Gold- And platinum-catalyzed cycloisomerization of enynyl esters versus alleneyl esters: An experimental and theoretical study

Article abstract of DOI:10.1002/chem.200801387

Experimental and theoretical studies on Au- and Pt-catalyzed cycloisomerization of a branched dienyne with an acetate group at the propargylic position are presented. The peculiar architecture of the dienyne precursor, which has both a 1,6- and a 1,5-enyne skeleton, leads, in the presence of alkynophilic gold catalysts, to mixtures of bicyclic compounds 3, 4, and 5. Formation of unprecedented bicyclo[3.1.0]hexene 5 is the main focus of this study. The effect of the ancillary ligand on the gold center was examined and found to be crucial for formation of 5. Further mechanistic studies, involving cyclization of an enantioenriched dienyne precursor, ¹⁸O-labeling experiments, and DFT calculations, allowed an unprecedented reaction pathway to be proposed. We show that bicyclo[3.1.0]hexene 5 is likely formed by a 1,3-OAc shift/allene-ene cyclization/ 1,2-OAc shift sequence, as calculated by DFT and supported by Au-catalyzed cyclization of isolated allenenyl acetate 7, which leads to improved selectivity in the formation of 5. Additionally, the possibility of OAc migration from allenyl acetates was supported by a trapping experiment with styrene that afforded the corresponding cyclopropane derivative. This unprecedented generation of a vinyl metal carbene from an allenyl ester supports a facile enynyl ester/allenenyl ester equilibrium. Further examination of the difference in reactivity between enynyl acetates and their corresponding [3,3]-rearranged allenenyl acetates toward Au- and Pt-catalyzed cycloisomerization is also presented.

Full text of DOI:10.1002/chem.200801387

FULL PAPER
DOI: 10.1002/chem.200801387
Gold- and Platinum-Catalyzed Cycloisomerization of Enynyl Esters versus
Allenenyl Esters: An Experimental and Theoretical Study
Nicolas Marion,[a, f] Gilles Lemiꢀre,[b] Andrea Correa,[c] Chiara Costabile,[c]
Rubꢁn S. Ramꢂn,[a, g] Xavier Moreau,[b] Pierre de Frꢁmont,[d] Rim Dahmane,[b]
Alexandra Hours,[b] Denis Lesage,[e] Jean-Claude Tabet,[e] Jean-Philippe Goddard,[b]
Vincent Gandon,[b] Luigi Cavallo,*[c] Louis Fensterbank,*[b] Max Malacria,*[b] and
Steven P. Nolan*[a, d, g]
Abstract: Experimental and theoretical
studies on Au- and Pt-catalyzed cyclo-
ACHTUNGTRENNUNGisomerization of a branched dienyne
with an acetate group at the propargyl-
ic position are presented. The peculiar
architecture of the dienyne precursor,
which has both a 1,6- and a 1,5-enyne
skeleton, leads, in the presence of alky-
nophilic gold catalysts, to mixtures of
bicyclic compounds 3, 4, and 5. Forma-
cyclization of an enantioenriched dien-
yne precursor, 18O-labeling experi-
ments, and DFT calculations, allowed
an unprecedented reaction pathway to
be proposed. We show that bicyclo-
in the formation of 5. Additionally, the
possibility of OAc migration from al-
lenyl acetates was supported by a trap-
ping experiment with styrene that af-
forded the corresponding cyclopropane
derivative. This unprecedented genera-
tion of a vinyl metal carbene from an
allenyl ester supports a facile enynyl
ester/allenenyl ester equilibrium. Fur-
ther examination of the difference in
reactivity between enynyl acetates and
their corresponding [3,3]-rearranged al-
lenenyl acetates toward Au- and Pt-cat-
alyzed cycloisomerization is also pre-
sented.
AHCTUNGTRENGN[UN 3.1.0]hexene 5 is likely formed by a
1,3-OAc shift/allene–ene cyclization/
1,2-OAc shift sequence, as calculated
by DFT and supported by Au-catalyzed
cyclization of isolated allenenyl acetate
7, which leads to improved selectivity
tion
of
unprecedented
bicyclo-
ACHTUNGTRENNUNG[3.1.0]hexene 5 is the main focus of this
study. The effect of the ancillary ligand
on the gold center was examined and
found to be crucial for formation of 5.
Further mechanistic studies, involving
Keywords: alkynes · allenes · cyclo-
isomerization · density functional
calculations · gold
[a] Dr. N. Marion, R. S. Ramꢀn, Prof. Dr. S. P. Nolan
[d] Dr. P. de Frꢃmont, Prof. Dr. S. P. Nolan
Institute of Chemical Research of Catalonia (ICIQ)
Av. Paꢁsos Catalans 16, 43007 Tarragona (Spain)
Fax : (+34)977-920-224
Department of Chemistry, University of New Orleans
2000 Lakeshore Drive, New Orleans, LA 70148 (USA)
[e] D. Lesage, Prof. Dr. J.-C. Tabet
E-mail: snolan@iciq.es
Laboratoire de Chimie Structurale Organique et Biologique
UMR CNRS 7613, Universitꢃ Pierre et Marie Curie (UPMC)
B. 229, 4 place Jussieu, 75005 Paris (France)
sn17@st-andrews.ac.uk
[b] G. Lemiꢂre, Dr. X. Moreau, R. Dahmane, Dr. A. Hours,
Dr. J.-P. Goddard, Dr. V. Gandon, Prof. Dr. L. Fensterbank,
Prof. Dr. M. Malacria
[f] Dr. N. Marion
Present address: Department of Chemistry
Massachusetts Institute of Technology
77 Massachusetts Avenue, Room 18-343, Cambridge, MA 02139
(USA)
Laboratoire de Chimie Organique, UMR CNRS 7611-FR2769
UPMC—Univ. Paris 06
B. 229, 4 place Jussieu, 75005 Paris (France)
E-mail: louis.fensterbank@upmc.fr
[g] R. S. Ramꢀn, Prof. Dr. S. P. Nolan
max.malacria@upmc.fr
Present address: School of Chemistry, University of St-Andrews
St-Andrews KY16 9ST (UK)
[c] A. Correa, C. Costabile, Prof. Dr. L. Cavallo
Department of Chemistry, University of Salerno
via Ponte don Melillo, Fisciano, 84084 (Italy)
E-mail: lcavallo@unisa.it
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200801387.
Chem. Eur. J. 2009, 15, 3243 – 3260
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3243

Introduction
Results and Discussion
Atom-economic reactions, as defined by Trost,[1] are of pri-
mary importance in modern organic chemistry. In this field,
polyunsaturated substrates have become ubiquitous and
have led to the discovery of a number of unprecedented
transformations in the presence of late transition-metal cata-
lysts.[2] Among available atom-economic transformations,[3]
the cycloisomerization of enynes remains unique because of
the increased molecular complexity achieved in one chemi-
cal step.[4] After extensive studies with palladium[5] and plati-
Access to extremely elaborate structural motifs, such as
fused polycyclic systems from acyclic substrates, is often de-
pendant on subtle modifications in the precursor architec-
ture. In this context, we have reported on the importance of
oxo substituents—and of their nature—at the propargylic
position in enyne systems.[11] Furthermore, we recently in-
vestigated the reactivity of dienyne 2, which formally exhib-
its 1,5- and 1,6-enyne scaffolds, and observed different selec-
tivity as a function of the nature of the catalyst employed
(Scheme 2).[12]
num,[6] gold(I) and gold
ACTHNUTRGNEU(GN III) salts have emerged lately as
powerful catalysts for a myriad of transformations involving
enynes.[7,8] In this context, enyne derivatives with an ester
group at the propargylic position can be regarded as a spe-
cial class of substrates due to their atypical reactivity pat-
tern.[9,10] More specifically, an appropriately situated O-ester
group undergoes an internal 1,2- or 1,3-shift upon electro-
ꢀ
philic activation of the C C bond (Scheme 1, I) to give rear-
ranged products II and III respectively, which can then fur-
ther evolve as a function of the remaining pendant groups
(Scheme 1).
Scheme 2. Pt- and Au-catalyzed cycloisomerization of 2. [a] Reaction
conditions: PtCl2 (5 mol%), toluene (0.025m), 808C. [b] Reaction condi-
tions: [(IPr)AuCl]/AgBF4 (2 mol%), CH2Cl2 (0.1m), RT.
Thus, formation of bicyclo
ACHTUNGTRENNUNG
A
4
used,[11a] while various gold catalysts afforded different prod-
uct ratios. More interestingly, the use of AuI salts in lieu of
PtII ones led to the formation of unexpected bicyclo-
AHCTUNGTRENNUNG
a particular set of steric and electronic parameters favoring
formation of compound 5, and the ability of N-heterocyclic
carbenes (NHCs)[13,14] to promote this unexpected transfor-
mation. Therefore, intrigued by this unprecedented mode of
cyclization, we thoroughly examined the reaction parame-
ters that might influence the formation of this type of carbo-
cycle. Moreover, we carried out 18O-labeling experiments
and theoretical calculations to gain further insight into the
mechanism of this cyclization. Herein we disclose the results
of these studies, which permit a new type of reactivity in the
enynyl ester series to be explained. Thus, we have shown
that the most likely pathway to 5 involves, after formation
of an allenyl ester, carbocyclization of an allenene core fol-
lowed by “retro-migration” of the ester moiety. The retro-
migration process was further studied on an extended set of
substrates.
Influence of reaction conditions: We began our study on
model substrate 2 by evaluating the influence of the ligand
using a series of [(NHC)AuCl] complexes[15] leading to the
formation of bicyclic product 5 (Table 1).[16] Overall, the
combined yields of cycloisomerized products range from
good to excellent with a slight decrease of activity for the
less encumbered ligands (Table 1, entries 14–16 and 21).
Scheme 1. Reactivity of propargylic acetates in the presence of gold cata-
lysts.
The wide diversity and complexity of accessible structures
from readily assembled precursors is probably the most ap-
pealing feature of enyne cycloisomerization reactions. How-
ever, it can also be seen as a disadvantage, since small
changes in the starting compound can lead to unexpected
outcomes, so that the reactivity of a specific substrate is
often rendered unpredictable. This is why insights into the
skeletal rearrangements of enynes catalyzed by gold salts
are still needed in order to understand the mechanism of
these transformations. Here we present our experimental
and theoretical findings on platinum- and gold-catalyzed cy-
cloisomerization of enynes bearing a propargylic acetate
group. Driven by theoretical findings, we focused our studies
on the reversibility of OAc migration between the propar-
gylic esters and their corresponding [3,3]-rearranged coun-
terparts, namely, the allenyl esters.
Simple “ligandless” gold(I) and gold
ACHTUNGERTN(NUNG III) chloride salts led
to the same selectivity,[17] yielded 3 and 4 in similar propor-
3244
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3243 – 3260

Cycloisomerization of Enynyl Esters versus Allenenyl Esters
FULL PAPER
Table 1. Effect of ligands on the cycloisomerization of 2.
Entry
[M]
3:4:5[a]
Total
Entry
[M]
3:4:5[a]
Total
yield [%][a]
yield [%][a]
1[b]
2[c]
3
PtCl2
PtCl2/AgBF4
AuCl
AuCl/AgBF4
AuCl3
AuCl3/AgBF4
1/0.1/0
1/0.1/0
1/0.7/0.2
1/0.5/0
1/0.8/0.2
1/0.3/0
1/0.8/0.1
1/0.1/0
1/0.5/1.5
1/0.4/1.7
1/0.4/1.4
98
33
86
84
88
77
80
81
78
72
84
12
N
1/0.3/0.9
1/0.5/0.4
1/0.1/0.2
1/0.2/0.1
1/0.2/0.3
1/0.6/0.4
1/0.1/0.6
1/0.1/0.7
1/0.2/0.5
1/0.1/0.2
1/0.6/0.1
90
94
62
70
63
88
73
95
76
64
54
13
14[g]
15
4
5[d]
6[e]
7[f]
8[d]
9
16[g]
17
T
18
19
N
N
20
10
11
T
21[g]
22
A
1
[a] Yields of isolated products, average of two runs, products ratio determined by H NMR spectroscopy. [b] Reaction performed with PtCl2 (5 mol%) in
toluene at 808C for 2 h. [c] Reaction performed with PtCl2 (5 mol%)/AgBF4 (10 mol%) in toluene at 808C for 2 h. 26% of 2 was recovered along with
8% of allenyl ester 7, and significant amounts of oligomerized byproducts were also formed. [d] Reaction stirred for 2 h. [e] Reaction performed with
AuCl3 (2 mol%)/AgBF4 (6 mol%) for 1 h. [f] Reaction stirred for 2 h, 8% of allenyl ester 7 was also formed. [g] No starting material remaining; signifi-
ACHTUNGTRENNUNGcant amounts of oligomerized byproducts were formed.
tion, and only gave small amounts of 5 (Table 1, entries 3
and 5). Interestingly, when these salts were employed in
combination with a silver(I) salt, formation of 5 was not ob-
served, and the 3:4 ratio increased (Table 1, entries 4 and 6)
and approached the selectivity displayed by PtCl2 (Table 1,
entry 1). Notably, platinum(II) chloride in conjunction with
2 equiv AgBF4 displayed poorer activity than PtCl2 alone.
The last “simple” and commercially available gold salt
tested was [(Me2S)AuCl] (Table 1, entries 7 and 8). As ex-
pected, it could be used without silver salt (Table 1, entry 7)
and, under these conditions, behaved as AuCl, supporting
decoordination of Me2S from the gold center prior to cata-
lytic activity.
On the other hand, the addition of AgBF4 to
[(Me2S)AuCl] led to a selectivity comparable with those of
AuCl/AgBF4 and AuCl3/AgBF4, whereby formation of 5 was
not observed (Table 1, entry 8). This can be interpreted in
two ways: either Me2S as ligand has little influence on the
course of the cycloisomerization or the catalytically active
species is simply the gold cation, formed upon chloride ab-
straction and Me2S decoordination.
If we examine the influence of the ancillary ligand on the
gold center by considering the formation of 5 as the main
parameter, the ligands employed here can be divided into
three classes. Those displaying low steric hindrance fur-
nished only minor amounts of bicycle 5 (Table 1, entries 13–
17 and 21; 3/5 1/<0.4), whereas highly hindered ligands led
to slightly higher 3/5 ratios (ca. 1/0.6; Table 1, entries 18–
20). The third class of ligands can be regarded as intermedi-
ate in terms of steric hindrance between the other two
classes.[18] It includes the widely used IPr and IMes ligands
and their saturated analogues SIPr and SIMes. Only these li-
gands produced mixtures in which 5 was the major com-
pound (Table 1, entries 9–12). Considering the minor differ-
ences in the electronic properties of the NHCs employed
here,[18] we believe that formation of unexpected bicyclo-
AHCTUNGTREG[NNNU 3.1.0]hexene 5 is mainly under steric control and responds
favorably to a medium/high steric demand.
Chem. Eur. J. 2009, 15, 3243 – 3260
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3245

S. P. Nolan, M. Malacria, L. Fensterbank, L. Cavallo et al.
Table 2. Au- and Pt-catalyzed cycloisomerization of enantioenriched 2.
Since prolonged reaction time, under strictly similar reac-
tion conditions, led to comparable ratios of cyclopropyl car-
bocycles, the possibility of chemical equilibrium between the
different products is excluded.
Subsequently, we attempted a trapping experiment using
methanol as solvent in order to possibly isolate relevant in-
termediates that would have allowed for a better under-
standing of the formation of product 5. Under these condi-
tions, enol ether 6, which results from addition of methanol
to the alkyne moiety, was the only observed product
[Eq. (1)].[19]
Entry
2
Conditions
Yield [%] (%ee)
4[a] 5[a]
3[a]
1
2
3
4
R=Ac, PtCl2 (5 mol%) toluene,
66% ee RT!808C, 2 h
88 (38) 8 (rac)
83 (42) 7 (rac)
78 (20) 6 (rac)
–
R=Ac, PtCl2 (5 mol%) toluene,
66% ee RT, 2 h
–
–
R=Piv, PtCl2 (5 mol%) toluene,
76% ee 808C, 48 h
R=Ac, AuCl3 (2 mol%) CH2Cl2, 46 (8)
60% ee RT, 10 min
36 (rac) 5 (4)
[a] Yields of isolated 3/4/5 mixtures, ratios, and ee values determined by
chiral GC. rac=racemic.
logue of 2, led to a further decrease of the ee of 3-Piv
Sequential acetate migration/cyclopropanation: A key fea-
ture of the reactivity of propargylic acetates in the presence
of soft p Lewis acidic metals such as gold and platinum lies
in the ability of the acetate moiety to act as an internal nu-
(Table 2, entry 3 versus entry 1).
Finally, using goldACHTNUGRTENUNG(III) chloride afforded 3 and 5 with
very low ee values and 4 as a racemic mixture (Table 2,
entry 4), and this suggests that with gold catalysts both se-
quences are likely to compete in the formation of all cyclo-
propanated products, with preference given to the migration
then cyclopropanation pathway.
At this point of our study, several mechanistic scenarios
were still to be considered,[23] including, in addition to the
aforementioned two sequential pathways, carbocyclization
ꢀ
cleophile that migrates in a 1,2 or 1,3 fashion onto the C C
bond. In the context of enyne cycloisomerization, this mi-
grating ability of the acetate group has raised questions
about the order of the migration/cyclopropanation se-
quence.[20,21] In our preliminary study,[12] we proposed a mi-
gration then cyclopropanation sequence, which appeared
ꢁ
more consistent with the formation of bicyclo
[3.1.0]hexene
onto the oxonium intermediate before rupture of the C O
5. Nevertheless, at the time we did not exclude the possibili-
ty that a cyclopropanation then migration sequence could be
involved in the formation of bicyclic compounds 3 and 4.
To gain insights into this crucial issue, we synthesized
enantioenriched dienyne 2 and subjected it to cycloisomeri-
zation conditions. In the case of cyclopropanation followed
by migration, some chiral information should be translated
into the product, whereas in the case of the opposite se-
quence, which involves an achiral intermediate of type II
(see Scheme 1), chirality transfer would be precluded.[22]
Enantioenriched 2, in the presence of 5 mol% PtCl2,
yielded rearranged products 3 and 4 in comparable yields
and ratio to those previously observed but with distinct
enantioselectivities (Table 2, entries 1 and 2). Cyclohexene
derivative 3 was formed with 40% ee, while cyclopentene 4
was found to be racemic. Starting from a 66% ee-enriched
dienyne, this could indicate that in the formation of 3, the
initial cyclopropanation would not be completely stereose-
lective[22b] or that both sequences are at play with a prefer-
ence for the scenario of cyclopropanation then migration.
On the contrary, the formation of 4 would appear to primar-
ily result from a migration then cyclopropanation sequence,
leading to complete loss of chirality transfer. However, in
this case, we cannot also exclude a totally nonselective ini-
tial cyclopropanation step. Additionally, the nature of the
migrating ester group seems to play an important role in the
choice of the sequence, since using 2-Piv, the pivalate ana-
bond to regenerate the ester functionality. This possibility,
first proposed by Fehr and Galindo,[22b] implies that the ste-
reochemical information, still present in the oxocarbenium
intermediate, is transferred in the cyclization step.
The additional information gathered so far on the cycliza-
tion of 2 and the formation of 3, 4, and 5 suggests competing
pathways, the balance of which can be shifted by means of
the ancillary ligand. Greater chirality transfer is observed
with platinum(II) chloride compared to gold catalysis.
Nevertheless, a rationale for the formation of the unpre-
cedented cyclopropyl derivative 5 remained elusive, notably
because of little precedent in the literature.[12,24] We there-
fore performed DFT calculations on a full catalytic cycle for
the formation of 3–5, focusing on 5.
DFT calculations on the formation of 3–5: For the sake of
clarity, we numbered the atoms in dienyne 2 and we labeled
structures from the DFT calculations with lowercase bold
letters. The only exceptions are starting dienyne 2, allene 7,
and bicyclic products 3, 4 and 5, all of which were character-
ized experimentally. For these species we use labels such as
2Au, 4Au, and so on, to indicate that these substrates are
coordinated to Au.
Since N-heterocyclic carbene ligands are necessary in
order to obtain satisfactory amounts of unprecedented car-
bocycle 5, we performed the calculations with IDM bound
to the gold(I) center (IDM=1,3-dimethylimidazol-2-yli-
3246
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3243 – 3260

Cycloisomerization of Enynyl Esters versus Allenenyl Esters
FULL PAPER
in a first approach (see Figures S1–S3, Supporting Informa-
tion). Furthermore, in order to give a general and more real-
istic picture of the mechanistic processes at play, which are
interdependent, we have gathered in one single scheme
every pathway considered here.
ꢁ
The catalytic cycle starts with displacement of a BF4
ꢁ
counterion from [(IDM)Au+]BF4 species a by the alkyne
group of 2 to furnish intermediate 2Au.[25] In CH2Cl2 dis-
ꢁ
dene); therefore, in all schemes and figures regarding DFT
calculations, the label “Au” in fact represents “[(IDM)Au]”.
placement of the BF4 counterion by the substrate is exer-
G
gonic by 26 kJmolꢁ1. Structure 2Au is then the branching
For the readerꢅs convenience, the different reaction path-
ways, which are interconnected, are discussed independently
and compared when relevant. Accordingly, a figure depic-
point for three different reaction paths (Scheme 3). The first
ꢁ
consists of nucleophilic attack of the C6 C7 double bond on
the C1 atom[26] (see Scheme 3 and Figure S1, Supporting In-
formation). This cyclopropanation step leads to intermediate
b, which already has the bicyclic skeleton of product 3. The
ACHTUNGTRENNUNG
Scheme 3. Possible mechanistic pathways based on calculations (energies given in kJmolꢁ1
)
Chem. Eur. J. 2009, 15, 3243 – 3260
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3247

S. P. Nolan, M. Malacria, L. Fensterbank, L. Cavallo et al.
Au carbene intermediate b is 60 kJmolꢁ1 lower in energy
than the starting alkyne-coordinated intermediate 2Au, and
is reached through transition state 2Au-b, with a rather low
energy barrier of 32 kJmolꢁ1 (Scheme 3 and Figure S1, Sup-
porting Information). Formation of 3 from b involves a 1,2-
shift of the ester group from C3 to C2. After a first step
leading to formation of intermediate c via transition state b-
c with a barrier of 25 kJmolꢁ1, the 1,2-shift is then complet-
ed via transition state c-3Au with a barrier of 13 kJmolꢁ1,
and finally affords the coordinated product 3Au. Both inter-
mediate c and 3Au are lower in energy than intermediate
2Au (by 82 and 147 kJmolꢁ1, respectively), and thus forma-
tion of 3Au is a substantially downhill path from the starting
alkyne-coordinated species 2Au to the coordinated product
species 3Au. Product release from 3Au assisted by a coordi-
26 kJmolꢁ1 lower in energy than starting alkyne-coordinated
intermediate 2Au, and is obtained via transition state 2Au-f
with a rather low energy barrier of 36 kJmolꢁ1.
Transition state 2Au-f is slightly higher in energy than the
similar transition state 2Au-b because of the higher steric
strain associated with formation of a more strained five-
membered ring in 2Au-f, compared to formation of a 6-
membered ring in 2Au-b. Formation of 4 from f then in-
volves 1,2-shift of the ester group from C3 to C2. The first
step of this 1,2-shift is the formation of intermediate g
through transition state f-g with a barrier of 14 kJmolꢁ1.
Final formation of coordinated product 4Au occurs through
transition state g-4Au with an energy barrier of 16 kJmolꢁ1.
Both intermediates g and 4Au are lower in energy than in-
termediate 2Au (by 61 and 127 kJmolꢁ1, respectively), and
thus formation of 4Au, similar to that of 3Au, is a substan-
tially downhill path from the starting alkyne-coordinated re-
actant 2Au to the coordinated product 4Au (Figure S2, Sup-
nating BF4 counterion is endergonic by 6 kJmolꢁ1, leads to
ꢁ
product 3, and closes the catalytic cycle by forming the start-
ing [(IDM)Au]BF4 species a. However, product 3 can be
reached from 2Au through an alternative reaction pathway.
Indeed, as mentioned before, in addition to the cyclopropa-
nation then migration sequence that was just examined, the
migration then cyclopropanation sequence must be consid-
ered. Thus, starting from alkyne-coordinated species 2Au, a
1,2-shift of the ester group corresponding to attack of O16
on C2 leads to intermediate d through transition state 2Au-
d with an energy barrier of 18 kJmolꢁ1 (Scheme 3). Inter-
mediate d, which is 35 kJmolꢁ1 lower in energy than 2Au, is
another branching point in this complex reaction manifold.
porting Information). Product release from 4Au assisted by
ꢁ
a
coordinating BF4
counterion is endergonic by
13 kJ.molꢁ1, leads to product 4, and closes the catalytic cycle
by forming the starting [(IDM)Au][BF4] species a.
AHCTUNGTRENNUNG
Like 3, 4 can be reached from 2Au through an alternative
path that involves initial 1,2-shift of the ester group. As dis-
cussed above, the branching point is intermediate d (see
ꢁ
Scheme 3). In fact, breaking of the C3 O13 bond can alter-
natively lead to intermediate (Z)e, with
a
ꢁ
barrier of
ꢁ1
33 kJmol , which, upon attack of the C11 C12 double
bond on C1, leads to coordinated product species 4Au via
transition state (Z)e-4Au with an energy barrier of
ꢁ
The branch that leads to 3 involves breaking of the C3 O13
bond to furnish intermediate (E)e, with
a
barrier of
34 kJmolꢁ1, and a subsequent cyclopropanation step, corre-
22 kJmolꢁ1. Bicyclo
ACTHNGUTERN[UNG 3.1.0]hexene 4 is then released from
ꢁ
ꢁ
sponding to attack of the C6 C7 double bond on C1
4Au by coordination of a BF4 anion, and the catalytic cycle
closed. Intermediate (Z)e is 13 kJmolꢁ1 lower in energy
than 2Au, and it is less stable than the (E)e isomer by
9 kJmolꢁ1. Considering that transition state 2Au-d is
18 kJmolꢁ1 lower in energy than transition state 2Au-f, we
believe that formation of 4 proceeds to a great extent
through a 1,2-shift followed by a cyclopropanation, confirm-
ing the results obtained with enantioenriched dienyne 2 (see
Table 2).
through transition state (E)e-3Au. A last step with an
almost negligible energy barrier of only 1 kJmolꢁ1 leads to
coordinated product species 3Au, from which 3 is released,
closing the catalytic cycle. Intermediate (E)e is 22 kJmolꢁ1
lower in energy than 2Au.
The actual pathway that is followed to reach 3 is deter-
mined by the energy difference between the transition states
of highest energy along the two alternative pathways, that is,
transition states 2Au-d and 2Au-b. According to our calcula-
tions, transition state 2Au-d is lower in energy than 2Au-b
by 14 kJmolꢁ1, which suggests that the main reaction chan-
nel leading to 3 involves 1,2-shift of the ester group first and
then a cyclopropanation step, although the alternative path
corresponding to the cyclopropanation then 1,2-shift se-
quence is competitive.
While the possible mechanistic pathways explaining for-
mation of 3 and 4 from 2Au were already proposed in the
literature,[11a,12,21,22] the formation of 5 from 2Au had only
little mechanistic explanation prior to this study.[12,24,27] We
anticipated that 5 could be formed by at least three different
reaction paths. In all cases allene-coordinated intermediate
7Au is the key intermediate for rationalizing formation of 5
(see Scheme 3). The simplest pathway involves a 1,3-shift of
the ester group from 2Au to give intermediate h through
transition state 2Au-h with an energy barrier of 41 kJmolꢁ1
(see Scheme 3 and Figure S3, Supporting Information). In-
termediate h, which is 14 kJmolꢁ1 lower in energy than 2Au,
then evolves, via transition state h-7Au with an almost negli-
gible energy barrier of 3 kJmolꢁ1, to allene-coordinated spe-
cies 7Au, which is 36 kJmolꢁ1 lower in energy than 2Au.
Several geometries, each rather similar in energy, can be
adopted by 7Au. For the sake of simplicity, in all reaction
Next we focused on the formation of 4, which can also be
reached through two alternative pathways (cyclopropanation
then 1,2-shift or the reverse sequence) that are very similar
to those described above to rationalize the formation of 3.
ꢁ
The first path again involves nucleophilic attack of a C C
ꢁ
double bond on C1, but in this case it is the C11 C12
double bond that we have to consider (see Scheme 3 and
Figure S2, Supporting Information). This cyclopropanation
step leads to intermediate f, which already has the bicyclic
skeleton of product 4. Gold carbene intermediate f is
3248
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3243 – 3260

Cycloisomerization of Enynyl Esters versus Allenenyl Esters
FULL PAPER
pathways presented, the most stable isomer of 7Au is dis-
cussed. Allene-coordinated species 7Au can evolve towards
coordinated product species 5Au through simultaneous
to intermediate (Z)k via transition state (Z)e-(Z)k with an
energy barrier of 20 kJmolꢁ1, and then to Au allene species
7Au through transition state (Z)k-7Au with an energy barri-
er of 50 kJmolꢁ1. Once intermediate 7Au has been reached,
the reaction can evolve to 5, as described before. In this
case, 1,2-shift transition state (Z)e-(Z)k must compete with
cyclopropanation transition state (Z)e-4Au, which is only
2 kJmolꢁ1 higher in energy. This implies that branching from
(Z)e along both reaction pathways is a very likely event.
We also explored whether the C1 and C3 atoms of the
ꢁ
attack of the C11 C12 double bond on C1 and C3 to form
the bicyclo[3.1.0]hexene skeleton of 5 in a single step lead-
G
ing to intermediate i through transition state 7Au-i with an
energy barrier of 24 kJmolꢁ1.
Finally, a 1,2-shift of the ester group of i, through inter-
mediate j and transition states i-j and j-5Au with energy bar-
riers of 16 kJmolꢁ1 and 21 kJmolꢁ1, respectively, leads to the
Au-coordinated product 5Au. Product release from 5Au as-
ꢁ
7Au allene species can be attacked by the C6 C7 double
ꢁ
sisted by a coordinating BF4 counterion is endergonic by
bond of the substrate (see Scheme 4). In this case, the
14 kJmolꢁ1, leads to product 5, and closes the catalytic cycle
bicycloACTHNGUTERNN[UG 3.2.0]heptene skeleton of m is formed through tran-
ꢁ
by forming the starting [(IDM)Au]BF4 species a. The BF4
sition state 7Au-m with a barrier of 55 kJmolꢁ1. Transition
state 7Au-m is 31 kJmolꢁ1 higher in energy than transition
state 7Au-i, which is in agreement with the experimental
counterion can also displace allene 7 from 7Au, yielding
starting [(IDM)Au]ACHTUNGTRENNUNG[BF4] species a and releasing the allene 7
into the reaction medium.
finding that no product having a bicyclo
framework was observed.
ACHTUNGTREN[NGNU 3.2.0]heptene
Formation of 5 can be also explained by two other path-
ways that branch from intermediates (E)e and (Z)e, which
we already introduced to rationalize formation of 3 and 4,
respectively (Scheme 3). We first discuss branching from in-
termediate (E)e. Instead of the cyclopropanation step corre-
For the sake of completeness, we calculated the full cata-
lytic cycle. Thus, intermediate m is 52 kJmolꢁ1 lower in
energy than 2Au, and is connected to the product o through
intermediate n at ꢁ80 kJmolꢁ1, and transition states m-n
and n-o at ꢁ48 kJmolꢁ1 and ꢁ69 kJmolꢁ1, respectively. The
coordinated product o is 144 kJmolꢁ1 lower in energy than
2Au.
ꢁ
sponding to nucleophilic attack of the C6 C7 double bond
on C1, (E)e can undergo a 1,2-shift of the ester group, lead-
ing initially to intermediate (E)k via transition state (E)e-
(E)k with an energy barrier of 33 kJmolꢁ1, and then to Au
allene species 7Au through transition state (E)k-7Au with
an energy barrier of 48 kJmolꢁ1. Once intermediate 7Au has
been reached, the reaction can
Finally, we also explored formation of 5 according to the
pathway shown in Scheme 5, which starts with displacement
ꢁ
of the BF4 counterion from [(IDM)Au]BF4 species a by the
evolve to 5, as described before
(Scheme 3 and Figure S3, Sup-
porting Information). However,
this branching is quite unlikely
considering that 1,2-shift transi-
tion state (E)e-(E)k must com-
pete with cyclopropanation
transition
state
(E)e-3Au,
which is 32 kJmolꢁ1 lower in
energy (Scheme 3).
We now discuss branching
from intermediate (Z)e. Instead
of the cyclopropanation step
corresponding to nucleophilic
ꢁ
attack of the C11 C12 double
bond on C1, (Z)e, similarly to
(E)e, can undergo a 1,2-shift of
the ester group, leading initially
Scheme 4. Alternative cyclization pathway from 7Au.
Scheme 5. Alternative pathway to form 5.
Chem. Eur. J. 2009, 15, 3243 – 3260
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3249

S. P. Nolan, M. Malacria, L. Fensterbank, L. Cavallo et al.
ꢁ
ꢁ
C11 C12 double bond of 2, rather than by the C1 C2 triple
formed with NHC ligands of low steric bulk such as ITM
(see Table 1).
bond.[28] Intermediate p, in which the C11 C12 double bond
ꢁ
is coordinated to the Au atom,[29] is only 2 kJmolꢁ1 higher in
energy than 2Au, which indicates that coordination of the
alkyne or of the alkene group of the substrate is scarcely se-
lective. Nucleophilic attack of the C1 C2 triple bond on the
coordinated C11 C12 double bond proceeds through transi-
Finally, 7Au is another excellent way off the carousel,
since the transition state 7Au-i that connects intermediate
7Au and product 5 is considerably lower in energy (by 33
and 37 kJmolꢁ1, respectively) than transition states 7Au-
(Z)k and 7Au-(E)k, which must be reached by 7Au to move
on the carousel. Thus, the relatively small amount of 5 pro-
duced with 2Au as entry point into the carousel is explained
by the relatively high-energy transition state (Z)k-7Au,
which does not allow easy formation of 7Au from (Z)e. On
the other hand, the direct 1,3-shift pathway from 2Au to
7Au is blocked by relatively high energy transition state
2Au-h. Finally, the golden carousel clearly explains the large
amount of 5 that is formed when allene species 7Au is used
as entry point.
Overall, the theoretical results presented allow us to ra-
tionalize the formation of unprecedented bicyclo-
AHCTUNGTREG[NNUN 3.1.0]hexene 5. According to our calculations, cyclization
would occur after formation of the allenyl ester, between
the allene and the ene part of the 1,4-allenene core.[30] To
verify this hypothesis, we prepared allenyl ester 7 by a
simple silver-catalyzed procedure[31] and subjected it to cycli-
zation conditions. In the presence of 2 mol% of
[(IPr)AuCl]/AgBF4, cycloisomerization of 7 led to the for-
mation of 5 as major product (Table 3, entry 1). The reaction
ꢁ
ꢁ
tion state p-q, which is 62 kJmolꢁ1 higher in energy than
2Au, so that this reaction path is excluded. In this case too,
for the sake of completeness, we calculated the full catalytic
cycle. Thus, intermediate q is 49 kJmolꢁ1 lower in energy
than p, and is connected to 5Au through transition state q-
5Au with an energy barrier of only 4 kJmolꢁ1.
Our calculations clearly indicate that this catalytic reac-
tion is characterized by numerous highly competitive reac-
tion pathways. We cannot exclude that there are other reac-
tion pathways which we were not able to envisage, and that
bulkier NHC ligands could result in somewhat different en-
ergetics. However, exploring this complex manifold of reac-
tions with a larger NHC is almost prohibitive. Nevertheless,
calculations clearly demonstrate that the high reactivity of
the starting structure 2Au is at the origin of this diversity.
As we already indicated, the alkyne-coordinated species can
undergo both 1,2 and 1,3-shift of the ester group, leading re-
spectively to Au carbene species (E)e and (Z)e, and to Au
allene species 7Au. All these species are connected in a cat-
alytic cycle we dubbed “Golden Carousel”. To better illus-
trate this, the most relevant sections of the numerous reac-
tion pathways are shown in Scheme 6.
Table 3. Cyclization of allenyl ester 7.
Clearly, alkyne-coordinated species 2Au is the intermedi-
ate of highest energy in the cycle, while the reservoir of
active species are intermediate d (which is not a way off the
cycle) and allene species 7Au. Intermediate d is easily
formed from 2Au. The Au carbene species (E)e and (Z)e
are ways off the carousel to products 3 and 4, respectively,
while Au-allene species 7Au is the way off to product 5.
Evolution of (E)e, (Z)e, and 7Au depends on the relative
energy of the three transition states (two of them corre-
spond to clockwise and counterclockwise movements on the
Golden Carousel, and the third is a way off the carousel)
that can be reached from each of these intermediates.
Entry
[M]
3:4:5[a]
Total
yield [%][a]
1
A
1/1.4/9.3
1/0.5/0.7
1/0.8/4.5
1/1.4/1.9
93
51
63
50
2[b]
3[b]
4[b]
ACHTUNGTRENNUNG
AuCl
AuCl3
[a] Yields of isolated products, average of two runs, products ratio deter-
mined by 1H NMR spectroscopy. [b] No starting material remained; sig-
nificant amounts of oligomerized by-products were formed.
Intermediate (E)e is an excellent way off the carousel,
since the transition state that connects intermediate (E)e
and product 3, (E)e-3Au, is quite lower in energy (by
20 kJmolꢁ1 and 32 kJmolꢁ1, respectively) than transition
states d-(E)e and (E)e-(E)k, which are the transition states
to be reached by (E)e in order to move on the carousel.
This explains the easy formation of 3. In contrast, (Z)e is
not a good way off the carousel, since the most likely event
is a counterclockwise move on the carousel, to yield the
highly stable intermediate d. The two other transition states
of highest energy accessible from intermediate (Z)e, namely,
(Z)e-4Au and (Z)k-7Au, correspond respectively to the way
off the carousel leading to 4Au and to a clockwise move to-
wards allene intermediate 7Au (Scheme 6). They are 11 and
23 kJmolꢁ1 higher in energy than transition state d-(Z)e,
which explains the small amount of products 4 and 5
was more selective in favor of 5 than the cyclization of 2 (3/
4/5, 1/1.4/9.3 vs. 1/0.4/1.4), which strongly supports allenyl
ester 7 as intermediate in the transformation 2!5. Impor-
tantly, this is the first reaction producing 3 as the minor
product.
Further examination of the reactivity of 7 in the presence
of [(PPh3)AuCl] revealed the importance of the ligand on
the gold center (Table 3, entry 2). The phosphine-containing
catalyst notably produced significant amounts of oligomer-
ized products, as previously observed in the cyclization of di-
enyne 2, and afforded a more contrasting ratio of cyclopro-
panated bicycles. Keeping in mind that products 3 and 4 are
likely produced via direct cyclopropanation of enyne 2 and/
or intermediate I, both arising from “retro-migration” of the
3250
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3243 – 3260

Cycloisomerization of Enynyl Esters versus Allenenyl Esters
FULL PAPER
Scheme 6. Golden Carousel linking 2, 3, 4, 5, and 7.
OAc group in 7, it appears that the competition between
acetate and alkene as internal nucleophiles in 7 is strongly
influenced by the ligand on gold. More precisely, allenyl
acetate 7 seems to be less prone to isomerization back to I
and 2 when activated by the cationic [(NHC)Au]+ species
than with the phosphine gold catalyst.[32] Finally, the absence
of rearranged product from the 1,5-allenene framework
seems to indicate a strong pref-
We also carried out the cycloisomerization of allenene 8,
which afforded a quantitative yield of novel cyclopropyl
compound 9 [Eq. (2)]. In this case, excellent selectivity was
already obtained when starting from the corresponding
propargylic acetate 10, but the yield of 9 was clearly im-
proved. While this total selectivity in favor of the new type
of product was unexpected at the time of our preliminary
erence for the 1,4-allenene scaf-
fold in the present catalytic
system, as was predicted in our
theoretical
studies
(see
Scheme 4).[33,34]
Chem. Eur. J. 2009, 15, 3243 – 3260
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3251

S. P. Nolan, M. Malacria, L. Fensterbank, L. Cavallo et al.
communication,[12] it can now be easily rationalized. Forma-
tion of allenyl acetate 8 from 10 is favored by the alkyl sub-
stituents of the alkyne, and therefore leads straightforwardly
to cyclized product 9.
1,2- versus 1,3-OAc migration: 18O-labeling experiments:
Upon p activation, propargylic esters can lead to allenyl
esters that are further activated by the metal center, giving
rise to a wide variety of products.[9] Even though it is also
observed in the presence of platinum,[35] this reactivity is
mostly triggered by gold catalysts.[36] We have established
Scheme 7. 18O-labeling experiments in the cycloisomerization of 2.
3252
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3243 – 3260

Cycloisomerization of Enynyl Esters versus Allenenyl Esters
FULL PAPER
that allenyl ester 7 is very likely a key intermediate in the
pathway leading to carbocycle 5. This crucial insight was
provided by our theoretical calculations. Additionally, the
computed values for the double 1,2-OAc migration were
found to be slightly higher than those for the direct 1,3-OAc
migration (i.e., 50 vs. 41 kJmolꢁ1), which, taking into consid-
eration the possible experimental error, renders both path-
ways competitive. Therefore, in an effort to try to under-
stand every aspect of the cyclization studied here, and will-
ing to consider thoroughly any possibility, we carried out la-
beling experiments with 2-18O, our model substrate with a
carbonyl oxygen atom 47% enriched in 18O, to address this
issue.[37]
We then analyzed, by high-resolution mass spectrometry,
the isotopic distribution of different mixtures of products
from reactions performed in the presence of PtCl2,
AuCl3,and [(IPr)AuCl]/AgBF4 (Scheme 7). Both products of
the Pt-catalyzed cycloisomerization showed nearly quantita-
tive transfer of the 18O content at the vinylic position
(Scheme 7, top bracket), consistent with a net 1,2-migration
of the acetate group. On the contrary, the reactions conduct-
ed in the presence of gold revealed a more complex behav-
ior (middle and bottom brackets). The three produced car-
bocycles 3-18O, 4-18O, and 5-18O displayed scrambled isotopic
contents carbonyl/enol ranging from 75/25 to 40/60. These
results indicate that gold catalysts, both AuCl3 and the cat-
ionic [(IPr)Au]+ fragment, isomerize propargylic acetates to
allenyl acetates by both double 1,2-migration (Scheme 1,
I!II!III) and direct 1,3-migration (I!III) randomly. No-
tably, under the present conditions, isomerization between
the alkyne and the allene presumably occurs several times
before the cyclization step is achieved.[38,39]
Scheme 8. 1,2-OAc migration via an oxirenium intermediate.
7-18O to 5-18O. The absence of scrambling in this case could
appear surprising and points to the faster rate of nucleophil-
ic attack by the alkene than by the carbonyl group. We be-
lieve that the improved selectivity observed in the formation
of 5 when allene 7 was used as cyclization precursor in place
of 2 and the absence of 18O scrambling from 7-18O both sup-
port this reactivity pattern.[43,44]
Interestingly, these findings correlate well with the results
of the cycloisomerization of the enantioenriched precursor
2. Better chirality transfer was observed with platinum(II)
catalysis, while gold catalysis, which shows significant 18O
scrambling, also furnished nearly racemic products.
Reversibility of acetate migration: One of the main findings
of the experimental and theoretical studies described above
on the model dienyne 2 is that an apparent 1,2-OAc migra-
tion in the final cyclized product can in fact result from a
1,3-shift followed by a “retro” 1,2-migration of the acetate
group. This “retro-migration” process has only been ob-
served adventitiously previously[45] and has not been the sub-
ject of further investigations. We now provide additional
data supporting this reactivity.
We also envisaged, as a possible alternative, that migra-
tion of the acetate group could occur via an oxirenium inter-
mediate, as proposed by Gevorgyan et al. for the isomeriza-
tion of propargylic acetates and phosphates,[40] and by
Hotha and Kashyap for the activation of propargyl glyco-
sides.[41,42] This possibility was ruled out by DFT calculations,
since the oxirenium-like transition state, connecting in a
single step the alkyne-coordinated species 2Au and the Au
carbene intermediate (E)e, lies 121 kJmolꢁ1 above 2Au (see
Scheme 8), which is much higher in energy than other tran-
sition states that can be reached from 2Au.
Intermolecular cyclopropanation: Intermolecular cyclopro-
panation of propargylic esters with alkene was first reported
by Ohe and Uemura using ruthenium(II) catalysts[46] and
subsequently studied by Toste et al. with gold catalysts.[21]
Our goal was to use allenyl acetates in this reaction and
prove “retro-migration” of the ester by intermolecular trap-
ping of the gold carbenoid by an olefin.
We carried out the reaction of allenyl acetate 11 with
3 equivalents of styrene in the presence of platinum(II),
gold(I), and goldACTHNUTRGNEU(GN III) catalysts (Table 4). Thus, under typical
“platinum conditions” (toluene, 808C), we observed forma-
tion of cyclopropane derivative 12 in moderate yield
(Table 4, entry 1). GoldACTHNUGRTENUNG(III) chloride led to decomposition
(Table 4, entry 2), while gold(I) catalysts showed compara-
ble activity to PtCl2 and had the advantage of being active
Next, we prepared allene 7-18O from 2-18O and subjected
it to cyclization conditions. Interestingly, we observed com-
plete loss of 18O [Eq. (3)], in strong support of 1,2-migration
of the acetate group through its carbonyl oxygen atom, from
Chem. Eur. J. 2009, 15, 3243 – 3260
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3253

S. P. Nolan, M. Malacria, L. Fensterbank, L. Cavallo et al.
Table 4. Intermolecular cyclopropanation of allenyl ester 11.
Table 5. Cycloisomerization of allenyl acetate 15.
Entry [M] (mol%)
Reaction conditions Yield of
16 [%][a]
1
2
3
PtCl2 (5 mol%)
PtCl2 (5 mol%)
AuCl3 (2 mol%)
toluene, 808C, 4 h
neat, 808C, 2 h
CH2Cl2, reflux, 2 h
49
61
24
51
20
Entry [M]
Reaction conditions
Yield of
12 [%][a]
4
ACHTNUGTERN[NUNG (PPh3)AuCl]/AgSbF6 (2 mol%) CH2Cl2, reflux, 2 h
5[b]
A
CH2Cl2, reflux, 8 h
1
PtCl2 (5 mol%)
AuCl3 (2 mol%)
toluene, 808C, 15 min 55
2[b]
3
CH2Cl2, RT, 2 h
–
[a] Yields of isolated products. [b] [AuI]=
A
46
51
62
4
5
N
CH2Cl2, RT, 4 h
CH2Cl2, RT, 2 h
T
1
[a] Yields of isolated products, d.r. determined by H NMR spectroscopy.
[b] Decomposition
of
11
was
mainly
observed.
[c] [AuI]=
transformation than its triphenylphosphine counterpart
(Table 5, entry 5 versus entry 4). The results presented here
show the viability of ester migration onto the allene moiety
in the context of intramolecular cyclization. As observed
previously in the cyclization of the corresponding propargyl-
ic acetate,[49] the reaction produced only one diastereomer.
In an attempt to extend the scope of this reaction, forma-
tion of tricyclic derivative 18 was envisioned. Unexpectedly,
we only recovered the unchanged allene 17, even after pro-
longed reaction in the presence of AuI and AuIII catalysts
[Eq. (4)].
at room temperature (Table 4, entries 3 and 4). This set of
experiments validates the aforementioned possibility of
forming transition metal vinyl carbenoid intermediates on
activation of an allenyl ester moiety, which, to the best of
our knowledge, is unprecedented.[47] In the present case, 12
is likely formed through the intermediacy of carbene species
14, resulting from a 1,2-OAc
shift in 13.[48] Encouraged by
these results, we decided to fur-
ther examine the behavior of
allenyl esters in the presence of
soft p-acid catalysts and espe-
cially the possible dichotomy
between allenenyl esters and
their corresponding enynyl
esters.
Cyclization of propargylic ace-
tates versus allenyl acetates:
scope and limitations: Acetate
15 having a 1,6-allenene core
was subjected to typical cycliza-
tion conditions in the presence
of gold and platinum catalysts
(Table 5).
Formation of tricyclic com-
pound 16 was observed in all
cases, whereby PtCl2 and
[(PPh3)AuCl]/AgSbF6 were the
most efficient catalytic systems
(Table 5, entries 1, 2, and 4).
The gold(I) catalyst bearing a
bulky ortho-biphenylphosphine
ligand was less effective in this
3254
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3243 – 3260

Cycloisomerization of Enynyl Esters versus Allenenyl Esters
FULL PAPER
Even though a thorough explanation of this observation is
lacking at the moment, it appears that, as in the case of
enyne, minute changes in the architecture of the allenene
precursor can be detrimental to the cycloisomerization pro-
cess.
various enynyl acetates with their corresponding allenenyl
acetates showed that the latter substrates could lead to for-
mation of alternative products.
This combined experimental and theoretical study permit-
ted the details of a novel cycloisomerization mode to be un-
covered and more generally to gain insight into mechanistic
aspects of the intricate cyclization mode of enynes bearing a
propargylic ester group. Further studies taking advantage of
this experimental/theoretical approach to complex synthetic
issues are ongoing.
Turning our attention to other substrates, we examined
the reactivity of different unsaturated cyclization precursors.
Interestingly, 1,5-allenene 19 and its 1,6-enyne counterpart
20 afforded similar yields and product ratio of 21 and 22
[Eq. (5)], whereby the cyclization from the allene was slight-
ly more efficient than from the alkyne.[50] The formation of
cyclized product 22, which is formally an enyne ring-closing
metathesis product,[51] from allene 19 implies in situ forma-
tion of alkyne 20 via 1,3-shift of the acetate moiety.
In light of the difference in reactivity observed between
15 and 17, we decided to examine close analogues of 19 and
20, namely, 23 and 26. In the presence of gold(I) catalyst,
alkyne 23 afforded a mixture of cyclopropane and diene de-
rivatives 24 and 25 [Eq. (6)].[52] In contrast to the cyclobutyl
derivatives, in this series the cyclopropanated product 24 is
favored and formed in a 3/1 ratio over the metathesis prod-
uct, diquinane 25.
Starting from the corresponding allenyl acetate 26 re-
vealed a different behavior of the gold(I) catalyst employed
here. In this case, tricyclic acetate 24 was only formed as the
minor product along with triene 27 as major compound
[Eq. (7)]. Additionally, no formal metathesis product 25
could be observed. The formation of 27 results presumably
from a [3,3] rearrangement catalyzed by the cationic gold(I)
center. A thermal [3,3] sigmatropic rearrangement is pre-
sumably ruled out since the reaction occurs at room temper-
ature.[53]
Experimental Section
General information: All reagents were used as purchased. Dry THF and
dry dichloromethane were purified by passing through a purification
column from Innovative Technology Inc. (SPS-400-6). Silver salts were
stored in a desiccator wrapped in aluminum foil. TLC analysis of reaction
mixtures was performed on EMD Chemicals silica gel 60 F254 plates and
visualized by UV. Flash chromatography was performed on silica gel 60
(230–400 mesh, Silicycle). 1H and 13C NMR spectra were recorded on a
Varian 300, Varian 400, or Bruker 300 MHz spectrometer at ambient
temperature in CDCl3 containing TMS. Chemical shifts were referenced
to the peak of TMS (0.0 ppm). Assignments of some 1H and 13C NMR
signals rely on COSY and/or HMBC experiments. Chiral GC analyses
were performed under isothermal conditions (1108C) on a chiral column
CP-Chirasil-DEX CB (25 m). [(NHC)AuCl][15] complexes and [AuI][54]
were synthesized according to literature procedures. Compounds 2,[11a]
7,[12] 10,[12] 15,[49] 17,[49] 19,[49] 20,[49] and 23[49] were synthesized according
to literature procedures, and their spectroscopic data found to be in good
agreement with previously reported characteristic data. Characterization
of cyclized compounds 3,[11a] 4,[12] 5,[12] 9,[12] 16,[49] 21,[49] and 24[49] was
based on previous reports, and their spectroscopic data were in good
agreement with previously reported characteristic data.
Synthesis of [(IDD)AuCl]: [(IDD)AgCl][55] (200 mg, 0.37 mmol, 1 equiv)
and dimethyl sulfide gold(I) chloride (130 mg, 0.44 mmol, 1.1 equiv) were
dissolved in a minimal amount of CH2Cl2 and stirred overnight at room
temperature. Charcoal was added and the reaction mixture stirred for a
further 3 h. After filtration over Celite, a clear greenish solution was ob-
tained. The volume of CH2Cl2 was reduced, and the desired complex pre-
cipitated and was washed with pentane to yield 180 mg (77% yield) of a
white powder. 1H NMR (400 MHz, CDCl3): d=6.94 (s, 2H, CHIm), 4.91
(pent, J=6.6 Hz, 2H, CHCyclododecyl), 2.06–1.96 (m, 4H), 1.69–1.60 (m,
8H), 1.56–1.5 (m, 2H), 1.46–1.28 ppm (m, 30H); 13C NMR (100 MHz,
CDCl3): d=169.8 (C, CAu), 117.6 (CH, CHIm), 58.0 (CH, CHCyclododecyl),
31.2 (CH2), 23.7 (CH2), 23.5 (CH2), 23.4 (CH2), 23.2 (CH2), 21.6 ppm
(CH2); HRMS calcd for C29H52AuN3 [MꢁCl+MeCN]+: 639.3821; found:
639.3825. CCDC 679915 contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_-
request/cif..
Conclusion
A thorough study of the reactivity pattern of a peculiar
branched dienyne in the presence of gold and platinum cata-
lysts has been presented. We examined the effect of the an-
cillary ligand bound to the gold center and found that bulky
N-heterocyclic carbenes are a key parameter for obtaining
significant ratios of bicycloACTHNUTRGENUGN[3.1.0]hexene 5 over compounds
3 and 4. Further experimental and theoretical mechanistic
studies allowed the formation of all cyclized products to be
accounted for. Notably, a reaction pathway leading to cy-
clized compound 5 has been proposed. The cycloisomeriza-
tion reaction leading from 2 to 5 would occur by a 1,3-OAc
shift/allene–ene cyclization/1,2-OAc shift sequence. Support-
ing this hypothesis, which translates the apparent 1,2-OAc
shift from 2 to 5 in a 1,3-OAc migration followed by a 1,2-
OAc “retro-migration”, cyclization of allenenyl acetate 7 af-
forded a greater proportion of 5 over 3 and 4. Additionally,
theoretical calculations suggested that the selectivity to-
wards the three cyclopropyl derivatives 3, 4, and 5 is dictat-
ed by the following equilibrium: [gold propargyl acetates
IQgold carbenoid vinyl acetates IIQgold allenyl acetates
III]. Finally, examination of the differences in reactivity of
Typical procedure for cycloisomerization of (2): To a solution of gold
[platinum] catalyst (2 mol% [5 mol%]) in CH2Cl2 [toluene] (5 mL) in a
scintillation vial equipped with a stir bar, AgBF4 (if required) was added.
The solution instantly became cloudy. A solution of dienyne 2 (44 mg,
0.2 mmol) in CH2Cl2 [toluene] (3 mL) was then added. After consump-
tion of the starting material, the solvent was removed. The resulting mix-
ture was dissolved in pentane, filtered through Celite, and the filtrate
concentrated in vacuo. Yields refer to mixtures of isolated products;
1
ratios are based on H NMR integration.
Trapping experiment in methanol: formation of 4-acetyl-6,6-dimethyloc-
ta-1,7-dien-4-yl acetate (6): Following the typical procedure for the cyclo-
isomerization of 2 (1 mmol) in methanol, the crude product was purified
by flash chromatography (pentane/tert-butyl methyl ether 8/2), which af-
forded 149 mg (59% yield) of 6. 1H NMR (CDCl3, 400 MHz): d=5.83
Chem. Eur. J. 2009, 15, 3243 – 3260
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3255

S. P. Nolan, M. Malacria, L. Fensterbank, L. Cavallo et al.
(dd, J=17.6, 10.8 Hz, 1H, H2), 5.62–
5.51 (m, 1H, H7), 5.01–4.96 (m, 2H,
7-Allyl-9,9-dimethylundeca-5,6,10-
trien-5-yl acetate (8): The general pro-
cedure using propargylic acetate 10[12]
H8), 4.84–4.77 (m, 2H, H1), 4.38 (d,
J=2.0 Hz, 1H, 1H10), 4.04 (d, J=
(442 mg, 1.6 mmol) yielded 420 mg
(95%) of 8. 1H NMR (400 MHz,
2.0 Hz, 1H, 1H10), 3.50 (s, 3H, OMe),
3.07 (dd, J=14.4, 8.8 Hz, 1H, 1H6),
CDCl3): d=5.83 (dd, J=17.4, 10.7 Hz,
1H, H2), 5.86–5.75 (m, 1H, H7), 5.05–
2.54–2.47 (m, 2H, 1H6 + 1H4), 1.97 (s,
5.00 (m, 2H, 2H8), 4.91 (dd, J=17.5, 1.4 Hz, 1H, 1H1), 4.89 (dd, J=11.4,
1.4 Hz, 1H, 1H1), 2.87–2.73 (m, 2H, 2H6), 2.25–2.13 (m, 2H, 2H11), 2.08
(s, 3H, OAc), 1.43–1.32 (m, 4H, 2H12 + 2H13), 1.04 (s, 3H, Me), 1.03 (s,
3H, Me), 0.89 ppm (t, J=7.1 Hz, 3H, 3H14); 13C NMR (100 MHz,
CDCl3): d=193.9 (C, C9), 168.6 (C, C=O), 148.0 (CH, C7), 135.6 (CH,
C2), 123.6 (C, C10), 116.0 (CH2, C8), 111.7 (CH2, C1), 110.3 (C, C5), 46.1
(CH2, C6), 40.3 (CH2, C4), 37.4 (C, C3), 31.5 (CH2, C11), 28.5 (CH2), 27.0
(CH3), 26.8 (CH3), 22.1 (CH2), 21.0 (CH3, OAc), 13.8 ppm (CH3, C14);
HRMS (ES+): m/z calcd for C18H28O2Na [M+Na]+: 299.1982; found:
299.1990.
3H, OAc), 1.87 (d, J=15.2 Hz, 1H,
1H4), 1.01 ppm (s, 6H, 2ꢆCH3); 13C NMR (CDCl3, 100 MHz): d=169.6
(C, OC(O)Me), 161.3 (C, COMe), 148.4 (CH, C2), 132.8 (CH, C7), 117.9
(CH2, C8), 109.1 (CH2, C1), 85.3 (C, C5), 82.2 (CH2, C10), 54.4 (CH3,
OMe), 44.8 (CH2, C4), 40.8 (CH2, C6), 37.0 (C, C3), 27.8 (CH3, Me), 27.4
(CH3, Me), 22.7 ppm (CH3, OAc); elemental analysis (%) calcd for
C15H24O3 (M=252.35): C 71.39, H 9.09, found: C 71.18, H 9.15.
Preparation of enantioenriched dienyne acetate (2):[56] A 2.2m solution of
nBuLi in hexanes was added to a solution of bis-oxazoline (322 mg,
1.05 mmol, 1.05 equiv) and 2,2’-bipyridine (0.2 mg) in THF (1.8 mL) at
08C till the solution turned red. After completion of addition, the reac-
tion mixture became a red-brown suspension. The reaction mixture was
warmed to room temperature and stirred for 15 min. A solution of allyl-
zinc bromide (1.5m, 0.7 mL, 1 mmol, 1 equiv) was added to the solution
of lithiated bis-oxazoline. After 30 min, ketone A[11a] (187 mg, 0.9 mmol,
0.9 equiv) in THF (1.3 mL) was added at ꢁ1008C. When the reaction
was complete, the mixture was quenched with MeOH/H2O (0.04 mL,
1/1). The solvent was removed under vacuum and the crude product puri-
2-(2-Allylcyclopentylidene)vinyl acetate (26): The general procedure
using propargylic acetate 23 (44 mg, 0.2 mmol) yielded 23 mg (51%) of
the title compound as a mixture of two diastereomers (1/1). 1H NMR
(CDCl3, 400 MHz): d=7.44 (m, 1H, one dia), 7.35 (m, 1H, one dia),
5.87–5.73 (m, 1H), 5.07–4.98 (m, 2H), 2.67–2.61 (m, 1H), 2.59–2.43 (m,
2H), 2.41–2.32 (m, 1H), 2.14 (s, 3H, one dia), 2.13 (s, 3H, one dia), 2.11–
1.99 (m, 1H), 1.97–1.89 (m, 1H), 1.84–1.75 (m, 1H), 1.70–1.59 (m, 1H),
1.45–1.31 ppm (m, 1H); 13C NMR (CDCl3, 100 MHz): d=190.4 (2C),
169.1 (2C), 137.3, 137.2, 124.9 (2C),
116.4, 116.3, 112.6, 112.1, 45.0, 44.2,
38.8, 38.4, 32.9, 32.8, 32.7, 32.6, 25.4,
25.3, 21.4, 21.3 ppm; IR (neat): n˜ =
3066, 2957, 1976, 1753, 1640,
1434 cmꢁ1; HRMS (ES+): m/z calcd
for C12H16O2Na [M+Na]+: 215.1042;
found: 215.1042.
Procedure for isotopic labeling: 18O-la-
beled water (95% 18O, 0.84 mL,
41.6 mmol) was added to a solution of
acetyl chloride (2.96 mL, 41.6 mmol)
in dry dichloromethane (20 mL) at room temperature with vigorous stir-
ring. After 2 h, 5.5 mL of this solution (corresponding to 9.85 mmol of
acetic acid) was added to a solution of thionyl chloride (0.790 mL,
10.83 mmol) in dry dichloromethane (2 mL), and the solution was stirred
at room temperature for 12 h. The resulting reaction mixture was trans-
ferred to a solution of dienyne 2 (350 mg, 1.97 mmol) in dry dichlorome-
thane (2 mL) with triethylamine (4.15 mL, 30 mmol) and 4-dimethylami-
nopyridine (120 mg, 0.98 mmol). The reaction was stirred at room tem-
perature for 50 h, and the resulting brown solution was diluted with di-
ethyl ether (100 mL) and washed successively with a saturated aqueous
solution of NH4Cl (2ꢆ50 mL), a saturated aqueous solution of NaHCO3
(2ꢆ30 mL), and brine (50 mL). The organic phase was dried over anhy-
drous magnesium sulfate and the solvent was removed by evaporation
under vacuum. The crude mixture was purified by flash chromatography
on silica gel with pentane/diethyl ether (98/2). The 18O-labeled acetate
was obtained as a colorless oil (312 mg, 71%). The spectroscopic data
were in agreement with those of the corresponding 16O acetate except for
the carbonyl signal, which appeared as two peaks in the 13C spectrum:
169.2 ppm for C=16O and 169.3 ppm for C=18O. The HRMS spectrum
confirmed the labeling of the acetate group: 16O-2: HRMS (ES+): m/z
calcd for C14H20O2Li [M+Li]+: 227.1618; found: 227.1620; 18O-2: HRMS
(ES+): m/z calcd for C14H20O18OLi [M+Li]+: 229.1660; found: 229.1661.
Integration of mass peaks gave an isotopic abundance of 47%. All the
isotopic labeling was localized on the oxygen atom of the carbonyl group,
which is in agreement with the preparation procedure.
fied by flash chromatography on silica gel (pentane/dichloromethane 7/3)
to afford B (105 mg, 47%). All spectroscopic data were in agreement
with a racemic sample. The enantiomeric excess was determined by
chiral gas chromatography (isotherm 1108C): enantiomer
1 (rt1=
32.4 min) and enantiomer 2 (rt2=33.2 min), 66% ee. [a]2D0 =+7.58 (c=
0.3, CHCl3). Desilylation and acylation of enantioenriched alcohol B ac-
cording to reported procedures[11a] led to formation of enantioenriched 2;
[a]2D0 =+3.28 (c=0.99, CHCl3).
Computational details: All DFT calculations were performed with the
Gaussian 03 package.[57] The BP86 GGA functional of Becke and Perdew
was used.[58] The TZVP triple-z basis set with one polarization function
was used for main group atoms,[59] while the relativistic SDD effective
core potential in combination with a triple-z basis set was used for the
Au atom.[60] All geometries were verified by frequency calculations that
resulted in 0 and 1 imaginary frequency for intermediates and transition
states, respectively. The reported energies include the vibrational gas-
phase zero-point energy term, and a solvation term that was obtained
through single-point calculations on the gas-phase optimized geometries.
The polarizable continuous solvation model IEF-PCM as implemented in
the Gaussian 03 package was used.[59] CH2Cl2 was chosen as model sol-
vent, with a dielectric constant of e=8.93. Standard nonelectrostatic
terms were also included.
Typical procedure for synthesis of allenyl acetates: A solution of propar-
gylic acetate (1 mmol) in CH2Cl2 (5 mL) was added to a suspension of
AgBF4 (5 mg, 0.05 mmol, 5 mol%) in CH2Cl2 (10 mL) in a scintillation
vial equipped with a stir bar in the absence of light. When TLC analysis
showed total consumption of the starting material, the solvent was re-
moved. The resulting mixture was dissolved in pentane, filtered through
Celite, and evaporated. The crude oil was purified, when necessary, by
flash chromatography on silica gel.
Determination of isotopic distribution: Due to experimental constraints,
the isotopic distribution was not determined directly on the enol acetates
obtained after the cycloisomerization reaction. The three products were
separated and immediately subjected to methanolysis conditions
(MeONa/MeOH solution). Neutralization was done by addition of solid
NH4Cl to avoid addition of water and hydration of the corresponding car-
3256
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3243 – 3260

Cycloisomerization of Enynyl Esters versus Allenenyl Esters
FULL PAPER
bonyl derivatives. MS analysis allowed for the 18O isotopic abundance
(the ratio between C12H18OLi and C12H18O18Li peaks) to be determined
on the carbonyl derivatives, and comparison with the starting material af-
forded the structural isotopic distribution (MS spectra for each experi-
ment are provided in the Supporting Information).
Acknowledgements
S.P.N. acknowledges the Petroleum Research Fund administered by the
ACS, the Ministerio de Educaciꢀn y Ciencia, the ICIQ Foundation, and
ICREA for financial support. SPN is an ICREA Research Professor.
N.M. (2007FI_A 01466) and R.S.R. (2008FI 01047) thank the AGAUR
for pre-doctoral fellowships. L.C. thanks the Regione Campania progetto
Legge5/2005 for financial support, and the INTSM (Cineca Grant) for
computer time. J.C.T. and D.L. thank the CNRS, the UPMC, and the MS
platform SM3P. L.F. and M.M. thank the MRES, the CNRS, the IUF, and
the ANR Blan 0302 “Allꢂnes” for financial support. N.M. thanks Umi-
core for a Ph.D award.
Carbene trapping experiment: formation of cyclopropyl derivative 12:
AgSbF6 (16.5 mg, 0.048 mmol) was added to a solution of [(PPh3)AuCl]
(24 mg, 0.048 mmol) in dry dichloromethane (10 mL). The solution
became cloudy. A solution of allenyl ester 11 (400 mg, 2.41 mmol) and
styrene (1.38 mL, 12.05 mmol) in dry dichloromethane (2 mL) was added
after 5 min and the mixture stirred under the conditions (time and tem-
perature) indicated in Table 4. When TLC analysis indicated total con-
sumption of the starting material, the crude product was immediately pu-
rified by flash chromatography (pentane/diethyl ether 95/5) to afford cy-
clopropane derivatives 12[45] (300 mg, 46%) as a 3/1 mixture of cis/trans
diastereomers. Application of the same procedure to AuCl3 catalysis
without silver salt and to [(IPr)AuCl]/AgBF4 afforded respectively less
than 10 and 51% of cyclopropane derivatives 12 with similar diastereo-
meric ratio. Platinum (PtCl2, 5 mol%) catalysis was carried out in tolu-
ene at 808C and the same products were isolated in 55% yield. All the
spectroscopic data were in agreement with those already reported.[45] The
cis diastereomer was isolated as an analytical sample: 1H NMR (CDCl3,
400 MHz): d=7.26–7.19 (m, 2H), 7.18–7.14 (m, 1H), 7.06 (d, J=8.4 Hz,
2H), 2.38–2.25 (m, 2H), 2.23–2.10 (m, 4H), 2.08–1.92 (m, 2H), 1.84–1.75
(m, 1H), 1.50–1.38 (m, 1H), 1.37–1.23 (m, 4H), 1.13–1.01 (m, 2H), 0.51–
0.39 ppm (m, 1H); 13C NMR (CDCl3, 100 MHz): d=169.5 (C), 139.4 (C),
135.2 (C), 130.6 (C), 127.6 (CH), 127.4 (CH), 125.6 (CH), 28.7 (CH2),
27.6 (CH2), 26.6 (CH2), 26.5 (CH2), 26.2 (C), 24.6 (CH), 21.6 (CH), 20.7
(CH3), 11.6 ppm (CH2).
[1] B. M. Trost, Science 1991, 254, 1471–1477.
[2] For recent general reviews on the chemistry of polyunsaturated sub-
strates, see: a) L. Fensterbank, J.-P. Goddard, M. Malacria, Compre-
hensive Organometallic Chemistry III, Vol. 10 (Eds.: R. Crabtree, M.
Mingos, I. Ojima), Pergamon Press, Oxford, 2006, Chapter 7,
pp. 299–368; b) C. Aubert, L. Fensterbank, V. Gandon, M. Malacria,
Top. Organomet. Chem. 2006, 19, 259–294.
[3] a) B. M. Trost, Angew. Chem. 1995, 107, 285–307; Angew. Chem.
Int. Ed. Engl. 1995, 34, 259–281; b) B. M. Trost, Acc. Chem. Res.
2002, 35, 695–705.
[4] For reviews on enyne cycloisomerization, 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) C. Nieto-Oberhuber, S.
Lꢀpez, E. Jimꢃnez-NfflÇez, A. M. Echavarren, Chem. Eur. J. 2006,
12, 5916–5923; c) Z. Zhang, G. Zhu, X. Tong, F. Wang, X. Xie, J.
Wang, L. Jiang, Curr. Org. Chem. 2006, 10, 1457–1478; d) C. Bru-
neau, Angew. Chem. 2005, 117, 2380–2386; Angew. Chem. Int. Ed.
2005, 44, 2328–2334; e) S. T. Diver, A. J. Giessert, Chem. Rev. 2004,
104, 1317–1382; f) A. M. Echavarren, C. Nevado, Chem. Soc. Rev.
2004, 33, 431–436; g) A. M. Echavarren, M. Mꢃndez, M. P. MuÇoz,
C. Nevado, B. Martín-Matute, C. Nieto-Oberhuber, D. J. Cµrdenas,
Pure Appl. Chem. 2004, 76, 453–463; h) G. Lloyd-Jones, Org.
Biomol. Chem. 2003, 1, 215–236; i) C. Aubert, O. Buisine, M. Mal-
acria, Chem. Rev. 2002, 102, 813–834.
Cycloisomerization of propargylic acetates versus allenyl acetates: Apply-
ing the general cycloisomerization procedure to 19 or 20 with [AuI] cata-
lyst gave 21 and 22 in 1/3 ratio and in 88% yield from 19 and 82% from
20. Dienyl derivative 22: 1H NMR (CDCl3, 400 MHz): d=6.33 (dd, J=
11.1, 18.0 Hz, 1H), 5.99 (brs, 1H), 5.29 (d, J=18.0 Hz, 1H), 5.03 (d, J=
11.1 Hz, 1H), 3.12–3.06 (m, 1H), 2.90–2.84 (m, 1H), 2.56–2.45 (m, 1H),
2.43–2.33 (m, 1H), 2.19–2.07 (m, 2H), 1.98 (s, 3H), 1.36–1.28 ppm (m,
1H); 13C NMR (CDCl3, 100 MHz): d=169.4, 140.9, 136.0, 129.7, 115.0,
90.0, 43.5, 38.8, 31.9, 22.8, 21.6 ppm: IR (neat): n˜ =2934, 2855, 1741, 1636,
1586, 1432 cmꢁ1; HRMS (ES+): m/z calcd for C11H14O2Na [M+Na]+,
201.0886; found, 201.0887.
Applying the general cycloisomerization procedure to 23 with [AuI] cata-
lyst gave 24 and formal metathesis product 25 in 70% yield as a 3/1 mix-
ture. Diquinane derivative 25: 1H NMR (CDCl3, 400 MHz): d=6.32 (dd,
J=11.1, 18.0 Hz, 1H), 5.81 (m, 1H), 5.34 (dd, J=1.0, 18.0 Hz, 1H), 5.05
(dd, J=1.0, 11.1 Hz, 1H), 2.97–2.90 (m, 1H), 2.88–2.82 (m, 1H), 2.13–
1.93 (m, 3H), 1.97 (s, 3H), 1.76–1.66 (m, 1H), 1.56–1.44 (m, 1H), 1.27–
1.25 ppm (m, 1H); 13C NMR (CDCl3, 100 MHz): d=170.1, 141.6, 134.6,
130.5, 114.4, 102.1, 47.3, 38.8, 38.3, 34.9, 25.0, 21.9 ppm: IR (neat): n˜ =
2999, 2951, 2856, 1752, 1711, 1643, 1451 cmꢁ1; HRMS (ES+): m/z calcd
for C12H16O2Na [M+Na]+: 215.1042; found: 215.1043.
Applying the general cycloisomerization procedure to 26 with [AuI] cata-
lyst gave 24 and triene 27 in 85% yield as a 35/65 mixture. Triene 27 was
isolated as a 2/1 (E/Z) isomer mixture: 1H NMR (CDCl3, 400 MHz): d=
7.23 (s, 1HM), 6.99 (s, 1Hm), 5.87–5.77 (m, 2H), 5.10–4.97 (m, 2H), 3.12
(d, J=6.3 Hz, 2 HM), 2.92 (dd, J=6.6, 1.6 Hz, 2Hm), 2.72–2.67 (m, 2Hm),
2.46–2.35 (m, 4HM + 2Hm), 2.17 (s, 3 HM), 2.14 (s, 3Hm), 1.95–1.86 ppm
(m, 2 HM + 2Hm); 13C NMR (CDCl3, 100 MHz): major diastereomer:
d=168.0, 139.9, 136.2, 133.3, 128.2, 121.1, 115.2, 33.0, 32.2, 31.1, 23.2,
21.0 ppm; minor diastereomer: d=168.0, 139.4, 136.8, 131.6, 130.4, 119.2,
116.2, 36.3, 34.9, 32.3, 23.6, 21.0 ppm; IR (neat): n˜ =3320, 3080, 2960,
2860, 1750, 1640, 1610 cmꢁ1; HRMS (ES+): m/z calcd for C12H16O2Na
[M+Na]+, 215.1042; found, 215.1043.
[5] For a review, see: B. M. Trost, Acc. Chem. Res. 1990, 23, 385–393.
[6] For reviews, see: a) L. AÇorbe, G. Domínguez, J. Pꢃrez-Castells,
Chem. Eur. J. 2004, 10, 4938–4943; b) M. Mꢃndez, V. Mamane, A.
Fürstner, Chemtracts 2003, 16, 397–425.
[7] For recent general reviews on homogeneous Au catalysis, see: a) R.
Skouta, C.-J. Li, Tetrahedron 2008, 64, 4917–4938; b) B. Crone, S. F.
Kirsch, Chem. Eur. J. 2008, 14, 3514–3522; c) N. Bongers, N.
Krause, Angew. Chem. 2008, 120, 2208–2211; Angew. Chem. Int. Ed.
2008, 47, 2178–2181; d) H. C. Shen, Tetrahedron 2008, 64, 3885–
3903; 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) A. S. K. Hashmi, Catal.
Today 2007, 122, 211–214; h) D. J. Gorin, F. D. Toste, Nature 2007,
446, 395–403; i) N. T. Patil, Y. Yamamoto, ARKIVOC 2007, 6–19;
j) E. Jimꢃnez-NfflÇez, A. M. Echavarren, Chem. Commun. 2007,
333–346; k) A. S. K. Hashmi, G. J. Hutchings, Angew. Chem. 2006,
118, 8064–8105; Angew. Chem. Int. Ed. 2006, 45, 7896–7936.
[8] For reviews on enynes in Au catalysis, see: a) L. Zhang, J. Sun, S. A.
Kozmin, Adv. Synth. Catal. 2006, 348, 2271–2296; b) S. Ma, S. Yu,
Z. Gu, Angew. Chem. 2006, 118, 206–209; Angew. Chem. Int. Ed.
2006, 45, 200–203.
[9] For reviews on propargylic esters in Pt and Au catalysis, see: a) N.
Marion, S. P. Nolan, Angew. Chem. 2007, 119, 2806–2809; Angew.
Chem. Int. Ed. 2007, 46, 2750–2752; b) J. Marco-Contelles, E. Soria-
no, Chem. Eur. J. 2007, 13, 1350–1357.
[10] For seminal reports, see: a) H. Strickler, J. B. Davis, G. Ohloff, Helv.
Chim. Acta 1976, 59, 1328–1332; b) V. Rautenstrauch, J. Org.
Chem. 1984, 49, 950–952.
[11] a) E. Mainetti, V. Mouriꢂs, L. Fensterbank, M. Malacria, J. Marco-
Contelles, Angew. Chem. 2002, 114, 2236–2239; Angew. Chem. Int.
Ed. 2002, 41, 2132–2135; b) Y. Harrak, C. Blaszykowski, M. Ber-
Chem. Eur. J. 2009, 15, 3243 – 3260
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3257

S. P. Nolan, M. Malacria, L. Fensterbank, L. Cavallo et al.
nard, K. Cariou, E. Mainetti, V. Mouriꢂs, A.-L. Dhimane, L. Fen-
sterbank, M. Malacria, J. Am. Chem. Soc. 2004, 126, 8656–8657;
c) V. Mamane, T. Gress, H. Krause, A. Fürstner, J. Am. Chem. Soc.
2004, 126, 8654–8655; d) C. Blaszykowski, Y. Harrak, M.-H. Gon-
calves, J.-M. Cloarec, A.-L. Dhimane, L. Fensterbank, M. Malacria,
Org. Lett. 2004, 6, 3771–3774; e) J. Marco-Contelles, N. Arroyo, S.
Anjum, E. Mainetti, N. Marion, K. Cariou, G. Lemiꢂre, V. Mouriꢂs,
L. Fensterbank, M. Malacria, Eur. J. Org. Chem. 2006, 4618–4633.
[12] N. Marion, P. de Frꢃmont, G. Lemiꢂre, E. D. Stevens, L. Fenster-
bank, M. Malacria, S. P. Nolan, Chem. Commun. 2006, 2048–2050.
[13] For a review on N-heterocyclic carbenes in gold catalysis, see: N.
Marion, S. P. Nolan, Chem. Soc. Rev. 2008, 37, 1776–1782.
[14] For general reviews on N-heterocyclic carbenes, see: a) N-Heterocy-
clic Carbenes in Transition Metal Catalysis, Top. Organomet. Chem.,
Vol. 21 (Ed.: F. Glorius), Springer, Heidelberg, 2007; b) N-Heterocy-
clic Carbenes in Synthesis (Ed.: S. P. Nolan), Wiley-VCH, Weinheim,
2006; c) D. Bourissou, O. Guerret, F. P. Gabbai, G. Bertrand, Chem.
Rev. 2000, 100, 39–92; d) P. de Frꢃmont, N. Marion, S. P. Nolan,
Coord. Chem. Rev. 2008, DOI: 10.1016/j.ccr.2008.05.018.
[24] We isolated this type of rearranged product in the presence of PtCl2
but in very small amount (3%): C. Blaszykowski, Y. Harrak, C.
Brancour, K. Nakama, A.-L. Dhimane, L. Fensterbank, M. Malac-
ria, Synthesis 2007, 2037–2049.
[25] For references on isolated and structurally characterized [(h -RC
2
ꢀ
CR)AuI] complexes, see: a) P. Schulte, U. Behrens, Chem. Commun.
1998, 1633–1634; b) N. D. Shapiro, F. D. Toste, Proc. Natl. Acad.
Sci. USA 2008, 105, 2779–2782; c) J. Wu, P. Kroll, H. V. R. Dias,
Inorg. Chem. 2009, DOI: 10.1021/ic8020854.
[26] Even though 5-exo-dig cyclizations are more common for 1,6-
enynes, there are numerous precedents of 6-endo-dig cyclizations in
the literature: a) C. Nevado, D. J. Cµrdenas, A. M. Echavarren,
Chem. Eur. J. 2003, 9, 2627–2635; b) C. Nieto-Oberhuber, M. P.
MuÇoz, E. BuÇuel, C. Nevado, D. J. Cµrdenas, A. M. Echavarren,
Angew. Chem. 2004, 116, 2456–2460; Angew. Chem. Int. Ed. 2004,
43, 2402–2406; c) S. I. Lee, S. M. Kim, M. R. Choi, Y. K. Chung, J.
Org. Chem. 2006, 71, 9366–9372; d) C. Ferrer, M. Raducan, C.
Nevado, C. K. Claverie, A. M. Echavarren, Tetrahedron 2007, 63,
6306–6316; e) S. M. Kim, J. H. Park, S. Y. Choi, Y. K. Chung,
Angew. Chem. 2007, 119, 6284–6287; Angew. Chem. Int. Ed. 2007,
46, 6172–6175.
[15] For the synthesis and full characterization of the [(NHC)AuCl] com-
plexes, see: P. de Frꢃmont, N. M. Scott, E. D. Stevens, S. P. Nolan,
Organometallics 2005, 24, 2411–2418. [(IDD)AuCl] has not been re-
ported previously; for details see the Experimental Section and the
Supporting Information.
[27] In a review on Au-catalyzed transformations, Echavarren proposed,
for the transformation 2!5, an alternative and interesting cyclopro-
panation/1,2-OAc shift/cationic rearrangement sequence.[7j]
[16] A portion of the results presented here has been communicated pre-
[28] For selected references on alkene activation in Au catalysis, see:
a) A. S. K. Hashmi, L. Schwarz, J.-H. Choi, T. M. Frost, Angew.
Chem. 2000, 112, 2382–2385; Angew. Chem. Int. Ed. 2000, 39, 2285–
2288; b) G. Dyker, E. Muth, A. S. K. Hashmi, L. Ding, Adv. Synth.
Catal. 2003, 345, 1247–1252; c) X. Yao, C.-J. Li, J. Am. Chem. Soc.
2004, 126, 6884–6885; d) M. Alfonsi, A. Arcadi, M. Aschi, G. Bian-
chi, F. Marinelli, J. Org. Chem. 2005, 70, 2265–2273; e) C.-G. Yang,
C. He, J. Am. Chem. Soc. 2005, 127, 6966–6967; f) D. Xing, B.
Guan, G. Cai, Z. Fang, L. Yang, Z. Shi, Org. Lett. 2006, 8, 693–696;
g) J. Zhang, C.-G. Yang, C. He, J. Am. Chem. Soc. 2006, 128, 1798–
1799; h) X.-Y. Liu, C.-H. Li, C.-M. Che, Org. Lett. 2006, 8, 2707–
2710; i) C. F. Bender, R. A. Widenhoefer, Chem. Commun. 2006,
4143–4144; j) C. F. Bender, R. A. Widenhoefer, Org. Lett. 2006, 8,
5303–5305; k) R.-V. Nguyen, X. Yao, C.-J. Li, Org. Lett. 2006, 8,
2397–2399; l) N. Marion, R. Gealageas, S. P. Nolan, Org. Lett. 2007,
9, 2653–2656 [Additions/Corrections:N. Marion, R. Gealageas, S. P.
Nolan, Org. Lett. 2008, 10, 1037]; m) R. Skouta, C.-J. Li, Angew.
Chem. 2007, 119, 1135–1137; Angew. Chem. Int. Ed. 2007, 46, 1117–
1119. For theoretical studies on Au-catalyzed alkene activation, see:
n) G. Kovµcs, G. Ujaque, A. Lledꢀs, J. Am. Chem. Soc. 2008, 130,
853–864; o) C. Gourlaouen, N. Marion, S. P. Nolan, F. Maseras, Org.
Lett. 2009, DOI: 10.1021/ol802430m.
viously.[12]
[17] We previously observed, in the context of Au-catalyzed cycloisome-
rization reactions, disproportionation of AuCl into Au0 and AuCl3 in
CH2Cl2 solution: a) G. Lemiꢂre, V. Gandon, N. Agenet, J.-P. God-
dard, A. de Kozak, C. Aubert, L. Fensterbank, M. Malacria, Angew.
Chem. 2006, 118, 7758–7761; Angew. Chem. Int. Ed. 2006, 45, 7596–
7599. See also: b) A. Vogler, H. Kundely, Coord. Chem. Rev. 2001,
219–221, 489–507; c) T. C. Wabnitz, J.-Q. Yu, J. B. Spencer, Chem.
Eur. J. 2004, 10, 484–493.
[18] For reviews on the stereoelectronic parameters of NHCs, see: a) S.
Díez-Gonzµlez, S. P. Nolan, Coord. Chem. Rev. 2007, 251, 874–883;
b) L. Cavallo, A. Correa, C. Costabile, H. Jacobsen, J. Organomet.
Chem. 2005, 690, 5407–5413; c) T. Strassner, Top. Organomet.
Chem. 2004, 7, 1–20.
[19] For a seminal report on AuI-catalyzed addition of alcohols to al-
kynes, see: a) J. H. Teles, S. Brode, M. Chabanas, Angew. Chem.
1998, 110, 1475–1478; Angew. Chem. Int. Ed. 1998, 37, 1415–1418.
For a general review on gold catalysis in the presence of alcohols,
see: b) J. Muzard, Tetrahedron 2008, 64, 5815–5849.
[20] For theoretical calculations on this issue, see: a) E. Soriano, P. Bal-
lesteros, J. Marco-Contelles, Organometallics 2005, 24, 3182–3191;
b) E. Soriano, J. Marco-Contelles, J. Org. Chem. 2005, 70, 9345–
9353; c) O. Nieto Faza, C. Silva Lꢀpez, R. lvarez, A. R. de Lera, J.
Am. Chem. Soc. 2006, 128, 2434–2437; d) E. Soriano, J. Marco-Con-
telles, J. Org. Chem. 2007, 72, 1443–1448.
[21] In the case of an intermolecular cyclopropanation between a propar-
gylic ester and an olefin, Toste et al. showed that a migration then
cyclopropanation sequence applied: M. Johansson, D. J. Gorin, S. T.
Staben, F. D. Toste, J. Am. Chem. Soc. 2005, 127, 18002–18003.
[22] This method has already been used by Fürstner and Hannen and by
Fehr and Galindo independently: a) A. Fürstner, P. Hannen, Chem.
Eur. J. 2006, 12, 3006–3019; b) C. Fehr, J. Galindo, Angew. Chem.
2006, 118, 2967–2970; Angew. Chem. Int. Ed. 2006, 45, 2901–2904.
For a theoretical approach, see: c) E. Soriano, J. Marco-Contelles, J.
Org. Chem. 2007, 72, 2651–2654. For a related example of chirality
transfer, see d) Y. Zou, D. Garayalde, Q. Wang, C. Nevado, A.
Goeke, Angew. Chem. 2008, 120, 10264–10267; Angew. Chem. Int.
Ed. 2008, 47, 10110–10113.
[23] Control reactions on possible isomerization between the different
products were performed. No isomerization was observed, and this
notably excludes a plausible pathway involving cyclopropane ring
opening: a) A. E. Hours, J.K. Snyder, Tetrahedron Lett. 2006, 47,
675–678; b) A. E. Hours, J.K. Snyder, Organometallics 2008, 27,
410–417.
[29] For isolated AuI adducts of olefins, see: a) D. Belli Dell’Amico, F.
Calderazzo, R. Dantona, J. Strähle, H. Weiss, Organometallics 1987,
6, 1207–1210; b) R. M. Davila, R. J. Staples, J. P. Fackler, Jr., Orga-
nometallics 1994, 13, 418–420; c) M.A. Cinellu, G. Minghetti, F.
Cocco, S. Stoccoro, A. Zucca, M. Manassero, Angew. Chem. 2005,
117, 7052–7055; Angew. Chem. Int. Ed. 2005, 44, 6892–6895;
d) M.A. Cinellu, G. Minghetti, F. Cocco, S. Stoccoro, A. Zucca, M.
Manassero, M. Arca, Dalton Trans. 2006, 5703–5716; e) H. V. R.
Dias, J. Wu, Angew. Chem. 2007, 119, 7960–7962; Angew. Chem.
Int. Ed. 2007, 46, 7814–7816; f) H. V. R. Dias, M. Fianchini, T. R.
Cundari, C. F. Campana, Angew. Chem. 2008, 120, 566–569; Angew.
Chem. Int. Ed. 2008, 47, 556–559; g) P. de Frꢃmont, N. Marion, S. P.
Nolan, J. Organomet. Chem. 2009, DOI: 10.1016/j.jorgan-
chem.2008.10.047. For a review, see: h) H. V. R. Dias, J. Wu, Eur. J.
Inorg. Chem. 2008, 509–522.
[30] Au-catalyzed cyclization of 1,4-allenenyl esters has been described
previously for related systems, but no migration of the ester group
was observed: a) A. Buzas, F. Gagosz, J. Am. Chem. Soc. 2006, 128,
12614–12615. For other reports on 1,4-allenene cyclization in the
presence of gold catalysts, see: b) X. Huang, L. Zhang, J. Am.
Chem. Soc. 2007, 129, 6398–6399; c) X. Huang, L. Zhang, Org. Lett.
2007, 9, 4627–4630.
3258
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3243 – 3260

Cycloisomerization of Enynyl Esters versus Allenenyl Esters
FULL PAPER
[31] For the formation of allenyl esters from propargylic esters using sil-
ver(I) salts, see: a) G. Saucy, R. Marbet, H. Lindlar, O. Isler, Helv.
Chim. Acta 1959, 42, 1945–1955; b) W. R. Benn, J. Org. Chem.
1968, 33, 3113–3118; c) H. Schlossarczyk, W. Sieber, M. Hesse, H.-J.
Hansen, H. Schmid, Helv. Chim. Acta 1973, 56, 875–944.
[32] We previously predicted by DFT calculations that, in the Au-cata-
lyzed isomerization [gold propargyl acetates IQgold carbenoid vinyl
acetates IIQgold allenyl acetates III], a [(NHC)Au] species would
favor the allene derivative, while a [(PR3)Au] species would isomer-
ize more easily: A. Correa, N. Marion, L. Fensterbank, M. Malacria,
S. P. Nolan, L. Cavallo, Angew. Chem. 2008, 120, 730–733; Angew.
Chem. Int. Ed. 2008, 47, 718–721.
mediate, see: L. Peng, X. Zhang, S. Zhang, J. Wang, J. Org. Chem.
2007, 72, 1192–1197.
[43] Nevertheless, a clear explanation accounting for the great differen-
ces in selectivity between pre-isolated allene and allene formed in
situ is still needed. We believe that, during the isomerization process
from the alkyne to the allene, the gold center is ideally bound to the
allene moiety to catalyze another migration of the acetate group.
On the other hand, with a preformed allene, coordination of the
gold center might be slightly different, and lead preferentially to
cyclization with the alkene. Further studies clarifying the coordina-
tion behavior of gold toward allenes, and more specifically allenyl
esters, are needed. See ref. [44].
[44] For a rare report on coordination of gold to allenes, see: a) J. B.
Chenier, J.A. Howard, B. Mile, J. Am. Chem. Soc. 1985, 107, 4190–
4191. For reports on the coordination of platinum(II) to allenes, see:
b) K. Vrieze, H. C. Volger, P. Praat, J. Organomet. Chem. 1970, 21,
467–475; c) F. L. Bowden, R. Giles, Coord. Chem. Rev. 1976, 20,
81–106; d) A. De Renzi, B. Di Blasio, A. Panunzi, C. Pedone, A.
Vitagliano, J. Chem. Soc. Dalton Trans. 1978, 1392–1397; e) P. Salva-
dori, G. Uccello-Barretta, R. Lazzaroni, A. M. Caporusso, J. Chem.
Soc. Chem. Commun. 1990, 1121–1123.
[45] 1,3-OAc migration in allenyl acetate substrates leading to formation
of propargylic acetates has been observed as a minor reaction path
in the Au-catalyzed cyclization of allenyl esters to indenyl esters: N.
Marion, S. Díez-Gonzµlez, P. de Frꢃmont, A. R. Noble, S. P. Nolan,
Angew. Chem. 2006, 118, 3729–3732; Angew. Chem. Int. Ed. 2006,
45, 3647–3650. See also ref. [37].
[46] a) K. Miki, K. Ohe, S. Uemura, Tetrahedron Lett. 2003, 44, 2019–
2022; b) K. Miki, K. Ohe, S. Uemura, J. Org. Chem. 2003, 68, 8505–
8513; c) K. Miki, M. Fujita, S. Uemura, K. Ohe, Org. Lett. 2006, 8,
1741–1743; d) Y. Ikeda, M. Murai, T. Abo, K. Miki, K. Ohe, Tetra-
hedron Lett. 2007, 48, 6651–6654; e) Y. Nakanishi, K. Miki, K. Ohe,
Tetrahedron 2007, 63, 12138–12148.
[47] For an interesting account on the formation of vinyl carbenoid spe-
cies from alkynes mediated by transition metals, see: a) K. Miki, S.
Uemura, K. Ohe, Chem. Lett. 2005, 34, 1068–1073. For a seminal
report, see: b) K. Miki, F. Nishino, K. Ohe, S. Uemura, J. Am.
Chem. Soc. 2002, 124, 5260–5261.
[48] Alternatively, and in agreement to our proposal of fast interconver-
sion alkyne/carbene/allene, 14 could be reached via the formation of
the corresponding alkyne upon 1,3-OAc shift from 11 and subse-
quent 1,2-shift. We thank a referee for this suggestion.
[49] X. Moreau, J.-P. Goddard, M. Bernard, G. Lemiꢂre, J. M. Lꢀpez-
Romero, E. Mainetti, N. Marion, V. Mouriꢂs, S. Thorimbert, L. Fen-
sterbank, M. Malacria, Adv. Synth. Catal. 2008, 350, 43–48.
[50] Cyclization of 20 with the same catalyst but under neat conditions
afforded, after 16 h, 59% of a mixture of 21/22 (1/2).
[51] For the observation of formal enyne ring-closing metathesis prod-
ucts in Au-catalyzed cycloisomerization, see: N. Cabello, C. Rodrí-
guez, A. M. Echavarren, Synlett 2007, 1753–1758. See also
Ref. [26b].
[52] Cyclization of 23 with the same catalyst but under neat conditions
afforded 51% of a mixture of 24/25 (3/1) along with 13% of starting
material 23 and 13% of the [3,3]-rearranged allenyl acetate.
[53] We previously reported a PtCl2-catalyzed 1,3-OAc shift followed by
[3,3]-sigmatropic rearrangement of 1,5-allenenes. In this case, [3,3]-
sigmatropic rearrangement was experimentally shown to be ther-
mal.[35]
[33] We previously reported the preference of gold catalysts for the cycli-
zation of 1,3-allenenes over 1,5-allenenes: a) V. Gandon, G. Le-
miꢂre, A. Hours, L. Fensterbank, M. Malacria, Angew. Chem. 2008,
120, 7644–7648; Angew. Chem. Int. Ed. 2008, 47, 7534–7538; b) G.
Lemiꢂre, V. Gandon, K. Cariou, T. Fukuyama, A.-L. Dhimane, L.
Fensterbank, M. Malacria, Org. Lett. 2007, 9, 2207–2209.
[34] To the best of our knowledge, there is no report on Au-catalyzed re-
actions of 1,5-allenenes. For Au-catalyzed cyclization of 1,3-alle-
nenes, see: a) J. H. Lee, F. D. Toste, Angew. Chem. 2007, 119, 930–
932; Angew. Chem. Int. Ed. 2007, 46, 912–914 and ref. [33]. For Au-
catalyzed cyclization of 1,4-allenenes, see ref. [30]. For Au-catalyzed
cyclization of 1,6-allenenes, see: b) M. R. Luzung, P. Mauleꢀn, F. D.
Toste, J. Am. Chem. Soc. 2007, 129, 12402–12403; c) M. A. Tarselli,
A. R. Chianese, S. L. Lee, M. R. Gagnꢃ, Angew. Chem. 2007, 119,
6790–6793; Angew. Chem. Int. Ed. 2007, 46, 6670–6673; d) M. A.
Tarselli, M. R. Gagnꢃ, J. Org. Chem. 2008, 73, 2439–2441.
[35] K. Cariou, E. Mainetti, L. Fensterbank, M. Malacria, Tetrahedron
2004, 60, 9745–9755.
[36] For examples of Au-mediated 1,3-shift of propargylic esters, see:
a) L. Zhang, J. Am. Chem. Soc. 2005, 127, 16804–16805; b) A.
Buzas, F. Istrate, F. Gagosz, Org. Lett. 2006, 8, 1957–1959; c) S.
Wang, L. Zhang, J. Am. Chem. Soc. 2006, 128, 8414–8415; d) S.
Wang, L. Zhang, Org. Lett. 2006, 8, 4585–4587; e) S. Wang, L.
Zhang, J. Am. Chem. Soc. 2006, 128, 14274–14275; f) H.-S. Yeom,
S.-J. Yoon, J. Shin, Tetrahedron Lett. 2007, 48, 4817–4820; g) T. Luo,
S. L. Schreiber, Angew. Chem. 2007, 119, 8398–8401; Angew. Chem.
Int. Ed. 2007, 46, 8250–8253; h) A. S. Dudnik, T. Schwier, V. Ge-
vorgyan, Org. Lett. 2008, 10, 1465–1468.
[37] Recently, Gevorgyan et al. studied the isomerization of propargylic
acetates and phosphates by 17O labeling experiments in the context
of cycloisomerization of ynones to furans: T. Schwier, A. W.
Sromek, D. M. L. Yap, D. Chernyak, V. Gevorgyan, J. Am. Chem.
Soc. 2007, 129, 9868–9878.
[38] We recently reported on this phenomenon of back-and-forth isomer-
ization of propargylic acetates to allenyl acetates in the presence of
gold, a phenomenon for which we coined the phrase “Golden Car-
ousel”.[32]
[39] A referee evoked the possibility of ionization of the propargylic car-
boxylate leading to scrambling. We cannot, at the present time, un-
equivocally exclude such a possibility. However, the following points
are indicative of the absence of ionization of the propargylic ace-
tate: 1) There is no solvolysis of the acetate group in MeOH under
gold catalysis [see Eq. (1)]. 2) Gevorgyan et al. did not observe ioni-
zation with gold catalysts.[37] 3) Under platinum catalysis, the ee of
enantioenriched starting material 2-Ac at half-conversion showed no
racemization of the substrate. This precludes ionization of the start-
ing propargyl carboxylate under these reaction conditions. We thank
this referee for this insightful remark.
[54] E. Herrero-Gꢀmez, C. Nieto-Oberhuber, S. Lꢀpez, J. Benet-Buch-
holz, A. M. Echavarren, Angew. Chem. 2006, 118, 5581–5585;
Angew. Chem. Int. Ed. 2006, 45, 5455–5459.
[40] Note that the architecture of the substrates of Gevorgyan et al. dif-
fers greatly from ours. The presence of a hydrogen atom at the
propargylic position, available for 1,3-H shift, and the particular
structural features of conjugated ynones allow for mechanistic path-
ways distinct from those proposed here.[37]
[41] S. Hotha, S. Kashyap, J. Am. Chem. Soc. 2006, 128, 9620–9621.
[42] For an interesting report on Au-catalyzed 1,2-migration of sulfur in
propargylic sulfides, presumably occurring via a thiirenium inter-
[55] For details on the synthesis of [(IDD)AgCl], see: P. de Frꢃmont,
N. M. Scott, E. D. Stevens, T. Ramnial, O. C. Lightbody, C. L. B.
Macdonald, J.A. C. Clyburne, C. D. Abernethy, S. P. Nolan, Organo-
metallics 2005, 24, 6301–6309.
[56] M. Nakamura, A. Hirai, M. Sogi, E. Nakamura, J. Am. Chem. Soc.
1998, 120, 5846–5847.
[57] Gaussian 03, Revision C.02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M.A. Robb, J. R. Cheeseman, J.A. Montgomer-
Chem. Eur. J. 2009, 15, 3243 – 3260
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3259

S. P. Nolan, M. Malacria, L. Fensterbank, L. Cavallo et al.
y, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyen-
gar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N.
Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J.
Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C.
Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J.V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cio-
slowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaro-
mi, R. L. Martin, D. J. Fox, T. Keith, M.A. Al-Laham, C. Y. Peng,
A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wall-
ingford CT, 2004.
[58] a) A. D. Becke, Phys. Rev. A 1988, 38, 3098–3100; b) J. P. Perdew,
Phys. Rev. B 1986, 33, 8822–8824; c) J. P. Perdew, Phys. Rev. B 1986,
34, 7406.
[59] a) A. Schaefer, H. Horn, R. Ahlrichs, J. Chem. Phys. 1992, 97, 2571–
2577; b) A. Schaefer, C. Huber, R. Ahlrichs, J. Chem. Phys. 1994,
100, 5829–5835.
[60] a) U. Häusermann, M. Dolg, H. Stoll, H. Preuss, Mol. Phys. 1993,
78, 1211–1224; b) W. Kuechle, M. Dolg, H. Stoll, H. Preuss, J.
Chem. Phys. 1994, 100, 7535; c) T. Leininger, A. Nicklass, H. Stoll,
M. Dolg, P. Schwerdtfeger, J. Chem. Phys. 1996, 105, 1052–1059.
[61] a) M. Cossi, V. Barone, R. Cammi, J. Tomasi, Chem. Phys. Lett.
1996, 255, 327–335; b) M. T. Cancꢂs, B. Mennucci, J. Tomasi, J.
Chem. Phys. 1997, 107, 3032–3041; c) M. Cossi, V. Barone, B. Men-
nucci, J. Tomasi, Chem. Phys. Lett. 1998, 286, 253–260.
Received: July 8, 2008
Revised: October 22, 2008
Published online: February 6, 2009
3260
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 3243 – 3260