Silica-Grafted Tris(neopentyl)aluminum: A Monomeric Aluminum Solid Co-catalyst for Efficient Nickel-Catalyzed Ethene Dimerization

Article abstract of DOI:10.1002/anie.202006285

A silica-supported monomeric alkylaluminum co-catalyst was prepared via surface organometallic chemistry by contacting tris(neopentyl)aluminum and partially dehydroxylated silica. This system, fully characterized by solid-state ²⁷Al NMR spectroscopy augmented by computational studies, efficiently activates (ⁿBu₃P)₂NiCl₂ towards dimerization of ethene, demonstrating comparable activity to previously reported dimeric diethylaluminum chloride supported on silica. Three types of aluminum surface species have been identified: monografted tetracoordinated Al species as well as two types of bisgrafted Al species—tetra- and pentacoordinated. Of them, only the monografted Al species is proposed to be able to activate the (ⁿBu₃P)₂NiCl₂ complex and generate the active cationic species.

Full text of DOI:10.1002/anie.202006285

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Silica-grafted Tris(neopentyl)aluminum: A Monomeric Aluminum
Solid Co-Catalyst for Efficient Nickel-Catalyzed Ethene
Dimerization
Authors: Christophe Copéret, Ilia Moroz, Pierre Florian, Jasmine Viger
Gravel, Christopher P Gordon, and Anne Lesage
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202006285
Link to VoR: https://doi.org/10.1002/anie.202006285

10.1002/anie.202006285
Angewandte Chemie International Edition
RESEARCH ARTICLE
Silica-grafted Tris-(neopentyl)aluminum: A Monomeric Aluminum
Solid Co-Catalyst for Efficient Nickel-Catalyzed Ethene
Dimerization
Ilia B. Moroz,[a] Pierre Florian,[b] Jasmine Viger-Gravel,[c] Christopher P. Gordon,[a] Anne Lesage,[c] and
Christophe Copéret*[a]
[a]
Dr. I. B. Moroz, C. P. Gordon, Prof. Dr. C. Copéret
Department of Chemistry and Applied Biosciences, ETH Zürich
Vladimir-Prelog-Weg 1-5, 8093 Zürich, Switzerland
E-mail: ccoperet@ethz.ch
[b]
[c]
Dr. P. Florian
CEMHTI CNRS UPR3079, Université d’Orléans
1D Avenue de la Recherche-Scientifique, F-45071 Orléans Cedex 2, France
Dr. J. Viger-Gravel, Dr. A. Lesage
Centre de RMN à Très Hauts Champs, Université de Lyon (CNRS/ENS Lyon/UCB Lyon 1)
69100 Villeurbanne, France
Supporting information for this article is given via a link at the end of the document.
Abstract: A silica-supported monomeric alkylaluminum co-catalyst
was prepared via Surface Organometallic Chemistry by contacting
tris-(neopentyl)aluminum and partially dehydroxylated silica. This
system, fully characterized by solid-state 27Al NMR augmented by
computational studies, efficiently activates (nBu3P)2NiCl2 towards
dimerization of ethene demonstrating comparable activity to
previously reported dimeric diethylaluminum chloride supported on
silica. Three types of aluminum surface species have been identified:
monografted tetracoordinated Al species as well as two types of bis-
grafted Al species – tetra- and pentacoordinated. Of them, only the
monografted Al species is proposed to be able to activate the
(nBu3P)2NiCl2 complex and generate the active cationic species.
species upon reaction with the metal pre-catalysts, e.g.
(nBu3P)2NiCl2. Such species can further dissociate, and free
monomeric AlR3 species can react with putative active [LMR]+
species to form a deactivated [LM(µ2-R)2AlR2]+ complex as shown
in Scheme 1.[6]
Scheme 1. The proposed mechanisms of (a) pre-catalyst activation by dimeric
trialkylaluminum co-catalysts, (b) dissociation of anionic co-catalyst species, (c)
inhibition of pre-activated catalyst, where [LMCl2] is a pre-catalyst, [AlR3]2 – a
dimeric alkylaluminum co-catalyst.[6]
Introduction
The oligomerization of ethene is one of the key technologies for
the production of α-olefins (e.g. 1-butene, 1-hexene, etc.) that are
co-monomers for the synthesis of linear low-density polyethylene.
This reaction is catalyzed by various homogeneous catalysts
based on early- (Ti, Ta, Cr) or late-transition metals (Ni) and
requires the addition of co-catalysts, typically alkyl aluminum
reagents to generate the active species.[1] Of them,
methylaluminoxane (MAO) shows the best co-catalytic
performance,[2] though a very large excess of >100 equiv. of MAO
with respect to the active metal is needed, making its use
problematic in industry.[3] While trialkylaluminums (Al2R6, R = Me
or Et) are used to activate Ti(OR’)4-pre-catalyst (Alphabutol
process) with only a small excess (<10 equiv. with respect to the
metal) compared to MAO,[4] such co-catalysts are inefficient for
Ni-based systems, and they need to be replaced by
alkylaluminum chloride – Al2R6-2nCl2n, n = 1 or 2 – or the
corresponding phenoxide – Al2R4(OR)2, that are all dimeric in
solution.[5]
The presence of bridging Cl ligands in alkylaluminum chloride
(AlR3-nCln, n = 1 or 2) prevents the inhibition of the activated
cationic species. The same behavior has been observed for the
corresponding
silica-supported
co-catalysts
based
on
triethylaluminum (TEA) and diethylaluminum chloride (DEAC) that
led to inactive and active Ni-oligomerization catalysts,
respectively.[7] The combined 27Al solid-state NMR and
computational studies indicated the presence of various dimeric
Al species on the surface, and the high catalytic performance of
the DEAC-based system (unlike inactive TEA-based) was
attributed to the preservation of µ2-Cl bridging ligands between
the two Al atoms that stabilize the Al dimeric surface structure
thus preventing deactivation of the Ni catalyst.[6-7]
The inefficiency of Al2R6 co-catalysts has been proposed to be
associated with the formation of anionic dimeric [AlR3][AlR2Cl2]–
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202006285
Angewandte Chemie International Edition
RESEARCH ARTICLE
t
t
Scheme 2. Possible surface species resulting from a reaction of Al(CH2 Bu)3 with a silica surface: (a) reaction of Al(CH2 Bu)3 with a surface silanol leading to the
t
formation of monografted species; (b) neopentyl transfer process leading to the formation of bis-grafted and tris-grafted Al species and Si–CH2 Bu surface moieties.
d, d1, d2, and d3 are Al–[O(Siº)2] bond lengths. The terms mono-, bis- and tris-grafted Al species refer to the number of bonds between Al and surface siloxy ligand
(ºSi-O-); it does not include an additional coordination of neutral siloxane bridges (ºSi-O-Siº), that only contribute to coordination number.
We reasoned that monomeric alkylaluminum isolated on a silica
surface should also be an efficient co-catalyst. They are predicted
by DFT computations to be able to activate late-transition metal
complexes,[6] their site-isolation should prevent the poisoning of
tentative active species.
and the formation of an additional Al–OSiº bond.[7, 10] Since these
processes lead to similar stoichiometries (evolution of 2,2-
dimethylpropane and overall elemental analysis), one cannot
preclude the formation of bis- (single alkyl transfer) and tris-
grafted neopentyl-free (transfer of two alkyls) aluminum species
(see Scheme2b).
We thus decided to investigate the grafting of a base-free
monomeric alkylaluminum, namely tris-(neopentyl)aluminum
t
Al(CH2 Bu)3, on a silica partially dehydroxylated at 700 °C, SiO2-
700, that contains isolated OH groups. We show that such
supported alkylaluminum can be an efficient solid co-catalyst for
the Ni-catalyzed ethene oligomerization, whose co-catalytic
activity is attributed to the presence of isolated alkylaluminum
species by enabling the formation of cationic Ni alkyl species from
(nBu3P)2NiCl2.[8]
Results and Discussion
t
The aluminum precursor Al(CH2 Bu)3 was synthesized by reaction
t
of AlBr3 with 3 equiv. of LiCH2 Bu in heptane, and isolated as a
colorless liquid in 89% overall yield (see Figure S1 for 1H NMR);
this compound is monomeric in the liquid state and in hydrocarbon
solutions.[9]
t
Grafting of Al(CH2 Bu)3 on SiO2-700 (1.08 OH nm–2) in C6H6
generates ~1 equiv. of 2,2-dimethylpropane per surface OH group.
The resulting solid (material 1) contains 0.97 wt% Al and 4.55 wt%
C that correspond to 1.08 Al nm–2 and 10.6 C per Al atom. These
data and the complete consumption of surface OH groups as
observed by IR (see Figure S2) are consistent with a simple
Figure 1. 27Al MAS NMR spectra (solid black lines) obtained at (a) 16.4 and (b)
23.5 T, along with three simulated components: site S1 (dotted line), S2
(dotdash line), and S3 (dashed) as well as the resulting line shape (solid grey
line).
t
t
protonolysis of one Al–CH2 Bu bond of Al(CH2 Bu)3 leading to the
formation of the monografted dialkyl Al species (Scheme 2a).
However, the grafting of alkylaluminum compounds on silica is
often quite complex because of the possible subsequent reaction
of the surface alkylaluminium species with adjacent siloxane
bridges through the transfer of an alkyl group from Al to a Si atom
Thus, in order to obtain information on the exact structure of Al
species, we performed 27Al solid-state NMR measurements
complemented by DFT computations.[7, 10c] 27Al NMR parameters
(chemical shift (d) and quadrupolar coupling constants, CQ) are
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202006285
Angewandte Chemie International Edition
RESEARCH ARTICLE
highly sensitive to the local environment of 27Al nuclei and allow
the discrimination of possible mono- and dimeric sites as well as
understanding the coordination sphere (which ligand and
geometry).[7, 10c, 11] 27Al spin-echo spectra were acquired at two
different high magnetic fields, namely 16.4 and 23.5 T, under very
fast Magic Angle Spinning (MAS >50 kHz). The spectra show a
broad line shape featuring resolved discontinuities, particularly at
16.4 T. Those features are highly dependent on the magnetic
field: the line width sharply decreases with increasing magnetic
field, as expected for signals dominated by quadrupolar
broadening (Figure 1).
MHz), not matching any of the experimental values (Figure 3 and
º
Table S2). Coordination of a siloxane bridge (M_[O(Si )2]_d)
leads to a significant decrease of both diso and CQ, and their values
depend on the Al–[O(Siº)2] bond length d (see Figure 2b): varying
d from 1.8 to 2.1 Å (around the bond length found by optimizing
the geometry) gives diso ranging from 154.8 to 173.5 ppm and CQ
– from 24.7 to 34.9 MHz (Figure 3).Note that the CQ value is
significantly affected upon varying the bond length, while the diso
value is much less sensitive.[12b] The experimentally observed site
S1 (diso = 154 ppm, CQ = 31 MHz) is most consistent with the
structure M_[O(Si )2]_d with a distance between Al and the
º
The simulation of the two data sets – carried out simultaneously
(see the Supporting Information for a detailed discussion on the
simulation of 27Al NMR spectra) – reveals three significant
contributions (Figure 1 and Table S1) that are necessary to
consistently model the two spectra with the exact same set of
NMR parameters (i.e. diso. CQ and hQ): the site S1 (cyan, 19% of
total area, Figure 1) characterized by diso /CQ /hQ of 154 ppm / 31
MHz / 0.4; the site S2 (orange, 58%, Figure 1) – 101 ppm / 19
MHz / 0.7; and the site S3 (purple, 23%, Figure 1) – 83 ppm / 19
MHz / 0.9. The need for three components, each characterized by
a specific set of NMR parameters, to simulate the obtained 27Al
NMR spectra, indicates the presence of at least three types of
surface Al species.
oxygen of the siloxane bridge of ca. 1.9 Å (d = 1.9 Å).
The bis-grafted tri-coordinated structure (B, Figure 2c), which
can be formed upon the reaction of the grafted neopentyl
aluminum species with a nearby siloxane bridge, shows further
decrease in diso down to 142.9 ppm, while CQ remains significantly
high – 36.4 MHz. Such combination of parameters does not match
the experimental data. However, an increase of the Al
coordination number from 3 to 4 by coordinating a siloxane bridge
(B_[O(Si )2]_d, Figure 2d) leads to diso of ca. 105-107 ppm and
º
CQ of 18.3–25.8 MHz, the CQ significantly depends on the Al–
[O(Siº)2] bond length d (varied between 1.8 and 2.0 Å), as
observed for the monografted structures. The 27Al NMR
parameters of site S2 (diso = 101 ppm, CQ = 19 MHz) match quite
well those of B_[O(Si )2]_d structure with the Al–[O(Siº)2] bond
º
length being close to 1.8 Å (Figure 3). Thus, sites S1 and S2 can
be assigned to tetracoordinated Al species, rather than tri-
coordinated, that are mono- and bis-grafted, respectively.
Figure 2. Optimized structures of alkylaluminum surface species (right) with a
schematic representation of Al coordination spheres (left): (a) monografted tri-
coordinated species (M); (b) monografted tetracoordinated species
(M_[O(Si
º
)2]_d) with Al–[O(Siº)2] bond length d of 1.9 Å as an example; (c) bis-
grafted tri-coordinated species (B); (d) bis-grafted tetracoordinated species
(B_[O(Si
(B_[O(Si
º
º
)2]_d) with d of 1.9 Å; (e) bis-grafted pentacoordinated species
)2]2_d1_d2) with d1 and d2 of 2.1 Å. Al, O, C, Si and H atoms are
Figure 3. Graphical representation of NMR parameters CQ (horizontal axis) and
represented as violet, red, black, grey, and white balls, respectively.
d
iso (vertical axis) of experimentally observed sites S1 (filled hexagon), S2 (filled
triangle) and S3 (filled diamond), and calculated structures shown as open
circles: monografted tri-coordinated M and tetracoordinated M_[O(Si )2]_d
structures (open circles), bis-grafted tri-coordinated B and tetracoordinated
º
To assign the observed 27Al NMR features to specific Al species,
we carried out DFT studies on cluster models representing
various Al environments, that are expected to be formed upon
B_[O(Si
B_[O(Si
º
º
)2]_d structures (open triangles), bis-grafted pentacoordinated
)2]2_d1_d2 structures (open squares).
t
reaction of Al(CH2 Bu)3 with surface OH groups and reactive
siloxane bridges present on SiO2-700 (vide supra). Representative
structures were optimized (Figure 2), and their NMR properties
were calculated (Figure 3), in particular 27Al diso and CQ, that are
important signatures for Al coordination and geometry in
aluminosilicates and related materials.[12] The computed NMR
parameters are also given in Table S2.
Regarding site S3 (diso = 83 ppm, CQ = 19 MHz), one can propose
that a second alkyl transfer from Al to Si occurs with the formation
of tris-grafted alkyl-free Al species, however the calculated diso
and CQ values of neither tri-coordinated (diso = 70-80 ppm, CQ
=
31-35 MHz) nor tetracoordinated (diso = 61-65 ppm, CQ = 7-21
MHz) tris-grafted Al structures match the observed experimental
values due to too high CQ or too low diso, respectively (Figure
First, the monografted tri-coordinated structure (M, see Figure
2a) shows both very high diso and CQ values (209.9 ppm and 40.7
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202006285
Angewandte Chemie International Edition
RESEARCH ARTICLE
3).[12b] Noteworthy, site S3 is characterized by CQ value, very
close to the one of site S2, but with a diso that is 20 ppm lower.
Such 20-ppm-decrease in diso while maintaining similar CQ values
is typically observed upon addition of the second siloxane bridge
in the coordination sphere of aluminum in a trans-position to the
first siloxane, increasing coordination of Al from 4 to 5 and
changing the geometry from tetrahedral to trigonal
bipyramidal.[12b] Indeed, for bis-grafted pentacoordinated
structures with two mutually trans siloxane bridges
silica-supported AlEt2Cl (DEAC@SiO2, see the Supporting
Information for synthetic details) demonstrates turnover numbers
(TON) of ca. 2800. Both systems show rather high selectivity in
butenes: 93% for material 1 vs. 92% for DEAC@SiO2. Together
with a desired product – 1-butene – a mixture of E/Z-2-butenes is
formed, with selectivity in 1-butene of 37% (vs. 17% for
DEAC@SiO2). The formation of E/Z-2-butenes is well-known for
Ni-catalyzed oligomerization of ethene that is due to a favorable
double-bond isomerization of 1-butene into thermodynamically
more stable 2-butenes.[7, 14] Other minor by-products are mostly
hexenes as identified by GC. The pre-activated material 2 also
demonstrates catalytic activity in ethene oligomerization, reaching
3900 mol(C2H4) mol(Ni)–1 after 1 h, and the selectivity in butenes
of 96%. Note that the solution left after isolation of material 2 is
inactive in the oligomerization of ethene, thus confirming that this
catalytic system is heterogeneous.
(B_[O(Si )2]2_d1_d2, Figure 2e), the calculated diso is 69-74 ppm,
º
which is rather close to the diso value of the site S3 (Figure 3). CQ
depends strongly on the Al–[O(Siº)2] bond lengths d1 and d2 (see
Figure 2e) and varies between 11.5 (d1=1.8 Å, d2 = 2.1 Å) and
20.5 (d1 = 2.1 Å, d2 = 2.1 Å) MHz (see Figure 3 and Table S2)
Note that CQ > 16 MHz (observed for site S3, see Figure 3) is
associated with both d1 and d2 longer than 2.0 Å. We thus assign
the site S3 to bis-grafted pentacoordinated alkylaluminum species
We further studied the reaction of (nBu3P)2NiCl2 and its
diamagnetic Pt equivalent with material 1 (1 equiv. of Ni or Pt per
surface Al, as calculated based on EA of material 1) to better
understand the activation process (see the Supporting
Information for synthetic details). Based on elemental analysis,
the resulting (catalytically active) material 2 contains 0.45 wt.% of
Ni and 0.85 wt.% of Al that corresponds to an Al/Ni ratio of 4.1 as
mentioned above. Material 2 also contains 0.59 wt.% of Cl and
0.52 wt.% of P. This corresponds to a Cl/Ni and P/Ni ratio of 2.2
with the structure B_[O(Si )2]2_d1_d2 (Figure 2e) with long d1
º
and d2 distances, greater than 2.0 Å.
The co-catalytic property of material 1 in the oligomerization of
ethene was then examined by contacting material 1 with a
solution of (nBu3P)2NiCl2 under constant ethene pressure of 5 bar
at 30 °C (in situ formation of active species).[5c, 13] We have also
isolated the pre-activated catalyst by reacting (nBu3P)2NiCl2 with
material 1 (material 2, ex situ activation, see the Supporting
Information for synthetic details and Figure S3 for IR spectra) and
examined its catalytic performance under the same reaction
conditions. An Al/Ni ratio of ca. 4.1 was used for material 1 that
corresponds to the ratio found in material 2 based on elemental
1
in both cases, as present in the initial Ni complex. H NMR of
combined benzene washings shows the presence of at least two
compounds, both containing nBu3P ligands, however their
identification is not trivial due to overlapping NMR signals (Figure
S9). 31P NMR is usually very characteristic for the metal
phosphine complexes, but the presence of paramagnetic species
typical for 4-coordinated Ni(II) complexes leads to very broad
signals that are difficult to interpret (Figure S11a).
analysis. Compared to
a
previously reported Cl-free
triethylaluminum on mesoporous silica SBA-15 (TEA@SBA),
which was reported as inactive,[7] material 1 does activate
(nBu3P)2NiCl2 complex converting 9400 mol(C2H4) per mol(Ni)
after 1 h of reaction (see Table S3). Under the same conditions,
Scheme 3. The proposed mechanisms for activation of (nBu3P)2NiCl2 on material 1 that yields material 2: (a) plausible activation pathway involving more reactive
monografted di(neopentyl)aluminum species; (b) unlikely activation pathway on less reactive bis-grafted mono(neopentyl)aluminum species.
We thus prepared the corresponding material from the
diamagnetic cis-(nBu3P)2PtCl2 complex (see Figure S5 for IR
spectra). In particular, the chemical shift and the magnitude of
1J(195Pt, 31P), observed in 31P NMR spectra of Pt(II) phosphine
complex, are known to be very sensitive to the nature of trans
ligand in square-planar Pt(II) complexes. The 31P NMR spectrum
of the combined washings exhibits two signals (Figure S11b): the
1
first one with d(31P) = 4.86 ppm and J(195Pt–31P) = 2400 Hz is
characteristic of trans-(nBu3P)2PtCl2,[15] which likely results from
the isomerization of the initial compound, while the second one
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202006285
Angewandte Chemie International Edition
RESEARCH ARTICLE
1
(with d(31P) = 4.30 ppm and J(195Pt–31P) = 3000 Hz) has a J-
equiv. of Ni with respect to Al species was found in material 2,
coupling constant characteristic of the alkylated compound,
namely trans-[(nBu3P)2PtRCl],[16] where R is neopentyl. This
indicates that alkyl-chloride exchange takes place upon reaction
of the supported (neopentyl)aluminum species and cis-
(nBu3P)2PtCl2. It is expected that a similar exchange occurs in the
case of (nBu3P)2NiCl2, that is in line with the presence of two Ni
suggesting that only 25% of (neopentyl)aluminum species can
t
further react with [(nBu3P)2Ni(CH2 Bu)Cl] complex and form an
t
ionic pair with [(nBu3P)2Ni(CH2 Bu)]+. Taking into account that
monografted tetracoordinated bis-alkyl Al species, representing
ca. 20% of all Al species, are expected to be more reactive than
bis-grafted pentacoordinated species (vide supra), we propose
that the monografted alkyl aluminum species generate the active
1
phosphine compounds observed in H NMR of washings (vide
supra). Note also, that 31P MAS NMR of the obtained solid
material contains signal at 7.3 ppm, which can be attributed to
adsorbed (nBu3P)2PtCl2, together with two shoulders at 15.7 and
-1.5 ppm (see Figure S12b). While alkylchloride complexes
[(R’3P)2PtRCl] are characterized by slightly more shielded d(31P)
than their dichloride analogues [(R’3P)2PtCl2] (see Table S4), a
deshielded signal at 15.7 ppm is likely associated with the
formation of cationic alkyl complex [(nBu3P)2PtR]+ on the surface.
cationic species by abstraction of Cl– from [(nBu3P)2Ni(CH2 Bu)Cl]
t
(Scheme 3a), while bis-grafted species (ca. 80%) do not
participate in the activation (Scheme 3b). The co-catalytic
properties of material 1 are thus likely related to the presence of
monografted di(neopentyl)aluminum species supported on silica.
Conclusion
t
Indeed, according to DFT computations, [(nBu3P)2Pt(CH2 Bu)]+
has ca. 15 ppm more deshielded 31P chemical shift than
(nBu3P)2PtCl2 (see Table S4). Formation of such cationic species
in the case of Ni complex would be consistent with the catalytic
properties of material 2 in ethene oligomerization.
To summarize, material 1 prepared by grafting Al(CH2 Bu)3 on
SiO2-700 contains three types of monomeric alkylaluminum
t
In conclusion, we have shown that grafting of Al(CH2 Bu)3 on silica
treated at 700 °C leads to the formation of supported monomeric
Al species that activate the (nBu3P)2NiCl2 complex towards
selective dimerization of ethene and demonstrates comparable
catalytic performances with the previously reported supported
dimeric DEAC co-catalyst systems. Three specific structures for
the grafted monomer have been identified using 27Al MAS NMR
augmented by DFT computations of NMR parameters: (i)
monografted di(neopentyl)aluminum species with coordinated
t
species.
The
first
type
represents
monografted
di(neopentyl)aluminum species that are tetracoordinated because
of the coordination of additional siloxane bridge to Al. These
t
t
species are formed via protonolysis of one Al–CH2 Bu bond of
siloxane bridge resulting from protonolysis of Al–CH2 Bu bond
t
Al(CH2 Bu)3 by surface OH groups, as expected for grafting of
with surface OH groups; and (ii-iii) two bis-grafted
mono(neopentyl)aluminum species having one and two siloxane
bridges coordinated to Al, respectively that are likely formed by
neopentyl transfer from Al to an adjacent siloxane bridge. Analysis
of the reaction between well-defined co-catalyst and
(nBu3P)2NiCl2 suggests that monografted species are responsible
for the efficient catalytic properties of the material, while bis-
grafted species are unlikely to activate the Ni complex. We are
currently exploring approaches to directly observe the active
species in these supported systems.
alkyl compounds on SiO2-700. However, only 19% of all Al on the
surface belong to this type. Two other types represent bis-grafted
mono(neopentyl)aluminum species that are tetra- (58%) and
pentacoordinated (23%) due to the coordination of one or two
additional siloxane bridges coordinated to Al, respectively. Taking
into account the high isolation of OH groups on the surface of
SiO2-700 and C/Al ratio of 10.6 (ca. 2 neopentyl per surface Al), we
propose that bis-grafted species result from transfer of a
neopentyl group from Al to a Si atom of an adjacent siloxane
bridge and the formation of additional Al–OSiº bond, that has
been previously observed for other alkylaluminum compounds
grafted on partially dehydroxylated silica (Scheme 2).[7, 10a-c, 10e, 10f]
Material 1 was shown to activate in situ (nBu3P)2NiCl2 towards
selective dimerization of ethene, demonstrating ca. 3.3 times
higher TON and higher selectivity in 1-butene (37% vs. 17%) as
compared to DEAC@SiO2 tested under the same conditions.
Acknowledgements
I.B.M. is grateful to Swiss Government Excellence scholarship for
financial support. We thank M. Pucino and J. de Jesus Silva for
their help with gas chromatography analyses and J. Meyet for his
help with catalytic tests.
These
results
suggest
that
tailor-made
monomeric
(neopentyl)aluminum species grafted on silica provide access to
an efficient co-catalyst for Ni-catalyzed ethene dimerization. We
believe that by stabilizing monomeric alkylaluminum species on
the surface of silica, poisoning of the putative active species with
free AlR3 is likely avoided, and as a result, high catalytic
performance of the system can be achieved.
Keywords: co-catalysts • oligomerization of ethene • nickel •
alkylaluminum • aluminum-27 solid state NMR • DFT
computations
[1]
a) E. Y. X. Chen, T. J. Marks, Chem. Rev. 2000, 100, 1391-1434; b) K.
P. Bryliakov, E. P. Talsi, Coord. Chem. Rev. 2012, 256, 2994-3007; c) K.
Ziegler, E. Holzkamp, H. Breil, H. Martin, Angew. Chem. 1955, 67, 426-
426; d) R. Gao, W.-H. Sun, C. Redshaw, Catal. Sci. Technol. 2013, 3,
1172-1179.
We have also investigated activation of Ni pre-catalyst by co-
catalyst 1. Based on our results, material 1 can alkylate Ni pre-
t
catalyst (vide supra), forming [(nBu3P)2Ni(CH2 Bu)Cl] complex,
similar to the first step of activation with homogeneous co-
catalysts (Scheme 3a).[17] The second step likely involves
abstraction of Cl– by surface Al and formation of cationic nickel-
[2]
[3]
a) E. Zurek, T. Ziegler, Prog. Polym. Sci. 2004, 29, 107-148; b) H. Sinn,
W. Kaminsky, H.-J. Vollmer, R. Woldt, Angew. Chem. 1980, 92, 396-402;
c) W. Kaminsky, Macromolecules 2012, 45, 3289-3297.
t
alkyl complex [(nBu3P)2Ni(CH2 Bu)]+, which is the proposed active
species in this Ni-catalyzed ethene dimerization. It is expected
that these species are strongly bound to the surface due to ionic
interactions with anionic Al species. Noteworthy, only ca. 0.25
a) G. J. P. Britovsek, V. C. Gibson, D. F. Wass, Angew. Chem., Int. Ed.
1999, 38, 428-447; b) S. D. Ittel, L. K. Johnson, M. Brookhart, Chem. Rev.
2000, 100, 1169-1204; c) V. C. Gibson, S. K. Spitzmesser, Chem. Rev.
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202006285
Angewandte Chemie International Edition
RESEARCH ARTICLE
2003, 103, 283-316; d) D. S. McGuinness, Chem. Rev. 2011, 111, 2321-
2341.
[4]
[5]
A. W. Al-Sa'doun, Appl. Catal., A 1993, 105, 1-40.
a) A. Kermagoret, P. Braunstein, Organometallics 2008, 27, 88-99; b) M.
Dötterl, H. G. Alt, ChemCatChem 2012, 4, 660-667; c) M. Eichmann, W.
Keim, M. Haumann, B. U. Melcher, P. Wasserscheid, J. Mol. Catal. A:
Chem. 2009, 314, 42-48; d) D. S. McGuinness, W. Mueller, P.
Wasserscheid, K. J. Cavell, B. W. Skelton, A. H. White, U. Englert,
Organometallics 2002, 21, 175-181.
[6]
[7]
Z. Boudene, A. Boudier, P. A. R. Breuil, H. Olivier-Bourbigou, P.
Raybaud, H. Toulhoat, T. de Bruin, J. Catal. 2014, 317, 153-157.
A. Kermagoret, R. N. Kerber, M. P. Conley, E. Callens, P. Florian, D.
Massiot, F. Delbecq, X. Rozanska, C. Copéret, P. Sautet, J. Catal. 2014,
313, 46-54.
[8]
[9]
F. Rascón, R. Wischert, C. Copéret, Chem. Sci. 2011, 2, 1449-1456.
O. T. Beachley, L. Victoriano, Organometallics 1988, 7, 63-67.
[10] a) J. Li, J. A. DiVerdi, G. E. Maciel, J. Am. Chem. Soc. 2006, 128, 17093-
17101; b) R. Anwander, C. Palm, O. Groeger, G. Engelhardt,
Organometallics 1998, 17, 2027-2036; c) R. N. Kerber, A. Kermagoret,
E. Callens, P. Florian, D. Massiot, A. Lesage, C. Copéret, F. Delbecq, X.
Rozanska, P. Sautet, J. Am. Chem. Soc. 2012, 134, 6767-6775; d) E.
Mazoyer, J. Trébosc, A. Baudouin, O. Boyron, J. Pelletier, J.-M. Basset,
M. J. Vitorino, C. P. Nicholas, R. M. Gauvin, M. Taoufik, L. Delevoye,
Angew. Chem., Int. Ed. 2010, 49, 9854-9858; e) B. Werghi, A.
Bendjeriou-Sedjerari, A. Jedidi, E. Abou-Hamad, L. Cavallo, J.-M. Basset,
Organometallics 2016, 35, 3288-3294; f) J. Pelletier, J. Espinas, N. Vu,
S. Norsic, A. Baudouin, L. Delevoye, J. Trébosc, E. Le Roux, C. Santini,
J.-M. Basset, R. M. Gauvin, M. Taoufik, Chem. Commun. 2011, 47,
2979-2981.
[11] A. Kermagoret, R. N. Kerber, M. P. Conley, E. Callens, P. Florian, D.
Massiot, C. Copéret, F. Delbecq, X. Rozanska, P. Sautet, Dalton Trans.
2013, 42, 12681-12687.
[12] a) J. A. Tang, J. D. Masuda, T. J. Boyle, R. W. Schurko, ChemPhysChem
2006, 7, 117-130; b) E. Lam, A. Comas-Vives, C. Copéret, J. Phys.
Chem. C 2017, 121, 19946-19957; c) E. Lam, C. Copéret, Helv. Chim.
Acta 2018, 101, e1800120.
[13] A. Kermagoret, P. Braunstein, Dalton Trans. 2008, 1564-1573.
[14] J. Zhang, H. Gao, Z. Ke, F. Bao, F. Zhu, Q. Wu, J. Mol. Catal. A: Chem.
2005, 231, 27-34.
[15] A. Pidcock, R. E. Richards, L. M. Venanzi, J. Chem. Soc. A 1966, 1707-
1710.
[16] R. L. Brainard, T. M. Miller, G. M. Whitesides, Organometallics 1986, 5,
1481-1490.
[17] I. E. Soshnikov, N. V. Semikolenova, K. P. Bryliakov, V. A. Zakharov, W.-
H. Sun, E. P. Talsi, Organometallics 2015, 34, 3222-3227.
6
This article is protected by copyright. All rights reserved.

10.1002/anie.202006285
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents
t
Isolation for Activity. Monomeric Al solid co-catalyst for Ni-catalyzed ethene dimerization was prepared by grafting Al(CH2 Bu)3 on
partially dehydroxylated silica and characterized by 27Al solid-state NMR augmented by DFT computations. The efficient co-catalytic
properties of the material were assigned to the presence of monografted alkylaluminum species, representing 19 % of all Al species.
7
This article is protected by copyright. All rights reserved.