Palladium-Catalyzed Carbonylative Annulation of Internal Alkynes: Synthesis of 3,4-Disubstituted Coumarins

Article abstract of DOI:10.1021/jo0350763

The palladium-catalyzed annulation of internal alkynes by o-iodophenols in the presence of CO results in exclusive formation of coumarins. No isomeric chromones have been observed. The best reaction conditions utilize the 2-iodophenol, 5 equiv of alkyne, 1 atm of CO, 5 mol% Pd(OAc)₂, 2 equiv of pyridine, and 1 equiv of n-Bu₄NCl in DMF at 120°C. The use of a sterically unhindered pyridine base is essential to achieve high yields. A wide variety of 3,4-disubstituted coumarins containing alkyl, aryl, silyl, alkoxy, acyl, and ester groups have been prepared in moderate to good yields. Mixtures of regioisomers have been obtained when unsymmetrical alkynes are employed. 2-Iodophenols with electron-withdrawing and electron-donating substituents and 3-iodo-2-pyridone are effective in this annulation process. The reaction is believed to proceed via (1) oxidative addition of the 2-iodophenol to Pd(0), (2) insertion of the alkyne triple bond into the aryl-palladium bond, (3) CO insertion into the resulting vinylic carbon-palladium bond, and (4) nucleophilic attack of the phenolic oxygen on the carbonyl carbon of the acylpalladium complex with simultaneous regeneration of the Pd(0) catalyst. This annulation process is the first example of intermolecular insertion of an alkyne occurring in preference to CO insertion.

Full text of DOI:10.1021/jo0350763

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: Syn th esis of 3,4-Disu bstitu ted Cou m a r in s
Dmitry V. Kadnikov and Richard C. Larock*
Department of Chemistry, Iowa State University, Ames, Iowa 50011
larock@iastate.edu
Received J uly 24, 2003
The palladium-catalyzed annulation of internal alkynes by o-iodophenols in the presence of CO
results in exclusive formation of coumarins. No isomeric chromones have been observed. The best
reaction conditions utilize the 2-iodophenol, 5 equiv of alkyne, 1 atm of CO, 5 mol % Pd(OAc)2, 2
equiv of pyridine, and 1 equiv of n-Bu4NCl in DMF at 120 °C. The use of a sterically unhindered
pyridine base is essential to achieve high yields. A wide variety of 3,4-disubstituted coumarins
containing alkyl, aryl, silyl, alkoxy, acyl, and ester groups have been prepared in moderate to good
yields. Mixtures of regioisomers have been obtained when unsymmetrical alkynes are employed.
2-Iodophenols with electron-withdrawing and electron-donating substituents and 3-iodo-2-pyridone
are effective in this annulation process. The reaction is believed to proceed via (1) oxidative addition
of the 2-iodophenol to Pd(0), (2) insertion of the alkyne triple bond into the aryl-palladium bond,
(3) CO insertion into the resulting vinylic carbon-palladium bond, and (4) nucleophilic attack of
the phenolic oxygen on the carbonyl carbon of the acylpalladium complex with simultaneous
regeneration of the Pd(0) catalyst. This annulation process is the first example of intermolecular
insertion of an alkyne occurring in preference to CO insertion.
In tr od u ction
various nucleophiles are among the most useful methods
of synthesis of aryl carbonyl compounds known. However,
the consecutive insertion of CO and unsaturated hydro-
carbons (alkenes and alkynes) into the aryl-palladium
bond (Scheme 1) has received relatively little attention,
and factors controlling the sequence of insertion are still
not completely understood. On one hand, intramolecular
versions of such processes in which a carbon-carbon
double or triple bond is tethered to the aryl or vinylic
iodide have been extensively studied.11 The order of
insertion in these processes depends on a variety of
factors, such as the CO pressure, the nature of the
multiple bond, the size of the ring formed, the reaction
temperature, solvent, and other parameters. On the other
hand, only a few examples of intermolecular three-
component processes have been reported.12 The pal-
The development of methods for the formation of
several carbon-carbon and/or carbon-heteroatom bonds
in one reaction is one of the most important goals of the
synthetic chemist, because such processes allow the
assembly of complex molecular structures from relatively
simple precursors in a single step. One of the fastest
growing areas of research directed toward this goal is
the development of transition-metal-catalyzed reactions
involving the insertion of unsaturated molecules, such
as alkenes, alkynes, and carbon monoxide, into a carbon-
metal bond. For example, the palladium-catalyzed an-
nulation of dienes and internal alkynes by aromatic and
vinylic halides bearing a neighboring nucleophilic sub-
stituent has been developed in our laboratories in the
last 15 years1 as an efficient way to synthesize a wide
variety of carbo- and heterocyclic compounds, including
indoles,2 isoquinolines,3 benzofurans,4 benzopyrans,4 iso-
coumarins,4,5 R-pyrones,5,6 indenones,7 naphthalenes,8
and phenanthrenes.9
(5) Larock, R. C.; Doty, M. J .; Han, X. J . Org. Chem. 1999, 64, 8770.
(6) Larock, R. C.; Han, X.; Doty, M. J . Tetrahedron Lett. 1998, 39,
5713.
(7) (a) Larock, R. C.; Doty, M. J .; Cacchi, S. J . Org. Chem. 1993, 58,
4579. (b) Larock, R. C.; Tian, Q.; Pletnev, A. A. J . Am. Chem. Soc.
1999, 121, 3238.
(8) Larock, R. C.; Tian, Q. J . Org. Chem. 1998, 63, 2002.
(9) Larock, R. C.; Doty, M. J .; Tian, Q.; Zenner, J . M. J . Org. Chem.
1997, 62, 7536.
(10) (a) Tsuji, J . Palladium Reagents and Catalysts: Innovations
in Organic Synthesis; J ohn Wiley & Sons, Ltd.: Chichester, U.K., 1995.
(b) Colquhoun, H. M.; Thompson, D. J .; Twigg, M. V. Carbonylation:
Direct Synthesis of Carbonyl Compounds; Plenum Press: New York
and London, 1991.
(11) (a) Negishi, E.; Cope´ret, C.; Ma, S.; Mita, T.; Sugihara, T.; Tour,
J . M. J . Am. Chem. Soc. 1996, 118, 5904. (b) Negishi, E.; Ma, S.;
Amanfu, J .; Cope´ret, C.; Miller, J . A.; Tour, J . M. J . Am. Chem. Soc.
1996, 118, 5919. (c) Sugihara, T.; Cope´ret, C.; Owczarczyk, Z.; Harring,
L. S.; Negishi, E. J . Am. Chem. Soc. 1994, 116, 7923. (d) Grigg, R.;
Sridharan, V. J . Organomet. Chem. 1999, 576, 65.
The insertion of CO into the aryl-palladium bond
leading to the formation of an acylpalladium complex is
now a ubiquitous process in organic synthesis.10 Reac-
tions of the resulting acylpalladium complexes with
(1) (a) Larock, R. C. J . Organomet. Chem. 1999, 576, 111. (b) Larock,
R. C. Pure Appl. Chem. 1999, 71, 1435.
(2) (a) Larock, R. C.; Yum, E. K.; Refvik, M. D. J . Org. Chem. 1998,
63, 7652. (b) Larock, R. C.; Yum, E. K. J . Am. Chem. Soc. 1991, 113,
6689.
(3) (a) Roesch, K. R.; Larock, R. C. J . Org. Chem. 1998, 63, 5306.
(b) Roesch, K. R.; Larock, R. C. Org. Lett. 1999, 1, 553.
(4) Larock, R. C.; Yum, E. K.; Doty, M. J .; Sham, K. K. C. J . Org.
Chem. 1995, 60, 3270.
10.1021/jo0350763 CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/31/2003
J . Org. Chem. 2003, 68, 9423-9432
9423

Kadnikov and Larock
SCHEME 1
react. Many other syntheses of coumarins also utilize
strong acidic or basic conditions, consist of several steps,
and are often limited to the synthesis of monosubstituted
coumarins.
Only a few Pd(0)-catalyzed approaches to the synthesis
of coumarins have been reported,14b,21 and the scope of
these processes is very limited. A more general approach
involves the Pd(II)-catalyzed reaction of phenols with
2-alkynoates.22 Still, only polyoxygenated phenols have
been used in this process, and the reaction is limited to
the synthesis of 4-monosubstituted coumarins.
We have previously reported that the reaction of
2-iodophenol with internal alkynes in the presence of 1
atm of CO and a palladium catalyst results in the
exclusive formation of coumarins 1, without any traces
of the isomeric chromones 2.23 Herein we report the full
details of the synthesis of 3,4-disubstituted coumarins
by the palladium-catalyzed carbonylative annulation of
internal alkynes by 2-iodophenols.
ladium-catalyzed reactions of 2-iodophenol with nor-
bornene,13 norbornadiene,14 allenes,15 and terminal
alkynes16 in the presence of CO produce coumaranone
or chromone derivatives. Two examples of the palladium-
catalyzed reactions of internal alkynes with aryl iodides
and CO have been reported.17 In all of the intermolecular
reactions, insertion of CO always precedes insertion of
the allene or alkyne, even when 1 atm of CO is employed,
and the order of insertion of CO and norbornene can be
controlled by the choice of the reaction temperature and
phosphine ligand.
Our interest in palladium-catalyzed annulations of
internal alkynes prompted us to explore the reaction of
2-iodophenols with internal alkynes in the presence of
carbon monoxide (eq 1). This reaction appeared to be a
potentially very efficient route to coumarins 1 or
chromones 2.
Resu lts a n d Discu ssion
Op tim iza tion of th e Rea ction Con d ition s. Initially,
the reactions of 2-iodophenol with diphenylacetylene and
4-octyne under 1 atm of CO were run employing the
standard conditions developed in our laboratories for the
annulation of internal alkynes,2,4 or the conditions usu-
ally employed in palladium-catalyzed carbonylation
reactions.11a,b,15,24 However, none of these systems af-
forded the desired product, except the reaction of 2-io-
dophenol with 4-octyne using NaOAc as a base at 120
°C, which afforded a small amount of 3,4-dipropylcou-
marin (3) (eq 2). We, thus, chose the reaction of 2-io-
dophenol and 4-octyne under 1 atm of CO as a model
system and thoroughly investigated the effect of various
reaction parameters on the outcome of this reaction.
Naturally occurring coumarins possess interesting
biological activity, including anticancer and HIV-1-
specific reverse transcriptase inhibitor properties.18 How-
ever, the existing methods for the synthesis of coumarins
suffer major disadvantages. Two classical methods, the
Perkin19 and Pechman20 reactions, require the use of
stoichiometric amounts of strong acids and often high
temperatures. Few functional groups in general can
tolerate the harsh reaction conditions, and moreover,
phenols with electron-withdrawing substituents fail to
The nature of the base employed in the reaction proved
to be the decisive factor determining the success of the
reaction (Table 1). Only very low yields of the desired
coumarin were obtained in reactions employing inorganic
bases (entries 1-11). Moreover, with these bases, the
presence or absence or nature of the phosphine ligand
(entries 4-8) or variation in the amount of the alkyne
employed (entries 6, 9, and 10) did not have any ap-
preciable effect on the yield of 3. Organic amines usually
used in palladium-catalyzed carbonylation reactions,
such as Et3N or i-Pr2NEt (entries 12 and 13), were
completely ineffective in this reaction. Speculating that
these amines can undergo â-hydride elimination upon
(12) Satoh, T.; Itaya, T.; Okuro, K.; Miura, M.; Nomura, M. J . Org.
Chem. 1995, 60, 7267.
(13) (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.
(14) (a) An, Z.; Catellani, M.; Chiusoli, G. P. J . Organomet. Chem.
1989, 371, C51. (b) An, Z.; Catellani, M.; Chiusoli, G. P. Gazz. Chim.
Ital. 1990, 120, 383.
(15) Okuro, K.; Alper, H. J . Org. Chem. 1997, 62, 1566.
(16) (a) Torii, S.; Okumoto, H.; Xu, L. H.; Sadakane, M.; Shostak-
ovsky, M. V.; Ponomaryov, A. B.; Kalinin, V. N. Tetrahedron 1993, 49,
9773. (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.
(17) (a) Cope´ret, C.; Sugihara, T.; Wu, G.; Shimoyama, I.; Negishi,
E. J . Am. Chem. Soc. 1995, 117, 3422. (b) Okuro, K.; Furuune, M.;
Miura, M.; Nomura, M. J . Org. Chem. 1992, 57, 4754.
(18) (a) Naser-Hijazi, B.; Stolze, B.; Zanker, K. S. Second Proceedings
of the International Society of the Coumarin Investigators; Springer:
Berlin, 1994. (b) Murray, R. D. H.; Me´ndez, J .; Brown, S. A. The
Natural Coumarins: Occurrence, Chemistry, and Biochemistry; J .
Wiley: New York, 1982.
(21) (a) Catellani, M.; Chuisoli, G. P.; Fagnola, M. C.; Solari, G.
Tetrahedron Lett. 1994, 35, 5919. (b) Catellani, M.; Chiusoli, G. P.;
Fagnola, M. C.; Solari, G. Tetrahedron Lett. 1994, 35, 5923.
(22) (a) Trost, B. M.; Toste, F. D. J . Am. Chem. Soc. 1996, 118, 6305.
(b) J ia, C.; Piao, D.; Oyamada, J .; Lu, W.; Kitamura, T.; Fujiwara, Y.
Science 2000, 287, 1992. (c) J ia, C.; Piao, D.; Kitamura, T.; Fujiwara,
Y. J . Org. Chem. 2000, 65, 7516.
(23) Kadnikov, D. V.; Larock, R. C. Org. Lett. 2000, 2, 3643.
(24) Ma¨gerlein, W.; Beller, M.; Indolese, A. F. J . Mol. Catal. A 2000,
156, 213.
(19) J ohnson, J . R. Org. React. 1942, 1, 210.
(20) Sethna, S.; Phadke, R. Org. React. 1953, 7, 1.
9424 J . Org. Chem., Vol. 68, No. 24, 2003

Pd-Catalyzed Carbonylative Annulation of Alkynes
TABLE 1. Op tim iza tion of th e Rea ction Con d ition s on
th e Mod el System (Eq 2)a
isolated
phosphine
amt of alkyne yield of
entry
base
TlOAc
NaHCO3
Na2CO3
K2CO3
(amount, mol %)
(equiv)
3 (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
PPh3 (10)
PPh3 (10)
PPh3 (10)
PPh3 (10)
2
2
2
2
2
2
2
2
1
5
2
2
2
2
2
2
3
5
2
5
2
9
12
17
21
19
23
23
21
26
25
30
12
13
15
37
43
50
63
37
57
20
run in other amide solvents, such as DMA or NMP,
afforded slightly lower yields of the desired product, while
DMSO proved to be completely ineffective, yielding the
desired product in only 8% yield. Good yields of the
desired product have been obtained only at 120 °C; only
a 39% yield of 3 has been achieved after 48 h at 100 °C.
The use of a higher pressure of CO proved to be
detrimental to the process; no identifiable products could
be isolated from the reaction mixture after 24 h at 120
°C under 5 atm of CO. The final optimized reaction
conditions for the reaction of 2-iodophenol, 4-octyne, and
carbon monoxide have thus been determined to be 0.5
mmol of 2-iodophenol, 5 equiv of 4-octyne, 5 mol % Pd-
(OAc)2, 2 equiv of pyridine, and 1 equiv of n-Bu4NCl in 5
mL of DMF at 120 °C for 24 h.
Scop e a n d Lim ita tion s. The scope and limitations
of this annulation reaction have been studied by allowing
a wide variety of internal alkynes to react with various
2-iodophenols under our optimized reaction conditions.
The results of this study are summarized in Table 2.
First, the carbonylative annulation of alkyl- and aryl-
substituted acetylenes was investigated. Both dialkyl-
alkynes (entries 1 and 2) and diarylalkynes (entries 3
and 4) afford coumarins in 55-63% yields, although a
higher temperature was required with diphenylacetylene
to drive the reaction to completion. A series of phenyl-
alkylacetylenes, where the alkyl group is methyl, ethyl,
or isopropyl, were also successfully employed in reactions
with 2-iodophenol (entries 5-7). Mixtures of regioisomers
were obtained in all cases with modest regioselectivity.
The regioisomers were identified by comparison of the
1H NMR chemical shifts of the hydrogens at C-5 in both
isomers. The signal of this hydrogen is shifted upfield in
3-alkyl-4-phenylcoumarins relative to 4-alkyl-3-phenyl-
coumarins, because of the deshielding effect of the
benzene ring.25 Thus, it was determined that in all cases
the major isomer is the 3-phenyl-4-alkylcoumarin. This
pattern of alkyne insertion, in which the palladium atom
adds to the more sterically hindered end of the triple
bond, is consistent with results obtained in all of our
previous annulation reactions.1,2,3a,5
K2CO3
NaOAc
NaOAc
NaOAc
NaOAc
NaOAc
Cs2CO3
Et3N
i-Pr2NEt
TMPPb
pyridine
pyridine
pyridine
pyridine
PPh3 (10)
dppp (5)
PPh3 (10)
PPh3 (10)
PPh3 (10)
PPh3 (10)
PPh3 (10)
PPh3 (10)
PPh3 (10)
2,4,6-collidine PPh3 (10)
2,4,6-collidine
DTBMPc
PPh3 (10)
a
The following procedure was employed: 2-iodophenol (0.5
mmol), 4-octyne, the base (1.0 mmol), n-Bu4NCl (0.5 mmol), and
Pd(OAc)2 (5 mol %, 0.025 mmol) in DMF (10 mL) under 1 atm of
b
CO were stirred at 120 °C for 24 h. TMPP ) 2,2,6,6-tetrameth-
ylpiperidine. c DTBMP ) 2,6-di-tert-butyl-4-methylpyridine.
coordination to the palladium atom, we tested a couple
of amines lacking such hydrogens (entries 14 and 15).
Only pyridine afforded a significantly better yield of the
desired product. A further increase in the yield was
achieved by removal of the phosphine ligand (entry 16)
and an increase in the amount of 4-octyne (entries 17
and 18). A similar effect was observed when another
relatively sterically unhindered pyridine, 2,4,6-collidine,
was used as the base (entries 19 and 20). Use of the more
sterically hindered base 2,6-di-tert-butyl-4-methylpyri-
dine (entry 21) resulted in a significant decrease in the
yield of 3.
These results suggest that pyridine acts not only as a
base, but also as a ligand, and perhaps its coordination
to the palladium atom generates a more stable complex.
Further support for this hypothesis has been obtained
in experiments with bidentate pyridine ligands. Thus,
when 2,2-bipyridyl was used as a base, only a 7% yield
of 3 was obtained, and no coumarin 3 could be detected
in the reaction employing 1,10-phenanthroline. However,
the addition of AgClO4 to the latter reaction results in
the formation of coumarin 3 in 33% yield. These results
can be rationalized by invoking the formation of complex
4 after oxidative addition of 2-iodophenol to the pal-
ladium(0) complex. Complex 4 lacks labile ligands that
can easily dissociate to open a coordination site necessary
for the reaction to proceed. The silver cation presumably
removes the iodine atom from this complex, thus opening
a coordination site and allowing the reaction to proceed
(eq 3).
Both the regioselectivity and the yields of the annu-
lations of phenylalkylacetylenes are affected by the steric
bulk of the alkyl substituent. An increase in the size of
the alkyl group generally results in a decrease in the
regioselectivity and the yield of the reaction (entries 5-8).
However, the dependence of the yield on the size of the
alkyl substituent is not linear. The yield of the coumarin
decreases in the following order: Me = Et > i-Pr . t-Bu.
The effect of the tert-butyl group is enormous. No product
was obtained when 3,3-dimethyl-1-phenyl-1-butyne was
employed as the alkyne (entry 8). The reaction with
another alkyne bearing a tert-butyl group, 4,4-dimethyl-
Using 0.5 mmol of 2-iodophenol, 5 equiv of 4-octyne,
and 2 equiv of pyridine as the base, other reaction
parameters (catalyst precursor, solvent, and reaction
temperature and time) have also been examined. Both
Pd(OAc)2 and Pd(dba)2 are equally effective, while Pd-
(PPh3)4 afforded a slightly lower yield of 3. The reactions
(25) Patil, V. O.; Kelkar, S. L.; Waida, M. S. Indian J . Chem., B
1987, 26, 674
J . Org. Chem, Vol. 68, No. 24, 2003 9425

Kadnikov and Larock
TABLE 2. Syn th esis of Cou m a r in s by th e 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 esa
9426 J . Org. Chem., Vol. 68, No. 24, 2003

Pd-Catalyzed Carbonylative Annulation of Alkynes
Ta ble 2. (Con tin u ed )
J . Org. Chem, Vol. 68, No. 24, 2003 9427

Kadnikov and Larock
Ta ble 2. (Con tin u ed )
a
The following representative procedure for the carbonylative annulation of internal alkynes has been employed unless indicated
otherwise: the 2-iodophenol (0.5 mmol), the alkyne (2.5 mmol), pyridine (1.0 mmol), n-Bu4NCl (0.5 mmol), Pd(OAc)2 (5 mol %, 0.025
mmol), and DMF (5 mL) were placed in a 4 dram vial. The vial was purged with CO for 2 min and then connected to a balloon of CO, and
b
d
the reaction mixture was stirred at 120 °C for 24 h. The reaction was run at 135 °C. c The reaction was complete in 10 h. The reaction
was complete in 12 h. e The reaction was complete in 6 h. f 3-Methoxyphenol was obtained in 80% yield. 2-Naphthol was obtained in
g
82% yield, and 12% of 1-iodo-2-naphthol was recovered.
2-pentyne (entry 9), resulted in only a 9% yield of the
desired product. This unusual sensitivity of the annula-
tion to steric hindrance is even more surprising consider-
ing that these alkynes have been among the best in our
previous annulation reactions.1,2,7a,8 The reaction tem-
perature does not have any major effect on the regiose-
lectivity. The carbonylative annulation of 1-phenyl-1-
butyne at 120 °C affords coumarins 10 and 11 as a 2.6:1
mixture in 78% yield, while the analogous reaction run
at 100 °C affords a 2.4:1 mixture in 47% yield. The
carbonylative annulation of 1-cyclohexyl-1-propyne also
produced a mixture of regioisomers (entry 10).
Alkynes bearing a variety of functional groups, such
as alkoxy, silyl, acyl, and ester groups, have also been
effectively annulated under our standard conditions
(entries 12-19). Surprisingly, the propargylic alcohol
9428 J . Org. Chem., Vol. 68, No. 24, 2003

Pd-Catalyzed Carbonylative Annulation of Alkynes
3-phenyl-2-propyn-1-ol failed to provide any coumarin in
this process (entry 11). A significant amount of the
starting acetylene was recovered from the reaction, along
with a small amount (∼0.2 mmol) of the corresponding
formate. Protection of the hydroxyl group circumvents
this problem. Thus, when 3-methoxy-1-phenyl-1-propyne
was employed, the desired product was obtained in 65%
yield as a 3:1 mixture of the regioisomers 17 and 18
(entry 12). More easily removed protecting groups, such
as benzyl, can also be employed successfully in this
process (entries 13 and 14).
Only one regioisomer has been obtained in the annu-
lation of 1-trimethylsilyl-1-propyne, albeit in only 43%
yield (entry 15). The improved regioselectivity is due to
the steric bulk of the trimethylsilyl group. Unfortunately,
the steric bulk also leads to a decrease in the overall yield
of the coumarin, similar to the effect of a tert-butyl group
(entries 8 and 9). However, it is noteworthy that the yield
from the reaction of this silylacetylene with a longer C-Si
bond is significantly higher than the yield from the
reaction of 4,4-dimethyl-2-pentyne (entry 9).
the annulation products is similar to that of the reaction
of ethyl phenylpropiolate. In this case, however, the
decarboxylated coumarin 30 is the major product (24%
yield). Coumarin 31 has been identified by comparison
1
of its H NMR spectral data with literature data.28 Thus,
the selectivity of this reaction is only (30 + 31):32 ) (24
+ 5):17 ) 1.7:1. This result is similar to the result
obtained with 1-benzyloxy-2-butyne (entry 14), and the
regiochemistry is, therefore, most probably due only to
the difference in size between the methyl and ethoxy-
carbonyl groups. Hence, unlike the reaction with 3-hexyn-
2-one, no major electronic effects have been detected in
the carbonylative annulation of alkynoates.
Various functionalized iodophenols have also been
successfully employed in the carbonylative annulation of
internal alkynes. Thus, the carbonylative annulation of
2-iodophenols bearing electron-withdrawing groups para
to the hydroxyl group have proven to be nearly as
successful as 2-iodophenol itself (entries 20-22). An
electron-withdrawing group can also be introduced at C-7
of the coumarin (entry 23). Again, the yield is comparable
to the yield obtained with the parent 2-iodophenol. In
addition, the reaction of iodophenol 38 is much faster,
reaching completion within 12 h. This result is consistent
with the higher reactivity of an electron-poor aryl iodide
in oxidative addition to Pd(0). These results are all the
more important, because coumarins possessing electron-
withdrawing groups cannot be prepared using classical
methods, such as the Pechman reaction.20
Electron-deficient alkynes (alkynones and alkynoates)
behave in these carbonylative annulation reactions quite
differently from other alkynes. 3-Hexyn-2-one (entry 16)
affords the desired coumarin in good yield with good
regioselectivity (9:1). The major product is coumarin 24,
1
identified by comparison of its H NMR spectrum with
literature data.26 The high regioselectivity of the annu-
lation is clearly governed not only by steric factors (ethyl
and acetyl groups are not very different in size), but also
by electronic factors. Electronic factors apparently favor
insertion of the alkyne into the aryl-palladium bond so
that the palladium moiety ends up on the carbon next to
the carbonyl group.
On the other hand, an electron-donating group, such
as a methoxy group, cannot be as easily introduced into
the coumarin. When 2-iodo-4-methoxyphenol was allowed
to react with 4-octyne, the desired coumarin was pro-
duced in 62% yield (entry 24). However, the annulation
of 4-octyne with 2-iodo-5-methoxyphenol (entry 25) af-
forded none of the desired coumarin. The only product
isolated from the reaction was 3-methoxyphenol, obtained
in 80% yield. The same unusual behavior was observed
in the reaction of 1-iodo-2-naphthol (entry 26), in which
the only product was 2-naphthol, isolated in 82% yield,
alongside 12% of the recovered starting material.
While the annulation of 3-hexyn-2-one was successful,
none of the desired coumarin was observed in the
annulation of 1-phenyl-2-butyn-1-one (entry 17). The
major product isolated from the reaction mixture was
identified as 2-benzoylmethyl-2-methyl-1-benzofuran-
3(2H)-one (26), obtained in 75% yield. The structure of
1
the product was confirmed by comparison of its H and
13C NMR spectra with those of 2-benzoylmethyl-2-n-
butyl-1-benzofuran-3(2H)-one.27 The mechanism for the
formation of this interesting product will be discussed
later.
The reaction of 4-octyne with 2,5-diiodo-1,4-hydro-
quinone (44) (entry 27) shows that double annulation can
be successfully accomplished using the standard reaction
conditions, even without increasing the alkyne-to-io-
dophenol ratio beyond the usual 5:1 ratio. The yield of
the doubly annulated product 45 (54%) is only slightly
lower than the yield of coumarin 3 in the reaction with
the parent 2-iodophenol (entry 1, 63%). This result and
the absence of the product of monoannulation seem to
indicate that the first pyrone ring activates the interme-
diate toward formation of the second pyrone ring.
Unexpected products were also observed in the reac-
tions of 2-alkynoates. The carbonylative annulation of
ethyl phenylpropiolate resulted in the formation of the
desired coumarins 27 and 28 in a 1:3 ratio and 45% yield
(entry 18). Here, the regioselectivity of the reaction is
most likely determined by the size of the substituents
on the carbon-carbon triple bond. Besides these two
coumarins, a small amount of 4-phenylcoumarin (29) has
Heterocyclic analogues of 2-iodophenol were also ex-
amined in the carbonylative annulation. 3-Hydroxy-2-
iodopyridine appeared to be unreactive in this process
(entry 28). Only a small amount of a bipyridine coupling
product was observed in the reaction, although the
starting material almost completely disappeared. On the
other hand, the substituted 3-iodo-2-pyridone 47 affords
the azacoumarin 48 in 70% yield, the highest yield
obtained thus far in the annulation of 4-octyne (entry 29).
1
also been detected by GC-MS and H NMR spectroscopy,
although we have been unable to isolate it in pure form
and fully characterize it. This coumarin is likely the
result of the in situ decarboalkoxylation of coumarin 27.
The annulation of ethyl 2-butynoate (entry 19) also
affords three coumarins, 30-32, and the total yield of
(26) Clinging, R.; Dean, F. M.; Houghton, L. E. J . Chem. Soc., Perkin
Trans. 1 1974, 66.
(27) Kokubo, K.; Matsumasa, K.; Nishinaka, Y.; Miura, M.; Nomura,
M. Bull. Chem. Soc. J pn. 1999, 72, 303.
(28) Awasthi, A. K.; Tewari, R. S. Synthesis 1986, 1061.
J . Org. Chem, Vol. 68, No. 24, 2003 9429

Kadnikov and Larock
SCHEME 2
SCHEME 3
Surprisingly, 2-iodoanisole is also reactive in the
carbonylative annulation, producing coumarin 3 in 27%
yield (entry 30). Although this is only half of the yield
obtained with 2-iodophenol, it is interesting to see that
a free hydroxyl group is not necessary for the reaction to
proceed.
Mech a n ism . A possible mechanism for the carbony-
lative annulation is shown in Scheme 2. First, Pd(OAc)2
is reduced to a Pd(0) complex, which is the actual catalyst
in the process. We believe that CO acts as the reducing
agent in this reaction, since it is known to reduce Pd(II)
to Pd(0) in the presence of alcohols and phenols.29 The
oxidative addition of 2-iodophenol to Pd(0) to form an
arylpalladium complex starts the catalytic cycle. Dis-
sociation of one of the neutral ligands from the palladium
opens a coordination site, which is then occupied by the
internal alkyne. Insertion of the alkyne into the aryl-
palladium bond provides a vinylic palladium species.
Insertion of CO into the vinyl-palladium bond gives rise
to an acylpalladium complex. Finally, intramolecular
nucleophilic attack on the carbonyl group of the latter
complex by the phenolic oxygen produces the desired
coumarin and regenerates Pd(0).
This process represents the first example of insertion
of an internal alkyne into an aryl-palladium bond in
preference to CO insertion. The reverse is true in all
examples previously reported in the literature.11a,17 In
fact, Negishi determined the relative rates of various
processes to be as follows: CO insertion = intramolecular
5-exo- or 6-exo-alkyne carbopalladation > intramolecular
5-exo-alkyne acylpalladation > trapping of an acylpalla-
dium with MeOH > intermolecular carbopalladation or
acylpalladation.11c Thus, the preferential insertion of an
alkyne and, moreover, the complete absence of products
arising from the initial insertion of CO are quite surpris-
ing. Therefore, we attempted to establish the origins of
such unusual behavior.
assumes that insertions of both CO and the alkyne
readily occur under our reaction conditions. The alkyne
inserts irreversibly to afford the vinylpalladium complex
51, which then reacts with CO, eventually forming the
coumarin. In contrast, CO inserts in a reversible fashion.
Therefore, if the acylpalladium complex 50 reacts with
the alkyne very slowly, or not at all, then, in the absence
of any other reaction pathways, the palladium complex
50 could undergo decarbonylation, regenerating the
original arylpalladium complex 49. Eventually, most of
the arylpalladium complex 49 reacts with the internal
alkyne, affording the desired coumarin.
In the case shown in Scheme 3, insertion of the alkyne
may be either faster or slower than the insertion of CO.
On the other hand, it is possible that no insertion of CO
into the aryl-palladium bond takes place under our
reaction conditions. Indeed, our reaction conditions [Pd-
(OAc)2, pyridine, n-Bu4NCl] are quite different from the
conditions usually employed in palladium-catalyzed car-
bonylations [PdCl2(PPh3)2, Et3N, no chloride source].11a,b,15,24
To establish whether the insertion of CO occurs under
our standard conditions, we attempted to trap the
acylpalladium complexes by reaction with a nucleophile.
Initially, alcohols were used as external nucleophiles.
Formation of three products is possible (eq 4). However,
in all cases only coumarin 3 was isolated in yields
comparable to the yield of the reaction without any
alcohol present. For example, when 1-pentanol was used,
3 was obtained in 66% yield.
The observed order of insertion can be rationalized in
two ways. The first explanation, shown in Scheme 3,
Since an attempt to use an external nucleophile failed,
we employed an aryl iodide with an internal nucleophile.
2-Iodobenzyl alcohol was the obvious choice. If the
(29) (a) Rivetti, F.; Romano, U. J . Organomet. Chem. 1979, 174, 221.
(b) Stromnova, T. A.; Vargaftik, M. N.; Moiseev, I. I. J . Organomet.
Chem. 1983, 252, 113.
9430 J . Org. Chem., Vol. 68, No. 24, 2003

Pd-Catalyzed Carbonylative Annulation of Alkynes
TABLE 3. Ca r bon yla tive An n u la tion of In ter n a l
Alk yn es w ith 2-Iod oben zylic Alcoh ols (Eq 5)
isolated yield (%)
temp
(°C)
time
(h)
entry
R
R′
53
54
1
2
3
4
5
H
H
H
H
n-Pr
n-Pr
n-Pr
Ph
120
100
80
120
120
4
8
24
4
24
45
67
46
78
43
28
<5
16
0
Me
n-Pr
2
insertion of CO into the aryl-palladium bond occurs, the
resulting acylpalladium complex should be easily trapped
by the adjacent hydroxyl group, leading to formation of
the five-membered ring lactone 53 (eq 5).30 Alternatively,
formation of the seven-membered ring lactone 54 might
occur upon insertion of an alkyne and CO.
the alkyne into the aryl-palladium bond of complex 49.
The factors affecting the relative reactivity of one of these
(or both) complexes change the ratio of the products of
the initial CO insertion and the initial alkyne insertion.
The more favorable conformation of the hydroxyl group
in 52b than in 52a apparently increases the rate of
trapping of the acylpalladium complex corresponding to
50. On the other hand, a decrease in the temperature
slows insertion of the alkyne into the complex corre-
sponding to 49. Diphenylacetylene is less reactive than
4-octyne; thus, the yield of 53a increases.
Thus, the apparent reason for the exclusive formation
of coumarins in our process is the slow insertion of an
internal alkyne into an acyl-palladium bond. Only in one
case, the reaction of 2-iodophenol with 1-phenyl-2-butyn-
1-one (Table 2, entry 17), have we observed a product
apparently arising from the insertion of an internal
alkyne into the acyl-palladium bond. Surprisingly, the
product of this reaction is not a chromone, but a 1-ben-
zofuran-3(2H)-one derivative 26. A probable mechanism
for its formation is shown in Scheme 4. Addition of the
acylpalladium complex to the carbon-carbon triple bond
generates a vinylpalladium intermediate that probably
exists in equilibrium with the corresponding palladium
enolate. The hydrolysis of this enolate by water present
in DMF leads to enone 55. Intramolecular Michael
addition of the phenol then affords the final product 26.
No coumarins were detected in this reaction, while no
products arising from the insertion of the alkyne into the
acylpalladium bond were observed in the reaction with
3-hexyn-2-one (Table 2, entry 16). The origins of this
complete reversal of the reactivity are not clear at this
point.
Under our standard carbonylative annulation condi-
tions, the reaction of 2-iodobenzyl alcohol (52a , R ) H)
with 4-octyne was complete within 4 h, and both possible
products 53a (R ) H) and 54a (R ) H, R′ ) n-Pr) were
isolated in 24% and 43% yields, respectively. The tem-
perature has a remarkable effect on the chemoselectivity
of this reaction (Table 3, entries 1-3). The yield of 54a
drops dramatically with a decrease in the reaction
temperature, while the yield of 53a increases. The nature
of the alkyne is also a very important factor in determin-
ing the selectivity of the reaction. Thus, when diphenyl-
acetylene was used (entry 4), the five-membered ring
lactone 53a was obtained in 46% yield, while the yield
of the seven-membered ring lactone 54b (R ) H, R′ )
Ph) was only 16%.
When the tertiary benzylic alcohol 52b (R ) Me) was
employed in the reaction, only the five-membered ring
lactone 53b (R ) Me) was obtained in 78% yield, and
none of the product of the alkyne insertion was detected
(entry 5). Apparently, the presence of two methyl groups
forces the alcohol into a conformation in which the
hydroxyl group is in close proximity to the iodine atom,
and consequently, trapping of the acylpalladium inter-
mediate is much more effective than in the case of
2-iodobenzyl alcohol.
We were interested in seeing whether the unusual
selectivity would be observed with other unsaturated
compounds, such as allenes. However, only the product
of initial CO insertion, 3-n-butylene-2-n-propyl-2,3-dihy-
dro-4H-1-benzopyran-4-one (56), was obtained in the
carbonylative annulation of 4,5-nonadiene under our
standard annulation conditions (eq 6). The same order
These results suggest that CO does insert into the
aryl-palladium bond under our reaction conditions.
Moreover, the outcome of the reaction with alcohol 52b
shows that even at 120 °C the insertion of CO is faster
than the insertion of an internal alkyne. Consequently,
the ratio of products arising from the initial CO insertion
to those of the initial alkyne insertion is determined by
the relative rates of two processes, the reaction of
acylpalladium complex 50 with either a nucleophile or
some other species (alkene or alkyne) and insertion of
of insertion has been observed by Alper under different
reaction conditions.15 This result shows that the carbon-
carbon double bond of the allene easily inserts into the
acyl-palladium bond under our reaction conditions, and
(30) Cowell, A.; Stille, J . K. J . Am. Chem. Soc. 1980, 102, 4193.
J . Org. Chem, Vol. 68, No. 24, 2003 9431

Kadnikov and Larock
SCHEME 4
(66), 187 (100). Anal. Calcd for C15H18O2: C, 78.23; H, 7.88.
Found: C, 78.05; H, 8.05.
therefore, the unusual order of insertion in our chemistry
is probably limited to internal alkynes.
Full characterization and spectral data for compounds 7-11,
14, 19, 20, 23-25, 37, 39, 41, and 48 can be found in ref 23.
Full characterization and spectral data for all other coumarins
prepared can be found in the Supporting Information.
Tr a p p in g of th e Acylp a lla d iu m Com p lexes w ith a n
Exter n a l Alcoh ol. 2-Iodophenol (110 mg, 0.5 mmol), 4-octyne
(275 mg, 2.5 mmol), pyridine (79 mg, 1.0 mmol), n-Bu4NCl (139
mg, 0.5 mmol), Pd(OAc)2 (5.6 mg, 5 mol %, 0.025 mmol), and
DMF (5 mL) were placed in a 4 dram vial. The vial was purged
with CO for 2 min and then connected to a balloon of CO. An
alcohol (10 mmol) was added to the reaction mixture in one
portion. The reaction mixture was stirred at 120 °C for 24 h,
then allowed to cool to room temperature, diluted with EtOAc,
washed with water, dried over anhydrous MgSO4, and con-
centrated under reduced pressure. The products were isolated
by flash chromatography on silica gel.
Ca r bon yla tive An n u la tion of In ter n a l Alk yn es w ith
2-Iod oben zylic Alcoh ols. The 2-iodobenzylic alcohol (0.5
mmol), the alkyne (2.5 mmol), pyridine (79 mg, 1.0 mmol),
n-Bu4NCl (139 mg, 0.5 mmol), Pd(OAc)2 (5.6 mg, 5 mol %, 0.025
mmol), and DMF (5 mL) were placed in a 4 dram vial. The
vial was purged with CO for 2 min and then connected to a
balloon of CO. Upon completion of the reaction (for the reaction
temperatures and times, see Table 3), the reaction mixture
was cooled to room temperature, diluted with EtOAc, washed
with water, dried over anhydrous MgSO4, and concentrated
under reduced pressure. The products were isolated by flash
chromatography on silica gel.
Con clu sion s
An efficient palladium-catalyzed synthesis of 3,4-
disubstituted coumarins has been developed. A wide
variety of alkynes containing alkyl, aryl, silyl, alkoxy,
acyl, and ester groups afford coumarins in moderate to
good yields. The process is sensitive to the steric bulk of
the alkynes, and alkynes bearing tertiary alkyl substit-
uents generally fail to undergo annulation. Unsym-
metrical alkynes produce mixtures of regioisomers with
generally only modest selectivity. The regioselectivity of
the process is governed by both steric and electronic
factors. Substituted 2-iodophenols with both electron-
donating and electron-withdrawing substituents, as well
as substituted pyridinones, are effective in this process,
thus creating a fast and efficient route to coumarins that
are not easily accessible by classical methods.
This carbonylative annulation process represents the
first example of a domino process in which the insertion
of an internal alkyne into an aryl-palladium bond occurs
prior to the insertion of CO. Trapping experiments
suggest that the reason for this unusual selectivity is not
the intrinsic inability of CO to undergo insertion under
our reaction conditions, but the absence of any species
able to react with the resulting acylpalladium complex
faster than it undergoes decarbonylation.
Da ta for 4,5-d ip r op yl-1,3-d ih yd r oben zo[c]oxep in -3-on e
(54a ): colorless oil; 1H NMR (CDCl3) δ 7.37-7.48 (m, 3H), 7.31
(ddd, J ) 1.2, 7.2, 7.6 Hz, 1H), 5.04 (d, J ) 7.6 Hz, 1H), 4.81
(d, J ) 7.6 Hz, 1H), 2.66-2.75 (m, 3H), 2.51-2.58 (m, 1H),
1.59-1.68 (m, 2H), 1.34-1.46 (m, 2H), 1.03 (t, J ) 7.2 Hz, 3H),
0.90 (t, J ) 7.2 Hz, 1H); 13C NMR (CDCl3) δ 171.0, 143.5, 139.5,
135.7, 133.2, 129.5, 128.5, 128.3, 127.3, 68.3, 35.1, 34.2, 23.3,
22.5, 14.5, 14.4; IR (CHCl3, cm-1) 2959, 2872, 1707; MS m/z
(rel intens) 244 (45, M+), 216 (28), 201 (71), 173 (100); HRMS
m/z calcd for C16H20O2 244.1463, found 244.1468.
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 Cou m a r in s. The 2-iodophenol (0.5 mmol), the
alkyne (2.5 mmol), pyridine (79 mg, 1.0 mmol), n-Bu4NCl (139
mg, 0.5 mmol), Pd(OAc)2 (5.6 mg, 5 mol %, 0.025 mmol), and
DMF (5 mL) were placed in a 4 dram vial. The vial was purged
with CO for 2 min, and then connected to a balloon of CO.
The reaction mixture was stirred at 120 °C for 24 h, then
allowed to cool to room temperature, diluted with EtOAc,
washed with water, dried over anhydrous MgSO4, and con-
centrated under reduced pressure. The product was isolated
by flash chromatography on silica gel.
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.
Data for 3,4-dipr opyl-2H-1-ben zopyr an -2-on e (3): white
solid; mp 58-60 °C; H NMR (CDCl3) δ 7.59 (dd, J ) 1.2, 8.0
1
Hz, 1H), 7.45 (ddd, J ) 1.2, 8.0, 8.4 Hz, 1H), 7.25-7.32 (m,
2H), 2.79 (m, 2H), 2.61 (m, 2H), 1.55-1.70 (m, 4H), 1.11 (t, J
) 7.6 Hz, 3H), 1.03 (t, J ) 7.6 Hz, 3H); 13C NMR (CDCl3) δ
162.1, 152.8, 150.1, 130.5, 126.6, 124.7, 124.2, 120.0, 117.2,
30.7, 29.9, 23.1, 22.5, 14.7, 14.5; IR (CHCl3, cm-1) 3073, 2962,
2872, 1716; MS m/z (rel intens) 230 (67, M+), 215 (77), 201
Su p p or tin g In for m a tion Ava ila ble: Procedures for the
preparation of starting materials and spectral data for all new
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
J O0350763
9432 J . Org. Chem., Vol. 68, No. 24, 2003