Cyclobutenes as isolable intermediates in the gold(I)-catalysed cycloisomerisation of 1,8-enynes

Article abstract of DOI:10.1002/adsc.200900056

The gold(I)-catalysed isomerisation of 1,8-enynes allows the efficient synthesis of function -alised bicyclo[5.2.0]nonenes. Notably, these cyclobutenes derivatives can be isolated as reactive intermediates that could undergo subsequent gold(I)-catalysed t

Full text of DOI:10.1002/adsc.200900056

COMMUNICATIONS
DOI: 10.1002/adsc.200900056
Cyclobutenes as Isolable Intermediates in the Gold(I)-Catalysed
Cycloisomerisation of 1,8-Enynes
Yann Odabachiana and Fabien Gagosza,*
a
Laboratoire de Synthꢀse Organique, UMR 7652 CNRS/Ecole Polytechnique, 91128 Palaiseau, France
Fax : (+33)-(0)1-6933-59 72; e-mail: gagosz@dcso.polytechnique.fr
Received: December 22, 2008; Revised: January 26, 2009; Published online: February 6, 2009
This work is dedicated with admiration to Professor William B. Motherwell
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/adsc.200900056.
Abstract: The gold(I)-catalysed isomerisation of
1,8-enynes allows the efficient synthesis of function-
alised bicycloACHTUNGTRENNUNG[5.2.0]nonenes. Notably, these cyclobu-
tenes derivatives can be isolated as reactive inter-
mediates that could undergo subsequent gold(I)-
catalysed transformations such as isomerisation,
fragmentation or ene reaction to furnish more
structurally complex products. This study also pro-
vides useful information related to the mechanism
leading to metathesis-type derivatives, examples of
which were shown to be produced, in the present
case, by a gold(I)-catalysed ring fragmentation of
the cyclobutene moiety.
Keywords: ACHTUNGTRENNUNG[2+2]cycloaddition; cyclobutenes; gold;
homogeneous catalysis
1,n-Enynes represent a privileged class of substrates
in the area of transition metal-catalysed reactions.[1]
They can indeed be cycloisomerised efficiently and
generally selectively under mild conditions into a
wide variety of products, depending on their substitu-
tion pattern and the nature of the metallic species
Scheme 1. Cyclobutenes as products in metal-catalysed cy-
cloisomerisation of 1,n-enynes.
used. Among these transformations, those related to skeletal rearrangement. This is the case with 1,7-
the formation of a cyclobutene product 2 following a enyne substrates of type 4 bearing an endocyclic
formal [2+2] cycloaddition of an 1,n-enyne 1 are alkene moiety that reacts with an electrophilic
much less common [Scheme 1, Eq. (1)].[2a]
ACTHNGUTERNNUG
Pd(II),[3a] Ga(III),[3b] Pt(II),[3c] or Au(I)[3d] catalyst to
This can be explained by the presence of a variety furnish tricyclic compounds of type 5 [Scheme 1, Eq.
of other skeletal rearrangements that are generally (2)]. Certain other 1,6- and 1,7-enynes 6–8, possessing
more favoured than the [2+2]cycloaddition process, an acyclic alkene and an arylalkyne or ynamide
which leads to potentially more strained bicyclic pro- moiety can also be transformed into the correspond-
ducts.[1,2a] A series of cyclobutenes derivatives have ing cyclobutenes 9–11 in the presence of a Pd(II),[4a,b]
however been isolated as the products of the reaction Pt(II),[4c–g] Au(I),[4h–j] or Ru(II)[4k] catalyst [Scheme 1,
when a specific substitution pattern of the alkene and/ Eqs. (3)–(5)]. Cyclobutenes 2 have also been postulat-
or alkyne disfavours the more commonly observed ed as intermediates in a series of metal-catalysed
Adv. Synth. Catal. 2009, 351, 379 – 386
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
379

Yann Odabachian and Fabien Gagosz
COMMUNICATIONS
[8]
transformations of 1,n-enynes 1 that furnish metathe- the stable (Ph3P)AuNTf2 in CH2Cl2 at room temper-
1
sis type compounds 3 [Scheme 1, Eq (1)].[5]
ature [Scheme 3, Eq. (6)].[9] After 24 h, H NMR anal-
In the specific case of the gold-catalysed single exo- ysis of the crude reaction mixture showed a complete
cleavage skeletal rearrangement of 1,6-enynes such as and clean conversion of 16 into cycloheptadienes 17
12, DFT calculations have, however, lent support to and 18, which were isolated as a 1.4:1 mixture in 91%
mechanistic pathways that do not involve the interme- yield. The formation of deconjugated cycloheptadiene
diacy of such cyclobutenes (Scheme 2).[2a,6]
17 as the major product was however rather surprising
Following our previous investigations in the field of and could not be directly explained following a mech-
gold-catalysed cycloisomerisations of enynes,[7] we anistic pathway analogous to that proposed for the cy-
report herein that a series of 1,8-enynes can be trans- cloisomerisation of 1,6-enynes. Due to the difficulties
formed into functionalised cyclobutenes. These can be encountered in monitoring the reaction by TLC, the
either isolated as the final products of the reaction or conversion was performed in CD2Cl2 and monitored
1
further transformed into various products including by H NMR spectroscopy [Scheme 3, Eq. (7)].
metathesis-type derivatives.
We were pleased to see that another species,
In an attempt to cycloisomerise 1,8-enynes into cy- namely cyclobutene 19, was involved in this transfor-
cloheptene derivatives following a reaction pathway mation. This intermediate rapidly accumulated to a
similar to that reported in the case of 1,6-enynes (see maximum of ca. 20% yield (after 3 h) and then gradu-
Scheme 2), substrate 16 was treated with 4 mol% of ally disappeared while the quantity of 17 and 18 con-
tinuously increased. In order to obtain more insight
into the mechanism of this reaction and unambigu-
ously establish the intermediacy of cyclobutene 19 in
the formation of 17 and 18, we envisaged its isolation
and various gold(I) catalysts were therefore screened
in order to optimise its yield.[10] While a faster trans-
formation was observed when the more electrophilic
gold(I) complex [(Ph3P)AuNCCH3]SbF6 was used as
the catalyst, a disappointing maximum ca. 8% yield of
cyclobutene 19 (after 30 min) was observed. The use
of the bulkier catalyst 20 had a remarkable impact on
the course of the transformation. Gratifyingly, the se-
Scheme 2. Mechanistic pathway for the gold(I)-catalysed
single exo-cleavage skeletal rearrangement of 1,6-enynes.
Scheme 3. Cycloisomerisation of enyne 16/optimisation of the formation of cyclobutene intermediate 19.
380
asc.wiley-vch.de
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 379 – 386

Cyclobutenes as Isolable Intermediates in the Gold(I)-Catalysed Cycloisomerisation
COMMUNICATIONS
lectivity was improved in this case and cyclobutene 19 lysed fragmentation of the cyclobutene moiety.[12] No-
could be isolated in 86% yield when the reaction was tably, even if the gold-catalysed cycloisomerisations of
quenched after 15 min. Resubmitting 19 to the same 1,6-enyne 12 and 1,8-enyne 16 furnish structurally
reaction conditions, furnished a mixture of cyclohep- similar metathesis type compounds 15 and 18, these
tadienes 17 and 18 in the same yield and ratio as are produced by two different mechanistic pathways.
those obtained starting from 1,8-enyne 16.[11] Addi- The formation of deconjugated cycloheptadiene 17 as
tionally, the 17:18 ratio remains unchanged when the the major product of the reaction is however difficult
mixture is re-treated with catalyst 20, indicating that to rationalise, more especially when contrasted with
deconjugated diene 17 is not converted into conjugat- the reaction of 1,6-enyne 12, which exclusively fur-
ed diene 18.[12] Cyclobutene 19 also proved to be nishes conjugated diene 15 via a postulated carbocat-
stable in solution in the absence of gold catalyst 20, ionic intermediate 14 that is similar to 21.[13]
thus excluding its conversion into conjugated diene 18
by conrotatory ring opening. We have hence estab- presence of the bicyclo
Given the novelty of this transformation and the
[5.2.0]nonane framework in
AHCTUNGTRENNUNG
lished the intermediacy of cyclobutene 19 in the for- the skeleton of a number of natural products, we de-
mation of cycloheptadienes 17 and 18, which could cided to study the scope of the [2+2]cycloaddition.
arise through carbocation 21 following a gold-cata- As seen from the results compiled in Table 1, the re-
Table 1. Scope of the Au(I)-catalysed [2+2]cycloaddition.
Entry
1
Substrate
Product
t
Yield [%][a]
90
22a
23a
0.25 h
2
3
22b
22c
23b
23c
3 h
83
24 h
0[b]
4
22d
23d
10 h
80
5
6
22e
22f
R5 =Me
23e
23f
8.5 h
7.5 h
57 (82)[c]
66
R5 =n-C5H11
Adv. Synth. Catal. 2009, 351, 379 – 386
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
381

Yann Odabachian and Fabien Gagosz
COMMUNICATIONS
Table 1. (Continued)
Entry
Substrate
Product
t
Yield [%][a]
7
8
22g
22h
R4 =H, R5 =Me
23g
23h
13.5 h
2 h
72
62
4
5
À
R
R =-(CH2)3-
9
22i
22j
22k
22l
R4 =H, R5 =Me
23i
23j
23k
23l
0.25 h
1.5 h
3 h
85
78
77
85
10
11
12
R4 =H,R5 =n-C5H11
4
4
5
À
R
R
R =-(CH2)3-
5
À
R =-(CH2)4-
5 min
13
14
15
22m
22n
23m
23n
10 min
22 h
86
62[b]
22o
22p
23o
23p
16 h
65
1 h
6 h
49[d] (1:1)[f]
41[e] (1:1)[f]
16
[a]
Isolated yields.
[b]
[c]
[d]
[e]
[f]
Under reflux.
NMR spectroscopic yield.
In refluxing 1,2-dichloroethane.
With [(2,4-di-t-BuPhO)3PAu]SbF6 as the catalyst (4 mol%).
Diastereoisomeric ratio.
action proved to be quite general and various 1,8- tries 9–13) were for instance rapidly formed and did
enynes 22a–p reacted using 4 mol% of gold complex not suffer subsequent transformation.[16] In contrast,
20 to give cyclobutenes 23a–p in generally good yields the reaction was very slow for substrates 22n–o bear-
(41–90%).[14]
ing a simple allyl moiety. A highly competitive iso-
The transformation could be performed with sub- merisation process took place in this case, that ulti-
strates possessing diversely substituted alkenes, with mately led to the formation of cyclobutene 23n–o in
the exception of 1,8-enyne 22c, bearing a neryl chain which the C=C bond of the cyclobutene had migrated
which, surprisingly, did not react (entry 3).[15] The at the ring junction. Finally, even if it makes the trans-
esters at C(4) could be replaced by diacetoxymethyl formation easier, the presence of an unsaturation con-
groups whithout loss of efficiency (entry 1) and other jugated with the alkyne is not essential. Substrate 22p
acyclic (entries 6, 10), carbocyclic (entries 8, 11, 12) or could for instance be transformed into cyclobutene
aromatic (entries 13, 15) susbtituents were tolerated 23p using 4 mol% of catalyst 20 in refluxing 1,2-DCE
at postions C(6) and/or C(7) of the 1,8-enyne. Nota- (49%) or 4 mol% of the more electrophilic [(2,4-di-t-
bly, the time necessary to reach a maximum yield of BuPhO)3PAu]SbF6 catalyst[6] at room temperature
the cylobutene was generally dependant on the nucle- (41%) (entry 16).
ophilicity of the alkene, while the yield was function
Certain 1,8-enynes could be ultimately transformed
of the presence of possible competitive reactions af- into metathesis-type derivatives by gold-catalysed
fecting the cyclobutene ring. Cyclobutenes 23i–m de- fragmentation of the intermediate cyclobutene, as
rived from enynes possessing a methallyl group (en- previously observed for enyne 16. For example,
382
asc.wiley-vch.de
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 379 – 386

Cyclobutenes as Isolable Intermediates in the Gold(I)-Catalysed Cycloisomerisation
COMMUNICATIONS
Scheme 4. Examples of cyclobutene ring fragmentation.
enynes 22e and 22g furnished cycloheptadienes 24
For substrates bearing a geranyl chain, the [2+2]
and 25 as single isomers in, respectively, 96% and cycloaddition could be followed by an impressive
81% yield [Scheme 4, Eq. (8)].[17] Thus, cyclobutenes gold-catalysed ene reaction that stereoselectively fur-
23f and 23h, possessing the same substitution pattern nished tetracylic compound 38 in an excellent 90%
as 23e and 23g, were stable in CD2Cl2 but were slowly yield in the case of enyne 36 and a mixture of com-
rearranged into 26 and 27 by traces of HCl contained pounds 40 and 41 in the case of enyne 39
in CDCl3 [Scheme 4, Eq. (9)]. A remarkable sequence (Scheme 6).[2c,18]
of [2+2]cycloaddition/ring fragmentation/Friedel–
A mechanistic proposal for the [2+2] cycloaddition
Crafts reaction/diene isomerisation/Wagner–Meer- of 1,8-enynes is presented in Scheme 7. The electro-
wein rearrangement was observed in the case of philic gold(I) activation of the alkyne in substrate 42
enyne 28 (Scheme 5).[18] The fragmentation of cyclo- might lead to the formation of a cyclopropyl gold car-
butene 29 produced the carbocationic species 30, bene (or gold-stabilised cyclopropylmethyl carbocat-
which was trapped by the pendant phenyl group. The ion)[20] 44. A subsequent ring expansion could then
resulting tricyclic compound 31 was subsequently iso- furnish the cyclobutyl carbocation 45 which would
merised into naphthalene derivative 33.[19] A similar evolve into cyclobutene gold complex 46. A final sub-
result was observed in the case of substrate 34.
stitution of the coordinated cyclobutene 47 by the
Scheme 5. Multiple-steps Au(I)-catalysed isomerisation of 28 and 34.
Adv. Synth. Catal. 2009, 351, 379 – 386
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
383

Yann Odabachian and Fabien Gagosz
COMMUNICATIONS
Scheme 6. Multiple-steps Au(I)-catalysed isomerisation of 36 and 39.
Scheme 7. Mechanistic proposal.
substrate would complete the cycle.[21]. The ring ex-
In summary, it was found that a variety of function-
pansion leading to 45 might be more favoured in the alized bicycloACTHNUTRGENUG[N 5.2.0]nonenes could be accessed by a
present case than in the cycloisomerisation of 1,6- gold(I)-catalysed [2+2]cycloaddition of 1,8-enynes.
enynes (Scheme 2) due to a reduction of geometric These cyclobutenes could be isolated as intermediates
constraints (7- versus 5-membered cycle) and a possi- that could suffer subsequent gold(I)-catalysed trans-
ble greater stabilisation of the resulting cationic formations such as isomerisation, fragmentation or
charge by the unsaturation between C(6) and C(7).
ene reaction to furnish more structurally complex
Even if the intermediacy of 44 could not be experi- products. This study provides useful information relat-
mentally proved,[22] it is nevertheless supported by the ed to the mechanism leading to metathesis-type deriv-
formation of conjugated diene 51 as a minor product atives, which were shown to be produced, in the pres-
in the cycloisomerization of enyne 22m (Scheme 8). ent case, by a gold(I)-catalysed ring fragmentation of
Cyclobutene 23m proved to be stable in CH2Cl2 in the cyclobutene moiety. It also highlights the diversity
the presence of catalyst 20 or HNTf2, thus excluding of products which can be formed from a single mech-
the formation of 51 by ring fragmentation. This latter anistic class depending on the relative stability of the
might, however, be competitively formed from a intermediates involved.
common cyclopropyl intermediate 52 by a rearrange-
ment similar to that proposed for the isomerisation of
1,6-enynes (Scheme 2).
384
asc.wiley-vch.de
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 379 – 386

Cyclobutenes as Isolable Intermediates in the Gold(I)-Catalysed Cycloisomerisation
COMMUNICATIONS
Scheme 8. Experimental result supporting the formation of a gold-stabilised cyclopropyl methyl carbocation intermediate.
Experimental Section
Kotsuma, S. Murai, J. Am. Chem. Soc. 2002, 124,
10294–10295; c) A. Fürstner, F. Stelzer, H. Szillat, J.
Am. Chem. Soc. 2001, 123, 11863–11869; d) C. Nieto-
Oberhuber, S. López, M. Paz MuÇoz, D. J. Cµrdenas, E.
BuÇuel, C. Nevado, A. M. Echavarren, Angew. Chem.
2005, 117, 6302–6304; Angew. Chem. Int. Ed. 2005, 44,
6146–6148.
Typical Procedure
To a solution of 16 (13.9 mg, 0.05 mmol) in 0.5 mL of
CD2Cl2 in an NMR tube was added 4 mol% of gold catalyst
20 (2 mg, 0.002 mmol). The reaction was monitored by
1H NMR spectroscopy. After 0.25 h, the reaction was
quenched by the addition of Et3N (0.1 mL), the mixture
concentrated under vacuum and the residue purified by
flash chromatography (SiO2, petroleum ether/Et2O=95/5)
to afford product 19: yield: 12 mg (86%).
[4] a) B. M. Trost, G. J. Tanoury, J. Am. Chem. Soc. 1988,
110, 1636–1638; b) B. M. Trost, M. K. Trost, Tetrahe-
dron Lett. 1991, 32, 3647–3650; c) A. Fürstner, P. W.
Davies, T. Gress, J. Am. Chem. Soc. 2005, 127, 8244–
8245; d) G. B. Bajracharya, I. Nakamura, Y. Yamamoto,
J. Org. Chem. 2005, 70, 892–897; e) F. Marion, J. Cou-
lomb, C. Courillon, L. Fensterbank, M. Malacria, Org.
Lett. 2004, 6, 1509–1511; f) F. Marion, J. Coulomb, A.
Servais, C. Courillon, L. Fensterbank, M. Malacria, Tet-
rahedron 2006, 62, 3856–3871; g) T. Matsuda, S. Kado-
waki, T. Goya, M. Murakami, Synlett 2006, 575–578;
h) C. Nieto-Oberhuber, S. López, A. M. Echavarren, J.
Am. Chem. Soc. 2005, 127, 6178–61795; i) S. Couty, C.
Meyer, J. Cossy, Angew. Chem. 2006, 118, 6878–6882;
Angew. Chem. Int. Ed. 2006, 45, 6726–6730; j) C. H.
Oh, A. Kim Synlett 2008, 777–781; k) A. Fürstner, A.
Schlecker, C. W. Lehmann, Chem. Commun. 2007,
4277–4279.
[5] a) A. Fürstner, H. Szillat, F. Stelzer, J. Am. Chem. Soc.
2000, 122, 6785–6786; b) E. M. Simmons, R. Sarpong,
Org. Lett. 2006, 8, 2883–2886; c) J. Blum, H. Beer-
Kraft, Y. Bradieh, J. Org. Chem. 1995, 60, 5567–5569;
see also Refs.[3b,c,4a]
[6] For DFT calculations on the skeletal rearrangement
and the [2+2]cycloaddition of 1,6-enynes, see
Refs.[2a,3d] See also: C. Nieto-Oberhuber, P. Pꢂrez-
Galµn, E. Herrero-Gómez, T. Lauterbach, C. Rodri-
guez, S. López, C. Bour, A. Rosellón, D. J. Cµrdenas,
A. M. Echavarren, J. Am. Chem. Soc. 2008, 130, 269–
279.
Acknowledgements
The authors thank F. Istrate for preliminary studies, Prof. S.
Z. Zard for helpful discussions and Rhodia Chimie Fine for
a generous gift of HNTf2.
References
[1] For recent reviews on the cycloisomerisation of 1,n-
enynes, see: a) V. Michelet, P. Y. Toullec, J. P. GenÞt,
Angew. Chem. 2008, 120, 4338–4386; Angew. Chem.
Int. Ed. 2008, 47, 4268–4315; b) E. Jimꢂnez-NfflÇez,
A. M. Echavarren, Chem. Rev. 2008, 108, 3326–3350;
c) L. Zhang, J. Sun, S. A. Kozmin, Adv. Synth. Catal.
2006, 348, 2271–2296; d) C. Bruneau, Angew. Chem.
2005, 117, 2380–2386; Angew. Chem. Int. Ed. 2005, 44,
2328–2334.
[2] For a review on the [2+2]cycloaddition of enynes, see:
a) C. Nieto-Oberhuber, S. López, E. Jimꢂnez-NfflÇez,
A. M. Echavarren, Chem. Eur. J. 2006, 12, 5916–5923.
For a selection of recent metal-mediated syntheses of
cyclobutenes, see: b) H. Xu, W. Zhang, D. Shu, J. B.
Werness, W. Tang, Angew. Chem. 2008, 120, 9065–
9068; Angew. Chem. Int. Ed. 2008, 47, 8933–8936;
c) O. Debleds, J.-M. Campagne, J. Am. Chem. Soc.
2008, 130, 1562–1563; d) A. Fürstner, C. Aïssa, J. Am.
Chem. Soc. 2006, 128, 6306–6307; e) Y. Liu, M. Liu, Z.
Song, J. Am. Chem. Soc. 2005, 127, 6662–6663.
[7] a) F. Gagosz, Org. Lett. 2005, 7, 4129–4132; b) A.
Buzas, F. Gagosz, J. Am. Chem. Soc. 2006, 128, 12614–
12615; c) A. K. Buzas, F. M. Istrate, F. Gagosz, Angew.
Chem. 2007, 119, 1159–1162; Angew. Chem. Int. Ed.
2007, 46, 1141–1144; d) S. Bçhringer, F. Gagosz, Adv.
Synth. Catal. 2008, 350, 2617–2630.
[8] N. Mezailles, L. Ricard, F. Gagosz, Org. Lett. 2005, 7,
4133–4136.
[3] a) B. M. Trost, M. Yanai, K. Hoogsteed, J. Am. Chem.
Soc. 1993, 115, 5294–5295; b) N. Chatani, H. Inoue, T.
Adv. Synth. Catal. 2009, 351, 379 – 386
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
385

Yann Odabachian and Fabien Gagosz
COMMUNICATIONS
[9] For recent reviews on gold catalysis, see: a) Z. Li, C.
Brower, C. He, Chem. Rev. 2008, 108, 3239–3245;
b) A. Arcadi, Chem. Rev. 2008, 108, 3266–3325; c) D. J.
Gorin, F. D. Toste, Chem. Rev. 2008, 108, 3351–3378;
d) A. S. K. Hashmi, M. Rudolph, Chem. Soc. Rev. 2008,
37, 1766–1775; e) A. S. K. Hashmi, Chem. Rev. 2007,
107, 3180–3211; f) A. Fürstner, P. W. Davies, Angew.
Chem. 2007, 119, 3478–3519; Angew. Chem. Int. Ed.
2007, 46, 3410–3449; g) E. Jimꢂnez-Nœnez, A. M.
Echavarren, Chem. Commun. 2007, 333–346.
[15] The lack of reactivity of 22c might be explained by a
steric interaction between Rcis and AuL in an inter-
mediate of type 45 (see Scheme 7).
[16] Compounds 23i–m were stable in the presence of gold
catalyst 20.
[17] Compounds 22e and 22g were stable in CD2Cl2 whitout
gold catalyst 20.
[18] See Supporting Information for a complete mechanistic
1
proposal based on H NMR monitoring.
[19] The formation of 29 and 31 was observed by H NMR
1
[10] HNTf2 and AgNTf2 did not catalyse the formation of
monitoring.
19.
[20] The exact nature of such an intermediate is not known.
For a discussion, see: A. Fürstner, L. Morency, Angew.
Chem. 2008, 120, 5108–5111; Angew. Chem. Int. Ed.
2008, 47, 5030–5033; see also: A. S. K. Hashmi, Angew.
Chem. 2008, 120, 6856–6858; Angew. Chem. Int. Ed.
2008, 47, 6754–6756.
[21] An alternative syn-cyclopropyl gold carbene undergo-
ing ring expansion could also be envisaged, see
Refs.[2a,3d]
[11] HNTf2 also catalysed the formation of 17 and 18 (1:1)
from 19.
[12] 1H NMR spectroscopy monitoring showed a gradual
disappearance of 19 to give 17 and 18 with a 17:18 ratio
remaining the same during the whole process. See Sup-
porting Information for deuterium-labelling experi-
ments supporting this mechanism.
[13] The involvement of a carbocationic species of type 21
is in agreement with the transformation of enynes 28
and 34 in which similar cationic species could partici-
pate in a Friedel–Crafts reaction (see Scheme 5).
[14] Reactions were performed in CD2Cl2 in order to moni-
tor the transformation by 1H NMR spectroscopy.
[22] All attempts to trap a gold carbene intermediate of
type 44 were unsuccessful (oxidation, cyclopropana-
tion).
386
asc.wiley-vch.de
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 379 – 386