Lewis acid assisted ruthenium-catalyzed metathesis reactions

Article abstract of DOI:10.1002/cctc.201300675

The combination of a ruthenium-arene complex, a noncoordinating salt, and a Lewis acid facilitates access to a highly active and selective in situ metathesis catalyst. The catalyst is formed from an inexpensive ruthenium precursor and olefins are used as the substrates. RCM=Ring-closing metathesis, acac=acetylacetonate. Copyright

Full text of DOI:10.1002/cctc.201300675

CHEMCATCHEM
COMMUNICATIONS
DOI: 10.1002/cctc.201300675
Lewis Acid Assisted Ruthenium-Catalyzed Metathesis
Reactions
Christa Lꢀbbe,[a] Andreas Dumrath,[a] Helfried Neumann,[a] Matthias Beller,*[a] and
Renat Kadyrov*[b]
The combination of a ruthenium–arene complex, a noncoordi-
nating salt, and a Lewis acid facilitates access to a highly active
and selective in situ metathesis catalyst. The catalyst is formed
from an inexpensive ruthenium precursor and olefins are used
as the substrates. RCM=Ring-closing metathesis, acac=acetyl-
acetonate.
[(L)(p-cymene)RuCl2]. Common concepts are photochemical ac-
tivation[6a–c] and activation by carbene precursors such as tri-
methyldiazomethane[6d,e] and alkyne.[6f–i] Interestingly, Delaude
and co-workers reported that the homobimetallic ruthenium
complex [(p-cymene)Ru(m-Cl)3RuCl(h2-C2H4)(IMes)] [IMes=1,3-
bis(2,4,6-trimethylphenyl)imidazol-2-ylidene] does not require
any aforementioned activators to initiate the metathesis reac-
tions.[8] Nonetheless, the benchmark reactions of diethyl diallyl-
malonate and N,N-diallyltosylamide again resulted in mixtures
of cycloisomerization and RCM products.
Since its discovery in the mid 1960s, metathesis has become
a valuable tool for organic syntheses and polymer chemistry,
particularly in industry. Therein, ill-defined catalysts[1] dominat-
ed for the first three decades, before well-defined catalysts es-
tablished metathesis as an indispensable tool for organic syn-
thesis.[2] In this regard, ruthenium-based complexes have nota-
bly experienced tremendous growth in interest.[3] These latter
complexes emerged to form efficient and highly tolerant cata-
lysts for all kinds of metathesis reactions. Hence, numerous ele-
gant synthetic applications have been realized.[4] However, de-
spite the clear advantages of well-defined activators, olefin
metathesis is still limited by irreversible catalyst deactivation,
which necessitates, in most cases, high catalyst loadings.[5]
Therefore, approaches that can be used to generate the active
catalyst from readily available and stable precursors significant-
ly extend the impact of olefin metathesis for industrial ap-
plications, which is demonstrated by a number of reported
examples.[6] In this respect, compounds of the type
[(NHC)(p-cymene)RuCl2] (NHC=N-heterocyclic carbene) consti-
tute easily available ruthenium complexes; complexes of this
type have been known since the pioneering report of Nolan
et al. in 1999.[7] Unfortunately, the use of such complexes, apart
from the desired ring-closing metathesis (RCM) reaction, also
results in unwanted cycloisomerizations in the benchmark re-
action of diethyl diallylmalonate (see below).[6i] Therefore, sev-
eral efforts have been undertaken to perform selective meta-
thesis transformations in the presence of arene complexes
such as
Herein, we report that the addition of noncoordinating salts
and Lewis acids enhances the metathesis activity of ruthenium
arene complexes of the type [(NHC)(p-cymene)RuCl2] to effec-
tively suppress such undesired cycloisomerization reactions ac-
companied with a very high productivity.
The RCM reaction of diethyl diallylmalonate (Scheme 1) was
chosen as a benchmark reaction to demonstrate the influence
of different additives by using complex 1 as the catalyst pre-
cursor. Preliminary studies showed that full conversion of the
Scheme 1. Ring-closing metathesis of diethyl diallylmalonate by using cata-
lyst 1 and additives.
substrate could be achieved in toluene at 808C for catalyst 5.[7]
However, in the presence of 1 (1 mol%), only 33% of desired
RCM product 3 and 32% of undesired cycloisomer 4 were ob-
tained within 1 h (Table 1, entry 1).
[a] C. Lꢀbbe, Dr. A. Dumrath, Dr. H. Neumann, Prof. Dr. M. Beller
Leibniz-Institut fꢀr Katalyse an der Universitꢁt Rostock e.V.
Albert-Einstein-Strasse 29a, 18059 Rostock (Germany)
Homepage: www.catalysis.de
To our delight, the addition of NaPF6 (5 mol%) raised the
productivity to 76%, and the cycloisomerization was complete-
ly suppressed. Any further increase or decrease in the noncoor-
dinating anion did not result in any increase in the amount of
product formed (Table 1, entries 2–4). Notably, the addition of
AgPF6 caused complete deactivation of the catalyst.
E-mail: matthias.beller@catalysis.de
[b] Dr. R. Kadyrov
Evonik Industries AG
Rodenbacher Chaussee 4, 63457 Hanau-Wolfgang (Germany)
Homepage: www.evonik.com
Interestingly, the addition of Lewis acids such as Fe(acac)3
(acac=acetylacetonate) and AlCl3 also successfully inhibited
the formation of cycloisomerization product 4 (Table 1, en-
tries 7 and 9). Consequently, we applied a combination of
E-mail: renat.kadyrov@evonik.com
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201300675.
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 105 – 108 105

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
Table 1. Effect of additives on the activity of 1 in the RCM of diethyl
diallylmalonate.
Entry[a]
Additive [mol%]
Conv. 2 [%][b]
Yield 3 [%][b]
1[e]
2
–
65
67
33
21
NaPF6 (1)
3
NaPF6 (5)
76
76
4
NaPF6 (10)
72
63
5
6
7
PdCl2(PPh3)2 (5)
PdCl2(PPh3)2 (5)/NaPF6 (5)
AlCl3 (5)
46
64
34
26
61
34
8
9
AlCl3 (5)/NaPF6 (5)
Fe(acac)3 (5)
94
48
94
46
10
11
12
13
Fe(acac)3 (5)/NaPF6 (5)
Fe(acac)3 (5)/NaPF6 (5)
Fe(acac)3 (5)/NaPF6 (5)
Fe(OAc)2 (5)/NaPF6 (5)
97
94
67
>99
97
87[c]
62[d]
97
[a] Reaction conditions: diethyl diallylmalonate (4 mmol), (p-cymene)Ru-
(Me2IMes)Cl2 (1, 1 mol%), additive, toluene (100 mL), Ar bubbling, 808C,
1 h. [b] Determined by GC with hexadecane as an internal standard.
[c] Technical-grade toluene (100 mL); reaction performed in air. [d] De-
gassed water (50 mL) was added. [e] See Figure S1.
Figure 1. Effect of the NHC on productivity in the RCM of diethyl diallylmalo-
nate.
Subsequently, we investigated the application of our new
catalyst system consisting of 1 with NaPF6 and Fe(acac)3 as ad-
ditives for different metathesis reactions (Table 2). The RCM of
1-allyl-2-(allyloxy)benzene (10) was performed with 1 and
NaPF6 and a Lewis acid, which resulted in excellent perfor-
mance; the RCM product was delivered in up to 97% yield
with exclusive selectivities (Table 1, entries 6, 8, 10, 13).
Moreover, the oxidation state of iron apparently has no in-
fluence on the activation; thus, Fe(acac)3 and Fe(OAc)2 behave
in the same manner (97%; Table 1, entries 10, 13). Inspired by
these excellent results, other M(acac)x salts (M=FeII, CoI,II, NiII,
CrIII, CuII, MnII,III) were employed together with NaPF6. However,
these combinations revealed a lower performance in RCM than
the use of only NaPF6.
Table 2. RCM and cross-metathesis reactions promoted by 1 and assisted
by NaPF6 and Fe(acac)3.
Entry[a] Substrate
t [h] Conv. [%][b] RCM Yield [%][b]
1
0.5
3
3
94
94
16
85
81
9
2[c]
3[d]
10
Given that all of the starting materials are air stable, we
were interested in keeping an industrial focus. Therefore, we
executed the RCM reaction in air and successfully obtained 3
in 87% yield (Table 1, entry 11), for which small amounts of
water were tolerated (Table 1, entry 12).
4
11[g]
12[h]
13[i]
3
3
3
14
50
85
12
24
41
5
6[e]
In further experiments, the conditions of the new catalyst
system were optimized for temperature and solvent as a result
of the properties of the additive; 808C in toluene was best,
whereas dichloromethane, dichloroethane, 1,4-dioxane, and
isobutyl alcohol resulted in full deactivation of the catalyst.
Noteworthy, the preparation of the ruthenium complex
in situ from [(p-cymene)RuCl2]2, Me2IMesHCl, and NaOtBu re-
sulted in comparable productivities (52% of 3). In doing so,
the experiment was also successfully conducted with the iso-
lated free carbene, Me2IMes, which confirmed no negative in-
fluence of the base on the catalytic activity (see the Support-
ing Information, Table S3).
7
8
14
15
3
3
90
95
81
94
9[f]
16
17
3
3
23
54
13
21
10[f]
[a] Reaction conditions: substrate (4 mmol), (p-cymene)Ru(Me2IMes)Cl2 (1,
1 mol% related to diene), NaPF6 (5 mol%), Fe(acac)3 (5 mol%), toluene
(100 mL), Ar bubbling, 1 h. [b] Determined by GC with hexadecane as an
internal standard. [c] Compound 1 (0.1 mol%), NaPF6 (0.5 mol%), and Fe-
(acac)3. [d] Compound 1 (0.01 mol%), NaPF6 (0.5 mol%), and Fe(acac)3,
110 8C.
[e] 1,2-Dibromocyclohexane
(1 mol%).[12]
[f] Compound
1 (10 mol%), NaPF6 (50 mol%), and Fe(acac)3. [g] Ts=tosyl. [h] E=CO2Et.
[i] Bn=benzyl.
Next, various ruthenium complexes of the type [(NHC)(p-
cymene)RuLCl2] were applied to the RCM reaction; mesitylene-
substituted imidazol-2-ylidenes 1 and 5 offered significantly
better results than 1,3-bis(2,6-diisopropylphenyl)-substituted
imidazol-2-ylidenes 6 and 7 (Figure 1). Interestingly, if the 4,5-
positions of the imidazol-2-ylidenes were substituted by
methyl groups, formation of the product dramatically in-
creased to 97%; thus, complex 1 served as the best ring-clos-
ing metathesis precursor.
0.1 mol% of 1, and excellent conversions and yields were ob-
tained (Table 2, entries 1 and 2). Amide 11 and sterically hin-
dered olefin 12 were converted to lesser degrees and provided
the desired product in only 12 and 24% yield, respectively,
(Table 2, entries 4 and 5). In the course of this, the formation
of five-membered rings 14 and 15 were more efficient than
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 105 – 108 106

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
the formation of six-membered ring 13 (Table 2, entries 6, 7,
and 8).
Recent reports showed that complete conversion of diethyl
diallylmalonate could be achieved with NHC-containing Ru–al-
kylidenes at catalyst loadings as low as 25–50 ppm.[5,14] Thus,
to effectively perform the RCM reaction, only trace amounts of
alkylidenes have to be generated, which are below the detec-
tion limit.
Moreover, in addition to ring-closing metathesis, we were
pleased to successfully apply the catalytic system to the self-
metathesis of styrene (16) and 6-bromohexene (17). However,
the yields were low owing to a missing enthalpic driving force
(ring strain), which is often observed for self- and cross-meta-
thesis reactions.[9]
In summary, the metathesis activity of catalysts of the type
[(p-cymene)Ru(NHC)Cl2] was significantly improved by the ad-
dition of NaPF6 and a Lewis acid, which effectively eliminate
undesired side reactions, specifically cycloisomerizations. To
the best of our knowledge, this is the first application of a new
type of metathesis catalyst in which the initiator is generated
by means of the alkene that is used as a substrate and not by
external alkyne or diazomethane activators. Notably, the cata-
lyst can be generated in situ from inexpensive, air-stable, and
commercially available starting materials, and the activity is
highly dependent on the structure of the imidazolylidene
ligand. In addition, the novel catalytic system was successfully
employed in ring-closing metathesis and cross-metathesis
reactions.
With respect to the mechanism, we believe that both com-
peting reaction pathways are induced by the formation of
a metallacycle intermediate formed through oxidative cycliza-
tion, as shown in Scheme 2.[10] So, in pathway A, subsequent
Experimental Section
Ring-closing metathesis of 2 and 10–17 with additives
Representative procedure: The noncoordinating salt, NaPF6 (0.50–
0.05 mmol), and Lewis acid, for example, Fe(acac)3 (0.50–
0.05 mmol), were weighed directly into a 50 mL, three-necked
flask. A magnetic stirrer bar, two septa, and a cooling unit were at-
tached, and after careful evaporation and purging with argon, sub-
strate 2 or 10–17 (1.0 mmol), hexadecane (0.5 mmol), and dry and
degassed toluene (25 mL) were added. (p-cymene)Ru(Me2IMes)Cl2
(0.001–0.10 mmol) was dissolved in toluene (4.0 mL) in a 5 mL
Schlenk tube and was added by syringe to the warm (T=808C) re-
action solution, which was bubbled continuously with argon and
stirred at 808C for 3 h. The reaction was monitored by GC. Details
on screening experiments and the preparation of the substrates
and complexes are reported in the Supporting Information.
Scheme 2. Cycloisomerization versus Ru–alkylidene formation.
b-hydride elimination forms the catalytically active ruthenium
hydride species, which releases undesired cycloisomerization
product 4 after reductive elimination. However, for the case in
which the b position is blocked, a-hydride elimination appears
to be the next most favorable process, which represents reac-
Keywords: carbene ligands · homogenous catalysis · Lewis
acids · metathesis · ruthenium
tion pathway B, and this results in desired product
3
[1] K. J. Ivin, J. C. Mol, Olefin Metathesis and Metathesis Polymerization, Aca-
demic Press, San Diego, 1997.
[2] a) Handbook of Metathesis (Ed.: R. H. Grubbs), Wiley-VCH, Weinheim,
2003; b) Green Metathesis Chemistry (Eds.: A. Demonceau, I. Dragutan,
E. S. Finkelstein), NATO Science for Peace and Security Series-A: Chemis-
try and Biology, Springer, Dordrecht, 2010.
[3] R. H. Grubbs, S. Chang, Tetrahedron 1998, 54, 4413–4450.
[4] a) D. Tindall, J. H. Pawlow, K. B. Wagner, Top. Organomet. Chem. 1998, 1,
180; b) L. L. Kiessling, L. E. Strong, Top. Organomet. Chem. 1998, 1, 196;
c) Alkene Metathesis in Organic Synthesis (Ed.: A. Fꢀrstner), Springer,
Berlin, 1998; d) M. Schuster, S. Blechert, Angew. Chem. 1997, 109, 2124–
2144; Angew. Chem. Int. Ed. Engl. 1997, 36, 2036–2056; e) S. K. J. Arm-
strong, Chem. Soc. Perkin Trans. 1 1998, 371–388; f) T. M. Trnka, R. H.
Grubbs, Acc. Chem. Res. 2001, 34, 18–29; g) S. Blechert, Pure Appl.
Chem. 1999, 71, 1393–1399.
(Scheme 2). This a-hydrogen abstraction is known for other M
alkylidene complexes (M=W, Mo, Ta, Re)[11] as well. We assume
that the presence of a Lewis acid (e.g., AlCl3), a noncoordinat-
ing salt (e.g., NaPF6), and an organic halogen-containing com-
pound[12] enable the commonly disfavored a-hydride elimina-
tion, which proceeds through desired reaction pathway B.
Here, a carbocation, for instance, R+BrÀAlCl3À, is formed from
the organic halogen compound and the Lewis acid (AlCl3),
which initiates the generation of a metathesis-active rutheni-
um alkylidene complex. In agreement with this proposal, in
the formation of their rhenium alkylidene complex, Gladysz
et al. also observed that the carbocation allows a-hydride elim-
ination despite the presence of b hydrogen atoms.[13] Hence, it
is reasonable to suspect that a-hydride elimination is more
rapid than b-hydride elimination in the presence of a carbocat-
ion.
[5] R. Kadyrov, Chem. Eur. J. Chem. 2013, 19, 1002–1012.
[6] a) A. Hafner, A. Mꢀhlebach, P. A. van der Schaaf, Angew. Chem. 1997,
109, 2213–2216; Angew. Chem. Int. Ed. Engl. 1997, 36, 2121–2124; b) L.
Delaude, A. Demonceau, A. F. Noels, Chem. Commun. 2001, 986–987;
c) A. Fꢀrstner, L. Ackermann, Chem. Commun. 1999, 95–96; d) A. W.
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 105 – 108 107

CHEMCATCHEM
COMMUNICATIONS
www.chemcatchem.org
Stumpf, E. Saive, A. Demonceau, A. F. Noels, Chem. Commun. 1995,
1127–1128; e) A. Demonceau, A. W. Stumpf, E. Saive, A. F. Noels, Macro-
molecules 1997, 30, 3127–3136; f) Y. Miyaki, T. Onishi, S. Ogoshi, H. Kur-
osawa, J. Organomet. Chem. 2000, 616, 135–139; g) J. Louie, R. H.
Grubbs, Angew. Chem. 2001, 113, 253–255; Angew. Chem. Int. Ed. 2001,
40, 247–249; h) D. Sꢂmeril, C. Bruneau, P. H. Dixneuf, Helv. Chim. Acta
2001, 84, 3335–3341; i) D. Sꢂmeril, C. Bruneau, P. H. Dixneuf, Adv. Synth.
Catal. 2002, 344, 585–595.
c) Some tantalum dialkyl complexes show a tendency to form alkyli-
dene complexes by a-hydrogen abstraction instead of olefin complexes
by b-abstraction: R. Toreki, R. R. Schrock, W. M. Davis, J. Am. Chem. Soc.
1992, 114, 3367–3380.
[12] Recently, we demonstrated that impurities in the olefinic substrate
have an dramatic influence on the activity and selectivity of catalyst
1 and additives. We showed that 1,2-dibromocyclohexane, which is
contained in commercially available diethyl diallylmalonate, enabled
a reproducible and highly productive RCM reaction. We assume that
1,2-dibromocyclohexane has a crucial influence on the formation of the
active ruthenium alkylidene species. C. Lꢀbbe, A. Dumrath, H. Neu-
mann, M. Beller, R. Kadyrov ChemCatChem 2013, DOI: 10.1002/
cctc.201300779.
[7] L. Jafarpour, J. Huang, E. D. Stevens, S. P. Nolan, Organometallics 1999,
18, 3760–3763.
[8] X. Sauvage, Y. Borguet, A. F. Noels, L. Delaude, A. Demonceau, Adv.
Synth. Catal. 2007, 349, 255–265.
[9] S. J. Connon, S. Blechert, Angew. Chem. 2003, 115, 1944–1968; Angew.
Chem. Int. Ed. 2003, 42, 1900–1923.
À
[13] The reactions of Ph3C+PF6 with complexes (h-C5H5)Re(NO)(PPh3)-
[10] Y. Yamamoto, Y. Nakagai, N. Ohkoshi, K. Itoh, J. Am. Chem. Soc. 2001,
123, 6372–6380.
(CH2CH3) and (h-C5H5)Re(NO)(PPh3)(CH2CH2CH3) give an a-hydride ab-
straction product [(h-C5H5)Re(=CHR)(NO)(PPh3)]+PF6À, despite the pres-
ence of b-hydrogen atoms: W. E. Kiel, G.-Y. Lin, J. A. Gladysz, J. Am.
Chem. Soc. 1980, 102, 3299–3301.
[11] Regarding M alkylidene complexes (M=W, Mo, Ta, Re), a) The highly
active metathesis alkylidene complex W=CHtBu(OCH2tBu)2Cl2 is formed
by transformation of W(OAlHal3)(OCH2tBu)2(CH2tBu)2 in the presence of
aluminum halides in dichloromethane by release of neopentane: J.
Kress, M. Wesolek, J. A. Osborn, J. Chem. Soc. Chem. Commun. 1982,
514–516; b) An excess of triflic acid induces a-hydrogen abstraction on
Mo-bis(alkyl) complexes Mo(NAr)2(CH2R)2 to give alkylidene complex
Mo=CHR(NAr)(OTf)2: R. R. Schrock, J. S. Murdzek, G. C. Bazan, J. Robbins,
M. DiMare, M. O’Regan, J. Am. Chem. Soc. 1990, 112, 3875–3886;
[14] K. M. Kuhn, J.-B. Bourg, C. K. Chung, S. C. Virgil, R. H. Grubbs, J. Am.
Chem. Soc. 2009, 131, 5313–5320.
Received: August 13, 2013
Published online on December 9, 2013
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 105 – 108 108