Synthesis of 2-quinolones via palladium-catalyzed carbonylative annulation of internal alkynes by N-substituted o-iodoanilines

Article abstract of DOI:10.1021/jo049149+

The palladium-catalyzed annulation of internal alkynes by N-substituted o-iodoanilines under 1 atm of carbon monoxide results in the formation of 3,4-disubstituted 2-quinolones. The nature of the substituent on the nitrogen is crucial to obtaining high yields of 2-quinolones. The best results are obtained using alkoxycarbonyl, p-tolylsulfonyl, and trifluoroacetyl substituents. The nitrogen substituent is lost during the course of the reaction resulting in the formation of N-unsubstituted 2-quinolones. A variety of internal alkynes, bearing alkyl, aryl, heteroaryl, hydroxyl, and alkoxyl substituents, are effective in this process. Electron-rich and electron-poor N-substituted o-iodoanilines, as well as heterocyclic analogues, can be employed as annulating agents.

Full text of DOI:10.1021/jo049149+

Syn th esis of 2-Qu in olon es via P a lla d iu m -Ca ta lyzed Ca r bon yla tive
An n u la tion of In ter n a l Alk yn es by N-Su bstitu ted o-Iod oa n ilin es
Dmitry V. Kadnikov and Richard C. Larock*
Department of Chemistry, Iowa State University, Ames, Iowa 50011
larock@iastate.edu
Received May 19, 2004
The palladium-catalyzed annulation of internal alkynes by N-substituted o-iodoanilines under 1
atm of carbon monoxide results in the formation of 3,4-disubstituted 2-quinolones. The nature of
the substituent on the nitrogen is crucial to obtaining high yields of 2-quinolones. The best results
are obtained using alkoxycarbonyl, p-tolylsulfonyl, and trifluoroacetyl substituents. The nitrogen
substituent is lost during the course of the reaction resulting in the formation of N-unsubstituted
2-quinolones. A variety of internal alkynes, bearing alkyl, aryl, heteroaryl, hydroxyl, and alkoxyl
substituents, are effective in this process. Electron-rich and electron-poor N-substituted o-
iodoanilines, as well as heterocyclic analogues, can be employed as annulating agents.
In tr od u ction
namely the intermolecular insertion of an internal alkyne
into an arylpalladium bond in preference to the insertion
of CO. We envisioned that this process might be extended
to o-iodoanilines, which would lead to the formation of
3,4-disubstituted 2-quinolones (eq 2).
In the last two decades, the development of transition-
metal-catalyzed multicomponent processes, which involve
the formation of several carbon-carbon and/or carbon-
heteroatom bonds, has significantly expanded the arsenal
of synthetic organic chemistry, allowing easy access to
complex molecular structures from fairly simple precur-
sors. In particular, palladium-catalyzed annulations of
unsaturated molecules, such as 1,3-dienes, allenes, and
internal alkynes, with various functionally substituted
aryl and vinylic halides provide a concise and efficient
route to a wide variety of hetero- and carbocycles.1 The
introduction of an additional component, such as carbon
monoxide, into these processes would further increase
their efficiency and synthetic utility. Several examples
of palladium-catalyzed reactions of o-iodophenols and
o-iodoanilines with norbornene,2 norbornadiene,3 allenes,4
and terminal alkynes3a,5 in the presence of CO have been
reported.
The 2-quinolone core is widely found in various alka-
loids, many of which possess interesting biological activ-
ity. There has been considerable interest in developing
2-quinolones as anticancer,7 antiviral,8 and antihyper-
tensive agents.8 4-Substituted 3-phenyl-2-quinolones ex-
hibit high affinity in binding to the glycine site of the
N-methyl-D-aspartate receptor, and such antagonists
have promise for the treatment of several central nervous
system disorders.9 Amides of 3-hydroxy- and 3-alkyl-4-
carboxylic acids of 2-quinolones also exhibit high affinity
for the 5-HT3 serotonin receptor.10 2-Quinolones are also
valuable intermediates in organic synthesis, since they
We have recently reported that the palladium-cata-
lyzed carbonylative annulation of internal alkynes with
o-iodophenols results in the exclusive formation of cou-
marins (eq 1).6 This process exhibits a unique reactivity,
(5) (a) Torii, S.; Okumoto, H.; Xu, L. H.; Sadakane, M.; Shostak-
ovsky, M. V.; Ponomaryov, A. B.; Kalinin, V. N. Tetrahedron 1993, 49,
6773. (b) Ciattini, P. G.; Morera, E.; Ortar, G.; Rossi, S. S. Tetrahedron
1991, 47, 9449. (c) Miao, H.; Yang, Z. Org. Lett. 2000, 2, 1765.
(6) (a) Kadnikov, D. V.; Larock, R. C. Org. Lett. 2000, 2, 3643. (b)
Kadnikov, D. V.; Larock, R. C. J . Org. Chem. 2003, 68, 9423.
(7) Ferrer, P.; Avendan˜o, C.; So¨llhuber, M. Liebigs Ann. 1995, 1895.
(8) Afonso, A.; Weinstein, J .; Gentles, M. J . (Schering Corp.) PCT
Int. Appl. WO 9203327, 1992.
(9) (a) McQuaid, L. A.; Smith, E. C. R.; Lodge, D.; Pralong, E.; Wikel,
J . H.; Calligaro, D. O.; O’Malley, P. J . J . Med. Chem. 1992, 35, 3423.
(b) Kugalowski, J . J .; Rowley, M.; Leeson, P. D.; Mawer, I. M. (Merck
Sharp and Dohme Ltd., U.K.) Eur. Pat. Appl. EP 481676, 1992. (c)
Carling, R. W.; Leeson, P. D.; Moore, K. W.; Moyer, C. R.; Duncton,
M.; Hudson, M. L.; Baker, R.; Foster, A. C.; Grimwood, S.; Kemp, J .
A.; Marshall, G. R.; Tricklebank, M. D.; Saywell, K. L. J . Med. Chem.
1997, 40, 754.
(1) (a) Larock, R. C. J . Organomet. Chem. 1999, 576, 111. (b) Larock,
R. C. Pure Appl. Chem. 1999, 71, 1435.
(2) (a) Moinet, C.; Fiaud, J .-C. Synlett 1997, 97. (b) Grigg, R.; Khalil,
H.; Levett, P.; Virica, J .; Sridharan, V. Tetrahedron Lett. 1994, 35,
3197.
(3) (a) An, Z.-W.; Catellani, M.; Chiusoli, G. P. J . Organomet. Chem.
1989, 371, C51. (b) An, Z.-W.; Catellani, M.; Chiusoli, G. P. Gazz. Chim.
Ital. 1990, 120, 383.
(4) (a) Okuro, K.; Alper, H. J . Org. Chem. 1997, 62, 1566. (b) Grigg,
R.; Liu, A.; Shaw, D.; Suganthan, S.; Woodall, D. E.; Yoganathan, G.
Tetrahedron Lett. 2000, 41, 7125.
(10) Hayashi, H.; Miwa, Y.; Miki, I.; Ichikawa, S.; Yoda, N.; Ishii,
A.; Kono, M.; Suzuki, F. J . Med. Chem. 1992, 35, 4893.
10.1021/jo049149+ CCC: $27.50 © 2004 American Chemical Society
Published on Web 08/31/2004
6772
J . Org. Chem. 2004, 69, 6772-6780

Pd-Catalyzed Synthesis of 2-Quinolones
TABLE 1. Effect of th e Su bstitu en t on th e Nitr ogen (eq
4)a
are easily converted into 2-chloro- and then 2-amino-
quinoline derivatives.11
The classic methods of synthesis of 2-quinolones in-
clude a base-catalyzed intramolecular aldol condensation
leading to formation of the C3-C4 bond (the Friedla¨nder
synthesis12 and the Camps modification of the Fried-
la¨nder synthesis13) and an acid-catalyzed cyclization of
â-ketoanilides forming the C9-C4 bond (the Knorr
synthesis14). While a wide variety of 3- or 4-substituted
2-quinolones can be synthesized using these methods, the
utility of these reactions for the synthesis of 3,4-disub-
stituted 2-quinolones is quite limited.15 Several other
methods for the synthesis of 3,4-disubstituted 2-quino-
lones, such as the titanium-promoted intramolecular
reductive cyclization of N-(2-acylphenyl) R-ketoamides,16
the AlCl3-promoted annulation of terminal and internal
alkynes,17 and the electrocyclic closure of 2-isocyana-
tostyrenes generated in situ by oxidation of the corre-
sponding 2-cyanostyrenes with m-CPBA,18 have been
reported. The yields of these processes, however, are often
low, and the variety of substrates is very limited. Very
few examples of the palladium-catalyzed synthesis of
2-quinolones have been reported, and the utility of these
processes is again severely limited. Two of these methods
are based on the Heck reaction of 2-iodoanilines19 or
2-bromonitrobenzenes20 with â-substituted acrylic acids,
followed by intramolecular cyclization, and another method
involves the intramolecular carbonylation of (Z)-2-ac-
etamino-R-bromostyrene.21
% recovery of
starting material
entry
R
% yield of 1
1
2
3
4
5
6
7
8
H
CH3
0
0b
26
55
12
0
0
20
0
0
0
0
SO2CF3
COCH3
CON(CH3)2
CHO
SO2CH3
SO2-p-Tol
COCF3
CO2CH2CH3
CO2C(CH3)3
traces
12
23c
33
40
56d
56
9
10
11
71e
75
a
Standard reaction conditions: N-substituted o-iodoaniline (0.5
mmol), 4-octyne (2.5 mmol), pyridine (1 mmol), n-Bu4NCl (0.5
mmol), and Pd(OAc)2 (5 mol %, 0.025 mmol) under 1 atm of CO in
b
DMF (5 mL) at 100 °C for 24 h. N-Methylaniline was isolated in
d
33% yield. c Isolated as a 70:30 mixture of 1 and 3. The reaction
was run for 48 h. e Isolated as an 85:15 mixture of 1 and 4.
these reactions, we isolated small amounts of byproducts
apparently arising from the attack of the amino group
of 2-iodoaniline on the acylpalladium complex formed
from another molecule of 2-iodoaniline. Thus, we specu-
late that the failure of the annulation process can be
attributed to the high nucleophilicity of the amino group.
Therefore, protection of the amino group was explored
as a possible strategy to alleviate the problem. Indeed,
the reaction of 4-octyne with N-tosyl-o-iodoaniline under
the standard coumarin reaction conditions afforded 3,4-
dipropyl-2-quinolone (1) in a 56% yield. Unexpectedly,
the tosyl group was removed during the course of the
reaction, and no traces of the quinolone containing the
tosyl group were detected. The same yield of 2-quinolone
1 was obtained when the reaction was run at 100 °C,
though it required 48 h to reach completion.
Herein, we report a new method for the synthesis of
2-quinolones via palladium-catalyzed carbonylative an-
nulation of internal alkynes by N-substituted o-iodoa-
nilines.
Resu lts a n d Discu ssion
Op tim iza tion of th e Rea ction Con d ition s. Applica-
tion of our standard reaction conditions for the synthesis
of coumarins to the reaction of o-iodoaniline and various
internal alkynes, such as 4-octyne and diphenylacetylene,
in the presence of 1 atm of CO afforded neither the
desired 2-quinolone product 1 nor the isomeric 4-qui-
nolone 2 (eq 3). Variations of the palladium catalyst, the
Encouraged by our initial success, we examined the
effect of a variety of protecting groups on this process
(eq 4 and Table 1), including an alkyl (entry 2), sulfonyl
(entries 3, 7, and 8), acyl (entries 4, 6, and 9), aminocar-
bonyl (entry 5), and alkoxycarbonyl groups (entries 10
and 11). None of the desired product was obtained when
base, and the temperature or the addition of phosphine
ligands did not result in any improvement. In some of
(11) (a) Anzini, M.; Cappelli, A.; Vomero, S. J . Heterocycl. Chem.
1991, 28, 1809. (b) Godard, A.; Fourquez, J . M.; Tamion, R.; Marsais,
F.; Que´guiner, G. Synlett 1994, 235.
a simple alkyl substituent was employed (entry 2). Only
the reduction product, N-methylaniline, was isolated in
a 33% yield, along with 26% of the unreacted starting
material. Analysis of the results with other substituents
(entries 3-11) shows that the yield of quinolone 1
depends on the electron-withdrawing ability of the sub-
stituent on the nitrogen, as measured by the pKa (NH)
(12) Quinolines; J ones, G., Ed.; J ohn Wiley & Sons: London, New
York, Sydney, Toronto, 1977; pp 181-187.
(13) Quinolines; J ones, G., Ed.; J ohn Wiley & Sons: London, New
York, Sydney, Toronto, 1977; pp 191-194.
(14) Quinolines; J ones, G., Ed.; J ohn Wiley & Sons: London, New
York, Sydney, Toronto, 1977; pp 151-158.
(15) Searles, A. L.; Ressler, D. J . Am. Chem. Soc. 1958, 80, 3656.
J . Org. Chem, Vol. 69, No. 20, 2004 6773

Kadnikov and Larock
endented. Bolotin and co-workers reported that the
condensation of N-tosyl-2-aminobenzaldehyde with phe-
nylacetic acid afforded only unprotected 3-phenyl-2-(1H)-
quinolone in a 90% yield.23 Similarly, the palladium-
catalyzed carbonylation of Z-2-acetamino-R-bromostyrene
also produced only the unprotected 2-quinolone.21
In our studies, only in the reactions of N,N-dimethyl-
N′-(2-iodophenyl)urea (entry 5) and ethyl N-(2-iodophe-
nyl)carbamate (entry 10) was the protected quinolone
isolated in 7% and 11% yields, respectively. Since the
reaction with ethyl N-(2-iodophenyl)carbamate afforded
one of the best yields of the desired product, ways to
achieve complete deprotection during the reaction have
been explored. Increasing the amount of pyridine to 5
equiv did not produce the desired result; the yields of 1
and 4 were 47% and 18%, respectively. Increasing the
temperature to 120 °C led to the disappearance of 4, but
the yield of 1 was only 57%. While variation of the
reaction conditions did not succeed in getting rid of 4,
compound 4 can be readily converted to 1 by treatment
with 1 M ethanolic NaOH for 30 min at room tempera-
ture. When the entire reaction mixture from the annu-
lation of 4-octyne was treated with 1 M ethanolic NaOH,
quinolone 1 was obtained in a 72% yield (eq 5). As
F IGURE 1. Dependence of the yield of 1 on the pKa (NH) of
PhNHR (the R group is shown next to the data points).
of the corresponding N-substituted anilines22 (Figure 1).
High yields of the quinolone were obtained when strongly
electron-withdrawing groups, such as tosyl or trifluoro-
acetyl, are employed (entries 8 and 9), while the use of
weaker electron-withdrawing groups, such as an acetyl
group (entry 4), results in a significant decrease in the
yield. It is worth noting that the pKa’s of N-tosylaniline
(8.40) and trifluoroacetanilide (9.51) are close to the pKa
of phenol (9.98). Therefore, the effect of the electron-
withdrawing strength of the nitrogen substituent likely
indicates that deprotonation of the NH group occurs
during the reaction. The best results, however, have been
obtained by employing alkoxycarbonyl substituents (en-
tries 10 and 11), even though these groups are not as
strong electron-withdrawing groups as tosyl or trifluo-
roacetyl (pKa for PhNHCO2Et ) 13.8, Figure 1). More-
over, the reaction of an o-iodoaniline bearing an ex-
tremely strong electron-withdrawing group, such as
trifluoromethanesulfonyl (entry 3), did not afford any of
the desired product. Due to its high acidity (pKa ) 4.55),
the N-(2-iodophenyl)trifluoromethanesulfonamide is prob-
ably significantly deprotonated under the reaction condi-
tions (pKa of pyridine is 5.20), and the anionic species is
apparently completely inactive in the annulation reac-
tions. Thus, a variety of factors determines the suitability
of the protecting group, and careful selection of such a
group is paramount for the success of the reaction.
It should be mentioned here that the reactions of all
but two N-substituted o-iodoanilines shown in Table 1
afforded only the unprotected quinolone 1. The in situ
deprotection of the 2-quinolone nitrogen is not unprec-
indicated in Table 1, the unprotected product 1 can be
obtained as the only product in a 75% yield by employing
tert-butyl N-(2-iodophenyl)carbamate (entry 11). Al-
though this higher yield using a simpler procedure makes
the tert-butyl carbamate appear to be the protecting
group of choice, in most of our studies the ethyl carbam-
ate moiety was employed because it is easier to prepare
and handle (ethyl N-(2-iodophenyl)carbamate is a solid,
while the tert-butyl carbamate is a viscous liquid).
The effects of the palladium catalyst, the base, the
chloride source, the stoichiometry of the reactants, the
reaction time, and the temperature were next examined
using N-ethoxycarbonyl and N-tosyl derivatives of 2-io-
doaniline. The results of this study are summarized in
Table 2.
(16) Fu¨rstner, A.; J umbam, D. N.; Shi, N. Z. Naturforsch. B 1995,
50, 326.
(17) Merault, G.; Bougeois, P.; Duffaut, N. Bull. Soc. Chim. Fr. 1974,
1949.
(18) Kobayashi, K.; Kitamura, T.; Yoneda, K.; Morikawa, O.; Kon-
ishi, H. Chem. Lett. 2000, 798.
(19) Cortese, N. A.; Ziegler, C. B., J r.; Hrnjez, B. J .; Heck, R. F. J .
Org. Chem. 1978, 43, 2952.
(20) Holzapfel, C. W.; Dwyer, C. Heterocycles 1998, 48, 215.
(21) Mori, M.; Chiba, K.; Ohta, N.; Ban, Y. Heterocycles 1979, 13,
329.
(22) (a) Homer, R. B.; J ohnson, C. D. In The Chemistry of Amides;
Zabicky, J ., Ed.; J ohn Wiley & Sons: New York, 1970; p. 239 (b) King,
J . F. In The Chemistry of Sulfonic Acids, Esters and Their Derivatives;
Saul Patai, Z. R., Ed.; J ohn Wiley & Sons: New York, 1991; p 253. (c)
Vontor, T.; Vecera, M. Collect. Czech. Chem. Commun. 1973, 38, 3139.
First, various palladium catalysts have been evaluated.
It can be seen from the results that Pd(OAc)2 and Pd-
(dba)2 are generally equally effective in this process
(compare entries 1 and 2, and 6 and 7). Surprisingly,
palladium complexes containing phosphine ligands are
completely ineffective in the reactions with ethyl N-(2-
iodophenyl)carbamate (entries 3-5). Even the simple
(23) Bolotin, B. M.; Chernova, N. I.; Zeryukina, L. S. Zh. Vses. Khim.
Obshchest. 1972, 17, 460; Chem. Abstr. 1977, 139760.
6774 J . Org. Chem., Vol. 69, No. 20, 2004

Pd-Catalyzed Synthesis of 2-Quinolones
TABLE 2. Effect of th e P a lla d iu m Ca ta lyst on th e Yield
of Qu in olon e 1 (eq 4)a
TABLE 3. Effect of th e Rea ction Tim e a n d
Tem p er a tu r ea
entry
R
catalyst
Pd(OAc)2
Pd(dba)2
Pd(PCy3)2
Pd(OAc)2 + 2PPh3
Pd(PPh3)4
Pd(OAc)2
Pd(dba)2
base
% yieldb
entry
R
T (°C)
time (h)
% yield of 1b
1
2
3
4
5
6
7
8
9
10
11
12
13
CO2Et
pyridine
71 (85:15)
66 (86:14)
13 (62:38)
6
2
56
56
50
1
2
3
4
5
6
7
8
9
Ts
80
100
48
24
48
12
24
12
24
6
30
42
56
47
120
80
56
Ts
pyridine
CO2Eta
37 (8:92)
52 (19:81)
71 (41:59)
75 (69:31)
71 (85:15)
57 (100:0)
Pd(OAc)2 + 2PPh3
Pd(OAc)2
100
NaOAc
NEt3
0
12
24
24
22c
10
11
CO2Et
NEt3
37d
120
pyridinee
pyridinef
66 (76:24)
72 (86:14)
a
Standard reaction conditions: N-substituted o-iodoaniline (0.5
mmol), 4-octyne (2.5 mmol), Pd(OAc)2 (5 mol %, 0.025 mmol),
pyridine (1 mmol), and n-Bu4NCl (0.5 mmol) under 1 atm of CO
a
Standard reaction conditions: N-substituted o-iodoaniline (0.5
mmol), 4-octyne (2.5 mmol), pyridine (1 mmol), n-Bu4NCl (0.5
mmol), and Pd catalyst (5 mol %, 0.025 mmol) with or without
PPh3 (10 mol %, 0.05 mmol) under 1 atm of CO in DMF (5 mL) at
b
in DMF (5 mL). Ratio of 1/4 in parentheses.
100 °C for 24 h. The ratio of 1/4 is given in parentheses. c N-
b
need to be run at 120 °C to reach completion in 24 h
(entries 1-5). On the other hand, the annulation is
complete after 6 h at 100 °C, when ethyl N-(2-iodophe-
nyl)carbamate is employed. However, the major product
after 6 h is the protected quinolone 4. This compound is
hydrolyzed to 1 fairly slowly under the reaction condi-
tions, and the hydrolysis is incomplete even after 24 h
at 100 °C. At higher temperatures (120 °C), the carbam-
ate affords only 1 after 24 h, but the yield is considerably
lower (entry 11). At lower temperatures (80 °C), the
carbonylative annulation is notably slower. The reaction
requires 24 h to reach completion, and even in this case
the overall yield is only 52% (entries 6 and 7). Interest-
ingly, little deprotection occurs at 80 °C.
Scop e a n d Lim ita tion s. The carbonylative annula-
tion of various internal alkynes with ethyl N-(2-iodophe-
nyl)carbamate was subsequently examined to determine
the scope and limitations of this reaction. The reactions
were run under the following “optimized” conditions: 0.5
mmol of the 2-iodoaniline carbamate derivative, 5 mol
% Pd(OAc)2, 2 equiv of pyridine, and 1 equiv of n-Bu4-
NCl in 5 mL of DMF at 100 °C for 12 h. The crude
products were treated for 30 min at room temperature
with 1 M ethanolic NaOH to complete the hydrolysis of
the carbamate protecting group. The results of this study
are summarized in Table 4.
The carbonylative annulation of dialkyl (entries 1-5),
as well as alkylaryl acetylenes (entries 6-9), affords the
desired quinolones in 56-82% yields. These yields are
comparable or slightly better than the yields of the
corresponding coumarins (for example, with 4-octyne 71%
vs 63% and with 1-phenyl-1-butyne 82% vs 78%, respec-
tively). Moreover, no significant decrease in the yield of
the quinolones is observed when the amount of the
internal alkyne is reduced to 3 equiv (compare entries 1
and 2, 4 and 5, and 6 and 7). Even the use of only 2 equiv
of 4-octyne in the reaction with ethyl N-(2-iodophenyl)-
carbamate afforded 2-quinolone 1 in a satisfactory yield
(entry 3). The carbonylative annulation of 1-phenyl-1-
propyne (entry 8) and 1-(2-methoxyphenyl)-1-butyne
(entry 9) employing 3 equiv of the alkynes also afforded
the desired quinolones in good yields.
d
Tosyl-2,3-dipropylindole was obtained in a 43% yield. N-Ethoxy-
carbonyl-2,3-dipropylindole was obtained in a 15% yield. e No
chloride source was added. f 1 equiv of LiCl was used as a chloride
source.
addition of 10 mol % of triphenylphosphine to the reaction
mixture results in almost complete suppression of the
reaction (entry 4). This dramatic effect, however, is not
observed in the reactions with the tosylamide derivative;
a 50% yield of 1 was obtained when 10 mol % of PPh3
was employed (entry 8).
The use of pyridine as a base was crucial for the success
of our coumarin synthesis.6 Therefore, several bases have
also been examined to determine if this is true for this
process as well (entries 9-11). None of the desired
quinolone 1 was detected in the reaction of the tosylamide
derivative when the inorganic base NaOAc was used
(entry 9). The use of triethylamine, in the reactions with
either the N-tosyl or N-ethoxycarbonyl derivative, re-
sulted in low yields of 1 (entries 10 and 11). Moreover,
the use of triethylamine leads to the formation of
significant amounts of 2,3-dipropylindole derivatives.
This is likely because NEt3 is a much stronger base than
pyridine, and it apparently deprotonates the amino group
much more easily, especially when a strong electron-
withdrawing substituent like a tosyl group is employed.
Thus, in the vinylpalladium complex formed after inser-
tion of the alkyne, the palladium atom is strongly
coordinated by the nitrogen atom, and reductive elimina-
tion leading to formation of the indole occurs faster than
the insertion of CO. Note that no deprotection of the
indole nitrogen occurs during the reaction. A small
amount of the protected quinolone 4 (about 5%) has also
been detected in the reaction of ethyl N-(2-iodophenyl)-
carbamate with Et3N as the base.
Elimination of the chloride source from the reaction
mixture slightly reduces the yield of the quinolone 1
(66%, entry 12), while the use of a different chloride salt,
LiCl (entry 13), gave virtually the same result as the use
of n-Bu4NCl (72% and 71%, respectively).
Next, the reaction temperature and time were varied
to find the optimal conditions. The results are sum-
marized in Table 3. N-(2-Iodophenyl)tosylamide is con-
siderably less reactive than ethyl N-(2-iodophenyl)car-
bamate, and the reactions of the sulfonamide derivatives
The limitations of the process are similar to those of
the coumarin synthesis. The annulation of unsymmetri-
cal alkynes, such as 1-phenyl-1-butyne, produces mix-
tures of regioisomers, which arise from the two possible
J . Org. Chem, Vol. 69, No. 20, 2004 6775

Kadnikov and Larock
TABLE 4. Syn th esis of 3,4-Disu bstitu ted 2-Qu in olon es via P a lla d iu m -Ca ta lyzed An n u la tion of In ter n a l Alk yn esa
6776 J . Org. Chem., Vol. 69, No. 20, 2004

Pd-Catalyzed Synthesis of 2-Quinolones
Ta ble 4. (Con tin u ed )
a
Standard reaction conditions: N-substituted o-iodoaniline (0.5 mmol), alkyne (1.5 mmol), pyridine (1.0 mmol), n-Bu4NCl (0.5 mmol),
Pd(OAc)2 (5 mol %, 0.025 mmol) under 1 atm of CO in DMF (5 mL) at 100 °C for 12 h, then the crude product was treated with 1 M
b
ethanolic NaOH (5 mL) at rt for 30 min. The reaction was run at 120 °C. c The treatment with 1 M ethanolic NaOH was omitted.
d
4-Nitroaniline was also isolated in a 64% yield. e The reaction was run for 36 h.
J . Org. Chem, Vol. 69, No. 20, 2004 6777

Kadnikov and Larock
SCHEME 1
yields from phenylalkyl acetylenes and with slightly
better regioselectivity.
We were surprised to discover that diphenylacetylene
reacts very poorly under the standard reaction conditions
and affords the desired 2-quinolone in only a 22% yield
(entry 13). Even though diphenylacetylene is less reactive
than 4-octyne, a 51% yield of the corresponding coumarin
was obtained in the carbonylative annulation of diphe-
nylacetylene by o-iodophenol.6 The reactions of dipheny-
lacetylene with other o-iodoaniline derivatives, such as
the tosylamide (entry 14) and the trifluoroacetamide
(entry 15), afforded 15 in even lower yields, even though
5 equiv of the alkyne were employed. The reason for such
remarkable differences in reactivity between the aniline
derivatives and the phenols is not clear.
The reactivity of internal alkynes bearing various
functional groups was examined next. To our delight, the
carbonylative annulation of 3-phenyl-2-propyn-1-ol (entry
16) afforded the desired 2-quinolones in a 42% yield. Even
though the yield is only modest, the carbonylative an-
nulation by o-iodophenol of this and other alkynes with
a hydroxyl group was completely ineffective, and no
coumarins were observed. Still, protection of the hydroxyl
group, for example, as a methyl ether (entry 17), im-
proves the yield of the reaction substantially. Mixtures
of regioisomers were obtained in both cases with poor
regioselectivity. No products were obtained in the reac-
tion of 2-butyne-1,4-diol (entry 18), while the correspond-
ing dibenzylated alkyne afforded the desired 2-quinolone
20 in a 51% yield (entry 19).
Only a 26% yield of 2-quinolone 21 was obtained when
a silyl-substituted alkyne, 1-trimethylsilyl-1-propyne,
was employed (entry 20). The use of a 5-fold excess of
the alkyne and a higher reaction temperature (120 °C)
did not improve the yield of the desired product (entry
21). We suspected that treatment of the crude product
with ethanolic NaOH might be causing decomposition of
the desired product. However, omission of this step
resulted only in the incomplete hydrolysis of 21, but no
increase in the overall yield. The products, 21 and the
still protected 2-quinolone 22, were obtained in a 28%
yield as a 39:61 mixture (entry 22).
Electron-deficient alkynes appear to be very poor
substrates for the carbonylative annulation by ethyl N-(2-
iodophenyl)carbamate. Only a 19% yield of 2-quinolone
23 was obtained in the reaction of 3-hexyn-2-one (entry
23), while the corresponding coumarin was obtained in
a 61% yield.6 Interestingly, only one regioisomer was
isolated, and its structure is assumed to be analogous to
the structure of the major coumarin isomer. This result
is consistent with the excellent regioselectivity (90:10)
observed in the carbonylative annulation of this alkyne
by o-iodophenol. Since only a 19% yield of 23 was
obtained, it is likely that the minor isomer was lost
during the workup, even if it was formed in the reaction.
We have also examined the reactivity of various
electron-rich and electron-poor alkyl N-(2-iodophenyl)-
carbamates as annulating agents in the reaction with
4-octyne. Introduction of an electron-donating substitu-
ent, for example a methoxy group, para to the amino
group (entry 24) does not affect the yield of the carbo-
nylative annulation. The corresponding 2-quinolone 25
was obtained in a 67% yield, comparable to the yield of
the parent system (entry 2). However, only a 36% yield
modes of alkyne insertion into the arylpalladium bond.
However, the palladium-catalyzed annulation of the same
alkynes with o-iodoaniline and its derivatives leading to
2,3-disubstituted indoles proceeds in most cases with
excellent regioselectivity, giving only one regioisomer.24
Since poor regioselectivity has also been observed in our
synthesis of coumarins, it is reasonable to assume that
the reaction conditions employed in the carbonylative
annulation lead to poor discrimination between the two
modes of alkyne insertion. All of our previous results on
the annulation of internal alkynes suggest that the
regioselectivity of the insertion is controlled almost
exclusively by steric factors. The insertion of an alkyne
into the arylpalladium bond results in the formation of
two vinylpalladium complexes A and B (Scheme 1). Since
the palladium-carbon bond is much longer than the
carbon-carbon bond, the steric repulsion between the
palladium atom and the larger substituent on the triple
bond (RL) in the complex A should be smaller than the
repulsion between the same substitutent and the aryl
group in the complex B. The major regioisomer in all
annulation reactions arises from the complex A.
While our previous annulation reactions have been run
either in the presence of triphenylphosphine or in the
absence of any ligands, carbonylative annulation requires
the use of pyridine, which most likely functions not only
as a base, but also as a ligand. Since a Pd-N bond is
much shorter than a Pd-P bond, coordination of pyridine
to the palladium atom significantly increases the steric
bulk of the palladium. Therefore, the difference in energy
between the complexes A and B is smaller when pyridine
is used, and consequently, the regioselectivity is worse.
The pyridine ligation also probably manifests itself in
the fact that, like the synthesis of coumarins, this process
is very sensitive to steric hindrance. Internal alkynes
with very bulky substituents react sluggishly and afford
the desired 2-quinolones in very low yields (entries 10
and 11). Only one regioisomer, however, is obtained in
each of these reactions, further supporting the assump-
tion that the regioselectivity is sterically controlled.
Internal alkynes bearing heterocyclic substituents,
such as 1-(5-pyrimidinyl)-1-hexyne (entry 12), are also
effective in the carbonylative annulation reaction giving
the desired products in a 66% yield, comparable to the
(24) (a) Larock, R. C.; Yum, E. K. J . Am. Chem. Soc. 1991, 113, 6689.
(b) Larock, R. C.; Yum, E. K.; Refvik, M. D. J . Org. Chem. 1998, 63,
7652.
6778 J . Org. Chem., Vol. 69, No. 20, 2004

Pd-Catalyzed Synthesis of 2-Quinolones
SCHEME 2
of the desired product was obtained from the o-iodoa-
niline derivative with a methoxy group para to the iodine
(entry 25). Still, this result is an improvement over the
reaction of the analogous 2-iodo-5-methoxyphenol, which
was completely unreactive in the carbonylative annula-
tion of 4-octyne, undergoing instead rapid deiodination
to produce 3-methoxyphenol in an 82% yield.6 It appears
that the presence of two strong electron-donating groups
in positions ortho and para to the iodine atom reduces
the stability of the aryl iodide, with deiodination occur-
ring much faster than oxidative addition to the palladium
complex. The presence of the carbamate protecting group
reduces the electron-donating ability of the amino sub-
stituent, thus increasing the stability of the aryl iodide
and allowing it to participate in carbonylative annulation
process.
Mech a n ism . The mechanism of this process is un-
doubtedly analogous to the mechanism proposed for the
synthesis of coumarins.6 It is shown in Scheme 2 and
involves (1) reduction of Pd(OAc)2 to the actual Pd(0)
catalyst; (2) oxidative addition of the N-substituted
o-iodoaniline to Pd(0), generating an arylpalladium com-
plex; (3) insertion of the alkyne triple bond into the
arylpalladium bond to give a vinylpalladium complex; (4)
insertion of CO to generate an acylpalladium complex;
and (5) nucleophilic attack of the nitrogen on the carbonyl
group of the acylpalladium complex leading to formation
of the 2-quinolone and regeneration of Pd(0). Removal
of the protecting group from the nitrogen occurs after
formation of the amide bond, at least in the case of the
carbamate derivatives. It is not clear whether deproto-
nation of the amino group is necessary for the nucleo-
philic attack to occur. It is likely that deprotonation
occurs in the case of the tosyl- and trifluoroacetamide
derivatives, but the situation with the carbamate deriva-
tives is not obvious. On one hand, the carbamate is
significantly less acidic than the tosylamide or trifluo-
roacetamide. On the other hand, the reaction with the
carbamate 35, lacking the amide hydrogen, (eq 6) did not
produce any of the annulation product.
A different picture emerged from the reactions of
electron-poor iodoanilines. Unlike the reactions of the
parent 2-iodoaniline, the reaction of unprotected 2-iodo-
4-nitroaniline afforded the desired 2-quinolone 29, albeit
only in 17% yield (entry 26). This reaction was very clean,
and the only other product of the reaction was the
deiodinated 4-nitroaniline, isolated in a 64% yield. The
formation of 2-quinolone 29 in this reaction further
supports the idea that the poor reactivity of the parent
2-iodoaniline relates to the high nucleophilicity of the
amino group. Even though unprotected 2-iodo-4-nitroa-
niline afforded some of the desired 2-quinolone, protection
of the amino group as an ethyl carbamate is still
necessary to obtain a satisfactory yield of 29 (49%, entry
27). The corresponding 4-methoxycarbonyl derivative 31
also afforded the 2-quinolone 32 in a similar 44% yield
(entry 28). Thus, the presence of electron-withdrawing
substituents on the aromatic ring of the 2-iodoanilines
lowers the yield of the carbonylative annulation. A
possible reason for this is the decreased stability of the
carbamates toward hydrolysis, since the presence of an
electron-withdrawing group should increase the positive
charge on the carbonyl carbon.
J ust as in the carbonylative annulation of internal
alkynes with o-iodophenols,6 in this process we have
never observed any 4-quinolones, which arise from the
reverse order of alkyne/CO insertion into the carbon-
palladium bond. Previously, we have established that
even though the insertion of CO into the arylpalladium
bond to form an acylpalladium complex occurs under our
Another electron-poor aniline, a carbamate derivative
of 4-amino-3-iodopyridine 33, also afforded a low 31%
yield of the corresponding 6-aza-2-quinolone 34 (entry
27). Moreover, this reaction was very slow and only
reached completion in 36 h.
J . Org. Chem, Vol. 69, No. 20, 2004 6779

Kadnikov and Larock
SCHEME 3
lecular trapping of the acylpalladium complex and the
irreversible insertion of the alkyne into the arylpalladium
bond. In most cases, insertion of the alkyne is a much
faster process, and only in the reactions with very
unreactive alkynes (Table 4, entries 10 and 11) does the
intramolecular attack of the carbamate oxygen predomi-
nate.
Con clu sion s
We have successfully extended the palladium-catalyzed
carbonylative annulation of internal alkynes to reactions
with o-iodoaniline derivatives, thus providing a new
synthesis of 3,4-disubstituted 2-quinolones. A crucial
aspect of the synthesis is the choice of the protecting
group on the nitrogen atom of the iodoaniline. The most
effective groups are alkoxycarbonyl, p-toluenesulfonyl,
and trifluoroacetyl. A wide variety of internal alkynes
bearing alkyl, aryl, heteroaryl, hydroxyl, and alkoxyl
substituents has been employed in this process affording
the desired 2-quinolones in 50 to 80% yields. Though the
use of unsymmetrical alkynes leads to the formation of
mixtures of regioisomers with low regioselectivity, these
products are easily separable by column chromatography.
Electron-rich iodoanilines with substituents para to the
amino group are also effective in the carbonylative
annulation. The carbonylative annulation of electron-poor
iodoaniline derivatives affords 2-quinolones in lower
yields than the parent system.
reaction conditions, the reaction of this acylpalladium
complex with an internal alkyne is apparently very
slow.6b Therefore, in the absence of any internal nucleo-
phile capable of trapping the acylpalladium complex, the
acylpalladium complex undergoes decarbonylation to the
original arylpalladium complex, which eventually reacts
with the internal alkyne, and is thus converted into the
coumarin or the 2-quinolone. Further evidence support-
ing this scheme has been obtained during our investiga-
tion of the carbonylative annulation of N-substituted
o-iodoanilines.
Exp er im en ta l Section
Gen er a l P r oced u r e for th e P a lla d iu m -Ca ta lyzed Syn -
th esis of 3,4-Disu bstitu ted 2-Qu in olon es. Ethyl N-(2-
iodophenyl)carbamate (0.5 mmol), the alkyne (1.5 mmol),
pyridine (1.0 mmol), n-Bu4NCl (0.5 mmol), and Pd(OAc)2 (5
mol %, 0.025 mmol) were placed in a 4 dram vial and then
dissolved in 5 mL of DMF. The vial was purged with CO for 2
min and then connected to a balloon of CO. The reaction
mixture was stirred at 100 °C for 12 h, allowed to cool to room
temperature, diluted with EtOAc, washed with water, and
concentrated under reduced pressure. The residue was treated
with 5 mL of 1 M ethanolic NaOH at room temperature for 30
min. Then satd aq NH4Cl (15 mL) was added, and the resulting
mixture was extracted with EtOAc. The organic extracts were
combined, washed with satd aq NH4Cl, and water, dried over
anhydrous MgSO4, and concentrated under reduced pressure.
The residue was separated by column chromatography on
silica gel.
3,4-Dip r op yl-2(1H)-qu in olin on e (1): white solid; mp 159-
160 °C; 1H NMR (CDCl3) δ 7.68 (d, J ) 8.0 Hz, 1H), 7.43 (ddd,
J ) 1.2, 6.8, 8.0 Hz, 1H), 7.38 (dd, J ) 1.2, 8.0 Hz, 1H), 7.20
(ddd, J ) 1.4, 6.8, 8.2 Hz, 1H), 2.86-2.90 (m, 2H), 2.74-2.78
(m, 2H), 1.59-1.71 (m, 4H), 1.05-1.12 (m, 6H); 13C NMR
(CDCl3) δ 164.3, 147.6, 137.5, 131.4, 129.2, 124.5, 122.3, 120.4,
116.5, 31.1, 29.3, 23.6, 22.9, 14.8, 14.7; IR (neat, cm-1) 2963,
2868, 1661; MS m/z (rel intensity) 229 (67, M+), 228 (52), 214
(100), 200 (47), 186 (41); HRMS calcd for C15H19NO 229.1467,
found 229.1470.
In most of the reactions employing ethyl N-(2-iodophe-
nyl)carbamate a minor byproduct can be isolated, along
with the desired 2-quinolone. The byproduct is either
isatoic anhydride (36) (in the reactions without ethanolic
NaOH treatment) or ethyl 2-aminobenzoate (37) (in the
reactions with a basic workup). In the reaction with
4-octyne the yield of the byproduct is around 10-12%.
However, in the reactions employing a very unreactive
alkyne, such as 4,4-dimethyl-2-pentyne or 1-phenyl-3,3-
dimethyl-1-butyne (Table 4, entries 10 and 11), ethyl
2-aminobenzoate was isolated in 30 and 39% yields,
respectively. In almost all reactions this byproduct was
detected in the crude product mixtures by 1H NMR
spectroscopy, but it was not isolated in a pure form and
the exact yield was not determined. The mechanism of
formation of these byproducts is shown in Scheme 3.
Insertion of CO into the arylpalladium bond, followed by
intramolecular attack by the oxygen atom of the carbam-
ate group on the carbonyl of the acylpalladium complex,
leads to the intermediate 38. Attack on the oxonium ion
38 by a water molecule and elimination of ethanol lead
to the formation of isatoic anhydride. Upon treatment
with ethanolic NaOH, the more electrophilic ester car-
bonyl group is attacked by an ethoxide anion generating
a carbamic acid anion, which spontaneously loses CO2
giving rise to ethyl 2-aminobenzoate. A similar intramo-
lecular attack of an amide oxygen on an acylpalladium
complex has been previously observed.25
Ack n ow led gm en t. We gratefully acknowledge the
donors of the Petroleum Research Fund, administered
by the American Chemical Society, for partial support
of this research, and Kawaken Fine Chemicals Co., Ltd.
and J ohnson Matthey, Inc. for donating palladium
acetate.
Su p p or tin g In for m a tion Ava ila ble: Procedures for the
preparation of starting materials; characterization data and
1H and 13C NMR spectra for all new compounds. This material
is available free of charge via the Internet at http://pubs.acs.org.
These data support the mechanistic scheme outlined
above. An acylpalladium complex is generated under our
reaction conditions, but it is not very reactive. The ratio
of the products from initial CO insertion and initial
alkyne insertion depends on the rates of the intramo-
(25) Cacchi, S.; Fabrizi, G.; Marinelli, F. Synlett 1996, 997.
6780 J . Org. Chem., Vol. 69, No. 20, 2004
J O049149+