An unusual access to medium sized cycloalkynes by a new goId(I)-catalysed cycloisomerisation of diynes

Article abstract of DOI:10.1002/chem.200901312

A study was conducted to demonstrate efficient cycloisomerization of a series of 1,9- and 1,10-diynes into medium sized cycloalkynes by a gold-catalyzed alkyne-alkyne coupling. The reaction was performed using 4 mol% of gold complex in CD₂Cl₂ at room temperature and was monitored by using ¹H NMR spectroscopy. The reaction of analogous diyne substrate was more rapid and a complete conversion was observed after 40 hours at room temperature. The cycloalkyne was isolated as a solid in 95% yield from which single crystals suitable for X-ray crystal structure determination was obtained. A rapid screening of catalysts and experimental conditions was undertaken to optimize the formation of cycloalkyne. A mechanistic proposal was also introduced for the cycloisomerization of diynes into cycloalkynes on the basis of the experimental observations.

Full text of DOI:10.1002/chem.200901312

DOI: 10.1002/chem.200901312
An Unusual Access to Medium Sized Cycloalkynes by a New
Gold(I)-Catalysed Cycloisomerisation of Diynes
Yann Odabachian,[a] Xavier F. Le Goff,[b] and Fabien Gagosz*[a]
Dedicated with admiration to Professor Samir Z. Zard
Over the last decade, gold catalysis has demonstrated a
high synthetic potential for the formation of various func-
tionalised structures by the addition of a wide range of nu-
cleophiles to gold-activated alkynes or allenes.[1] While a
series of carbon nucleophiles (alkene, allene, aryl) have
been used for this purpose, it is surprising that only little at-
tention has been given to the use of alkynes.[2] Such a lack
of interest is probably due to the inherent reduced nucleo-
philicity of the alkyne functionality combined with con-
straints related to its linear geometry in the case of intramo-
lecular transformations. Herein we report that a series of
1,9- and 1,10-diynes can indeed be efficiently cycloisomer-
ised into medium sized cycloalkynes by an apparently un-
Scheme 1. First experimental observation of an unprecedented gold(I)-
catalysed alkyne–alkyne coupling. Yield of 2a: 53% in CD2Cl2 and 79%
in CDCl3.
precedented
(Scheme 1)
gold-catalysed
alkyne–alkyne
coupling
In the course of our recent study of the [2+2] cycloaddi-
tion of 1,8-enynes,[3,4] we attempted to transform symmetri-
cal substrate 1a into the corresponding cyclobutene 3. The
reaction was performed by using 4 mol% of gold complex
4[5] in CD2Cl2 at room temperature and was monitored by
formed that could be accumulated to a maximum of 53%
yield after three days of reaction (79% yield in CDCl3).[6]
This new compound could not be isolated in a pure form[7]
but its structure could, however, be assigned as cycloalkyne
2a by NMR spectroscopic analysis of the crude reaction
mixture.[8] This unexpected gold-catalysed alkyne–alkyne
coupling is remarkable given the mildness of the reaction
conditions under which cycloalkyne 2a is formed.[9] This
highly unsaturated compound did not suffer further transfor-
mation and was stable in the presence of electrophilic gold
complex 4. It should also be noted that the formation of cy-
cloalkyne 2a from diyne 1a is rather unique in the field of
gold catalysis, in view of the extremely limited number of
known transformations that involve alkynes as nucleo-
philes[2] or lead to medium-sized ring products.[10]
The reaction of analogous diyne substrate 1b was more
rapid and a complete conversion was observed after 40 h at
room temperature (Table 1, entry 1). The corresponding cy-
cloalkyne 2b was isolated as a solid in 95% yield, from
which single crystals suitable for X-ray crystal structure de-
termination could be obtained.[11] A rapid screening of cata-
lysts and experimental conditions was next undertaken to
optimise the formation of cycloalkyne 2b (Table 1). Heating
1
using H NMR spectroscopy. The conversion of substrate 1a
was surprisingly slow, whereas a series of other 1,8-enynes
reacted rapidly to give the corresponding [2+2] cycloaddi-
tion products.[3] Moreover, no trace of cyclobutene forma-
tion could be observed. Instead, another product was cleanly
[a] Y. Odabachian, Dr. F. Gagosz
Y. Odabachian, F. Gagosz
Laboratoire de Synthꢀse Organique
UMR 7652, CNRS, Ecole Polytechnique
91128 Palaiseau (France)
Fax : (+33)169-33-38-51
E-mail: gagosz@dcso.polytechnique.fr
[b] Dr. X. F. Le Goff
Laboratoire Heteroelements et Coordination
UMR 7653, CNRS, Ecole Polytechnique
91128 Palaiseau (France)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200901312.
8966
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 8966 – 8970

COMMUNICATION
Table 1. Optimisation of the reaction conditions.[a]
Table 2. AuI-catalysed formation of ten-membered cycloalkynes by cyclo-
isomerisation of symmetrical diynes.[a]
Entry
Substrate[b]
Product
t [h]
Yield[c] [%]
1
2
3
1a R=Me
1c R=Ph
1d R=p-FC6H4
2a
2c
2d
8
17
2
85
95
89
Entry
Catalyst
T [8C]
t [h]
Yield[b] [%]
1
2
3
4
5
6
4
4
20
65
65
65
20
65
40
2
24
24
24
24
95
98
12[c]
–
–
–
[Au
[Au
U
E
ACHTUNGTRENNUNG
HNTf2
HNTf2
4
5
1e n=1
1 f n=2
2e
2 f
48
18
70[d,e]
82
[a] Reaction monitored by 1H NMR spectroscopy; X=C
ACTHNUGTRNEGNU(CO2Me)2.
[b] Isolated yields. [c] Yield determined by using NMR spectroscopy.
the reaction mixture to 658C in the presence of 4 mol% of
gold complex 4 increased the rate of the reaction (2 h) with-
out affecting the yield (Table 1, entry 2), whereas the use of
triphenylphosphine–gold complexes led to poor or no con-
version (Table 1, entries 3–4). A control experiment per-
formed in the presence of HNTf2 also showed that the for-
mation of 2b was not catalysed by a Brønsted acidic species
(Table 1, entries 5–6).
To explore the scope of this new transformation, a series
of symmetrical diynes 1a–i were subjected to the optimal re-
action conditions (4 mol% of 4 in CDCl3 heated to reflux).
The reaction proved to be general, and cycloalkynes 2a–i
could be isolated in yields ranging from 70% to 98%
(Table 2). The formation of cycloalkyne 1a could be im-
proved to a yield of 85% with a reduced reaction time
(Scheme 1 vs. Table 2, entry 1). The methyl substituent on
the alkene moiety could be substituted by an aromatic ring
(Table 2, entries 2, 3) and the esters on the linker by diace-
toxymethyl groups whithout loss of efficiency (Table 2,
entry 7). Other carbocyclic (Table 2, entries 4, 5) or aromatic
residues (Table 2, entries 6–9) were also tolerated at posi-
6
7
8
9
1b R=H
1g R=H[f]
1h R=F
2b
2g
2h
2i
2
1
1
8
98
90
98
81
1i R=OMe
[a] Reaction conditions: Catalyst 4 (4 mol%), c=0.1m, CDCl3, reflux.
[b] X=C(CO2Me)2. [c] Isolated yields. [d] 8 mol% of catalyst 4. [e] Yield
determined by using NMR spectroscopy, not isolated. [f] X=C-
(CH2OAc)2.
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
entries 1–3).[13] Interestingly, the removal of an unsaturated
bond on one alkynyl chain did not affect the course and the
efficiency of the reaction but greatly increased the selectivi-
ty (Table 3, entries 4–7).[14] For instance, diynes 1m,n, which
have an aromatic ring, were highly selectively transformed
into cycloalkynes 2m,n (Table 3, entries 4, 5), whereas cyclo-
alkyne 2o was the only detectable product in the reaction of
diyne 1o (Table 3, entry 6). To further highlight the synthet-
ic possibilities of the process, the transformation was at-
tempted with diyne substrates in which one alkynyl chain
has been shortened by one carbon (Table 3, entries 8–10).
We were pleased to see that a series of nine-membered cy-
cloalkynes 2q–s could also be produced in high yields (82–
88%), even if the selectivities were lower than those ob-
tained for the formation of the analogous ten-membered cy-
cloalkynes (Table 3, entries 5, 7 vs. entries 8, 9). It is interest-
ing to note that all the transformations attempted (Table 2
and Table 3) specifically furnished alkyne–alkyne coupling
products with no detectable formation of byproducts de-
rived from a cyclisation onto an aromatic ring or an alkene
moiety. While the transformation proved to be fairly gener-
al, limits in reactivity were found with symmetrical diyne 1t
and internal alkynes 1u,v, which failed to afford the corre-
sponding cycloalkynes (Figure 1).
ꢀ
ꢀ
tions C(3) C(4) and C(8) C(9) of the 1,10-enyne. Surprins-
ingly, the conversion of diyne 1e, which has a cyclopentenyl
motif, proved to be more difficult and a higher loading of
catalyst was required to produce the corresponding cycloal-
kyne 2e in 70% yield.[12] It is also interesting to note the in-
fluence of the aromatic ring substituents on the course of
the reaction (Table 2, entries 2, 3 and 6, 8, 9). The presence
of an electron-withdrawing group (fluorine) tended to speed
up the reaction whereas a slower reaction was observed
with an electron-donating group (methoxy).
We next focused our attention on the case of unsymmetri-
cal diynes. The results of this study are shown in Table 3.
Unsymmetrical diynes 1j–s reacted cleanly and gave isomer-
ic mixtures of cyclodiynes 2j–s and 2’j–s in good yield (76–
93%) with modest to excellent selectivity.
The transformations of substrates 1j–l, which have two
unsaturated bonds conjugated to the alkyne functionalities,
were rapid and efficient but the selectivity was low (Table 3,
To gain more insight into the reaction mechanism, deuter-
ated symmetrical diyne [D]-1i was subjected to the opti-
mised conditions and its conversion monitored by H NMR
1
spectroscopy (Table 4).
Chem. Eur. J. 2009, 15, 8966 – 8970
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8967

F. Gagosz et al.
Table 3. AuI-catalysed formation of nine- and ten-membered cycloalkynes by cycloisomerisation of unsymmet-
rical diynes.[a]
The electrophilic gold(I) acti-
vation of the alkyne in sub-
strate 5 could lead to the for-
mation of gold acetylide species
7.[17] The involvement of 7 is
supported by the decrease in
the deuterium label in unreact-
ed diyne [D]-1i, which may be
attributed to the presence of
traces of water in the solvent
Entry
Substrate[b]
Products
t [h]
Yield[c] [%]
2:2’[d]
1
2
3
1j R1 =H, R2 =Me
1k R1–R2 =-(CH2)4-
1l R1–R2 =-(CH2)3-
2j
2k
2l
2’j
2’k
2’l
6
0.5
1
85
85
93
1:1
1:1.5
1:2.4
(see Table 4).[18,19]
A further
gold(I) activation of the free
alkyne in 7 would induce the
nucleophilic addition of the
gold acetylide and the forma-
tion of vinylic carbocation 9
possessing two vinyl–gold units.
The alkyne group is generally
considered to be a poor nucleo-
phile, but the formation of the
gold acetylide could enhance its
nucleophilic character and thus
allow the formation of the new
4
5
1m R=H
1n R=F
2m
2n
2’m
2’n
6
6
86
84
10:1
20:1
6
7
1o R1 =H, R2 =Ph
1p R1–R2 =-(CH2)4-
2o
2p
2’o
2’p
13
5
87
76
1:0
5.7:1
ꢀ
C C bond. A subsequent loss
8
1q
2q
2’q
5
88
3.3:1
of one cationic gold species
would then give vinyl–gold in-
termediate 10. A final protode-
metalation of 10 by a protic
source from the reaction
medium[20] would produce cy-
cloalkyne 11 with regeneration
of the second cationic gold spe-
cies. This mechanism is in
agreement with the stereoiso-
9
10
1r R1–R2 =-(CH2)4-
2r
2s
2’r
2’s
5
3.5
88
82
1.7:1
5.2:1
1s R1 =H, R2 =p-FC6H4
[a] Reaction conditions: Catalyst 4 (4 mol%), c=0.1m, CDCl3, reflux. [b] X=C
[d] Isomeric ratio determined by using H NMR spectroscopy.
(CO2Me)2. [c] Isolated yields.
1
topic distribution observed in cycloalkyne [D]-2i. The per-
centage deuteration at position a of the alkene in [D]-2i was
indeed found to be comparable with that of unreacted diyne
[D]-1i during the whole process (compare column 2 with
column 8 in Table 4).[21] The protodemetalation process (10
to 11) is also supported by the low percentage of deutera-
tion observed at position b of the alkene in [D]-2i (see
Table 4, column 9).[22,23]
Figure 1. Unsuitable substrates for the formation of cycloalkynes.
X=CACHTUNGTRENNUNG(CO2Me)2.
The probable greater preference of cationic gold species
to coordinate to gold acetylides rather than to alkynes (12
vs. 8) does not, however, favour this mechanism.[24] An alter-
native mechanism involving the nucleophilic addition of an
alkyne to a gold(I)-activated gold acetylide (12 to 13) could
also be envisaged (Scheme 3).[25,26] Similar gold species have
been recently proposed by Toste, Houk et al. for the cyclo-
After reacting for 6 h, cycloalkyne [D]-2i was formed in
85% yield but with a great decrease in the amount of deute-
rium label (61% D on average). Surprisingly, stereoisotopic
isomer [Da,Db]-2i was not the major component. The stereo-
isotopic distribution of cycloalkyne [D]-2i[15] slowly evolved
with time, with a decrease in the major isomer [Da,Hb]-2i
and an increase in isomer [Da,Db]-2i. However, the propor-
tions of the two minor isomers, [Ha,Db]-2i and [Ha,Hb]-2i,
remain constant. Intriguingly, part of the deuterium label on
unreacted starting diyne [D]-1i was washed out after 6 h
(73% D on average).
AHCTUNGTRENNUNG
cycloalkynes could be accessed by an unusual gold(I)-cata-
lysed cycloisomerisation of 1,9- and 1,10-diynes. This study
highlights the dual reactivity of the alkyne functionality,
which can be used in gold catalysis as either a nucleophile
(perhaps as a gold acetylide) or an electrophile (gold(I)-acti-
vated alkyne). Further studies into the use of alkynes as nu-
On the basis of these experimental observations, a tenta-
tive mechanistic proposal for the cycloisomerisation of
diynes into cycloalkynes is presented in Scheme 2.[16]
8968
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 8966 – 8970

Medium-Sized Cycloalkynes
COMMUNICATION
Table 4. Summary of the reaction of symmetrical diyne [D]-1i (X=C
C
1H NMR spectroscopy. Once the sub-
strate has been consumed (8 h), the re-
action was quenched by the addition
of triethylamine, concentrated in
vacuo and purified by flash chroma-
tography (SiO2, PE/AcOEt 9:1) to
give
cycloalkyne
2a
(12.2 mg).
1HNMR (400 MHz, CDCl3): d=5.69
(td, J=7.9 Hz, J=1.5 Hz, 1H), 5.49 (s,
1H), 5.45 (s, 1H), 5.06 (dd, J=
12.7 Hz, J=5.2 Hz, 1H), 3.77 (s, 3H),
3.73 (s, 3H), 3.65 (dd, J=14.3 Hz, J=
12.8 Hz, 1H), 2.96 (dd, J=13.2 Hz, J=
8.9 Hz, 1H), 2.69 (dd, J=13.2 Hz, J=
7.0 Hz, 1H), 2.54 (ddt, J=14.3 Hz, J=
3.5 Hz, J=1.7 Hz, 1H), 1.95 (s, 3H),
1.81 ppm (s, 3H); 13C NMR (100 MHz,
CDCl3): d=170.3, 170.0, 136.3, 133.7,
128.4, 124.1, 123.7, 119.6, 95.9, 94.3,
58.7, 51.6, 51.4, 32.1, 30.6, 23.6,
19.7 ppm. IR (CCl4): n˜ = 2953, 1742,
1440, 1205 cmꢀ1; MS (EI+): m/z: 288
[M]+, 256, 229, 213; MS (HRMS EI):
m/z calcd for C17H20O4: 288.1372;
found: 288.1372.
t [h]
Unreacted[a]
[D]-1i [% D]
[D]-2i[a]
[%]
Stereotopic distribution of
[D]-2i [%][a]
ACTHNGUTERNNUG
[Da,Hb] [Ha,Db] [Ha,Hb]
Da [%][a]
Db [%][a]
[Da,Db]
R
G
0
0.5
1
2
3.5
6
100
85
84
82
77
73
0
–
–
–
4
6
6
7
8
–
–
–
32
48
58
73
85
25
27
30
32
34
59
56
52
49
46
12
11
12
12
12
84
83
82
81
80
29
33
36
39
42
[a] Determined by using NMR spectroscopy.
Acknowledgements
The authors thank C. Gronnier and R. Guignard for their participation
to this work, Prof. S. Z. Zard and Dr. C. Fehr for stimulating discussions
and Rhodia Chimie Fine for a generous gift of HNTf2.
Keywords: alkyne–alkyne coupling
·
cycloalkynes
·
cycloisomerization · gold · homogeneous catalysis
[1] For recent reviews on gold catalysis, see: a) V. Michelet, P. Y. Toul-
lec, J. P. GenÞt, Angew. Chem. 2008, 120, 4338–4386; Angew. Chem.
Int. Ed. 2008, 47, 4268–4315; b) E. Jimꢂnez-NfflÇez, A. M. Echavar-
ren, Chem. Rev. 2008, 108, 3326–3350; c) Z. Li, C. Brower, C. He,
Chem. Rev. 2008, 108, 3239–3245; d) A. Arcadi, Chem. Rev. 2008,
108, 3266–3325; e) D. J. Gorin, F. D. Toste, Chem. Rev. 2008, 108,
3351–3378; f) A. S. K. Hashmi, M. Rudolph, Chem. Soc. Rev. 2008,
37, 1766–1775; g) A. S. K. Hashmi, Chem. Rev. 2007, 107, 3180–
3211; h) A. Fürstner, P. W. Davies, Angew. Chem. 2007, 119, 3478–
3519; Angew. Chem. Int. Ed. 2007, 46, 3410–3449; i) E. Jimꢂnez-
Nœnez, A. M. Echavarren, Chem. Commun. 2007, 333–346; j) L.
Zhang, J. Sun, S. A. Kozmin, Adv. Synth. Catal. 2006, 348, 2271–
2296.
Scheme 2. Mechanistic proposal for the formation of cycloalkyne via a
double activation by the gold catalyst.
Scheme 3. Alternative mechanistic proposal.
[2] The [3+2] cycloaddition of arenyne–yne functionalities leading to
polyclic products that feature a final 1,3-diene unit has been report-
ed, see: J.-J. Lian, P.-C. Chen, Y.-P. Lin, H.-C. Ting, R.-S. Liu, J. Am.
Chem. Soc 2006, 128, 11372–11373; for a similar transformation,
see: T. Shibata, R. Fujiwara, D. Takano, Synlett 2005, 2062–2064.
[3] Y. Odabachian, F. Gagosz, Adv. Synth. Catal. 2009, 351, 379–386.
[4] For earlier contributions by our group in the field of gold(I)-cata-
lysed cycloisomerisation, see: 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.
[5] N. Mꢂzailles, L. Ricard, F. Gagosz, Org. Lett. 2005, 7, 4133–4136.
[6] The conversion of 1a reached a maximum of 84% after 4 d in
CDCl3.
cleophiles in gold-catalysed reactions and DFT calculations
that allow the mechanistic elucidation of the present trans-
formation are underway and will be reported elsewhere.
Experimental Section
Gold catalyst 4 (1.9 mg, 0.002 mmol, 4 mol%) was added to a solution of
diyne 1a (14.4 mg, 0.05 mmol) in CDCl3 (0.5 mL, c=0.1m) in an NMR
tube. The NMR tube was placed into an oil bath and the mixture was
heated to reflux. The reaction was periodically monitored by using
[7] Cycloalkyne 2a was isolated as a mixture with unreacted diyne 1a.
Chem. Eur. J. 2009, 15, 8966 – 8970
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8969

F. Gagosz et al.
[8] 1H and 13C NMR spectroscopy analysis attested to the presence of
an exo-methylene group (d(C)=119.6 ppm, d(H)=5.49, 5.45 ppm)
and an internal cyclic alkyne with signals shifted upfield (d(C)=
95.9, 94.3 ppm).
Abad, A. Corma, H. Garcia, M. Iglesias, F. Sµnchez Angew. Chem.
2007, 119, 1558–1560; Angew. Chem. Int. Ed. 2007, 46, 1536–1538;
f) P. H.-Y. Cheong, P. Morganelli, M. R. Luzung, K. N. Houk, F. D.
Toste, J. Am. Chem. Soc. 2008, 130, 4517–4526.
[9] For other Pd- or Y-catalysed formations of macrocyloalkynes from
diynes, see: a) B. M. Trost, S. Matsubara, J. J. Caringi, J. Am. Chem.
Soc. 1989, 111, 8745–8746; b) B. Bulic, U. Luecking, A. Pfaltz, Syn-
lett 2006, 1031–1034; c) K. Komeyama, T. Kawabata, K. Takehira,
K. Takaki, J. Org. Chem. 2005, 70, 7260–7266; d) V. Gevorgyan, N.
Tsuboya, Y. Yamamoto, J. Org. Chem. 2001, 66, 2743–2746.
[10] A single example of a gold(I)-catalysed cyclisation of a 1,9-enyne
into a ten-membered cycle by using 50 mol% of catalyst has been
reported, see: E. Comer, E. Rohan, L. Deng, J. A. Porco, Org. Lett.
2007, 9, 2123–2126.
[18] Compound 6 is presumably regenerated from 7 by demetalation
with HNTf2.
[19] The loss of the D label in [D]-1i was even more pronounced when
the reaction was performed in CDCl3 saturated with water (35% of
D label in unreacted [D]-1i after 6 h of reaction). See the Support-
ing Information for more details and other deuterium labelling ex-
periments.
[20] This demetalation should occur formally with the HNTf2 generated
during the formation of the gold acetylide species (6!7).
[21] See the Supporting Information for more details.
[11] The representation of 2b in the solid state (see Table 1) shows a dis-
torted structure with a strained alkyne functionality that deviates
from linearity by 118. Notably, the p system of the exo-methylene
moiety is not conjugated with that of the adjacent aromatic ring.
See the Supporting Information for more details. CCDC 726515
(2b) contain 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.
[12] Geometric constraints due to the presence of the cyclopentene units
could be hindering the desired cyclisation.
[13] The observed poor selectivity could be due to the small differentia-
tion between the two alkynes in the formation of the gold acetylide
(see Table 4).
[22] The extensive loss of deuteration at position b of the alkene in [D]-
2i might be the result of a mass and/or isotopic effect that favours
the exchange of the metal fragment with a hydrogen atom rather
than with a deuterium atom. This loss of deuteration could be mini-
mised by performing the reaction in CDCl3 saturated with D2O
(ꢁ80% of D label at position b). See the Supporting Information
for more details on deuterium labelling experiments and other ex-
periments supporting the mechanistic proposal.
[23] This mechanism also accounts for the failure of diynes 1u and 1v to
give cycloalkynes. This unreactivity could be explained by the prob-
able presence of an unfavoured steric interaction between the two
fragments at the alkyne termini in the transition state that leads to
9.
[24] This assumption is based on calculations reported by Toste, Houk
et al. (see ref. [17f]). Coordination of a phosphine–gold cation to a
phosphine–gold acetylide was calculated to be more favoured (by
ꢁ22 kcalmolꢀ1) over coordination to an allene, whereas calculations
showed only a small difference in the preference for a phosphine–
gold cation to coordinate to an alkyne or an allene.
[14] The observed selectivity might be the result of a better differentia-
tion of the two alkyne units with formation of the gold acetylide on
the more acidic alkyne chain and the gold(I) activation of the more
electron rich alkyne (see Table 4).
1
[15] Determined by using H NMR spectroscopy.
[16] This mechanism appears to be more in tune with the experimental
observations. See the Supporting Information for other mechanistic
proposals.
[17] For a summary of gold-catalysed transformations involving gold ace-
tylide species, see section 3.1 in ref. [1c]. Gold acetylides have been
proposed as intermediates in nucleophilic additions to C=N or C=O
[25] This mechanism, involving a stereoselective intramolecular protode-
metalation (13!10), is also in agreement with the stereoisotopic
distribution observed in the case of diyne [D]-1i. See the Supporting
Information for more details.
[26] A mechanism of nucleophilic addition of the pendant alkyne to a
gold(I)-activated alkyne, such as 6, could also be envisaged. Its oc-
currence remains relatively improbable given the stereoisotopic dis-
tribution observed in the case of diyne [D]-1i. See the Supporting
Information for more details.
ꢀ
bonds, Sonogashira-type coupling reactions and other C C bond for-
mations; for examples, see: a) C. Wei, C.-J. Li, J. Am. Chem. Soc.
2003, 125, 9584–9585; b) X. Yao, C.-J- Li, Org. Lett. 2006, 8, 1953–
1955; c) X.-Y. Liu, P. Ding, J.-S. Huang, C.-M. Che, Org. Lett. 2007,
9, 2645–2648; d) Y. Fuchita, Y. Utsunomiya, M. Yasutake, J. Chem.
Soc. Dalton Trans. 2001, 2330–2334; e) C. Gonzµlez-Arellano, A.
Received: May 16, 2009
Published online: July 31, 2009
8970
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 8966 – 8970