Palladium-Catalyzed Cyclization of o-Alkynyltrifluoroacetanilides with Allyl Esters. A Regioselective Synthesis of 3-Allylindoles

Article abstract of DOI:10.1021/jo971237p

The reaction of readily available o-alkynyltrifluoroacetanilides 1 with allyl esters provides a valuable new route to 3-allylindoles 3. Three basic procedures have been developed: a stepwise method based on the isolation of the N-allyl derivative 4 and its subsequent cyclization to 3 (procedure a), a one-pot reaction omitting the isolation of 4 (procedure b), and a procedure which most probably leads to the formation of 3 through a mechanism not involving the intermediacy of 4 (procedure c). In the presence of the electron-rich sterically encumbered ligand tris(2,4,6-trimethoxyphenyl)-phosphine (ttmpp) the reaction exhibits remarkable regioselectivity and almost exclusive formation of 3-allylindoles with the indolyl moiety bound to the less substituted allyl terminus is usually observed. However, some loss of olefin geometry is observed.

Full text of DOI:10.1021/jo971237p

J . Org. Chem. 1998, 63, 1001-1011
1001
P a lla d iu m -Ca ta lyzed Cycliza tion of o-Alk yn yltr iflu or oa ceta n ilid es
w ith Allyl Ester s. A Regioselective Syn th esis of 3-Allylin d oles
Sandro Cacchi,* Giancarlo Fabrizi, and Paola Pace
Dipartimento di Studi di Chimica e Tecnologia delle Sostanze Biologicamente Attive, Universita` La
Sapienza, P.le A. Moro 5, I-00185 Roma, Italy
Received J uly 8, 1997
The reaction of readily available o-alkynyltrifluoroacetanilides 1 with allyl esters provides a valuable
new route to 3-allylindoles 3. Three basic procedures have been developed: a stepwise method
based on the isolation of the N-allyl derivative 4 and its subsequent cyclization to 3 (procedure a),
a one-pot reaction omitting the isolation of 4 (procedure b), and a procedure which most probably
leads to the formation of 3 through a mechanism not involving the intermediacy of 4 (procedure c).
In the presence of the electron-rich sterically encumbered ligand tris(2,4,6-trimethoxyphenyl)-
phosphine (ttmpp) the reaction exhibits remarkable regioselectivity and almost exclusive formation
of 3-allylindoles with the indolyl moiety bound to the less substituted allyl terminus is usually
observed. However, some loss of olefin geometry is observed.
In tr od u ction
clization reactions promoted by η3-allylpalladium com-
plexes are limited to the pioneering synthesis of γ-(E)-
alkylidene-γ-butyrolactones from pentynoic acid and allyl
acetates (or from pentynoic acid allyl esters) described
by Tsuda, Saegusa et al.5 and to the synthesis of
R-substituted-γ-methylene-γ-butyrolactones from pen-
tynoic acid derivatives described by Mandai, Tsuji et al.6
Our continuing interest in the application of this chem-
istry to the synthesis of heterocycles1a-c,g-i,2b and the
attractiveness of widening the scope of the methodology
to include the vastly important class of nitrogen hetero-
cycles led us to investigate the utilization of η3-allylpal-
ladium complexes as promoters of cyclization reactions
of alkynes containing nitrogen nucleophiles. In particu-
lar, because of the intense interest in indole-based
biologically active molecules1e,7 and, consequently, in the
selective syntheses of the indole nucleus,1a,i,2b,8 our efforts
were devoted to the development of a new approach to
the preparation of indole derivatives. o-Alkynyltrifluo-
roacetanilides 1, previously employed by us in new
syntheses of variously substituted indoles,1i,2b were se-
lected as the acetylenic building blocks that might
participate in a new cyclization process leading to 3-al-
lylindoles 3 (eq 2).
The cyclization of terminal and internal alkynes con-
taining proximate nucleophiles, promoted by organopal-
ladium complexes, is expanding into one of the most
effective strategies for ring construction. It provides in
fact a straightforward approach to the synthesis of
functionalized carbo- and heterocycles through the regio-
and stereoselective addition of a nucleophile and of an
unsaturated carbon unit across the carbon-carbon triple
bond (eq 1).
Most of this chemistry, however, deals with the utiliza-
tion of σ-organopalladium complexes.1-4 Reports of cy-
(1) For cyclizations promoted by σ-vinyl- and σ-arylpalladium
complexes, see: (a) Cacchi, S.; Fabrizi, G.; Marinelli, F, Moro, L.; Pace,
P. Synlett 1997, 1363. (b) Cacchi, S.; Fabrizi, G.; Moro, L. J . Org. Chem.
1997, 62, 5327. (c) Arcadi, A.; Cacchi, S.; Del Rosario, M.; Fabrizi, G.;
Marinelli, F. J . Org. Chem. 1996, 61, 9280. (d) Cavicchioli, M.;
Decortiat, S.; Bouyssi, D.; Gore, J .; Balme, G. Tetrahedron 1996, 52,
11463. (e) Saulnier, M. G.; Frennesson, D. B.; Deshpande, M. S.; Vyas,
D. M. Tetrahedron Lett. 1995, 36, 7841. (f) Balme, G.; Bouyssi, D.
Tetrahedron 1994, 50, 403. (g) Arcadi, A.; Cacchi, S.; Larock, R. C.;
Marinelli, F. Tetraherdon Lett. 1993, 34, 2813. (h) Arcadi. A.; Burini,
A.; Cacchi, S.; Delmastro, M.; Marinelli, F.; Pietroni, B. R. J . Org.
Chem. 1992, 57, 976. (i) Arcadi, A.; Cacchi, S.; Marinelli, F. Tetrahedron
Lett. 1992, 33, 3915. (j) Luo, F.-T.; Wang, R.-T. Tetrahedron Lett. 1992,
33, 6835. (k) Luo, F.-T.; Schreuder, I.; Wang, R.-T. J . Org. Chem. 1992,
57, 2213. (l) Fournet, G.; Balme, G.; Gore, J . Tetrahedron 1991, 47,
6293. (m) Bouyssi, D.; Balme, G.; Gore, J . Tetrahedron Lett. 1991, 32,
6541. (n) Fournet, G.; Balme, G.; Van Hemelryck, B.; Gore, J .
Tetrahedron Lett. 1990, 31, 5147. (o) Inoue, Y.; Itoh, Y.; Imaizumi, Y.;
Imaizumi, S. J . Mol. Catal. 1990, 60, L1-L3.
(3) For cyclizations promoted by σ-alkynylpalladium complexes,
see: Bouyssi, D.; Gore, J .; Balme, G. Tetrahedron Lett. 1992, 33, 2811.
(4) For cyclizations promoted by σ-allenylpalladium complexes,
see: Bouyssi, D.; Gore, J .; Balme, G.; Louis, D.; Wallach, J . Tetrahedron
Lett. 1993, 34, 3129.
(5) Tsuda, T.; Ohashi, Y.; Nagahama, N.; Sumiya, R.; Saegusa, T.
J . Org. Chem. 1988, 53, 2650.
(2) For cyclizations promoted by σ-acylpalladium complexes, see: (a)
Arcadi, A.; Rossi, E. Tetrahedron Lett. 1996, 37, 6811. (b) Arcadi, A.;
Cacchi, S.; Carnicelli, V.; Marinelli, F. Tetrahedron 1994, 50, 437. (c)
Okuro, K.; Furuune, M.; Miura, M.; Nomura, M. J . Org. Chem. 1992,
57, 4754.
(6) Mandai, T.; Ohta, K.; Baba, N.; Kawada, M.; Tsuji, J . Synlett
1992, 671.
S0022-3263(97)01237-1 CCC: $15.00 © 1998 American Chemical Society
Published on Web 02/05/1998

1002 J . Org. Chem., Vol. 63, No. 4, 1998
Cacchi et al.
The preparation of 3-allylindoles has previously been
described by Utimoto.9 His approach is based on the
cyclization of N-(methoxycarbonyl)-o-alkynylanilines and
allyl chlorides in the presence of a palladium(II) catalyst
not containing σ-carbon-palladium bonds, PdCl2(MeCN)2.
The reaction was presumed to proceed via trapping of
the σ-indolylpalladium intermediate, generated in situ,
with the allyl chloride, which has to be used in large
excess (10 equiv). The new carbon-carbon bond is
generated regioselectively at the γ-position in an SN2′
fashion. We surmised that our procedure might employ
a limited excess of the allylating agent and, more
significantly, might provide a chance to synthesize 3-al-
lylindoles not available by means of the PdCl2-catalyzed
cyclization/allylation sequence.
Resu lts
Th e Rea ction of o-(P h en yleth yn yl)tr iflu or oa ceta -
n ilid e 1a w ith Allyl Aceta te. Our initial attempt
explored the reaction of o-(phenylethynyl)trifluoroaceta-
nilide 1a with allyl acetate (1.2 equiv) under conditions
similar to those employed by us for the cyclization of 1
promoted by σ-vinyl- and σ-arylpalladium species [Pd-
(PPh3)4 (0.05 equiv), K2CO3 (5.0 equiv), MeCN, 60 °C].1i
However, the nucleophilic attack of the nitrogen on the
η3-allylpalladium intermediate derived from allyl acetate
and palladium proved to be faster than the desired
organopalladium-promoted cyclization. Only the corre-
sponding N-allyl derivative 4a was isolated in 96% yield
after 1.5 h (eq 3). Nonetheless, because of the contem-
porary presence of the aryl and trifluoroacetyl groups on
the nitrogen atom of 4a , we thought that under suitable
reaction conditions it might undergo a palladium-
promoted ionization of the N-Callyl bond and generate
an η3-allylpalladium complex which, by taking advantage
of the proximity of the π-electrons of the carbon-carbon
triple bond, could in turn promote the desired heterocy-
clization to the indole nucleus. We were pleased to find
that heating 4a in MeCN at 90 °C for 1 h in the presence
of Pd(PPh3)4 (0.05 equiv) afforded 2-phenyl-3-allylindole
3a and its N-trifluoroacetyl derivative 5a in 47 and 27%
yield, respectively (eq 3). Complete conversion into 3a
(91% isolated yield) was achieved by allowing 4a to react
under the same conditions in the presence of K2CO3 (5.0
equiv).
Hereafter we report our observations on the scope of
the reaction, the influence of the ligands on the regio-
chemical outcome, and a few mechanistic insights.
(7) For recent leading works in the area of biologically active indole-
based molecules, see: Ohkubo, M.; Kawamoto, H.; Ohno, T.; Nakano,
M.; Morishima, H. Tetrahedron 1997, 53, 585. Ishikura, M.; Yaginuma,
T.; Agata, I.; Miwa, Y.; Yanada, R.; Taga, T. Synlett 1997, 214. Mayer,
G.; Wille, G.; Steglich, W. Tetrahedron Lett. 1996, 37, 4483. Kuehne,
M. E.; Wang, T.; Seaton, P. J . J . Org. Chem. 1996, 61, 6001. Mayer,
G.; Wille, G.; Steglich, W. Tetrahedron Lett. 1996, 37, 4483. Pawlak,
J . M.; Khau, V. V.; Hutchison, D. R.; Martinelli, M. J . J . Org. Chem.
1996, 61, 9055. Kawasaki, I.; Yamashita, M, Ohta, S. Chem. Pharm.
Bull. 1996, 44, 1831. Choshi, T.; Yamada, S.; Sugino, E.; Kuwada, T.;
Hibino, S. J . Org. Chem. 1995, 60, 5899. Choshi, T.; Yamada, S.;
Sugino, E.; Kuwada, T.; Hibino, S. Synlett 1995, 147. Fisher, L. E.;
Labadie, S. S.; Reuter, D. C.; Clark, R. D. J . Org. Chem. 1995, 60,
6224. Cornec, O.; J oseph, B.; Me´rour, J .-Y. Tetrahedron Lett. 1995,
38, 8587. Shin, K.; Ogasawara, K. Synlett 1995, 859; Achab, S.; Guyot,
M.; Potier, P. Tetrahedron Lett. 1995, 36, 2615. Knight, S. D.; Overman,
L. E.; Pairaudeau, G. 1995, 117, 5776. Xie, G.; Lown, J . W. Tetrahedron
Lett.. 1994, 35, 5555. Tietze, L. F.; Grote, T. J . Org. Chem. 1994, 59,
192. Xie, G.; Lown, J . W. Tetrahedron Lett. 1994, 35, 5555. Davidsen,
S. K.; Summers, J . B.; Albert, D. H.; Holms, J . H.; Heyman, H. R.;
Magoc, T. J .; Conway, R. G.; Rhein, D. A.; Carter, G. W. J . Med. Chem.
1994, 37, 4423. Koerber-Ple´, K.; Massiot, G. Synlett 1994, 759. Horwell,
D. C.; Nichols, P. D.; Ratcliffe, G. S.; Roberts, E. J . Org. Chem. 1994,
59, 4418. Harris, W.; Hill, C. H.; Keech, E.; Malsher, P. Tetrahedron
Lett. 1993, 34, 8361. Sakamoto, T.; Kondo, Y.; Uchiyama, M.; Ya-
manaka, H. J . Chem. Soc., Perkin Trans. 1 1993, 1941. Black, D. St.
C.; Keller, P. A.; Kumar, N. Tetrahedron 1993, 49, 151. Agarwal, A.;
Pearson, P. P.; Taylor, E. W.; Li, H. B.; Dahlgren, T.; Hersloˆf, M.; Yang,
Y.; Lambertr, G.; Nelson, D. L.; Regan, J . W.; Martin, A. R. J . Med.
Chem. 1993, 36, 4006. Semmelhack, M. F.; Rhee, H. Tetrahedron Lett.
1993, 34, 1395. Kozikowski, A. P.; Ma, D.; Brewer, J .; Sun, S.; Costa,
E.; Romeo, E.; Guidotti, A. J . Med. Chem. 1993,36, 2908. Malleron, J .
L.; Gue´re´my, C.; Mignami, S.; Peyronel, J .-F.; Truchon, A.; Blanchard,
J .-C.; Doble, A.; Laduron, P.; Piot, O.; Zundel, J .-L.; Betschart, J .;
Canard, H.; Chaillou, P.; Ferris, O.; Huon, C.; J ust, B.; Kerphirique,
R.; Martin, B.; Mouton, P.; Renaudon, A. J . Med. Chem. 1993, 36, 1194.
Williams, T. M.; Ciccarone, T. M.; MacTough, S. C.; Rooney, C. S.;
Balani, S. K.; Conra, J . H.; Emini, E. A.; Goldman, M. E.; Greenlee,
W. J .; Kauffman, L. R.; O’Brien, J . A.; Sardana, V. V.; Schleif, W. A.;
Theoarides, A. D.; Anderson, P. S. J . Med. Chem. 1993, 36, 1291.
(8) For recent leading references on the construction of the indole
nucleus, see: Chen, C.-y.; Lieberman, D. R.; Larsen, R. D.; Veroeven,
T. R.; Reider, P. J . J . Org. Chem. 1997, 62, 2676. Ezquerra, J .; Pedregal,
C.; Lamas, C.; Barluenga, J .; Pe´rez, M.; Garcia-Martin, M. A.;
Gonzalez, J . M. J . Org. Chem. 1996, 61, 5804. Desarbre, E.; Me´rour,
J .-Y. Tetrahedron Lett. 1996, 37, 43. Larock, R. C.; Zenner, J . M. J .
Org. Chem. 1995, 60, 482. Edstrom, E. D. Synlett 1995, 49. Samizu,
K.; Ogasawara, K. Heterocycles 1995, 41, 1627. Aulwurm, U. R.;
Melchinger, J . U.; Kisch, H. Organometallics 1995, 14, 3385. Iwao, M.
Heterocycles 1994, 38, 45. Akazome, M.; Kondo, T, Watanabe, Y. J .
Org. Chem. 1994, 59, 3375. Fukuyama, T.; Chen, X.; Peng, G. J . Am.
Chem. Soc. 1994, 116, 3127. Samizu, K.; Ogasawara, K. Synlett 1994,
499. Tidwell, J . H.; Peat, A. J .; Buchwald, S. L. J . Org. Chem. 1994,
59, 7164. Yang, C.-C.; Maruoka, K.; Oishi, M. Yamamoto, H. 1993,
58, 7638. Chang, H.-T.; Fang, J .-M. 1993, 58, 3100. Satomura, M. J .
Org. Chem. 1993, 58, 3757. Caube`re, C.; Caube`re, P.; Renard, P.; Bizot-
Espiart, J .-G.; J amart-Gre´goire, B. Tetrahedron Lett. 1993, 34, 6889.
J eshke, T.; Wensbo, D.; Annby, U.; Gronowitz, S.; Cohen, L. A.
Tetrahedron Lett. 1993, 34, 6471. Gray, M.; Parsons, P. J .; Neary, A.
P. Synlett, 1993, 281. Ishibashi, H.; Tabata, T.; Hanaoka, K.; Iriyama,
H.; Akamatsu, S.; Ikeda, M. Tetrahedron Lett. 1993, 34, 489.
The reaction of 1a with allyl acetate at higher tem-
perature (90 °C; 3.5 h), but otherwise keeping all other
parameters the same, also produced 3a in good yield,
although lower than the overall yield of the two-step
protocol (75 vs 87%) (eq 3). The lower yield appears to
be mainly associated with the formation of 2-phenylin-
dole, 6a , very likely arising from a palladium-catalyzed
reaction. It was in fact generated only in trace amounts
when the reaction was carried out with the omission of
the Pd(PPh3)4.
The best result was obtained by treating 1a with allyl
acetate (1.2 equiv) at 60 °C in MeCN in the presence of
Pd(PPh3)4 (0.05 equiv) and K2CO3 (5 equiv) for 1.5 h (after
which time it is completely converted into 4a ) and then
raising the reaction temperature to 90 °C (2 h): 3a was
isolated in 81% yield. However, the application of these
conditions to substituted allyl acetates such as 2-hexen-
(9) Iritani, K.; Matsubara, S.; Utimoto, K. Tetrahedron Lett. 1988,
29, 1799.

A Regioselective Synthesis of 3-Allylindoles
J . Org. Chem., Vol. 63, No. 4, 1998 1003
1-yl and 1-octen-1-yl acetates met with failure. For
example, neither 3-(2-hexen-1-yl)indole nor the corre-
sponding N-(2-hexen-1-yl) derivative was discernible
when 1a was treated with 2-hexen-1-yl acetate (1.2 equiv)
for 24 h [MeCN, 60 °C, Pd(PPh3)4 (0.05 equiv), K2CO3 (5
equiv); 1a was recovered in 76% yield].
(2,6-dimethoxyphenyl)phosphine (tdmpp),11 showed the
same regioselectivity,12 but it was less rewarding in terms
of yield (Table 1, entries 12 and 13). The ligand effect
on the stereoselectivity appears to be minor because, with
the exception of a slightly greater loss of olefin geometry
with dppe (Table 1, entry 6), E:Z ratios change little upon
switching from electron-withdrawing to electron-donating
or from monodentate to bidentate ligands. Solvent effects
(as well as temperature effects), while affecting the
reaction rate and yields, do not seem to play a significant
role in the regio- and stereoselectivity.
Step w ise P r ep a r a tion of 3-Allylin d oles 3 (P r oce-
d u r e a ). Having established satisfactory conditions for
the N-allylation of 1 [Pd2(dba)3, dppb, THF, 60 °C] and
the subsequent cyclization of the corresponding o-alkynyl-
N-allyltrifluoroacetanilide 4 [Pd2(dba)3, ttmpp, DME, 100
°C or Pd(PPh3)4, K2CO3, MeCN, 90 °C], the stepwise
protocol was applied to the synthesis of some 3-allylin-
doles. Our preparative results are summarized in Table
2. No attempts have been made to optimize the N-
allylation and cyclization conditions for any example.
Thus, further improvement for a particular case of
importance appears likely. Nevertheless, Pd(PPh3)4 has
been successfully employed in the cyclization of allylic
derivatives generating symmetric η3-allylpalladium com-
plexes and even when the two allylic termini are mark-
edly different from a steric point of view (Table 2, entries
1-7, 10-12, 15, 17-21). The Pd2(dba)3/ttmpp combina-
tion has been found to be the catalyst of choice when
steric differences between the two allylic termini are
small, as it is the case of 2-hexen-1-yl and 1-octen-1-yl
carbonates (Table 2, entries 8, 9, 13).
N-Allyla tion of o-Alk yn yltr iflu or oa ceta n ilid es 1
w ith Allyl Ca r bon a tes 2. Because allyl carbonates
have been reported to be very reactive substrates, more
prone to undergo a number of reactions difficult to
achieve with allyl acetates under mild conditions,10 we
decided to investigate their utilization in the present
reaction. Taking into account our results with allyl
acetates, we started this part of the study by exploring
each of the two steps of the process separately. We
initially tried the N-allylation of 1a with ethyl 2-hexen-
1-yl carbonate, 2b, and Pd(PPh3)4 (0.05 equiv) in MeCN
(60 °C, 24 h) but omitting K2CO3.10 Under these condi-
tions, 1a was allylated to afford 4b as one regioisomer
in 33% yield. A remarkable increase in yield to 80% (of
approximately an 86:14 E:Z mixture) was achieved in
THF (60 °C, 5 h). Even faster reaction and higher yield
were observed in the presence of Pd2(dba)3 and dppb [1,4-
bis(diphenylphosphino)butane] in THF (60 °C, 1.5 h): 4b
was isolated in 87% yield (∼88:12::E:Z ). The use of dppp
[1,3-bis(diphenylphosphino)propane] was found equally
effective (87% yield, ∼ 86:14::E:Z, 1.5 h) whereas adding
dppe [1,2-bis(diphenylphosphino)ethane] to the reaction
mixture resulted in the isolation of 4b in a sparing 27%
yield (48 h) along with 2-phenylindole (45% yield).
Cycliza tion of o-Alk yn yl-N-a llyltr iflu or oa ceta n il-
id es 4. We next investigated the cyclization step (eq 4).
o-(Phenylethynyl)-N-(2-hexen-1-yl)trifluoroacetanilide 4b
(R ) Ph, R2 ) R3 ) H, R4 ) Prn) was selected as the model
system, and the effect of temperature and a variety of
ligands and solvents on the product distribution was
examined.
On e-P ot P r ep a r a tion of 3-Allylin d oles 3 (P r oce-
d u r e b). To keep the procedure as simple as possible,
we turned our attention to the development of a protocol
not involving the isolation of 4. We explored the effect
on the reaction outcome of temperature and of a variety
of ligands and solvents and arrived at the following one-
pot protocol: Pd(PPh3)4 in THF at 60 °C till the disap-
pearance of 1, addition of K2CO3, and then raising the
reaction temperature to 80 °C. The results obtained are
summarized in Table 3.
Th e Rea ction of o-Alk yn yltr iflu or oa ceta n ilid es 1
w ith Allyl Ca r bon a tes 2 in th e P r esen ce of P d 2
(d ba )3 a n d Tr is(2,4,6-tr im eth oxyp h en yl)p h osp h in e
(ttm p p ) (P r oced u r e c). When the development of a
one-pot protocol to 3 was attempted, the utilization of
ttmpp in the reaction of o-alkynyltrifluroacetanilides with
allyl carbonates was evaluated first. The aim was to find
conditions allowing the employment of this ligand and
Pd2(dba)3 in both steps, so as to evolve a simplified
protocol with the beneficial effect of ttmpp on the regio-
chemistry of the new carbon-carbon bond. To this end,
the reaction of 1a with ethyl 2-hexen-1-yl carbonate 2b,
our model system, was reexamined in the presence of Pd2-
(dba)3 and ttmpp in THF at 60 °C (procedure c). Surpris-
ingly, under these conditions 1a was found to undergo a
rapid (3 h) conversion into the corresponding indole
derivative, isolated in 76% yield (3′b:3′′b ) 97:3). The
use of a 1:6 Pd:ligand ratio proved to be crucial for the
The use of the same ligand that gives the highest yields
in the N-allylation, dppb, led to acceptable results only
in DME or DMSO at 100 °C (Table 1, entries 1-4).
Similar overall yields and higher 3′b :3′′b ratios were
obtained with dppp and dppe (Table 1, entries 5 and 6).
The popular Pd(PPh3)4 led to the cyclization derivative
in good yield (Table 1, entries 21 and 22), but the best
result with regard to both yield and regioselectivity was
observed upon going to Pd2(dba)3 and an electron-rich
sterically encumbered ligand, tris(2,4,6-trimethoxyphe-
nyl)phosphine (ttmpp),11 in a 1:6 Pd:ligand ratio (Table
1, entry 11). Decreasing the amount of ligand relative
to palladium resulted in incomplete conversion (Table 1,
entries 9 and 10). Notably, regioisomeric distribution
remains unchanged moving from a 1:2 to a 1:6 Pd:ligand
ratio. The less electron-rich, but equally crowded tris-
(12) Tris(2,6-dimethoxyphenyl)phosphine has been described to play
an important role in the regioselective addition of terminal alkynes to
either terminal or activated internal alkynes: Trost, B. M.; Sorum,
M. T.; Chan, C.; Harms, A. E.; Ruˆhter, G. J . Am. Chem. Soc. 1997,
119, 698. Trost, B. M.; McIntosh, M. C. J . Am. Chem. Soc. 1995, 117,
7255.
(10) Tsuji, J .; Minami, I. Acc. Chem. Res 1987, 20, 140.
(11) Wada, M.; Higashizaki, S. J . Chem. Soc., Chem. Commun. 1984,
482.

1004 J . Org. Chem., Vol. 63, No. 4, 1998
Cacchi et al.
success of the reaction (decreasing the Pd:ligand ratio to
1:2 resulted in the isolation of the indole derivative in
15% yield; 1a was recovered in 73% yield and 2-phe-
nylindole was isolated in 11% yield). No N-allyl inter-
mediate, 4b, however, was discernible in the reaction
mixture. Furthermore, 4b was recovered in 78% yield
when allowed to react under the same conditions [Pd2-
(dba)3, ttmpp, THF, 60 °C] for 10 h, and the indole
derivative was isolated in 16% yield only. On the other
hand, the cyclization of 4b in the presence of Pd2(dba)3
and ttmpp took about 15 h in DME at 100 °C to reach
completion (Table 1, entry 11). The results obtained by
using procedure c are summarized in Table 4.
Discu ssion
Mechanistically, the present synthesis of 3-allylindoles
is likely to proceed as outlined in eq 5. Path a, typical of
stepwise and one-pot protocols, consists of the following
four basic phases: (1) the nucleophilic attack of the
nitrogen atom on the allylic portion of the η3-allylpalla-
dium complex 7 to give 4 (only N-allyl derivatives bearing
the nitrogen fragment on the less substituted allyl
terminus have been isolated); (2) the formation of 10,
resulting from the ionization of the η2-olefinpalladium
complex 9 (η2-olefinpalladium complexes are usually
believed to be the first intermediates in the palladium-
catalyzed allylations)13 and the displacement of one
ligand to the palladium by the carbon-carbon triple bond
(that may be favored by the proximity of the acetylenic
moiety to the metal and that may also take place before
ionization); (3) the intramolecular nucleophilic attack of
the nitrogen atom across the activated carbon-carbon
triple bond to afford 11; (4) the reductive excision of Pd-
(0) through the transfer of the indolyl fragment to the
allyl group in a cis fashion,14 which produces the indole
derivative and regenerates the active catalytic species.
According to this scheme, the nitrogen atom intervenes
in the process as nucleophile in the N-allylation step and
as leaving group15,16 in the cyclization step. Such an
ambivalent behavior of amino groups has previously been
described.17 However, while in the present reaction the
reversible formation of the nitrogen-carbon bond does
not involve any change in the nature of the amino group,
known applications of this chemistry use free amino
groups in the preparation of N-allylamines, and allylam-
monium salts, generated in situ through protonation of
the allylamines, in the formation of η3-allylpalladium
complexes.
(Table 2, entries 5 and 6; Table 3, entry 3) whereas
substitution at both termini (Table 2, entry 3) or steri-
cally encumbered substituents at one end of the alkyne
moiety (Table 2, entry 11) hamper the cyclization reac-
tion. The examples reported in entries 18-21 (Table 2)
and 4, 5 (Table 3) illustrate the compatibility of both
electron-donating and electron-withdrawing substituents.
Path b differs from path a in the way 10 is generated
from 7. It entails the coordination of the carbon-carbon
triple bond to the palladium of the η3-allylpalladium
fragment of 7 being faster than the nucleophilic attack
of the nitrogen on the allylic portion of the complex, and
10 is generated without the intermediacy of 4. The
existence of path b is supported by the absence of the
N-allyl derivative 4b as intermediate when 1a and ethyl
2-hexen-1-yl carbonate 2b were subjected to Pd2(dba)3
and ttmpp: none of 4b was detected under these condi-
tions. The observation that the reaction of 1a with 2b
in the presence of Pd2(dba)3 and ttmpp to afford the indole
nucleus is much faster than the cyclization of 4b in the
presence of the same catalytic system also argues in favor
of a reaction mechanism not involving the N-allylation
step. In fact, it rules out the possibility that 4b reacts
as it is formed, thus making detection impossible. In
practice, the use of ttmpp completely changes the course
of the reaction providing an alternative, straightforward
route to 3-allylindoles.
Stepwise and one-pot protocols gave good results with
a variety of o-alkynyltrifluoroacetanilides and allylic
derivatives. The presence of a substituent on the central
carbon atom of the allylic system seems to be tolerated
(13) Trost, B. M.; Van Vranken, D. L. Chem. Rev. (Washington, D.C.)
1996, 96, 395. Williams, J . M. J . Synlett 1996, 705.
(14) Kurosawa, H.; Shiba, K.; Hirako, K.; Kakiuchi, K.; Ikeda, I. J .
Chem. Soc., Chem. Commun. 1994, 1099; Trost, B. M.; Weber, L.;
Strege, P. E.; Fullerton, T. J .; Dietsche, T. J . J . Am. Chem. Soc. 1978,
100, 3416.
(15) For the utilization of trialkylallylammonium salts as precursors
of η3-allylpalladium complexes, see: Doi, T.; Yanagisawa, A.; Miyaza-
wa, M.; Yamamoto, K. Tetrahedron Asymmetry 1995, 6, 389; Hunter,
R.; Simon, C. D. Tetrahedron Lett. 1988, 29, 2257.
(16) For the utilization of 3-vinylaziridines as precursors of η3-
allylpalladium complexes, see: Spears, G. W.; Nakanishi, K.; Ohfune,
Y. Synlett 1991, 91.
(17) Lemaire-Audoire, S.; Savignac, M.; Dupuis, C.; Genet, G. P.
Bull. Soc. Chim. Fr. 1995, 132, 1157. Garro-Helion, F.; Merzouk, A.;
Guibe´, F. J . Org. Chem. 1993, 58, 6109.
This route, however, does not invalidate procedures a
and b. Its scope, in fact, has been briefly investigated,

A Regioselective Synthesis of 3-Allylindoles
J . Org. Chem., Vol. 63, No. 4, 1998 1005
Ta ble 1. Liga n d s, Solven t, Tim e a n d Tem p er a tu r e Stu d ies in th e P a lla d iu m -Ca ta lyzed Cycliza tion of
o-(P h en yleth yn yl)-N-(h ex-2-en -1-yl)tr iflu or oa ceta n ilid e 4ba
3-allylindole
reaction
temp (°C)
reaction
time (h)b
overall
E:Z ratio
for 3′b
entry
catalyst
solvent
yield %c,d
3′b/3′′bc
1
2
Pd2(dba)3/dppb
Pd2(dba)3/dppb
Pd2(dba)3/dppb
Pd2(dba)3/dppb
Pd2(dba)3/dppb
Pd2(dba)3/dppe
Pd2(dba)3/(iPrO)3P
Pd2(dba)3/nBu3P
Pd2(dba)3/ttmppg,h
Pd2(dba)3/ttmppg,i
Pd2(dba)3/ttmppg,j
Pd2(dba)3/tdmppk,l
Pd2(dba)3/tdmppk,m
Pd2(dba)3/tmppn
Pd2(dba)3/(p-tol)3P
Pd2)dba)3/(o-tol)3P
Pd2(dba)3/Ph3P
Pd2(dba)3/(p-Cl-C6H4)3P
Pd2(dba)3/(C6F5)3P
Pd2(dba)3/(PhO)3P
Pd(PPh3)4
THF
80
90
100
100
100
100
90
100
100
100
100
100
100
100
100
100
100
100
100
100
80
24
24
24
3.5
24
24
24
24
6
11
15
6
15
6.5
8
- (84)
35 (34)f
64
-
-
MeCN
DME
DMSO
DME
DME
MeCN
DME
DME
DME
DME
DME
DME
DME
DME
DME
DME
DME
DME
DME
THF
64/36
65/35
71/29
75/25
81/19
-
88:12
88:12
88:12
88:12
78:22
-
3
4
62
5
52
60
- (92)
83
42 (31)
69 (18)
84
6
7
8
71/29
97/3
97/3
97/3
97/3
97/3
81/19
65/35
86:14
83:17
83:17
83:17
87:17
84:16
86:14
86:14
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
25 (69)
71 (12)
85
68
24
8
24
24
24
20
4
- (86)o
63
57/43
54/46
86:14
87:13
44 (53)
- (98)
- (75)p
73
71
64
65/35
66/34
66/34
88:12
88:12
86:14
Pd(PPh3)4
MeCN
DME
90
100
Pd(PPh3)4
2.5
a
All of the reactions were carried out under an argon atmosphere, in the presence of 0.05 equiv of the palladium catalyst, 0.1 equiv of
b
the phosphorus ligand (when added), and 5 equiv of K2CO3. In most cases, monitoring by TLC or HPLC showed that the reaction ceased
after less time than the time alloted. c Yields are given for isolated products and refer to single runs. Figures in parentheses refer to the
d
recovered starting alkyne 4b. e Relative percentages were calculated by NMR analysis. f 2-Phenylindole was isolated in 15% yield. g ttmpp
h
i
j
k
l
) tris[2,4,6-(MeO)3-C6H2]3P. Pd:ttmpp ) 1:2. Pd:ttmpp ) 1:4. Pd:ttmpp ) 1;6. tdmpp ) tris[2,6-(MeO)2-C6H3]3P. Pd:tdmpp ) 1:2.
Pd:tdmpp ) 1:6. tmpp ) tris(p-MeO-C6H4)3P. o A similar result was obtained by using a 1:6 Pd:(o-tol)3P ratio. 2-Phenylindole was
m
n
p
isolated in 14% yield.
and the examples examined have shown that the reaction
suffers from some limitations that can be overcome by
using procedures a and b. For example, the reaction of
1e (R ) n-C5H11) with 2-hexen-1-yl and cinnamyl carbon-
ates failed to afford the corresponding indole derivatives
3′l and 3′v (Table 4, entries 6 and 7) whereas they were
isolated in 69 and 75% overall yield, respectively, by
using the stepwise and one-pot protocols (Table 2, entry
13 and Table 3, entry 6). In effect, the presence of an
aryl substituent on the alkyne moiety appears to be
crucial for the success of procedure c. The substitution
pattern of the allylating agents also plays a role. Despite
the presence of the phenyl group, no indole derivative,
3f, was isolated when 1a was allowed to react with
isobutenyl carbonate, 2f, bearing a substituent on the
central carbon of the allylic system (Table 4, entry 4).
Conversely, the stepwise protocol gave 3f in an overall
89% yield (Table 2, entry 5).
amount of the indole derivative increased with time after
4b became a component of the reaction mixture.
We explored the question of whether the effect of ttmpp
on the reaction mechanism is dependent on steric or
electronic factors. Therefore, as model, we examined the
employment of Pd2(dba)3 in the presence of a sterically
demanding ligand, P(o-tol)3, and of an electron-rich ligand
that does not exhibit steric bulk, tris(p-methoxyphenyl)-
phosphine (tmpp). Using 1a and 2b as substrates, we
exposed them to Pd2(dba)3 and P(o-tol)3 [Pd:P(o-tol)3 )
1:6] and found that the reaction gave the indole nucleus
in 78% overall yield (eq 6a). None of the N-allyl inter-
mediate was discernible. Subjection of 1a and 2b to the
Pd2(dba)3/tmpp combination [Pd:tmpp ) 1:6] resulted
instead in the isolation of the N-allyl derivative, 4b, in
66% yield and of the indole derivative in 21% yield, with
the latter arising largely if not exclusively from 4b (eq
6b). Strong support for this view is provided by the
observation that monitoring the reaction showed that no
indole was present before 4b was formed, but that the
These data argue in favor of the idea that steric factors
may play a major role on the reaction mechanism. A
working hypothesis accounting for this effect envisages
that the use of a sterically demanding ligand may create

1006 J . Org. Chem., Vol. 63, No. 4, 1998
Cacchi et al.
Ta ble 2. P a lla d iu m -Ca ta lyzed N-Allyla tion of o-Alk yn yltr iflu or oa ceta n ilid es 1 a n d Cycliza tion of
o-Alk yn yl-N-a llyltr iflu or oa ceta n ilid es 4 (p r oced u r e a )

A Regioselective Synthesis of 3-Allylindoles
J . Org. Chem., Vol. 63, No. 4, 1998 1007
Ta ble 2 (Con tin u ed ) P a lla d iu m -Ca ta lyzed N-Allyla tion of o-Alk yn yltr iflu or oa ceta n ilid es 1 a n d Cycliza tion of o-Alk yn yl-
N-a llyltr iflu or oa ceta n ilid es 4 (p r oced u r e a )
an η3-allylpalladium complex 7 more prone to relieving
steric crowding in the coordination sphere of palladium
rather than allowing a nucleophilic attack by the nitrogen
atom on the allylic complex. Consequently, in the pres-
ence of ttmpp [or even tdmpp or P(o-tol)3] the ambident
electrophilic part of 7 prefers to complex with the alkyne
unit, generating 10, instead of producing the η2-olefin-
palladium complex 8.
It may be worth emphasizing that the result obtained
in the reaction of 1a and 2b in the presence of Pd2(dba)3
and (o-tol)3P (eq 6a) and the observation that the N-allyl
derivative 4b failed to cyclize to the corresponding indole
in the presence of the same catalytic system (Table 1,
entry 16), provide a further, convincing proof in favor of
path b and of the notion that the ligand may have a
strong directing effect on the reaction mechanism. In
fact, if the reaction of 1a and 2b in the presence of Pd2-
(dba)3 and (o-tol)3P proceeded through an N-allylation
step, the N-allyl derivative would be recovered unreacted
and no indole could be isolated.
As to the regioselectivity of the intramolecular nucleo-
philic attack of the indolyl carbon atom on the η3-allyl
unit,18 several factors have been found to play a role: the
nature of the allylating agent (compare, for example,
(18) Previous work on the palladium-catalyzed allylation of carbon
nucleophiles (phenyl and linear vinyl fragments) which react by
transfer from the palladium to the carbon14,19 showed that the new
carbon-carbon bond is generated preferentially at the more crowded
allyl terminus in the presence of triphenylphosphine. This tendency
has been rationalized by assuming that the sterically demanding
triphenylphosphine prefers to reside trans to the bulkier allyl terminus,
so as to minimize steric strain. In our case, even in the presence of
triphenylphosphine, the main regioisomer contains the indolyl unit
bound to the less crowded allyl terminus. This might reflect a less clear-
cut differentiation between the steric requirements of the triph-
enylphosphine and the substituted indolyl unit.

1008 J . Org. Chem., Vol. 63, No. 4, 1998
Cacchi et al.
Ta ble 3. On e-P ot Syn th esis of 3-Allylin d oles 3 fr om
o-Alk yn yltr iflu or oa ceta n ilid es 1 a n d Allyl Ester s 2
(p r oced u r e b)a
(Table 1, compare entry 14 with entry 18) effects relative
to monodentate ligands. The remarkable regioselectivity
exhibited by ttmpp most probably depends upon a
combination of steric and electronic effects. Noteworthy,
steric and electronic effects exerted by the substituent
of the indolyl unit, R, which have been found to play an
important role in the regiochemistry in the presence of
PPh3 (Table 1, entry 17; Table 2, entries 7, 12, 15, 20,
21), turn out to be unimportant in the presence of ttmpp.
The stereochemistry of the process reveals that it is
accompanied by some loss of the olefin geometry and that
this is not influenced by solvent, temperature, or ligand
effects to any extent.
In conclusion, we have developed a new, highly efficient
palladium-catalyzed synthesis of 2-substituted- and 2-un-
substituted-3-allyl indoles from readily available o-alky-
nyltrifluoroacetanilides and allyl esters which shows a
wide structural flexibility in both the alkyne and the allyl
component. In every case, almost complete regioselec-
tivity toward the 3-allylindole bearing the indolyl unit
on the less substituted allyl terminus has been achieved.
The overall yields are generally high. Despite some
limitations such as the partial loss of olefin geometry,
the process holds promise as a useful tool for the
construction of complex heterocycles containing the indole
fragment.
Exp er im en ta l Section
Melting points were determined with a Bu¨chi apparatus and
are uncorrected. All the starting materials (catalysts, ligands,
bases, alcohols, alkynes and solvents) commercially available
were used as purchased, without further purification. o-
Ethynyltrifluoroacetanilide 1f was prepared according to the
ref 2b. Allyl esters were prepared according to standard
methods. Reaction products were purified on axially com-
pressed columns (packed with SiO2, 25-40 nm, Macherey
Nagel) connected to a Gilson solvent delivery system and to a
Gilson refractive index detector. n-Hexane/EtOAc mixtures
were used for elution. 1H NMR spectra (CDCl3, unless
otherwise stated; TMS as internal standard) were recorded
at 200 MHz. 13C NMR spectra were recorded at 50.3 MHz.
Gen er a l P r oced u r e for th e P r ep a r a tion of o-Alk yn yl-
tr iflu or oa ceta n ilid es (1) fr om o-Iod oa n ilin e. o-(P h en yl-
eth yn yl)tr iflu or oa ceta n ilid e (1a ). To a solution of 2-io-
doaniline (0.850 g, 3.87 mmol) in DMF (3 mL) were added
phenylacetylene (0.51 mL, 4.65 mmol), diethylamine (2.5 mL),
Pd(OAc)2(PPh3)2 (0.029 g, 0.038 mmol), and CuI (0.015 g, 0.077
mmol). The reaction mixture was stirred at room temperature
under argon for 3 h and then poured into a separatory funnel
containing EtOAc and water saturated with NH4Cl. The
organic layer was separated and washed with water, dried
(Na2SO4), and concentrated under vacuum. The residue was
connected to a vacuum line, kept at 1 mm/Hg for 2 h, dissolved
in anhydrous THF (5 mL), and cooled in an ice bath. Trifluo-
roacetic anhydride (1.1 mL, 7.74 mmol) was added, and the
reaction mixture was stirred under argon for 0.25 h and
extracted with EtOAc and water saturated with NaHCO3. The
organic phase was dried (Na2SO4) and concentrated under
vacuum. The residue was chromatographed on silica gel
eluting with a n-hexane/EtOAc 90/10 (v/v) mixture to give 0.82
g (73% yield) of 1a . Spectral data are in good agreement with
those described in the literature.2b
Table 1, entry 22 with Table 2, entry 2); steric and
electronic effects exerted by the substituents on the
indolyl fragment, R (Table 1, entry 17; Table 2, entries
7, 12, 15, 20, 21); the number of methylene units between
the phosphorus atoms with bidentate ligands (Table 1,
entries 3, 5, 6); steric (compare, for example, Table 1,
entry 15 with the result reported in eq 6a) and electronic
Gen er a l P r oced u r e for th e P r ep a r a tion of o-Alk yn yl-
tr iflu or oa ceta n ilid es (1) fr om o-Eth yn yla n ilin e. o-[(p-
Acetylp h en yl)eth yn yl]tr iflu or oa ceta n ilid e (1h ). To a
stirred solution of 2-ethynylaniline (0.300 g, 2.56 mmol) in
DMF (1.5 mL) were added 4-iodoacetophenone (0.757 g, 3.07
mmol), diethylamine (5 mL), Pd(OAc)2(PPh3)2 (0.020 g, 0.025
mmol), and CuI (0.010 g, 0.051 mmol). The reaction mixture
was stirred at room temperature for 3 h and extracted with
(19) Keinan, E.; Sahai, M. J . Chem. Soc., Chem. Commun. 1984,
648. Temple, J . S.; Riediker, M.; Schwartz, J . J . Am. Chem. Soc. 1982,
104, 1310. Hayasi, Y.; Riediker, M.; Temple, J . S.; Schwartz, J .
Tetrahedron Lett. 1981, 22, 2629.

A Regioselective Synthesis of 3-Allylindoles
J . Org. Chem., Vol. 63, No. 4, 1998 1009
Ta ble 4. Th e Rea ction of o-Alk yn yltr iflu or oa ceta n ilid es 1 w ith Allyl Ca r bon a tes 2 in th e P r esen ce of th e P d 2(d ba )3/
ttm p p Com bin a tion (p r oced u r e c)a
EtOAc and water saturated with NH4Cl. The organic layer
was separated and washed with water, dried (Na2SO4), and
concentrated under vacuum. The crude coupling product was
kept under vacuum (1 mm/Hg) for 2 h and dissolved in
anhydrous THF (3 mL), and the resultant solution was cooled
in an ice bath. Trifluoroacetic anhydride (0.71 mL, 5.12 mmol)
was added and the reaction mixture was stirred under nitrogen
for 0.25 h and extracted with a saturated NaHCO3 solution
and ethyl acetate. The organic layer was dried (Na2SO4) and
evaporated under vacuum. The residue was chromatographed
on silica gel eluting with a n-hexane/EtOAc 85/15 (v/v) mixture
to give 0.76 g (89% yield) of 1h : mp 135-6 °C; IR (KBr) 3303,
P r ep a r a tion of o-Eth yn yla n ilin e. To a stirred solution
of o-iodoaniline (3.0 g, 13.70 mmol) in DMF (2 mL) and Et2-
NH (2 mL) were added (trimethylsilyl)acetylene (2.84 mL,
20.54 mmol), Pd(PPh3)4 (0.079 g, 0.068 mmol), and CuI (0.026
g, 0.137 mmol). The reaction mixture was stirred overnight
under argon at room temperature and poured into a separatory
funnel containing diethyl ether and 0.1 N HCl. The organic
layer was separated, and the aqueous layer was extracted
twice with diethyl ether. The combined organic phases were
dried (Na2SO4) and evaporated under vacuum. The residue
was dissolved in methanol (50 mL), and K2CO3 (0.178 g, 1.37
mmol) was added. The reaction mixture was stirred for 1 h
at room temperature, methanol was partially evaporated to
around 1/4 of its original volume, and the resultant mixture
was extracted with diethyl ether and water. The organic phase
was dried (Na2SO4), evaporated, and chromatographed on
silica gel. Eluting with a n-hexane/EtOAc 95/5 (v/v) mixture
afforded 1.31 g (82% yield) of o-ethynylaniline. Spectral data
were in good agreement with those reported in the literature.20
1
2212, 1712, 826, 769 cm-1; H NMR δ 8.83 (bs, 1 H), 8.38 (d,
J ) 8.0 Hz, 1 H), 7.99 (d, J ) 8.3 Hz, 2 H), 7.63-7.57 (m, 3
H), 7.47 (t, J ) 7.6 Hz, 1 H), 7.28-7.21 (m, 1 H), 2.64 (s, 3 H);
13C NMR δ 197.3, 158.2 (q, J ) 36 Hz), 137.2, 136.4, 132.1,
131.8, 130.7, 128.8, 127.6, 125.9, 120.0, 116.5 (q, J ) 288 Hz),
113.1, 97.2, 86.1, 26.9; MS m/e (relative intensity) 331 (M+,
68), 316 (100). Anal. Calcd for C18H12F3NO2: C, 65.24; H,
3.65; N, 4.23. Found C, 65.01; H, 3.66; N. 4.21.

1010 J . Org. Chem., Vol. 63, No. 4, 1998
Cacchi et al.
Typ ica l P r oced u r e for th e Syn th esis of o-Alk yn yl-N-
a llyltr iflu or oa ceta n ilid es (4). o-(P h en yleth yn yl)-N-(2-
h exen -1-yl)tr iflu or oa ceta n ilid e (4b). A mixture of Pd2-
(dba)3 (0.016 g, 0.017 mmol) and (diphenylphosphino)butane
(0.030 g, 0.069 mmol) in anhydrous THF (3 mL) was stirred
under argon at 60 °C for 10 min. After cooling at room
temperature, o-(phenylethynyl)trifluoroacetanilide 1a (0.200
g, 0.69 mmol) and 2-hexen-1-yl ethyl carbonate 2b (0.143 g,
0.83 mmol) were added, and the solution was stirred at 60 °C
for 1.5 h. Then, the THF was evaporated under vacuum, and
the oily residue was subjected directly to chromatography on
silica gel. Elution with a n-hexane/EtOAc 96/4 (v/v) mixture
gave 0.223 g (87% yield) of 4b, (∼88:12:: E:Z): IR (liquid film)
intensity) 247 (M+, 79), 232 (19), 206 (100). Anal. Calcd for
C
18H17N: C, 87.41; H, 6.93; N, 5.66. Found C, 87.58; H, 6.91;
N, 5.68.
Con d ition s B. 2-P h en yl-3-(2-h exen -1-yl)in d ole (3′b)
a n d 2-P h en yl-3-(h ex-1-en -3-yl)in d ole (3′′b). A mixture of
Pd2(dba)3 (0.009 g, 0.010 mmol) and tris(2,4,6-(MeO)3C6H4)3P
(0.065 g, 0.12 mmol) in DME (3 mL) was stirred under argon
at 60 °C for 10 min. The solution was cooled at room
temperature, and 0.150 g (0.40 mmol) of o-(phenylethynyl)-
N-(2-hexen-1-yl)trifluoroacetanilide 4b and 0.278 g (2.02 mmol)
of K2CO3 were added. The reaction mixture was stirred at
100 °C for 15 h and workup as before to afford a residue which
was purified by chromatography eluting with a n-hexane/
EtOAc 94/6 (v/v) mixture to provide 0.093 g (84% yield) of 3′b
and 3′′b as a 97:3 mixture. The presence of 3′′b was indicated
by a multiplet (ddd) at 6.28-6.18 (H2CdCHCH) and a mul-
tiplet at 5.09-4.99 ppm (H2CdCHCH); 3′b was obtained as
an approximately 83:17 E:Z mixture: IR (liquid film) 3329,
1
2221, 1696, 761 cm-1; H NMR δ 7.65-7.58 (m, 1 H), 7.52-
7.47 (m, 2 H), 7.37-7.30 (m, 5 H), 7.20-7.16 (m, 1 H), 5.60-
5.41 (m, 2 H), 4.80 (dd, J ) 13.5 Hz, J ) 2.8 Hz, 1 H), 4.02-
3.91 (m, 1 H), 1.96-1.79 (m, 2 H), 1.37-1.11 (m, 2 H), 0.82 (t,
J ) 7.2 Hz, 3 H); 13C NMR δ 157.0 (q, J ) 35 Hz, CO-CF3),
140.0, 137.5, 132.8, 131.8, 130.0, 129.0, 128.9, 128.8, 128.6,
123.7, 122.5, 116.5 (q, J ) 288 Hz, CF3), 95.5, 85.0, 53.1
(NCH2CH)CH, 4b E isomer) and 47.2 (NCH2CH)CH, 4b Z
isomer), 34.2, 22.2, 13.7; MS m/e (relative intensity) 371 (M+,
20), 342 (23), 328 (65), 302 (100), 220 (73), 165 (69).
1
1688, 740, 695 cm-1; H NMR δ 8.42 (bs, 1 H), 7.96-7.08 (m,
9 H), 5.69 (dt, J ) 15.2 Hz, J ) 5.6 Hz, 1 H), 5.50 (dt, J ) 15.2
Hz, J ) 6.4 Hz, 1 H), 3.61 (d, J ) 6.5 Hz, 0.28 H, CCH2CH),
3′b Z isomer), 3.57 (dd, J ) 5.6 Hz, J ) 1.6 Hz, 1.72 H,
CCH2CH), 3′b E isomer), 2.24-2.14 (m, 0.34 H, dCHCH2CH2,
3′b Z isomer), 2.02-1.92 (m, 1.72 H, dCHCH2CH2, 3′b E
isomer),1.50-1.12 (m, 2H), 0.95 (t, J ) 7.3 Hz, 0.42 H, CH3-
CH2, 3′b Z isomer), 0.85 (t, J ) 7.3 Hz, 2.58 H, CH3CH2, 3′b
E isomer); 13C NMR δ 136.0, 134.5, 122.2, 119.5, 111.5, 110.8,
34.6 (CCH2CH), 3′b E isomer) and 29,7 (CCH2CH), 3′b Z
isomer), 22.7, 13.7; MS m/e (relative intensity) 275 (M+, 22),
242 (36), 208 (100), 193 (54).
o-(P h en yleth yn yl)-N-[(3-(p-m eth oxyp h en yl)-2(E)-p r o-
p en -1-yl]tr iflu or oa ceta n ilid e (4p ) was obtained by allowing
4a (0.150 g, 0.45 mmol) to react with p-methoxyphenyl iodide
(0.160 g, 0.68 mmol), in the presence of K2CO3 (0.078 g, 0.57
mmol), KOAc (0.045 g, 0.45 mmol), and Pd(OAc)2 (0.005 g,
0.023 mmol) in DMF (2 mL), at 80 °C, under argon for 2 h.
The usual workup and purification by chromatography on
silica gel, and elution with a n-hexane/EtOAc 90/10 (v/v)
mixture, gave 4p in 46% yield as a colorless oil: IR (liquid
On e-P ot Syn th esis of 3-Allylin d oles fr om o-Alk yn yl-
tr iflu or oa ceta n ilid es (1) a n d Allyl Ester s (2) (P r oced u r e
b). 2-[o-(2-Tetr a h yd r op yr a n yloxy)p h en yl]-3-cin n a m ylin -
d ole (3′y). To a solution of 2-[o-(2-tetrahydropyranyloxy)-
phenyl]trifluoroacetanilide 1i (0.200 g, 0.514 mmol) and
cinnamyl ethyl carbonate 2c (0.127 g, 0.617 mmol) in anhy-
drous THF (3 mL) was added, under argon, Pd(PPh3)4 (0.029
g, 0.026 mmol). The mixture was stirred at 60 °C for 3 h, till
the disappearance of 1i. Then, K2CO3 (0.355 g, 2.57 mmol)
was added, and the reaction was stirred at 80 °C for 24 h. After
the usual workup, the residue was chromatographed on silica
gel. Eluting with a n-hexane/EtOAc (95/5 v/v) mixture gave
0.135 g (64% yield) of 3′y: mp 40-2 °C; IR (KBr) 3427, 744
cm-1; 1H NMR δ 9.03 (bs, 1 H), 7.65 (d, J ) 7.6 Hz, 1 H), 7.54
(dd, J ) 7.7 Hz, J )1.2 Hz, 1 H), 7.44-6.95 (m, 11 H), 6.56
(dt, J ) 15.8 Hz, J ) 4.0 Hz, 1 H), 6.40 (d, J ) 15.8 Hz, 1 H),
5.40-5.35 (m, 1 H), 3.90-3.82 (m, 1 H), 3.76 (d, J ) 4.0 Hz, 2
H), 3.56-3.47 (m, 1 H), 1.80-1.90 (m, 6 H); 13C NMR δ 154.5,
138.0, 135.8, 132.3, 131.4, 130.2, 130.0, 129.3, 128.6, 128.1,
127.0, 126.2, 123.1, 122.7, 122.1, 119.5, 119.3, 117.1, 111.1,
110.7, 98.7, 63.2, 30.8, 28.8, 25.2, 19.6; MS m/e (relative
intensity) 409 (M+, 8), 325 (100). Anal. Calcd for C28H27NO2:
C, 82.11; H, 6.65; N, 3.42. Found C, 82.30; H, 6.67; N, 3.41.
film) 2221, 1696, 752 cm-1 1H NMR δ 7.63-7.31 (m, 9 H),
;
7.20 (d, J ) 8.7 Hz, 2 H), 6.78 (d, J ) 8.7 Hz, 2 H), 6.36 (d, J
) 15.8 Hz, 1 H), 6.19-6.05 (m, 1 H), 4.90 (dd, J ) 14.3 Hz, J
) 6.1 Hz, 1 H), 4.18 (dd, J ) 14.3 Hz, J ) 7.7 Hz, 1 H), 3.75
(s, 3 H); 13C NMR δ 159.6, 157.2 (q, J ) 37 Hz), 140.1, 134.9,
132.9, 131.8, 129.8, 129.3, 129.1, 129.0, 128.6, 127.9, 123.6,
122.4, 119.6, 116.4 (q, J ) 289 Hz), 113.6, 95.6, 85.0, 53.3,
53.6; MS m/e (relative intensity) 435 (M+, 81), 147 (100). Anal.
Calcd for C26H20F3NO2: C, 71.72; H, 4.63; N, 3.22. Found C,
71.53; H, 4.62; N, 3.20.
o-(P h en yleth yn yl)-N-[3-(p-a cetylp h en yl)-2(E)-p r op en -
1-yl]tr iflu or oa ceta n ilid e (4q) was prepared as described for
4p from 4a and p-acetylphenyl iodide, in 66% yield: mp 64-6
°C; IR (KBr) 2221, 1704, 1696, 752 cm-1; 1H NMR δ 7.77 (d, J
) 8.4 Hz, 2 H), 7.55-7.21 (m, 11 H), 6.45 (d, J ) 15.4 Hz, 1
H), 6.39-6.32 (m, 1 H), 4.81 (dd, J ) 13.4 Hz, J ) 5.3 Hz, 1
H), 4.28 (dd, J ) 13.4 Hz, J ) 6.2 Hz, 1 H), 2.51 (s, 3 H); 13C
NMR δ 197.6, 157.3 (q, J ) 38 Hz), 140.8, 140.0, 136.4, 134.1,
133.0, 131.8, 129.6, 129.1, 129.1, 128.8, 128.6, 127.5, 126.7,
125.0, 123.6, 122.2, 116.3 (q, J ) 289 Hz), 95.8, 84.9, 53.6,
26.7; MS m/e (relative intensity) 447 (M+, 31), 404 (49), 350
(100). Anal. Calcd for C27H20F3NO2: C, 72.46; H, 4.51; N, 3.13.
Found C, 72.28; H, 4.49; N, 3.15.
Rea ction of o-Alk yn yltr iflu or oa ceta n ilid es 1 w ith Al-
lyl Ca r bon a tes 2 in th e P r esen ce of th e P d 2(d ba )3/ttm p p
Com bin a tion (P r oced u r e c). 2-P h en yl-3-(2-octen -1-yl)-
in d ole (3′z) a n d 2-P h en yl-3-(1-octen -3-yl)in d ole (3′′z). A
mixture of Pd2(dba)3 (0.012 g, 0.013 mmol) and ttmpp (0.083
g, 0.156 mmol) in anhydrous THF (3 mL) was stirred at 60 °C
for 10 min, under an argon atmosphere. The solution was
cooled at room temperature and o-(phenylethynyl)trifluoroac-
etanilide 1a (0.150 g, 0.519 mmol) and 2-octenyl ethyl carbon-
ate 2h (0.125 g, 0.623 mmol) were added, and the mixture was
stirred at 60 °C for 3 h. THF was evaporated under vacuum,
and the residue was chromatographed on silica gel eluting with
a n-hexane/EtOAc 94/6 (v/v) mixture to give 0.126 g (80% yield)
of 3′z and 3′′z as a 97:3 mixture. The presence of 3′′z was
indicated by a multiplet (ddd) at 6.23-6.20 ppm (H2CdCHCH)
and a multiplet at 5.09-4.99 ppm (H2CdCHCH). 3′z was
obtained as an approximately 88:12 E:Z mixture: IR (liquid
Cycliza tion of o-Alk yn yl-N-a llyltr iflu or oa ceta n ilid es
(P r oced u r e a ). Con d ition s A. 2-P h en yl-3-(2-Meth yl-2-
p r op en -1-yl)in d ole (3f). To a solution of o-(phenylethynyl)-
N-(2-methyl-2-propen-1-yl)trifluoroacetanilide 4f (0.120 g, 0.35
mmol) in acetonitrile (3 mL) were added K2CO3 (0.241 g, 1.75
mmol) and Pd(PPh3)4 (0.020 g, 0.017 mmol). The resultant
mixture was refluxed under argon, at 90 °C, for 3 h. Then it
was cooled at room temperature, diluted with ethyl acetate,
and washed with water. The organic layer was dried (Na2-
SO4) and evaporated under vacuum, and the residue was
chromatographed on silica gel. Eluting with a n-hexane/EtOAc
97/3 (v/v) mixture afforded 0.082 g (95% yield) of 3f as a white
1
solid: mp 62-3 °C; IR (KBr) 3344, 769, 744 cm-1; H NMR δ
8.05 (bs, 1 H), 7.60-7.31 (m, 7 H), 7.23-7.07 (m, 2H), 4.83
(bs, 1 H), 4.66 (bs, 1 H), 3.53 (s, 2 H), 1.83 (s, 3 H); 13C NMR
δ 145.0, 122.4, 119.8, 111.2, 110.9, 33.1, 23.2; MS m/e (relative
1
film) 3329, 1687, 1622, 741, 695 cm-1; H NMR δ 7.92 (bs, 1
H), 7.53-7.09 (m, 9 H), 5.67 (dt, J ) 15.1 Hz, J ) 5.5 Hz, 1
H), 5.49 (dt, J ) 15.1 Hz, J ) 6.6 Hz, 1 H), 3.54 (d, J ) 5.5
(20) Arcadi, A.; Cacchi, S.; Marinelli, F. Tetrahedron Lett. 1989, 30,
2581.

A Regioselective Synthesis of 3-Allylindoles
J . Org. Chem., Vol. 63, No. 4, 1998 1011
Hz, 2 H), 1.99-1.82 (m, 2 H), 1.38-0.60 (m, 9 H); 13C NMR δ
142.1, 113.9 (H2C)), 41.0, 34.5, 32.7, 31.8, 31.6, 29.4, 28.0,
27.7, 22.7, 14.3, 14.2; MS m/e (relative intensity) 303 (M+, 33),
232 (100), 208 (57).
chetto of the Istituto Superiore di Sanita` for obtaining
the mass spectra of new products.
Su p p or tin g In for m a tion Ava ila ble: Characterization
data for 1b-e,g, i,j, 4a , c-o, r , s, 3a , 3′c, ′e, ′g-j, ′l-x, ′za -
-zc (11 pages). This material is contained in libraries on
microfiche, immediately follows this article in the microfilm
version of the journal, and can be ordered from the ACS; see
any current masthead page for ordering information.
Ack n ow led gm en t. The authors gratefully acknowl-
edge the Consiglio Nazionale delle Ricerche (CNR) and
the Ministero dell′Universita` e della Ricerca Scientifica
(MURST) for financial support of this research. The
authors are also greatly indebted to Dr. Luciana Tur-
J O971237P