Magnitude and consequences of or ligand σ-donation on alkene metathesis activity in d0 silica supported (≡SiO)W(NAr)(=CHtBu) (OR) catalysts

Article abstract of DOI:10.1039/c4sc00230j

Well-defined silica supported W catalysts with the general formula [(SiO)W(NAr)(CHtBu)(OR)] (OR = OtBu_F9, OtBu_F6, OtBu _F3, OtBu and OSi(OtBu)₃), prepared by grafting bis-alkoxide complexes [W(NAr)(CHtBu)(OR)₂] on silica dehydroxylated at 700 °C (SiO₂₍₇₀₀₎), display unexpectedly high activity in comparison to their Mo homologues. In this series, the activity of the self-metathesis of cis-4-nonene increases as a function of the OR ligand as follows: OtBu < OtBu_F3 < OSi(OtBu)₃ < OtBu _F6 < OtBu_F9. In addition, the ratio of the two parent metallacyclobutane intermediates, trigonal bipyramidal (TBP)/square pyramidal (SP), which were formed by the metathesis of ethylene and observed by solid-state NMR, follows the same order: OtBu < OtBu_F3 < OSi(OtBu)₃ < OtBu_F6 < OtBu_F9. This provides clear evidence of the decreasing σ-donating ability of the OR ligand with an increasing number of fluorine atoms and the positioning of a siloxy ligand in between OtBu_F3 and OtBu_F6. This study provides the first detailed structure-activity relationship analysis of a series of well-defined heterogeneous catalysts, showing that weaker σ-donor OR ligands lead to higher activity, and that the surface siloxy ligand is a rather small and weak σ-donor ligand overall, thus providing highly active yet stable catalysts. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c4sc00230j

Chemical
Science
View Article Online
View Journal | View Issue
EDGE ARTICLE
Magnitude and consequences of OR ligand
s-donation on alkene metathesis activity in d0 silica
supported (^SiO)W(NAr)(]CHtBu)(OR) catalysts†
Cite this: Chem. Sci., 2014, 5, 2475
*
V. Mougel and C. Coperet
´
Well-defined silica supported W catalysts with the general formula [(^SiO)W(NAr)(]CHtBu)(OR)] (OR ¼
OtBuF9, OtBuF6, OtBuF3, OtBu and OSi(OtBu)3), prepared by grafting bis-alkoxide complexes [W(NAr)-
(]CHtBu)(OR)2] on silica dehydroxylated at 700 ꢀC (SiO2(700)), display unexpectedly high activity in
comparison to their Mo homologues. In this series, the activity of the self-metathesis of cis-4-nonene
increases as a function of the OR ligand as follows: OtBu < OtBuF3 < OSi(OtBu)3 < OtBuF6 < OtBuF9. In
addition, the ratio of the two parent metallacyclobutane intermediates, trigonal bipyramidal (TBP)/square
pyramidal (SP), which were formed by the metathesis of ethylene and observed by solid-state NMR,
follows the same order: OtBu < OtBuF3 < OSi(OtBu)3 < OtBuF6 < OtBuF9. This provides clear evidence of
the decreasing s-donating ability of the OR ligand with an increasing number of fluorine atoms and the
positioning of a siloxy ligand in between OtBuF3 and OtBuF6. This study provides the first detailed
structure–activity relationship analysis of a series of well-defined heterogeneous catalysts, showing that
weaker s-donor OR ligands lead to higher activity, and that the surface siloxy ligand is a rather small and
weak s-donor ligand overall, thus providing highly active yet stable catalysts.
Received 21st January 2014
Accepted 27th February 2014
DOI: 10.1039/c4sc00230j
www.rsc.org/chemicalscience
structure–activity relationship studies over the past 40 years.
One would thus expect that a better understanding of the
Introduction
reaction mechanism and the life cycle of the corresponding
heterogeneous catalysts would be a key step towards improving
their performances (activity, selectivity and stability).
The past 50 years have witnessed extensive alkene metathesis
research in both academia and industry that has resulted in
numerous chemical processes and applications. While the
current petrochemical processes rely on heterogeneous cata-
lysts based on supported transition metal oxides,1,2 the recent
development of homogeneous catalysts based on well-dened
Mo, W, Re and Ru alkylidenes has led to ourishing applica-
tions in the areas of polymers and ne chemicals.3–6 The most
recent outcome even allows Z-selective metathesis to be carried
out, which is a major milestone for the use of metathesis
towards the synthesis of advanced intermediates and complex
molecules.7–13 Furthermore, Mo- and W-based homogeneous
catalysts, when combined with Ir-dehydrogenation cata-
lysts,14–16 can be used for the upgrade of low molecular weight
alkanes into high-grade diesel fuel via a tandem alkane
metathesis process, as an alternative to the single component
system.17–24 All of these applications have recently emerged from
the rational development of homogeneous catalysts through
The mechanism for alkene metathesis, proposed by Chau-
vin, involves the [2 + 2] cycloaddition of an alkene and a metal
carbene to afford a metallacyclobutane intermediate.25,26 The
metallacyclobutane then undergoes cycloreversion to either
release the original reagents (degenerate), or produce a new
alkene/alkylidene (productive). In Schrock-type catalysts with
the general formula [M(NR)(]CHR)(X)(Y)] (M ¼ Mo, W), the
anionic X and Y substituents have tremendous inuence on the
stability of the reaction intermediates, the catalytic activity27–29
and the stereoselectivity.7,30–32 A metal alkylidene complex con-
taining X and Y ligands with notably different electron donating
abilities (good and poor) facilitates the coordination of the
incoming alkene, and thereby the subsequent cycloaddition
process.33 The formation, structure and stability of the resulting
metallacycles have been extensively reported for tungsten-based
homogeneous catalysts29,34–38 and a few Mo systems.35,39,40
Detailed solution NMR, X-ray and computational studies have
shown that these metallacycles can adopt either trigonal
bipyramidal (TBP) or square pyramidal (SP) geometries that
interconvert via a turnstile process (Scheme 1).27,28,36,37
¨
Department of Chemistry and Applied Biosciences, Vladimir Prelog Weg 2, ETH Zurich,
¨
CH-8093 Zurich, Switzerland. E-mail: ccoperet@ethz.ch
† Electronic supplementary information (ESI) available: Complete synthetic and
catalytic procedures, crystal structure of [W(NAr)(]CHtBu)(OtBuF9)2], and 1H
and 13C solid state NMR and IR spectra. CCDC 977064. For ESI and
crystallographic data in CIF or other electronic formats, see DOI:
10.1039/c4sc00230j
DFT studies have demonstrated that the TBP isomer is in the
metathesis reaction pathway and participates in the alkylidene
exchange process, whilst the SP isomer is usually more stable
This journal is © The Royal Society of Chemistry 2014
Chem. Sci., 2014, 5, 2475–2481 | 2475

View Article Online
Chemical Science
Edge Article
s-donating OR ligands (stronger electron withdrawing) exhibit
higher activity.
Results and discussion
Synthesis of molecular precursors
The molecular precursors [W(NAr)(]CHtBu)(OR)2], WOR (OR ¼
OtBuF6, OtBu), were synthesized by salt metathesis from the bis-
triate complex [W(NAr)(]CHtBu)(OTf)2(DME)] and lithium
Scheme 1 Catalytic pathway proposed for d0 alkene metathesis.
alkoxides as previously described.38 Note that for OR ¼ OtBuF3
,
we prepared the DME adduct [W(NAr)(]CHtBu)(Ot-
BuF3)2(DME)] (WOtBu ) because it was much more stable and
F3
and corresponds to a catalytic resting state. In addition, SP
metallacycles may also be involved in deactivation
processes.41,42
more convenient to use for graing (vide infra). In addition, we
prepared WOtBu and WOSi(OtBu) via a similar route in 69% and
F9
3
88% yield, respectively, using LiOtBuF9 and KOSi(OtBu)3.59
These two unreported compounds were characterized by NMR
Yet, for heterogeneous catalysts, observing the expected
reaction intermediates, i.e. metallocarbenes and metal-
lacyclobutanes, has remained a challenge, with only a single
example for a well-dened silica supported tungsten catalyst
reported.43 In recent years, numerous well-dened supported
alkylidene metathesis catalysts that exhibit similar, or in some
cases greater, catalytic activity than their homogeneous
analogues have been developed.22,44–58 In particular, the
controlled graing of well-dened Mo and W precursors onto
silica partially dehydroxylated at 700 ꢀC (SiO2(700)) provides
“single-site” silica-supported systems, where the site isolation
prevents binuclear decomposition, thus increasing the stability
of the catalyst.56,57 The resulting silica-supported alkylidene
complexes [(^SiO)M(NAr)(]CHR)(X)] have intrinsically
dissymmetric metal centers and produce highly reactive surface
species. The reactivity and stability of these complexes can be
tuned by varying the X-ligand. Whilst the most active catalysts
are based on M ¼ Mo with X ¼ pyrrolyl,54 the hexauoro-tert-
butoxide (X ¼ OtBuF6) showed promising performance.55 We
also showed that the analogous W-imido catalysts (X ¼ CH2tBu
and 2,5-dimethylpyrrolyl) are less active than their Mo-
analogues by ca. one order of magnitude in the self-metathesis
of propene, but display much greater stability. In fact, we
exploited the stability of the W-imido species to observe the
metallacyclobutane intermediates on a supported catalyst for
the rst time.43 Clearly, the electronic environment around the
metal center plays a crucial role in the reactivity of the sup-
ported alkylidene. Quantication of the s-donation properties
of the anionic X ligand is critical to the elucidation of the
structure–activity relationship, but has remained elusive.
Here, we investigate a family of well-dened silica supported
tungsten alkylidene alkoxide complexes, obtained from graing
[W(NAr)(]CHtBu)(OR)2] complexes {OR ¼ OC(CF3)3 (OtBuF9);
OCMe(CF3)2 (OtBuF6); OCMe2(CF3) (OtBuF3); OtBu; OSi(OtBu)3;
Ar ¼ 2,6-iPr2C6H3} onto silica partially dehydroxylated at 700 ꢀC,
SiO2(700). These supported catalysts display activities and
stabilities that in most cases surpass their molecular precur-
sors, and even the best silica-supported Mo-analogues. Evalua-
tion of the s-donating abilities of the OR-ligands, from the
spectroscopically observed ratio of TBP/SP metallacycles, allows
us to establish a detailed structure–reactivity relationship for
[(^SiO)W(NAr)(]CHtBu)(OR)], where catalysts with weaker
and elemental analysis, and the crystal structure of WOtBu was
F9
resolved (see ESI†).60
Graing of the molecular precursors
Graing [W(NAr)(]CHtBu)(OR)2] (OR ¼ OtBuF9, OtBuF6, OtBu,
OSi(OtBu)3) or [W(NAr)(]CHtBu)(OtBuF3)2(DME)] on SiO2(700)
yielded the corresponding materials (WOR@SiO2). Monitoring
the graing step by IR spectroscopy showed that the isolated
surface silanols at 3747 cmꢁ1 for SiO2(700) disappeared, while
the stretching and bending (C–H) vibrations of the ligands
appeared in the 3100–2800 and 1500–1350 cmꢁ1 regions,
respectively (Fig. S2†). A broad band at 3650 cmꢁ1 was present
in each material, and is associated with the remaining
unreacted OH groups interacting with the adjacent ligands. The
shi in the maximum of the band to a higher wavenumber with
increased uorine content of the OR ligand indicates that this
interaction involves the tBu ligands and the uorine when they
are present. W elemental analysis is consistent with the partial
consumption of the OH groups (0.18–0.21 mmol W gꢁ1 vs.
0.26 mmol gꢁ1 OH). In all cases, graing is accompanied by the
release of ROH, consistent with the formation of the expected
surface complexes [(^SiO)W(NAr)(]CHtBu)(OR)] (OR
OtBuF9 (1), OtBuF6 (2), OtBuF3 (3), OtBu (4), OSi(OtBu)3 (5)
¼
(Scheme 2)).
More specically, graing of WOR (OR ¼ OtBuF9 and OtBuF6
)
releases ca. 0.8 equiv. of HOR per surface silanol, which is good
evidence for 1 and 2 being formed as the major surface species.
Scheme
2 Grafting of WOR molecular complexes on SiO2(700),
affording WOtBu @SiO2, WOtBu @SiO2, WOtBu @SiO2, WOtBu@SiO2 or
F9
F6
F3
WOSi(OtBu) @SiO2. For OR ¼ OtBu, grafting led to ca. 40% [(^SiO)
3
W(NAr)(CH2tBu)(OtBu)2] byproduct.
2476 | Chem. Sci., 2014, 5, 2475–2481
This journal is © The Royal Society of Chemistry 2014

View Article Online
Edge Article
Chemical Science
For [W(NAr)(]CHtBu)(OtBuF3)2(DME)], graing is accompa- substrate. All catalytic tests involved placing the solid catalysts
nied by the release of ca. 0.8 equiv. of HOtBuF3 and 0.9 equiv. of (0.1 mol% or 0.02 mol%) in contact with a 0.8 M toluene solu-
DME, consistent with the formation of 3 as a major species. In tion of cis-4-nonene at 30 ꢀC. All of the supported catalysts
the case of [W(NAr)(]CHtBu)(OtBu)2], only 0.4 equiv. of tBuOH
WOR@SiO2 are active in metathesis under the conditions
is released when graing is performed under the same condi- described above, resulting in the formation of 4-octenes and 5-
tions. This value increases to 0.6 equiv. when graing is carried decenes as the only products which could be observed by gas
out at a lower temperature (ꢁ40 ꢀC in toluene). In this case, chromatography (GC) (Table 2). Whilst thermodynamic
[(^SiO)W(NAr)(]CHtBu)(OtBu)] (4) is formed, along with conversion is reached for all catalysts at 0.1 mol% loading, their
[(^SiO)W(NAr)(CH2tBu)(OtBu)2] (ca. 30–40%) (vide infra), activity increases with an increase in the electron withdrawing
which probably results from the subsequent reaction of the character of the alkoxide ligands (OtBu < OtBuF3 < OtBuF6
<
released tBuOH with the surface species 4.43,58,61 For [W(NAr)- OtBuF9), and OSi(OtBu)3 is placed in between OtBuF3 and
(]CHtBu)(OSi(OtBu)3)2], 0.5 equiv. of HOSi(OtBu)3 per surface OtBuF6. For comparison, the catalytic activity of the molecular
silanol is released, and a low loading is obtained upon graing precursors [W(NAr)(]CHtBu)(OR)2] (OR ¼ OtBuF9, OtBuF6
,
of WOSi(OtBu) , probably because of the large size of OSi(OtBu)3. OtBuF3, OtBu, OSi(OtBu)3) was investigated under the same
3
The compositions of WOR@SiO2 were conrmed by elemental reaction conditions. A similar trend is observed for OR ¼ OtBu,
38
analysis of C, N, H and F (Table 1).
OtBuF3 and OtBuF6
,
but not for OtBuF9, which is placed in
The 1H Magic Angle Spinning (MAS) NMR spectra of between OtBuF3 and OtBuF6 (Fig. S28 and Table S5†).
compounds WOR@SiO2 (Fig. S3–11†) displayed signals with a Surprisingly, the molecular complex [W(NAr)(]CHtBu)-
chemical shi very close to those of the molecular precursors, in (OSi(OtBu)3)2] is inactive (<1% conversion aer 24 h), despite
agreement with the formation of [(^SiO)W(NAr)- the high activity of the graed complex. One possible explana-
(]CHtBu)(OR)] as the major surface species. The resonance of tion for this deviation in activity between the graed complexes
alkylidene W]CHtBu is shown at 8.92 ppm for 1, 8.60 ppm for and their molecular analogues is that the electron withdrawing
2, 8.29 ppm for 3, 7.99 ppm for 4 and 8.35 ppm for 5, following ability of the OtBuF9 and OSi(OtBu)3 ligands is compensated for
the same trend which was observed for the chemical shi of the by their relatively large size, as they have a larger buried volume
parent molecular complexes in solution (Table 1). The 13C cross
polarisation magic angle spinning (CP MAS) solid state NMR
spectra of WOR@SiO2 (Fig. S4–12†) also show good agreement
Table
2 Initial TOF and time to equilibrium of the supported
with the proposed structures. Whilst it was not possible to get
complexes towards the self-metathesis of cis-4-nonene in toluene,
with a catalyst loading of 0.1 mol% at 30 ꢀCa
direct evidence of the alkylidene ligand by 13C NMR without 13
C
enrichment at this position,43,61 the spectrum for each material
contains two sharp signals at 46.1 and 32.4 ppm for 1, 45.2 and
32.2 ppm for 2, 45.0 and 32.4 ppm for 3, 44.5 and 33.0 ppm for 4
and 45.5 and 33.4 ppm for 5, assignable to the b- and g- carbons
of the neopentylidene group. An extra signal at 72 ppm is
observed in the 13C CP MAS NMR spectrum of WOtBu@SiO2,
consistent with the formation of [(^SiO)W(NAr)(CH2t-
Bu)(OtBu)2], likely from the partial protonation of the alkylidene
by tBuOH.43,61 Such a signal is not observed for the other sup-
ported alkylidenes, showing that if it is present, it corresponds
only to minor surface species (<10%).
Silica supported
Unsupported
b
Precursor
TOF3 min
Timec
TOF3 min
Timec
b
WOtBu
115
75
15
5
41
46
<10 min
<30 min
<120 min
<480 min
<30 min
<30 min
93
>166
61
<1
<1
<10 min
<3 min
F9
WOtBu
F6
WOtBu
24 h (39%)d
24 h (<2%)d
24 h (<1%)d
<3 min
F3
WOtBu
WOSi(OtBu)
3
MoOtBu
>166
F6
a
0.8 M solution of cis-4-nonene in toluene, containing heptane as the
internal standard (0.1 M); the loading in mol% was calculated based
b
on the W wt%. TOF is dened here as the number of new alkenes
produced per W atom and unit of time, without accounting for
c
isomerization of the precursor to trans-4-nonene.62
Time to
Catalytic activity
d
equilibrium conversion. Values in parentheses are given when the
system did not reach the equilibrium conversion, and correspond to
the measured conversion aer 24 h.
We evaluated the catalytic activity of all the complexes in the
self-metathesis of cis-4-nonene as
a prototypical liquid
Table 1 Comparative mass balance analysis and NMR chemical shifts observed for the samples after grafting
Entry
Wa (wt%) ROHreleasedb (RO/Wsurf) C (wt%) C/Wc
H (wt%) H/Wc
N (wt%) N/Wc F (wt%) F/Wc 1Halkylidened (ppm)
WOtBu @SiO2
3.47 (0.19) 0.76
3.88 (0.21) 0.80
3.71 (0.20) 0.85
3.32 (0.18) 0.60
4.84
5.39
5.18
4.78
4.72
21.3 (21) 0.51
26.8 (27) 0.38
30.6 (30) 0.37
35.4 (33) 0.38
37.4 (36) 0.48
55.9 (54) 0.29
1.4 (1) 3.14
1.3 (1) 2.05
1.3 (1) 1.09
1.9 (1) —
8.8 (9) 8.92 (9.46)
5.1 (6) 8.60 (8.83)
2.8 (3) 8.29 (8.38)
F9
WOtBu @SiO2
21.3 (21) 0.65
21.4 (21) 0.72
22.0 (21) 0.68
29.1 (29) 0.76
F6
WOtBu @SiO2
F3
WOtBu@SiO2
—
7.99 (8.01)
8.35 (8.19)
WOSi(OtBu) @SiO2 2.48 (0.15) 0.54
1.5 (1) —
—
3
a
Value in mmol gꢁ1 is given in brackets. b Protonated form of the OR ligand released upon graing of the precursor, and quantied by 1H NMR
c
d
analysis of the ltrate aer graing with internal standard. The theoretical value is given in brackets. Chemical shi observed for the parent
molecular complex is given in brackets.
This journal is © The Royal Society of Chemistry 2014
Chem. Sci., 2014, 5, 2475–2481 | 2477

View Article Online
Chemical Science
Edge Article
according to Cavallo's model (Table S4†).63 The lower affinity of recovered, likely due to slower kinetics. The resulting solids
the peruorinated OtBuF9 ligand towards the hydrocarbon64 were evacuated under high vacuum (10ꢁ5 mbar) for 1.5 h and
1
might also inuence its reactivity by preventing the alkane from analyzed by H and 13C solid state NMR. Whilst the signals of
approaching and the formation of the metallacyclobutane. Note the alkylidene proton were clearly observed in the 1H solid state
that graing also dramatically enhanced the activity of the OtBu NMR spectra of the supported catalyst precursors, these signals
complex, converting a virtually inactive complex into an active disappeared upon exposure to ethylene, conrming the reaction
catalyst, as already observed for the Mo system.55 The graing of the parent alkylidene (Fig. 1a and S12–22†). Moreover, the 13
C
on silica allowed the properties of alkoxide ligand series to be CP MAS NMR spectra of the complexes contained new peaks
ranked, which was not possible for the molecular precursors associated with the formation of metallacyclobutane interme-
due to steric constraints.
diates. It is noteworthy that no signals attributable to the
These silica supported tungsten arylimido alkylidene methylidene species were observed in this series.43 The signals
complexes are highly active alkene metathesis catalysts. In fact, at about 100 ppm and ꢁ5 ppm correspond to the TBP metal-
it is noteworthy that WOtBu @SiO2 displays greater catalytic lacyclobutane, and the signals at about 50 ppm are associated
F6
activity than the corresponding Mo complex bearing the same with the SP isomer. The resonances corresponding to the a and
ligands [(^SiO)Mo(NAr)(]CHtBu)(OtBuF6)] (Table 2). It should b carbon signals of the TBP metallacycle were observed only in
be noted that the catalysts WOtBu @SiO2 and WOtBu @SiO2 the cases of 1–3 and 5. These systems also show 13C signals
F9
F6
readily catalyze the metathesis of cis-4-nonene with loadings as corresponding to the SP isomer, whilst this is the only signal
low as 0.02 mol%, reaching equilibrium in less than 3.5 h for observed in the spectra for 4 (Fig. 1b–f and Table 3). All of these
WOtBu @SiO2 and 12 h for WOtBu @SiO2 (Table S7 and assignments were conrmed by heteronuclear correlation
F9
F6
Fig. S30†). The activity for cis-4-nonene metathesis at very low (HETCOR) experiments (Fig. S22–S27†). The TBP/SP isomer
loading was investigated for the most active complex WOt- ratio can be obtained by the integration of the metallacycle
BuF9@SiO2, and full conversion was obtained with a catalyst signals in the 13C CP MAS experiments. Whilst CP MAS exper-
ꢀ
loading as low as 50 ppm in 48 h at 30 C.
iments are not ideal for the quantitative integration of carbon
signals, the similar natures of the carbon species (all CH2)
allows for the TBP/SP ratio to be properly estimated. In the tert-
butoxide ligand series, this ratio increases with the increasing
number of ligand triuoromethyl groups, gradually varying
from a TBP/SP ratio of 8 with OR ¼ OtBuF9 to ca. 0 for the OtBu
Stability of the graed complexes
Dimerization of the molecular precursors during a reaction has
been proposed to be one of the main routes for the deactivation
of molecular tungsten alkylidene complexes.65 Schrock reported
that the molecular complex [W(NAr)(]CHtBu)(OtBuF3)2] has
high initial catalytic activity towards cis-alkenes, although this
catalyst rapidly deactivates via dimerization.15,18 We have here
investigated the stability of the supported analogue WOtBu
@
F3
SiO2. The exposure of WOtBu @SiO2 to 20 equivalents of cis-3-
F3
hexene in toluene for 12 h66 prior to a catalytic test with cis-4-
nonene did not result in any signicant change in the catalytic
activity of the catalyst (see ESI†). This is in sharp contrast to its
molecular precursor [W(NAr)(]CHCMe2Ph)(OtBuF3)2], which
dimerizes under the same conditions to yield the inactive
species {W(NAr)(OtBuF3)2}2,38,67 conrming that it has a higher
stability towards bimolecular deactivation, due to site isolation
on a silica surface. In fact, WOtBu @SiO2 can be used and
F3
recycled more than 4 times in the presence of 1000 equivalents
of cis-4-nonene without any noticeable ageing of the catalyst
(similar initial rate and time to nal conversion – see ESI†).
Metallacycles
In order to obtain more insight into the catalyst active site, we
investigated the structures of the reaction intermediates by
solid state NMR. The WOR@SiO2 catalystsꢀwere exposed to
13
10 equiv. of C di-labeled ethylene at ꢁ196 C. Aer standing
for 1 h at room temperature, a slight color change to red–orange
was observed, and the volatiles were analyzed by gas chroma-
tography-mass spectrometry (GC-MS), which indicated the
presence of 0.7–0.8 equiv. of mono-labeled 13C 3,3-dime-
thylbutene (see ESI for details†). Note that for OR ¼ OtBu, only
Fig. 1 1H MAS solid state NMR spectra of 2 before (black) and after
(red) exposure to 10 equivalents of 13C ethylene (a). 13C CP MAS solid
state NMR spectra of WOtBu (b), WOtBu (c), WOSi(OtBu) (d), WOtBu (e)
F3
3
F6
and WOtBu (f) after exposure to 10 equivalents of 13C ethylene. The
F9
MAS frequency was set to 10 kHz, and the contact time of the CP was
0.2 equiv. of mono-labeled 13C 3,3-dimethylbutene was 250 ms. * indicates spinning side bands.
2478 | Chem. Sci., 2014, 5, 2475–2481
This journal is © The Royal Society of Chemistry 2014

View Article Online
Edge Article
Chemical Science
Table 3 Solid state NMR data for the metallacyclobutane surface
complexes
stable unsubstituted metallacycles into active species, thus
indicating the importance of the substitution of the metalla-
cycles on the reaction rates.69
Compound
dTBP (Ca)a
dTBP (Cb)a
dSP (Ca)a
Ratio TBP/SPb
WOtBu
WOtBu
WOtBu
WOtBu
102.6 (4.1)
100.7 (4.3)
102.3 (4.1)
—
ꢁ2.7 (ꢁ0.7)
ꢁ5.4 (ꢁ0.7)
ꢁ4.2 (0.0)
—
50 (1.9)
83/17
43/57
9/91
0/100
22/78
F9
Conclusions
49.5 (1.6)
47.3 (1.8)
42.5 (2.1)
48.6 (1.7)
F6
F3
A series of well-dened silica-supported alkylidene alkoxide
tungsten catalysts [(^SiO)W(NAr)(]CHtBu)(OR)] (OR
¼
WOSi(OtBu)
101.5 (4.2)
ꢁ6.3 (ꢁ0.1)
3
OtBuF9, OtBuF6, OtBuF3, OtBu, OSi(OtBu)3) were prepared via
surface organometallic chemistry. These catalysts display cata-
lytic performances, including rates, which can exceed those of
the corresponding well-dened silica supported Mo-equivalents
with OtBuF6. The order of reactivity follows the electron with-
a
Values given in ppm, determined by CP MAS experiments at 10 kHz.
The corresponding proton chemical shi is given in parentheses.
b
Values given in %, determined as detailed in ESI.†
drawing character of the alkoxide ligands (OtBuF9 > OtBuF6
>
analogue (Table 3). With a TBP/SP ratio of 2.5, the OSi(OtBu)3
complex lies in between its OtBuF3 and OtBuF6 analogues.36
The formation of metallacyclobutane from the ve alkyli-
dene WOR@SiO2 precursors highlights the higher stability of
the metallacycles with respect to the separated reactants in
these systems, as suggested previously by theoretical
studies.28,42 Stronger s-donor ligands favor SP- over TBP-geom-
etries, which could be explained by their stronger trans-inu-
ence.68 The SP/TBP isomer ratios thus appear to be a good
measure of the trans-inuence of the OR-ligand (s-donation).
The SP/TBP ratio decreases for this series of dissymmetric
OtBuF3 > OtBu), and the siloxy ligand (OSi(OtBu)3) is placed in
between the OtBuF6 and OtBuF3 ligands. The increased activity
of the complexes bearing OSi(OtBu)3 and OtBuF9 ligands upon
graing highlights the fact that silica acts as a rather small
electron withdrawing ligand. It is noteworthy that the supported
catalysts, including the most reactive ones, display very good
stability, typically higher than their molecular analogues,
allowing for them to be recycled and the use of very low catalyst
loadings down to 50 ppm. In addition, these well-dened sup-
ported systems allowed for the direct observation of the met-
allacyclobutane intermediates of the supported alkene
metathesis catalysts, and the quantication of the SP/TBP ratio
of the metallacyclobutane intermediates, which provides a way
to evaluate the trans-inuence (s-donation) of the OR-ligand
(OtBuF9 < OtBuF6 < OSi(OtBu)3 < OtBuF3 < OtBu). In this series,
strong electron-withdrawing OR ligands favor TBP intermedi-
ates and afford more active catalysts, thus establishing the
structure–activity relationship at the molecular level and paving
the way for the more rational improvement of heterogeneous
alkene metathesis catalysts.
surfaceꢁcomplexes, following the order tBuOꢁ > tBuF3Oꢁ
>
tBuF6O > tBuF9Oꢁ, in line with the decreasing s-donating
ability of each ligand. The TBP/SP ratio obtained for OR ¼
OSi(OtBu)3 allows its s-donating ability to be ranked in between
OtBuF3 and OtBuF6. Due to the similarity between this ligand
and the isolated silanols found at the surface of SiO2(700), one
can assume that silica would lie in a similar position in this
spectroscopic series, although as previously mentioned, this
does not take into account the larger steric volume of the
OSi(OtBu)3 ligand compared to a silanol on the silica surface.
Moreover, it is noteworthy that the metathesis rates for the
series of catalysts follow the same trend, directly connecting Acknowledgements
higher rates with fewer s-donor ligands in these systems.
V.M. was supported by an ETH fellowship (co-funded by the
¨
ETH Zurich-Marie Curie action for people, FEL-08 12-2). Dr
Catalytic activity of the metallacycles
Matthew P. Conley and Maxence Valla are thanked for helpful
The catalytic activity of these metallacycles were investigated discussions and NMR spectroscopy assistance.
and compared to the parent alkylidene complexes. As indicated
in Table S8 and Fig. S31,† all of the reaction proles show an S-
type shape: a period with a low TOF just aer the addition of cis-
Notes and references
4-nonene, followed by a second period with a TOF following the
same trend and magnitude as observed for the parent
complexes, with the exception of WOtBu@SiO2. These two pha-
ses (and the initiation period in particular) show that the parent
metallacycles are very stable, do not readily participate in the
metathesis reaction, and only slowly undergo cycloreversion, in
contrast to the tri-substituted metallacyclobutanes involved in
the metathesis of cis-4-nonene. For WOtBu@SiO2, the SP-metal-
lacycle does not show any signicant activity towards cis-4-
nonene, which probably results from its excessively high
stability or its propensity to undergo deactivation.41 Overall, the
initiation phase observed for the metallacycles suggests that the
induction period is due to the slow conversion of the more
1 J. C. Mol, J. Mol. Catal. A: Chem., 2004, 213, 39–45.
2 K. Ivin and H. Mol, in Olen Metathesis and Metathesis
Polymerization (2), ed. K. J. Ivin and J. C. Mol, Academic
Press, London, 1997.
3 R. R. Schrock and A. H. Hoveyda, Angew. Chem., Int. Ed.,
2003, 42, 4592–4633.
4 R. H. Grubbs, Angew. Chem., Int. Ed., 2006, 45, 3760–3765.
5 A. H. Hoveyda, S. J. Malcolmson, S. J. Meek and
A. R. Zhugralin, Angew. Chem., Int. Ed., 2010, 49, 34–44.
6 S. Kress and S. Blechert, Chem. Soc. Rev., 2012, 41, 4389–
4408.
7 I. Ibrahem, M. Yu, R. R. Schrock and A. H. Hoveyda, J. Am.
Chem. Soc., 2009, 131, 3844–3845.
This journal is © The Royal Society of Chemistry 2014
Chem. Sci., 2014, 5, 2475–2481 | 2479

View Article Online
Chemical Science
Edge Article
8 A. J. Jiang, Y. Zhao, R. R. Schrock and A. H. Hoveyda, J. Am. 35 R. R. Schrock, A. J. Jiang, S. C. Marinescu, J. H. Simpson and
Chem. Soc., 2009, 131, 16630–16631. P. Muller, Organometallics, 2010, 29, 5241–5251.
9 S. J. Meek, R. V. O'Brien, J. Llaveria, R. R. Schrock and 36 J. Feldman, W. M. Davis and R. R. Schrock, Organometallics,
A. H. Hoveyda, Nature, 2011, 471, 461–466. 1989, 8, 2266–2268.
10 D. V. Peryshkov, R. R. Schrock, M. K. Takase, P. Muller and 37 J. Feldman, W. M. Davis, J. K. Thomas and R. R. Schrock,
A. H. Hoveyda, J. Am. Chem. Soc., 2011, 133, 20754–20757. Organometallics, 1990, 9, 2535–2548.
11 R. K. M. Khan, S. Torker and A. H. Hoveyda, J. Am. Chem. 38 R. R. Schrock, R. T. Depue, J. Feldman, K. B. Yap, D. C. Yang,
Soc., 2013, 135, 10258–10261.
12 K. Endo and R. H. Grubbs, J. Am. Chem. Soc., 2011, 133,
8525–8527.
W. M. Davis, L. Park, M. Dimare, M. Schoeld, J. Anhaus,
E. Walborsky, E. Evitt, C. Kruger and P. Betz,
Organometallics, 1990, 9, 2262–2275.
13 B. K. Keitz, K. Endo, P. R. Patel, M. B. Herbert and 39 R. R. Schrock, J. S. Murdzek, G. C. Bazan, J. Robbins,
R. H. Grubbs, J. Am. Chem. Soc., 2012, 134, 693–699.
14 A. S. Goldman, A. H. Roy, Z. Huang, R. Ahuja, W. Schinski
and M. Brookhart, Science, 2006, 312, 257–261.
M. Dimare and M. Oregan, J. Am. Chem. Soc., 1990, 112,
3875–3886.
40 G. C. Bazan, R. R. Schrock and M. B. Oregan,
Organometallics, 1991, 10, 1062–1067.
15 B. C. Bailey, R. R. Schrock, S. Kundu, A. S. Goldman,
Z. Huang and M. Brookhart, Organometallics, 2009, 28, 41 A. M. Leduc, A. Salameh, D. Soulivong, M. Chabanas,
´
355–360.
J. M. Basset, C. Coperet, X. Solans-Monfort, E. Clot,
16 M. C. Haibach, S. Kundu, M. Brookhart and A. S. Goldman,
Acc. Chem. Res., 2012, 45, 947–958.
O. Eisenstein, V. P. W. Bohm and M. Roper, J. Am. Chem.
Soc., 2008, 130, 6288–6297.
´
17 V. Vidal, A. Theolier, J. ThivolleCazat and J. M. Basset, 42 X. Solans-Monfort, C. Coperet and O. Eisenstein, J. Am.
Science, 1997, 276, 99–102. Chem. Soc., 2010, 132, 7750–7757.
´
´
18 C. Coperet, O. Maury, J. Thivolle-Cazat and J. M. Basset, 43 F. Blanc, R. Berthoud, C. Coperet, A. Lesage, L. Emsley,
Angew. Chem., Int. Ed., 2001, 40, 2331–2334.
19 E. Le Roux, M. Chabanas, A. Baudouin, A. de Mallmann,
R. Singh, T. Kreickmann and R. R. Schrock, Proc.Natl.
Acad. Sci. U. S. A., 2008, 105, 12123–12127.
´
C. Coperet, E. A. Quadrelli, J. Thivolle-Cazat, J. M. Basset, 44 W. A. Herrmann, A. W. Stumpf, T. Priermeier, S. Bogdanovic,
W. Lukens, A. Lesage, L. Emsley and G. J. Sunley, J. Am.
Chem. Soc., 2004, 126, 13391–13399.
V. Dufaud and J. M. Basset, Angew. Chem., Int. Ed., 1996, 35,
2803–2805.
´
´
20 E. Le Roux, M. Taouk, C. Coperet, A. de Mallmann, 45 F. Blanc, M. Chabanas, C. Coperet, B. Fenet and
J. Thivolle-Cazat, J. M. Basset, B. M. Maunders and
G. J. Sunley, Angew. Chem., Int. Ed., 2005, 44, 6755–6758.
E. Herdweck, J. Organomet. Chem., 2005, 690, 5014–5026.
46 S. I. Wolke and R. Buffon, J. Mol. Catal. A: Chem., 2000, 160,
181–187.
´
21 F. Blanc, C. Coperet, J. Thivolle-Cazat and J.-M. Basset,
´
Angew. Chem., Int. Ed., 2006, 45, 6201–6203.
47 M. Chabanas, A. Baudouin, C. Coperet and J. M. Basset,
´
22 F. Blanc, J. Thivolle-Cazat, J. M. Basset and C. Coperet, Chem.
J. Am. Chem. Soc., 2001, 123, 2062–2063.
´
– Eur. J., 2008, 14, 9030–9037.
48 M. Chabanas, A. Baudouin, C. Coperet, J.-M. Basset,
´
23 J.-M. Basset, C. Coperet, D. Soulivong, M. Taouk and
W. Lukens, A. Lesage, S. Hediger and L. Emsley, J. Am.
Chem. Soc., 2003, 125, 492–504.
J. T. Cazat, Acc. Chem. Res., 2010, 43, 323–334.
´
´
24 C. Coperet, Chem. Rev., 2010, 110, 656–680.
49 M. Chabanas, C. Coperet and J.-M. Basset, Chem. – Eur. J.,
25 Y. Chauvin, Angew. Chem., Int. Ed., 2006, 45, 3740–3747.
2003, 9, 971–975.
26 J. L. Herisson and Y. Chauvin, Makromol. Chem., 1971, 141, 50 H. Balcar, N. Zilkova, J. Sedlacek and J. Zednik, J. Mol. Catal.
161–176. A: Chem., 2005, 232, 53–58.
27 A. Poater, X. Solans-Monfort, E. Clot, C. Coperet and 51 F. Blanc, C. Coperet, J. Thivolle-Cazat, J. M. Basset, A. Lesage,
´
´
O. Eisenstein, J. Am. Chem. Soc., 2007, 129, 8207–8216.
L. Emsley, A. Sinha and R. R. Schrock, Angew. Chem., Int. Ed.,
2006, 45, 1216–1220.
´
28 X. Solans-Monfort, C. Coperet and O. Eisenstein,
Organometallics, 2012, 31, 6812–6822.
52 B. Rhers, A. Salameh, A. Baudouin, E. A. Quadrelli,
´
29 J. Feldman, R. T. Depue, C. J. Schaverien, W. M. Davis and
R. R. Schrock, Advances in Metal Carbene Chemistry, 1989.
30 S. J. Malcolmson, S. J. Meek, E. S. Sattely, R. R. Schrock and
A. H. Hoveyda, Nature, 2008, 456, 933–937.
M. Taouk, C. Coperet, F. Lefebvre, J. M. Basset, X. Solans-
Monfort, O. Eisenstein, W. W. Lukens, L. P. H. Lopez,
A. Sinha and R. R. Schrock, Organometallics, 2006, 25,
3554–3557.
´
31 S. C. Marinescu, D. S. Levine, Y. Zhao, R. R. Schrock and 53 F. Blanc, R. Berthoud, A. Salameh, J. M. Basset, C. Coperet,
A. H. Hoveyda, J. Am. Chem. Soc., 2011, 133, 11512–11514.
32 S. C. Marinescu, R. R. Schrock, P. Muller, M. K. Takase and
A. H. Hoveyda, Organometallics, 2011, 30, 1780–1782.
R. Singh and R. R. Schrock, J. Am. Chem. Soc., 2007, 129,
8434–8435.
´
54 F. Blanc, J. Thivolle-Cazat, J. M. Basset, C. Coperet,
´
33 X. Solans-Monfort, E. Clot, C. Coperet and O. Eisenstein,
A. S. Hock, Z. J. Tonzetich and R. R. Schrock, J. Am. Chem.
Soc., 2007, 129, 1044–1045.
J. Am. Chem. Soc., 2005, 127, 14015–14025.
34 J. Feldman and R. R. Schrock, Prog. Inorg. Chem., 1991, 39, 55 N. Rendon, R. Berthoud, F. Blanc, D. Gajan, T. Maishal,
´
1–74.
J. M. Basset, C. Coperet, A. Lesage, L. Emsley,
2480 | Chem. Sci., 2014, 5, 2475–2481
This journal is © The Royal Society of Chemistry 2014

View Article Online
Edge Article
Chemical Science
S. C. Marinescu, R. Singh and R. R. Schrock, Chem. – Eur. J., 61 M. P. Conley, V. Mougel, D. V. Peryshkov, W. P. Forrest,
´
2009, 15, 5083–5089.
D. Gajan, A. Lesage, L. Emsley, C. Coperet and
´
56 C. Coperet, Dalton Trans., 2007, 5498–5504.
R. R. Schrock, J. Am. Chem. Soc., 2013, 135, 19068–19070.
´
57 N. Rendon, F. Blanc and C. Coperet, Coord. Chem. Rev., 2009, 62 J. L. Bilhou, J. M. Basset, R. Mutin and W. F. Graydon, J. Am.
253, 2015–2020. Chem. Soc., 1977, 99, 4083–4090.
58 D. Gajan, N. Rendon, K. M. Wampler, J. M. Basset, 63 https://www.molnac.unisa.it/OMtools/sambvca.php.
´
C. Coperet, A. Lesage, L. Emsley and R. R. Schrock, Dalton 64 D. O'Hagan, Chem. Soc. Rev., 2008, 37, 308–319.
Trans., 2010, 39, 8547–8551.
59 It should be noted that the synthesis of the neophylidene
65 A. K. Rappe and W. A. Goddard, J. Am. Chem. Soc., 1982, 104,
448–456.
analogue of WOtBuF9 [W(NAr)(CHMe2Ph)(OtBuF9)2] was 66 Conditions used in ref. 67 for the quantitative dimerization
recently described in ref. 15. of the molecular precursor WOtBuF3
60 The overall geometry of WOtBuF9 is pseudotetrahedral with 67 L. P. H. Lopez, R. R. Schrock and P. Muller, Organometallics,
a syn-alkylidene, two symmetric alkoxides and a linear imido 2006, 25, 1978–1986.
ligand, similarly to the previously described crystal 68 B. J. Coe and S. J. Glenwright, Coord. Chem. Rev., 2000, 203,
.
structures of its WOR analogues (OR ¼ OtBuF6, OtBuF3 and
5–80.
˚
OtBu), with a W]C bond of 1.870 (8) A, in the range of 69 C. Wang, F. Haeffner, R. R. Schrock and A. H. Hoveyda,
previously described W-carbene bond (Fig. S1 and Tables
Angew. Chem., Int. Ed., 2013, 52, 1939–1943.
S1–3†).
This journal is © The Royal Society of Chemistry 2014
Chem. Sci., 2014, 5, 2475–2481 | 2481