Ethene-induced temporary inhibition of grubbs metathesis catalysts

Article abstract of DOI:10.1002/adsc.201100509

Grubbs-type catalysts were immobilized in form of supported ionic liquid phase (SILP) materials and applied in the gas-phase cross-metathesis of various substrates in order to study the effect of ethene on catalyst activity and stability in detail. From our theoretical and experimental findings we suggest that ethene causes temporary inhibition effects due to a reversible shift of ruthenium complexes from the productive catalytic cycle towards unproductive, dormant ruthenium species. By avoiding the presence of ethene, constant catalytic activity and selectivity for more than 20 days of continuous operation could be realized. Copyright

Full text of DOI:10.1002/adsc.201100509

FULL PAPERS
DOI: 10.1002/adsc.201100509
Ethene-Induced Temporary Inhibition of Grubbs Metathesis
Catalysts
a
a
b
c
Judith Scholz, Soebiakto Loekman, Normen Szesni, Wolfgang Hieringer,
c
d,
a,
Andreas Gçrling, Marco Haumann, * and Peter Wasserscheid *
Lehrstuhl fꢀr Chemische Reaktionstechnik, Universitꢁt Erlangen-Nꢀrnberg, Egerlandstr. 3, 91058 Erlangen, Germany
Fax : (+49)-9131-852-7421; phone: (+49)-9131-852-7420; e-mail: wasserscheid@crt.cbi.uni-erlangen.de
Sꢀd-Chemie AG, Catalytic Technologies R&D, Waldheimer Str. 13, 83052 Bruckmꢀhl, Germany
Lehrstuhl fꢀr Theoretische Chemie, Universitꢁt Erlangen-Nꢀrnberg, Egerlandstr. 3, 91058 Erlangen, Germany
Chemical Reaction Engineering, Friedrich-Alexander-Universitꢁt, Campus Busan, 1276 Jisa-Dong, Gangseo-Gu, Busan
a
b
c
d
618-230, South Korea
E-mail: marco.haumann@busan-fau.org
Received: June 30, 2011; Published online: October 10, 2011
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/adcs.201100509.
Abstract: Grubbs-type catalysts were immobilized in unproductive, dormant ruthenium species. By avoid-
form of supported ionic liquid phase (SILP) materi- ing the presence of ethene, constant catalytic activity
als and applied in the gas-phase cross-metathesis of and selectivity for more than 20 days of continuous
various substrates in order to study the effect of operation could be realized.
ethene on catalyst activity and stability in detail.
From our theoretical and experimental findings we
suggest that ethene causes temporary inhibition ef- Keywords: DFT calculations; ionic liquids; metathe-
fects due to a reversible shift of ruthenium com- sis; ruthenium; SILP catalysis; supported ionic liquid
plexes from the productive catalytic cycle towards phase (SILP)
Introduction
process, the applied catalyst complexes were immobi-
lized in a thin film of ionic liquid dispersed on a
The development of highly active homogeneous meta- porous support. Such supported ionic liquid phase
thesis catalysts that do not require co-catalyst activa-
tion has been a major breakthrough for organic syn-
[
1]
thesis in the past decades. Ruthenium-carbene com-
plexes of the general type [RuCl L (=CHR)], known
AHCTUNGTRENNUNG
2
2
as Grubbs catalysts, are tolerant towards a variety of
functional groups in the substrate molecules and are
[
2]
kinetically stable against water and air. Additionally,
the development of modified, phosphine-free Grubbs
catalysts by Hoveyda and co-workers has given access
to even more reactive and also somewhat more stable
[
3]
metathesis catalysts.
In this contribution, we report kinetic studies deal-
ing with immobilized Grubbs metathesis catalysts in
continuous gas-phase cross-metathesis reactions of
short-chain, unfunctionalized alkenes under very mild
conditions. The study applies the Ru complexes 2, 4
and 5 (Scheme 1) and focuses on the effect of ethene
on catalyst activity and stability. To realize homogene- Scheme 1. Selected Grubbs-type metathesis catalysts. PCy =
3
ous catalysis conditions in the continuous gas-phase tricyclohexylphosphine; Mes=2,4,6-trimethylphenyl.
Adv. Synth. Catal. 2011, 353, 2701 – 2707
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2701

Judith Scholz et al.
FULL PAPERS
(
SILP) materials have proven to be highly suitable for fluoroborate ([BMIM][BF ]) on calcined silica 60.
4
a variety of gas-phase and batch reactions as demon- The so obtained 2-SILP, 4-SILP and 5-SILP materials
[
4]
strated by us and other groups. Recently, Hagiwara were tested in the propene self-metathesis at 408C
and co-workers reported on the synthesis and applica- and 1 bar using an average residence time of 32 s in
[
5]
tion of SILP metathesis catalysts. In Hagiwaraꢃs the reactor. Note that propene metathesis affords the
work, SILP catalysts based on complexes 1, 3 and 4 products ethene and 2-butene under the applied reac-
(
(
Scheme 1) were applied in slurry phase reactions tion conditions in a maximum, equilibrium-limited
solid SILP material dispersed in liquid reaction mix- conversion of 34%. Disappointingly, all three SILP
ture) and the systems proved to be active in a variety catalysts showed a rapid decrease in conversion
of ring-closing metathesis reactions. After reaction, versus time on stream, the 5-SILP catalyst being the
the SILP catalysts were isolated by filtration and most active representative.
could be reused for three times before showing clear
signs of deactivation.
In seeking to understand the influence of the im-
mobilizing ionic liquid on the catalyst performance,
Although SILP catalysts are applied in this work, the 5-SILP catalyst was prepared by using a number
the scope of our contribution goes far beyond demon- of different ionic liquids, namely [BMIM][BF ],
4
strating the SILP technology for a gas-phase metathe- [BMIM][PF ],
[BMIM][SbF ],
[BMIM][OTf],
6
6
sis process. Our aim is to apply SILP catalysis to dem- [BMIM][NTf ], and [Me AHCTUNGRETG(NNUN C H O) MIM][BF ]. The
2
2
4
3
4
À
onstrate the influence of ethene on the activity and lowest catalyst stability was obtained in [NTf ] -based
2
stability of immobilized, Ru-based metathesis cata- ionic liquids whereas the most stable systems con-
À
lysts in a very general sense. Our surprising experi- tained the [BF4] anion. The experimentally estab-
À
À
3
mental findings have been rationalized with the help lished order of stability was [BF ] >[CF SO ] @
4
3
À
À
À
2
of density-functional theory (DFT) calculations.
[PF ] >[SbF ] >[NTf ] . Note that the observed sta-
6 6
bility order does neither correlate to the IL anionꢃs
[
6]
[7]
coordination strength nor to the anionꢃs thermal
[
8]
Results and Discussion
or hydrolytic stability.
Therefore, a deactivation mechanism based on the
All kinetic studies were carried out as continuous gas- coordination of the IL counter-ion or on IL decompo-
phase reactions in classical fixed-bed reactors with sition products is unlikely. In searching for other pos-
direct product analysis via online gas chromatography. sible correlations, we realized that the SILP catalyst
Details about the reactor set-up and all chemicals ap- stability order does indeed match surprisingly well
plied in this study can be found in the Experimental with the order of ethene solubility in the different
Section. Figure 1 shows the catalytic performance of ionic liquids. While the least stable 5-SILP catalyst
the three single-component ruthenium ylidene com- was obtained with [BMIM][NTf ], having the highest
2
[9]
plexes 2, 4 and 5 immobilized as SILP systems in the solubility of ethene (k =873 MPa at 308C), the
H
ionic liquid (IL) 1-butyl-3-methylimidazolium tetra- most stable catalyst was based on [BMIM][BF ], ex-
4
hibiting the lowest solubility of ethene (k =
H
[
10]
2
065 MPa at 308C).
liquid cation, it was found that the ethylene glycol-
modified ionic liquid [Me(C H O) MIM][BF ] and
BMIM][BF ] gave the best results with respect to ac-
Upon variation of the ionic
AHCTUNGTRENNUNG
2
4
3
4
[
4
tivity and stability (see the Supporting Information
for all details of the IL variation experiments).
To further test the hypothesis that the propene
metathesis product ethene is indeed responsible for
the observed catalyst deactivation, a flow of ethene
was passed through the 5-SILP {ionic liquid: [Me-
AHCTUNGTRNNEU(G C H O) MIM][BF ]} catalyst bed prior to starting
2 4 3 4
the propene metathesis reaction. Different times of
À1
ethene purge (10 mLmin for 0, 1, 3, 5 and 15 h)
were realized in independent experiments, corre-
sponding to an ethene exposure of the catalyst (3.69ꢄ
À2
10
mmol) of 0, 23, 69, 115 and 350 mmol of ethene.
As shown in Figure 2, after 1 h (23 mmol) ethene ex-
posure, the initial catalyst activity remained nearly
the same as for no ethene purge. For longer ethene
Figure 1. Conversion versus time plots for different SILP
metathesis catalysts in propene cross-metathesis at 408C,
1
bar, residence time=32 s, 0.4 g SILP, Ru loading 0.2 wt%, exposure, however, a direct correlation between
calcined silica gel 100, [BMIM][BF ], IL loading 5 vol%. ethene exposure time and initial activity was ob-
4
2702
asc.wiley-vch.de
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2011, 353, 2701 – 2707

Ethene-Induced Temporary Inhibition of Grubbs Metathesis Catalysts
Figure 2. Ethene exposure of 5-SILP prior to propene meta-
thesis. 408C, 1 bar, residence time=8 s, 0.4 g SILP, Ru load-
ing 1 wt%, silica gel 60, [Me
ing 5 vol%.
ACHTUNGRTNNGE(U C H O) MIM][BF ], IL load-
2 4 3 4
served, giving support to the assumption of ethene-in-
duced catalyst deactivation.
Although the initial activity of the 5-SILP catalyst
decreased upon exposure to ethene, it is a very inter-
esting observation that the catalyst regained part of
its activity when the substrate was switched to pro-
pene (as monitored by the 2-butene formed in the re-
action; note that the situation regarding unproductive
ethene+propene metathesis is independent on the
duration of ethylene exposure). In fact, a maximum in
activity was observed shortly after the substrate
Scheme 2. Catalytic cycle of the Ru-catalyzed metathesis re-
action displaying the calculated intermediates of Table 1;
species G, H, I are unproductive in propene cross-metathe-
sis.
change. Remarkably, the 5-SILP catalysts with higher favored dissociative pathway and a hypothetical asso-
ethene exposure required longer times to reach its ciative pathway (see the Supporting Information for
maximum activity and a direct correlation between details on the calculated species). In the latter case,
the amount of ethene and the lag time to tmax was ob- the O-donor ligand in 4 or 5 remains coordinated to
served. Based on this finding, it appeared highly un- the Ru center, while in the former case it is not. We
likely to us that the observed maximum stems from take both pathways into account here because it is un-
reactor instationarities. Instead, we assumed a kind of clear to what extent dissociation occurs in an ionic
catalyst re-activation process going along with the liquid environment. Test calculations have shown that
substrate change.
the gas-phase dissociation energies of the O-coordi-
To provide a possible explanation for the ethene in- nated moiety in 4 or 5 can be substantial (50–
À1
fluence on the ruthenium alkylidene catalysts, we per- 100 kJmol ) at the present dispersion-corrected DFT
formed DFT calculations on the possible reaction in- level. In agreement with previous investigations, we
termediates shown in Scheme 2. For the intermediates consider the metallacycle in a cis position with respect
C and E all the different regio- and stereoisomers to the NHC ligand in the associative path, while it
were calculated which originate from catalyst 4 (see was found to be more favorable in the trans position
[
11c,e,l,n,s]
the Supporting Information for details). The mecha- in the dissociative pathway.
Among the possi-
nism of ligand effects in olefin metathesis using ble isomers, we here discuss only the lowest-energy
Grubbs-type catalysts has been studied theoretically metallacyclobutane intermediates as predicted by the
[
11]
in great detail before (including certain aspects of present DFT method (details on the other intermedi-
[
12]
propene cross-metathesis ) and a dissociative path- ates will be published elsewhere). In the associative
way has been found to be favorable compared to an pathway, the highest formation energy (lowest rela-
associative pathway in the gas-phase. In this study, we tive energy) of all intermediates was found for the
considered the lowest energy metallacyclobutane in- RuC H metallacycle G (see Table 1, Scheme 2, and
4
8
termediates as possible resting states along both the the Supporting Information) with one methyl sub-
Adv. Synth. Catal. 2011, 353, 2701 – 2707
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2703

Judith Scholz et al.
FULL PAPERS
À1
Table 1. Formation energies in kJmol of the lowest energy isomers of the various metallacyclobutane intermediates shown
in Scheme 2 according to DFT calculations (see the Supporting Information for further information).
Complex
Associative pathway
Dissociative pathway
[
a]
[b]
[d]
[a]
[b]
[d]
C [Ru]C H Me
À38 /À73
À109 /À131
3
4
2
[
[c]
[c]
E [Ru]C H Me
À31 /À65
À87
À114 /À135
3
5
[
e]
F [Ru]C H6
À130
À124
À119
À115
3
e]
G [Ru]C H Me
À90
3
5
[e]
H [Ru]C H Me
À16
3
4
2
e]
[
I [Ru]C H Me
3
À18
3
3
[
[
[
[
[
a]
b]
c]
d]
e]
Formation from B and propene.
Formation from D and trans-2-butene.
Formation from B and ethene.
Formation from D and propene.
Metallacycles F–I are unproductive in propene metathesis, cf. Scheme 2.
stituent in the 2-position. This intermediate would un- should re-activate the metathesis system as more Ru
dergo an unproductive cycle to yield propene and catalyst species would become part of the productive
complex B. The second lowest energy was found for cycle again.
the non-substituted RuC H metallacycle F which is
To check this hypothesis we performed the follow-
formed by addition of ethene to complex D. In the ing experiment. Our fixed-bed reactor was filled with
dissociative pathway, metallacycles E, C, and F are 2 g of 5-SILP (IL: [Me(C H O) MIM][BF ], support:
3
6
AHCTUNGTRENNUNG
2
4
3
4
lowest in energy, i.e., a significant amount of catalyst silanized silica gel 60) and contacted at 408C first
can be trapped in the low-energy species F also here, with a pre-mixed, highly diluted gas mixture of 1-
although the energetic differences in metallacycles butene and 2-butenes in butanes (2.5% 1-butene,
are less pronounced (Table 1). The other possible in- 39.0% 2-butenes, 58.5% butanes; residence time 8 s).
termediates have lower formation energies as sum- Under the applied conditions, 5-SILP promoted
marized in Table 1.
almost complete 1-butene conversion (X1-butene =96%,
The difference in energy of the metallacycle inter- note that the equilibrium of 1-butene/2-butene cross-
mediates can help to understand our experimental ob- metathesis under the applied conditions is 96%) in
servations. The initial activity of the 5-SILP catalyst the 1-butene/2-butene cross-metathesis leading to pro-
system is high and propene is converted with the pene and 2-pentene formation. Due to the high dilu-
equilibrium conversion of 34% at 408C. The formed tion of the feed consecutive metathesis of the primary
ethene will then shift the equilibrium between D and product propene was insignificant during the resi-
F towards the low energy complex F.
dence time of 8 s as demonstrated by our GC product
This complex is not part of the productive metathe- analysis (ethene concentrations in the product gas
sis cycle and overall lowers the effective concentra- mixture were always below 0.2%). The very high con-
tion of productive ruthenium complexes.
version of 1-butene found in this experiment is very
The observation that the [BMIM][NTf ] containing remarkable given its low concentration in the feed
2
5
-SILP catalysts deactivated faster than the (low feedstock concentration should slow down the
[
BMIM][BF ] based system can now be explained, reaction rate) and the fact that unproductive 2-butene
4
since the two different ionic liquids act as ethene res- self-metathesis is proceeding in parallel to the produc-
ervoirs of different efficiency, thus creating different tive 1-butene/2-butene reaction. Most important in
ethene concentrations at the catalyst center.
the context of our catalyst stability study is, however,
Assuming that the presence of ethene primarily the finding, that the immobilized Grubbs catalyst
shifts the equilibrium between complexes D and F, showed no deactivation over 20 h time-on-stream in
the effective concentration of the active catalyst is this 1-butene/2-butene cross-metathesis reaction
merely lowered without deactivating the ruthenium (Figure 3).
species irreversibly. Deactivation of ruthenium based
After 20 h time-on-stream we added ethene to the
metathesis catalysts by ethene had been proposed by feedstock entering the reactor so that the molar ratio
[
13,14]
several authors in the past.
In contrast, we con- of ethene and 2-butene in the feedstock was equal to
cluded from our experiments and calculations that re- 1. As expected from our previous results (e.g.,
placing propene by another alkene feed that would Figure 1) an immediate deactivation of the immobi-
hardly produce any ethene as a product of metathesis lized Grubbs catalyst took place when ethene entered
should be able to shift the equilibrium back to the the reactor. After 140 h time-on-stream the 2-butene
active species B–E. As a consequence such a shift conversion had decreased to a value below 10% start-
2704
asc.wiley-vch.de
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2011, 353, 2701 – 2707

Ethene-Induced Temporary Inhibition of Grubbs Metathesis Catalysts
versible catalyst modification is still under investiga-
tion in our groups.
Thus, our work provides strong evidence for the
conclusion that active Grubbs-type metathesis cata-
lysts suffer from temporary inhibition caused by the
presence of large amounts of ethene. Our theoretical
work suggests that the observed ethene-induced in-
hibition effect is due to a reversible shift of Ru com-
plexes from the productive catalytic cycle towards un-
productive, dormant Ru species. This shift is reversi-
ble, equilibrium driven and thus dependent on the
actual ethene concentration at the catalyst. By replac-
ing ethene in the feedstock and by largely avoiding
the catalytic formation of ethene, the described
Grubbs-type SILP-metathesis catalyst showed con-
stant catalytic activity and selectivity for more than
2
0 days of continuous gas-phase operation reacting a
Figure 3. Conversion vs. time plot for the investigation of
substrate influence on catalyst stability of 5-SILP at 408C,
small concentration of 1-butene (2.5%) in a very
short residence time (8 s) to equilibrium concentra-
tions of 1-butene, propene and 2-pentenes. Apart
from these specific conclusions on Ru-catalyzed cross-
1
bar, residence time=8 s, 2 g SILP, Ru loading 0.1 wt%,
silanized silica gel 60, [Me(C H O) MIM][BF ], IL loading
vol%, full points: 1-butene metathesis in a diluted C mix-
AHCTUNGTRENNUNG
2
4
3
4
5
4
ture; open points: ethene/2-butene (molar 1:1) metathesis; metathesis, our work clearly demonstrates the poten-
for TOF information see Supporting Information.
tial of continuous gas-phase reactions using SILP cat-
alyst systems for systematic mechanistic studies of
complex, transition metal complex-catalyzed reac-
ing from 48% at the moment of the feedstock switch tions.
note that the thermodynamic equilibrium for the for-
(
mation of propene from ethene and 2-butene is 66%
under the applied conditions). Then we changed the Experimental Section
feedstock entering the reactor back to the same dilut-
ed C feed that was used for the first 20 h time-on- Chemicals
4
stream. With this switch in feedstock the system re-
All syntheses were carried out using standard Schlenk tech-
turned quickly to the same high 1-butene conversion
as observed during the first 20 h indicating the cata-
lyst deactivation observed with ethene feed to be per-
niques under argon (99.999%, Linde AG). The applied cata-
lysts 2, 4 and 5 were purchased from STREM Chemicals.
Silica gel 100 (Merck KgaA) was partly dehydroxylated at a
fectly reversible and feedstock-dependent. To be ab- temperature of 3708C in air for 12 h followed by storage
solutely sure about this finding we repeated the feed- under vacuum prior to use. Silanized silica gel 60 and di-
stock switch back to ethene and diluted C , and then chloromethane were purchased from Merck KgaA and
4
stored under argon prior to use. The ionic liquids
to the diluted C another two times with different re-
4
[BMIM][BF4] (99.9%), [BMIM][PF6] (99.8%) and
action periods leading to a total operation time of the
immobilized Grubbs catalyst of more than 500 h time-
on-stream. During this prolonged operation time no
sign of catalyst deactivation was observed with the di-
[
BMIM][CF SO ] (99,7%) were purchased from Merck
3
3
[15]
KGaA.
The
syntheses
BMIM][NTf ], [Me ACHTUNGRTENUN(NG C H O) MIM][BF4] were carried
2
of
[BMIM][SbF ],
6
[17]
[16]
[
2
4
3
out according to the literature.
luted C feed while clearly each period using ethene
4
and diluted C feed showed repeatedly an apparent
catalyst deactivation. Although the catalyst obviously
4
SILP Catalyst Preparation
maintained its full initial conversion for the diluted C4 The applied SILP materials were prepared by impregnation
feed over 20 days of operation with full recovery of of the silica support with a solution of dichloromethane and
the respective ionic liquid containing the relevant metathesis
the activity after ethene-rich operation (cf. Figure S4
catalyst under an argon atmosphere using standard Schlenk
in the Supporting Information), our results also indi-
techniques, followed by subsequent removal of dichlorome-
cate a slow irreversible transformation of the catalyst
thane under vacuum to obtain a free flowing SILP catalyst.
over time that exclusively affects the metathesis of
ethene and 2-butene. While the system provides up to
Continuous-Flow Gas-Phase Metathesis Experiments
48% 2-butene conversion at the beginning of the first
ethene-rich operation phase, the system has signifi-
The continuous-flow gas-phase metathesis experiments were
cantly lower initial conversions in ethene/2-butene performed in a screening-rig consisting of ten parallel, inde-
metathesis for the later slots. The nature of this irre- pendently operating reactor lines. The solid SILP catalysts
Adv. Synth. Catal. 2011, 353, 2701 – 2707
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2705

Judith Scholz et al.
FULL PAPERS
were used as a fixed-bed in a stainless-steel tubular reactor References
at atmospheric pressure and a temperature of 408C. Before
filling the catalyst into the reactor, the rig was flushed with
nitrogen to generate an inert atmosphere. As soon as the re-
action temperature was reached the respective gas was fed
into the rig by means of mass flow controllers (Brooks In-
struments, Brockhorst) for starting the reaction. The reactor
lines were connected to the FID-gas chromatograph (Agi-
lent 6890 N) in a switchable manner via a multi-port valve.
Gaps in the data acquisition (see Figure 1 and Figure 2 as
well as Figures S2 and S3 in the Supporting Information)
originate from analysis of parallel operated reactors during
the continuous experiment.
[1] a) G. C. Vougioukalakis, R. H. Grubbs, Chem. Rev.
2010, 110, 1746–1787; b) R. H. Grubbs, Angew. Chem.
2006, 118, 3845–3850; Angew. Chem. Int. Ed. 2006, 45,
3760–3765.
[2] a) S. T. Nguyen, L. K. Johnson, R. H. Grubbs, J. Am.
Chem. Soc. 1992, 114, 3974–3975; b) S. T. Nguyen,
R. H. Grubbs, J. W. Ziller, J. Am. Chem. Soc. 1993, 115,
9858–9859; c) P. Schwab, M. B. France, J. W. Ziller,
R. H. Grubbs, Angew. Chem. 1995, 107, 2179–2181;
Angew. Chem. Int. Ed. Engl. 1995, 34, 2039–2041; d) P.
Schwab, R. H. Grubbs, J. W. Ziller, J. Am. Chem. Soc.
1
996, 118, 100–110; e) T. R. Belderrain, R. H. Grubbs,
Computational Details
Organometallics 1997, 16, 4001–4003.
[
3] a) J. S. Kingsbury, J. P. A. Harrity, P. J. Bonitatebus Jr,
A. H. Hoveyda, J. Am. Chem. Soc. 1999, 121, 791–799;
b) S. B. Garber, J. S. Kingsbury, B. L. Gray, A. H. Hov-
eyda, J. Am. Chem. Soc. 2000, 122, 8168–8179.
All DFT calculations were performed using the Turbomole
[18]
program package. The exchange-correlation functional ac-
[19]
[20]
cording to Becke
and Perdew
augmented with
[21]
Grimmeꢃs dispersion correction
(BP-D) has been used
[
4] a) C. P. Mehnert, R. A. Cook, N. C. Dispenziere, M.
Afeworki, J. Am. Chem. Soc. 2002, 124, 12932–12933;
b) A. Riisager, K. M. Eriksen, P. Wasserscheid, R. Fehr-
mann, Catal. Lett. 2003, 90, 149–153; c) A. Riisager, R.
Fehrmann, S. Flicker, R. van Hal, M. Haumann, P.
Wasserscheid, Angew. Chem. 2005, 117, 826–830;
Angew. Chem. Int. Ed. 2005, 44, 815–819; d) A. Riisag-
er, R. Fehrmann, M. Haumann, B. S. K. Gorle, P. Was-
serscheid, Ind. Eng. Chem. Res. 2005, 44, 9853–9859;
e) S. Werner, M. Haumann, P. Wasserscheid, Annu.
Rev. Chem. Biomol. Engin. 2010, 1, 203–230.
[22]
throughout in combination with the SV(P) basis set. The
multipole-accelerated resolution of the identity (MARIJ)
[23]
technique
has been employed. In addition, test calcula-
tions using the same functional (BP-D) but the TZVP basis
[
22]
set
as well as the B3LYP-D functional and the SV(P)
basis set have been performed, which all resulted in the
same qualitative trends as the BP-D/SV(P) method in the
present case. Geometry optimizations were performed with-
out symmetry constraints, and the identity of minima was
confirmed by frequency calculations.
The theoretical calculations in this study focus on the for-
mation energies of the metallacyclobutane intermediates as
possible resting species in propene metathesis to yield cis/
trans-2-butene and ethene. No activation barriers have been
considered. For both the dissociative and the associative
pathways (cf. Supporting Information, Scheme S1), all possi-
ble regio- and stereoisomers for the metallacyclobutane spe-
cies with 0–3 methyl substituents that can potentially occur
in the metathesis reaction have been calculated. In the
paper, we discuss only those metallacycles according to
Scheme 2 which are lowest in energy as compared to their
isomers (see Scheme 2 and the Supporting Information for
further information). Table 1 lists the formation energies of
these lowest-energy isomers from B, D, ethene, propene,
and cis/trans-2-butene, see Scheme S4 in the Supporting In-
formation for an example). Note that species E in principle
can be formed by reaction of D with propene or (alterna-
tively) by reaction of B with ethene. Likewise, species C can
either be formed by B with 2-butene (cis or trans) or by re-
action of B with propene.
[
5] a) H. Hagiwara, N. Okunaka, T. Hoshi, T. Suzuki, Syn-
lett 2008, 1813–1816; b) H. Hagiwara, T. Nakamura, N.
Okunaka, T. Hoshi, T. Suzuki, Helv. Chim. Acta 2010,
9
3, 175–182.
[
6] a) M. J. Muldoon, C. M. Gordon, I. R. Dunkin, J.
Chem. Soc. Perkin Trans. 1 2001, 2, 433–435; b) P. Was-
serscheid, C. M. Gordon, C. Hilgers, M. J. Muldoon,
I. R. Dunkin, Chem. Commun. 2001, 1186–1187.
[
7] J. G. Huddleston, A. E. Visser, W. M. Reichert, H. D.
Willauer, G. A. Broker, R. D. Rogers, Green Chem.
2
001, 3, 156–164.
[
8] a) J. S. Wilkes, M. J. Zaworotko, J. Chem Soc. Chem.
Commun. 1992, 965–967; b) J. Fuller, R. T. Carlin,
H. C. Delong, D. Haworth, J. Chem. Soc. Chem.
Commun. 1994, 299–300.
[
9] J. Zhang, Q. Zhang, B. Qiao, Y. Deng, J. Chem. Eng.
Data 2007, 52, 2277–2283.
[
[
10] D. Camper, C. Becker, C. Koval, R. Noble, Ind. Eng.
Chem. Res. 2005, 44, 1928–1933.
11] a) O. M. Aagaard, R. J. Meier, F. Buda, J. Am. Chem.
Soc. 1998, 120, 7174–7182; b) T. Weskamp, F. J. Kohl,
W. Hieringer, D. Gleich, W. A. Herrmann, Angew.
Chem. 1999, 111, 2573–2576; Angew. Chem. Int. Ed.
Acknowledgements
1
999, 38, 2416–2419; c) C. Adlhart, P. Chen, Angew.
Chem. 2002, 114, 4668–4671; Angew. Chem. Int. Ed.
The authors acknowledge financial support by the Deutsche
Forschungsgemeinschaft (DFG) within the Excellence Cluster
2
2
002, 41, 4484–4487; d) L. Cavallo, J. Am. Chem. Soc.
002, 124, 8965–8973; e) S. F. Vyboishchikov, M. Bꢀhl,
“Engineering of Advanced Materials” in the framework of
W. Thiel, Chem. Eur. J. 2002, 8, 3962–3975; f) F. Ber-
nardi, A. Bottoni, G. P. Miscione, Organometallics
the excellence initiative.
2
003, 22, 940–947; g) S. Fomine, S. M. Vargas, M. A.
Tlenkopatchev, Organometallics 2003, 22, 93–99; h) C.
Adlhart, P. Chen, J. Am. Chem. Soc. 2004, 126, 3496–
2706
asc.wiley-vch.de
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2011, 353, 2701 – 2707

Ethene-Induced Temporary Inhibition of Grubbs Metathesis Catalysts
3
510; i) C. Costabile, L. Cavallo, J. Am. Chem. Soc.
004, 126, 9592–9600; j) C. H. Suresh, N. Koga, Orga-
[13] a) S. H. Hong, A. G. Wenzel, T. T. Salguero, M. W. Day,
R. H. Grubbs, J. Am. Chem. Soc. 2007, 129, 7961–7968;
b) Z. Lysenko, B. R. Maughon, T. Mokhtar-Zadeh,
M. L. Tulchinsky, J. Organomet. Chem. 2006, 691, 5197–
5203.
2
nometallics 2004, 23, 76–80; k) W. J. van Rensburg, P. J.
Steynberg, W. H. Meyer, M. M. Kirk, G. S. Forman, J.
Am. Chem. Soc. 2004, 126, 14332–14333; l) D. Benitez,
W. A. Goddard III, J. Am. Chem. Soc. 2005, 127,
[14] W. J. van Rensburg, P. J. Steynberg, W. H. Meyer,
M. M. Kirk, G. S. Forman, J. Am. Chem. Soc. 2004, 126,
1
6
2218–12219; m) B. F. Straub, Angew. Chem. 2005, 117,
129–6132; Angew. Chem. Int. Ed. 2005, 44, 5974–5978;
1
4332–14333.
15] Y. Chauvin, H. Olivier-Bourbigou, CHEMTECH, 1995,
5, 26–30.
16] E. ꢅchsner, K. Schneiders, K. Junge, M. Beller, P. Was-
serscheid, Appl. Catal. A 2009, 364. 8–14.
17] C. Kolbeck, M. Killian, F. Maier, N. Paape, P. Wasser-
n) A. Correa, L. Cavallo, J. Am. Chem. Soc. 2006, 128,
3352–13353; o) M. Piacenza, I. Hyla-Kryspin, S.
[
[
[
[
1
2
Grimme, J. Comput. Chem. 2007, 28, 2275–2285;
p) P. E. Romero, W. E. Piers, J. Am. Chem. Soc. 2007,
1
3
1
Goddard III, Chem. Commun. 2008, 6194–6196; t) S.
Torker, D. Merki, P. Chen, J. Am. Chem. Soc. 2008,
1
Goddard III, Organometallics 2009, 28, 2643–2645;
v) I. C. Stewart, D. Benitez, D. J. OꢃLeary, E. Tkatch-
ouk, M. W. Day, W. A. Goddard III, R. H. Grubbs, J.
Am. Chem. Soc. 2009, 131, 1931–1938; w) A. Poater, F.
Ragone, A. Correa, A. Szadkowska, M. Barbasiewicz,
K. Grela, L. Cavallo, Chem. Eur. J. 2010, 16, 14354–
29, 1698–1704; q) B. F. Straub, Adv. Synth. Catal. 2007,
49, 204–214; r) C. E. Webster, J. Am. Chem. Soc. 2007,
29, 7490–7491; s) D. Benitez, E. Tkatchouk, W. A.
AHCTUNGTRENNUNGs cheid, H.-P. Steinrꢀck, Langmuir 2008, 24, 9500–9507.
18] a) R. Ahlrichs, F. Furche, C. Hꢁttig, W. Klopper, M.
Sierka, F. Weigend, Turbomole program package, ver-
sion 5.10, University of Karlsruhe, Karlsruhe, since
30, 4808–4814; u) D. Benitez, E. Tkatchouk, W. A.
1
988, http://www.turbomole.de; b) R. Ahlrichs, M. Bꢁr,
M. Hꢁser, H. Horn, C. Kçlmel, Chem. Phys. Lett. 1989,
62, 165–169.
1
[19] A. D. Becke, Phys. Rev. A 1988, 38, 3098–3100.
[20] J. P. Perdew, Phys. Rev. B 1986, 33, 8822–8824.
[21] S. Grimme, J. Comput. Chem. 2006, 27, 1787–1799.
[22] F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys.
1
4364; x) F. Ragone, A. Poater, L. Cavallo, J. Am.
2
005, 7, 3297–3305.
23] M. Sierka, A. Hogekamp, R. Ahlrichs, J. Chem. Phys.
003, 118, 9136–9148.
Chem. Soc. 2010, 132, 4249–4258.
[
[
12] N. Bahri-Laleh, R. Credendino, L. Cavallo, Beilstein J.
2
Org. Chem. 2011, 7, 40–45.
Adv. Synth. Catal. 2011, 353, 2701 – 2707
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2707