Gold(i)-catalyzed cycloisomerization of alkynyl hydroxyallyl tosylamides to 4-oxa-6-azatricyclo[3.3.0.02,8]octanes

Article abstract of DOI:10.1039/c1cc10907c

Reaction of alkyne allyl alcohols tethered with N-(p-tolylsulfonamide) in the presence of a cationic gold(i) catalyst gave new cycloisomerization products, 4-oxa-6-azatricyclo[3.3.0.0^2,8]octanes.

Full text of DOI:10.1039/c1cc10907c

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2011, 47, 5190–5192
www.rsc.org/chemcomm
COMMUNICATION
Gold(I)-catalyzed cycloisomerization of alkynyl hydroxyallyl tosylamides
to 4-oxa-6-azatricyclo[3.3.0.02,8]octanesw
Yujung Park,a Sun Young Kim,a Ji Hoon Park,a Jieun Cho,a Youn Kyung Kangb and
Young Keun Chung*a
Received 16th February 2011, Accepted 11th March 2011
DOI: 10.1039/c1cc10907c
Reaction of alkyne allyl alcohols tethered with N-(p-tolylsulfon-
amide) in the presence of a cationic gold(I) catalyst gave new
cycloisomerization products, 4-oxa-6-azatricyclo[3.3.0.02,8]octanes.
far as we are aware. We herein communicate our preliminary
results.
The transition metal-catalyzed cycloisomerization reactions
have attracted much attention because they can provide many
useful cyclized compounds from readily available starting
materials.1 The selective synthesis of different cyclic products
from the same starting materials by subtle modifications of the
catalytic conditions is an interesting but often troublesome
topic for chemists.2 In this regard, the gold-catalyst has
widened its application by its versatile catalytic activities.3,4
ð1Þ
When 1 was treated with 7 mol% of Au (7 mol% Au(PPh3)Cl
with 10 mol% of AgSbF6) in THF at 50 1C, 1a was obtained in
20% yield with a concomitant formation of 1b and 1c in 51%
and 3% yields, respectively.11 Compound 1b was the product
reported by Yeh et al. recently10 and 1c was a hydration
product.12 In contrast to Yeh’s result where only syn-1b was
formed, we obtained both syn and anti isomers of 1b. We
tentatively attribute the cause of this observed discrepancy to
the reaction temperature; the reaction at elevated temperature
might cause a loss of stereoselectivity. Hydration of alkynes to
form the respective ketones by Au(I)-phosphine catalyst has
been demonstrated.13 The structure of 1a was assigned based
on a series of 1D and 2D NMR studies and was further
confirmed by X-ray crystallography (Fig. 1).11 According to
the X-ray diffraction study, two isomeric (1 : 1) mixtures exist
in crystals.
Recently, Muller and coworkers have reported cyclo-
¨
isomerization reactions of alkyne allyl alcohols (AAAs) with
rhodium,5 iridium,6 and palladium7 catalysts. Nicolaou et al.
also reported8 a rhodium-catalyzed cycloisomerization using
AAAs with terminal alkynes. In all cases, Alder-ene-type
reaction products were isolated. The Montgomery group has
reported a base-promoted, nickel-catalyzed cycloisomerization
of AAAs and they also observed the formation of Alder-ene-
type products.9 We have studied the use of this substrate in an
Au(I)-catalyzed cycloisomerization. While we were engaged in
this study, Yeh et al. reported the result of such reactions.10
The structure of the product in their work differs slightly from
a typical Alder-ene-type product; 9-endo-dig cyclization was
suggested as a plausible reaction pathway by the authors.
Interestingly, however, we observed the formation of a new
cycloisomerized product 5-methyl-6-tosyl-4-oxa-6-azatricyclo-
[3.3.0.02.8]octane (eqn (1), 1a) from the Au(I)-catalyzed cyclo-
isomerization of a virtually same AAA starting material as
that of Yeh’s work yet under only subtly different reaction
conditions. The structure of the 4-oxa-6-azatricyclo[3.3.0.02.8]-
octane skeleton is exposed for the first time in this study as
Encouraged by this observation, we began the optimization
study to obtain a maximum yield of 1a. Several parameters
including counter-anion, solvent, temperature, and reaction
time were examined. The most relevant results are presented in
Table 1. It should be noted that the major difference in the
a Intelligent Textile System Research Center,
Department of Chemistry, College of Natural Sciences,
Seoul National University, Seoul 151-747, Korea.
E-mail: ykchung@snu.ac.kr; Fax: +82 2 889 0310;
Tel: +82 2 880 6662
b Department of Chemistry, Sangmyung University, Seoul 110-743,
Korea. E-mail: younkang@smu.ac.kr
w Electronic supplementary information (ESI) available: CCDC
804791 (1a). For ESI and crystallographic data in CIF or other
electronic format see DOI: 10.1039/c1cc10907c
Fig. 1 X-Ray structure of 1a. C, N, O, and S atoms are shown in
gray, violet, red, and yellow, respectively. Hydrogen atoms are omitted
for clarity.
c
5190 Chem. Commun., 2011, 47, 5190–5192
This journal is The Royal Society of Chemistry 2011

View Article Online
Table 1 Reaction of 1 under various reaction conditionsa
Yieldb (%)
Table 2 Gold-catalyzed cycloisomerizationsa
Entry
Reactant
Productb (%)
1a
1b
1c
Entry Au (mol%) X (mol%) T/1C t/h
1c
2d
3
4
5
7
5
5
5
5
5
5
5
10
5
SbF6 (10)
SbF6 (7)
SbF6 (7)
PF6 (7)
50
40
50
50
50
50
50
50
85
85
85
115
19
5
4
24
3
24
24
2
20
45
42
9
31
14
0
17
12
55
72
51
22
12
6
20
28
0
43
68
15
11
13
3
10
12
7
18
17
0
2
0
0
0
BF4 (7)
1
2
3
4
5
6
7
8
R = phenyl
4-C6H4OCH3
4-C6H4CH3
3,5-C6H3(CH3)2
4-C6H4Cl
Naphthyl
Methyl
Ethyl
n-Butyl
Cyclopropyl
Cyclohexenyl
Vinyl
72
55
65
65
62
65c
48
50
29
62
42
30
11
32
18
7
10
9
0
9
10
10
7
6
ClO4 (7)
NO3 (7)
OTf (7)
OTf (14)
SbF6 (7)
SbF6 (14)
SbF6 (14)
7e
8
9
1/6
0.5
1/6
10
11
12
10
10
1/12 60
0
a
9
0.15 mmol (or 0.3 mmol) of 1, gold catalyst, and silver salt were
b
reacted in 6.0 ml dichloroethane. Isolated yield. Solvent = THF.
d
c
10
11d
12e
e
Solvent = CH2Cl2. 83% of the reactant recovered.
Trace
reaction conditions between Yeh’s and ours was the reaction
temperature. Therefore, we kept the reaction temperature at
above the room temperature. We first tested the reaction in
dichloromethane, THF, and 1,2-dichloroethane (DCE) solvents
(entries 1–3). The yield was highly sensitive to the solvent
(THF, 20%; CH2Cl2, 45%; DCE, 42%). Although the yield of
1a was slightly higher in CH2Cl2 than in DCE, that of the
side product 1b was much lower in DCE. Thus DCE was
our choice. With this solvent in hand, we screened the silver
salts (entries 4–7). Reactions with AgSbF6, AgPF6, AgBF4,
and AgClO4 salts produced 1a in 55%, 9%, 31%, and 14%
yields, respectively, and AgSbF6 was thus the best. With
AgNO3, however, the reaction did not produce any products
and 83% of the reactant was recovered. An interesting result
was observed when AgOTf was used; the reaction produced 1b
as a major product (43%) with a minor formation of 1a in
17% yield (entry 8). This result is reminiscent of the observa-
tion of Yeh’s group.10 Their [(PPh3)AuCl]/AgOTf system
at room temperature solely gave 1b without producing 1a.
The observed difference strongly suggests that the reaction
pathway is highly dependent upon the reaction temperature.
Congruent with this, the yield of 1a increased to 55% when the
reaction temperature increased to 85 1C (entry 10) with a 30 min
reaction time. Additional increase of the yield was achieved
with the increase of the catalyst amount up to 10 mol% at
the same temperature (72%, entry 11). With these reaction
conditions, the reaction time was further shortened to 10 min.
However, a reaction temperature higher than 85 1C was not
helpful to enhance the yield of 1a; at 115 1C (entry 12), for
example, 1a was isolated in 60% within 5 min. Likewise, the
reaction with AgOTf at 85 1C gave rise to the decrease of the
yield of 1a with concomitant increase of 1b (12% and 68%
yields, respectively) relative to that at 50 1C. It is obvious that
both temperature and counter anion play important roles in
determining the reaction pathway. With the results described
above, our optimized reaction conditions were established as
follows: 10 mol% Au(PPh)3Cl, 14 mol% AgSbF6, DCE, 85 1C,
and 10 min.
13
14
15f
a
Reaction conditions: 0.15 mmol of substrate, 10 mol% AuPPh3Cl,
and 14 mol % AgSbF6 were reacted in 6 ml DCE at 85 1C for
10 min. Isolated yields. Two isomers are mixed. 51% reactant
b
c
d
e
recovered. 28% of the reactant recovered and a trace amount of b
f
was obtained. 20 mol% catalyst used.
All the reactions were completed within 10 min. A certain type
of substrates only gave the desired product in low yield with
considerable amounts of reactants recovered (entries 11 and
12, 51% and 28% recovered) even under the prolonged reac-
tion time. A variety of substrates with different substituents at
the terminal alkyne have been proved to be useful candidates
for the synthesis of 4-oxa-6-azatricyclo[3.3.0.02.8]octanes. In
particular, substrates with all kinds of aryl groups tested in this
study (entries 1–6) gave 4-oxa-6-azatricyclo[3.3.0.02.8]octanes in
good yields (55–72%). It seems that the electronic or steric
effect of the substituent(s) on the para-position of the arene
ring was rather insensitive to the yield of the reaction. Substrates
with alkyl substituents at the terminal alkyne afforded the
corresponding 4-oxa-6-azatricyclo[3.3.0.02.8]octanes in moderate
yields (42–62%, entries 7–10) except substrate 10 (29%). However,
substrates with alkenyl–alkynyl (entries 11 and 12) gave a
rather poor yield (30–42%). With a methyl substituent, in
addition to 9a (48%), a concomitant formation of 9c (18%)
was observed. When two methyl groups were introduced to an
With the optimal conditions in hand, the substrate scope of
the Au(I)-catalyzed cyclization reaction was investigated with
a series of AAAs, and the results are summarized in Table 2.11
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 5190–5192 5191

View Article Online
olefin group (entry 13), 13a and 13b were isolated in 21% and
49% yields, respectively. When an alkynyl tosylamide bearing
a secondary alcohol was used as a substrate (entry 14), a
mixture of 14a and 14b was isolated in 19% and 48% yields,
respectively, and further purification of 14b leads to separation
of syn and anti isomers in a ratio of 1 : 2. When a methyl
group was introduced to the 3-position of 1,6-enyne (15), a
cyclization reaction was found not to take place in the
presence of 10 mol% Au catalyst and the reactant was
recovered; 15a was isolated in 35% yield with a trace amount
of 15b only in the presence of 20 mol% of Au catalyst.
Interestingly, when the protecting group on the N tether was
changed from a tosyl to Ph group, the catalytic system was
ineffective and the reactant was recovered. It seems that the
electron-donating substituent further increased the coordinating
ability of the amino group, which might block the alkyne from
coordinating with the gold species. Therefore, no cycloisomeri-
zation product was observed with this substrate.
equilibrate with an intermediate IV. Because of the significant
carbocationic character at the alkynyl carbon adjacent to the
N-Ts group in IV, a nucleophilic addition of the hydroxy
oxygen occurs to give an intermediate V.15 Proton transfer
followed by protodemetalation in V affords the product and
regenerates the gold(I) catalyst in the catalytic cycle.
In conclusion, we have reported a novel gold-catalyzed
cycloisomerization of alkynyl hydroxyallyl tosylamides,
leading to the synthesis of oxaazatricyclic compounds in a
concise manner. Change in the reaction conditions completely
alters the reaction pathway. On the basis of the experimental
results, a possible reaction mechanism has been proposed.
Further studies are needed to delineate the intimate mecha-
nistic steps and expand the scope of this cyclization process.
Notes and references
1 M. Chen, Y. Weng and A. W. Lei, Prog. Chem., 2010, 22, 1341;
A. Furstner, Chem. Soc. Rev., 2009, 38, 3208; V. Michelet,
¨
The formation of a new 4-oxa-6-azatricyclo[3.3.0.02.8]octane
skeleton appeared to be quite interesting in terms of its
reaction mechanism. In order to determine the position where
the 13C atom is located in the cycloisomerized product we
performed a 13C-labeling experiment using the simple enyne
1-13C. The reaction of 1-13C in the presence of Au(PPh3)Cl/
AgSbF6 in DCE at 85 1C for 10 min gave 1a-13C in 55% yield,
showing that the C3–N bond cleavage occurred during the
reaction (Scheme 1). It is important to note that the C3–N
bond remained intact in 1b-13C as the 13C atom was found to
be in the aza-5-membered ring.
P. Y. Tollec and J. P. Genet, Angew. Chem., Int. Ed., 2008, 47,
4268; C. Bruneau, Angew. Chem., Int. Ed., 2005, 44, 2328;
A. M. Echavarren and C. Nevado, Chem. Soc. Rev., 2004, 33,
431; C. Aubert, O. Buisine and M. Malacria, Chem. Rev., 2002,
102, 813.
2 H. Gao and J. Zhang, Adv. Synth. Catal., 2009, 351, 85.
3 E. Soriano and J. Macro-Contelles, Acc. Chem. Res., 2009, 42,
1026; A. S. K. Hashmi and M. Rudolph, Chem. Soc. Rev., 2008,
37, 1766; N. Marion and S. P. Nolan, Chem. Soc. Rev., 2008, 37,
1776; A. S. K. Hashmi, Angew. Chem., Int. Ed., 2008, 47, 6754.
4 Y. T. Lee, Y. K. Kang and Y. K. Chung, J. Org. Chem., 2009, 74,
7922; S. M. Kim, J. H. Park, Y. K. Kang and Y. K. Chung, Angew.
Chem., Int. Ed., 2009, 48, 4532; S. M. Kim, J. H. Park, S. Y. Choi
and Y. K. Chung, Angew. Chem., Int. Ed., 2007, 46, 6172; S. I. Lee,
J. Y. Baek, S. H. Sim and Y. K. Chung, Synthesis, 2007, 2107;
S. I. Lee, S. M. Kim, S. Y. Kim, M. R. Choi and Y. K. Chung,
J. Org. Chem., 2006, 71, 9366; S. I. Lee, S. M. Kim, S. Y. Kim and
Y. K. Chung, Synlett, 2006, 2256.
On the basis of the above results and previous study,14
a
plausible reaction mechanism was proposed (Scheme 2). The
first step is a gold(I)-catalyzed cycloisomerization of an enyne
to a cyclopropanated compound (I), followed by a transannula-
tion between the nitrogen atom and the carbon atom attached
to the gold cation to give an intermediate II. The ring
contraction through intermediate II is intriguing and might
explain the different reactivity observed in this system. Shift of
the gold leads to formation of an intermediate III which may
5 N. Korber, F. Rominger and T. J. J. Muller, Synlett, 2010, 782.
¨
6 M. Kummeter, C. M. Ruff and T. J. J. Muller, Synlett, 2007, 717.
¨
¨
7 C. J. Kressierer and T. J. J. Muller, Synlett, 2005, 1721;
¨
C. J. Kressierer and T. J. J. Muller, Org. Lett., 2005, 7,
¨
2237; C. J. Kressierer and T. J. J. Muller, Synlett, 2004, 655;
¨
C. J. Kressierer and T. J. J. Muller, Tetrahedron Lett., 2004, 45,
2155.
¨
8 K. C. Nicolaou, A. Li, D. J. Edmond, G. S. Tria and S. P. Ellery,
J. Am. Chem. Soc., 2009, 131, 16905; K. C. Nicolaou, A. Li and
S. P. Ellery, Angew. Chem., Int. Ed., 2009, 48, 6293.
9 J. H. Phillips and J. Montgomery, Org. Lett., 2010, 12, 4556.
10 M.-C. P. Yeh, M.-N. Lin, W.-J. Chang, J.-L. Liou and Y.-F. Shih,
J. Org. Chem., 2010, 75, 6031.
11 The synthetic procedures and spectroscopic data of the new
compounds are summarized in ESI.w CCDC 804791 (1a) contains
the supplementary crystallographic data for this paper.
12 C. Sperger and A. Fiksdahl, Org. Lett., 2009, 11, 2449; H. Teles,
S. Brode and M. Chabanas, Angew. Chem., Int. Ed., 1998, 37, 1415.
13 For reviews, see: F. Alonso, I. P. Beletskaya and M. Yus, Chem.
Rev., 2004, 104, 3079; I. V. Kozhernikov, Chem. Rev., 1998, 98,
171. For gold-catalyzedhydration of alkynes, see: Y. Fukuda and
Scheme 1 Gold-catalyzed cycloisomerization of 13C labelled enyne 1.
´
K. Utimoto, J. Org. Chem., 1991, 56, 3729; A. Almassy,
C. E. Nagy, A. C. Benyei and F. Joo, Organometallics, 2010, 29,
´
2484; S. K. Schneider, W. A. Herrmann and E. Herdtweck,
´
Z. Anorg. Allg. Chem., 2003, 629, 2363; A. Leyva and A. Corma,
J. Org. Chem., 2009, 74, 2067; N. Marion, R. S. Ramo
S. P. Nolan, J. Am. Chem. Soc., 2009, 131, 448.
14 C. H. M. Amijs, V. Lopez-Carrillo, M. Raducan, P. Pe
´
n and
´
´
rez-Gala
´
n,
C. Ferrer and A. M. Echavarren, J. Org. Chem., 2008, 73, 7721.
15 Instead of the formation of the intermediate V, one of the referees
suggested that the enamine tautomerizes to the iminium and gets
trapped by the hydroxyl group. R. L. Lalonde, W. E. Brenvovich,
Jr., D. Benitez, E. Tkatchouk, K. Kelley, W. A. Goddard, III and
F. D. Toste, Chem. Sci., 2010, 1, 226.
Scheme 2 A plausible reaction mechanism.
5192 Chem. Commun., 2011, 47, 5190–5192
c
This journal is The Royal Society of Chemistry 2011