Csp3-H bond activation with triel metals: Indium and gallium zwitterions through internal hydride abstraction in rigid salan ligands

Article abstract of DOI:10.1002/chem.201402358

The hydropyrimidine salan (salan=N,N'-dimethyl-N,N'-bis[(2-hydroxyphenyl) methylene]-1,2-diaminoethane) proteo-ligands with a rigid backbone {ON^(CH ₂)^NO}H₂ react with M(CH₂SiMe₃) ₃ (M=Ga, In) to yie

Full text of DOI:10.1002/chem.201402358

DOI: 10.1002/chem.201402358
Full Paper
&
Zwitterions
Csp3ÀH Bond Activation with Triel Metals: Indium and Gallium
Zwitterions through Internal Hydride Abstraction in Rigid Salan
Ligands**
Nicolas Maudoux,[a] Jian Fang,[b] Thierry Roisnel,[c] Vincent Dorcet,[c] Laurent Maron,[b] Jean-
FranÅois Carpentier,*[a] and Yann Sarazin*[a]
Abstract: The hydropyrimidine salan (salan=N,N’-dimethyl-
N,N’-bis[(2-hydroxyphenyl)methylene]-1,2-diaminoethane)
proteo-ligands with a rigid backbone {ON^(CH2)^NO}H2
react with M(CH2SiMe3)3 (M=Ga, In) to yield the zwitterions
{ON^(CH+)^NO}MÀ(CH2SiMe3)2 (M=Ga, 2; In, 3) by abstrac-
tion of a hydride from the ligand backbone followed by
elimination of dihydrogen. By contrast, with Al2Me6, the neu-
tral-at-metal bimetallic complex [{ON^(CH2)^NO}AlMe]2 ([1]2)
is obtained quantitatively. The formation of indium zwitter-
ions is also observed with sterically more encumbered li-
gands containing o-Me substituents on the phenolic rings,
or an N(CHPh)N moiety in the heterocyclic core. Overall, the
ease of Csp3ÀH bond activation follows the order Al !Ga<
In. Experimental data based on model complexes, XRD stud-
ies, and 2H NMR spectroscopy show that the formation of
the Ga/In zwitterion involves rapid release of SiMe4 followed
by evolution of H2, and suggest the formation of a transient
metal-hydride species. DFT calculations indicate that the sys-
tems {ON^(CH2)^NO}H2 +M(CH2SiMe3)3 (M=Al, Ga, In) all ini-
tially lead to the formation of the neutral monophenolate di-
hydrocarbyl species through a single protonolysis. From
here, the thermodynamic product, the model neutral-at-
metal complex 1, is formed in the case of aluminum after
a second protonolysis. On the other hand, lower activation
energy pathways lead to the generation of zwitterionic com-
plexes 2 and 3 in the cases of gallium and indium, and the
formation of these zwitterions obeys a strict kinetic control;
the computations suggest that, as inferred from the experi-
mental data, the reaction proceeds through an instable
metal-hydride species, which could not be isolated
synthetically.
blocks.[1] The a-alkylation of HNMe2 with alkenes catalyzed by
M(NMe2)x (M=Zr, Nb, Ta; x=4,5) was reported in the 1980s,[2]
and the hydroaminoalkylation catalyzed by d0 complexes of
group 4–5 metals is now a key reaction.[3,4] Late transition
metal catalysts are also known to activate the a-amino Csp3ÀH
bond,[5] but examples of a-CÀH addition of alkylamines to al-
kenes mediated by main-group metals are seldom. The activa-
tion of a-amino Csp3ÀH bonds leads to the formation of by-
products during the hydroamination of vinylarenes catalyzed
by alkali amides,[6] but similar reactivity is not reported for
group 13 metals: aluminum, gallium, and indium (referred to
as triel metals).
Introduction
The activation of a-amino Csp3ÀH bonds, notably by a-lithiation
and transition-metal-catalyzed activations in which an N-het-
erocycle participates as the nucleophilic coupling partner, af-
fords convenient access to nitrogen-containing building
[a] N. Maudoux, Prof. Dr. J.-F. Carpentier, Dr. Y. Sarazin
Organometallics: Materials and Catalysis Department
Institut des Sciences Chimiques de Rennes
UMR 6226 CNRS—Universitꢀ de Rennes 1
Campus de Beaulieu, 35042 Rennes (France)
E-mail: jean-francois.carpentier@univ-rennes1.fr
yann.sarazin@univ-rennes1.fr
b-Hydride elimination has long been known to be a key pro-
cess in organometallic chemistry, including for Al and related
triel elements.[7] Hydrogen transfer from the a position of nitro-
gen in 2-substituted indoline derivatives has been used to
reduce imines in metal-free transfer hydrogenation reac-
tions.[8,9] Uhl and co-workers have reported a unique case of
Csp2ÀH bond activation upon treatment of di(tert-butyl)buta-
diyne with AlR2H (R=Me, tBu).[10] However, we are not aware
of intramolecular abstraction of hydride from a-amino Csp3ÀH
moieties (an umpolung vis-ꢀ-vis the above examples of a-
amino Csp3ÀH bond activations) and concomitant generation of
[b] Dr. J. Fang, Prof. Dr. L. Maron
Laboratoire de Physique et Chimie de Nano-objets
UMR 5215 CNRS—Universitꢀ de Toulouse
135 avenue de Rangueil, 31077 Toulouse (France)
[c] Dr. T. Roisnel, Dr. V. Dorcet
Centre de Diffractomꢀtrie des Rayons X
Institut des Sciences Chimiques de Rennes
UMR 6226 CNRS—Universitꢀ de Rennes 1
35042 Rennes (France)
[**] Salan=N,N’-dimethyl-N,N’-bis[(2-hydroxyphenyl)methylene]-1,2-diamino-
ethane.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201402358.
Chem. Eur. J. 2014, 20, 7706 – 7717
7706
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
stable carbocations mediated by
triel metals, which act as hydride
acceptors.
Herein, we report such an ex-
ample and highlight the role of
the triel metal when metal hy-
drocarbyls are reacted with rigid
hydropyrimidine-bridged salan
proteo-ligands, with effective
Csp3ÀH activation mediated by
gallium and indium but not by
aluminum. Experimental data
Scheme 1. Preparation of the salan proteo-ligand {ON^(CH2)^NO}H2 and the substituted analogues {MeON^-
(CH2)^NOMe}H2 and {ON^(CHPh)^NO}H2.
combined with DFT computations demonstrate that the larger
triel atoms act as hydride acceptors to eventually release dihy-
drogen, and that the formation of zwitterions with gallium and
indium is dictated by a kinetic regime.
Results and Discussion
Experimental studies
The new salan proteo-ligand {ON^(CH2)^NO}H2 rigidified by
a hydropyrimidine core was prepared straightforwardly in 75%
yield by the Mannich reaction between salicylaldehyde and
2,2-dimethyl-1,3-diaminopropane. The substituted analogues
Scheme 2. Synthesis of neutral aluminum complexes.
{
MeON^(CH2)^NOMe}H2 and {ON^(CHPh)^NO}H2 were prepared
These observations apply to some extent to sterically more
congested ligands and/or when the hydride-donating capabili-
ty is modified. The addition of ortho-Me groups on the phenol-
ic rings (as in {MeON^(CH2)^NOMe}H2) or the introduction of a 2-
Ph group in the heterocyclic fragment (as in {ON^-
(CHPh)^NO}H2) affords rigid and bulky proteo-ligands,[16] which
upon treatment with Al2Me6 or In(CH2SiMe3)3 give the bimetal-
lic Al complex [4]2 (Scheme 2) and the zwitterions {ON^-
similarly in 81 and 90% yields, respectively, starting from the
corresponding salicylaldehyde and diamines (Scheme 1).
The stoichiometric reaction of {ON^(CH2)^NO}H2 with Al2Me6
at 708C yields [{ON^(CH2)^NO}AlMe]2 ([1]2; Scheme 2); the pro-
duction of this bimetallic complex is not unexpected in view
of recent reports on related bis(phenolate) aluminum com-
plexes.[11] By contrast, the reactions with M(CH2SiMe3)3 (M=Ga,
In) based on the larger (rionic: Al3+, 0.32 ꢂ; Ga3+, 0.47 ꢂ; In3+
,
(CPh+)^NO}InÀ(CH2SiMe3)2 (5) and
{
MeON^(CH+)^NOMe}InÀ-
0.62 ꢂ) metals neatly afford the monometallic zwitterions
{ON^(CH+)^NO}MÀ(CH2SiMe3)2 (M=Ga, 2; In, 3) through ab-
straction of a hydride from the ligand backbone and elimina-
tion of dihydrogen (experimentally detected, see below;
Scheme 3). These air-stable zwitterions feature both a carbocat-
ion stabilized by two adjacent nitrogen atoms and a formal
negative charge located on the metal center. To our knowl-
edge, such Ga-/In-mediated intramolecular Csp3ÀH activation is
not documented.[12,13] Beyond sheer acid–base Lewis adducts,
heavier triel (Ga, In) zwitterions
(CH2SiMe3)2 (6) (Scheme 3), respectively. The formation of the
indium zwitterion 6 requires more forcing conditions (90 h at
908C) than for 3 (22 h at 708C) in which the ligand is less cum-
bersome, whereas the gallium congener of 6 cannot actually
be prepared, as the reaction halts after the first protonolysis to
yield [{MeON^(CH2)^NOMe}H]Ga(CH2SiMe3)2 (7-H, see below). The
formation of 5 takes place at 808C, that is, steric considerations
are preponderant for ortho substituents on the phenolic moiet-
ies.[17]
can form through bimolecular
addition processes,[14] but intra-
molecular activation is limited to
the use of phosphine–gallanes
and –indanes for the formation
of MÀ–AuI+ species (M=Ga, In)
following chloride transfer from
gold to the triel metal.[15] We
have verified that the treatment
of Al(CH2SiMe3)3 (instead of
Al2Me6) with {ON^(CH2)^NO}H2
equally failed to return zwitter-
ionic complexes.
Scheme 3. Synthesis of zwitterionic gallium/indium complexes.
www.chemeurj.org ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2014, 20, 7706 – 7717
7707

Full Paper
The complexes have been characterized by NMR spectrosco-
py and elemental analysis, and by X-ray diffraction crystallogra-
meric and monometallic, whereas our repeated attempts only
afforded the bimetallic, dimeric [1]2.
1
phy in several cases. In the H NMR spectra of the charge-neu-
The gallium and indium complexes 2 and 3 are isostructural,
and only the latter is presented here. In the structure of 3
(Figure 2), the four-coordinated indium atom lies in a distorted
tetrahedral environment; bond lengths to the coordinated O
and C atoms are not impacted by the formal negative charge
tral aluminum complexes [1]2 and [4]2 in C6D6, the AlÀMe
group is identified by a sharp singlet at high field (d1H =À0.33
1
and À0.25 ppm, respectively). The H NMR spectra of zwitter-
ions 2, 3, and 6 exhibit a sharp, deshielded singlet at about
d=7.92–8.13 ppm characteristic of the NCHN+ hydrogen in
the carbocation. Compared with that of the proteo-ligand
{ON^(CHPh)^NO}H2, the solid-state FTIR spectrum of the
indium complex 3 shows a new, intense absorption band at
n=1674 cmÀ1, that is, in the region typical of the stretching
frequency for C=N imine bonds; the corresponding band in
the gallium analogue 2 appears at n=1682 cmÀ1.[16]
The molecular solid-state structures of [1]2, 2, 3, and 5 have
been established by X-ray diffraction crystallography. In the
structure of [1]2 depicted in Figure 1, each aluminum atom is
four-coordinated and lies in a tetrahedral environment through
Figure 2. Molecular structure of {ON^(CH+)^NO}InÀ(CH2SiMe3)2 (3) in the
solid state. Hydrogen atoms and noninteracting solvent (Et2O) are omitted
for clarity. Ellipsoids are drawn at the 50% probability level. Pertinent bond
lengths [ꢂ] and angles [8]: In(1)ÀO(11) 2.110(2), In(1)ÀO(1) 2.116(2), In(1)À
C(30) 2.167(3), In(1)ÀC(40) 2.171(3), N(8)ÀC(20) 1.323(3), N(18)ÀC(20)
1.304(3); N(18)-C(20)-N(8) 124.2(3), N(18)-C(20)-H(20) 117.9, N(8)-C(20)-H(20)
117.9, C(20)-N(8)-C(9) 120.5(2), C(20)-N(8)-C(7) 120.8(2), C(9)-N(8)-C(7)
118.7(2).
at the metal and fall in the expected range.[22] The unusually
stable half-chair conformation of the pentahydropyrimidinium
core in 3 differs from the chair conformation of the hydropyri-
midine in {ON^(CH2)^NO}H2.[16] The C(21) atom rests 0.65 ꢂ
above the perfect plane formed by C(9), N(8), C(20), N(18), and
C(19), and which also contains the C(7) and C(17) atoms. The
geometries around the now sp2-hybridized N(8), N(18), and
positively charged C(20)+ atoms are perfectly trigonal planar
(Sangles =3608). The C(20)–N(8) and C(20)–N(18) distances in 3
(1.32 and 1.30 ꢂ) are much shorter than the corresponding
ones in the {ON^(CH2)^NO}H2 (1.47 ꢂ). They in fact come close
to that for the C=N imine bond (ca. 1.28 ꢂ), thus indicating
substantial donation of the nitrogen lone pairs into the formal-
ly empty p orbital on C(20)+; this is consistent with the FTIR
spectrum for this complex (see above). The other bond lengths
in the heterocyclic core are not affected upon formation of
3.[23]
Figure 1. Molecular structure of [{ON^(CH2)^NO}AlMe]2 ([1]2) in the solid
state. Hydrogen atoms are omitted for clarity. Ellipsoids are drawn at the
50% probability level. Pertinent bond lengths [ꢂ] and angles [8]: C(1)ÀAl(1)
1.943(2), O(2)ÀAl(1) 1.748(1), O(11)ÀAl(1) 1.734(1), N(21)ÀAl(1)#1 2.008(1);
O(11)-Al(1)-O(2) 112.63(6), O(11)-Al(1)-C(1) 115.04(7), O(2)-Al(1)-C(1) 112.69(7),
O(11)-Al(1)-N(21)#1 100.11(5), O(2)-Al(1)-N(21)#1 100.30(5), C(1)-Al(1)-N(21)#1
114.52(6); #1=symmetry related.
coordination of two Ophenolate and only one Nheterocycle atoms
(N(21))[18] in addition to the alkyl moiety (C(1)); this is unusual
because aluminum–salen and –salan complexes are known to
generally adopt five-coordinated geometries.[19] The central
pyrimidine core retains the chair conformation found in the
parent proteo-ligand. The bond lengths in [1]2 are unexcep-
tional, and match those found in Gibson’s Al–salan alkyl com-
plexes featuring an N,N-disubstituted ethylenediamine
bridge,[20] and those reported by Fulton for his {ONNO}AlMe
complex supported by a piperazine-bridged dianionic salan
ligand.[21] Each {ON^(CH2)^NO}2À ligand in [1]2 binds to the
two Al atoms through AlÀOphenolate s bonds. Interestingly, Ful-
ton’s complex featuring a rigid backbone (in which the pipera-
zine adopts a boat conformation) akin to that in [1]2 is mono-
The structure of 5 (Figure 3) resembles that of 3, and indeed
the InÀC and InÀO bond lengths in the two complexes are
near-identical. The impact of the phenyl substituent on the
+
Csp2 carbocation on the overall geometry of the zwitterion is
minimal. A smaller distance between the metal atom and the
two CH2C(CH3)2CH2 carbon atoms at the back of the hydropyri-
midine ring are measured in 5 (5.38 and 5.48 ꢂ in 3; 4.27 and
4.46 ꢂ in 5). Accordingly, the angle between the two O-In-O
and C-N-C+-N-C average planes is larger in 3 (78.88) than in 5
Chem. Eur. J. 2014, 20, 7706 – 7717
www.chemeurj.org
7708
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
tensity at d1H =4.47 ppm was attributed to solubilized H2.[24]
The formation of large amounts of dihydrogen in the course of
the reaction was further corroborated by gas chromatography
1
2
analysis.[16] The monitoring by H and H NMR spectroscopies
of the reaction of the deuterated {ON^(CH2)^NO}D2 (d2H
=
10.19 ppm in C6H6; obtained by treating {ON^(CH2)^NO}Li2
with excess D2O) and In(CH2SiMe3)3 (1.0 equiv) carried out at
258C in C6H6 revealed quantitative release of Me3SiCH2D and
formation of [{ON^(CH2)^NO}D]In(CH2SiMe3)2 (d2H =9.82 ppm in
C6H6) within the first point of analysis (Scheme 4). No further
change was noted until the temperature was raised to 708C,
when quantitative production of 3 was observed over about
8 h; the evolution of HD was not detected spectroscopically.
The scenario was identical with gallium, but for the fact that
the formation of 2 was slower than that of its indium congener
and required heating to 1008C in (deuterated) THF. The reac-
tion of the bulkier proteo-ligand {MeON^(CH2)^NOMe}H2 with In-
(CH2SiMe3)3 proceeded more slowly (in comparison with the
non o-Me-substituted case) but still returned 6 (slightly conta-
minated with 9; see below), whereas with Ga(CH2SiMe3)3 it
stopped at the formation of [{MeON^(CH2)^NOMe}H]Ga-
(CH2SiMe3)2 (7-H). This gallium complex was fully characterized,
including by XRD studies (Figure 4); it is robust and does not
evolve towards zwitterionic or neutral-at-metal bis(phenolate)
species even upon heating to 1008C for 48 h. Correspondingly,
the reaction of {MeON^(CH2)^NOMe}D2 (d2H =10.32 ppm) with
Ga(CH2SiMe3)3 (1.0 equiv) yielded [{MeON^(CH2)^NOMe}D]Ga-
(CH2SiMe3)2 (7-D, d2H =9.40 ppm), see Scheme 4.[25] Finally, the
Figure 3. Molecular structure of {ON^(CPh+)^NO}InÀ(CH2SiMe3)2 (5) in the
solid state. Hydrogen atoms are omitted for clarity. Ellipsoids are drawn at
the 50% probability level. Pertinent bond lengths [ꢂ] and angles [8]: In(1)À
O(21) 2.107(1), In(1)ÀO(1) 2.109(1), In(1)ÀC(41) 2.161(2), In(1)ÀC(45) 2.171(2),
N(9)ÀC(31) 1.324(2), C(8)ÀN(9) 1.477(2), N(29)ÀC(31) 1.326(2), C(28)ÀN(29)
1.477(2); C(31)-N(9)-C(10) 121.20(12), C(31)-N(9)-C(8) 123.23(12), C(10)-N(9)-
C(8) 115.28(11).
(57.18), and the distance to the carbocation is greater in the
latter (In···C+ = 4.19 ꢂ in 5 and 4.08 ꢂ in 3). Besides, the angle
between the two mean planes defined by the phenolic rings is
wide in 3 (54.78), but it is considerably narrower in 5 (21.18).
1H NMR monitoring of the formation of 3 in C6D6 (708C) indi-
cated quantitative release of SiMe4 whereas a singlet of low in-
proteo-ligand
{Me^MeON^-
(CH2)^NOMe}H (obtained follow-
ing a seven-step synthesis)[16]
gave {Me^MeON^(CH2)^NOMe}In-
(CH2SiMe3)2 (8-Me) upon reac-
tion
with
In(CH2SiMe3)3
(1.0 equiv). This complex did
not evolve further upon heating
to 130–1508C for 12–24 h, and
in particular the abstraction of
hydride resulting in the forma-
tion of
a hydropyrimidinium
moiety was not detected. Yet,
formation of an indium zwitter-
ion is thermodynamically fa-
vored with a bisphenol, as the
equimolar reaction of 8-Me
with {ON^(CH2)^NO}H2 returned
3 through ligand exchange and
release
of
{Me^MeON^-
(CH2)^NOMe}H (Scheme 4). Note
that formation of a zwitterion
does not occur upon mixing 8-
Me and o-cresol.
The solid-state structure of
the monophenoxide gallium 7-
H (Figure 4) is informative. The
four-coordinated metal atom
rests in a near-ideal tetrahedral
Scheme 4. Intermediates and products in the synthesis of Ga/In zwitterions.
Chem. Eur. J. 2014, 20, 7706 – 7717
www.chemeurj.org
7709
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 4. Molecular structure of [{MeON^(CH2)^NOMe}H]Ga(CH2SiMe3)2 (7-H) in
the solid state. Hydrogen atoms are omitted for clarity. Ellipsoids are drawn
at the 50% probability level. Pertinent bond lengths [ꢂ] and angles [8]:
N(32)ÀGa(1) 2.148(2), O(41)ÀGa(1) 1.873(1), C(51)ÀGa(1) 1.962(2), C(55)ÀGa(1)
1.977(2); N(12)-C(4)-N(32) 112.5(2), N(12)-C(4)-H(4 A) 109.1, N(32)-C(4)-H(4 A)
109.1, N(12)-C(4)-H(4B) 109.1, N(32)-C(4)-H(4B) 109.1, H(4 A)-C(4)-H(4B) 107.8,
C(4)-N(32)-C(31) 109.4(1), C(4)-N(32)-C(33) 112.5(2), C(31)-N(32)-C(33) 111.4(2),
C(4)-N(32)-Ga(1) 108.6(1), C(31)-N(32)-Ga(1) 108.6(1), C(33)-N(32)-Ga(1)
106.1(1), O(41)-Ga(1)-C(51) 108.14(8), O(41)-Ga(1)-C(55) 107.91(8), C(51)-Ga(1)-
C(55) 126.68(9), O(41)-Ga(1)-N(32) 95.05(6), C(51)-Ga(1)-N(32) 106.86(7), C(55)-
Ga(1)-N(32) 107.62(7).
Figure 5. Molecular structure of the mixed zwitterionic/neutral-at-metal bi-
metallic indium complex 9 in the solid state, showing only the main compo-
nent of the disordered SiMe3 fragment. Hydrogen atoms and noninteracting
solvent (toluene) are omitted for clarity. Ellipsoids are drawn at the 50%
probability level. Pertinent bond lengths [ꢂ] and angles [8]: In(1)ÀO(11)
2.050(2), In(1)ÀO(36) 2.058(2), In(1)ÀO(41) 2.087(2), In(1)ÀC(1) 2.142(3), In(2)À
O(61) 2.095(2), In(2)ÀC(51) 2.153(4), In(2)ÀC(55) 2.155(4), In(2)ÀN(70) 2.357(3);
C(27)-N(20)-C(21) 119.6(3), C(27)-N(20)-C(19) 121.4(3), C(21)-N(20)-C(19)
118.3(3), C(27)-N(26)-C(25) 121.9(3), C(27)-N(26)-C(28) 119.4(3), C(25)-N(26)-
C(28) 118.6(3).
geometry in this inert complex. The coordination sphere
around gallium comprises the two alkyl groups (C(51) and
C(55)), the Ophenolate atom (O(41)), and a single nitrogen atom
from the hydropyrimidine cycle (N(32)). In particular, the po-
tentially reactive phenol moiety does not bind to the metal
atom, but is instead very remote, HOphenol···Ga=6.41 ꢂ; such an
arrangement does not favor further intramolecular reactivity,
towards either a zwitterion or a neutral-at-metal complex, al-
though intermolecular reactivity remains feasible with such
complexes. For instance, the mixed bimetallic indium complex
direct vicinity of the metal center is required to remove the hy-
dride from the hydropyrimidine core in an intramolecular pro-
cess; 3) the passage from complexes such as 8-H to the corre-
sponding zwitterions (6 in this case) does not involve a stable,
long-lived metal-hydride transient species, from which H2
would be delivered by recombination with the remaining phe-
nolic proton; 4) in order to generate coordinatively saturated
tetrahedral In (or Ga) complexes such as 7-H and the like, bind-
{
MeON^(CH+)^NOMe}InÀ(CH2SiMe3)–{MeON^(CH2)NOMe}–In-
(CH2SiMe3)2 (9) having zwitterionic and neutral-at-metal In
atoms was isolated in small quantity during the synthesis of 6.
The arrangement of the salan ligand linking the two indium
atoms in the byproduct
9
(Figure 5) is very similar to that
of the ligand in 7-H whereas
the geometrical features of the
zwitterionic part match those in
3; its formation is attributed to
the deleterious reaction of 6
with 8-H,[26] the structure of
which must be similar to that of
7-H (Scheme 5). Attempts to
isolate substantial amounts of
the putative 8-H and of 9 were
unsuccessful.
Based on all the above exper-
imental observations, we con-
cluded that: 1) elimination of
SiMe4 by protonolysis occurs
before the abstraction of hy-
dride; 2) the presence of
a second phenolic moiety in the
Scheme 5. Proposed pathway for the formation of the mixed zwitterionic/neutral-at-metal bimetallic indium com-
plex 9.
Chem. Eur. J. 2014, 20, 7706 – 7717
www.chemeurj.org
7710
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ing of the phenol OH moiety is less favorable than that of
a Nhydropyrimidine atom; and 5) the ability to form zwitterions
varies with metal size and polarizing ability: the observed reac-
tivity trend contrasts with these three metals’ matching ability
to promote halide abstraction in phosphine–trielane gold(I)
complexes.[15,27]
tion state AlR1T3, DGꢀ =À0.9 kcalmolÀ1); moreover, the com-
puted product AlR4T2f is also much more stable than AlR1T3f,
with a difference of 18.0 kcalmolÀ1 in favor of the former. Note
that only the monometallic AlR4T2 was modeled for the econo-
my of computing time, and the experimentally observed bi-
metallic [1]2 is likely to be substantially more stable, even if
the energy of the associated transition state is expected to be
of the same magnitude as that determined for AlR4T2. DFT cal-
culations thus indicate that the formation of a zwitterionic alu-
minum complex is unfavorable, which agrees with the experi-
mental observations. Note that the transition state AlR4T2 lead-
ing to the formation of the neutral-at-metal product by proto-
nolysis of an AlÀCH3 bond is predicted to consist of a five-co-
ordinated aluminum atom, with coordination of the phenolic
OH moiety onto the metal.
Computational studies
Theoretical computations performed on the {ON^(CH2)^NO}H2/
MR3 systems corroborate these experimental observations.[16]
In the case of aluminum (Figure 6), starting from the entry
point E0 (viz. the proteo-ligand+Al2Me6), the first protonolysis
leading to AlR1T1f takes place relatively easily and proceeds
via a transition state AlR1T1 (the monophenol–AlMe3 adduct)
of relatively low free energy, approximately 17.4 kcalmolÀ1
;
The scenario is different for gallium (Figure 7). In this case
the formation of the four-coordinated neutral-at-metal com-
plex GaR4T2f (a putative monometallic complex of formula
{ON^(CH2)^NO}GaCH2SiMe3) is again thermodynamically slight-
ly favored over the formation of the zwitterionic GaR1T3f (that
is, 2 in its optimized geometry). However, the activation barrier
leading to the formation of the thermodynamically more
stable complex from the four-coordinated intermediate
GaR1T1f (Figure 7, plain line) is now much greater (DGꢀ =
30.8 kcalmolÀ1) than that affording the zwitterionic hydride
GaR1T2f (DGꢀ =18.9 kcalmolÀ1, dashed line). From GaR1T2f,
formation of the final zwitterion GaR1T3f occurs readily (DGꢀ =
12.9 kcalmolÀ1).
note that the aluminate intermediate AlR1T1f with a protonated
ammonium is kinetically more available than its neutral mono-
phenolate derivative (i.e., the aluminum analogue of 7-H).
From AlR1T1f, further reactivity can potentially follow two
pathways. Computations predict that the route leading to the
formation of the neutral-at-metal AlR4T2f (a product of general
formula {ON^(CH2)^NO}AlMe, i.e., the monometallic version of
[1]2; Figure 6, plain line) is favored over that leading to the
zwitterionic AlR1T3f (viz. the Al analogue of 2 and 3; dashed
line) both on kinetic and thermodynamic grounds. The transi-
tion state AlR4T2 leading to the neutral-at-metal AlR4T2f is of
lower energy (DGꢀ =À11.6 kcalmolÀ1) than that leading to
AlR1T2f (a putative Al-hydride intermediate; transition state
AlR1T2, DGꢀ =À7.2 kcalmolÀ1) and ultimately AlR1T3f (transi-
On the basis of these computations, one can therefore
assume that the formation of zwitterion 2 observed experi-
mentally occurs owing to
a largely favorable kinetic con-
trol. Note that: 1) starting from
GaR1T1f, attempts to converge
towards a stable four- or five-
coordinated neutral Ga···OHphenol
adduct (a requisite for the pro-
duction of a neutral-at-metal Ga
complex akin to GaR4T2f) failed,
as the system systematically re-
laxed back to more stable four-
coordinated Ga···N species free
of interactions with the phenol-
ic OH moiety; and 2) the geom-
etry around the tetrahedral
metal in the intermediate
GaR1T1f is similar to that ob-
served experimentally by X-ray
diffraction crystallography in
the related 7-H.
Computations suggest that
kinetic control also governs the
formation of the indium zwitter-
ion 3 (Figure 8). Whereas the
neutral-at-metal InR4T2f is esti-
mated to be nominally more
stable than the zwitterion
Figure 6. Computed pathways for the formation of aluminum zwitterionic (dashed line) versus neutral-at-metal
model monometallic complex (plain line); E0 =1= Al2Me6 +{ON^(CH2)^NO}H2.
2
Chem. Eur. J. 2014, 20, 7706 – 7717
www.chemeurj.org
7711
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
energy, DGꢀ =24.3 kcalmolÀ1),
we have been unable to identi-
fy resilient In···HOphenol adducts.
Figure S26 (see the Support-
ing Information) summarizes
the outcome of DFT computa-
tions, highlighting the different
paths leading to neutral-at-
metal aluminum and zwitterion-
ic gallium and indium com-
plexes. These results are fully
compatible with the experimen-
tal data. All proceed through
a first four-coordinated mono-
phenolate neutral complex, in
which
coordination
of
a Nhydropyrimidine atom is favorable
over that of the HOphenolic
moiety. Note that in the opti-
mized metal-hydride intermedi-
ates GaR1T2f and InR1T2f, the
phenolic proton is ideally ori-
ented towards the MÀH hydritic
moiety, which facilitates the
elimination of H2. The calcula-
tions suggest that these inter-
mediates are less stable than
their neutral-at-metal monophe-
nolate parents GaR1T1f and
InR1T1f, respectively; this is con-
sistent with the fact that the for-
mation of an In-hydride species
could not be detected upon
heating 8-Me, for which the for-
mation of a zwitterion is pre-
cluded owing to the absence of
a second phenolic proton.
Figure 7. Computed pathways for the formation of gallium zwitterionic (dashed line) versus neutral-at-metal
model monometallic complex (plain line); E0 =Ga(CH2SiMe3)3 +{ON^(CH2)^NO}H2.
The DFT computations are
also in agreement with the ex-
periments in the case of Ga-
(CH2SiMe3)3 and the bulkier
ortho-Me-substituted
proteo-
ligand
{
MeON^(CH2)^NOMe}H2.
Figure 8. Computed pathways for the formation of indium zwitterionic (dashed line) versus neutral-at-metal
model monometallic complex (plain line); E0 =In(CH2SiMe3)3 +{ON^(CH2)^NO}H2.
The robust 7-H (the analogue to
GaR1T1f in Figures 7 and S26 in
the Supporting Information)
formed with Ga(CH2SiMe3)3 does
InR1T3f (that is, the geometrically optimized version of 3) not evolve further towards either the expected zwitterion or
within the limits of the calculation method, from the tetrahe- a neutral-at-metal complex (see Scheme 4). The computed acti-
dral N-coordinated intermediate InR1T1f the activation energy vation barriers for both reactions (DGꢀ =30.0 and 40.9 kcal
required to yield the former is again substantially higher molÀ1, respectively) are much greater than those for the corre-
(Figure 8, plain line; DGꢀ =29.8 kcalmolÀ1) than that for the sponding events calculated above for the less bulky {ON^-
generation of the zwitterionic [In]-H transient species InR1T2f (CH2)^NO}H2, and kinetic control thus prevents further evolu-
(DGꢀ =23.5 kcalmolÀ1) and, from there, to the final zwitterion tion of the system in either direction. Regarding the system In-
InR1T3f following release of dihydrogen (DGꢀ =14.7 kcalmolÀ1, (CH2SiMe3)3/{MeON^(CH2)^NOMe}H2, from the first neutral-at-
dashed line). As for Ga, starting from InR1T1f (which is accessed metal monophenolate intermediate both the formation of the
from the entry point via a transition state relatively high in free neutral-at-metal (DG=À34.0 kcalmolÀ1) and zwitterion
6
Chem. Eur. J. 2014, 20, 7706 – 7717
www.chemeurj.org
7712
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
(DG=À15.5 kcalmolÀ1) are thermodynamically feasible, but
the production of the zwitterion is actually favored because of
a lower activation energy (DGꢀ =21.0 and 18.0 kcalmolÀ1, re-
spectively).
process and eventual formation of a carbocationic moiety
opens the route towards nucleophilic functionalization of the
pentahydropyrimidinium core.[32]
Experimental Section
Conclusion
All manipulations were performed under inert atmosphere using
standard Schlenk techniques or in a dry, solvent-free glovebox (Ja-
comex; O2 <1 ppm, H2O <5 ppm). AlMe3 (2.0m in toluene, Al-
drich) was used as purchased. THF was distilled under argon from
Na/benzophenone prior to use. Other solvents (pentane, toluene,
dichloromethane, Et2O) were collected from MBraun SPS-800 purifi-
cation alumina columns. Deuterated solvents (Eurisotop, Saclay,
France) were stored in sealed ampoules over activated 3 ꢂ molecu-
lar sieves and degassed by several freeze–thaw cycles. Ga-
(CH2SiMe3)3 and In(CH2SiMe3)3 were prepared following literature
procedures.[33]
Indium and, to a lesser extent, gallium hydrocarbyls afford
original zwitterionic complexes upon treatment with salan
proteo-ligands rigidified by a cyclic hydropyrimidine core
through an unusual process of intramolecular hydride abstrac-
tion from a-amino Csp3ÀH moieties. These Ga and In zwitterions
are obtained by release of H2, which was detected experimen-
tally. Since they are structurally and configurationally very rigid,
it is tempting to consider the whole process as the reverse of
H2 activation by frustrated Lewis pairs (FLPs),[28] and it is partic-
ularly related to the metal FLPs described by Wass and co-
workers[29] and Uhl and co-workers.[30] Major differences in the
reactivity are observed on moving from aluminum, which
simply affords the (expected) neutral-at-metal complexes by
double protonolysis, to gallium and indium, for which the ease
of hydride abstraction, release of dihydrogen, and formation of
zwitterionic species varies with the metal size and correspond-
ing polarizing ability. It is intriguing that the ability to carry out
this hydride abstraction is apparently inversely proportional to
the MÀH bond enthalpies for Al (67 kcalmolÀ1), Ga (62 kcal
molÀ1), and In (54 kcalmolÀ1).[31] These findings contrast with
previous reports in which the three metals displayed compara-
ble ability in promoting activation reactions.[12,16,17] DFT compu-
tations suggest that the outcome of the reactions is governed
by a strict kinetic control, and a plausible reaction pathway
based on experimental and DFT results is illustrated in
Scheme 6 for the system M(CH2SiMe3)3/{ON^(CH2)^NO}H2 (M=
Ga, In), which generates the zwitterionic complexes 2 and 3.
Despite repeated attempts and the bespoke synthesis of the
monophenol {Me^MeON^(CH2)^NOMe}H, we have not been able
to isolate metal-hydride intermediates or models; this is in line
with the high reactivity of such hydrometalate species suggest-
ed by the DFT computations. Yet, this unusual CÀH activation
NMR spectra were recorded on Bruker AM-400 and AM-500 spec-
trometers. All chemical shifts (given in ppm) were determined
using residual signals of the deuterated solvents and were calibrat-
ed versus SiMe4. Assignment of the signals was carried out using
1D (1H, 13C{1H}) and 2D (COSY, HMBC, HMQC) NMR experiments.
Elemental analyses were performed on a Carlo Erba 1108 Elemental
Analyser instrument at the London Metropolitan University by Ste-
phen Boyer and were the average of a minimum of two independ-
ent measurements.
FTIR spectra were recorded between 400 and 4000 cmÀ1 as
a powder on a Shimadzu IRAffinity-1 spectrometer equipped with
an attenuated total reflectance module.
Crystals of {ON^(CH2)^NO}H2, [1]2, 2, 3·Et2O, and 7-H suitable for X-
ray diffraction analysis were obtained by recrystallization of the pu-
rified compound; crystals of 9 were isolated as byproducts during
the synthesis of 6. Diffraction data were collected at 150(2) K by
using a Bruker APEX CCD diffractometer with graphite-monochro-
mated MoKa radiation (l=0.71073 ꢂ). A combination of w and
F scans was carried out to obtain at least a unique data set. The
crystal structures were solved by direct methods, and remaining
atoms were located from difference Fourier synthesis followed by
full-matrix least-squares refinement based on F2 (programs SIR97
and SHELXL-97).[34] Carbon- and oxygen-bound hydrogen atoms
were placed at calculated positions and forced to ride on the at-
tached atom. All non-hydrogen atoms were refined with anisotrop-
ic displacement parameters. The
locations of the largest peaks in
the final difference Fourier map
calculation as well as the magni-
tude of the residual electron den-
sities were of no chemical signifi-
cance. Relevant collection and re-
finement data are given in the
Supporting Information. CCDC-
980041
({ON^(CH2)^NO}H2),
980042 ([1]2), 980043 (2), 980044
(3·Et2O), 980045 (7-H), 987689
({tBuON^(CH2)^NOtBu}H2),
987690
(5), 987691 ({ON^(CHPh)^NO}H2),
987692 ({MeON^(CHPh)^NOMe}H2),
987693
987694
({MeON^(CH2)^NOMe}H2),
({{MeON^(CH2)^NOMe}Li–
Li{THF}2}2), and 987695 (9·toluene)
contain the supplementary crystal-
lographic data for this paper.
Scheme 6. Plausible reaction pathway for the formation of zwitterionic gallium/indium complexes.
Chem. Eur. J. 2014, 20, 7706 – 7717
www.chemeurj.org
7713
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_
request/cif.
124.6 (Carom-CH2N), 120.7 (CaromH), 118.3 (CaromH), 74.2 (NCH2N), 63.9
(NCH2-Carom), 58.9 (CH2C(CH3)2CH2), 31.4 (CH2C(CH3)2CH2), 27.3
(CH2CCH3(CH3)CH2), 26.2 ppm (CH2CCH3(CH3)CH2), no visible Al-
CH3; elemental analysis calcd (%) for C42H54Al2N4O4 (732.85 gmolÀ1):
C 68.8, H 7.4, N 7.6; found: C 68.9, H 7.1, N 7.9.
{ONNO}H2
2,2-Dimethylpropane-1,3-diamine (5.1 g, 50 mmol) was dissolved in
ethanol (120 mL) and salicylaldehyde (10.7 mL, 100 mmol) was
added slowly. The reaction mixture was heated at reflux for 1 h.
After cooling to 08C, NaBH4 was added (7.6 g, 200 mmol; 4 equiv)
in portions and the reaction mixture was stirred at room tempera-
ture for 1 h. Water (200 mL) was added and the mixture was stirred
for another hour. The product was extracted with CH2Cl2, the or-
ganic phase was dried over MgSO4, and the volatiles were pumped
off to give {ONNO}H2 as a colorless solid. Yield: 15.2 g, 97%.
{ON^(CH+)^NO}GaÀ(CH2SiMe3)2 (2)
{ON^(CH2)^NO}H2 (140 mg, 0.43 mmol) and Ga(CH2SiMe3)3 (142 mg,
0.43 mmol) were dissolved in THF (10 mL). The reaction mixture
was stirred at 1008C for 90 h. Then, the volatiles were pumped off,
the product was extracted with diethyl ether (2ꢃ5 mL), the solvent
was removed, and the solid was washed with pentane (3ꢃ3 mL)
and dried under high dynamic vacuum affording 2 as a colorless
solid. Yield: 112 mg, 46%. 1H NMR (C6D6, 315 K, 500.13 MHz): d=
7.92 (s, 1H; NCHN), 7.31 (td, 3JHH =8.3, 4JHH =1.9 Hz, 2H; CaromH),
7.03 (dd, 3JHH =8.3, 4JHH =1.1 Hz, 2H; CaromH), 6.82 (dd, 3JHH =7.3,
4JHH =1.9 Hz, 2H; CaromH), 6.66 (td, 3JHH =7.3, 4JHH =1.1 Hz, 2H;
CaromH), 3.76 (s, 4H; NCH2-Carom), 1.93 (s, 4H; CH2C(CH3)2CH2), 0.45 (s,
18H; Si(CH3)3), 0.03 (s, 6H; CH2C(CH3)2CH2), À0.08 ppm (s, 4H;
{
MeONNOMe}H2 and {tBuONNOtBu}H2 were synthesized by following
the same procedure with the appropriate substituted salicylalde-
hyde.
{ON^(CH2)^NO}H2
GaCH2Si); 13C{1H} NMR (C6D6, 298 K, 125.75 MHz): d=165.3 (Carom
-
{ONNO}H2 (6.00 g, 19.1 mmol) was dissolved in acetonitrile (ca.
150 mL), and formaldehyde (14.3 mL of a 37% aqueous solution,
191 mmol, 10 equiv) was added by syringe. The reaction mixture
was stirred for 24 h at room temperature, which resulted in the
precipitation of a colorless solid. The solid was isolated by filtration
and washed with cold acetonitrile (3ꢃ15 mL) and pentane (2ꢃ
5 mL). {ON^(CH2)^NO}H2 was obtained as a colorless powder upon
OGa), 153.7 (NCHN), 130.9 (CaromH), 130.2 (CaromH), 120.4 (CaromH),
120.2 (Carom-CH2N), 114.1 (CaromH), 57.8 (NCH2-Carom), 51.9 (CH2C-
(CH3)2CH2), 26.0 (CH2C(CH3)2CH2), 22.8 (CH2C(CH3)2CH2), 2.5 (Si(CH3)3),
À0.4 ppm (InCH2Si); 15N NMR (C6D6, 298 K, 40.5 MHz): 113.4 ppm
(NCH2-Carom); elemental analysis calcd (%) for C28H45GaN2O2Si2
(567.56 gmolÀ1): C 59.3, H 8.0, N 4.9; found: C 59.0, H 8.0, N 5.1.
1
drying to constant weight. Yield: 4.68 g, 75%. H NMR (C6D6, 343 K,
{ON^(CH+)^NO}InÀ(CH2SiMe3)2 (3)
500.13 MHz): d=9.94 (s, 2H; Carom-OH), 7.09 (t, 3JHH =7.7 Hz, 2H;
3
3
CaromH), 7.04 (d, JHH =7.3 Hz, 2H; CaromH), 6.78 (d, JHH =7.9 Hz, 2H;
CaromH), 6.71 (t, 3JHH =7.3 Hz, 2H; CaromH), 3.17 (s, 4H; NCH2-Carom),
2.72 (br s, 2H; NCH2N), 1.83 (s, 4H; CH2C(CH3)2CH2), 0.71 ppm (s,
{ON^(CH2)^NO}H2 (200 mg, 0.61 mmol) and In(CH2SiMe3)3 (230 mg,
0.61 mmol) were dissolved in THF (10 mL). The reaction mixture
was stirred at 708C for 22 h. The volatiles were then pumped off to
afford a solid, which was dried under vacuum. The title product
was extracted using diethyl ether (3ꢃ3 mL), washed with pentane
(2ꢃ3 mL), and dried under high vacuum. Yield: 288 mg, 77%.
1H NMR (C6D6, 315 K, 400.16 MHz): d=7.96 (s, 1H; NCHN), 7.27–
6H; CH3); 13C{1H} NMR (C6D6, 343 K, 125.75 MHz): d=158.3 (Carom
-
OH), 129.1 (CaromH), 128.4 (CaromH), 120.6 (Carom-CH2N), 118.9 (CaromH),
116.6 (CaromH), 74.5 (NCH2N), 63.4 (CH2C(CH3)2CH2), 58.5 (NCH2-Carom),
30.4 (CH2C(CH3)2CH2), 25.1 ppm (CH2C(CH3)2CH2); ESI-HRMS: m/z
calcd for [M+H]+: 327.20725; found: 327.2070 (1 ppm); elemental
analysis calcd (%) for C20H26N2O2 (326.46 gmolÀ1): C 73.6, H 8.0, N
8.6; found: C 73.4, H 7.7, N 8.4. The other proteo-ligands were ob-
tained by following a similar procedure using various aldehydes
and ortho-substituted phenols.[16]
3
4
7.25 (m, 2H; CaromH), 6.88 (dd, JHH =8.2, JHH =1.0 Hz, 2H; CaromH),
3
4
3
6.82 (dd, JHH =7.26, JHH =1.82 Hz, 2H; CaromH), 6.60 (td, JHH =7.28,
4JHH =1.11 Hz, 2H; CaromH), 3.82 (s, 4H; NCH2-Carom), 1.99 (s, 4H;
CH2C(CH3)2CH2), 0.39 (s, 18H; Si(CH3)3), 0.07 (s, 6H; CH2C(CH3)2CH2),
0.01 ppm (s, 4H; InCH2Si); 13C{1H} NMR (C6D6, 315 K, 100.62 MHz):
d=167.2 (Carom-OIn), 153.0 (NCHN), 131.0 (CaromH), 130.4 (CaromH),
120.3 (Carom-CH2N), 120.2 (CaromH), 113.2 (CaromH), 57.5 (NCH2-Carom),
52.0 (CH2C(CH3)2CH2), 25.9 (CH2C(CH3)2CH2), 22.7 (CH2C(CH3)2CH2),
2.5 (Si(CH3)3), 0.1 ppm (InCH2Si); elemental analysis calcd (%) for
C28H45InN2O2Si2 (612.66 gmolÀ1): C 54.9, H 7.4, N 4.7; found: C 54.0,
H 7.6, N 4.9. Satisfactory elemental analysis was not obtained for C,
most likely because of the presence of a substantial amount of Si
leading to the formation of SiC.
[{ON^(CH2)^NO}AlMe]2 ([1]2)
{ON^(CH2)^NO}H2 (300 mg, 0.92 mmol) was dissolved in toluene
(15 mL) by heating the solution at 608C. Then, this solution was
added dropwise to AlMe3 (0.54 mL of a 2.0m toluene solution,
0.92 mmol) diluted in toluene (2 mL). The reaction mixture was
stirred at 708C for 7 h. The volatiles were then pumped off to give
a colorless solid, which was washed with pentane (3ꢃ7 mL). The
resulting colorless powder was dried under high vacuum. Finally,
the expected product was crystallized from hot benzene by slow
decrease of the temperature, from 908C to room temperature.
Yield: 650 mg, 97%. This reaction was also carried out in THF with
[{MeON^(CH2)^NOMe}AlMe]2 ([4]2)
{
MeON^(CH2)^NOMe}H2 (100 mg, 0.28 mmol) was dissolved in tolu-
ene (4 mL), a 2.0m AlMe3 solution in toluene (140 mL, 0.28 mmol)
was diluted in toluene (1 mL), and then the solution of proligand
was added dropwise onto the solution of AlMe3 at room tempera-
ture resulting in the release of methane. The reaction mixture was
stirred at 708C for 7 h. After cooling the reaction medium to room
temperature, the volatiles were pumped off. The resulting solid
was washed with pentane (2ꢃ1 mL) and dried under dynamic
vacuum for several hours. 1H NMR (C6D6, 298 K, 500.13 MHz): d=
7.15–7.13 (m, 4H; CaromH), 6.74–6.68 (m, 8H; CaromH), 4.01 and 2.57
1
the same outcome. H NMR (C6D6, 343 K, 500.13 MHz): d=7.14 (dd,
3JHH =8.5, 4JHH =0.9 Hz, 4H; CaromH), 7.07 (d, 3JHH =8.5 Hz, 4H;
3
4
3
CaromH), 6.74 (dd, JHH =7.2, JHH =0.9 Hz, 4H; CaromH), 6.66 (td, JHH
=
=
7.2, 4JHH =0.9 Hz, 4H; CaromH), 3.96 and 2.62 (AX system, JHH
2
2
11.6 Hz, 4H; NCH2N), 3.62 and 3.02 (AB system, JHH =13.5 Hz, 8H;
NCH2-Carom), 2.46 and 1.83 (AX system, 2JHH =12.8 Hz, 8H; CH2C-
(CH3)2CH2), 0.73 (s, 6H; CH2CCH3(CH3)CH2), 0.39 (s, 6H; CH2CCH3-
(CH3)CH2), À0.33 ppm (s, 6H; Al-CH3); 13C{1H} NMR (C6D6, 343 K,
125.75 MHz): d=160.2 (Carom-OAl), 130.6 (CaromH), 129.4 (CaromH),
2
(AX system, JHH =11.5 Hz, 4H; NCH2N), 3.55 and 2.98 (AX system,
Chem. Eur. J. 2014, 20, 7706 – 7717
www.chemeurj.org
7714
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
2JHH =13.7 Hz, 8H; NCH2-Carom), 2.50 (s, 12H; Carom-CH3), 2.35 and
1.70 (AX system, 2JHH =12.9 Hz, 8H; CH2C(CH3)2CH2), 0.68 (s, 6H;
CH2CCH3(CH3)CH2), 0.31 (s, 6H; CH2CCH3(CH3)CH2), À0.25 ppm (s,
[{MeON^(CH2)^NOMe}H]Ga(CH2SiMe3)2 (7-H)
MeON^(CH2)^NOMe}H2 (250 mg, 0.71 mmol) and Ga(CH2SiMe3)3
{
(235 mg, 0.71 mmol) were dissolved in THF (9 mL) and heated at
708C for 50 h. The volatiles were next pumped off to afford
a sticky solid. Stripping with pentane (3ꢃ2 mL) followed by drying
in vacuo afforded 7-H as a white solid. Yield: 400 mg, 95%.
1H NMR (C6D6, 298 K, 500.13 MHz): d=9.49 (s, 1H; CaromOH), 7.19 (d,
6H; Al-CH3); 13C{1H} NMR (C6D6, 298 K, 125.75 MHz): d=158.3 (Carom
-
OAl), 131.6 (CaromH), 128.3 (Carom-CH3), 127.4 (CaromH), 123.8
(CaromCH2N), 118.0 (CaromH), 74.0 (NCH2N), 63.4 (CH2C(CH3)2CH2), 58.9
(NCH2-Carom), 31.2 (CH2C(CH3)2CH2), 27.1 (CH2CCH3(CH3)CH2), 25.9
(CH2CCH3(CH3)CH2), 17.1 (Carom-C(CH3), À11.6 ppm (br, Al-CH3); ele-
mental analysis calcd (%) for C46H62Al2N4O4 (788.97 gmolÀ1): C 70.0,
H 7.9, N 7.1; found: C 69.9, H 7.8, N 7.0.
3
3JHH =7.2 Hz, 1H; CaromH), 7.05 (d, JHH =7.5 Hz, 1H; CaromH), 6.92 (d,
3
3JHH =7.2 Hz, 1H; CaromH), 6.74 (t, JHH =7.5 Hz, 1H; CaromH), 6.66 (m,
2H; CaromH), 4.16 (s, 2H; NCH2-Carom), 3.55 and 2.73 (AX system,
2
2JHH =11.1 Hz, 2H; NCH2N), 3.10 and 2.73 (AX system, JHH =13.4 Hz,
2H; CH2C(CH3)2CH2), 2.58 and 2.25 (AB system, 2JHH =14.1 Hz, 2H;
NCH2-Carom), 2.46 (s, 3H; Carom-CH3), 2.43 (s, 3H; Carom-CH3), 2.25 and
1.23 (AX system, 2JHH =11.6 Hz, 2H; CH2C(CH3)2CH2), 0.92 (s, 3H;
CH2CCH3(CH3)CH2), 0.33 (s, 9H; Si(CH3)3), 0.27 (s, 9H; Si(CH3)3), 0.21
(s, 3H; CH2CCH3(CH3)CH2), À0.40 (m, 2H; GaCH2Si), À0.51 ppm (m,
2H; GaCH2Si); 13C{1H} NMR (C6D6, 298 K, 125.75 MHz): d=161.4
(Carom-OGa), 155.9 (Carom-OH), 131.9 (CaromH), 131.3 (CaromH), 129.0
(CaromH), 128.7 (Carom-CH3), 127.0 (CaromH), 125.6 (Carom-CH3), 119.2
(CaromH), 119.6 (Carom-CH2N), 118.4 (Carom-CH2N), 116.4 (CaromH), 72.1
(NCH2N), 63.6 (CH2C(CH3)2CH2), 58.8 (NCH2-Carom), 58.4 (CH2C-
(CH3)2CH2), 57.4 (N’CH2-Carom), 30.5 (CH2C(CH3)2CH2), 28.2 (CH2CCH3-
(CH3)CH2), 27.4 (CH2CCH3(CH3)CH2), 17.5 (Carom-CH3), 16.0 (Carom-CH3),
2.5 (Si(CH3)3), 2.4 (Si(CH3)3), À0.3 ppm (GaCH2Si); elemental analysis
calcd (%) for C30H51GaN2O2Si2 (597.63 gmolÀ1): C 60.3, H 8.6, N 4.7;
found: C 59.9, H 8.7, N 4.6.
{ON^(CPh+)^NO}InÀ(CH2SiMe3)2 (5)
{ON^(CHPh)^NO}H2 (250 mg, 0.62 mmol) and In(CH2SiMe3)3
(233 mg, 0.62 mmol) were dissolved in THF (8 mL) and stirred at
808C. After 3.5 days, the volatiles were removed under vacuum.
The isolated solid was washed with pentane (3ꢃ3 mL) and dried
under high vacuum, affording 5 as a colorless solid. Yield: 275 mg,
66%. 1H NMR (C6D6, 298 K, 500.13 MHz): d=9.25 (dt, 3JHH =7.9,
4JHH =1.0 Hz, 1H; CaromH), 7.48 (td, 3JHH =7.5, 4JHH =0.9 Hz, 1H;
3
4
3
CaromH), 7.28 (td, JHH =7.7, JHH =1.8 Hz, 2H; CaromH), 7.11 (tt, JHH
=
4
3
4
7.5, JHH =1.2 Hz, 1H; CaromH), 7.04 (td, JHH =7.5, JHH =0.9 Hz, 1H;
CaromH), 6.80 (dd, 3JHH =7.9, 4JHH =1.0 Hz, 2H; CaromH), 6.70 (dd,
4
3JHH =7.5, JHH =1.8 Hz, 2H; CaromH), 6.60 (m, 3H; CaromH), 4.74 and
2
2.99 (AX system, JHH =11.7 Hz, 4H; NCH2-Carom), 2.79 and 2.19 (AB
system, 2JHH =12.4 Hz, 4H; CH2C(CH3)2CH2), 0.51 (s, 9H; Si(CH3)3),
0.49 (s, 3H; CH2CCH3(CH3)CH2), 0.35 (s, 9H; Si(CH3)3), 0.24 (s, 2H;
InCH2Si), 0.16 (s, 3H; CH2CCH3(CH3)CH2), À0.40 ppm (s, 2H; InCH2Si);
13C{1H} NMR (C6D6, 298 K, 125.75 MHz): d=166.9 (Carom-OIn), 162.8
(i-C6H5), 133.6 (CaromH), 131.0 (2ꢃCaromH), 130.8 (NC(Ph)N), 130.7
[{MeON^(CH2)^NOMe}D]Ga(CH2SiMe3)2 (7-D)
Compound 7-D was synthesized quantitatively in a J-Young NMR
tube as described for 7-H starting from [{MeON^(CH2)^NOMe}D2 and
Ga(CH2SiMe3)3 in C6D6. All resonances matched those of 7-H except
for CaromOD. 2H NMR (C6H6, 298 K, 61.42 MHz): d=9.32 ppm (br,
Carom-OD).
(CaromH), 129.7 (CaromH), 128.4 (CaromH), 126.7 (CaromH), 121.4 (Carom
-
CH2N), 120.4 (CaromH), 114.1 (CaromH), 58.5 (NCH2-Carom), 55.7 (CH2C-
(CH3)2CH2), 26.8 (CH2C(CH3)2CH2), 24.5 (CH2CCH3(CH3)CH2), 22.8
(CH2CCH3(CH3)CH2), 2.8 (Si(CH3)3), 2.8 (Si(CH3)3), 2.3 (InCH2Si),
À1.2 ppm (InCH2Si); elemental analysis calcd (%) for C34H49InN2O2Si2
(668.75 gmolÀ1): C 59.3, H 7.2, N 4.1; found: C 59.4, H 7.1, N 4.2.
{MeMeON^(CH2)^NOMe}In(CH2SiMe3)2 (8-Me)
{Me^MeON^(CH2)^NOMe}H2 (200 mg, 0.54 mmol) and In(CH2SiMe3)3
(203 mg, 0.54 mmol) were dissolved in THF (6 mL) and the reaction
medium was stirred at 708C for 20 h. After cooling to room tem-
perature, the volatiles were pumped off under vacuum. The result-
ing solid was purified by stripping with pentane (3ꢃ2 mL) and
dried under dynamic vacuum to afford 8-Me as a viscous oil. Yield:
{
MeON^(CH+)^NOMe}InÀ(CH2SiMe3)2 (6)
MeON^(CH2)^NOMe}H2 (185 mg, 0.52 mmol) and In(CH2SiMe3)
{
1
3
(196 mg, 0.52 mmol) were dissolved in THF (5 mL) and the reaction
mixture heated at 908C for 90 h. After cooling to room tempera-
ture, the volatiles were pumped off, and the resulting solid was
washed with pentane (2ꢃ3 mL) and dried under dynamic vacuum
to afford a colorless powder. Despite several attempts at recrystalli-
zation, 6 could not be obtained with purity higher than 95%;
a minor impurity, most likely corresponding to 9, was always de-
tected spectroscopically. 1H NMR (C6D6, 298 K, 500.13 MHz): d=
8.13 (s, 1H; NCHN), 7.28 (dd, 3JHH =7.3, 4JHH =1.8 Hz, 2H; CaromH),
6.82 (dd, 3JHH =7.3, 4JHH =1.8 Hz, 2H; CaromH), 6.67 (t, 3JHH =7.3 Hz,
2H; CaromH), 3.84 (s, 4H; NCH2-Carom), 2.50 (s, 6H; Carom-CH3), 1.94 (s,
4H; CH2C(CH3)2CH2), 0.32 (s, 18H; Si(CH3)3), 0.16 (s, 6H; CH2C-
(CH3)2CH2), À0.05 ppm (s, 4H; InCH2Si); 13C{1H} NMR (C6D6, 298 K,
125.75 MHz): d=166.2 (Carom-OIn), 153.9 (NCHN), 132.9 (CaromH),
129.2 (CaromH), 128.4 (Carom-CH3), 120.7 (Carom-CH2N), 114.0 (CaromH),
58.5 (NCH2-Carom), 53.3 (CH2C(CH3)2CH2), 26.7 (CH2C(CH3)2CH2), 23.4
(CH2C(CH3)2CH2), 19.6 (Carom-CH3), 2.9 (Si(CH3)3), 1.4 ppm (InCH2Si).
Satisfactory elemental analysis could not be obtained for this com-
plex, most probably due to contamination by 9 and also the pres-
ence of a substantial amount of Si leading to the formation of SiC.
315 mg, 96%. H NMR (C6D6, 298 K, 500.13 MHz): d=7.34 (d, JHH =
3
7.3 Hz, 1H; CaromH), 7.25 (d, JHH =6.3 Hz, 1H; CaromH), 6.97–6.89 (m,
3H; CaromH), 6.74 (t, 3JHH =7.3 Hz, 1H; CaromH), 4.05 and 2.83 (AX
2
system, one broad signal, JHH =12.2 Hz, 2H; NCH2-Carom), 3.65 and
2
2.43 (AX system, one broad signal, JHH =9.3 Hz, 2H; NCH2N), 3.32
and 3.17 (AB system, 2JHH =13.3 Hz, 2H; NCH2-Carom), 3.24 (s, 3H;
2
Carom-OCH3), 2.59 and 1.51 (AX system, one broad signal, JHH
=
12.6 Hz, 2H; CH2C(CH3)2CH2), 2.55 (s, 3H; Carom-CH3), 2.16 and 1.45
(br AX system, 2JHH =11.3 Hz, 2H; CH2C(CH3)2CH2), 0.90 (s, 3H;
CH2CCH3(CH3)CH2), 0.47 (s, 3H; CH2CCH3(CH3)CH2), 0.41 (s, 9H;
2
Si(CH3)3), 0.31 (s, 9H; Si(CH3)3), 0.08–0.00 (AB system, JHH =12.6 Hz,
2H; InCH2Si), À0.12 to À0.23 ppm (AB system, 2JHH =12.5 Hz, 2H;
InCH2Si); 13C{1H} NMR (C6D6, 298 K, 125.75 MHz): d=165.1 (Carom
-
OIn), 157.8 (Carom-OH), 132.1 (CaromH), 131.3 (Carom-CH3), 130.9
(CaromH), 129.8 (Carom-CH2N), 129.5 (CaromH), 129.1 (Carom-CH3), 128.4
(CaromH), 124.2 (CaromH), 119.8 (Carom-CH2N), 114.7 (CaromH), 77.1
(NCH2N), 66.0 (CH2C(CH3)2CH2), 62.8 (CH2C(CH3)2CH2), 61.6 (NCH2-
Carom), 60.2 (Carom-OCH3), 53.0 (NCH2-Carom), 31.5 (CH2C(CH3)2CH2),
27.0 (CH2CCH3(CH3)CH2), 26.9 (CH2CCH3(CH3)CH2), 17.5 (Carom-CH3),
16.2 (Carom-CH3), 2.9 (InCH2), 2.7 (Si(CH3)3), 2.6 (Si(CH3)3), 2.1 ppm
Chem. Eur. J. 2014, 20, 7706 – 7717
www.chemeurj.org
7715
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[11] a) E. L. Whitelaw, G. Loraine, M. F. Mahon, M. D. Jones, Dalton Trans.
2011, 40, 11469; b) Y. Wang, H. Ma, Chem. Commun. 2012, 48, 6729.
[12] Hydride abstraction with [Ph3C][B(C6F5)4] in the ligand backbone of a bi-
s(oxazolinato)gallium complex gave a cationic gallium species: S. Dag-
orne, S. Bellemin-Laponnaz, A. Maisse-FranÅois, M.-N. Rager, L. Jugꢉ, R.
Welter, Eur. J. Inorg. Chem. 2005, 4206. For another similar example of
hydride abstraction from a ligand backbone with [Ph3C][B(C6F5)4] and
generation of cationic zirconium complexes, see: B. Lian, L. Toupet, J.-F.
Carpentier, Chem. Eur. J. 2004, 10, 4301 and references cited therein.
[13] Of interest, intramolecular CÀH activation of o-tBu groups in Mes*-sub-
stituted (Mes=mesityl) phosphaalkene–MCl3 (M=Al, Ga, In) adducts
leads to coordinated ylide–MCl3 zwitterions: C.-W. Tsang, C. A. Rohrick,
T. S. Saini, B. O. Patrick, D. P. Gates, Organometallics 2004, 23, 5913.
[14] a) W. Petz, C. Kutschera, S. Tschan, F. Weller, B. Neumꢄller, Z. Anorg. Allg.
Chem. 2003, 629, 1235; b) F. Molinos de Andrade, W. Massa, C. Peppe,
W. Uhl, J. Organomet. Chem. 2005, 690, 1294.
(InCH2Si). No satisfactory elemental analysis was obtained for 8-Me,
most likely because of the presence of a substantial amount of Si
leading to the formation of SiC.
DFT computations
All calculations were carried out at the DFT level using the B3PW91
functional in the Gaussian 09 code.[35] The equilibrium and transi-
tion structures were fully optimized at the Becke’s three-parameter
hybrid functional[36] combined with the nonlocal correlation func-
tional provided by Perdew/Wang.[37] Aluminum, gallium, and
indium atoms were represented by the relativistic effective core
potential SDD (augmented by a d polarization function).[38] For the
rest of the atoms the 6-31G(d,p) basis set was used.[39] Gibbs free
energies were obtained at T=298.15 K within the harmonic ap-
proximation. Intrinsic reaction paths[40] were followed from the vari-
ous transition structures to verify the reactant-to-product linkage.
[15] a) M. Sircoglou, M. Mercy, N. Saffon, Y. Coppel, G. Bouhadir, L. Maron, D.
Bourissou, Angew. Chem. 2009, 121, 3506; Angew. Chem. Int. Ed. 2009,
48, 3454; b) E. J. Derrah, M. Sircoglou, M. Mercy, S. Ladeira, G. Bouhadir,
K. Miqueu, L. Maron, D. Bourissou, Organometallics 2011, 30, 657.
[16] See the Supporting Information for details.
Acknowledgements
[17] This is further corroborated by the fact that the formation of an indium
zwitterion also seems to occur if phenyl is replaced by an anthracenyl
(anth) substituent on the pyrimidine heterocycle in {ON^(CH-
(anth))^NO}H2 (the resulting compound could not be adequately char-
acterized owing to its very poor solubility), whereas no zwitterion what-
Y.S. and J.-F.C. warmly acknowledge financial support from
Total Raffinage-Chimie (grant to N.M.). L.M. is grateful to CINES
and CalMip (CNRS, Toulouse) for the grant of computing time.
soever could be synthesized with the very hindered
{
tBu2ON^-
(CH2)^NOtBu2}H2 proteo-ligand bearing tBu substituents in ortho and
para positions of the phenolic rings (see the Supporting Information).
Attempts to obtain {ON^(CH(CF3))^NO}H2, {ON^(CH(tBu))^NO}H2, and
{ON^(CH(mes))^NO}H2 (mes=mesityl) were unsuccessful, thus preclud-
ing the further evaluation of electronic and steric considerations.
[18] The distances from N(19) to the two Al atoms of 3.66 and 4.65 ꢂ rule
out coordination to the metal centers.
Keywords: C–H activation · density functional calculations ·
gallium · hydride abstraction · indium · zwitterions
[1] a) K. R. Campos, Chem. Soc. Rev. 2007, 36, 1069; b) P. W. Roesky, Angew.
Chem. 2009, 121, 4988; Angew. Chem. Int. Ed. 2009, 48, 4892.
[2] W. A. Nugent, D. W. Ovenall, S. J. Holmes, Organometallics 1983, 2, 161.
[3] a) S. B. Herzon, J. F. Hartwig, J. Am. Chem. Soc. 2007, 129, 6690; b) S. B.
Herzon, J. F. Hartwig, J. Am. Chem. Soc. 2008, 130, 14940; c) P. Eisen-
berger, R. O. Ayinla, J. M. P. Lauzon, L. L. Schafer, Angew. Chem. 2009,
121, 8511; Angew. Chem. Int. Ed. 2009, 48, 8361; d) A. L. Reznichenko,
K. C. Hultzsch, J. Am. Chem. Soc. 2012, 134, 3300.
[19] For leading reviews on Al–salen complexes, see: a) D. A. Atwood, M. J.
Harvey, Chem. Rev. 2001, 101, 37; b) S. Bellemin-Laponnaz, S. Dagorne,
‘Coordination chemistry and applications of salen, salan and salalen
metal complexes’, in The Chemistry of Metal Phenolates (Ed. Zabicky, J.),
John Wiley & Sons, Ltd, Chichester, 2012.
[4] a) R. Kubiak, I. Prochnow, S. Doye, Angew. Chem. 2009, 121, 1173;
Angew. Chem. Int. Ed. 2009, 48, 1153; b) J. A. Bexrud, P. Eisenberger,
D. C. Leitch, P. R. Payne, L. L. Schafer, J. Am. Chem. Soc. 2009, 131, 2116;
c) R. Kubiak, I. Prochnow, S. Doye, Angew. Chem. 2010, 122, 2683;
Angew. Chem. Int. Ed. 2010, 49, 2626; d) I. Prochnow, P. Zark, T. Mꢄller, S.
Doye, Angew. Chem. 2011, 123, 6525; Angew. Chem. Int. Ed. 2011, 50,
6401.
[5] a) J. F. Hartwig, J. Am. Chem. Soc. 1996, 118, 7010; b) S. Sakaguchi, T.
Kubo, Y. Ishii, Angew. Chem. 2001, 113, 2602; Angew. Chem. Int. Ed.
2001, 40, 2534; c) S. Doye, Angew. Chem. 2001, 113, 3455; Angew.
Chem. Int. Ed. 2001, 40, 3351; d) K. Yamaguchi, N. Mizuno, Angew. Chem.
2003, 115, 1518; Angew. Chem. Int. Ed. 2003, 42, 1480; e) J. S. M. Samec,
A. H. ꢅll, J. E. Bꢆckvall, Chem. Eur. J. 2005, 11, 2327; f) J. Cꢇmpora, I.
Matas, P. Palma, E. Alvarez, C. Graiff, A. Tiripicchio, Organometallics
2007, 26, 3840; g) M. Z. Wang, C. Y. Zhou, M. K. Wong, C. M. Che, Chem.
Eur. J. 2010, 16, 5723.
[6] P. Horrillo-Martꢈnez, K. C. Hultzsch, A. Gil, V. Branchadell, Eur. J. Org.
Chem. 2007, 3311.
[7] a) B. E. Bent, R. G. Nuzzo, L. H. Dubois, J. Am. Chem. Soc. 1989, 111,
1634; b) Comprehensive Organometallic Chemistry II, Volumes 1 and 11
(Eds.: E. W. Abel, F. G. A. Stone, G. Wilkinson), Elsevier Science Oxford,
1995.
[8] a) K. Saito, Y. Shibata, M. Yamanaka, T. Akiyama, J. Am. Chem. Soc. 2013,
135, 11740; b) C. Zhu, T. Akiyama, Adv. Synth. Catal. 2010, 352, 1846.
[9] In nature, NADPH is a known reducing agent that switches to the co-
enzyme NAPD+ (nicotinamide adenine dinucleotide phosphate)
through formal loss of hydride in several biological processes; the
transformation is driven by aromatization to a pyridinium moiety.
[10] a) W. Uhl, J. Grunenberg, A. Hepp, M. Matar, A. Vinogradov, Angew.
Chem. 2006, 118, 4465; Angew. Chem. Int. Ed. 2006, 45, 4358; b) W. Uhl,
A. Vinogradov, S. Grimme, J. Am. Chem. Soc. 2007, 129, 11259.
[20] P. Hormnirun, E. L. Marshall, V. C. Gibson, A. J. P. White, D. J. Williams, J.
Am. Chem. Soc. 2004, 126, 2688.
[21] N. C. Johnstone, E. S. Aazam, P. B. Hitchcock, J. R. Fulton, J. Organomet.
Chem. 2010, 695, 170.
[22] a) I. Peckermann, A. Kapelski, T. P. Spaniol, J. Okuda, Inorg. Chem. 2009,
48, 5526; b) I. Yu, A. Acosta-Ramꢈrez, P. Mehrkhodavandi, J. Am. Chem.
Soc. 2012, 134, 12758; c) D. C. Aluthge, B. O. Patrick, P. Mehrkhodavandi,
Chem. Commun. 2013, 49, 4295.
[23] For the description of a related pyrimidinium core, see: T. W. Hudnall,
C. W. Bielawski, J. Am. Chem. Soc. 2009, 131, 16039.
[24] G. R. Fulmer, A. J. M. Miller, N. H. Sherden, H. E. Gottlieb, A. Nudelman,
B. M. Stoltz, J. E. Bercaw, K. I. Goldberg, Organometallics 2010, 29, 2176.
[25] The precursor
{ {
MeON^(CH2)^NOMe}Li2 required to prepare MeON^-
(CH2)^NOMe}D2 and, from here, 7-D, crystallized as the tetrametallic
complex [{ON^(CH2)^NO}Li·Li(THF)2]2 from THF. The four four-coordinat-
ed Li atoms present an unusual linear arrangement; see the Supporting
Information for details.
[26] The formation of 6 is slow, and requires a long reaction time at high
temperature; it is very likely that 6 and 8-H, which need not be present
in large amounts, are in the presence of each other at some stage of
the reaction, and react with one another to yield 9.
[27] M. Sircoglou, N. Saffon, K. Miqueu, G. Bouhadir, D. Bourissou, Organo-
metallics 2013, 32, 6780.
[28] D. W. Stephan, G. Erker, Angew. Chem. 2010, 122, 50; Angew. Chem. Int.
Ed. 2010, 49, 46 and references therein.
[29] A. M. Chapman, M. F. Haddow, D. F. Wass, J. Am. Chem. Soc. 2011, 133,
18463.
[30] C. Appelt, H. Westenberg, F. Bertini, A. W. Ehlers, J. C. Slootweg, K. Lam-
mertsma, W. Uhl, Angew. Chem. 2011, 123, 4011; Angew. Chem. Int. Ed.
2011, 50, 3925.
Chem. Eur. J. 2014, 20, 7706 – 7717
www.chemeurj.org
7716
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[31] The Group 13 Metals Aluminium, Gallium, Indium and Thallium: Chemical
Patterns and Peculiarities (Eds: S. Aldridge, A. J. Downs), Wiley, Hoboken,
2011.
[32] Attempts to deprotonate the pentahydropyrimidinium moiety (by treat-
ment of 3 with potassium bis(trimethylsilyl)amide (KHMDS) or other
strong bases) to access the corresponding NHC carbene species were
unsuccessful.
gar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
slowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.
[36] A. D. Becke, J. Chem. Phys. 1993, 98, 5648.
[33] a) O. T. Beachley Jr., R. N. Rusinko, Inorg. Chem. 1979, 18, 1966; b) O. T.
Beachley Jr., R. G. Simmons, Inorg. Chem. 1980, 19, 1021.
[34] a) G. M. Sheldrick, SHELXS-97, Program for the Determination of Crystal
Structures; University of Goettingen: Germany, 1997; b) G. M. Sheldrick,
SHELXL-97, Program for the Refinement of Crystal Structures; University
of Goettingen: Germany, 1997.
[35] Gaussian 09, Revision A.02, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
[37] J. P. Perdew, Y. Wang, Phys. Rev. B 1992, 45, 13244.
[38] a) A. Bergner, M. Dolg, W. Kuchle, H. Stoll, H. Preuss, Mol. Phys. 1993, 80,
1431; b) T. Leininger, A. Berning, A. Nicklass, H. Stoll, H.-J. Werner, H.-J.
Flad, Chem. Phys. 1997, 217, 19.
[39] a) R. Ditchfield, W. J. Hehre, J. A. Pople, J. Chem. Phys. 1971, 54, 724;
b) W. J. Hehre, R. Ditchfield, J. A. Pople, J. Chem. Phys. 1972, 56, 2257;
c) P. C. Hariharan, J. A. Pople, Theor. Chim. Acta. 1973, 28, 213.
[40] a) C. Gonzalez, H. B. Schlegel, J. Chem. Phys. 1989, 90, 2154; b) C. Gonza-
lez, H. B. Schlegel, J. Phys. Chem. 1990, 94, 5523.
Received: February 26, 2014
Published online on May 19, 2014
Chem. Eur. J. 2014, 20, 7706 – 7717
www.chemeurj.org
7717
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim