Atom economy. Palladium-catalyzed formation of coumarins by addition of phenols and alkynoates via a net C-H insertion

Article abstract of DOI:10.1021/ja0286573

A strategy to achieve ortho substitution of phenols initiated by an ortho-palladation to create coumarins was examined. Indeed, treatment of alkynoates with electron-rich phenols in the presence of a palladium catalyst and an acid does generate coumarins. The scope of the reaction with respect to the phenol and the alkynoates is defined. With unsymmetrical aromatic substrates, generally good regioselectivity that reflects the HOMO coefficients can be observed. In the course of these studies, numerous important naturally occurring coumarins have been synthesized, including fraxinol methyl ether, ayapin, herniarin, xanthoxyletin, and alloxanthoxyletin. The fact that a Pd(0) is the precatalyst rather than a Pd(+2) species and that an acid that reduces Pd(+2) salts, formic acid, functions better than other carboxylic acids raises doubts about the initial working hypothesis. A novel mechanism involving a palladium phenoxide formed from a hydridopalladium carboxylate and phenol is invoked to rationalize the results.

Full text of DOI:10.1021/ja0286573

Published on Web 03/20/2003
Atom Economy. Palladium-Catalyzed Formation of Coumarins
by Addition of Phenols and Alkynoates via a Net C-H
Insertion
Barry M. Trost,* F. Dean Toste, and Kevin Greenman
Contribution from the Department of Chemistry, Stanford UniVersity,
Stanford, California, 94305-5080
Received September 23, 2002; E-mail: Bmtrost@stanford.edu.
Abstract: A strategy to achieve ortho substitution of phenols initiated by an ortho-palladation to create
coumarins was examined. Indeed, treatment of alkynoates with electron-rich phenols in the presence of a
palladium catalyst and an acid does generate coumarins. The scope of the reaction with respect to the
phenol and the alkynoates is defined. With unsymmetrical aromatic substrates, generally good regioselectivity
that reflects the HOMO coefficients can be observed. In the course of these studies, numerous important
naturally occurring coumarins have been synthesized, including fraxinol methyl ether, ayapin, herniarin,
xanthoxyletin, and alloxanthoxyletin. The fact that a Pd(0) is the precatalyst rather than a Pd(+2) species
and that an acid that reduces Pd(+2) salts, formic acid, functions better than other carboxylic acids raises
doubts about the initial working hypothesis. A novel mechanism involving a palladium phenoxide formed
from a hydridopalladium carboxylate and phenol is invoked to rationalize the results.
Introduction
turned our attention to the use of palladium catalysts to
accomplish the transformation.
The conversion of aromatic C-H bonds into C-C bonds
classically has involved Friedel-Crafts reactions that most
normally require stoichiometric quantities of the Lewis acids,
although recent advances notably with rare earth triflates are
addressing this problem.1 The harshness of the conditions has
led to increasing use of a two-step protocolsformation of a
C-X bond by electrophilic aromatic substitution followed by
replacement of the C-X bond with a C-C bond by a transition-
metal-catalyzed coupling reaction. Formation of substituted
arenes by direct substitution of an aromatic C-H bond with a
C-C bond under mild conditions where nothing else is needed
catalytically streamlines this process. Our efforts directed toward
the synthesis of the coumarin natural products inspired us to
envision utilizing such a process to directly prepare coumarins
by the reaction of phenols with alkynoates. The general
biological importance of coumarins makes the development of
milder strategies for their direct synthesis significant. Since the
first isolation of coumarin in 1820, over 1400 natural coumarins
have been isolated. Their biological activities include antico-
agulation, antibiotic, antipsoriasis, antitumor, anti-HIV, etc.
Surprisingly, despite their importance, few mild ways for their
direct synthesis exist. The most common is undoubtedly the
Pechmann condensation and its variants; however, harshness
of the quite strong acid conditions limits its scope. Our initial
approach to this problem focused on the well-documented
electrophilic metalation (ortho-palladation) of the aromatic ring.2
Furthermore, the insertion of an alkyne into the palladium-
aryl bond, a carbopalladation, has also been reported.3 We thus
Results
Our preliminary catalytic system4 employed palladium(+2)
acetate, in acetic acid, as the catalyst.5 However, under these
conditions no coupling occurred between electron-rich phenol61
and ethyl propynoate (2) (Table 1, entries 1,2). Remarkably,
conducting the reaction using formic acid in place of acetic acid
afforded 5,7-dimethoxycoumarin (3)7 in 40% yield (entry 3),
which could be improved to 62% by lowering the palladium
acetate loading to 10 mol % (entry 5). The dramatic difference
in the reactivity of palladium acetate in formic acid, compared
to acetic acid, led us to conjecture that perhaps the formic acid
was reducing the palladium(+2) salt to palladium(0).8 In support
of this deduction, using 5 mol % of (dba)3Pd2‚CHCl3 as catalyst,
in formic acid, afforded 88% yield of coumarin 3 (entry 11).
(2) For reviews of cyclometalation (intramolecular C-H activation), see: (a)
Ryabov, A. D. Chem. ReV. 1990, 90, 403. (b) Canty, A. J. In ComprehensiVe
Organometallic Chemistry II; Able, E. W., Stone, F. G. A., Wilkinson, G.,
Puddephatt, R. J., Eds.; Pergamon: Oxford, 1995; Vol. 9, Chapter 5, pp
242-8. (c) Omae, I. Chem. ReV. 1987, 79, 287. (d). Rybov, A. D. Synthesis
1985, 233.
(3) (a) Pfeffer, M. Pure Appl. Chem. 1992, 64, 335. (b) Vicente, J.; Saura-
Llamas, I.; Turp´ın, J.; de Arellano, M. C. R. Organometallics 1999, 18,
2683 and references therein.
(4) For a preliminary communication of the results described in this section,
see: Trost, B. M.; Toste, F. D. J. Am. Chem. Soc. 1996, 118, 6305.
(5) Stoichiometric Pd(II) has been employed, under similar conditions, for the
coupling of aromatics to olefins: (a) Moritani, I. Synthesis 1973, 524. b.
Trost, B. M.; Fortunak, J. M. Organometallics 1982, 1, 7. (c) Fuchita, Y.;
Hiraki, K.; Kamogawa, Y.; Suenga, M.; Tohgoh, K.; Fujuwara, Y. Bull.
Chem. Soc. Jpn. 1989, 62, 1081.
(6) For substituent effects in electrophilic palladation, see: Fujiwara, Asano,
R.; Moritani, I.; Teranishi, S J. Org. Chem. 1976, 41, 1681.
(7) (a) Mali, R. S.; Yadav, V. J. Synthesis 1977, 464. (b) Calwell, A. G.; Jones,
E. R. H. J. Chem. Soc. 1945, 540.
(8) Trost, B. M.; Li, Y. J. Am. Chem. Soc. 1996, 118, 6625. See also: Tsuji,
J.; Mandai, T. Synthesis 1996, 1.
(1) Kobayashi, S.; Suguira, M.; Kitagawa, H.; Lam, W.W. L. Chem. ReV. 2002,
102, 2227. Also see: Matsuo, J.; Odashima, K.; Kobayashi, S. Synlett 2000,
403.
9
4518
J. AM. CHEM. SOC. 2003, 125, 4518-4526
10.1021/ja0286573 CCC: $25.00 © 2003 American Chemical Society

Atom Economy
A R T I C L E S
Table 1. Palladium Catalyzed Preparation of
5,7-Dimethoxycoumarin
For example, the reaction of 1 with ethyl carbonate protected
alkynol 5c (Table 2, entries 4 and 5) produced coumarin 6c in
58% yield with the Pd2(dba)3 catalyst and 41% using Pd(OAc)2.
Using the optimized palladium(0) catalyzed conditions, we also
found that it is not necessary that the phenol be substituted with
methyl ethers. Thus, phloroglucinol (7) reacts with ethyl
propynoate (2) and ethyl butynoate (5b), catalyzed by Pd2(dba)3,
to afford 5,7-dihydroxy coumarins 8a11 and 8b12 in 79% and
83% yield, respectively (eq 1).
mol %
entry Pd source (mol %) NaOAc
solvent, temp, °C
3
4
1
2
3
4
5
6
7
8
9
Pd(OAc)2 (30)
Pd(OAc)2 (30)
Pd(OAc)2 (30)
50
50
50
20
20
20
20
20
20
40
20
20
20
20
CH3CO2H, 25
CH3CO2H, 70
HCOOH, 50
HCOOH, 50
HCOOH, 50
1:1 HCOOH/DMF, 35
1 equiv of HCOOH/CH2Cl2, 50 18
1:1 HCOOH/CH2Cl2, 25
1:1 HCOOH/PhH, 50
1:1 HCOOH/EtOAc, 50
HCOOH, 50
40
Pd(OAc)2 (10)
Pd(OAc)2 (10)
Pd(OAc)2 (10)
Pd(OAc)2 (10)
Pd(OAc)2 (10)
62
35
42
45 16
10 Pd(OAc)2 (20)
11 Pd2(dba)3 (5)
12 Pd2(dba)3 (2.5)
13 Pd2(dba)3 (5.0)
14 Pd2(dba)3 (5.0)
Increasing the electron richness of the aromatic ring, as in
the case of 3,4,5-trimethoxyphenol 9, has somewhat of a
deleterious effect. Thus, fraxinol methyl ether (10), a coumarin
isolated from Pelargonium reniforme and Pelargonium si-
doides,13 could be prepared in 46% yield by the Pd2(dba)3
catalyzed addition of phenol 9 to ethyl propynoate (eq 2). In
contrast, the Pechmann condensation of 3,4,5-trimethoxyphenol
(9) is often complicated by mixtures of coumarins and
chromones,14 which can be circumvented using a four-step
protocol from 9.15a Envisioning that increased acid sensitivity
in the fomation of 10 due to its increased electron richness may
be responsible for the lower yields led to examination of any
base effect. However, varying the base from sodium acetate to
sodium formate or sodium tetraborate had only a minor, albeit
deleterious, effect on the isolated yield of 10.
88
77
41
HCOOH, 25
CH3CO2H, 25
TFA/DMF, 25
Returning to acetic acid as solvent, now with a palladium(0)
catalyst, decreased the isolated yield of 3 (entry 13). Since
formic acid is a slightly stronger acid than acetic acid, this
difference in the acidity could be the source of the enhanced
reactivity in formic acid. However, use of a stronger acid,
trifluoroacetic acid (pKa ) -0.25), did not afford any of the
desired coumarin (entry 14). This is consistent with formic acid
serving the role of a reductant.
The tolerance of the palladium-catalyzed coumarin-forming
reaction to substitution on the alkynoate was examined by using
alkynes 5a-g (Table 2). The palladium-catalyzed reactions
proceed well with unsubstituted (2) as well as aryl- and alkyl-
substituted (5a-e, h) alkynoates. Most notably is the successful
coupling reaction of methyl 6-cyanohex-2-ynoate (5d) with
phenol 1. The traditional Pechmann cyclization precludes the
use of nitrile groups as they can be hydrolyzed under the
strongly acidic reaction conditions. In contrast to such strong
acid conditions, these reactions are best performed in the
presence of sodium acetate which presumably plays the role of
a general base cocatalyst to promote shuttling of protons required
in these reactions (vide infra). On the other hand, silyl-
substituted alkynoate 5e also participates in the palladium-
catalyzed coupling, but in this case the isolated coumarin (3)
no longer contains the silyl group. The removal of the silyl group
presumably occurs by protonolysis10 of the sp carbon-silicon
bond prior to coumarin formation. The palladium-catalyzed
reaction fails to proceed with alkynoates substituted with
potential nucleophiles, such as carboxylic acids (5f) or free
alcohols (5g). In both cases, the phenol is recovered unreacted
and the â-ketoester derived from hydration of the alkynoate is
isolated.
The question of regioselectivity is of both synthetic and
mechanistic interest. An initial test case was sesamol (11) whose
coumarin analogue 12 is ayapin,16 a naturally occurring cou-
marin exhibiting hemostatic and antibiotic activity.17 Preparation
of ayapin (12) by the reaction of sesamol (1) with ethyl 3,3-
diethoxypropanoate using the traditional Pechmann condensa-
(11) (a) de la Hoz, A.; Moreno, A. Va´squez, E. Synlett 1999, 609. (b) Ganguly,
A. K.; Joshi, B. J.; Kamat, V. N. Manmade, A. H. Tetrahedron 1967, 23,
4377. (c) Kaufmann, K. D.; Kelly, R. C. J. Heterocycl. Chem. 1965, 2, 91.
(12) (a) John, E. V. O.; Israekstam, S. S. J. Org. Chem. 1961, 26, 240. (b) Raj,
H. G.; Gupta, S.; Singh, S.; Singh, A.; Jha, A.; Biht, K. S.; Jain, S. C.;
Parmar, V. S. Bioorg. Med. Chem. 1996, 4, 2225. (c) Li, T.-S.; Zhang,
Z.-H.; Yang, F.; Fu, C.-G. J. Chem. Res. 1998, 38.
(13) (a) Wagner, H.; Bladt, S.; Abraham, D. J.; Lother, H. Tetrahedron Lett.
1974, 3807. (b) Kayser, O.; Kolodziej, H. Phytochemistry 1995, 39, 1181.
For studies on cytotoxicty and antibacterial activity, see: (c) Kolodziej,
H.; Kayser, O.; Woerdenbergm H. J.; Van Uden, W.; Pras, N. Z.
Natuforsch., C 1997, 52, 240; Chem. Abstr. 1997, 127, 60 203. (d) Kayser,
O.; Kolodziej, H. Z. Natuforsch., C 1999, 54, 169; Chem. Abstr. 1999,
131, 2695.
(14) McGarry, L. W.; Detty, M. R. J. Org. Chem. 1990, 55, 4349.
(15) (a). Ahluwalia, V. K.; Prakash, C.; Gupta, Y. K. Natl. Acad. Sci. Lect.
(India) 1978, 1, 215; Chem. Abstr. 1978, 89, 179 800. (b) Ahluwalia, V.
K.; Prakash, C.; Gupta, M. C. Indian J. Chem., Sec. B. 1978, 16, 286. (c)
Paknikar, S. K.; Bhattacharjee, J.; Nadkarni, K. K. J. Indian Inst. Sci. 1994,
74, 277; Chem. Abstr. 1995, 122, 290496.
(16) (a) Soman, S. S.; Trivedi, K. N. Indian J. Chem. B 1993, 32, 372. (b)
Castillo, D.; Rodriguez-Ubis, J. C.; Rodriguez, F. Synthesis 1986, 839. (c)
Yoshida, Y.; Nagai, S.; Oda, N.; Sakakibara, J. Synthesis 1986, 1026. (d)
Kelkar, S. L.; Phadke, C. P.; Marina, S. Indian J. Chem. B 1984, 23, 458.
(e) Fukui, K.; Nakayama, M. Bull. Chem. Soc. Jpn. 1962, 35, 1321.
(17) Chakraborty, D. P.; Sen, M.; Bose, P. K. Trans. Bose. Res. Inst. 1961, 24,
31; Chem. Abstr. 1962, 56, 1835.
Despite the lower catalyst loading, the palladium(0) system
was generally more efficient than the palladium(+2) catalyst.
(9) (a) Boland, G. M.; Donnelly, D. M. X.; Finet, J.-P.; Reaz, M. D. J. Chem.
Soc. 1996, 2591. (b) Ahluwalia, V. K.; Singh, D.; Singh, R. P. Monatsch.
Chem. 1985, 116, 869. (c) Ahluwalia, V. K.; Sunita Indian J. Chem. B
1978, 16, 528. (d) Parmar, V. S.; Bisht, K. S.; Jain, R.; Singh, S.; Sharma,
S. K.; Gupta, S. Indian J. Chem. B 1996, 35, 220. (e) Ahluwalia, V. K.;
Kumar, D. Indian J. Chem. B 1977, 15, 514. (f) Dreyer, D. L.; Munderich,
K. P. Tetrahedron 1975, 31, 287.
(10) Bach, J.; Berenguer, R.; Garcia, J.; Loscertales, T.; Vilarrass, J. J. Org.
Chem. 1996, 61, 9021.
9
J. AM. CHEM. SOC. VOL. 125, NO. 15, 2003 4519

A R T I C L E S
Trost et al.
Table 2. Palladium-Catalyzed Preparation of 4-Substituted 5,7-Dimethoxycoumarins
entry
R
reaction conditions
% yield (product)
1
2
3
4
5
6
7
8
9
a, C6H5
b, CH3
10% Pd(OAc)2, 20% NaOAc, HCOOH, 35 °C
10% Pd(OAc)2, 20% NaOAc, HCOOH, 35 °C
2.5% Pd2(dba)3, 10% NaOAc, HCOOH, 25 °C
10% Pd(OAc)2, HCOOH, 25 °C
2.5% Pd2(dba)3, HCOOH, 25 °C
2.5% Pd2(dba)2 HCOOH, 25 °C
10% Pd(OAc)2, 20% NaOAc,HCOOH, 25 °C
10% Pd(OAc)2, 20% NaOAc,HCOOH, 50 °C
10% Pd(OAc)2, 20% NaOAc,HCOOH, 50 °C
2.5% Pd2(dba)3, 10% NaOAc,HCOOH, 35 °C
69 (6a)
51 (6b)
63 (6b)
41 (6c)
58 (6c)
67 (6d)
53 (4)
c, (CH2)2OCO2Et
d, (CH2)3CN
e, Si(C6H5)(CH3)2
f, (CH2)2CO2t-Bu
g, CH2OH
10
tion is complicated by formation of an ayapin dimer.18 Thus, a
three-step synthesis in which the coumarin ring is formed by a
Wittig reaction16d or a four-step synthesis using an intramo-
lecular Claisen condensation is used to prepare coumarin 12.16e
Using the palladium-catalyzed coumarin formation, ayapin (12)
is available in a single step from sesamol (11) and ethyl
propynoate (2) in 67% yield (eq 3). In this reaction, the coumarin
formation occurred with complete regioselectivity for the
addition of the alkynoate to form the linear coumarin product.
Figure 1. Confirmation of regioselectivity of coumarin formation by NOE.
marins 14b22 and 15b,22a i.e., ortho to the sterically more
demanding methyl group (eq 4). In comparison, the two-step
modified Pechmann condensation of 13a with 3-ethoxyacryloyl
chloride affords a 9:1 mixture of 14a and 15a in slightly lower
yield.18b The fact that the presence of the methyl group did not
significantly alter the regioselectivity suggests that electronic
effects (vide infra) have a greater impact on the regioselectivity
than do steric factors. The regioselectivity of the reaction was
confirmed by NOE experiments on the product coumarins as
summarized in Figure 1.
The coumarin formation starting from 5-methoxyresorcinol
(16) represents a more challenging example of regioselectivity.
Palladium-catalyzed reaction of 16 with ethyl propynoate (2)
produced a 71% yield of a 1.2:1 mixture of regioisomeric
coumarins 17a23 and 18a24 (eq 5). The palladium-catalyzed
reaction of 16 and ethyl butynoate (5b) afforded a 62% yield
One way to interpret the result of eq 3 considers the addition
of the alkynoate to sesamol (11) to have occurred selectively
to the sterically less demanding ortho position. To further
examine the factors responsible for the regioselectivity of
coumarin formation, the palladium-catalyzed coupling of 3-meth-
oxyphenols 13a,b with ethyl propynoate (2) was examined (eq
4). Palladium(+2) acetate catalyzed reaction of phenol 13a
produced a 7.4:1 mixture of herniarin (14a),19 a coumarin
possessing antiinflammatory activity,20 to the regioisomeric
coumarin 15a.21 Switching to the palladium(0) catalyst system
diminished both the yield and regioselectivity of the reaction.
(20) (a) Silva´n, A. M.; Abad, M. J.; Bermejo, P.; Sollhuber, M.; Villar, A. J.
Nat. Prod. 1996, 59, 1183. (b) Harayama, T.; Nakatsuka, K.; Nishioka,
H.; Murakami, K.; Hayashida, N.; Ishii, H. Chem. Pharm. Bull. 1994, 42,
2170. (c) Bhattacharjee, J.; Paknikar, S. K. Indian J. Chem. B 1989, 28,
205. (d) Chatterjee, A.; Bhattacharya, S.; Banerji, J.; Ghosh, P. C. Indian
J. Chem. B 1977, 15, 214. (e) Steck, W.; Bailey, B. K. Can. J. Chem.
1967, 47, 3577.
(21) (a) Ahluwalia, V. K.; Sunita Indian J. Chem. 1977, 15, 936. (b) Ahluwalia,
V. K.; Prakash, C. Ind. J. Chem. B 1977, 15, 423. (c) Ahluwalia, V. K.;
Sachdev. G. P.; Seshardi, T. R. Ind. J. Chem. 1967, 5, 461. (d) Rehder, K.
S.; Kepler, J. A. Synth. Commun. 1996, 26, 4005. (d) Dubuffet, T.; Loutz,
A.; Lavielle, G. Synth. Commun. 1999, 29, 929.
(22) (a) Ahluwalia, V. K.; Dhingra, S.; Kapur, K. Indian J. Chem. B 1979, 18,
79. (b) Malik, O. P. Indian J. Chem. B 1977, 15, 194.
(23) (a) Xie, L.; Takeuchi, Y.; Cosentino, L. M.; Lee, K.-H. J. Med. Chem.
1999, 42, 2662. (b) Ahluwalia, V. K.; Venkateswarlu, P.; Seshardi, T. R.
Indian J. Chem. 1969, 7, 115. (c) Ahluwalia, V. K.; Prakash, C.; Gupta,
Y. K. Natl. Acad. Sci. Lett. (India) 1978, 1, 215; Chem. Abstr. 1979, 89,
179 800.
(24) (a) Kozawa, M.; Baba, K.; Matsuyama, Y.; Hata, K. Chem. Pharm. Bull.
1980, 28, 1782. (b) Murray, R. D. H.; Hogg, T. C. Tetrahedron Lett. 1972,
185.
The reaction of orcinol monomethyl ether (13b) showed a
similar 7.2:1 regioselectivity for coumarin formation of cou-
(18) (a) Crosby, D. G.; Berthold, R. V. J. Org. Chem. 1962, 27, 3083. (b) Zeigler,
T.; Mo¨hler, H.; Effenderger, F. Chem. Ber. 1987, 120, 373.
(19) Ishii, H.; Kaneko, Y.; Miyazaki, H.; Harayama, T. Chem. Pharm. Bull.
1991, 39, 3100.
9
4520 J. AM. CHEM. SOC. VOL. 125, NO. 15, 2003

Atom Economy
A R T I C L E S
Table 3. Alternative Metals for Coupling of Phenols and
of coumarins 17b and 18b21c,25 with a slightly improved
regioselectivity of 1.6:1. In contrast, the traditional Pechmann
condensation of ethyl acetoacetate and 16 affords a 65% yield
of a 1.1:1 mixture of hydroxycoumarins 17b:18b.25 The
regiochemistry of the products was confirmed by NOE experi-
ments (Figure 1). Unfortunately, irradiation of the 4-methyl
substituent showed enhancement of only the 3-hydrogen.
Irradiation of the methoxy group, however, clearly confirms the
regiochemistry of the products. Reaction of 16 with the sterically
more demanding alkynoate (5h) is more regioselective, affording
a 2.4:1 mixture of coumarins 17h:18h in 37% combined yield.
This yield could be improved to 47% by portionwise addition
of 2.5 equiv of the alkynoate, without deterioration of the
regioselectivity (2.4:1).
Alkynoates
entry
reaction conditions
% yield 3
1
2
3
10% Ni(OAc)2, 40% NaOAc, HCOOH, 50 °C
10% PtCl2, 40% NaOAc, HCOOH, 50 °C
1% (PPh3)RhCl, CH3CN, 70 °C
49
4
5
6
10% RuCl3, 30% NH4PF6, THF, 60 °C
10% CpRu(CH3CN)3PF6, acetone, 25 °C
1eq In(OTf)3, THF, 25 °C
7
1eq Hg(OAc)2, THF, 25 °C
8
9
10
11
12
13
1eq AgBF4, THF, 25 °C
1eq Ag(O2CCF3), THF, 25 °C
10% AgBF4, THF, 25 °C
10% AgBF4, 10% o-DPPB, THF, 25 °C
10% AgBF4, 10% o-DPPPy,10% HOAc, THF, 25 °C
10% AgBF4, 10% o-DPPPy,10% CSA, THF, 25 °C
92
65
30
40
7
59
products. Thus, bromine, ortho to the hydroxyl group suffers
hydrogenolysis competitive with coumarin formation.
â-Naphthol (19) is reported to be an unsuitable substrate for
the traditional Pechmann condensation.26 Therefore, preparation
of 5,6-benzocoumarins have generally relied on multistep
sequences based on formylation of 19 followed by addition to
the aldehyde.27 Alternatively, a two-step protocol involving
addition of 19 to triethyl orthoacrylate followed by dehydro-
genation is reported to afford 20.28 Utilizing the palladium-
catalyzed reaction, benzocoumarin 2029 is prepared in 46% yield
as a single regioisomer, by the reaction of 19 with ethyl
propynoate (2) (eq 6). Sonication of the reaction mixture, which
increases the solubility of â-naphthol (19) in formic acid,
improved the isolated yield of 20 to 61%. Surprisingly, all
attempts to react â-naphthol (19) with substituted alkynoates
5a and b failed to produce any of the desired 4-substituted
benzocoumarins.
As expected, electron-deficient phenols, such as 2-acetylphe-
nol, fail to participate in the palladium-catalyzed coumarin
formation. In contrast to 3-methoxy-5-methylphenol (13b) (eq
4), no coupling is observed between 3-methylphenol and ethyl
propynoate (2). Similarly, while 3-methoxyphenol (13a) reacts
with ethyl propynoate (eq 4), changing the phenol to 4-meth-
oxyphenol failed to produce any coupling reaction with the same
alkynoate. The reaction of 3-acetamidophenol with ethyl pro-
pynoate, however, produces a complex mixture in which both
reactants have been consumed. Thus far, the palladium-catalyzed
reaction has been limited to the addition of phenols to alkynoate
esters. Changing the acceptor to 3-butyn-2-one failed to produce
any coupling products in the Pd2dba3-catalyzed coupling with
3,5-dimethoxyphenol (1). Similarly, the reaction of 1 with
phenylacetylene also resulted in recovery of the starting phenol.
However, Pd(OAc)2 catalyzed addition of 1 to ethynyl sulfone
23a32 did produce an 18% yield of (E)-vinyl sulfone 24 (eq 8).
The analogous Pd2dba3-catalyzed reaction produced only a trace
of 24. The instability of ethynyl sulfone 23a to the acidic
reaction conditions prompted us to utilize trimethylsilyl protected
alkyne 23b. However, under the reaction conditions examined,
no coupling products between 23b and 1 were isolated.
The potential utility of bromocoumarins for further elaboration
led to consideration of the chemoselectivity with respect to aryl
bromides.30 As shown in eq 7, palladium acetate catalyzed
reaction of bromophenol 2131 with ethyl propynoate (2) afforded
a 1:5 mixture of brominated 22 and debrominated 3 coumarin
(25) (a) Tera´n, C.; Miranda, R.; Santana, L.; Teijeira, M.; Uriarte, E. Synthesis
1997, 1385. (b) Gia, O.;
(26) Sethna, S. M.; Shah, N. M. Chem. ReV. 1945, 36, 1.
(27) (a) Black, M.; Cadogan, J. I. G.; McNab, H.; MacPherson, A. D.; Roddam,
V. P.; Smith, C.; Swenson, H. R. J. Chem. Soc. 1997, 2483. (b) Taylor, R.
T.; Cassell, R. A. Synthesis 1982, 672. (c) Narasimhan, N. S.; Mali, R. S.
Tetrahedron 1975, 31, 1005. (d) Shriner, R. L. Org. React. 1942, 1, 1.
(28) Panetta, J. A.; Rapoport, H. J. Org. Chem. 1982, 47, 946.
(29) Das Gupta, A. K.; Chatterje, R. M.; Das Carp. K. R. J. Chem. Soc. C 1969,
2618. For an additional alternative, see: Subba Raju, K. V.; Sriman-
narayana, G.; Subba Rao, N. V. Tetrahedron Lett. 1977, 473.
(30) For examples of palladium-catalyzed hydrogenolysis of aryl halides and
triflates with formates, see: (a) Pews, R. G.; Hunter, J. E.; Wehmeyer, R.
M. Tetrahedron Lett. 1991, 32, 7191. (b) Anwer, M. K.; Sherman, D. B.;
Roney, J. G.; Spatola, A. F. J. Org. Chem. 1989, 54, 1284. (c) Dolk, R.;
Schmidt, S. J.; Kruse, L. I. Tetrahedron Lett. 1988, 29, 1581. (d) Numazawa,
M.; Kimura, K.; Ogata, M.; Nagaoka, M. J. Org. Chem. 1985, 50, 5421.
For a review, see: Ram, S.; Ehrenkaufer, R. E. Synthesis 1988, 91.
(31) Kiehlmann, E.; Lauener, R. W. Can. J. Chem. 1989, 67, 335.
Other metal catalysts were screened for their ability to
catalyze the coupling of 3,5-dimethoxyphenol (1) and ethyl
propynoate (2) (Table 3). The other group 10 metals, nickel
and platinum, were examined under conditions similar to those
(32) Bhattacharya, S. N.; Josiah, B. M.; Walton, D. R. M. Organomet. Chem.
Synth. 1971, 1, 145; Chem. Abstr. 1971, 75, 36214.
9
J. AM. CHEM. SOC. VOL. 125, NO. 15, 2003 4521

A R T I C L E S
Trost et al.
used in the palladium-catalyzed reaction (entries 1 and 2). The
nickel catalyst failed to produce any of the desired coumarin.
However, platinum(+2) chloride afforded the coumarin in only
slightly lower yield than the palladium(+2)-catalyzed reaction
(see Table 1, entry 5).33 Rhodium and ruthenium catalysts (Table
3, entries 3-5) proved completely ineffective, mainly resulting
in consumption of the alkynoate.
a wide range of pharmacological actions including anticancer38
and anti-HIV activity.23,39 Xanthoxyletin (26) and its angular
isomer alloxanthoxyletin (27) are examples of this class of
coumarins, which have been isolated from a number of plant
sources40 and even from a marine organism.41
Electrophilic metal reagents, which are known to mediate
additions of nucleophiles to alkynes, were also examined.
Indium(+3) triflate failed to produce any of the desired coumarin
(entry 6).34 Similarly, no reaction occurred between phenol 1
and alkynoate 2 in the presence of mercury(+2) acetate.35a After
these experiments, a similar thallium(+3) polyphosphoric acid
mediated coumarin formation was reported.35b This result
suggests that a strong acid may be required in combination with
these electrophilic metals, such as indium(+3) and mercury-
(+2), to promote the addition of phenols to alkynoates.
The silver(+1) promoted coupling of phenol 1 and alkynoate
2 was also examined (entries 8-10).36 Carrying out the reaction
in the presence of stoichiometric silver(+1) tetrafluoroborate
(entry 8) or trifluoroacetate (entry 9) afforded coumarin 2 in
92% and 65%, respectively. However, unlike other silver-
catalyzed reactions,37 lowering the amount of silver(+1) to 10
mol % dramatically reduced the yield to 30% (entry 10).
We postulated that the reaction was promoted by coordination
of silver(+1) to the alkyne, followed by nucleophilic addition
of the phenol (eq 9). This mechanism would result in the
formation of the vinyl silver intermediate, which could be stable
enough to prevent turnover of the silver. We therefore examined
the possibility that this problem could be alleviated by the use
of phosphine ligands o-DPPB and o-DPPPy. These phosphines
could potentially serve not only as ligands for silver(+1) but
also as facilitators of protonation of the silver-sp2 carbon bond.
Addition of these ligands did slightly enhance the catalytic
activity of the silver(+1) salt (entries 11-13), affording the
coumarin in 59% yield in the presence of 10% silver tetrafluo-
roborate and a strong acid, camphorsulfonic acid (entry 13). It
remains unclear, however, if they are acting as depicted in eq
9 or simply serving as a general acid source. Another important
distinction between the palladium-catalyzed and the silver(+1)-
mediated coumarin forming reaction is that the latter does not
tolerate substitution on the alkynoate. For example, the reaction
of 1 with ethyl butynoate (5b) proceeds smoothly under
palladium catalysis (see Table 2). Conversely, the silver-
mediated process is completely unproductive.
In our initial approach to the xanthoxyletins, we wanted to
examine the palladium-catalyzed coumarin-forming reaction
between chroman-4-one 2942 and ethyl propynoate (2). Unfor-
tunately, the reaction of phenol 29 with 2 was unsuccessful;
both in the presence of Pd2dba3 and Pd(OAc)2, 29 was recovered
unchanged.
The presence of the electron-withdrawing ketone functionality
was presumably responsible for the lack of reactivity of 29.
Our initial solution to this problem was to convert the chrom-
4-one into the chromene functionality present in the natural
product. To this end, selective tosylation of the less hindered
oxygen, followed by methylation of the remaining phenol,
afforded a selectively protected chrom-4-one. Reduction of the
ketone with borane followed by the acidic elimination of the
resulting borate provided the chromene, which was deprotected
with ethanolic potassium hydroxide to afford the desired phenol
30. Disappointingly, reaction of 30 with ethyl propynoate (2),
catalyzed by Pd2dba3 or Pd(OAc)2, did not afford any of the
desired coumarins. In this case, unlike the reaction of chrom-
4-one 29, chromene 30 was consumed under the reaction
conditions. Analysis of the crude reaction mixture indicated that
the chromene olefin was not stable to the reaction conditions.
(35) (a) For Hg(+2)-mediated addition to alkynes, see: (a) Subramanian, R.
S.; Balasubramanian, K. K. J. Chem. Soc., Chem. Commun. 1990, 1469.
Huang, H.; Forsyth, C. J. J. Org. Chem. 1997, 62, 4746. Carollo, L.; Floris,
B. J. Organomet. Chem. 1999, 583, 80 and references therein. (b) For Tl-
(+3) mediated addition, see: Shamsuddin, K. M.; Jamshed Ahmed Siddiqui,
M. J. Chem. Res. Synop. 1998, 392.
(36) For silver(+1) catalyzed addition to alkynes, see: Kotora, M.; Neghsi, E.
Synthesis 1997, 121. (b) Ogawa, Y.; Maruno, M.; Wakamatsu, T. Synlett
1995, 871.
(37) Marshall, J. A.; Bartley, G. S.; Wallace, E. M. J. Org. Chem. 1996, 61,
5728 and references therein.
(38) (a) Guntatilaka, A. A. L.; Kingston, D. G. I.; Wijertane, E. M. K.; Bandara,
B. M. R.; Hofmann, G. A.; Johnason. R. K. J. Nat. Prod. 1994, 57, 518.
(b) Magiatis, P.; Melliou, E.; Skaltsounis, A.; Mitaku, S.; Leonce, S.;
Renard, P.; Pierre, A.; Alassi, G. J. Nat. Prod. 1998, 61, 982 and references
therein.
(39) Takeuchi, Y.; Xie, L.; Cosentino, L. M.; Lee, K.-H. Bioorg. Med. Chem.
Lett. 1997, 7, 2573.
(40) (a) Mohan, P. S.; Ramesh, M.; Shanmugam, P. J. Nat. Prod. 1985, 48,
501. (b) Lasak, E. V.; Pinhey, J. T. Aust. J. Chem. 1969, 22, 2175. c. King,
F. E.; Housely, J. R.; King, S. J. J. Chem. Soc. 1954, 1392. (c) Tsukayama,
M.; Sakamoto, T.; Horie, T.; Masamura, M.; Nakayama, M. Heterocycles
1981, 16, 955.
(41) Tillekeratne, L. M. V.; De Silva, E. D.; Mahindaratne, M. P. D.; Schmitz,
F. J.; Gunasekera, S. P.; Alderslade, P. J. Nat. Prod. 1989, 52, 1303.
(42) (a) Tima´r, T.; Ja´szbere´nyl, J. C. J. Heterocycl. Chem. 1988, 25, 871. (b)
Wolfrom, M. L.; Koos, E. W.; Bhat, H. B. J. Org. Chem. 1967, 32, 1056.
(c) Sowmithran, D.; Rajendra Prasad, R. J. Synthesis 1985, 545.
Synthesis of the Xanthoxyletins. The pyranocoumarins,
several of which have been isolated from natural sources, display
(33) For an example of Pt(+2)-catalyzed addition to alkynes, see: Katoaka,
Y.; Matsumoto, O.; Tani, K. Organometallics 1996, 15, 5246 and references
therein.
(34) For examples of carboindination of alkynes, see: (a) Klaps, E.; Schmid,
N. J. Org. Chem. 1999, 64, 7537. (b) Fujiwara, N.; Yamamoto, Y. J. Org.
Chem. 1999, 64, 4095.
9
4522 J. AM. CHEM. SOC. VOL. 125, NO. 15, 2003

Atom Economy
A R T I C L E S
The DDQ oxidation42 of the chromans 33 and 34, to the
chromenes, completes the synthesis of xanthoxyletin (26) and
alloxanthoxyletin (27) (eq 11 and 12). Using the palladium-
catalyzed coumarin synthesis, alloxanthoxyletin (27) is available
in four steps, and 36% overall yield, from chrom-4-one 29. The
regioisomeric coumarin xanthoxyletin (26) is also produced in
13% overall yield.
Table 4. Regioselectivity of Pyranocoumarin Formation
entry
phenol
conditions
yield
33:34:35
1
31
10% Pd(OAc)2, 20% NaOAc,
HCOOH, 25 °C
63%
1.0:3.3:0
Discussion
The initial mechanistic proposal was based on the notion that
an electron-rich phenol could be ortho-palladated by an elec-
trophilic palladium(+2) source.3 Therefore, the first phenols
examined in the palladium-catalyzed coumarin forming reaction
all possessed strong electron-donating groups. Indeed, the
coupling reaction of phenols 1, 7, 9, 11, and 16, all of which
contain at least two activating groups, with a variety of
alkynoates, afford coumarins in good to excellent yields (46-
88%). On the other hand, the electron-deficient 2-acetylphenol
failed to produce any coumarin product. These results clearly
indicate that the reaction is favored by electron-donating
substituents on the phenol; however, it was unclear how electron
rich the phenol had to be for the reaction to proceed.
The successful reaction of phenols 13a and 13b with ethyl
propynoate (2) (eq 4) indicates that only a single additional
electron-donating substituent is required. The failure of 3-me-
thylphenol to participate in the palladium-catalyzed reaction
clearly demonstrates that at least one additional strong electron-
donating group, such as methoxy, is required and that alkyl is
not a strong enough activating substituent. Moreover, the
position of the electron-donating group is also important, as
demonstrated by the difference in reactivity between m-methoxy
(13a) and p-methoxy phenols. 3-Methoxyphenol (13a) reacts
with ethyl propynoate (2a) at the position para to the methoxy
group (eq 4), while 4-methoxyphenol fails to react with the same
alkynoate.
2
3
1eq AgBF4, THF, 25 °C
2.5% Pd2(dba)3, HCOOH, 30 °C
then CH3I, K2CO3, acetone
2.5% Pd2(dba)3, 10% NaOAc,
HCOOH, 25 °C
54%
56%
1.0:4.3:0
1.0:6.6:1.3
32
4
68%
1.0:2.8:0
then CH3I, K2CO3, acetone
To avoid this complication, the differentially protected
chromone derived from 29 was subjected to a Clemmenson
reduction, followed by deprotection of the tosyl group to afford
chroman 31.43 Reaction of phenol 31 with ethyl propynoate (2),
catalyzed by palladium(+2) acetate, produced a 1:3.3 mixture
of dihydroxanthoxyletin (kanzanol Q, 33) and dihydroalloxan-
thoxyletin (34) (Table 4, entry 1). The regiochemistry of the
coumarin formation was established by conversion of 33 and
34 to the natural products (eq 11 and 12). The silver-mediated
addition of ethyl propynoate to 2 also favored reaction at the
8-position of the chroman (Table 4, entry 2). Similar selectivity
for reaction at the 8-position of 31 has been observed in other
electrophilic reactions, including the Pechmann condensation
(42% yield, ratio 1:6:trace).43a
Unprotected chroman-4-one 29 was similarly reduced to
afford chroman 32.44 Chroman 32 offers an additional regio-
selectivity issue, since addition of the alkynoate to the 6-position
can lead either to 33 or to 35, depending on which hydroxyl
group cyclizes onto the intermediate cinnamate ester. The
reaction of 32 with ethyl propynoate catalyzed by Pd2dba3, and
subsequent methylation of the crude reaction mixture, afforded
a mixture of three coumarins 33:34:35 (Table 4, entry 3). The
regioselectivity of addition to the 8-position versus the 6-position
(2.9:1) is similar to that obtained for reaction with phenol 31
(entry 1). Addition of 10 mol % tetrabutylammonium chloride
to the reaction mixture did not have any observable impact on
the yield or selectivity of the reaction. In contrast, addition of
10 mol % sodium acetate increased the yield to 68% while
maintaining the regioselectivity for addition to the 8-position
to produce dihydroalloxanthoxyletin (34) (entry 4). Notably,
under the latter conditions, only two out of the three possible
regioisomeric coumarins were produced.
The palladium-catalyzed reactions proceed well with unsub-
stituted (2) as well as aryl and alkyl substituted (5a-e, h)
alkynoates. However, the palladium catalyzed reaction fails to
proceed with alkynoates substituted with potential nucleophiles,
such as carboxylic acids (5f) or free alcohols (5g). In these cases,
the â-keto ester derived from hydration of the alkynoate is
isolated.
The failure of alkynoate 5f to react with 3,5-dimethoxyphenol
(1) may arise from palladium(+2)-catalyzed intramolecular
addition of the homopropargylcarboxylic acid, resulting from
acidic deprotection of the tert-butyl ester, to the alkynoate (eq
13).45 Alternatively, insertion of palladium(0) into the carboxylic
(45) (a) Arcardi, A.; Burini, A.; Cacchi, S.; Delmastro, M.; Marinelli, F.; Pietroni,
B. R. J. Org. Chem. 1992, 57, 976. (b) Yanagihara, N.; Lambert, C.; Iritani,
K.; Utimoto, K.; Nozaki, H. J. Am. Chem. Soc. 1986, 108, 2753. (c)
Lambert, C.; Utimoto, K.; Nozaki, H. Tetrahedron Lett. 1984, 25, 5323. A
similar reaction is reported for propargyl alcohols in the presence of carbon
dioxide: Iritani, K.; Yanagihara, N.; Utimoto, K. J. Org. Chem. 1986, 51,
5501.
(43) (a) Matsui, T.; Nishimura, S.; Nakayama, M.; Hayashi, S.; Fukui, K. Bull.
Chem. Soc. Jpn. 1977, 50, 1975. (b) Lahey, F. N.; Stick, R. V. Aust. J.
Chem. 1973, 26, 2291.
(44) Bridge, W.; Heyes, R. G.; Robertson, A. J. Chem. Soc. 1937, 279.
9
J. AM. CHEM. SOC. VOL. 125, NO. 15, 2003 4523

A R T I C L E S
Trost et al.
Scheme 1. Mechanistic Proposals for Palladium Catalyzed
acid O-H bond, followed by hydropalladation-reductive
elimination, leads to a similar methylene lactone.46 Presumably,
the intermolecular addition of formic acid to the other alkynoates
is also occurring;47 however, the addition of phenol is competi-
tive. Indeed, with alkynoates 5a and 5d the corresponding â-keto
esters, 37 and 38, were detected along with the coumarins.
Coumarin Formation
The isolation of â-keto esters 37 and 38 from the reactions
depicted in Table 2 raises the possibility that they were
intermediates in the palladium-catalyzed coumarin formation.
To examine this possibility, phenol 1 was reacted with ethyl
acetoacetate (29) under the standard palladium-catalyzed condi-
tions (eq 14). Coumarin 6b was not obtained in this reaction,
thus excluding â-keto esters as intermediates in the mechanism
of the palladium-catalyzed coumarin formation.
to palladium(0), leads to a productive coupling reaction (see
Table 1). Indeed, in the absence of substrate, treatment of
palladium acetate with formic acid led to the formation of
palladium black. Furthermore, a palladium(0) source, Pd2dba3,
is an effective catalyst in both acetic and formic acid, although
the latter solvent gave better results. Finally, reaction with
2-bromophenol produced the debrominated coumarin as the
major product (eq 7). The fact that 8-bromocoumarin (22) is
isolated in this reaction suggests that coumarin formation is
competitive with oxidative addition of palladium into the aryl
bromide bond. This observation also led us to exclude a
mechanism, similar to that proposed for silver(+1) or thallium-
(+3), in which coordination of the palladium(+2) to the alkyne
activates it for nucleophilic addition of the phenol.
A mechanistic rational derives from the effectiveness of
Pd2dba3, in formic acid, to promote the cycloisomerization of
enynes.7 Since this reaction involves the hydropalladation of
an alkyne, the formation of HPdX from palladium(0) and formic
appears likely. Furthermore, other palladium(0) catalyzed ad-
ditions of phenols and alcohols to alkynes, in the presence of
acid, have been described.45 These reactions are proposed to
involve hydropalladation by a HPdX species, to produce a
vinylpalladium intermediate, which is then carbonylated. In
support of this proposal, a hydridopalladium species (HPd-
O2CF3) has recently been observed in the 1H NMR of a mixture
of Pd(PPh3)4 and trifluoroacetic acid.46a
Similarly, the proposed mechanism for the coumarin forma-
tion involves the formation of a hydridopalladium intermediate
from the reaction of palladium(0) and formic acid (Scheme 1).
One mechanistic possibility involves addition of the hydridopal-
ladium complex to the alkynoate (a hydropalladation) to form
a vinylpalladium intermediate 40. The regiochemistry of the
hydropalladation is consistent with that observed for the
hydropalladation of ethyl octynoate. The different polarization
of the H-PdX bond compared to C-PdX bond accounts for
the different regioselectivity of addition to alkynoates for the
two processes. The fact that protonation of Pd(0) is highly
reversible supports this argument. Nucleophilic addition of
phenol on 40 affords either the C-bound (41) or O-bound (42)
vinylpalladium complex. In any case, 41 and 42 are intercon-
vertible by a keto-enol type tautomerization. In a related
system, CO has been reported to insert into intermediates such
as 42 to afford R,â-unsaturated esters.46b In the absence of
carbon monoxide, reductive elimination of 41 regenerates the
palladium(0) catalyst and initially affords E-cinnamate ester 43.
Olefin E-Z isomerization of 43 followed by lactonization
The palladium-catalyzed coumarin formation proceeds with
excellent (>95:5, eqs 3 and 6) to moderate regioselectivity (1.2-
2.4:1, eq 5). In an attempt to rationalize the origin of the
regioselectivity of addition, ab intio (HF-6-31G**) calculations
were performed to determine both the atomic charges and the
HOMO of the phenols. In general, reactions of aromatic
compounds with hard (high lying LUMO’s) electrophiles are
charged controlled, while reactions with soft (low lying
LUMO’s) electrophiles are under frontier molecular orbital
(FMO) control. The calculated atomic charges are not consistent
with the experimentally determined regioselectivity. For ex-
ample, in the case of sesamol (11) the greatest amount of
negative charge is found at the 2-position (atomic charge )
-0.44), while reaction occurs selectively at the 6-position
(atomic charge ) -0.29) (eq 3). On the other hand, the observed
regioselectivity fits well with the calculated HOMO of the
phenols. In accord with the observed regioselectivity, the HOMO
coefficient at the 6-position of sesamol (11) is significantly larger
than that at the 2-position.
Our original mechanistic proposal was based on the idea than
an electrophilic palladium(+2) source could ortho-palladate48
an electron-rich phenol. Several experimental observations led
to the exclusion of this mechanism of coumarin formation. First,
insertion of alkynoates into aryl carbon-palladium bonds is
reported to proceed with regioselectivity opposite to that
observed in the coumarin formation.46 Second, a number of
results suggest that the reaction involves a palladium(0) rather
than a palladium(+2) catalyst. The reaction does not proceed
with a palladium(+2) source in acetic acid. However, changing
the solvent to formic acid, which may reduce palladium(+2)
(46) (a) Kushino, Y.; Itoh, K.; Miura, M.; Nomura, M. J. Mol. Catal. 1994, 89,
151. (b) Itoh, K.; Miura, M.; Nomura, M. Tetrahedron Lett. 1992, 33, 5369.
(47) Lu, X.; Zhu, G.; Ma, S. Tetrahedron Lett. 1992, 33, 7205.
(48) Fujiwara et. al. favor an ortho-palladation (C-H activation) mechanism
for their Pd(OAc)2-TFA catalyst system: (a) Jia, C.; Piao, D.; Oyamada,
J.; Lu, W.; Kitamura, T.; Fujiwara, Y. Science 2000, 287, 1992. (b) Jia,
C.; Lu, W.; Oyamda, J.; Kitamura, T.; Matsuda, K.; Irie, M.; Fujiwara, Y.
J. Am. Chem. Soc. 2000, 122, 7252. (c) Jia, C.; Piao, D.; Kitamura, T.;
Fujiwara, Y. J. Org. Chem. 2000, 65, 7516. (d) Oyamada, J.; Jia, C.;
Fujiwara, Y.; Kitamura, T. Chem. Lett. 2002, 380.
9
4524 J. AM. CHEM. SOC. VOL. 125, NO. 15, 2003

Atom Economy
A R T I C L E S
distilled from magnesium methoxide. Acetic acid, formic acid, and 95%
ethanol were used as obtained. All phenols were obtained from Aldrich
or Fluka and recrystallized or distilled prior to use. Palladium acetate,
Pd(OAc)2, was recyrstallized from hot acetic acid under a nitrogen
atmosphere. Tris(dibenzylidideneacetone) monochloroform complex,
Pd2dba3‚CHCl3, was prepared according to the literature.50All other
reagents were used as obtained unless otherwise noted.
affords coumarin 48. In support of the isomerization, subjecting
cinnamate ester 4 to the reaction conditions affords coumarin 1
in 75% yield (eq 15).
5,7-Dimethoxycoumarin (3). The procedures described for prepara-
tion of 5,7-dimethoxycoumarin (3) from 3,5-dimethoxyphenol (1) are
representative of those used to prepare subsequent coumarins.
Method A. To a mixture of 3,5-dimethoxyphenol (1) (50 mg, 0.324
mmol), palladium acetate (7 mg, 0.032 mmol), and sodium acetate (5
mg, 0.065 mmol) under a nitrogen atmosphere, formic acid (3 mL)
was added and the resultant mixture placed into a preheated to 50 °C
in an oil bath. After 2 min at this temperature, ethyl propynoate (2)
(66 µL, 0.649 mmol) was added and the resulting brown solution stirred
at 50 °C for 14 h. The reaction mixture was cooled to room temperature,
diluted with methylene chloride (10 mL), and washed with water (15
mL), 5% aqueous sodium bicarbonate (15 mL), and brine (15 mL).
The organic layer was dried (MgSO4) and concentrated in vacuo to
give a brown solid. Flash chromatography eluting with methylene
chloride afforded 51 mg (62%) of 5,7-dimethoxycoumarin (3).
Method B. To a mixture of 3,5-dimethoxyphenol (1) (2.00 g, 13.0
mmol), Pd2dba3‚CHCl3 (0.34 g, 0.325 mmol), and sodium acetate (0.11
g, 0.725 mmol) under a nitrogen atmosphere formic acid (13 mL) was
added followed by ethyl propynoate 2 (2.6 mL, 26 mmol). The resulting
brown/purple solution was stirred at room temperature for 16 h, then
diluted with methylene chloride (20 mL), and washed with water (25
mL), 5% aqueous sodium bicarbonate (25 mL), and brine (25 mL).
The organic layer was dried (MgSO4) and concentrated in vacuo to
give a brown solid. Flash chromatography eluting with methylene
chloride followed by recrystallization from CH2Cl2/pentanes afforded
2.51 g (77%) of 5,7-dimethoxycoumarin (3) as a slightly yellow solid,
mp 143-145 °C (lit.6 mp 143-144 °C). IR(film): 2931, 1723, 1616,
1228, 1152, 1113, 827 cm-1. 1H NMR (300 MHz, CDCl3): δ 7.96 (d,
J ) 9.6 Hz, 1 H), 6.41 (d, J ) 2.2 Hz, 1 H), 6.27 (d, J ) 2.2 Hz, 1 H),
6.15 (d, J ) 9.6 Hz, 1 H), 3.88 (s, 3 H), 3.85 (s, 3 H). 13C NMR (75
MHz, CDCl3): δ 164.1, 161.8, 157.3, 157.1, 139.0, 111.0, 104.2, 94.9,
93.1, 56.1, 55.9.
5,7-Dihydroxycoumarin (8a). Using method B, phloroglucinol (7)
(100 mg, 1.30 mmol), Pd2dba3‚CHCl3 (41 mg, 0.04 mmol), sodium
acetate (7 mg, 0.08 mmol), and ethyl propynoate (2) (100 µL, 1.59
mmol) were reacted in formic acid (0.8 mL) at room temperature for
16 h. After 16 h, the brown solid which precipitated from the reaction
mixture was collected by suction filtration and recrystallized from hot
water to afford 8a (140 mg, 79%) as a light yellow solid, mp 270-
272 °C (lit. mp11c 280 °C). IR(KBr): 3246, 1686, 1641, 1618, 1479,
1375, 1292, 1152, 1070 cm-1. 1H NMR (300 MHz, DMSO-d6): δ 8.75
(br s, 2 H), 7.95 (d, J ) 9.9 Hz, 1 H), 6.23 (d, J ) 2.1 Hz, 1 H), 6.12
(d, J ) 2.1 Hz, 1 H), 5.91 (d, J ) 9.9 Hz, 1 H). 13C NMR (75 MHz,
DMSO-d6): δ 162.1, 160.8, 158.8, 157.3, 156.1, 107.0, 101.2, 98.9,
93.1.
Alternatively, the hydridopalladium intermediate can undergo
exchange with the phenol to produce hydridopalladium phe-
noxide 44.49 Coordination of the alkyne to 44 affords 45, which
can undergo either hydrometalation or carbometalation of the
alkyne. Hydrometalation of the alkyne generates intermediate
42, whose conversion to coumarin 48 has already been
discussed. Carbometalation of the alkynoate, perhaps via 46,
generates the vinyl hydridopalladium intermediate 47, which
can undergo a reductive elimination to regenerate palladium(0)
and afford cinnamate ester 43.
The equilibrium between an O- (41 or 45) and C-bonded (42
or 46) phenol is consistent with the preference for addition of
the alkynoate to the position ortho to the hydroxyl group. In
some coupling reactions, such as with sesamol (11) (eq 3), the
alkynoate addition occurs with exceptional selectivity for one
of the ortho positions. The ab intio calculations strongly suggest
that the regioselectivity of coumarin formation is FMO con-
trolled. This is not unexpected, since both palladium(+2) (as
in 41) and the palladium-coordinated alkynoate (as in 45) would
be predicted to have low-lying LUMO’s. This further suggests
that, like in most FMO controlled electrophilic aromatic
substitutions, the addition of the phenol to the alkynoate is both
the regiochemistry- and rate-determining step.
Conclusion
In conclusion, the combination of palladium catalysts and
formic acid is an effective system for the catalytic ortho
vinylation of phenols with activated alkynes. This catalyst
system provides a new atom economic synthesis of coumarins
through the simple addition of phenols with alkynoate esters,
the first criterion for an atom economical reaction. While the
yields are not quantitative, they are higher than for other
coumarin syntheses. This is demonstrated by the application of
the palladium-catalyzed coumarin synthesis to the efficient
preparation of various coumarin natural products including
faxinol methyl ether, ayapin, herniarin, xanthoxyletin, and
alloxanthoxyletin. The mechanism of this new reaction appears
to involve a palladium(0) species. Several mechanistic pos-
sibilities involving hydridopalladium intermediates are pre-
sented. Much like electrophilic aromatic substitutions, the
coumarin formation with unsymmetrical phenols appears to be
frontier molecular orbital directed.
4-Methyl-5,7-dihydroxycoumarin (8b). Using method B, phloro-
glucinol 7 (100 mg, 1.30 mmol), Pd2dba3‚CHCl3 (41 mg, 0.04 mmol),
sodium acetate (7 mg, 0.08 mmol) and ethyl butynoate 5b (100 µL,
1.50 mmol) were reacted in formic acid (0.8 mL) at room temperature
for 16h. After 16h, the brown solid which precipitated from the reaction
mixture is collected by suction filtration and recrystallized from hot
water to afford 8b (135 mg, 83%) as a light yellow solid, mp
282-4 °C (lit.12a mp 284-285 °C). IR(KBr): 3158, 1669, 1622, 1558,
1386, 1300, 1160, 1097, 1024 cm-1. 1H NMR (300 MHz, DMSO-d6):
δ 6.25 (d, J ) 1.8 Hz, 1H), 6.16 (d, J ) 1.8 Hz, 1H), 5.83 (s, 1H),
4.2-3.8 (br s, 2H), 2.48 (s, 3H). 13C NMR (75 MHz, DMSO-d6): δ
161.7, 160.7, 158.5, 157.0, 155.6, 109.4, 102.7, 98.6, 95.1, 24.0.
Experimental Section
All reactions were performed under a nitrogen atmosphere unless
otherwise indicated. Solvents were generally freshly distilled before
use: methylene chloride, benzene, and toluene from calcium hydride;
DMF from barium oxide; THF and diethyl ether from sodium
benzophenone ketyl; acetone from cacium sulfate. Methanol was
(49) For examples of hydridometal phenoxides, see: (a) Bergman, R. G.
Polyhedron 1995, 14, 3227. (b) Kaplan, A. W.; Bergman, R. G. Organo-
metallics 1998, 17, 5072 and references therein.
(50) Ukai, T.; Kawazura, H.; Iishii, Y.; Bonnett, J. J.; Ibers, J. A. J. Organomet.
Chem. 1974, 65, 263.
9
J. AM. CHEM. SOC. VOL. 125, NO. 15, 2003 4525

A R T I C L E S
Trost et al.
6,7-(Methylenedioxy)coumarin (Ayapin, 12). Using method B,
phenol 11 (200 mg, 1.45 mmol), Pd2dba3‚CHCl3 (38 mg, 0.036 mmol),
sodium acetate (12 mg, 0.14 mmol), and ethyl propynoate 2 (290 µL,
2.9 mmol) were reacted in formic acid (1.5 mL) at room temperature
for 16 h. Flash chromatography eluting with 30% ethyl acetate:hexanes
followed by recrystallization from carbon tetrachloride afforded 12 (128
mg, 67%) as a tan solid, mp 225-227 °C (lit.16a mp 223 °C). IR(film):
2938, 1723, 1634, 1453, 1333, 1272, 1223, 1100, 1038 cm-1. 1H NMR
(500 MHz, CDCl3) δ 7.56 (d, J ) 9.8 Hz, 1 H), 6.81 (s, 2 H), 6.26 (d,
J ) 9.8 Hz, 1 H), 6.06 (s, 2 H).
34: mp 150-154 °C (lit.52b mp 155 °C). IR (film): 2963, 2849,
1
1730, 1619, 1366, 1252, 1160, 1112, 819, 796 cm-1. H NMR (300
MHz, CDCl3): δ 7.97 (d, J ) 9.6 Hz, 1 H), 6.34 (s, 1 H), 6.10 (d, J
) 9.6 Hz, 1 H), 3.85 (s, 3 H), 2.60 (t, J ) 6.8 Hz, 2 H), 1.79 (t, J )
6.8 Hz, 2 H), 1.34 (s, 6 H). 13C NMR (75 MHz, CDCl3): δ 162.0,
161.1, 155.2, 150.8, 139.2, 110.4, 105.6, 103.7, 90.4, 75.6, 55.7, 31.7,
26.5, 16.5. Anal. Calcd for C15H16O4: C, 69.22; H, 6.20. Found C,
69.95; H, 6.39.
35: mp 155-158 °C (lit.53 mp 162-163 °C). IR (film): 2973, 2855,
1
1730, 1602, 1367, 1205, 1140, 1112, 819 cm-1. H NMR (300 MHz,
Dihydroxanthoxyletin (Kanzanol Q, 33), Dihydroalloxanthoxy-
letin (34), and 5-Methoxydihydroseslin (35). From 2,2-Dimethyl-
7-hydroxy-5-methoxychroman (31). (a) Silver Mediated. To a
solution of phenol 31 (86 mg, 0.417 mmol) and silver tetrafluoroborate
(89 mg, 0.96 mmol) in THF (0.4 mL) was added ethyl propynoate (84
µL, 0.83 mmol) and the resulting orange solution was stirred at room
temperature for 2 h. After 2 h, the reaction mixture was diluted with
ether (10 mL) and the precipitated silver removed by filtration. The
filtrate was washed with 1 N sodium hydroxide (10 mL), brine (10
mL), dried (MgSO4), and concentrated in vacuo. Flash chromatography
eluting with 2:1 petroleum ether:diethyl ether afforded 33 (11 mg, 10%)
and 34 (47 mg, 44%).
(b) Palladium-Catalyzed. To a test tube containing phenol 31 (50
mg, 0.242 mmol), palladium acetate (5 mg, 0.022 mmol), and sodium
acetate (4 mg, 0.05 mmol) was added formic acid (0.3 mL) and the
solution stirred at room temperature for 5 min. To this solution was
added methyl propynoate (50 µL, 0.49 mmol), and the solution was
stirred at room temperature for 8 h. The reaction mixture is then diluted
with ether (5 mL), washed with 1M sodium hydroxide (5 mL) and
brine (5 mL), dried (MgSO4), and concentrated to afford a yellow solid.
Flash chromatography eluting with 2:1 petroleum ether:diethyl ether
afforded 33 (9 mg, 15%) and 34 (30 mg, 48%).
From 5,7-Dihydroxy-2,2-dimethylchroman (32). To a solution of
chroman 32 (100 mg, 0.51 mmol), Pd2(dba)3 (14 mg, 0.013 mmol),
and sodium acetate (4 mg, 0.05 mmol) in formic acid (0. 5 mL) was
added ethyl propynoate (2) and the reaction mixture stirred at room
temperature. Two additional equivalents of ethyl propynoate (2, 50 µL,
0.49 mmol) were added after 3 and 6 h, and stirring was continued for
a total of 10 h. After 10 h, the reaction mixture was flushed through a
pad of silica eluting with 1% methanol/methylene chloride and the
filtrate concentrated in vacuo to afford a brown solid (390 mg). The
solid was taken into acetone (5 mL) and treated with potassium
carbonate (211 mg, 1.53 mmol) followed by iodomethane (48 µL, 0.77
mmol), and the resulting suspension was heated at reflux for 6 h. The
reaction mixture was cooled to room temperature, the potassium salts
were removed by filtration and the filtrate concentrated in vacuo. Flash
chromatography eluting with 2:1 petroleum ether:diethyl ether gave
33 (22 mg, 18%) and 34 (62 mg, 50%).
CDCl3): δ 7.98 (d, J ) 9.6 Hz, 1 H), 6.18 (s, 1 H), 6.12 (d, J ) 9.6
Hz, 1 H), 3.83 (s, 3 H), 2.79 (t, J ) 6.8 Hz, 2 H), 1.82 (t, J ) 6.8 Hz,
2 H), 1.35 (s, 6 H). 13C NMR (75 MHz, CDCl3): δ 161.9, 158.1, 155.1,
154.0, 139.1, 109.9, 103.4, 101.5, 95.6, 76.0, 55.8, 31.7, 26.6, 15.3.
Xanthoxyletin 26. A solution of 33 (22 mg, 0.08 mmol) in benzene
(1 mL) was treated with DDQ (97 mg, 0.44 mmol) and the resulting
brown solution heated at reflux for 20 h. The reaction mixture was
cooled to room temperature and diluted with ether (15 mL), washed
with 1 N sodium hydroxide (10 mL), water (10 mL), and brine (10
mL), dried (MgSO4), and concentrated in vacuo. Flash chromatography
eluting with 1:1 ether:pentanes and recrystallization from ethanol
afforded xanthoxyletin 26 (16 mg, 75%), mp 130-132 °C (lit. mp54
134-135 °C). IR (film): 2927, 2855, 1727, 1614, 1462, 1370, 1240,
1
1140, 1024 cm-1. H NMR (300 MHz, CDCl3): δ 7.97 (d, J ) 9.6
Hz, 1 H), 6.61 (d, J ) 10.0 Hz, 1 H), 6.34 (s, 1 H), 6.16 (d, J ) 9.6
Hz, 1 H), 5.55 (d, J ) 10.0 Hz, 1 H), 3.87 (s, 3 H), 1.46 (s, 6 H). 13
C
NMR (75 MHz, CDCl3): δ 161.1, 157.5, 155.5, 152.8, 138.5, 115.8,
112.3, 111.3, 107.4, 100.8, 77.9, 63.6, 28.1, 14.1.
Alloxanthoxyletin (27). A solution of 34 (43 mg, 0.167 mmol) in
benzene (5 mL) was treated with DDQ (189 mg, 0.79 mmol) and the
resulting brown solution heated at reflux for 16 h. The reaction mixture
was diluted with ether (15 mL), washed with 1 N sodium hydroxide
(15 mL), water (15 mL), and brine (15 mL), dried (MgSO4), and
concentrated in vacuo. Flash chromatography eluting with 1:1 ether:
pentanes and recrystallization from hexanes afforded alloxanthoxyletin
(27) (33 mg, 76%), mp 113-6 °C (lit.40c mp 114-115 °C). IR (film):
2977, 2850, 1738, 1615, 1370, 1120 cm-1 1H NMR (300 MHz,
.
CDCl3): δ 7.97 (d, J ) 9.6 Hz, 1 H), 6.61 (d, J ) 10.0 Hz, 1 H), 6.34
(s, 1 H), 6.16 (d, J ) 9.6 Hz, 1 H), 5.55 (d, J ) 10.0 Hz, 1 H), 3.87
(s, 3 H), 1.46 (s, 6 H). 13C NMR (75 MHz, CDCl3): δ 161.5, 158.2,
155.8, 150.2, 138.6, 116.0, 111.1, 106.4, 103.7, 91.5, 77.7, 55.9, 27.9,
15.3.
Acknowledgment. We thank National Science Foundation
and the National Institutes of Health, General Medical Sciences
(GM13598), for their generous support of our programs. F.D.T.
was a Stanford Graduate Fellow during part of his Ph.D. studies.
Mass spectra were provided by the Mass Spectrometry Regional
Center of the University of California - San Francisco,
supported by the NIH Division of Research Resources.
The identical reaction of chroman 32 (200 mg, 0.51 mmol), Pd2-
(dba)3 (25 mg, 0.024 mmol), and ethyl propynoate 2 (3 additions of 95
µL, 0.96 mmol) in formic acid (2 mL) in the absence of sodium acetate
afforded 33 (15 mg, 6% yield), 34 (100 mg, 40%), and 35 (21 mg,
8%).
33: mp 140-144 °C (lit.52a mp 143 °C). IR (film): 2928, 2856,
1
1734, 1618, 1560, 1457, 1385, 1256, 1139, 1120, 829, 825 cm-1. H
Supporting Information Available: Experimental procedures
for 6a-d, 8, 14a,b, 17a,b,h, 18a,b,h, 20, 22, 24, and 29-32
(PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
NMR (300 MHz, CDCl3): δ 7.85 (d, J ) 9.8 Hz, 1 H), 6.54 (s, 1 H),
6.17 (d, J ) 9.8 Hz, 1 H), 3.84 (s, 3 H), 2.76 (t, J ) 6.8 Hz, 2 H), 1.80
(t, J ) 6.8 Hz, 2 H), 1.35 (s, 6 H). 13C NMR (75 MHz, CDCl3): δ
161.4, 158.5, 155.3, 154.2, 138.6, 112.0, 111.6, 106.6, 100.9, 75.7,
62.0, 31.8, 26.8, 17.0.
JA0286573
(51) Bhat, H. B.; Venkataraman, K. Tetrahedron 1963, 19, 77.
(52) (a) Roberston, A.; Subramaniam, T. S. J. Chem. Soc. 1937, 286. (b)
Roberston, A.; Subramaniam, T. S. J. Chem. Soc. 1937, 1545.
(53) Wu, T. S.; Kuoh, C. S.; Furukawa, H. Phytochemistry 1983, 22, 1493.
(54) Murray, R. D. H.; Jorge, Z. D. Tetrahedron Lett. 1984.
9
4526 J. AM. CHEM. SOC. VOL. 125, NO. 15, 2003