A Ru catalyzed divergence: Oxidative cyclization vs cycloisomerization of bis-homopropargylic alcohols

Article abstract of DOI:10.1021/ja011840w

During the course of investigating the development of catalytic reactions involving ruthenium vinylidene intermediates, a novel divergence of reactivity was discovered. The oxidative cyclization of bis-homopropargylic alcohols with Ru(+2) complexes as catalysts and N-hydroxysuccinimide as oxidant, which requires formation of a ruthenium vinylidene intermediate, is complicated by the simple electrophilically initiated direct attack of the hydroxyl group on π-complex of the alkyne and ruthenium. A catalytic system composed of CpRu[(p-CH₃O₆H₄)₃P]₂Cl and excess (p-CH₃O-C₆H₄)₃P directs the reaction toward the oxidative cyclization to form δ-lactones in good yields. Significantly, a simple switch of catalyst to CpRu[(p-FC₆H₄)₃P]₂Cl redirects the reaction to a cycloisomerization to form dihydropyrans in good yields. The synthetic utility of the oxidative cyclization is illustrated by the synthesis of oviposition attractant pheromone of the mosquito Culex pipens. The utility of the cycloisomerization to dihydropyrans is demonstrated by an iterative process leading to the antiviral agent narbosine B. A rationale for this dramatic switch by simple ligand modification is proposed.

Full text of DOI:10.1021/ja011840w

Published on Web 02/22/2002
A Ru Catalyzed Divergence: Oxidative Cyclization vs
Cycloisomerization of Bis-homopropargylic Alcohols
Barry M. Trost* and Young Ho Rhee
Contribution from the Department of Chemistry, Stanford UniVersity,
Stanford, California 94305-5080
Received July 30, 2001
Abstract: During the course of investigating the development of catalytic reactions involving ruthenium
vinylidene intermediates, a novel divergence of reactivity was discovered. The oxidative cyclization of bis-
homopropargylic alcohols with Ru(+2) complexes as catalysts and N-hydroxysuccinimide as oxidant, which
requires formation of a ruthenium vinylidene intermediate, is complicated by the simple electrophilically
initiated direct attack of the hydroxyl group on a π-complex of the alkyne and ruthenium. A catalytic system
composed of CpRu[(p-CH3O6H4)3P]2Cl and excess (p-CH3O-C6H4)3P directs the reaction toward the
oxidative cyclization to form δ-lactones in good yields. Significantly, a simple switch of catalyst to CpRu-
[(p-FC6H4)3P]2Cl redirects the reaction to a cycloisomerization to form dihydropyrans in good yields. The
synthetic utility of the oxidative cyclization is illustrated by the synthesis of oviposition attractant pheromone
of the mosquito Culex pipens. The utility of the cycloisomerization to dihydropyrans is demonstrated by an
iterative process leading to the antiviral agent narbosine B. A rationale for this dramatic switch by simple
ligand modification is proposed.
I. Introduction
Although these complexes are reported to be inert to chemical
reactions,6 our continuing interest in this area and the potential
synthetic utility of oxygen-heterocycles prompted us to inves-
tigate catalytic reactions by using such oxacarbene complexes
as reactive intermediates. Recently, we successfully developed
a novel oxidative cyclization of homopropargylic alcohols to
form γ-butyrolactones, using N-hydroxysuccinimide as a mild
oxidant.7 In an effort to extend the scope of the reaction, we
turned our attention to bis-homopropargylic alcohols.
Organometal-vinylidene complexes have been widely stud-
ied for decades.1 The facility with which such species can be
generated by direct reaction of terminal alkynes makes these
species atom economical intermediates for catalytic transforma-
tions if the metal can be used catalytically. The difficulty in
achieving this objective is suggested by the fact that such species
have been known, but catalytic reactions remain almost
unknown. Some years ago, we initiated programs toward
developing catalytic reactions involving ruthenium vinylidene
complexes as reactive intermediates and reported a ruthenium-
catalyzed two-component coupling of allyl alcohols and terminal
alkynes to form â,γ-unsaturated ketones invoking such com-
plexes as the key reactive intermediates.2,3 In these studies we
noted the electrophilic nature of such ruthenium-vinylidene
complexes.
An interesting series of molybdenum- and tungsten-mediated
cycloisomerizations of homopropargylic and bis-homopropar-
gylic alcohols to dihydrofurans and dihydropyrans has been
reported by McDonald et al.8 In these studies, bis-homopro-
pargylic alcohols are converted into dihydropyrans generally
in moderate yield.8a,b Recent studies from the McDonald group
show much improved results in terms of yield, but typically
require photolysis at elevated temperatures as well as high
catalyst loading (∼25%).8c,d In considering the oxidative cy-
clization of bis-homopropargylic alcohols to valerolactones, we
Because of this electrophilic nature, ruthenium-vinylidene
complexes generated from homopropargylic and bis-homopro-
pargylic alcohols form stable cyclic oxacarbene complexes.4,5
(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.
(5) For a review on the synthesis and chemistry of cyclic metal-complexed
oxacarbene complexes, see: Weyerhausen, B.; Dotz, K. H. Eur. J. Inorg.
Chem. 1999, 1057. For a review on reactions that use such a concept, see:
McDonald, F. E. Chem. Eur. J. 1999, 5, 3103.
(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. Soc. 1990, 112, 7809.
(3) For other examples of catalytic reactions of terminal acetylenes via
vinylidene intermediates, see: Mahe´, R.; Sasaki, Y.; Bruneau, C.; Dixneuf,
P. H. J. Org. Chem. 1989, 54, 1518. Wakatsuki, Y.; Yamazaki, H.;
Kumegawa, N.; Satoh, T.; Satoh, J. Y. J. Am. Chem. Soc. 1991, 113, 9604.
Murakami, M.; Ubukata, M.; Ito, Y. Tetrahedron Lett. 1998, 39, 7361.
Merlic, C. A.; Pauly, M. E. J. Am. Chem. Soc. 1996, 118, 11319. Gemel,
C.; Trimmel, G.; Slugovc, C.; Kremel, S.; Mereiter, K.; Schmid, R.;
Kirchner, K. Organometallics 1996, 15, 3998.
(6) Davies, G.; McNally, J. P.; Smallridge, A. J. AdV. Organomet. Chem. 1990,
30, 1.
(7) Trost, B. M.; Rhee, Y. H. J. Am. Chem. Soc. 1999, 121, 11680.
(8) (a) For example, see: McDonald, F. E.; Bowman, J. L. Tetrahedron Lett.
1996, 37, 4675. (b) McDonald, F. E.; Zhu, H. Y. H. J. Am. Chem. Soc.
1998, 120, 4246. (c) McDonald, F. E.; Reddy, K. S.; Diaz, Y. J. Am. Chem.
Soc. 2000, 122. 4304. (d) While this manuscript was in preparation, a new
cycloisomerization process was reported which needs 10% tungsten catalyst.
Although this result represents considerable improvement in terms of
catalyst loading, the scope of such a process has yet to be determined:
McDonald, F. E.; Reddy, K. S. J. Organomet. Chem. 2001, 617, 444.
(4) Bruce, M. I.; Swincer, A. G.; Thomson, B. J.; Wallis, R. C. Aust. J. Chem.
1980, 33, 2605.
9
2528 VOL. 124, NO. 11, 2002 J. AM. CHEM. SOC.
10.1021/ja011840w CCC: $22.00 © 2002 American Chemical Society

Ru Catalyzed Divergence
A R T I C L E S
Table 1. Optimization of Oxidative Cyclization and Cycloisomerizationa
entry
Ru complex (mol %)
ligand (mol %)
4(mol %)
time (h)
conversion (%)
yield 6a (isolated yield)
yield 7b (isolated yield)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
9a (10%)
9a (10%)
9a (10%)
9b (10%)
9c (10%)
9d (10%)
9e (10%)
9f (10%)
9g (10%)
9c (10%)
9c (10%)
9c (10%)
9c (5%)
2a (20%)
2a (40%)
2a (60%)
2b (40%)
2c (40%)
2d (40%)
2e (40%)
2f (40%)
2g (40%)
2c (40%)
2c (40%)
2c (40%)
2c (20%)
2e (40%)
2e (30%)
2e (20%)
2e (20%)
300c
300c
300c
300c
300 c
300c
300c
300 c
300c
450c
600c
1000c
600d
200 d
100d
50d
15
17
25
20
17
17
17
26
26
20
23
28
28
17
20
25
25
100
100
80
100
100
100
100
91
20%
25%
17%
33%
31%
24%
38% (34%)
60% (57%)
51%
19% (16%)
13%
40 (35%)
20% (17%)
23%
57% (52%)
56%
53%
(11%)
(7%)
(trace)
(6%)
(60%)
(61%)
(64%)
(53%)
81
9%
100
100
46
(65%)
(69%)
(30%)
(45%)
(11%)
(7%)
73
9e (10%)
9e (7.5%)
9e (5%)
100
100
99
(4%)
(5%)
9e (5%)
25d
84
a All reactions were run at 0.4 M bis-homopropargylic alcohol in DMF at 85 °C with 200 mol % NaHCO3 unless otherwise noted. b Yield determined by
gas chromatography with n-tetradecane as an internal standard, except where indicated it was an isolated yield. c N-Hydroxysuccinimide used. d Preformed
sodium salt of N-hydroxysuccinimide used, and sodium bicarbonate was not added.
uncovered an unusual flexibility to form either valerolactones
(eq 1, path a) or dihydropyrans (eq 1, path b) by simple ligand
Although removing water eliminated the formation of methyl
ketone 8, the combined yield of 6 and 7 was not much affected.
Presumably, path a is still occurring but the resultant enol ether
simply decomposes in the absence of water. On the other hand,
increasing the amount of phosphine 2a to 40 mol % under the
anhydrous condition had a significant increase in the combined
amount of 6 and 7. As a result we switched to use of complexes
of type 9.16,17
variation with the ruthenium-catalyzed process.
II. Results
A. Optimization of Reaction Conditions. Extrapolation of
the oxidative cyclization to six-membered rings was not
straightforward. Ruthenium (+2) complexes are known to
initiate addition of oxygen nucleophiles to alkynes.9 Homopro-
pargylic alcohols would not be so prone to such a competing
event, because it would require a 4-exo transition state. On the
other hand, bis-homopropargylic alcohols should be able to
undergo the related 5-exo cyclization more readily.10 Indeed,
this fear was justified as illustrated in the preliminary attempt
shown in eq 2. By using the conditions optimized for formation
Table 1 summarizes some of our efforts to optimize formation
of the six-membered ring products. Increasing the amount of
excess ligand to 60 mol % (entry 3) slowed the reaction
of butyrolactones, 10 mol % CpRu(COD)Cl (1), 15 mol % tris
(2-furyl)phosphine (2a), 30 mol % tetra-n-butylammonium
hexafluorophosphate (3), 3 equiv of N-hydroxysuccinimide (4),
and 2 equiv of sodium bicarbonate in 7:1 dimethylformamide
(DMF)/water at 85 °C gave an approximately coequal mixture
of lactone 6,11 dihydropyran 7,12 and ketone 813 from bis-
homopropargylic alcohol 5.14 The latter presumably arises via
5-exo cyclization of the alcohol onto the alkyne followed by
hydrolysis (eq 3, path a);15 whereas, the two former compounds,
(9) (a) Mitsudo, T.; Hori, Y.; Yamakawa, Y.; Watanabe, Y. J. Org. Chem.
1987, 52, 2230. (b) Neveux, M. I.; Seiller, B.; Hagedorn, F.; Bruneau, C.;
Dixneuf, P. H. J. Organomet. Chem. 1993, 451, 133.
(10) (a) For such a pathway in a Pd-catalyzed cyclization, see: Wang, Z.; Lu,
X. J. Org. Chem. 1996, 61, 2254. (b) This exo-mode cyclization has been
addressed in tungsten-catalyzed cycloisomerization of bis-homopropargylic
alcohols. See ref 8c.
(11) Mino, T.; Masuda, S.; Nishio, M.; Yamashita, M. J. Org. Chem. 1997, 62,
2633.
(12) All new compounds were fully characterized.
(13) Vaskan, R. N.; Kovalev, B. G. Zh. Org. Khim. 1975, 11, 1818.
(14) Yamaguchi, M.; Nobayashi, Y.; Hirao, I. Tetrahedron Lett. 1983, 24, 5121.
(15) This result could be alternatively explained by simple hydration of
ruthenium-alkyne complex with water. However, this possibility was
excluded by the experimental result with 1-octyne, which did not produce
the corresponding ketone under the same reaction condition.
(16) Complex 1 forms catalysts 9 in the presence of excess amount of the
corresponding phosphines 2. In fact, the use of 10% complex 1 with 60%
ligand 2a gave a result similar to entry 1 (Table 1). However, because of
the low yield in the preparation of complex 1, the use of in situ-generated
catalyst 9 was not investigated further. For the synthesis and the use of
complex 1, see: Albers, M. O.; Robinson, D. J.; Shaver, A.; Singleton, E.
Organometallics 1986, 5, 2199.
6 and 7, arise via the desired vinylidene complexes (eq 3, path
b).
(17) For the synthesis of these types of Ru complexes, see: Hartwig, J. F.;
Bhandari, S.; Rablen, P. R. J. Am. Chem. Soc. 1994, 116, 1839.
9
J. AM. CHEM. SOC. VOL. 124, NO. 11, 2002 2529

A R T I C L E S
Trost and Rhee
Table 2. Examples
a Method A: 10 mol % 9c, 40 mol % 2c, 30 mol % 3, 6 equiv of 4, 2 equiv of NaHCO3. Method B: 15mol % 9c, 60 mol % 2c, 45 mol % 3, 6 equiv
of 4, 2 equiv of NaHCO3. Method C: 5 mol % 9e, 20 mol % 2e, 15 mol % 3, 50 mol % 4 sodium salt. Method D: 7.5 mol % 9e, 30 mol % 2e, 23 mol
% 3, 1 equiv of 4 sodium salt. Method E: 10 mol % 9e, 40 mol % 2e, 30 mol % 3, 2 equiv of 4 sodium salt. b p-TsNH2 was obtained (∼10%). c p-TsNH2
was obtained (a trace amount). d Yield based on recovered starting material.
considerably. A significant improvement arises by switching
to triphenylphosphine as the ligand (entry 4). Placing an
electron-donating substituent18 in the para position not only
retained good yields but tilted the ratio to favor the lactone 6
(entries 5 and 6). Conversely, electron-withdrawing substituents
in the para (entries 7 and 8) or meta (entry 9) position18 favored
the formation of the dihydropyran 7. A further increase in the
amount of lactone is observed by increasing the amount of
N-hydroxysuccinimide (entries 10 and 11) but too much has a
strong negative impact on conversion (entry 12). Decreasing
the amount of catalyst maintained the same 6:7 ratio but with
incomplete conversion (entry 13). On the other hand, decreasing
the amount of N-hydroxysuccinimide (used as its preformed
sodium salt) increased the amount of dihydropyran compared
with lactone while maintaining good yields (entry 14). Signifi-
cantly, decreasing the amount of catalyst as well as sodium salt
of 3 maintained good yields and 6:7 ratios (entries 15-17).
Thus, we adopted the conditions of entry 10 for the synthesis
of lactones and those of entry 16 for dihydropyrans.
B. Scope and Limitation. With the optimized conditions in
hand, we tested several bis-homopropargylic alcohols to deter-
mine the scope and limitation of the reaction. As summarized
in Table 2, formation of dihydropyrans generally required lower
catalyst loading than those for lactone formation, as low as 5
mol % for complete conversion. In entry 6c vs 6d, an increased
catalyst loading from 5 mol % to 7.5 mol % did increase the
conversion from 94% to quantitative, but the isolated yield of
dihydropyran increased very little. On the contrary, yields of
lactones almost invariably increased by running the reactions
with 15 mol % compared to 10 mol % (entries 2a, 3a and 6a vs
2b, 3b, and 6b). The reactions show excellent chemoselectivity.
In entry 6, only the six-membered ring products formed. The
(18) For quantitative data for electron density of phosphine ligands, see: Tolman,
C. A. Chem. ReV. 1977, 77, 313.
9
2530 J. AM. CHEM. SOC. VOL. 124, NO. 11, 2002

Ru Catalyzed Divergence
A R T I C L E S
Scheme 1. A Synthesis of Narbosine B
a See Table 2, entry 3c. b (i) TsOH, PhCH3, r.t., (ii) TBAF, THF, r.t. c Method D, Table 2. d CSA, CH3OH, r.t. e DDQ, CH2Cl2, H2O, r.t.
yield for entries 2c and 3c are considerably higher than the only
ones reported for a related process that uses stoichiometric
amounts of tungsten to cycloisomerize to dihydropyrans.8b
Substrates bearing heteroatom substituent at propargylic
position such as 26 are particularly interesting because these
types of compounds form allenylidene complexes with the
concomitant elimination of propargylic substituents.23 In fact,
propargylic alcohol derivatives 25a-25c failed to produce the
corresponding dihydropyrans and valerolactones under both
conditions (eq 4). In all tests, only unreacted starting materials
D-amicetose24 and D-rhodinose,25 which are constituents of more
complex natural products. Compounds 17, 18, 20, and 21 can
serve in the construction of polyethers represented by the
brevetoxins, ciguatoxins, dactomelynes, maitotoxins, etc.20 The
synthesis of compound 27 gives a novel access to aminosugars.26
The oviposition attractant pheromone of the mosquito Culex
pipens fatigans 3027 was synthesized from the lactone 23 as
shown in eq 5 as demonstrated previously.21,28a The absolute
stereochemistry derived from an asymmetric dihydroxylation
of an E-olefin precursor and the ee was established as 93% by
the comparison of optical rotation for both 23 and the natural
product itself.28b Analogous to the work of McDonald et al., an
iterative process can be performed.8b Thus, a synthesis of
narbosine B, an antiviral secondary metabolite from Strepto-
myces,29 was performed (see Scheme 1). Compound 32 formed
as a single stereoisomer and underwent cycloisomerization to
glycal 33 under our standard protocol. Acid-catalyzed addition
of methanol to this compound produced acetal 34 in 64% yield
accompanied by 9% of the epimer. For analytical purposes, the
major isomer was separated and reacted with DDQ to provide
35 as a single anomer, whereas previously it was only isolated
as an anomeric mixture.29
were recovered. In substrate 25b benzyl alcohol was isolated
(∼10% yield). Unlike 25, substrate 26 provided the desired
dihydropyran 27 in moderate yield under cycloisomerization;
whereas it remained unreactive under oxidative cyclization. In
both cases, p-toluenesulfonamide was obtained (∼10% for entry
7a, a trace amount for entry 7b). Another interesting disparity
is seen with tertiary alcohol 28. Although oxidative cyclization
failed to produce the corresponding lactone, cycloisomerization
with 10% catalyst produced the dihydropyran 29 in modest
yield. This result is surprising because tertiary alcohols previ-
ously proved to be the best substrates for oxidative cyclization
of homopropargylic alcohols.7
(24) For some recent syntheses, see: Noecker, L. A.; Martino, J. A.; Foley, P.
J.; Rush, D. M.; Giuliano, R. M.; Villani, F. J., Jr. Tetrahedron: Asymmetry
1998, 9, 203. Pedersen, C.; Jensen, H. S. Acta Chem. Scand. 1994, 48,
222. Lajsic, S.; Miljkovic, D.; Cetkovic, G. Carbohydr. Res. 1992, 233,
261.
Several of the examples already provide interesting and useful
structural motifs. For example, dihydropyrans 12 and 15 or their
lactones 11 and 14 can be converted to the deoxysugars
(25) For more recent syntheses, see: Rohr, J.; Wohlert, S.-E.; Oelkers, C.;
Kirschning, A.; Ries, M. Chem. Commun. 1997, 973. Sobti, A.; Sulikowski,
G. A. Tetrahedron Lett. 1995, 36, 4193.
(26) For a review on the synthesis of aminosugars, see: Hauser, F. M.;
Ellenberger, S. R. Chem. ReV. 1986, 86, 35.
(27) (a) Lin, G.; Xu, H.; Wu, B.; Guo, G.; Zhou, W. Tetrahedron Lett. 1985,
26, 1233. (b) Machiya, K.; Ichimoto, I.; Kirihata, M.; Ueda, H. Agric. Biol.
Chem. 1985, 49, 643.
(28) (a) Kotuski, H.; Kadota, I.; Ochi, M. J. Org. Chem. 1990, 55, 4417. (b)
Eliel, E. L.; Ko, K.-Y. J. Org. Chem. 1986, 51, 5353.
(29) Henkel, T.; Breiding-Mack, S.; Zeeck, A.; Grabley, S.; Hammann, P. E.;
Hutter, K.; Till, G.; Thiericke, R.; Wink, J. Liebigs Ann. Chem. 1991, 575.
(19) Leeuwenburgh, M. A.; Litjens, R. E. J. N.; Codee, J. D. C.; Overkleeft, H.
S.; van der Marel, G. A.; van Boom, J. H. Org. Lett. 2000, 2, 1275.
(20) Van Boom, J. H.; Leeuwenburgh, M. A.; Overkleeft, H. S.; van der Marel,
G. A. Synlett 1997, 1263.
(21) Depezay, J.-C.; Gravier-Pelletier, C.; Le Merrer, Y. Tetrahedron 1995, 51,
1663.
(22) Jefford, C. W.; Jaggi, D.; Boukouvalas, J. Tetrahedron Lett. 1986, 27, 4011.
(23) (a) For a general review on the formation of allenylidene complexes, see:
Bruce, M, I. Chem. ReV. 1991, 91, 197. (b) For a catalytic reaction using
such complexes as reactive intermediates, see: Trost, B. M.; Flygare, J. A.
J. Am. Chem. Soc. 1992, 114, 5476.
9
J. AM. CHEM. SOC. VOL. 124, NO. 11, 2002 2531

A R T I C L E S
Trost and Rhee
Scheme 2. Mechanistic Rationale
III. Mechanism
The reactions of bis-homopropargylic alcohols show striking
differences from those of homopropargylic alcohols. Ru(II)
activates the trans-addition of oxygen nucleophile to alkynes.9
In this pathway ruthenium simply coordinates to alkynes and
promotes the addition of the nucleophile, as mentioned previ-
ously. Phosphine in excess to Ru is needed to divert the reaction
from such reactions to the formation of the key vinylidene
complexes.
Scheme 2 rationalizes the divergent behavior. Cycle A is
favored by more electron-rich ligands and larger excess of
ligands; whereas, cycle B is favored by less electron-rich ligands
and lower amounts of ligands. Thus, a ligand possessing suitable
electronic properties to facilitate formation of the pivotal
intermediate I is needed. If it is electron rich enough, it can
promote protonation at carbon to form the oxacarbene complex
II leading to lactone. On the other hand, if it is less electron
rich, ligand exchange to form an anionic complex III may occur
and allow simple protonation of the C-bound ligand to liberate
the dihydropyran. The formation of dihydropyran looks par-
ticularly intriguing because this transformation has been inves-
tigated by others with no success in the presence of ruthenium
complexes and conventional (noncoordinating) bases.30 Our
extensive effort to substitute other bases for N-hydroxysuccin-
imide also failed to give any dihydropyrans.31 This result indeed
supports the mechanistic rationale; simple protonation of
complex I occurs only at the carbon to generate complex II.
Therefore, the use of coordinating bases to promote ligand
exchange in the cycloisomerization pathway seems to be crucial.
In substrates 25 and 26, the elimination of propargylic
substituents turned out to be very facile. As depicted in eq 6,
such elimination could be explained either by the formation of
allenylidene complexes a (pathway a, eq 6), or the formation
of plausible intermediate c after the formation of complex b
(pathway b, eq 6).32 In the oxidative cyclization, such elimina-
tion will be promoted by excess protonic acid (N-hydroxysuc-
cinimide). In the cycloisomerization, on the other hand, the
leaving group capability of the propargylic substituent plays a
role. In substrate 25, the elimination still predominates because
of the presence of better leaving groups. Presumably, such a
process is promoted by the coordination of Lewis acidic
ruthenium complexes. In 26, the desired cycloisomerization
overrides the elimination because of the presence of a poorer
leaving group (which is also a poor Lewis base).
The disparity with tertiary alcohol 28 can be explained by
the activating effect of electron-poor phosphine ligand 2e over
2c in the formation of complex I (Scheme 2), as well as the
steric interaction in forming intermediate (e) in oxidative
cyclization, as described in eq 7. This type of steric interaction
(30) For example, see 6: Also see, McDonald, F. E.; Bowman, J. L. J. Org.
Chem. 1998, 63, 3680.
(31) Examples include organic bases (triethylamine, 2,6-di-tert-butylamine,
proton sponge) and inorganic bases (NaHCO3, Na2CO3, sodium p-
nitrophenoxide, NaOMe, NaOH). On the other hand, other N-hydroxyimides
(such as N-hydroxyphthalimide, N-hydroxymaleimide) also showed con-
siderable conversion.
is not seen in cycloisomerization.33
(32) McDonald, F. E.; Gleason, M. M. J. Am. Chem. Soc. 1996, 118, 6648.
9
2532 J. AM. CHEM. SOC. VOL. 124, NO. 11, 2002

Ru Catalyzed Divergence
A R T I C L E S
2. General Procedures for the Synthesis of Valerolactones. 2.a.
Preparation of Valerolactone 6 from Alcohol 5. By using method A
in Table 2, a mixture of alcohol 5 (135 mg, 0.742 mmol), catalyst 9c
(67 mg, 0.074 mmol), ligand 2c (104 mg, 0.296 mmol), N-hydroxy-
succinimide (511 mg, 4.44 mmol), sodium bicarbonate (124 mg, 1.48
mmol), tetra-n-butylammonium hexafluorophosphate (86 mg, 0.222
mmol) in DMF (1.9 mL, 0.4M) was placed in a preheated oil bath at
85 °C and stirred at that temperature under N2 for 23 h. The reaction
mixture was cooled to room temperature and diluted with ether (30
mL) and washed with water (2 × 10 mL). The aqueous layer was
extracted with ether (2 × 25 mL). The organic layers were combined,
dried over MgSO4, and concentrated. The residual oil was purified by
column chromatography (with deactivated silica gel, eluted with pet-
ether/EtOAc 20:1) to provide dihydropyran 7 (10 mg, 0.052 mmol,
7%). Further elution (with pet-ether/EtOAc 4:1) provided a trace amount
of starting material (<1 mg) and valerolactone 6 (101 mg, 0.509 mmol)
as clear oil (69% yield). The spectral data are in full accord with the
reported value.11 Rf 0.50 (pet-ether/EtOAc 7:1).
IV. Summary
In this article, we described a divergent ruthenium-catalyzed
reaction from bis-homopropargylic alcohols that provides a
convenient access to either dihydropyrans or valerolactones.
Unlike homopropargylic alcohols, the reactions of bis-homopro-
pargylic alcohols required the use of excess phosphines to avoid
the undesired exo-cyclization pathway. Furthermore, either
cycloisomerization to dihydropyrans or oxidative cyclization to
δ-valerolactones may be performed. The divergence described
here is remarkable because it derives only by choice and amount
of phosphine ligands, a flexibility not previously observed.
V. Experimental
1. General Procedures for the Synthesis of Dihydropyrans. 1.a.
Preparation of Dihydropyran 7 from Alcohol 5. By using method
C in Table 2, a mixture of alcohol 5 (140 mg, 0.768 mmol), catalyst
9e (32 mg, 0.038 mmol), ligand 2e (49 mg, 0.15 mmol), 4 sodium salt
(53 mg, 0.38 mmol), and tetra-n-butylammonium hexafluorophosphate
(39 mg, 0.10 mmol) in DMF (1.9 mL, 0.4 M) was placed in a preheated
oil bath at 85 °C and stirred at that temperature under N2 for 25 h. The
reaction mixture was cooled to room temperature and diluted with ether
(30 mL) and washed with water (2 × 10 mL). The aqueous layer was
extracted with ether (2 × 25 mL). The organic layers were combined,
dried over MgSO4, and concentrated. The residual oil was purified by
column chromatography (with deactivated silica gel, eluted with pet-
ether/EtOAc 20:1) to provide dihydropyran 7 (91 mg, 0.50 mmol) as
clear oil (64% yield) Rf 0.70 (pet-ether/EtOAc 20:1).
IR (neat film): ν 2928, 2857, 1736, 1465, 1378, 1342, 1241, 1178,
1
1046 cm-1. H NMR (CDCl3, 300 MHz): δ 4.26(m, 1H), 2.41-2.62
(m, 2H), 1.20-2.00(m, 16H), 0.87(t, J ) 6.4 Hz, 3H). 13C NMR
(CDCl3, 75 MHz): δ 172.0, 80.6, 35.8, 31.7, 29.5, 29.3, 29.1, 27.8,
24.9, 22.6, 18.5, 14.1.
2.b. Preparation of Valerolactone 17 from Alcohol 16. By using
the same procedure, a mixture of 16 (260 mg, 0.551 mmol), catalyst
9c (50 mg, 0.0551 mmol), ligand 2c (78 mg, 0.22mol), N-hydroxysuc-
cinimide (380 mg, 3.31 mmol), sodium bicarbonate (93 mg, 1.1 mmol),
tetra-n-butylammonium hexafluorophosphate (64 mg, 0.17 mmol) in
DMF (1.4 mL, 0.4M) was placed in a preheated oil bath at 85 °C and
stirred at that temperature for 26 h. The reaction mixture was cooled
to room temperature, and DMF was removed under reduced pressure.
The reaction mixture was purified directly by column chromatograhy
(with deactivated silica gel, eluted with pet-ether/EtOAc 20:1) to give
the dihydropyran (18 mg, 0.038, 7% yield). Further elution (with pet-
ether/EtOAc 4:1) provided the starting material (2 mg, ∼1%) and the
valerolactone 17 (175 mg, 0.358 mmol) as a clear oil (65% yield).
[R]D 70.7 (c 1.11, CHCl3). Rf 0.55 (pet-ether/EtOAc 4:1).
IR (neat film): ν 2928, 2856, 1730, 1461, 1260, 1121, 1033 cm-1
.
1H NMR (CDCl3, 300 MHz): δ 6.36(d, J ) 5.7 Hz, 1H), 4.65(m, 1H),
3.84(m, 1H), 1.85-2.10(m, 2H), 1.15-1.80(m, 14H), 0.88(t, J ) 6.5
Hz, 3H). 13C NMR (CDCl3, 75 MHz): δ 143.8, 100.3, 75.2, 35.3, 31.8,
29.6, 29.3, 27.8, 25.3, 22.7, 19.9, 14.1. HRMS: Calcd for C12H22O
(M+): 182.1671. Found: 182.1673.
Further elution (with pet-ether/EtOAc 4:1) provided unreacted
starting material 5 (∼1 mg) and the lactone 6 as clear oil (6 mg, 0.03
mmol, 4% yield).
1.b. Preparation of Dihydropyran 18 from Alcohol 16. With the
same method, a mixture of 16 (238 mg, 0.503 mmol), ruthenium catalyst
9e (21 mg, 0.025 mmol), phosphine ligand 2e (32 mg, 0.10 mmol), 4
sodium salt (35 mg, 0.25 mmol), and tetra-n-butylammonium hexafluo-
rophosphate (29 mg, 0.076 mmol) in DMF (1.3 mL, 0.4M) was placed
in a preheated oil bath at 85 °C and stirred at that temperature for 26
h. The reaction mixture was cooled to room temperature. DMF was
removed under reduced pressure, and the reaction mixture was purified
directly with column chromatography (with deactivated silica gel, eluted
with pet-ether/EtOAc 20:1) to give the title compound 6b as clear oil
(162 mg, 0.344 mmol, 68% yield). The spectral data are in full accord
with the reported value.20 [R]D 50.5 (c 2.10, CHCl3). Rf 0.75 (pet-ether/
EtOAc 10:1).
IR (neat film): ν 3030, 2867, 1950, 1749, 1595, 1497, 1454, 1366,
1
1310, 1252, 1203 cm-1. H NMR (CDCl3, 300 MHz): δ 7.03-7.40
(m, 15H), 5.03(d, J ) 11.0 Hz, 1H), 4.85(d, J ) 10.5 Hz, 1H), 4.78(d,
J ) 11.0 Hz, 1H), 4.60(d, J ) 12.1 Hz, 1H), 4.56(d, J ) 12.1 Hz,
1H), 4.48(d, J ) 10.5 Hz, 1H), 4.11(m, 1H), 3.60-3.82(m, 6H), 2.80-
(ddd, J ) 17.6, 9.5, 5.9 Hz, 1H), 2.61(m, 1H), 2.20(m, 1H), 1.96(m,
1H). 13C NMR (CDCl3, 75 MHz): δ 170.3, 138.1, 137.8, 137.7, 128.4,
128.1, 127.9, 127.8, 127.7, 94.0, 83.7, 81.3, 79.2, 77.2, 75.4, 73.5, 72.1,
68.6, 27.6, 24.3. HRMS: Calcd for C30H32O6 (M+): 488.2199.
Found: 488.2193.
IR (neat film): ν 3064, 3030, 2859, 1649, 1497, 1454, 1362, 1320,
1238, 1209, 1107, 1068 cm-1. H NMR (CDCl3, 300 MHz): 7.11-
1
Acknowledgment. We thank the National Institutes of Health
and the National Science Foundation for their generous support
of our program. Mass spectra were provided by the Mass
Spectrometry Facility at the University of California-San
Francisco supported by NIH Division of Research Resources.
7.41(m, 15H), 6.39(d, J ) 5.6 Hz, 1H), 4.99(d, J ) 11.0 Hz, 1H),
4.90(d, J ) 10.7 Hz, 1H), 4.86(d, J ) 10.7 Hz, 1H), 4.79(d, J ) 11.0
Hz, 1H), 4.71(m, 1H), 4.62(d, J ) 12.2 Hz, 1H), 4.54(d, J ) 12.2 Hz,
1H), 3.40-3.60(m, 7H), 2.40(m, 1H), 2.20(m, 1H). 13C NMR (CDCl3,
75 MHz): δ 142.8, 138.7, 138.1, 137.9, 128.4, 128.3, 127.9, 127.8,
127.7, 127.6, 127.5, 98.3, 84.2, 79.1, 78.8, 77.4, 75.2, 74.9, 73.5, 72.2,
68.9, 26.7.
Supporting Information Available: Detailed descriptions of
experimental procedures and spectral data for all new com-
pounds (PDF). This material is available free of charge via the
Internet at http://pubs.acs.org.
Further elution (with pet-ether/EtOAc :1) provided starting material
(2 mg, 0.005 mmol, 1%) and lactone 17 (17 mg, 0.026 mmol, 5%).
(33) The poor reactivity of tertiary bis-homopropargylic alcohols with stoichio-
metric amount of molybdenum complexes has been addressed: Weyer-
shausen, B.; Nieger, M.; Dotz, K. H. J. Organomet. Chem. 2000, 602, 37.
JA011840W
9
J. AM. CHEM. SOC. VOL. 124, NO. 11, 2002 2533