A Rh(I)-catalyzed cycloisomerization of homo- and bis-homopropargylic alcohols

Article abstract of DOI:10.1021/ja0344258

The ability to form rhodium-vinylidene complexes in situ from terminal alkynes has led to the development of a catalytic process, the cycloisomerization of homopropargylic and bis-homopropargylic alcohols to dihydrofurans and dihydropyrans. Among the tran

Full text of DOI:10.1021/ja0344258

Published on Web 05/31/2003
A Rh(I)-Catalyzed Cycloisomerization of Homo- and Bis-homopropargylic
Alcohols
Barry M. Trost* and Young Ho Rhee
Department of Chemistry, Stanford UniVersity, Stanford, California 94305-5080
Received January 30, 2003; E-mail: bmtrost@stanford.edu
While organometallic vinylidene complexes generated from
terminal alkynes have been examined from the point of view of
preparation and properties, catalytic processes have been slow to
develop until recently.1 Only very few metals capable of forming
such complexes have led to useful catalytic processes. The first
example involved the Ru-catalyzed addition of terminal alkynes
and allyl alcohols to generate â,γ-unsaturated ketones.2 Subse-
quently, a number of other processes involving Ru catalysis have
Figure 1. List of Rh(I) complexes and phosphine ligands.
emerged.3 The formation of oxygen heterocycles mediated/catalyzed
by carbonyl complexes of molybdenum and tungsten4 as well as
catalyzed by phosphine ruthenium complexes5 illustrated in eq 1
represents a particularly useful synthetic reaction given the biologi-
addition of ligand 2a (60%) significantly improved the yield (entry
3). To explore the electronic effect of the phosphine, we employed
complex 4. Although electron-rich phosphines such as triisopro-
cal importance of many such compounds. The limitations posed
pylphosphine have been often used for stoichiometric preparation
by the existing catalysts make the discovery of new catalysts that
of Rh(I)-vinylidene complexes (Figure 1), using such ligands in
might be more chemoselective and give higher turnover desirable.
the catalytic process slowed the reaction. Placing an electron-
donating substituent on the aryl rings of triphenyl phosphine, as in
ligand 2b, also harmed the yield of the cycloisomerization (entry
6). On the other hand, a significant improvement arises by
employing electron-poor ligands 2c, 2d, and 2e (entries 7-9). In
the best case, using 55 mol % ligand 2e with precatalyst 4 (2.5
The wide utility of Rh in catalytic synthetic organic chemistry
mol %) in DMF provided the dihydrofuran in 69% isolated yield
combined with the ability of Rh to form vinylidene complexes from
(entry 9). Under this condition, only trace amounts of dimeric
compounds were detected. On the other hand, using 2f (entry 6), a
more electron-poor and smaller ligand, significantly slowed the
reaction. Addition of triethylamine (0.5-2 equiv) had no effect.
Thus, the reaction is performed under totally neutral conditions.
The optimized conditions employ pregenerated9 catalyst 3d (5%)
in the presence of the ligand 2e (30%), which provided 5 in 68%
terminal alkynes6 makes its complexes natural targets for the
evaluation for cycloisomerizations as shown in eq 1. Little is known
about catalytic reactions involving Rh(I)-vinylidene complexes as
reactive intermediates.7 In the course of our continuing study in
this field, we uncovered a cycloisomerization of homo- and bis-
homopropargylic alcohols catalyzed by Rh(I) complexes that has
excellent chemoselectivity.
yield (entry 11).10
Initial attempts to use 1 as a model substrate in the presence of
the Rh complex 3a in toluene led to the formation of dimeric as
well as unidentified oligomeric compounds.8 Surprisingly, employ-
ing a polar solvent such as DMF gave dihydrofuran 5 in moderate
yield, as shown in Table 1, entry 1. Because the dimeric compound
(∼15%) was still obtained, we reasoned that the undesirable
processes could be suppressed by excess phosphines. Indeed, the
With the optimized conditions in hand, a number of homopro-
pargylic alcohols were tested in DMF at 85 °C. As summarized in
Table 2, secondary (entries 1, 3, and 4) as well as tertiary alcohols
(entry 2) gave the products in good yields. The catalyst loading
for the cyclization is considerably lower than that of related
cycloisomerizations.4,5 In the case of substrate 8 (entry 4), only 3
mol % catalyst loading was needed for complete conversion to give
16 in 67% yield.
Under the same conditions, bis-homopropargylic alcohols are
also successfully converted into dihydropyrans. Remarkably, the
scope of the substrate is much broader than that of related metal-
catalyzed cycloisomerizations. For example, tertiary alcohol 10
(entry 6) and propargylic ether 11 (entry 7) were converted into
dihydropyrans in good yields. Unlike molybdenum-mediated cyclo-
isomerizations, a pyrrole derivative was not formed in the case of
substrate 12 (entry 8).11 Interestingly, using a less electron-poor
Table 1. Optimization of Cycloisomerization
entry
Rh complex
ligand
time (h)
conv
yielda
25%
46%
53%
51%
55%
42%
61%
1
2
3
4
5
6
7
8
9
3a (10%)
3a (10%)
3a (10%)
3a (10%)
4 (2.5%)
4 (2.5%)
4 (2.5%)
4 (2.5%)
4 (2.5%)
4 (2.5%)
3d (5%)
3d (3%)
2
2
2
2
2
2
2
2
2
4
1
2
100%
100%
100%
100%
100%
100%
100%
100%
100%
40%
2a (40%)
2a (60%)
2a (80%)
2a (55%)
2b (55%)
2c (55%)
2d (55%)
2e (55%)
2f (55%)
2e (30%)
2e (20%)
69%
75% (69%)b
29%
10
11
12
100%
83%
73% (68%)b
55%
a GC yield. b Isolated yield.
9
7482
J. AM. CHEM. SOC. 2003, 125, 7482-7483
10.1021/ja0344258 CCC: $25.00 © 2003 American Chemical Society

C O M M U N I C A T I O N S
Table 2. Syntheses of Dihydrofurans and Dihydropyrans by
Scheme 1. A Working Mechanism
Cycloisomerization
by using N-hydroxysuccinimide3b only led to cycloisomerization.
To the best of our knowledge, such Rh(I)-complexed oxacarbenes
remain unknown.14
In summary, we demonstrated that the ease of generating
rhodium-vinylidene complexes from terminal alkynes in situ can
lead to a useful catalytic process. The cycloisomerization proceeds
under nonbasic conditions. The rhodium catalysts demonstrate the
best chemoselectivity and turnover numbers to date. Thus, the
effectiveness of this new catalyst enhances the generality and
contributes to making this useful approach to five- and six-
membered ring heterocyles a powerful tool in organic synthesis.
Investigation of the mechanism as well as applications in synthesis
are underway.
Acknowledgment. We thank the National Institutes of Health
(GM 33049) for their generous support of our programs. Mass
spectra were provided by the Mass Spectrometry Facility of the
University of California-San Francisco, supported by the NIH
Division of Research Resources.
Supporting Information Available: Detailed experimental pro-
cedure and characterization data for compounds 5, 14-21, and 22
(PDF). This material is available free of charge via the Internet at http://
pubs.acs.org.
References
a Method A: catalyst 4 (2.5%), ligand 2e (55%) were used. Method B:
catalyst 4 (1.5%), ligand 2e (33%) were used. Method C: catalyst 3d (5%),
ligand 2e (30%) were used. Method D: catalyst 3c (7.5%), ligand 2d (45%)
were used. b Isolated yield.
(1) For reviews, see: Bruce, M. I.; Swincer, A. G. AdV. Organomet. Chem.
1983, 22, 59. Bruce, M. I. Chem. ReV. 1991, 91, 197. Bruneau, C.;
Dixneuf, P. H. Acc. Chem. Res. 1999, 32, 311. See also: Weyerhausen,
B.; Dotz, K. H. Eur. J. Inorg. Chem. 1999, 1057.
(2) Trost, B. M.; Kulawiec, R. J. J. Am. Chem. Soc. 1992, 114, 5579. Trost,
B. M.; Dyker, G.; Kulawiec, R. J. J. Am. Chem. 1990, 112, 7809.
(3) (a) Merlic, A. A.; Pauly, M. E. J. Am. Chem. Soc. 1996, 118, 11319. (b)
Trost, B. M.; Rhee, Y. H. J. Am. Chem. Soc. 1999, 121, 11680. (c) For
catalytic processes involving other metals, see: Ohe, K.; Yokoi, T.; Miki,
K.; Nishino, F.; Uemura, S. J. Am. Chem. Soc. 2002, 124, 526. Miura,
T.; Iwasawa, N. J. Am. Chem. Soc. 2002, 124, 518.
(4) (a) For a review, see: McDonald, F. E. Chem.-Eur. J. 1999, 5, 3103. (b)
McDonald, F. E.; Reddy, K. S.; Diaz, Y. J. Am. Chem. Soc. 2000, 122,
4304.
ligand 2d significantly improved the yield for propargylic ethers
11 and 12 (method C vs method D for entries 7, 8).
To illustrate the synthetic utility of the Rh-catalyzed process,
the glycosylation of the dihydroyran 20 which cannot be made by
the other catalytic systems was pursued as an approach to amino
sugars. Methylation followed by Ph3P‚HBr-catalyzed glycosylation
gives â-anomer 22 predominantly.12 This type of deoxyaminogly-
coside is present in several bioactive natural products, such as
kedarcidin.13
(5) Trost, B. M.; Rhee, Y. H. J. Am. Chem. Soc. 2002, 124, 2528.
(6) (a) Wolf, J.; Werner, H.; Serhadli, O.; Ziegler, M. L. Angew. Chem., Int.
Ed. Engl. 1983, 22, 414. (b) Alonso, F. J.; Hohn, A.; Wolf, J.; Otto, H.;
Werner, H. Angew. Chem., Int. Ed. Engl. 1985, 24, 406.
(7) Ohmura, T.; Yamamoto, Y.; Miyaura, N. J. Am. Chem. Soc. 2000, 122,
4990.
(8) For Rh(I)-catalyzed dimerization of terminal alkynes, see: Schmit, H. J.;
Singer, H. J. Organomet. Chem. 1978, 153, 165.
(9) Richter, B. D.; Spek, A. L.; van Koten, G.; Deelman, B. J. J. Am. Chem.
Soc. 2000, 122, 3945.
(10) Employing bidentate ligands (BINAP, dppf) resulted in significant
conversion problems.
(11) McDonald, F. E.; Zhu, Y. H. Tetrahedron 1997, 53, 11061.
(12) Bolitt, V.; Mioskowski, C.; Lee, S.-G.; Falck, J. R. J. Org. Chem. 1990,
55, 5812.
Although the exact mechanism awaits further study, we suggest
Scheme 1 as a working hypothesis. It appears that the protonation
of intermediate I occurs exclusively at the metal to liberate the
product and that the Rh-oxacarbene complex II is not formed in
the cycloisomerization. In fact, all attempts to generate lactones
(13) Myers, A. G.; Hogan, P. C.; Hurd, A. R. Goldberg, S. D. Angew. Chem.,
Int. Ed. 2002, 41, 1062.
(14) Stang, P. J.; Huang, Y. H. J. Organomet. Chem. 1992, 432, 247.
JA0344258
9
J. AM. CHEM. SOC. VOL. 125, NO. 25, 2003 7483