Ruthenium-catalyzed cycloisomerization - Oxidation of homopropargyl alcohols. A new access to γ-butyrolactones

Article abstract of DOI:10.1021/ja992013m

Vinylidenemetal species, which readily form from terminal alkynes under mild conditions, have rarely been utilized as reactive intermediates in a catalytic cycle. The conversion of homopropargyl alcohols via such intermediates to metal-complexed oxacarbenes led to the development of an oxidant compatible with a ruthenium complex capable of performing the cycloisomerization, that would convert them to lactones. None of the oxidants known to stoichiometrically convert isolated metallooxacarbenes to esters are effective. The unconventional oxidants , N-hydroxyimides, proved to be capable of effecting the desired transformation, with N-hydroxysuccinimide being the oxidant of choice. The procedure of choice employs cyclopentadienyl (1,4-cyclooctadiene) ruthenium chloride and trifuryl phosphine as the precatalyst in the presence of tetra-n-butylammonium bromide or hexafluorophosphate with N-hydroxysuccinimide as the oxidant in DMF-water at 95°. In this way, a wide diversity of homopropargyl alcohols were converted to γ-butyrolactones with excellent chemoselectivity. Lactones synthesized include an intermediate toward a platelet aggregation inhibitor, a fruit flavor principle, an inhibitor of binding of phorbol esters to PKC-α, a tobacco constituent, a wood constituent (quercus lactone), an aldosterone antagonist (spironolactone) precursor, and an acetogenin known for pesticidal and antitumor activities (muricatacin).

Full text of DOI:10.1021/ja992013m

11680
J. Am. Chem. Soc. 1999, 121, 11680-11683
Ruthenium-Catalyzed Cycloisomerization-Oxidation of
Homopropargyl Alcohols. A New Access to γ-Butyrolactones
Barry M. Trost* and Young H. Rhee
Contribution from the Department of Chemistry, Stanford UniVersity, Stanford, California 94305-5080
ReceiVed June 14, 1999
Abstract: Vinylidenemetal species, which readily form from terminal alkynes under mild conditions, have
rarely been utilized as reactive intermediates in a catalytic cycle. The conversion of homopropargyl alcohols
via such intermediates to metal-complexed oxacarbenes led to the development of an “oxidant” compatible
with a ruthenium complex capable of performing the cycloisomerization, that would convert them to lactones.
None of the oxidants known to stoichiometrically convert isolated metallooxacarbenes to esters are effective.
The unconventional “oxidants”, N-hydroxyimides, proved to be capable of effecting the desired transformation,
with N-hydroxysuccinimide being the “oxidant” of choice. The procedure of choice employs cyclopentadienyl
(1,4-cyclooctadiene) ruthenium chloride and trifuryl phosphine as the precatalyst in the presence of tetra-n-
butylammonium bromide or hexafluorophosphate with N-hydroxysuccinimide as the oxidant in DMF-water
at 95°. In this way, a wide diversity of homopropargyl alcohols were converted to γ-butyrolactones with
excellent chemoselectivity. Lactones synthesized include an intermediate toward a platelet aggregation inhibitor,
a fruit flavor principle, an inhibitor of binding of phorbol esters to PKC-R, a tobacco constituent, a wood
constituent (quercus lactone), an aldosterone antagonist (spironolactone) precursor, and an acetogenin known
for pesticidal and antitumor activities (muricatacin).
Organometallic vinylidene complexes derived from acetylenic
compounds have been widely studied, but few catalytic cycles
have evolved.1-4 The facility by which such complexes are
generated makes them attractive targets for development of more
atom economical methodology. The importance of five-
membered ring oxygen heterocycles directed our attention to
the facility by which homopropargylic alcohols 1 react with a
ruthenium complex to form an oxacarbene species 3, presumably
via nucleophilic addition to a vinylidene carbene intermediate
2 (eq 1).2b Release of the tetrahydrofuran moiety is required in
order to convert this cycloisomerization into a catalytic cycle.
The potential susceptibility of carbene intermediates toward
nucleophilic addition and the importance of γ-butyrolactones
led us to consider a catalytic cycle as outlined in Scheme 1.
Such a task appeared daunting because extensive efforts toward
stoichiometric decomplexation of the intermediate proved
fruitless.5 Nevertheless, oxidation with strong oxidants such as
ceric ammonium nitrate and dimethyldioxirane has been re-
ported.3 The critical issue is to discover an oxidant that will
maintain the catalytic activity of the ruthenium for the cyclo-
isomerization. We wish to report a mild oxidative cyclization
of homopropargylic alcohols to γ-butyrolactones, catalyzed by
a ruthenium complex.
To investigate the feasibility of the process, the transformation
illustrated in eq 2 was pursued. We initiated our studies of the
(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.
(2) (a) McDonald, F. E.; Gleason, M. M. J. Am. Chem. Soc. 1996, 118,
6648 and references therein. (b) Bruce, M. I.; Swincer, A. G.; Thomson,
B. J.; Wallis, R. C. Aust. J. Chem. 1980, 33, 2605.
(3) (a) Quayle, P.; Rahman, S.; Ward, E. L. M. Tetrahedron Lett. 1994,
35, 3801. Quayle, P.; Rahman, S. Tetrahedron Lett. 1995, 36, 8087. (b)
Gilbert, M.; Ferrer, M.; Lluch, A.-M.; Sanchez-Baeza, F.; Messeguer, A.
J. Org. Chem. 1999, 64, 1591. (c) Wulff, W. D.; Bauta, W. E.; Kaesler, R.
W.; Lankford, P. J.; Miller, R. A.; Murray, C. K.; Yang, D. C. J. Am. Chem.
Soc. 1990, 112, 3642.
(4) (a) For a recent general review for ruthenium catalyzed reactions,
see: Murahashi, S.-I.; Takaya, H.; Naota, T. Chem. ReV. 1998, 98, 2599.
(b) For 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. Trost, B. M.; Dyker, G.; Kulawiec, R. J. J.
Am. Chem. Soc. 1990, 112, 7809. 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.
oxidative cyclization of homopropargyl alcohol 56 with the
ruthenium complex 4 based upon Bruce’s work2b and the use
of a polar medium such as DMF-water to promote ionization
of the catalyst. With conventional oxidants (hydrogen peroxide,
tert-butyl hydroperoxide, MCPBA, pyridine N-oxide, DMSO),
only starting material 5 was recovered. On the other hand, the
unconventional “oxidant”, N-hydroxyphthalimide (7), did give
some of the desired lactone 67 (Table 1, entry 1). A conversion
(5) For unsuccessful examples, see: Davies, G.; McNally, J. P.;
Smallridge, A. J. AdV. Organomet. Chem. 1990, 30, 1 and references therein.
(6) Danheiser, R. L.; Carini, D. J.; Kwasigroch, C. A. J. Org. Chem.
1986, 51, 3870.
(7) Olivier, P. Tetrahedron 1994, 50, 13687.
10.1021/ja992013m CCC: $18.00 © 1999 American Chemical Society
Published on Web 12/07/1999

New Access to γ-Butyrolactones
J. Am. Chem. Soc., Vol. 121, No. 50, 1999 11681
Table 2. Variation of the Ligand Systema
Scheme 1. Proposed Catalytic Cycle
Ru amt
(%)
ligand reaction convrsn yieldb brsmc
entry
ligand amt (%) time (h)
(%)b
(%)
(%)
1
2
3
4
5
6
7
8
9
15
15
15
15
15
15
15
10
5
PPh3
11
12
13
14
15
16
16
16
23
23
23
23
23
23
23
15
13
13
13
13
13
13
13
18
23
95
100
100
81
79
77
100
100
100
60
29
22
59
55
50
66
73
64
29
22
72
69
65
66
73
74
Table 1. Variation of the Oxidants Using Precatalyst 4a
7.5
74
(63d)
66
10
4
16
6
29
93
71
a Reaction performed as in eq 2 with N-hydroxysuccinimide (9) as
oxidant and tetra-n-butylammonium bromide as additive. b Determined
by GC using tetradecane as an internal standard. c Yield based upon
reacted starting material. d Isolated yield.
concn time convrsn yieldb brsmc
entry oxidant
additive
(M) (h)
(%)b
(%) (%)
1
2
3
4
5
7
8
9
9
9
0.1
0.1
0.1
0.4
0.4
72
72
72
42
22
33
58
97
96
100
17
38
61
52
65
63
63
66
61
Bu4NBr
66
(59d)
45
6
7
8
9
9
9
9
9
9
Bu4NBr
Bu4NCl
Bu4NHSO4
0.8
0.4
0.4
22
22
22
22
22
28
100
100
100
72
71
100
45
64
60
68
58
71
64
60
49
41
phosphines per ruthenium allowed reaction to proceed to
completion after 17 h. As summarized in Table 2, bulkier as
well as less bulky9 and more electron rich10 phosphines gave
poorer results (entries 2-5). A phosphite ligand gave poorer
conversions, although the yield based upon recovered starting
material was nearly unchanged (entry 6). On the other hand,
trifurylphosphine (16), a small and electron-poor ligand,11 gave
somewhat improved results (entry 7). Interestingly, decreasing
the amount of the catalyst increased the yield somewhat (entries
8 and 9). Dropping the catalyst loading below 5% still
maintained good yields but saw the conversion drop somewhat
(entry 10). These results are consistent with the nucleophilic
addition step being product (rate?) determining; therefore,
increasing the electrophilicity of the ruthenium12 and minimizing
steric hindrance increases the reactivity of the catalyst.
9
C16H33(Me)3NBr 0.4
10
(CH3)4NBr
Bu4NPF6
0.4
0.4
11e
71
(65d)
a Reaction performed as outlined in eq 2 with Ru complex 4 (15
mol %) as precatalyst and 45 mol % of any additive. b Determined by
GC using tetradecane as an internal standard. c Yield based upon reacted
starting material. d Isolated yield. e In this run, 18 mol % Ru complex
4 employed.
Using the conditions of 5-10% of ruthenium complex 10 in
the presence of trifurylphosphine and tetra n-butylammonium
bromide with sodium bicarbonate as base and 9 as oxidant, a
wide range of homopropargylic alcohols was converted to their
corresponding γ-butyrolactones (see Table 3).13 The substrates
problem was apparent. Using a sterically less demanding imide,
the maleimide 8, did improve the conversion (entry 2). The best
result was obtained with the more nucleophilic N-hydroxysuc-
cinimide (9) (entry 3), although, in principle, it should be the
poorer oxidant (i.e., stronger N-O bond).
Since nucleophilicity seems to be a significant factor,
modifying the counterion was examined. Use of a tetra-n-
butylammonium salt significantly shortened the reaction time
(entries 5-8). No apparent dependence on the nature of the
anion associated with the tetra-n-butylammonium salt was noted
(entries 5-8). On the other hand, the structure of the ammonium
ion did have a significant effect (entries 9 and 10). Performing
the reaction in air has no effect. Using the conditions of entry
5, a 59% isolated yield of lactone 6 was obtained within 1 day.
To explore the effect of a phosphine ligand, we switched to
complex 10 as the precatalyst.8 Better conversions were obtained
by using less than 2 phosphines per ruthenium. Using 1.5
(8) Albers, M. O.; Robinson, D. J.; Shaver, A.; Singleton, E. Organo-
metallics 1986, 5, 2199.
(9) (a) Cone angle data for the ligands: (i) PPh3, 145°; (ii) 12, 170°;
(iii) 13, 136°; (iv) 14, 122°; (v) 15, 109°. (b) For extensive compilation of
cone angles of phosphine ligands, see: Rahman, M. M.; Liu, H.-Y.; Eriks,
K.; Prock, A.; Giering, W. P. Organometallics 1989, 8, 1.
(10) For quantitative data for electron density of phosphine ligands,
see: Tolman, C. A. Chem. ReV. 1977, 77, 313.
(11) For the use of tri-o-furylphosphine in cross-coupling reactions,
see: Farina, V.; Krishnan, B. J. Am. Chem. Soc. 1991, 113, 9585. Farina,
V.; Baker, S. R.; Benigni, D. A.; Sapino, C., Jr. Tetrahedron Lett. 1988,
29, 5739.
(12) For analogous discussion in the intermolecular reaction of methanol
with Ru-vinylidene complexes, see ref 5. Also, see: Lagadec, R. L.; Roman,
E.; Toupet, L.; Muller, U.; Dixneuf, P. H. Organometallics 1994, 13, 5030.
Bozec, H. L.; Dixneuf, P. H. Organometallics 1991, 10, 2768.

11682 J. Am. Chem. Soc., Vol. 121, No. 50, 1999
Trost and Rhee
Table 3. Representative Examples of Ru-Catalyzed Oxidative Cyclization of Homopropargyl Alcohols
a Method A, as in eq 2 using 5% 10, 7.5% 16, 15% tetra-n-butylammonium bromide, and 9 as oxidant. Method B, as in eq 2 using 10% 10, 15%
16, 30% tetra-n-butylammonium bromide, and 9 as oxidant. Method C, as in eq 2 using 7% 10, 10% 16, 45% tetra-n-butylammonium
hexafluorophosphate, and 9 as oxidant. Method D, as in eq 2 using 13% 10, 20% 16, 60% tetra-n-butylammonium hexaflurophosphate, and 9 as
oxidant. Method E, as in eq 2 using 10% 10, 15% 16, 50% tetra-n-butylammonium hexafluorophosphate, and 9 as oxidant. b Isolated yields. c GC
yield in this case is 84°).
are easily accessed by adding allenylmagnesium bromide to
aldehydes or ketones (entries 1-3, 6, 7, 9-11) or by opening
epoxides (entries 4, 5 (after alcohol inversion), and 8).14
Excellent chemoselectivity is observed. Both cis- and trans-
fused lactones are equivalently accessed. In many examples of
formation of the vinylidene ruthenium complexes, ammonium
hexafluorophosphate is employed.1,4,16 To check whether the
presence of hexafluorophosphate anion plays any role, the
reaction of entry 1 was repeated using tetra-n-butylammonium
hexafluorophosphate. Remarkably, the yield increased signifi-
cantly. Using this latter quaternary salt did increase the yield
in several additional cases (entries 2, 4, 5, and 8) but had no
effect on entry 7. A number of the lactones in Table 3 are of
significant interest, including 25, an intermediate toward a
platelet aggregation inhibitor;15a 26, a fruit flavor principle; 29,
an inhibitor of binding of a phorbol ester to PKC-R; 30, a
(15) (a) For the discussion of the usage of this compound, see: Vijay,
N.; Jaya, P.; Tesmol, G. Tetrahedron 1998, 53, 15061. (b) Krief, A.;
Rouvaux, A.; Tuch, A. Tetrahedron 1998, 54, 6903. (c) Fristad, W. E.;
Peterson, J. R. J. Org. Chem. 1985, 50, 10. (d) Lee, J.; Marquez, V. E.;
Lewin, N. E.; Blumberg, P. M. Synlett 1994, 206. (e) Wahlberg, I.;
Pettersson, T.; Eklund, A.-M. J. Agric. Food Chem. 1993, 41, 2097. (f)
For other syntheses of Quercus lactone, see: Masuda, M.; Nishimura, K.
Chem. Lett. 1981, 133 and references therein. (g) An intermediate in the
synthesis of spironolactone: Cella, J. A.; Brown, E. A.; Burtner, R. R. J.
Org. Chem. 1959, 24, 743. For the discussion of the biological activity of
spironolactone as an aldosterone antagonist, see: Sutter, J. L.; Lau, E. P.
K. In Analytical Profiles of Drug Substances; Florez, K., Ed.; Academic:
New York, 1975; Vol. 4, pp 431-451.
(13) For the discussion of biological activity of γ-butyrolactones, see:
Chamberlin, R.; Koch, C. J. Org. Chem. 1993, 58, 2725 and references
therein.
(14) (a) Bieber, L. W.; Silva, M. F.; Costa, R. C.; Silva, L. O. Tetrahedron
Lett. 1998, 39, 3655. (b) Fukuda, Y.; Matsubara, S.; Utimoto, K. J. Org.
Chem. 1991, 56, 5812. (c) Murray, T. F.; Samsel, E. G.; Varma, V.; Norton,
J. R. J. Am. Chem. Soc. 1981, 103, 7520. (d) Alper, B.; Hendrix, M.; Sears,
P.; Wong, C.-H. J. Am. Chem. Soc. 1998, 120, 1965. (e) This compound
was synthesized from commercially available 5-acetyl-2-methylfuran by
treatment with propargylmagnesium bromide. (f) Hoffmann, R. W.; Giesen,
V. F.; Fuest, M. Liebigs Ann. Chem. 1993, 6, 629. (g) Mauvais, A.; Burger,
A.; Roussel, J. P.; Hetru, C.; Luu, B. Bioorg. Chem. 1994, 22, 36.
(16) Trost, B. M.; Kulawiec, R. J. J. Am. Chem. Soc. 1992, 114, 5579.
Ruiz, N.; Pe´ron, D.; Dixneuf, P. H. Organometallics 1995, 14, 1095.

New Access to γ-Butyrolactones
J. Am. Chem. Soc., Vol. 121, No. 50, 1999 11683
Scheme 2. Sunthesis of (-)-Muricatacin
optical rotations. This example highlights the chemoselectivitys
only the γ-butyrolactone is obtained, with no detectable amount
of the corresponding six-membered ring lactone. Thus, free
hydroxyl groups do not interfere.
The question of mechanism must be reserved until more
detailed studies are performed. The working hypothesis pre-
sented in Scheme 1 appears consistent with our observations.
The steric and electronic influences on the ruthenium that affect
the nucleophilic addition appear to be quite significant. Further
developments of this new catalytic lactone synthesis and its
broader implications for additional catalytic cycles based upon
the principles outlined in this scheme are the subject of future
studies.
tobacco constituent; 31, a wood constituent;15f and 32,15g an
intermediate toward spironolactone, an aldosterone antagonist.
Entries 10 and 11 illustrate chemoselectivity issues. Finally, a
concise synthesis of the acetogenin (-)-muricatacin,17 com-
pounds known for their pesticidal and antitumor activities, was
accomplished as shown in Scheme 2 from the known bromide
33.18,19 The asymmetric dihydroxylation of enyne 34 proceeded
in about 94% ee, established on the basis of the enantiomeric
purity of the final product as determined by comparison of
Acknowledgment. We thank the National Science Founda-
tion and the General Medical Sciences Institute of the National
Institutes of Health for their generous support of our work. Mass
spectra were provided by the Mass Spectrometry Regional
Center of the University of CaliforniasSan Francisco, supported
by the NIH Division of Research Resources.
Supporting Information Available: General procedure for
the synthesis of γ-butyrolactones and characterization data for
25-32, 36 (PDF). This material is available free of charge via
the Internet at http://pubs.acs.org.
(17) (a) For the isolation and biological activity, see: Rieser, M. J.;
Kozlowski, J. F.; Wood, K. V.; McLaughlin, J. L. Tetrahedron Lett. 1991,
32, 1137. (b) For other syntheses of this natural product, see: Yoon, S.-H.;
Moon, H.-S.; Hwang, S.-K.; Choi, S.; Kang, S.-K. Bioorg. Med. Chem.
1998, 6, 1043 and references therein.
(18) Cravatt, B. F.; Lerner, R. A.; Boger, D. L. J. Am. Chem. Soc. 1996,
118, 580.
JA992013M
(19) Enantiomeric excess of 36 was determined by the comparison of
optical rotation data. See: Sharpless, K. B.; Zhang, X.-L.; Wang, Z.-M.;
Keinan, E.; Sinha, S. Tetrahedron Lett. 1992, 33, 6407.