
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13095 – 13108

N-Heterocyclic Carbene (NHC)–Phosphenium Adducts
FULL PAPER
Table 2. Selected values of Fukui functions f + condensed on ELF or
AIM basins for the PdCl2 complexes of BIMINAP and BIMIONAP com-
plex 2. See atom labelling in Scheme 11.
[Pd
ELF basin
(BIMINAP)Cl2]
[PdACHTNGUTERN(UNG BIMIONAP)Cl2] (2)
f +
DfELF
ELF basin
f +
DfELF
ELF
ELF
C(Pd)
0.12
0.03
0.04
0.06
0.04
0.04
0.02
0.02
0.10
0.03
0.04
0.06
0.04
0.04
0.01
0.01
C(Pd)
0.11
0.15
0.03
0.03
0.03
0.03
0.02
0.01
0.10
0.15
0.03
0.03
0.02
ꢀ0.18
0.01
0.01
V
A
V
V
V
V
ACHTUNGTRENNUNG
V
V
V
V
V
V
N
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
V(Cl)
V
V
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
Scheme 11. Atom labelling for Fukui indices for BIMINAP. Values of
ELF basin populations, AIM charges and Fukui indices are given in
Table 1. The same labeling is adopted for the Pd complex in Table 2.
AIM basin
f +
DfAIM
AIM basin
f +
DfAIM
AIM
AIM
Pd
C1
C9
C10
C12
P1
0.12
0.10
0.03
0.12
0.07
0.09
0.02
0.02
Pd
C1
N1
C9
C10
P1
0.12
0.20
0.09
0.04
0.05
0.02
0.04
0.10
0.20
0.09
0.03
0.04
0.02
0.03
0.03
0.12
0.07
0.09
0.03
0.04
even harder through its interaction with the very hard Li+
cation in nBuLi. According to Pearson indeed the hardness
(h) of anionic atoms is approximately equal to the hardness
of the corresponding radicals[42] and h(Cl)=4.7 eV<h(C)=
5.0 eV. The soft chloride ion would thus react at the C1
atom, in agreement with the mechanism proposed in
Scheme 7a. In contrast, the harder carbanionic center of
nBuLi could react in a charge-controlled manner at the
P2
P2
moiety. It is much lower in energy than the LUMO+2 of
the parent BIMINAP resulting from the same type of frag-
ment orbital overlap. This is in agreement with the high
electrophilicity of the P1 C1 bond of 1 and further suggests
that this ligand should possess sizeable p-accepting proper-
ties.
The HOMO of BIMIONAP (occupied by the lone pair of
P2) and the HOMO-1 and HOMO-3 (exhibiting a sizeable
weight on P1) are much deeper in energy than the corre-
sponding HOMO and HOMO-1 of the parent BIMINAP.
The BIMIONAP ligand is, therefore, expected to be much
less s-donating than the BIMINAP ligand.
ꢀ
harder P1 center of the C1 P1 bond. The P1 atom being 2.5
ꢀ
times more positively charged than the C1 atom
(QAIM(P1)=1.6@QAIM(C1)=0.64, Scheme 9), the attack of
the hard nBuLi nucleophile is charge-controlled at the P1
center, in agreement with the mechanism proposed in
Scheme 7b.
ꢀ
In the neutral parent BIMINAP, both the P C bonds are
equivalent and are predicted to be insensitive to nucleophil-
ic attack because of their low computed f + values (Table 1),
in agreement with experimental observations.[43]
In the [Pd
G
The LUMOs of the complexes [Pd
ACHTUNGTRENN(GNU BIMIONAP)Cl2] (2)
f+
are the largest for the N2C P bond (0.15) and the Pd
and [Pd(BIMINAP)Cl2] are displayed in Figure 5 (for the
ACHTUNGTRENNUNG
ELF
center (0.11), and almost vanish at the P Pd bonds
(Table 2). The P1 Pd bond is thus predicted to be cleavage-
XRD geometry and at the B3PW91/6-31G**/DGDZVP
level of calculation). The LUMO of 2 is much lower in
energy and thus much more reactive than that of the neutral
BIMINAP complex, in agreement with both expectation
and experimental observations. The LUMO of 2 is compara-
ble in shape and location to the LUMO of the free ligand 1.
resistant in the early stage of the process, as proposed in
Scheme 4a-C. The largest f +
(0.20; it is almost two times smaller at the Pd atom: 0.12).
As in the case of free BIMIONAP (Scheme 7a), the initial
attack of 2 by the soft Clꢀ anion is predicted to occur at the
C1 center. All these results finally support the mechanism
proposed in Scheme 4a-C. For comparison, in the neutral
ꢀ
This suggests that the P1 C1 bond of 2 is the most sensitive
to attacks by sufficiently soft nucleophiles, in agreement
with the mechanism proposed in Scheme 4a-C, and with the
predictions of the ELF-condensed Fukui indices (see the
section entitled: ELF- and AIM-based electrophilic Fukui
indices of BIMIONAP and Table 2).
[Pd
ACHTUNGTRENNUNG
and f +
2, but the reactivity of the C1 P1 bond is found to vanish:
f + (P1,C1))=0.03 (!0.15 for 2), f +
(VACHTUNGTRENNUNG (C1)=0.03
ELF
(!0.20 for 2).
Molecular electrostatic potential of BIMIONAP: The molec-
ular electrostatic potential (MESP) has been extensively
used for probing molecular structure and reactivities.[44] It
corresponds to the potential generated by the molecular
charge distribution as experienced by a positive point
charge [see Eq. (4)]:
Frontier molecular orbitals of BIMIONAP: The near-fron-
tier molecular orbitals calculated for the experimental ge-
ometry of BIMIONAP (1) and BIMINAP at the B3PW91/6-
31G** level are displayed in Figure 4.
The LUMO of BIMIONAP (1) results from the overlap
Z
1ðr0Þ
X
ZA
jr ꢀ RAj
VðrÞ ¼
ꢀ
d3r0
of the p*
ACHTUNGTRENNUNG(NHC) antibonding orbital of the benzimidazolium
ð4Þ
jr ꢀ r0j
A
moiety with the pp HOMO of the diphenylphosphenium
Chem. Eur. J. 2010, 16, 13095 – 13108
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13103

Y. Canac, R. Chauvin et al.
attack.[45] For nucleophilic at-
tacks, since the maxima of
MESP are exclusively located
at nuclear positions, it is con-
venient to use MESP textured
on isodensity surfaces for ex-
ploring the regions of positive
MESP and finding the most
electrodeficient sites. Such sur-
faces have been obtained for
BIMINAP, BIMIONAP, and
their
PdCl2
complexes
(Figure 6).
These plots indicate that the
most positive MESP region is
ꢀ
located near the P1 C1 bond of
both the free BIMIONAP (1)
and its palladium complex 2.
This result is in agreement with
the electrophilic reactivity of
ꢀ
the P1 C1 bond towards the
Clꢀ and nBuꢀ anions (see the
section entitled: Experimental
ꢀ
reactivity of the N2C P bond of
BIMIONAP). Finally, both the
orbitals and the charges tend to
control nucleophilic attack at
ꢀ
the C1 P1 bond. In the corre-
sponding regions of the parent
free
BIMINAP
or
[Pd-
AHCTUNTGRENN(GNU BIMINAP)Cl2] complex, the
Figure 4. Near-frontier orbitals of BIMIONAP (1) and BIMINAP (B3PW91/6-31G** level of calculation).
MESP is much less positive, in
agreement with the observed
ꢀ
insensitivity of the P1 C1 bond
towards nucleophilic attack.
All the computed reactivity indices are, therefore, consis-
tent with the mechanisms of Schemes 4a, 7a, and 7b reflect-
ꢀ
ing that the P1 C1 bond of free or Cl2Pd–P-coordinated BI-
MIONAP is the most sensitive site to nucleophilic attack.
These indications could serve as guides for a complete
mechanistic study (of much higher computational cost)
through the exploration of the potential energy surface
versus suitable reaction coordinates.
Conclusion
Addition of anionic nucleophiles (Clꢀ, nBuꢀ) to free BI-
MIONAP (1) and BIMIONAP-containing palladium com-
Figure 5. LUMOs of palladium complexes of BIMIONAP complex 2 and
BIMINAP (B3PW91/6-31G**/DGDZVP level of calculation).
ꢀ
plexes (2, 4) results in selective cleavage of the N2C P
bond, from which the released NHC fragment can be
trapped by protonation, hydrolysis, sulfurization, or coordi-
nation to PdII centers. The displacement of a highly reactive
in which ZA denotes the nuclear charges, RA the position of
the nuclei, and 1(r) the electron density.
The MESP minima indicate the concentration sites of
electron density and the most favorable sites of electrophilic
NHC fragment by a so called “weak” chloride nucleophile
ꢀ
ꢀ
and the unusual transformation scheme C P+Cl !C:+
ꢀ
Cl P are convincing signatures of a donor–acceptor covalent
interaction in amidiniophosphines, that are, therefore, better
13104
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13095 – 13108

N-Heterocyclic Carbene (NHC)–Phosphenium Adducts
FULL PAPER
CaH2. All other reagents were used as commercially available. All reac-
tions were carried out under an argon atmosphere, by using schlenk and
vacuum line techniques. Column chromatography was carried out on
silica gel (60 ꢂ, C.C 70–200 mm). The following analytical instruments
were used. 1H, 13C, and 31P NMR spectroscopy: Bruker ARX 250, DPX
300, and AV 500. NMR chemical shifts d are in ppm, with positive values
to high frequency relative to the tetramethylsilane reference for 1H and
13C and to H3PO4 for 31P.
Compound 3:
0.66 mmol) and [Pd
A
mixture of tetraethylammonium chloride (0.109 g,
(BIMIONAP)Cl2] (2) (0.418 g, 0.44 mmol) was dis-
AHCTUNGTRENNUNG
solved in CH3CN (20 mL) and stirred at 508C for 3 h. After evaporation
of the solvent, the solid residue was purified by flash chromatography on
silica gel (CH2Cl2/acetone) to afford a yellow solid (yield: 0.17 g, 62%).
Recrystallisation at room temperature from a CH2Cl2/MeOH/Et2O mix-
ture gave complex 3 as yellow crystals (dec. p. 2508C); 31P{1H} NMR
(CDCl3, 258C): d=+23.16 ppm; 1H NMR (CDCl3, 258C): d=8.08 (d,
J
J
HH =8.4 Hz, 1H; Har), 8.03 (d, JHH =8.2 Hz, 1H; Har), 7.75 (pseudot,
HH =7.2, 7.6 Hz, 1H; Har), 7.67 (d, JHH =8.5 Hz, 1H; Har), 7.62–7.56 (m,
5H; Har), 7.53–7.50 (m, 2H; Har), 7.28–7.24 (m, 4H; Har), 7.17–7.12 (m,
3H; Har), 7.04 (pseudot, JHH =8.7, JHP =9.5 Hz, 1H; Har), 6.74 (d, JHH
=
8.2 Hz, 1H; Har), 4.36 ppm (s, 3H; CH3N); 13C{1H} NMR (CDCl3, 258C):
d=170.5 (s, NCN), 138.9 (d, JCP =8.6 Hz; Car), 136.9 (s, Car), 136.0 (d,
J
J
CP =1.9 Hz; Car), 135.6 (d, JCP =10.5 Hz; CHar), 133.5 (s, Car), 132.5 (d,
CP =3.0 Hz; CHar), 131.7 (d, JCP =2.4 Hz; CHar), 129.8 (d, JCP =8.6 Hz;
CHar), 129.2 (s, CHar), 128.9 (s, CHar), 128.6 (d, JCP =12.0 Hz; CHar),
128.3 (brs, CHar), 128.2 (s, CHar), 127.2 (d, JCP =6.4 Hz; Car), 126.9 (d,
J
CP =57.9 Hz; Car), 126.5 (d, JCP =5.5 Hz; CHar), 124.7 (s, CHar), 124.3 (s,
CHar), 124.1 (s, CHar), 122.4 (d, CP =55.3 Hz; Car), 121.6 (d, JCP
J
=
54.3 Hz; Car), 112.7 (s, CHar), 110.9 (s, CHar), 35.9 ppm (s, CH3N); MS
(FAB+): m/z: 583 [MꢀCl]+; HRMS (ES+): m/z: calcd for
C32H26ClN3PPd: 624.0596; found: 624.0632 [MꢀCl+CH3CN]+.
Compound 4: A mixture of complex bis(acetonitrile)-h3-allylpalladium
triflate (0.106 g, 0.28 mmol) and BIMIONAP (1) (0.22 g, 0.28 mmol) was
dissolved in CH2Cl2 (10 mL) and stirred at room temperature for 2 h.
After filtration and evaporation of the solvent, the residue was washed
with Et2O (3ꢅ10 mL) to afford a yellow solid as a mixture (60:40) of two
diastereoisomers 4a,b. Yield: 0.255 g, 84%; major isomer=A, minor
isomer=B: 31P{1H} NMR (CD3CN, 258C): d=+24.09 (d, JPP =41.5 Hz;
Figure 6. MESP visualization of BIMINAP and BIMIONAP ligands and
complexes. Isodensity surfaces (0.02 a.u.) colorcoded with the MESP.
Color scale: red: MESP<0.0; yellow: MESP=0.1; green: MESP=0.15;
light-blue: MESP=0.215; dark-blue: MESP> 0.26.
PB), +21.07 (d,
+17.49 ppm (d, JPP =40.8; PA); 1H NMR (CD3CN, 258C): d=8.26 (dd,
HH =7.6, JHP =13.6 Hz, 1.5H; Har), 8.18 (dd, JHH =7.3, JHP =13.5 Hz,
J J
PP =40.8 Hz; PA), +18.04 (d, PP =41.5 Hz; PB),
J
described as NHC!phosphenium adducts, as a priori sug-
gested before by several authors.[17,18] This experimental evi-
dence of a dative bonding is theoretically supported by the
comparison of the homolytic and heterolytic DFT-calculated
modes of dissociation in the gas phase and by ELF and
AIM analysis. Mechanistic issues on the observed electro-
philic reactivity of BIMIONAP are also gained from analy-
ses of the molecular orbitals, Fukui indices condensed on
ELF- and AIM basins, and MESP. Beyond the fundamental
information gained on the nature of NHC!phosphenium
interactions, the disclosed results pave the way for numerous
developments, for example, by using other weak nucleo-
philes and the possibly neutral nucleophiles. Finally, finding
applications of the disclosed reactivity on the enantiomeri-
cally resolved atropochiral BIMIONAP ligand is a natural
challenge in the field of stereoselective synthesis and asym-
metric catalysis.
2.2H; Har), 7.93–7.30 (m, 35H; Har), 7.26–7.10 (m, 6H; Har), 7.06–7.00
(m, 3.4H; Har), 6.93–6.89 (m, 1.7H; Har), 6.75–6.71 (m, 3.7H; Har), 6.63
(d, JHH =8.4 Hz, 1H; Har), 6.31–6.22 (m, 1.8H; CHallyl), 5.29–5.25 (m,
1.8H; CH2allyl), 4.56 (pseudot, JHH =JHP =5.8 Hz, 1H; CH2allyl), 4.46 (dd,
J
HH =10.2, JHP =13.9 Hz, 0.9H; CH2Allyl), 4.39 (dd, JHH =11.3, JHP =
11.8 Hz, 0.7H; CH2allyl), 4.27 (pseudot, JHH =JHP =5.1 Hz, 0.7H; CH2allyl),
4.01 (pseudot, JHH =JHP =11.8 Hz, 0.7H; CH2allyl), 3.68 (s, 3H; CH3N),
3.65 (s, 2.5H; CH3N), 3.46 ppm (pseudot,
CH2allyl); 13C{1H} NMR (CD3CN, 258C): d=148.5 (d, JCP =15.7 Hz; Car),
148.1 (d, JCP =18.0 Hz; Car), 135.3 (s, Car), 135.2 (s, Car), 135.0 (d, JCP
JHH =JHP =11.4 Hz, 0.8H;
=
16.3 Hz; CHar), 134.9 (d, JCP =12.6 Hz; CHar), 134.3–133.9 (m, Car; CHar),
133.5 (d, JCP =13.5; CHar), 133.3 (Car), 133.2 (Car), 132.9 (d, JCP =2.3 Hz;
CHar), 132.5–132.2 (m, Car; CHar), 131.6 (d, JCP =6.7 Hz; Car), 131.2 (d,
J
CP =6.5 Hz; Car), 130.6 (d, JCP =11.3 Hz; CHar), 130.5 (d, JCP =12.6 Hz;
CHar), 129.7 (d, JCP =11.3 Hz; CHar), 129.5–129.0 (m; Car, CHar), 128.8 (d,
CP =12.6 Hz; CHar), 128.6–128.5 (m; CHar), 127.7–127.4 (m; Car, CHar),
125.8 (d, JCP =47.8 Hz; Car), 125.5 (t, JCP =6.3 Hz; CHallyl), 125.1 (t, JCP
J
=
6.3 Hz; CHallyl), 124.4 (d, JCP =43.0 Hz; Car), 123.7–123.2 (m; Car, CHar),
122.0 (d, JCP =47.8 Hz, Car), 121.1 (q, JCF =320.0 Hz; CF3SO3ꢀ), 115.1 (s,
CHar), 115.0 (s, CHar), 113.4 (s, CHar), 113.3 (s, CHar), 84.9 (d, JCP
=
25.0 Hz; CH2allyl), 80.7 (d, JCP =24.4 Hz; CH2allyl), 79.0 (d, JCP =30.0 Hz;
CH2allyl), 75.8 (d, JCP =25.0 Hz; CH2allyl), 36.7 (s, CH3N), 36.4 ppm (s,
CH3N); MS (FAB): m/z: 923 [M2+OTfꢀ]+; HRMS (ES+): m/z: calcd for
C46H38N2P2PdF3SO3: 923.1065; found: 923.1052.
Experimental Section
Compound 5:
0.35 mmol) and [Pd
was dissolved in CH3CN (10 mL) and stirred at room temperature for
A
mixture of tetraethylammonium chloride (0.058 g,
General remarks: THF and diethyl ether were dried and distilled over
sodium/benzophenone, pentane, dichloromethane, and acetonitrile over
(allyl)(BIMIONAP)
R
ACHTUNGTRENNUNG
Chem. Eur. J. 2010, 16, 13095 – 13108
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13105

Y. Canac, R. Chauvin et al.
+
3 h. After evaporation of the solvent, the solid residue was purified by
flash chromatography on silica gel (CH2Cl2/acetone) to afford the carben-
ic complex 3 (yield: 0.052 g, 29%) and the carbenic complex 5 (0.170 g,
60%), respectively, as yellow solids. Recrystalisation of complex 5 at
ꢀ208C from a CH2Cl2/pentane mixture gave yellow crystals. M.p. 169–
507.1139
[M+H]
;
elemental
analysis
calcd
(%)
for
C30H23N2PS2·0.9H2O: C 68.91, H 4.78, N 5.35; found: C 69.20, H 5.17, N
4.55.
Compound 8: Butyllithium (2.5m, 0.04 mL, 0.1 mmol) was added drop-
wise to a solution of BIMIONAP (1) (0.081 g, 0.10 mmol) in THF (5 mL)
at ꢀ788C. Then the mixture was slowly warmed to room temperature
and stirred for 3 h. After evaporation of the solvent, the solid residue
was purified by chromatography on silica gel (CH2Cl2) to afford 8 as a
white solid. Yield: 0.04 g, 87%; m.p. 228–2308C; 31P{1H} NMR (CDCl3,
258C): d=ꢀ16.31 ppm; 1H NMR (CDCl3, 258C): d=8.28 (s, 1H; CHO),
7.89 (d, JHH =8.1 Hz, 1H; Har), 7.79 (d, JHH =8.1 Hz, 1H; Har), 7.71 (d,
1728C; 31P{1H} NMR (CDCl3, 258C): d=+17.76 (d,
JPP =457.9 Hz),
+15.13 ppm (d, JPP =457.9 Hz); 1H NMR (CDCl3, 258C): d=8.25–8.21
(m, 2H; Har), 7.85 (t, JHH =7.5 Hz, 1H; Har), 7.74–7.41 (m, 12H; Har),
7.30–7.09 (m, 12H; Har), 7.03 (t, JHH =7.2 Hz, 2H; Har), 6.66 (d, JHH
=
8.1 Hz, 1H; Har), 5.68–5.63 (m, 1H; CHallyl), 4.97 (d, JHH =17.4, 1H;
CH2allyl), 4.65 (d, JHH =10.2 Hz, 1H; CH2allyl), 3.54 (s, 3H; CH3N), 3.52–
3.38 ppm (m, 2H; CH2P); 13C{1H} NMR (CDCl3, 258C): d=171.2 (dd,
J
HH =8.4 Hz, 1H; Har), 7.56–7.51 (m, 1H; Har), 7.47–7.41 (m, 2H; Har),
7.40–7.28 (m, 9H; Har), 7.05 (d, JHP =2.4 Hz, 1H; Har), 7.02 (m, 1H; Har),
6.96–6.91 (m, 1H; Har), 6.77–6.72 (m, 1H; Har), 6.14 (d, JHH =8.1 Hz, 1H;
J
CP =7.5, 10.6 ppm; NCN), 137.8 (dd, JCP =3.7, 6.3 Hz; Car), 136.1 (s, Car),
136.0 (s, Car), 135.5 (dd, JCP =5.3, 7.5; CHar), 133.9 (dd, JCP =4.5, 7.4 Hz;
CHar), 133.5 (s, Car), 132.8 (s, CHar), 132.3 (brs, CHar), 132.0 (dd, JCP =4.5,
7.1 Hz; CHar), 131.4 (s, CHar), 131.1 (dd, JCP =2.7, 4.8 Hz; CHar), 130.3–
129.6 (m; Car), 129.6 (s, CHar), 129.5 (s, CHar), 129.4–129.1 (m; CHar),
129.0 (s, CHallyl), 128.9–128.8 (m; CHar), 128.6 (dd, JCP =3.3, 6.7 Hz;
CHar), 128.3 (d, JCP =11.5 Hz; CHar), 127.5 (dd, JCP =1.6, 3.5 Hz; Car)
127.4 (dd, JCP =31.3, 18.1 Hz; Car), 126.5 (brs, CHar), 125.5 (s, CHar),
125.1 (s, CHar), 125.0 (d, JCP =53.8 Hz; Car), 124.9 (d, JCP =15.8 Hz; Car),
Har), 5.87 (d,
13C{1H} NMR (CDCl3, 258C): d=163.7 (s, CHO), 143.0 (s, Car), 139.8 (d,
CP =21.6 Hz; Car), 135.7 (d, JCP =10.2 Hz; Car), 134.8 (s, Car), 134.2–133.4
JHP =1.8 Hz, 1H; NH), 3.13 ppm (s, 3H; CH3N);
J
(brs, CHar), 132.4 (d, JCP =10.1 Hz; Car), 130.5 (d, JCP =2.6 Hz; Car), 129.5
(s, CHar), 129.3 (s, CHar), 129.2–128.5 (m; CHar), 128.3 (s, CHar), 128.2 (s,
CHar), 127.4 (s, Car), 126.9 (s, CHar), 126.8 (s, CHar), 126.7 (s, CHar), 123.9
(d, JCP =2.7 Hz; CHar), 118.8 (s, CHar), 32.2 ppm (s, CH3N); MS (ES+):
m/z: 461 [M+H]+; HRMS (ES+): m/z: calcd for C30H26N2OP: 461.1783;
122.8 (dd,
JCP =31.3, 14.7 Hz; Car), 122.7 (s, CHar), 121.4–121.1 (m;
CHallyl), 121.0 (q, JCF =321.0 Hz; CF3SO3ꢀ), 112.1 (s, CHar), 112.0 (s,
CHar), 35.5 (s, CH3N), 33.0 ppm (dd, JCP =10.0, 17.8; CH2P); MS (ES+):
m/z: 809 [M]+; HRMS (ES+): m/z: calcd for C45H38ClN2P2Pd: 809.1234;
found, 809.1270.
found: 461.1793; elemental analysis calcd (%) for C30H25N2OP·ACHTUNGTRENNUNG(C4H8O):
C 76.67, H 6.25, N 5.26; found: C 76.85, H 6.27, N 4.90.
Computational details: Geometries of the model and fragments of BI-
MIONAP (1) were fully optimized at the (U)B3PW91/6–31G** level of
calculation by using Gaussian 03.[46] Vibrational analysis was performed
at the same level of calculation as the geometry optimization. AIM and
ELF analyses of the experimental structures were performed with the
TopMoD program[47] on the basis of the B3PW91/6–31G** wavefunction
for free ligands (BIMIONAP and BIMINAP) and of the B3PW91/6–
31G**/DGDZVP(Pd) one[46] for palladium complexes of BIMIONAP
and BIMINAP. The gas-phase molecular electrostatic potential (MESP)
was computed for the experimental geometries at the same level as
above by using Gaussian 03.[46] The MESP has been shown to be weakly
sensitive to the level of calculation.[48] Visualization of the isodensity sur-
faces colorcoded with the MESP were performed by using Molden.[49]
Compound 6:
A mixture of tetraethylammonium chloride (0.094 g,
0.56 mmol) and BIMIONAP (1) (0.250 g, 0.32 mmol) was dissolved in
CH3CN (6 mL) and stirred at 508C for 3 h. After evaporation of the sol-
vent, the solid residue was purified by flash chromatography on silica gel
(CH2Cl2/acetone) to afford 6 as a white powder. Yield: 0.126 g, 66%;
m.p. 92–948C; 31P{1H} NMR (CD2Cl2, 258C): d=ꢀ15.61 ppm; 1H NMR
(CD2Cl2, 258C): d=9.39 (d, JHP =1.8 Hz, 1H; NCHN), 8.16 (d, JHH
=
8.6 Hz, 1H; Har), 8.09 (d, JHH =8.2 Hz, 1H; Har), 7.90 (d, JHH =10.0 Hz,
1H; Har), 7.73–7.68 (m, 2H; Har), 7.59–7.56 (m, 1H; Har), 7.49–7.42 (m,
4H; Har), 7.37 (d, JHP =12.5 Hz, 1H; Har), 7.34 (d, JHP =3.2 Hz, 1H; Har),
7.35–7.31 (m, 2H; Har), 7.28–7.25 (m, 3H; Har), 7.21–7.17 (m, 2H; Har),
6.89 (d, JHH =8.5 Hz, 1H; Har), 4.29 ppm (s, 3H, CH3N); 13C{1H} NMR
(CD2Cl2, 258C): d=143.6 (s, NCHN), 136.8 (d, JCP =21.1 Hz; Car), 134.4
(s, Car), 134.3 (d, JCP =18.9 Hz; Car), 134.2 (d, JCP =22.0 Hz; CHar), 133.8
Crystal-structure determination of 3, 5, and 8: Intensity data for com-
pounds 3, 5, and 8 were collected at 180 K on a Bruker Apex2 or an Xca-
libur Oxford Diffraction diffractometer by using a graphite-monochro-
mated MoKa radiation source and equipped with an Oxford Cryosystems
Cryostream Cooler Device. Structures were solved by direct methods by
using SIR92, and refined by full-matrix least-squares procedures on F by
using the programs of the PC version of CRYSTALS. Atomic scattering
factors were taken from the International tables for X-ray crystallogra-
phy. All non-hydrogen atoms and nonsolvent molecules were refined ani-
sotropically. Hydrogen atoms were refined by using a riding model. Ab-
sorption corrections were introduced by using the program MULTI-
SCAN.
(d, JCP =20.5 Hz; CHar), 133.6 (s, Car), 132.5 (s, Car), 131.9 (d, JCP
=
22.8 Hz; Car), 131.8 (s, CHar), 131.5 (s, Car), 129.9 (s, CHar), 129.8 (s,
CHar), 129.7 (d, JCP =2.1 Hz; Car), 129.2 (s, CHar), 129.1 (d, JCP =7.3 Hz;
CHar), 129.0 (s, CHar), 128.8 (d, JCP =8.0 Hz; CHar), 128.6 (s, CHar), 128.3
(s, CHar), 127.8 (s, CHar), 127.6 (s, CHar), 121.4 (d, JCP =1.4 Hz; CHar),
120.8 (q,
J
CF =313.3 Hz; CF3SO3ꢀ), 113.8 (s, CHar), 113.1 (s, CHar),
34.1 ppm (s, CH3N); MS (ES+): m/z: 443 [M]+; HRMS (ES+): m/z: calcd
for C30H24N2P: 443.1677; found: 443.1686.
Compound 7: Butyllithium (2.5m, 0.033 mL, 0.082 mmol) was added
dropwise to a solution of BIMIONAP (1) (0.064 g, 0.082 mmol) in THF
(3 mL) at ꢀ788C and stirred for 20 min. Then a suspension of elemental
sulphur S8 (0.105 g, 0.41 mmol) in THF (3 mL) was added at ꢀ608C to
the solution. The mixture was slowly warmed to room temperature and
stirring was continued for 3 h. After evaporation of the solvent, the solid
residue was purified by chromatography on silica gel (CH2Cl2) to afford
7 as a white solid. Yield: 0.029 g, 71%; m.p. 268–2708C; 31P{1H} NMR
(CDCl3, 258C): d=+43.53 ppm; 1H NMR (CDCl3, 258C): d=8.14–8.06
(m, 3H; Har), 7.99–7.92 (m, 2H; Har), 7.63–7.39 (m, 7H; Har), 7.27–7.11
(m, 3H; Har), 6.99 (d, JHH =8.4 Hz; 1H; Har), 6.92–6.83 (m, 4H; Har),
3.43 ppm (s, 3H; CH3N); 13C{1H} NMR (CDCl3, 258C): d=170.7 (s, CS),
135.6 (d, JCP =2.2 Hz; Car), 134.2 (d, JCP =81.0 Hz; Car), 133.9 (s, Car),
Crystal data for complex 3: C30H23Cl2N2PPd, CH2Cl2, 0.5CH4O, 0.17H2O;
M=723.75 gmolꢀ1
;
trigonal;
a=32.7191(3),
b=32.7191(3),
c=
15.5065(4) ꢂ; V=14376.3(4) ꢂ3; T=180 K; space group Rꢀ3; Z=18; m-
AHCTUNGTERNNUNG =
(MoKa)=0.992 mmꢀ1; 133082 reflections measured, 13004 unique (Rint
0.02), 10169 reflections used in the calculations [I>3s(I)], 356 parame-
ters, R1=0.0402, wR2=0.0450.
Crystal data for complex 5: C26H26N2PPd, CF3O3S, M=652.95 gmolꢀ1; tri-
clinic; a=12.6140(3), b=12.2030(2), c=16.8654(7) ꢂ; b=119.913(2)8;
V=2250.23(12) ꢂ3; T=180 K; space group Pc; Z=2;
mACHTUNGTRENNUNG(MoKa)=
0.765 mmꢀ1; 21663 reflections measured, 8732 unique (Rint =0.03), 7596
reflections used in the calculations [I>3s(I)], 520 parameters, R1=
0.0548, wR2=0.0636.
133.1 (d, JCP =2.0 Hz; Car), 133.0 (d, JCP =87.8 Hz; Car), 132.9 (d, JCP
=
10.9 Hz; CHar), 132.1 (d, JCP =11.7 Hz; CHar), 131.6 (d, JCP =3.0 Hz;
CHar), 130.4 (d, JCP =12.4 Hz; CHar), 130.2 (d, JCP =3.0 Hz; CHar), 130.1
(s, CHar), 129.9 (d, JCP =11.5 Hz; CHar), 129.5 (d, JCP =82.9 Hz; Car),
128.7 (s, CHar), 128.5 (s, CHar), 128.3 (s, CHar), 128.2 (s, CHar), 128.1 (s,
CHar), 126.3 (d, JCP =13.1 Hz; CHar), 123.6 (s, CHar), 123.4 (s, CHar),
111.0 (s, CHar), 108.5 (s, CHar), 30.6 ppm (s, CH3N); MS (ES+): m/z: 507
[M+H]+; HRMS (ES+): m/z: calcd for C30H24N2PS2: 507.1119; found:
Crystal data for 8: C30H25N2OP; M=460.52 gmolꢀ1
; triclinic; a=
9.8337(4), b=11.1319(4), c=11.6248(4) ꢂ; a=71.097(1), b=83.413(1),
g=85.522(2)8; V=1194.83(8) ꢂ3; T=180 K; space group Pꢀ1; Z=2; m-
AHCTUNGTERNNUNG =
(MoKa)=0.141 mmꢀ1; 26857 reflections measured, 7572 unique (Rint
0.02), 5218 reflections used in the calculations [I>3s(I)], 307 parameters,
R1=0.0497, wR2=0.0556.
13106
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13095 – 13108

N-Heterocyclic Carbene (NHC)–Phosphenium Adducts
FULL PAPER
nium adduct has been prepared and characterized, see: N. J. Hard-
man, M. B. Abrams, M. A. Pribisko, T. M. Gilbert, R. L. Martin,
G. J. Kubas, R. T. Baker, Angew. Chem. 2004, 116, 1989; Angew.
Chem. Int. Ed. 2004, 43, 1955.
Acknowledgements
The authors thank the french Ministꢆre de l’Enseignement Supꢀrieur et
de la Recherche, the french Centre National de la Recherche Scientifi-
que, and the Agence Nationale de la Recherche (program ANR-08-
JCJC-0137-01) for financial support. The authors also thank the
CALMIP (Calcul intensif en Midi-Pyrꢀnꢀes, Toulouse, France), the
IDRIS (Institut du Dꢀveloppement et des Ressources en Informatique
Scientifique, Orsay, France), and the CINES (Centre Informatique de
l’Enseignement Supꢀrieur, Montpellier, France) for computing facilities.
[15] a) C. W. Schultz, R. W. Parry, Inorg. Chem. 1976, 15, 3046; b) N.
Burford, T. S. Cameron, P. J. Ragogna, J. Am. Chem. Soc. 2001, 123,
7947; c) N. Burford, P. J. Ragogna, R. McDonald, M. J. Ferguson, J.
Am. Chem. Soc. 2003, 125, 14404.
[16] a) J. M. Slattery, C. Fish, M. Green, T. N. Hooper, J. C. Jeffery, R. J.
Kilby, J. M. Lynam, J. E. McGrady, D. A. Pantazis, C. A. Russel,
C. E. Willans, Chem. Eur. J. 2007, 13, 6967; b) N. Burford, D. E. Her-
bert, P. J. Ragogna, R. McDonald, M. J. Ferguson, J. Am. Chem.
Soc. 2004, 126, 17067.
[1] If the dissociated fragments of the dative bond are neutral species
M and L, the a-zwitterionic Lewis symbol Mꢀ–L+ is equivalent to
[17] B. D. Ellis, P. J. Ragogna, C. L. B. Macdonald, Inorg. Chem. 2004, 43,
7857.
!
the dative symbol M L; a) R. S. Mulliken, W. B. Person, Ann. Rev.
[18] a) N. Debono, Y. Canac, C. Duhayon, R. Chauvin, Eur. J. Inorg.
Chem. 2008, 2991; b) M. Azouri, J. Andrieu, M. Picquet, H. Cattey,
Inorg. Chem. 2009, 48, 1236; c) Y. Canac, N. Debono, L. Vendier, R.
Chauvin, Inorg. Chem. 2009, 48, 5562.
Phys. Chem. 1962, 13, 107; b) R. S. Mulliken, J. Am. Chem. Soc.
1952, 74, 811; c) A. Haaland, Angew. Chem. 1989, 101, 1017; Angew.
Chem. Int. Ed. Engl. 1989, 28, 992; d) V. I. Minkin, Pure Appl.
Chem. 1999, 71, 1933.
[19] a) A. J. Arduengo III, C. J. Carmalt, J. A. C. Clyburne, A. H.
Cowley, R. Pyati, Chem. Commun. 1997, 981; b) N. Burford, T. S.
Cameron, D. J. LeBlanc, P. Losier, S. Sereda, G. Wu, Organometal-
lics 1997, 16, 4712.
[20] a) I. V. Komarov, M. Y. Kornilov, A. A. Tolmachev, A. A. Yurchen-
ko, E. B. Rusanov, A. N. Chernega, Tetrahedron 1995, 51, 11271;
b) N. Kuhn, J. Fahl, D. Blꢌser, R. Boese, Z. Anorg. Allg. Chem.
1999, 625, 729; c) N. Kuhn, M. Gçhner, G. Henkel, Z. Anorg. Allg.
Chem. 1999, 625, 1415; d) A. A. Tolmachev, A. A. Yurchenko, A. S.
Merculov, M. G. Semenova, E. V. Zarudnitskii, V. V. Ivanov, A. M.
Pinchuk, Heteroat. Chem. 1999, 10, 585.
[21] a) Y. Canac, R. Chauvin, Eur. J. Inorg. Chem. 2010, 2325; b) S. Nan-
chen, A. Pfaltz, Helv. Chim. Acta 2006, 89, 1559; c) A. T. Normand,
K. J. Cavell, Eur. J. Inorg. Chem. 2008, 2781.
[22] CCDC-780614 (3), -780612 (5), and -780613 (8) contain the supple-
mentary crystallographic data for this paper. These data can be ob-
tained free of charge from The Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
[23] F. E. Hahn, M. C. Jahnke, T. Pape, Organometallics 2006, 25, 5927.
[24] I. Abdellah, N. Debono, Y. Canac, C. Duhayon, R. Chauvin, Dalton
Trans. 2009, 7196.
[25] a) S. A. MacGregor, Chem. Soc. Rev. 2007, 36, 67; b) R. Chauvin,
Eur. J. Inorg. Chem. 2000, 577.
[26] L. A. Evans, N. Fey, J. N. Harvey, D. Hose, G. C. Lloyd-Jones, P.
Murray, A. G. Orpen, R. Osborne, G. J. J. Owen-Smith, M. Purdie, J.
Am. Chem. Soc. 2008, 130, 14471.
[2] a) E. M. Arnett, K. E. Molter, Acc. Chem. Res. 1985, 18, 339; b) S.
Burck, K. Gçtz, M. Kaupp, M. Nieger, J. Weber, J. Schmedt auf der
Gꢇnne, D. Gudat, J. Am. Chem. Soc. 2009, 131, 10763; c) Z. Benkꢈ,
S. Burck, D. Gudat, M. Hofmann, F. Lissner, L. Nyulꢉszi, U. Zen-
neck, Chem. Eur. J. 2010, 16, 2857; d) C. A. Caputo, A. L. Brazeau,
Z. Hynes, J. T. Price, H. M. Tuononen, N. D. Jones, Organometallics
2009, 28, 5261; e) D. Gudat, A. Haghverdi, H. Hupfer, M. Nieger,
Chem. Eur. J. 2000, 6, 3414; f) N. Burford, P. J. Ragogna, J. Chem.
Soc. Dalton Trans. 2002, 4307.
[3] a) M. C. Jahnke, F. E. Hahn, Top. Organomet. Chem. 2010, 30, 95;
b) S. Dꢊez-Gonzꢉlez, N. Marion, S. P. Nolan, Chem. Rev. 2009, 109,
3612; c) F. E. Hahn, M. C. Jahnke, Angew. Chem. 2008, 120, 3166;
Angew. Chem. Int. Ed. 2008, 47, 3122; d) D. Pugh, A. A. Danopou-
los, Coord. Chem. Rev. 2007, 251, 610; e) S. P. Nolan, N-Heterocyclic
Carbenes in Synthesis, Wiley-VCH, Weinheim, 2006; f) N. Kuhn, A.
Al-Sheikh, Coord. Chem. Rev. 2005, 249, 829; g) E. Peris, R. H.
Crabtree, Coord. Chem. Rev. 2004, 248, 2239.
[4] Y. Wang, Y. Xie, P. Wei, R. B. King, H. F. Schaefer III, P. von R.
Schleyer, G. H. Robinson, Science 2008, 321, 1069.
[5] A. Sidiropoulos, C. Jones, A. Stasch, S. Klein, G. Frenking, Angew.
Chem. 2009, 121, 9881; Angew. Chem. Int. Ed. 2009, 48, 9701.
[6] Y. Wang, Y. Xie, P. Wei, R. B. King, H. F. Schaeffer III, P. von R.
Schleyer, G. H. Robinson, J. Am. Chem. Soc. 2008, 130, 14970.
[7] a) O. Back, G. Kuchenbeiser, B. Donnadieu, G. Bertrand, Angew.
Chem. 2009, 121, 5638; Angew. Chem. Int. Ed. 2009, 48, 5530;
b) J. D. Masuda, W. W. Schoeller, B. Donnadieu, G. Bertrand, J. Am.
Chem. Soc. 2007, 129, 14180; c) J. D. Masuda, W. W. Schoeller, B.
Donnadieu, G. Bertrand, Angew. Chem. 2007, 119, 7182; Angew.
Chem. Int. Ed. 2007, 46, 7052.
[8] a) C. A. Dyker, V. Lavallo, B. Donnadieu, G. Bertrand, Angew.
Chem. 2008, 120, 3250; Angew. Chem. Int. Ed. 2008, 47, 3206; b) A.
Fꢇrstner, M. Alcarazo, R. Goddard, C. W. Lehmann, Angew. Chem.
2008, 120, 3254; Angew. Chem. Int. Ed. 2008, 47, 3210.
[9] a) P. A. Rupar, V. N. Staroverov, P. J. Ragogna, K. M. Baines, J. Am.
Chem. Soc. 2007, 129, 15138; b) A. J. Ruddy, P. A. Rupar, K. J.
Bladek, C. J. Allan, J. C. Avery, K. M. Baines, Organometallics 2010,
29, 1362.
[10] a) R. S. Ghadwal, H. W. Roesky, S. Merkel, J. Henn, D. Stalke,
Angew. Chem. 2009, 121, 5793; Angew. Chem. Int. Ed. 2009, 48,
5683; b) A. C. Filippou, O. Chernov, G. Schnakenburg, Angew.
Chem. 2009, 121, 5797; Angew. Chem. Int. Ed. 2009, 48, 5687.
[11] T. Matsumoto, F. P. Gabbaꢋ, Organometallics 2009, 28, 4252.
[12] J. L. Dutton, H. M. Tuononen, P. J. Ragogna, Angew. Chem. 2009,
121, 4473; Angew. Chem. Int. Ed. 2009, 48, 4409.
[13] R. W. Alder, M. E. Blake, L. Chaker, J. N. Harvey, F. Paolini, J.
Schꢇtz, Angew. Chem. 2004, 116, 6020; Angew. Chem. Int. Ed. 2004,
43, 5896.
[27] Instead of the expected 31P NMR chemical shift of Ph2PCl, two new
signals at d=+37.2 (d) and ꢀ22.9 ppm (d), with JPP =227.2 Hz,
were observed. They were attributed to Ph2PP(O)Ph2, which should
result from a formal hydrolytic dimerization of Ph2PCl, the exact
mechanism of which under the present conditions has not been in-
vestigated.
[28] M. Yoshifuji, T. Ishizuka, Y. J. Choi, N. Inamoto, Tetrahedron Lett.
1984, 25, 553.
[29] G. A. Gray, S. E. Cremer, K. L. Marsi, J. Am. Chem. Soc. 1976, 98,
2109.
[30] Y. J. Bai, C. Y. Li, W. Sun, G. Zhao, Z. Shi, Huaxue Shiji 2008, 30,
409.
[31] a) Y. Canac, S. Conejero, M. Soleilhavoup, B. Donnadieu, G. Ber-
trand, J. Am. Chem. Soc. 2006, 128, 459; b) S. Conejero, M. Song, D.
Martin, Y. Canac, M. Soleilhavoup, G. Bertrand, Chem. Asian J.
2006, 1–2, 155.
[32] a) A. D. Becke, K. E. Edgecombe, J. Chem. Phys. 1990, 92, 5379;
b) B. Silvi, A. Savin, Nature 1994, 371, 683.
[33] R. F. W. Bader, Atoms in Molecules, Clarendon Press, Oxford, 1990.
[34] a) C. Lepetit, B. Silvi, R. Chauvin, J. Phys. Chem. A 2003, 107, 464;
b) B. Silvi, Phys. Chem. Chem. Phys. 2004, 6, 256.
[14] a) A. H. Cowley, R. A. Kemp, Chem. Rev. 1985, 85, 367; b) D.
Gudat, Coord. Chem. Rev. 1997, 163, 71; c) H. Nakazawa, Adv. Or-
ganomet. Chem. 2004, 50, 108; d) a cationic NHC–diaminophosphe-
[35] The valence populations are scaled to the actual number of elec-
trons involved in the Lewis structures.
Chem. Eur. J. 2010, 16, 13095 – 13108
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13107

Y. Canac, R. Chauvin et al.
[36] a) A. Borgoo, D. J. Tozer, P. Geerling, F. De Proft, Chem. Phys. Lett.
Phys. Chem. Chem. Phys. 2009, 11, 2862; b) A. Michalak, Chem.
Phys. Lett. 2004, 386, 346.
[37] R. G. Parr, W. Yang, J. Am. Chem. Soc. 1984, 106, 4049.
[38] H. Chermette, P. Boulet, S. Portmann, Rev. Mod. Quantum Chem.
2002, 2, 992.
[39] F. A. Bulat, E. Chamorro, P. Fuentealba, A. Toro-Labbꢀ, J. Phys.
Chem. 2004, 108, 342.
[40] W. Tiznado, E. Chamorro, R. Contreras, P. Fuentealba, J. Phys.
Chem. 2005, 109, 3220.
[41] a) C. Morell, A. Grand, A. Toro-Labbe, J. Phys. Chem. A 2005, 109,
205; b) C. Morell, A. Grand, S. Gutirrez-Oliva, A. Toro-Labbe,
Theor. Comput. Chem. 2007, 19, 101; c) C. Cardenas, N. Rabi, P. W.
Ayers, C. Morell, P. Jaramillo, P. Fuentealba, J. Phys. Chem. A 2009,
113, 8660.
[42] R. G. Pearson, Inorg. Chem. 1988, 27, 734.
[43] I. Abdellah, N. Debono, Y. Canac, L. Vendier, R. Chauvin, Chem.
Asian J. 2010, 5, 1225.
[44] a) E. Sorocco, J. Tomasi, Adv. Quantum Chem. 1978, 11, 116; b) E.
Sorocco, J. Tomasi in Chemical Applications of Atomic and Molecu-
lar Electrostatic potentials (Eds.: P. Politzer, D. G. Thruhlar),
Plenum, New York, 1981; c) P. Politzer, J. S. Murray, Theor. Chem.
Acc. 2002, 134.
[45] a) S. R. Gadre, S. A. Kulkarni, I. H. Shrivastava, J. Chem. Phys.
1992, 96, 5253; b) S. R. Gadre in Computational Chemistry Reviews
in Current Trends, Vol. 4 (Ed.: J. Leczynski), World Scientific, Singa-
pore, 2000, p. 1; c) S. R. Gadre, R. N. Shirsat in Electrostatics of
Atoms and Molecules, Universities Press, Hyderabad, 2000.
[46] Gaussian 03, Revision C.02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgom-
ery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S.
Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani,
N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y.
Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg,
V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O.
Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Fores-
man, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski,
B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L.
Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Na-
nayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wallingford
CT, 2004.
[47] S. Noury, X. Krokidis, F. Fuster, B. Silvi, Comput. Chem. Eng. 1999,
23, 597.
[48] S. A. Kulkarni, S. R. Gadre, Chem. Phys. Lett. 1997, 274, 255.
[49] Molden software, G. Schaftenaar, CMBI: http://www.cmbi.ru.nl/
molden/molden.html.
Received: June 17, 2010
Published online: September 24, 2010
13108
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13095 – 13108