Gold-catalyzed diastereoselective cycloisomerization of alkylidene-cyclopropane-bearing 1,6-diynes

Article abstract of DOI:10.1002/anie.201405147

An unprecedented gold-catalyzed diastereoselective cycloisomerization of 1,6-diynes bearing an alkylidene cyclopropane moiety has been developed. This methodology enables rapid access to a variety of 1,2-trimethylenenorbornanes, which are important building blocks in the preparations of abiotic and sesquiterpene core structures. When gold met alkylidenecyclopropane: Cationic gold catalysts can mediate the highly exothermic (≈60 kcal mol^-1) cycloisomerization of 1,6-diynes bearing an alkylidene cyclopropane moiety. This diastereoselective methodology efficiently generates 1,2- trimethylenenorbornanes, an important building block for abiotic targets and sesquiterpene natural products. DCE=1,2-dichloroethane, Tf= trifluoromethanesulfonyl, Tol=Tolyl.

Full text of DOI:10.1002/anie.201405147

.
Angewandte
Communications
DOI: 10.1002/anie.201405147
Gold Catalysis
Gold-Catalyzed Diastereoselective Cycloisomerization of Alkylidene-
Cyclopropane-Bearing 1,6-Diynes**
Hongchao Zheng, Laura L. Adduci, Ryan J. Felix, and Michel R. Gagnꢀ*
Abstract: An unprecedented gold-catalyzed diastereoselective
cycloisomerization of 1,6-diynes bearing an alkylidene cyclo-
propane moiety has been developed. This methodology
enables rapid access to a variety of 1,2-trimethylenenorbor-
nanes, which are important building blocks in the preparations
of abiotic and sesquiterpene core structures.
Taking advantage of the ring strain relief strategy, our group
recently reported the first enantioselective Cope rearrange-
ment (gold-catalyzed) from achiral 1,5-dienes (Sche-
me 2A).[9b] This success led us to investigate ACP-containing
cyclic 1,5-dienes and to the discovery of a ring-expanding
I
ncreasing pressure on our natural resources has made
sustainability a key theme in many research areas, which in
turn has led to a premium being applied to methods able to
generate complex target molecules in a step- and atom-
economical manner.[1] Since catalysis provides the where-
withal to improve efficiency, selectivity, complexity, and
rate,[2] it provides a powerful tool for the efficient construction
of complex chemical architectures that would be difficult to
achieve using traditional reaction schemes.[3]
Illustrative of this point is 1,2-trimethylenenorbornane 2,
a structural motif that occurs in numerous synthetic precur-
sors to abiotic adamantanes[4] and [3.3.3]-propellanes,[5] in
addition to numerous classes of sesquiterpene natural prod-
ucts, including the cedrenes,[6] and the pentalenene/isoco-
menes (Scheme 1).[7] The traditional approaches to 2 have
relied on intramolecular [4+2] cycloaddition reactions,[4–7] but
the laborious process for preparing the needed starting
materials has limited the applicability of 2 in complex
molecule synthesis. Efficient methods for the construction
of the tricyclic ring system of 2 under catalyst control are thus
desirable and would enable their application in synthesis. To
this end, we report a gold-catalyzed, remarkably complex
cycloisomerization of readily synthesized 1,6-diynes 1 to
a collection of products containing the 1,2-trimethylenenor-
bornane core (Scheme 1).
Scheme 1. 1,2-Trimethylenenorbornane 2 as synthetic intermediate.
The inherently high ring strain of alkylidene cyclopro-
panes (ACPs) ( ꢀ 40 kcalmolÀ1) make them an especially
useful and reactive class of synthetic intermediates.[8] When
this strain can be released under the careful guidance of
a catalyst, it is possible for this potent thermodynamic driving
force to be harnessed for otherwise unfavorable reactions.[9]
[*] Dr. H. Zheng, L. L. Adduci, Dr. R. J. Felix, Prof. Dr. M. R. Gagnꢀ
Department of Chemistry
University of North Carolina at Chapel Hill
Chapel Hill, NC 27599-3290 (USA)
E-mail: mgagne@unc.edu
[**] We thank the National Institutes of Health, General Medicine (GM-
60578) for generous support, Dr. Mee-Kyung Chung for help with
HRMS analyses, and Prof. David Nicewicz (UNC-Chapel Hill) for
helpful discussion on the reaction mechanism.
Scheme 2. Gold-catalyzed: A) Enantioselective Cope rearrangement of
achiral 1,5-dienes. B) Ring-expanding cycloisomerization of 1,5-dienes.
C) Cycloisomerization of cyclopropylidene-bearing 1,6-diyne 1a.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201405147.
7904
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 7904 –7907

Angewandte
Chemie
cycloisomerization that yields tricyclic compounds incorpo-
rating the bicyclo[4.2.0]oct-1-ene core (Scheme 2B).[9c] Based
on this result, we reasoned that gold catalysis of ACP-bearing
1,5-enynes such as 1a, would yield bicyclo[4.2.0] dienes (3)
through a sequential 6-endo-dig-cyclization/ring expansion/
net 1,2-hydrogen shift process (Scheme 2C).[10] However,
when 1a (R2 = 1-propynyl) was treated with 10 mol%
PPh3AuNTf2 in 1,2-dichloroethane at 508C, the expected
bicyclo product 3 was not formed. Instead, the tricyclic
compound 2a was obtained as a single diastereomer in 70%
yield (48 h, Scheme 2C). DFT calculations indicated that the
conversion of 1a to 2a was exothermic by approximately
63 kcalmolÀ1, making the high selectivity and yield even more
remarkable.[11] The structure of 2a was elucidated by a 2D-
INADEQUATE experiment (see the Supporting Informa-
tion).
In the first optimization round, a number of gold catalysts
were evaluated for their ability to accelerate the cycloisome-
rization of 1a (entries 1–9, Table 1). The simple Lewis acid
AuCl3 and the common cationic gold precursor PPh3AuCl
showed almost no catalytic activity for this rearrangement
(entries 1–2, Table 1). In addition, the data indicated that the
ligand has a significant impact on the catalytic activity, with
triaryl phosphines (entries 3, 6, and 7, Table 1) exhibiting
better catalytic efficiency than trialkyl (entry 4, Table 1) and
mixed aryl–alkyl phosphine ligands (entry 5, Table 1). Other
ligands, including dialkyl sulfide (entry 8), and an N-hetero-
cyclic carbene (entry 9) showed no improvements. Electron-
rich triaryl phosphine ligands (entry 7) were superior to
electron-poor variants (entry 6). The catalyst (p-Tol)3-
PAuNTf2 4, derived from the activation of (p-Tol)3PAuCl
with AgNTf2 provided the highest yield (entry 7, Table 1). A
subsequent screen of silver salts (entries 7 and 10–12)
confirmed that AgNTf2 and AgSbF6 provided optimal yields
(entries 7 and 11). Since 4 is an air-stable white solid, this
catalyst was chosen for additional optimization.
A solvent optimization study using 10 mol% 4 showed
a preference for halogenated solvents (see the Supporting
Information). The need for slightly elevated temperatures led
to 1,2-dichloroethane (DCE) being chosen (entry 7). A screen
of catalyst loadings showed that 10 mol% was preferable
(entries 7, 13, and 14). Increasing the concentration of the
reaction mixture had little impact on the reaction yield
(entries 7 and 15). Finally, a control run in the absence of 4 led
to no product formation (entry 16).
With these optimized conditions (10 mol% 4 in DCE at
508C), the scope of substrates was explored. As shown in
Table 2, a variety of aryl substitutions at R1 were tolerated,
with electron-rich through electron-poor substituents
(entries 1–5, Table 2) successfully generating the expected
tricyclic compounds 2a–2e in good yield. Worth noting is the
tolerance of the cyano group in 1 f (entry 6), which has the
potential for side reactions through nitrile activation.[12] The
Table 1: Optimization of reaction conditions for the Au-catalyzed cyclo-
isomerization of 1a.[a]
Table 2: Gold-catalyzed cycloisomerizations of 1,6-diynes 1.[a,b]
Entry
LAuCl
AgX
Loading
(mol%)
Yield [%][b]
1
2
3
4
5
6
7
8
AuCl3
Ph3PAuCl
Ph3PAuCl
Me3PAuCl
(tBuXPhos)AuCl
(F5C6)3PAuCl
(p-Tol)3PAuCl
Me2SAuCl
–
–
10
10
10
10
10
10
10
10
10
10
10
10
20
5
trace
0
70
17
52
36
75
complex mixture
trace
65
71
61
AgNTf2
AgNTf2
AgNTf2
AgNTf2
AgNTf2
AgNTf2
AgNTf2
AgBF4
AgSbF6
AgPF6
9
(IPr)AuCl
10
11
12
13
14
15[d]
16[e]
(p-Tol)3PAuCl
(p-Tol)3PAuCl
(p-Tol)3PAuCl
(p-Tol)3PAuNTf2 (4)
(p-Tol)3PAuNTf2 (4)
(p-Tol)3PAuNTf2 (4)
74
32[c]
72
10
–
–
0
[a] Reaction conditions: LAuCl (0.01 mmol) was added to a solution of
AgX (0.01 mmol) in DCE (1.0 mL) at RT. The solution was stirred at RT
for 15 min and the precipitate was filtered over Celite. To the filtrate was
added 1a (22 mg, 0.1 mmol) and the resulting mixture was warmed to
508C and stirred for 48 h. [b] Yields of 2a purified by column chroma-
tography. [c] The yield was not increased even if a prolonged reaction
time (96 h) was employed. [d] A more concentrated reaction mixture
(0.4m) was employed. [e] Control: no gold catalyst added. IPr=1,3-
bis(2,6-diisopropylphenyl)-imidazolidene, Tf=trifluoromethane-sulfo-
nyl, p-Tol=para-toluenyl. XPhos=2-dicyclohexyl-phosphino-2’,4’,6’-tri-
isopropylbiphenyl.
[a] See the Experimental Section for the typical procedure. [b] Yields of
isolated 2 purified by column chromatography on silica gel. [c] A yield of
28% of bicyclo[4.2.0]diene 3h was isolated (see Scheme 1).
Angew. Chem. Int. Ed. 2014, 53, 7904 –7907
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 7905

.
Angewandte
Communications
use of an aliphatic substituent in place of the aryl moiety at R1
afforded the desired tricyclic compound, 2g, but in a slightly
lower yield (entries 1–5 versus entry 7). While a substrate
bearing a terminal alkynyl group successfully rearranged to
the tricyclic compound, the product was too volatile to isolate
from the reaction mixture. The introduction of a benzyl group
at R3 led to the desired product, 2h, and enabled its isolation
in a synthetically useful yield (entry 8). 1,6-Diynes with
sterically hindered substituents at the R2 position were also
suitable, provided a longer reaction time was employed
(entries 9 and 10). In all cases, the desired tricyclic products
were obtained as a single diastereomer (Table 2).
The mechanism of this cycloisomerization is undoubtedly
complex, but some preliminary observations are included
here.[13] When 1h, which bears a terminal alkyne, was
subjected to the optimal reaction conditions, it afforded 2h
together with the bicyclo compound 3h (Scheme 3A). As
Scheme 4. Proposed mechanism; the black dot denotes the movement
of the 13C label in 5.
kinetically controlled phase of the cycloisomerization
(Scheme 4).[13]
As mentioned in Scheme 2C, the bicyclic diene 3 is
reasonably generated through a sequential 6-endo-dig cycli-
zation/ring expansion/net 1,2-hydrogen shift sequence from
1.[10] Alkyne activation of 3 by the gold complex triggers the
cyclogeneration of allylic carbocation 10, which then under-
goes a 1,2-alkyl shift to afford a second allylic carbocation 11,
followed by elimination to 12. Reactivation of 12 by H+
generates yet another allyl cation 13, which experiences
a 1,2-alkyl shift to furnish the final product. Although
alternative sequences are possible,[14] this mechanism cor-
rectly predicts the kinetic preference for 6 and the conversion
of 8 to 9.
In summary, we have developed a novel gold-catalyzed
high yielding, highly diastereoselective cycloisomerization of
ACP-containing 1,6-diynes leading to tricyclic compounds
containing the 1,2-trimethylenenorbornane core. The reac-
tion is highly exothermic, and yet the catalyst exercises near
perfect control over the product identity and selectivity. The
now straightforward synthesis of the useful 1,2-trimethylene-
norbornane core should enable its applicability in complex
molecule synthesis.
Scheme 3. Mechanistic investigation of the cycloisomerization of alky-
lidene-cyclopropane-bearing 1,6-diynes using isotopic labeling experi-
ments.
discussed above, 3h is reasonably generated through a sequen-
tial 6-endo-dig cyclization/ring expansion/net 1,2-hydrogen
shift sequence from 1h (Scheme 2C).[10] When 3h was treated
with 20 mol% 4 in DCE at 508C, it slowly converted into 2h
(Scheme 3A),[14] suggesting that 3h might be an intermediate
in the conversion of 1h to 2h. The cycloisomerization of
isotopically labeled substrates, 5 and 8, also provided
beneficial mechanistic information.[15] As shown in Sche-
me 3B, when 13C-labeled substrate 5 was treated with
10 mol% 4 in 1,2-dichloroethane at 508C, two isotopomers
were obtained, 6 and 7, in a 3:1 ratio. Resubjecting these
purified products to the reaction conditions (10 mol% 4,
DCE, 508C) converged the mixture to a 1:1 ratio of 6 and 7.
Those experiments suggest that gold catalysis of 5 initially
affords 6 as the kinetic product, but that a secondary process
acts to interconvert these two positions in the product.
Although the mechanism for interconversion of 6 and 7 is not
known, this 13C-labeling experiment together with the con-
version of 8 to 9 (Scheme 3C) suggests a mechanism for the
Experimental Section
Typical procedure for the gold-catalyzed formation of 7-methylene-4-
phenyl-2,3,6,7-tetrahydro-3a,6-methanoindene (2a, Table 2): To a so-
lution of 1a (22 mg, 0.1 mmol) in DCE (1.0 mL) at RT was added (p-
Tol)3PAuNTf2 4 (7.8 mg, 0.01 mmol). The resulting solution was
stirred at 508C for 48 h. Upon evaporation of the solvent under
reduced pressure, the residue was purified by silica gel column
chromatography (100% hexanes) to afford 2a (16.5 mg, 75% yield)
in pure form. 1H NMR (600 MHz, CD2Cl2): d = 7.36 (t, J = 7.6 Hz,
7906 www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 7904 –7907

Angewandte
Chemie
2H), 7.28 (t, J = 7.4 Hz, 1H), 7.22 (d, J = 7.7 Hz, 2H), 6.02 (d, J =
3.2 Hz, 1H), 5.70 (t, J = 2.4 Hz, 1H), 5.22 (s, 1H), 5.08 (s, 1H), 3.52–
3.48 (m, 1H), 3.10–3.00 (m, 1H), 2.97–2.90 (m, 1H), 2.37 (ddd, J =
14.1, 8.5, 1.9 Hz, 1H), 2.12 (d, J = 7.4 Hz, 1H), 1.98–1.90 (m, 1H),
1.65 ppm (d, J = 7.7 Hz, 1H). 13C NMR (150 MHz, CD2Cl2): d =
153.2, 150.8, 146.1, 137.2, 132.2, 128.2, 126.8, 126.2, 114.9, 103.9,
70.1, 58.8, 54.3, 38.4, 24.3 ppm. HRMS (EI) for C17H16: calcd.
220.1252, found 220.1253. 2a–2j were similarly prepared and fully
characterized (see the Supporting Information).
[6] a) E. G. Breitholle, A. G. Fallis, Can. J. Chem. 1976, 54, 1991 –
1993; b) E. G. Breitholle, A. G. Fallis, J. Org. Chem. 1978, 43,
1964 – 1968; c) K. Makita, K. Fukumoto, M. Ihara, Tetrahedron
Lett. 1997, 38, 5197 – 5200; d) M. Ihara, K. Makita, K. Takasu, J.
Org. Chem. 1999, 64, 1259 – 1264.
[7] M. Hatanaka, F. Ueno, I. Ueda, Tetrahedron Lett. 1996, 37, 89 –
90.
[8] For a review of cyclopropylidenes, see: a) H. Pellissier, Tetrahe-
dron 2010, 66, 8341 – 8375; b) D. J. Mack, J. T. Njardarson, ACS
Catal. 2013, 3, 272 – 286; c) D.-H. Zhang, X.-Y. Tang, M. Shi, Acc.
Chem. Res. 2014, 47, 913 – 924; For recent examples on chemical
transformations involving cyclopropylidenes, see: d) L.-X. Shao,
Y.-P. Zhang, M.-H. Qi, M. Shi, Org. Lett. 2007, 9, 117 – 120;
e) P. A. Evans, P. A. Inglesby, J. Am. Chem. Soc. 2008, 130,
12838 – 12839; f) S. Mazumder, D. Shang, D. E. Negru, M.-H.
Baik, P. A. Evans, J. Am. Chem. Soc. 2012, 134, 20569 – 20572;
g) P. A. Evans, P. A. Inglesby, K. Kilbride, Org. Lett. 2013, 15,
1798 – 1801; h) S. Cui, Y. Zhang, Q. Wu, Chem. Sci. 2013, 4,
3421 – 3426; i) T. Kippo, K. Hamaoka, I. Ryu, J. Am. Chem. Soc.
2013, 135, 632 – 635; j) B. Meng, X. Huang, L. Wu, Adv. Synth.
Catal. 2013, 355, 2637 – 2650.
Received: May 9, 2014
Published online: June 11, 2014
Keywords: 1,6-diynes · cycloisomerization · cyclopropylidenes ·
.
gold catalysis · synthetic methods
[1] a) B. M. Trost, Science 1991, 254, 1471 – 1478; b) B. M. Trost,
Angew. Chem. 1995, 107, 285 – 307; Angew. Chem. Int. Ed. Engl.
1995, 34, 259 – 281; c) B. M. Trost, Acc. Chem. Res. 2002, 35, 695 –
705; d) T. Gaich, P. S. Baran, J. Org. Chem. 2010, 75, 4657 – 4673.
[2] C. A. Busacca, D. R. Fandrick, J. J. Song, C. H. Senanayake, Adv.
Synth. Catal. 2011, 353, 1825 – 1864.
[9] a) B. M. Trost, Top. Curr. Chem. 1986, 133, 3 – 82; b) R. J. Felix,
D. Weber, O. Gutierrez, D. J. Tantillo, M. R. Gagnꢁ, Nat. Chem.
2012, 4, 405 – 409; c) R. J. Felix, O. Gutierrez, D. J. Tantillo,
M. R. Gagnꢁ, J. Org. Chem. 2013, 78, 5685 – 5690.
[3] For selected reviews, see: a) A. Fꢀrstner, P. W. Davies, AAngew.
Chem. 2007, 119, 3478 – 3519; Angew. Chem. Int. Ed. 2007, 46,
3410 – 3449; b) A. S. K. Hashmi, Chem. Rev. 2007, 107, 3180 –
3211; c) Z. Li, C. Brouwer, C. He, Chem. Rev. 2008, 108, 3239 –
3265; d) A. Arcadi, Chem. Rev. 2008, 108, 3266 – 3325; e) E.
[10] a) H. Zheng, R. J. Felix, M. R. Gagnꢁ, Org. Lett. 2014, 16, 2272 –
2275; for other examples of gold(I)-catalyzed chemical trans-
formations that proceed via a cyclopropylmethyl cation inter-
mediate, see b) G. A. Olah, V. P. Reddy, G. K. S. Prakash, Chem.
Rev. 1992, 92, 69 – 95; c) S. G. Sethofer, S. T. Staben, O. Y. Hung,
F. D. Toste, Org. Lett. 2008, 10, 4315 – 4318; d) A. Pitaval, D.
Leboeuf, J. Ceccon, A. M. Echavarren, Org. Lett. 2013, 15, 4580 –
4583; for other examples of gold(I)-catalyzed chemical trans-
formations through a 6-endo-dig cyclization, see: e) F. Barabꢁ,
G. Bꢁtournay, G. Bellavance, L. Barriault, Org. Lett. 2009, 11,
4236 – 4238; f) F. Barabꢁ, P. Levesque, I. Korobkov, L. Barriault,
Org. Lett. 2011, 13, 5580 – 5583; g) B. Sow, G. Bellavance, F.
Barabꢁ, L. Barriault, Beilstein J. Org. Chem. 2011, 7, 1007 – 1013.
[11] Calculations were carried out on Macspartan (B3LYP).
[12] For a recent example of nitrile activation by gold, see: R. S.
Ramꢃn, N. Marion, S. P. Nolan, Chem. Eur. J. 2009, 15, 8695 –
8697.
[13] For recent reviews of gold(I) mechanisms, see: a) H. Schmid-
baur, A. Schier, Organometallics 2010, 29, 2 – 23; b) referen-
ce [3j]; c) L.-P. Liu, G. B. Hammond, Chem. Soc. Rev. 2012, 41,
3129 – 3139; d) A. Gꢃmez-Suꢄrez, S. P. Nolan, Angew. Chem.
2012, 124, 8278 – 8281; Angew. Chem. Int. Ed. 2012, 51, 8156 –
8159; e) L. Braun, A. M. Asiri, A. S. K. Hashmi, ACS Catal.
2013, 3, 1902 – 1907; f) C. Obradors, A. M. Echavarren, Chem.
Commun. 2014, 50, 16 – 28; g) A. S. K. Hashmi, Acc. Chem. Res.
2014, 47, 864 – 876.
´
˜
Jimꢁnez-Nunez, A. M. Echavarren, Chem. Rev. 2008, 108, 3326 –
3350; f) D. J. Gorin, B. D. Sherry, F. D. Toste, Chem. Rev. 2008,
108, 3351 – 3378; g) A. S. K. Hashmi, M. Rudolph, Chem. Soc.
Rev. 2008, 37, 1766 – 1775; h) V. Michelet, P. Y. Toullec, J. P.
GenÞt, Angew. Chem. 2008, 120, 4338 – 4386; Angew. Chem. Int.
Ed. 2008, 47, 4268 – 4315; i) A. Fꢀrstner, Chem. Soc. Rev. 2009,
38, 3208 – 3221; j) F. Gagosz, Tetrahedron 2009, 65, 1757 – 1757;
k) A. S. K. Hashmi, Angew. Chem. 2010, 122, 5360 – 5369;
Angew. Chem. Int. Ed. 2010, 49, 5232 – 5241; l) B. Alcaide, P.
Almendros, J. M. Alonso, Org. Biomol. Chem. 2011, 9, 4405 –
4416; m) B. Alcaide, P. Almendros, J. M. Alonso, Molecules
2011, 16, 7815 – 7843; n) C. Aubert, L. Fensterbank, P. Garcia, M.
Malacria, A. Simonneau, Chem. Rev. 2011, 111, 1954 – 1993;
o) N. Krause, C. Winter, Chem. Rev. 2011, 111, 1994 – 2009;
p) S. P. Nolan, Acc. Chem. Res. 2011, 44, 91 – 100; q) M.
Rudolph, A. S. K. Hashmi, Chem. Soc. Rev. 2012, 41, 2448 –
2462; r) B. L. Lu, L. Dai, M. Shi, Chem. Soc. Rev. 2012, 41,
3318 – 3339; s) C. Obradors, A. M. Echavarren, Acc. Chem. Res.
2014, 47, 902 – 912; t) Y.-M. Wang, A. D. Lackner, F. D. Toste,
Acc. Chem. Res. 2014, 47, 889 – 901.
[4] a) E. M. Engler, M. Farcasiu, A. Sevin, J. M. Cense, P. v. R.
Schleyer, J. Am. Chem. Soc. 1973, 95, 5769 – 5771; b) F. J. Jꢂggi,
C. Ganter, Helv. Chim. Acta 1980, 63, 866 – 871; c) A. M. Klester,
F. J. Jꢂggi, C. Ganter, Helv. Chim. Acta 1980, 63, 1294 – 1295;
d) M. Farcasiu, E. W. Hagaman, E. Wenkert, P. v. R. Schleyer,
Tetrahedron Lett. 1981, 22, 1501 – 1504; e) A. M. Klester, C.
Ganter, Helv. Chim. Acta 1985, 68, 734 – 744.
[14] In practice, the conversion of 3h to 2h occurs considerably more
slowly than the direct transformation in situ and requires
20 mol% 4.
[15] See the Supporting Information for more details.
[5] H. A. Patel, J. B. Stothers, S. E. Thomas, Can. J. Chem. 1994, 72,
56 – 68.
Angew. Chem. Int. Ed. 2014, 53, 7904 –7907
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 7907