Anion-accelerated palladium-catalyzed intramolecular coupling of phenols with aryl halides

Full text of DOI:10.1021/jo961876k

2
J . Org. Chem. 1997, 62, 2-3
Communications
An ion -Acceler a ted P a lla d iu m -Ca ta lyzed
In tr a m olecu la r Cou p lin g of P h en ols w ith
Ar yl Ha lid es
Sch em e 1
D. David Hennings, Seiji Iwasa, and
Viresh H. Rawal*
Department of Chemistry, University of Chicago,
Chicago, Illinois 60637
Received October 4, 1996
The aryl-aryl linkage is found in a wide range of com-
pounds, from natural products such as vancomycin1 and
aporphine alkaloids2 to important unnatural compounds
such as 1,1′-bi-2-naphthol3 and BINAP.4 Although sev-
eral procedures have been developed for the coupling of
aryl moieties, many of them have limited applicability
due to the harshness or functional group intolerance of
the conditions required.5 In connection with our interest
in asymmetric synthesis of biaryls,6 we have examined
several different procedures for constructing the aryl-
aryl linkage. We report here a mild, selective procedure
for the palladium-catayzed intramolecular coupling of
phenols with aryl halides.
The coupling of carbon nucleophiles with aryl halides
has been accomplished with palladium catalysis.7 The
reaction is believed to proceed via attack of the nucleo-
phile to a σ-bound palladium species, followed by reduc-
tive elimination of palladium to give the coupled product.8
Carbon nucleophiles that have been used successfully
include enolates, particularly those derived from â-di-
carbonyl compounds.9,10 Aromatic rings are also effective
as nucleophiles, although high temperatures are gener-
ally required for the reaction. We have found, however,
that when the aryl unit has a hydroxyl group attached,
then the ambident nature of the corresponding phenolate
anion renders the aryl ring more electron-rich and, hence,
more reactive in the coupling reaction.11
with t-BuOK (3 equiv) and Pd(PPh3)4 (0.2 equiv) in di-
methylacetamide (DMA) at 95 °C (bath), bromide 1
smoothly cyclized to predominently the “ortho” product
2, isolated in 87% yield. The reaction is expected to
proceed by the pathway shown in Scheme 1. Oxidative
addition of palladium to the aryl bromide would give
σ-aryl palladium intermediate 4. Nucleophilic attack of
the phenolate on the palladium would yield, after tau-
tomerization, diaryl palladium species 5, which on reduc-
tive elimination of the palladium would give the observed
product.
Experiments were performed to exclude the possibility
of a benzyne intermediate and to confirm that the
cyclization is palladium catalyzed (Table 1). When the
above reaction was carried out using the weaker base K2-
CO3, which is unlikely to produce benzyne under the
conditions used, the same results were obtained (entry
2). With t-BuOK as the base and with no palladium
catalyst, the starting material was slowly consumed, but
gave none of the cyclized product, thus ruling out the
benzyne mechanism (entry 3). The cyclization also did
not take place if base was omitted from the reaction
mixture, supporting the notion that the reaction is
accelerated using the phenolate as the nucleophile (entry
4). DMA was the solvent of choice for this transforma-
tion. In other solvents that were examined (e.g., DME,
THF, CH3CN, toluene), the yield of the cyclized product
was lower and included a higher percentage of the “para”
product 3. Interestingly, reduction of the halide, to afford
3-(benzyloxy)phenol (7), was by far the major pathway
when the reaction was carried out in DMF (entry 5).12
Of the other commonly used palladium catalysts that
were examined (e.g., Pd(OAc)2, Pd2(dba)3, PdCl2), none
were as effective as Pd(PPh3)4 for this cyclization. Her-
rmann recently reported a robust, discrete palladacycle
The initial studies were carried out on bromide 1
using Pd(PPh3)4 as the catalyst (eq 1). On treatment
(Eq 1)
(1) Rama Rao, A. V.; Gurjar, M. K.; Reddy, L.; Srinivasa Rao, A.
Chem. Rev. 1995, 95, 2135-2167.
(9) Cuifolini, M. A.; Browne, M. E. Tetrahedron Lett. 1987, 28, 171-
174.
(10) Fournet, G.; Balme, G.; Van Hemelryck, B.; Gore, J . Tetrahe-
dron Lett. 1990, 31, 5147-5150.
(11) Wiegand, S.; Schafer, H. J . Tetrahedron 1995, 51, 5341-5350.
(12) Zask, A.; Helquist, P. J . Org. Chem. 1978, 43, 1619-1620. The
authors have reported the reduction of aryl bromides in DMF in the
presence of Pd(PPh3)4 and NaOCH3. They proposed that the reducing
hydrogen arises from palladium-mediated decomposition of the meth-
oxide to formaldehyde, a process which cannot take place in the present
example. It is likely that in the present case saponification of DMF
yields a formate ion, the decomposition of which to PdH, which provides
the reducing hydrogen, and CO2 has ample precedent.
(2) The Alkaloids; Brossi, A., Ed.; Academic Press: Orlando, 1985;
Vol. 24.
(3) Rosini, C.; Franzini, L.; Raffaelli, A.; Salvadori, P. Synthesis
1992, 503-517.
(4) Noyori, R.; Takaya, H. Acc. Chem. Res. 1990, 23, 345-350.
(5) Ames, D. E.; Opalko, A. Tetrahedron 1984, 40, 1919-1925.
(6) Rawal, V. H.; Florjancic, A. S.; Singh, S. P. Tetrahedron Lett.
1994, 35, 8985-8988.
(7) Uno, M.; Seto, K.; Takahashi, S. J . Chem. Soc., Chem. Commun.
1984, 932-933.
(8) Cuifolini, M. A.; Qi, H. B.; Browne, M. E. J . Org. Chem. 1988,
53, 4149-4151.
S0022-3263(96)01876-2 CCC: $14.00 © 1997 American Chemical Society

Communications
J . Org. Chem., Vol. 62, No. 1, 1997
3
Ta ble 1
entry
substrate
base (3 equiv)
catalyst (mol %)
solvent
temp (°C)
time (days)
yield cyclized product
1
2
3
4
5
6
7
8
1
1
1
1
1
9
10
9
t-BuOK
K2CO3
t-BuOK
none
t-BuOK
Cs2CO3
Cs2CO3
Cs2CO3
Pd(PPh3)4 (20)
Pd(PPh3)4 (10)
none
Pd(PPh3)4 (10)
Pd(PPh3)4 (20)
HC, 8, (5)
HC, 8, (5)
Pd(OAc)2 (5)
(o-tolyl)3P (5)
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMA
95
95
95
95
95
80
80
84
2
2
2
2
2
1
1
1
87a
88a
0 (56% recovered 1)
0 (77% recovered 1)
0 (77% reduction product 7)
97
10 (90% recovered 10)
94
a
A small amount (<5%) of the para product 3 was also observed.
Ta ble 2a
(8),13 formed in high yield from Pd(OAc)2 and P(o-tolyl)3.
Herrmann’s catalyst (HC) has been shown to efficiently
catalyze the Heck14 and Suzuki15 cross-coupling reactions.
We have found HC to be as good or better than other
catalysts and have utilized it for all subsequent cycliza-
tions (Table 1). Unlike Pd(PPh3)4, HC can be stored and
handled without taking special precautions. The cycliza-
tion of iodide 913 proceeded in high yield at 75-80 °C
using just 5 mol % of HC (entry 6). The reaction was
faster with Cs2CO3 than with K2CO3 as the base.16
Whereas bromide 1 and iodide 9 reacted at comparable
rates,17 the corresponding aryl chloride was much less
reactive under the same conditions. It should be noted
that other than being convenient to use there appears to
be nothing special about using preformed catalyst 8 for
this reaction. Subjection of iodide 9 to the standard
reaction conditions using a 1:1 ratio of Pd(OAc)2 and P(o-
tolyl)3 gave the cyclization product in about the same
yield as using the preformed catalyst. This observation
differs from Herrmann’s, who found the preformed
catalyst to be “more active and productive than a
conventional ‘in situ catalyst’” for the Heck reaction.14
The main disadvantage with the in situ catalyst is that
the 1:1 stoichiometry of the two components needs to be
maintained. The reaction was found to be significantly
slower when Pd(OAc)2 and the phosphine were used in a
1:2 ratio. The accelerating effect of the phenolate anion
is evident from entry 7. Under the same conditions used
for the cyclization of phenol 9, the corresponding methyl
ether 1013 gave primarily recovered starting material,
along with the “para” cyclized product in 10% yield.
This anion-accelerated cyclization procedure using HC
appears to be general (Table 2). When the ortho position
is blocked, as in 11, the substrate was forced to cyclize
at the “para” position to yield 12. Cyclization of naphthyl
bromide 13 required longer reaction time but proceeded
in excellent yield to afford a mixture of the “ortho” and
“para” products. An additional heteroatom on the phe-
nolic component is not necessary for successful cycliza-
tion, although it may accelerate the reaction. The deoxy-
analog of 1, dihydrostilbenoid 16, required a higher
a
Conditions: All reactions done with 5 mol % of palladacycle
b
and 3 equiv of Cs2CO3 in DMA. Based on recovered starting
material.
temperature for the cyclization and gave the correspond-
ing dihydrophenanthranol, 17, in 75% yield (96% based
on recovered starting material), with good “ortho” selec-
tivity. The ability to use nitrogen-containing compounds
opens up this methodology to applications in alkaloid
synthesis. The cyclization of iodide 18 gave the “ortho”
cyclization product, dihydrophenanthridine derivative 19.
None of the cyclized product was isolated from reaction
of the corresponding free amine of 18. Presumably, the
intermediate aryl palladium species is tied up through
coordination with the amine.
In conclusion, the intramolecular coupling of phenols
with aryl halides is efficiently promoted under basic con-
ditions with palladium catalysts, particularly Herrmann’s
catalyst. Extension of this coupling protocol to natural
product synthesis is currently under investigation.
(13)
Ack n ow led gm en t. We thank the University of
Chicago for partial support of this work. Generous
financial support, in the form of faculty awards, from
Pfizer and Merck are also gratefully acknowledged.
(14) Herrmann, W. A.; Brossmer, C.; Ofele, K.; Reisinger, C. P.;
Priermeier, T.; Beller, M.; Fischer, H. Angew. Chem., Int. Ed. Engl.
1995, 34, 1844-1848.
(15) Beller, M.; Fischer, H.; Herrmann, W. A.; Ofele, K.; Brossmer,
C. Angew. Chem., Int. Ed. Engl. 1995, 34, 1848-1849.
(16) Little or no cyclization product was observed with Na2CO3 or
Li2CO3. Presumably, the larger cations are better solvated, resulting
in a more nucleophilic, “naked” phenolate anion.
(17) On the basis of the available data, it seems that the rate-
determining step for aryl bromides and iodides is not the oxidative
addition but the nucleophilic attack of the phenolate on the transient
palladium species 3.
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedure for the cyclization of compound 1, and spectral data,
including copies of 1H NMR and 13C NMR spectra, for
compounds 1-3 and 9-21 (35 pages).
J O961876K