Nickel-Catalyzed Cyclopropanation with NMe4OTf and nBuLi

Article abstract of DOI:10.1002/anie.201505482

Nickel was identified as a catalyst for the cyclopropanation of unactivated olefins by using in situ generated lithiomethyl trimethylammonium triflate as a methylene donor. A mechanistic hypothesis is proposed in which the generation of a reactive nickel carbene explains several interesting observations. Additionally, our findings shed light on a report by Franzen and Wittig published in 1960 that had been retracted later owing to irreproducibility, and provide a rational basis for the systematic development of the reaction for preparative purposes as an alternative to diazomethane or Simmons-Smith conditions.

Full text of DOI:10.1002/anie.201505482

.
Angewandte
Communications
International Edition: DOI: 10.1002/anie.201505482
German Edition: DOI: 10.1002/ange.201505482
Transition-Metal Catalysis
Nickel-Catalyzed Cyclopropanation with NMe4OTf and nBuLi
Stefan A. Künzi, Juan Manuel Sarria Toro, Tim den Hartog, and Peter Chen*
Abstract: Nickel was identified as a catalyst for the cyclo-
propanation of unactivated olefins by using in situ generated
lithiomethyl trimethylammonium triflate as a methylene donor.
A mechanistic hypothesis is proposed in which the generation
of a reactive nickel carbene explains several interesting
observations. Additionally, our findings shed light on
a report by Franzen and Wittig published in 1960 that had
been retracted later owing to irreproducibility, and provide
a rational basis for the systematic development of the reaction
for preparative purposes as an alternative to diazomethane or
Simmons–Smith conditions.
reactions owing to the inherent hazards of the reagent
itself.[5,6] A safe, scalable, and cost-effective method for
these substrates would therefore be highly desirable.
Our group recently reported the addition of lithiomethyl
trimethylammonium triflate to unsaturated bonds to give
aziridines, epoxides, and cyclopropanes.[7,8] Even though
styrenes and stilbenes could be cyclopropanated in good to
excellent yields, the substrate scope was limited to these
activated alkenes; attempts to cyclopropanate an unactivated
olefin, cyclohexene, resulted only in recovery of the starting
material.
We report now that a variety of nickel complexes catalyze
the cyclopropanation of unactivated olefins. In the absence of
catalyst, no reaction takes place (Table 1, entry 1). Different
I
n 1960, in a short Communication in Angewandte Chemie,
Franzen and Wittig reported the cyclopropanation of cyclo-
hexene with tetramethylammonium bromide (1) to give
norcarane (2).[1] Dropwise addition of an ethereal solution
of phenyllithium/phenylsodium (1:10) to a suspension of 1 in
neat cyclohexene resulted in the formation of 2 in 5–18%
yield. Franzen and Wittig proposed the decomposition of
trimethylammonium methylide into a free carbene that is
trapped by cyclohexene to give 2. In 1964, in Liebigs Annalen,
Wittig and Krauss reported that the 1960 cyclopropanation
could not be reproduced under any conditions, and explicitly
retracted the earlier claim.[2] Since then, this verdict has
seemingly remained unchallenged; there have been no
published reports of any attempts to use tetramethylammo-
nium salts as methylene donors and no citations of the work.[3]
Herein, we report the nickel-catalyzed cyclopropanation of
unactivated olefins with tetramethylammonium triflate and
nBuLi, which suggests that Franzenꢀs work on cyclopropana-
tion might have worked after all, albeit with the unsuspected
influence of a metal catalyst. Moreover, we report unusual
features of the reaction that would have hindered reproduc-
tion of the original work.
Table 1: Screening of catalysts.
Entry[a]
Catalyst
Yield [%][b]
1
2
3
4
None
0[c,d]
16
0.5 mol% [(dme)NiBr2]
1 mol% [(PPh3)2NiBr2]
1 mol% [Ni(cod)2]/2 mol% PPh3
23 (25)[d,e]
19[d]
[a] Conditions: NMe4OTf (1 equiv, 0.1 mmol), BuLi (1.05 equiv), COE
(5 equiv) in THF (0.05m). COE=cyclooctene, dme=dimethoxyethane,
cod=1,5-cyclooctadiene. [b] Calibrated GC yields with undecane as
internal standard. Average of two independent runs. [c] 15 equiv COE
used. [d] Single run only. [e] Slow addition of nBuLi over 5.5 h, 0.5 mmol
scale.
nickel sources, with and without added phosphine ligands,
proved to be effective (Entries 2 to 4), with [(PPh3)2NiBr2]
giving bicyclo[6.1.0]nonane (3) from cyclooctene, for exam-
ple, in 23% yield based on the ylide (Entry 3). Interestingly,
there seems to be no major difference between employing
a Ni0 or NiII complex (Entry 3 vs. 4). Surprisingly, no other
metal catalyst (Fe, Co, Cu, Zn, Rh, Pd, Re) tested yielded any
significant cyclopropanation product, which suggests that
nickel is unique in this reaction. (Table S2 in the Supporting
Information).
There are only a few methods for the cyclopropanation of
electron-rich olefins known, especially for simple methylena-
tion. The Simmons–Smith reaction is the method of choice in
industry but has the drawback of generating a large amount of
waste.[4] The decomposition of diazomethane with transition
metals, although very versatile, is limited to small-scale
[*] S. A. Künzi, Dr. J. M. Sarria Toro,[+] Dr. T. den Hartog,[$]
Prof. Dr. P. Chen
Table 2 shows that the yield for the cyclopropanation of
electron-rich olefins parallels the proclivity of the substrates
for forming olefin p-complexes.[9] Based on our previous
observations,[7,8] one would expect direct carbolithiation for
electron-poor olefins. As noted previously,[8] we found that
the lithiomethyl trimethylammonium reagent decomposes
when the reaction vessel is open to an argon line, which allows
gas exchange. The same decomposition occurs when using Ni
as a catalyst. When the flask is open to an argon line, a distinct
Laboratorium für Organsiche Chemie, ETH Zürich
Vladimir-Prelog-Weg 2, 8093 Zürich (Switzerland)
E-mail: peter.chen@org.chem.ethz.ch
[+] Present address: ICIQ
Av. Països Catalan 16, 43007 Tarragona (Spain)
[$] Present address: Institut für Technische und Makromol. Chemie,
RWTH Aachen University, Worringerweg 1, 52075 Aachen (Germany)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201505482.
10670
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 10670 –10674

Angewandte
Chemie
Table 2: Cyclopropanation yield for other alkenes.
Entry[a]
Alkene
Yield [%][b]
1
2
3
4
5
Norbornene
4-Phenyl-1-butene
Cyclohexene
3,4-Dihydro-2H-pyrane
a-Pinene
79
40
8
0[c,d]
0[c,d]
[a] For conditions, see Table 1; 5 equiv alkene. [b] Calibrated GC yields
with undecane as internal standard. Average of two independent runs.
[c] Single run only. [d] Only starting material observed.
Table 3: Difference between open and closed reaction vessels for the
cyclopropanation of cyclooctene.
Entry[a]
Catalyst
Open/closed[b]
Yield [%][c]
Figure 1. Dependence of the yield of the reaction cyclooctene!3 on
[(PPh3)2NiBr2] loading. Each point represents the average of two
independent runs.
1
2
3
4
0.5 mol% [(dme)NiBr2]
0.5 mol% [(dme)NiBr2]
1 mol% [(PPh3)2NiBr2]
1 mol% [(PPh3)2NiBr2]
Open
Closed
Open
<1
16
16
Closed
23
[a] For conditions see Table 1. [b] Open/closed refers to whether the
reaction was run in a closed vessel or open to an argon line. [c] Calibrated
GC yields with undecane as internal standard. Average of two
independent runs.
decrease in the cyclopropanation yield is obtained (Table 3,
entry 1 vs. 2, 3 vs. 4) relative to the case where an identical
flask is kept closed after the addition of nBuLi. Interestingly,
with catalytic amounts PPh3, there is a much reduced
dependence on whether the flask is open or closed.
Despite this finding, no or very little influence on product
yield has been observed for the addition of NMe3 (Table S5),
at least in the molar excesses we tried. This holds true for
catalysts with and without phosphine ligands. We note that the
reaction catalyzed by [(dme)NiBr2] alone is unpredictable
and yields between 2% and 20% were obtained under
otherwise identical conditions, whereas [(PPh3)2NiBr2] cata-
lyzed the reaction reproducibly and reliably.[10] Especially for
catalysts with no phosphine ligand, this seems surprising given
the marked influence on yield of whether the flask is open or
closed (Table 3).
Scheme 1. Proposed catalytic cycle for the nickel-catalyzed cyclopropa-
nation with lithiomethyl trimethylammonium triflate as a methylene
donor.
A
proposed schematic catalytic cycle is shown in
A third, important observation was made based on the
catalyst loading. With no catalyst, there is no cyclopropana-
tion. With 5 mol% catalyst, there is also no cyclopropanation.
A maximum in the cyclopropanation yield is obtained with
a catalyst loading of between 0.1 and 1.0 mol% (Figure 1).[11]
Importantly, no other products were observed that corre-
spond to alternative CH2 insertions (e.g. methylenecyclooc-
tane or 1-methylcyclooctene from cyclooctene)[12] or Stevens
rearrangement of the lithiomethyl trimethylammonium to
N,N-dimethylethylamine.[13] The only identifiable byproducts
Scheme 1. Under the reaction conditions, the precatalyst is
likely first reduced by the alkyllithium species RLi (R = Bu or
CH2NMe3) that are present in solution since NiII halides are
known to be reduced to Ni0 by nBuLi.[9] The active Ni0
catalyst 4 (L = PPh3 or COE) then adds lithiomethyl trime-
thylammonium triflate to give the niccolate 5. After the
elimination of NMe3, nickel carbene 6 is formed, which can
react further to an intermediate nickelacyclobutane 7 (inner-
sphere mechanism) or react directly via a transition state
(outer-sphere mechanism; not shown) with an olefin to give
the cyclopropane product and regenerate 4. A DFT study by
Geng et al. indicated that both cyclopropanation pathways
could be plausible for Ni carbenes.[14] Reversible trapping by
a nucleophile present in solution, either NMe3 or PPh3, leads
to 5ꢀ and 8, respectively, thereby reducing the steady-state
1
were a polymer, identified as polyethylene by GPC, IR, H
and 13C NMR spectroscopy (Figures S14–S19), which was
isolated in 17% yield, and cyclopropane, presumably from the
cyclopropanation of ethylene, which was also observed in
significant amounts by 13C NMR (Figure S12).
Angew. Chem. Int. Ed. 2015, 54, 10670 –10674
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 10671

.
Angewandte
Communications
concentration of carbene intermediate 6. Given the volatility
(b.p. 38C) of NMe3, this proposed trapping might also explain
why reactions with catalytic PPh3 in addition to Ni are less
susceptible to whether the flask is open to an argon line or
closed (Table 3). Stoichiometric methylenation of PPh3 had
been observed by Franzen and Wittig.[1] While the catalytic
cycle proposed in Scheme 1 is admittedly speculative at this
time, it does account adequately for the oddities of the
reaction, which may have been the cause of the difficulties
encountered 50 years ago. Moreover, each individual step in
Scheme 1 has reasonable precedent in the literature.
Although, to our knowledge, there are no reported
adducts of trimethylammonium methylides to nickel, the
formally isoelectronic ZnII and HgII adducts have been
synthesized. They are ill-characterized in the literature, and
no attempts were made to cyclopropanate alkenes.[15] In situ
prepared allyl ammonium ylide adducts of Cu and Zn have
been used in [2,3] rearrangements, but no cyclopropanation
derived from the ylides was observed.[16] Nevertheless, the
behavior of Ni0 as a weak Lewis acid (the L2Ni0 fragment is
isolobal to DCH2) has been studied in detail. Pçrschke et al.
reported the addition of MeLi and Ph3PCH2 to Ni0 to give the
corresponding niccolates(0) (Figure 2).[17] The compound
Figure 3. Nickel carbenes isolated and structurally characterized (non-
Fischer) by Hillhouse et al.[18]
Even though we have not detected the formation of
ethylene under the conditions where we observe cyclopropa-
nation, we had previously observed the decomposition of
lithiomethyl trimethylammonium to give ethylene.[8]
A
homocoupling of two nickel carbenes can be expected to be
second-order with respect to the carbene, as had been shown
by Gladysz et al. for an electrophilic rhenium carbene and by
Schrock et al. for a nucleophilic Ta carbene,[20] while cyclo-
propanation is first-order, thereby accounting for the strongly
non-monotonic dependence on catalyst loading (Figure 1), in
analogy to the persistent radical effect.[21] Additionally, the
formation of polymer and the lack of ethylene could be
explained by a subsequent polymerization and/or cyclopro-
panation of ethylene by nickel.[22] While ethylene in our
reaction could, in principle, derive from decomposition of
THF induced by nBuLi, this decomposition has been reported
under conditions involving higher temperature and longer
reaction times than we employed.[23] Moreover, in our
previous work, we observed C2H4 in a closed NMR tube
experiment where the solvent was [D8]THF. Finaly, the
decomposition of THF by nBuLi was reported to produce,
besides ethylene, the lithium enolate of acetaldehyde, which
was trapped by benzophenone. We found no evidence for this
adduct in our previous work, in which our lithiomethyl
trimethylammonium triflate was prepared in situ and reacted
with benzophenone.[7] Accordingly, we consider it most
probable that the polyethylene and cyclopropane side prod-
ucts in our reaction derive from homocoupling of nickel
carbene 6.
Nickelacyclobutanes formed by the reaction of a nickel
carbene and an alkene have been invoked as reactive
intermediates in cyclopropanation reactions to explain the
observed stereochemistry.[24] Grubbs and Miyashita published
a series of papers on the synthesis and reactivity of nickel-
acyclobutanes of the type [(R3P)nNi(CH2C(Me)2CH2)], for
example, 15 (R = Ph, n = 2; Figure 4).[25] These complexes are
stable at low temperature. Thermal decomposition of 15 or
oxidation with O2 results in the formation of 1,1-dimethylcy-
clopropane in 47%[25b] and 80%[25c] yield, respectively.
Experimental observations indicate that nickelacyclobutane
15 might be in equilibrium with the corresponding carbene-
and alkene-containing complex. Thermal decomposition of 15
in the presence of cyclohexene produced norcarane (2) in
10% yield.[25d] In the present case, labeling studies with 13C-
enriched NMe4OTf showed no scrambling of the label. Only
enriched NMe3 and isotopic incorporation into the CH2 of the
cyclopropane product were observed, that is, there is no
indication for metathesis (see the Supporting Information).
Hillhouse et al. reported on the similar complexes 16a and
Figure 2. Structurally characterized niccolates(0) relevant to this study.
Compound 9 has been identified by X-ray crystallography and com-
pound 10 has been studied in solution by NMR spectroscopy.[17]
[(PMDTA)(LiMe)Ni(C2H4)2] (9; PMDTA = N,N,N’,N’,N’’-
pentamethyldiethylenetriamine) was characterized by X-ray
crystallography and is analogous to our proposed intermedi-
ate 5. The P-ylide adduct [(Ph3PCH2)Ni(C2H4)2] (10) was
identified in solution by NMR spectroscopy. Complex 10 is
stable in solution under 08C and stable in the solid state at
room temperature for several hours. Under an atmosphere of
ethylene at higher temperature, 10 decomposes to give
cyclopropane (not quantified), which led the authors to
propose a carbenoid character for the methylene group in 10.
There are only a few terminal nickel carbene complexes
(non-Fischer) known in the literature (Figure 3). Hillhouse
et al. reported the synthesis of nickel carbene complexes 11,
12, and 13 through decomposition of the appropriate diazo
precursor.[18] Heating complex 11 under an atmosphere of
ethylene resulted in the formation of 1,1-diphenylcyclopro-
pane (14) in 85% yield of isolated product.[19] Additionally,
catalytic cyclopropanation was observed with 10 mol% 11
=
and N2 CPh2 under an atmosphere of ethylene to give 14 in
41% yield. A [2+2] cycloaddition of ethylene to the nickel
carbene was proposed to explain the formation of 14.
10672 www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 10670 –10674

Angewandte
Chemie
reaction from Prof. Dr. W. Leitner, RWTH Aachen Univer-
sity.
Keywords: carbenes · cyclopropanation · nickel · ylides
How to cite: Angew. Chem. Int. Ed. 2015, 54, 10670–10674
Angew. Chem. 2015, 127, 10817–10821
Figure 4. Isolated nickelacyclobutanes that undergo reductive elimina-
tion to form cyclopropanes.[25,26]
[1] V. Franzen, G. Wittig, Angew. Chem. 1960, 72, 417 – 417.
[2] G. Wittig, D. Krauss, Justus Liebigs Ann. Chem. 1964, 679, 34 –
41.
[3] We have summarized these historical events elsewhere, see: S. A.
Künzi, J. M. Sarria Toro, T. den Hartog, P. Chen, Isr. J. Chem.,
DOI: 10.1002/ijch.201500041.
[4] a) M. Pasco, N. Gilboa, T. Mejuch, I. Marek, Organometallics
2013, 32, 942 – 950; b) “Cyclopropanation mediated by zinc
organometallics”: A. B. Charette in The Chemistry of Organo-
zinc Compounds, Vol. 1 (Eds.: Z. Rappoport, I. Marek), Wiley,
Chichester, 2006, pp. 237 – 286.
[5] Pd-catalyzed cyclopropanation using CH2N2: a) R. McCrindle,
D. W. J. Sneddon, Organomet. Chem. 1985, 282, 413 – 419;
b) Y. V. Tomilov, A. B. Kostitsyn, O. M. Nefedov, Izv. Akad.
Nauk SSSR Ser. Khim. 1987, 2878; c) E. V. Guseva, N. V.
Volchkov, Y. V. Tomilov, O. M. Nefedov, Eur. J. Org. Chem.
2004, 3136 – 3144, and references therein; d) M. D. Khanova,
R. M. Sultanova, S. S. Zlotskii, V. A. Dokichev, Y. V. Tomilov,
Russ. Chem. Bull. 2005, 54, 1003 – 1007; e) S. E. Denmark, R. A.
Stavenger, A.-M. Faucher, J. P. J. Edwards, Org. Chem. 1997, 62,
3375 – 3389; f) G.-H. Fang, Z.-J. Yan, M.-Z. Deng, Org. Lett.
2004, 6, 357 – 360; g) M. A. Emerzian, W. Davenport, J. Song, J.
Li, I. Erden, Adv. Synth. Catal. 2009, 351, 999 – 1004.
[6] Ni-catalyzed cyclopropanation using CH2N2: A. Nakamura, T.
Toshikatsu, M. Cowie, S. Otsuka, J. A. Ibers, J. Am. Chem. Soc.
1977, 99, 2108 – 2117. Only trace yields of cyclopropanes were
obtained when alkyl- and aryl-substituted olefins were used. For
an alternative Ni-catalyzed cyclopropanation using lithiated
sulfones, see: Y. Gai, M. Julia, J.-N. Verpeaux, Synlett 1991, 56 –
57.
16b (Figure 4).[26] Both undergo facile reductive elimination
to form the corresponding cyclopropanes in quantitative
yield.
Each individual step in the proposed mechanism has
precedent in analogous stoichiometric reactions, although the
present report is the first to demonstrate catalytic turnover.
While the best present cyclopropanation yields are admittedly
only modest to good, we have identified the main side
reactions that consume the lithiomethyl trimethylammonium
triflate in an otherwise very clean reaction. The side reactions
limit the chemical yield of the desired cyclopropane product,
which for the better substrates already corresponds to
turnover numbers on the order of 102 when the catalyst
loading is near or slightly above 0.1 mol%.
In conclusion, we have shown that nickel catalyzes the
cyclopropanation of unactivated olefins with lithiomethyl
trimethylammonium triflate as the methylene donor, albeit in
modest to good chemical yield. Based on our experimental
observations and the presented literature, we propose a plau-
sible mechanistic hypothesis involving a nickel carbene as
a reactive intermediate. Our findings offer a potential explan-
ation for the original 1960 report by Franzen and Wittig, and
identify odd features of the reaction that might have made its
reproduction difficult at the time. The mechanism further-
more provides a rational basis for the systematic development
of the reaction for preparative purposes. Detailed mechanistic
studies are ongoing.
[7] T. den Hartog, J. M. Sarria Toro, E. P. A. Couzijn, P. Chen,
Chem. Commun. 2014, 50, 10604 – 10607.
[8] J. M. Sarria Toro, T. den Hartog, P. Chen, Chem. Commun. 2014,
50, 10608 – 10610.
[9] P. W. Jolly, G. Wilke, The Organic Chemistry of Nickel, Vol. 1,
Academic Press, New York, 1974.
[10] This could be explained by spontaneous formation of an inactive
Ni0 species when no stabilizing PPh3 ligand is present.
[11] [(dme)NiBr2] without an additional ligand showed identical
behavior, see Supporting Information.
[12] T. den Hartog, J. M. Sarria Toro, P. Chen, Org. Lett. 2014, 16,
1100 – 1103.
[13] T. S. Stevens, E. M. Creighton, A. B. Gordon, M. MacNicol, J.
Chem. Soc. 1928, 3193 – 3197.
[14] X. Zhang, Z.-Y. Geng, Y.-C. Wang, W.-Q. Li, Z. Wang, F.-X. Liu,
J. Mol. Struct.: THEOCHEM 2009, 893, 56 – 66.
[15] a) G. Wittig, K. Schwarzenbach, Justus Liebigs Ann. Chem. 1961,
650, 1 – 20; b) H. Daniel, J. Paetsch, Chem. Ber. 1965, 98, 1915 –
1922; c) W. K. Musker, R. R. Stevens, Inorg. Chem. 1969, 8, 255 –
264.
[16] a) V. K. Aggarwal, G. Y. Fang, J. P. H. Charmant, G. Meek, Org.
Lett. 2003, 5, 1757 – 1760; b) K. Honda, H. Shibuya, H. Yasui, Y.
Hoshino, S. Inoue, Bull. Chem. Soc. Jpn. 2008, 81, 142 – 147.
[17] a) K.-R. Pçrschke, K. Jonas, G. Wilke, R. Benn, R. Mynott, R.
Goddard, C. Krüger, Chem. Ber. 1985, 118, 275 – 297; b) K.-R.
Pçrschke, G. Wilke, R. Mynott, Chem. Ber. 1985, 118, 298 – 312.
[18] a) D. J. Mindiola, G. L. Hillhouse, J. Am. Chem. Soc. 2002, 124,
9976 – 9977; b) V. M. Iluc, G. L. Hillhouse, J. Am. Chem. Soc.
2014, 136, 6479 – 6488.
Experimental Section
General procedure for cyclopropanation: An oven-dried 5 mL
Schlenk flask was charged with NMe4OTf (22.3 mg, 0.1 mmol),
cyclooctene (65 mL, 0.5 mmol), catalyst in anhydrous THF, and
anhydrous THF (total volume 2 mL, 0.05m) inside the glove box.
The flask was removed and cooled to 08C, and nBuLi (66 mL,
0.105 mmol) was added dropwise. The flask was sealed and the
reaction was allowed to warm to RT overnight. After 16 h, undecane
was added as an internal standard and the yield was determined by
GC–FID.
Acknowledgements
Funding from the ETH Zürich, the Swiss National Science
Foundation, and the European Union Seventh Framework
Programme (FP7/2007-2013) under grant agreement PIEF-
GA-2010-275400 (IEF-Marie Curie post-doctoral grant to
TdH) are gratefully acknowledged. We also acknowledge
assistance with regard to independent reproduction of the
Angew. Chem. Int. Ed. 2015, 54, 10670 –10674
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 10673

.
Angewandte
Communications
[19] R. Waterman, G. L. Hillhouse, J. Am. Chem. Soc. 2003, 125,
13350 – 13351.
[24] R. Noyori, H. Kawauchi, H. Takaya, Tetrahedron Lett. 1974, 15,
1749 – 1752.
[25] a) R. H. Grubbs, A. Miyashita, J. Am. Chem. Soc. 1978, 100,
7418 – 7420; b) A. Miyashita, R. H. Grubbs, Tetrahedron Lett.
1981, 22, 1255 – 1256; c) A. Miyashita, M. Ohyoshi, H. Shitara,
H. Nohira, J. Organomet. Chem. 1988, 338, 103 – 111; d) R.
Grubbs, A. Miyashita in Fundamental Research in Homogeneous
Catalysis (Ed.: M. Tsutsui), Springer, New York, 1979, pp. 151 –
163.
[20] a) J. H. Merrifield, G.-Y. Lin, W. A. Kiel, J. A. Gladysz, J. Am.
Chem. Soc. 1983, 105, 5811 – 5819; b) R. R. Schrock, P. R. Sharp,
J. Am. Chem. Soc. 1978, 100, 2389 – 2399.
[21] a) H. Fischer, J. Am. Chem. Soc. 1986, 108, 3925 – 3927; b) K.-S.
Focsaneanu, J. C. Scaiano, Helv. Chim. Acta 2006, 89, 2473 –
2482.
[22] K. Ziegler, E. Holzkamp, H. Breil, H. Martin, Angew. Chem.
1955, 67, 541 – 547.
[23] a) A. Maercker, W. Theysohn, Justus Leibigs Ann. Chem. 1971,
747, 70 – 83; b) P. Tomboulian, D. Amick, S. Beare, K. Dumke, D.
Hart, R. Hites, A. Metzger, R. Nowak, J. Org. Chem. 1973, 38,
322 – 325.
[26] K. D. Kitiachvili, D. J. Mindiola, G. L. Hillhouse, J. Am. Chem.
Soc. 2004, 126, 10554 – 10555.
Received: June 15, 2015
Published online: July 23, 2015
10674 www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 10670 –10674