Stereocontrolled syntheses of the nemorensic acids using 6-diazoheptane-2,5-dione in carbonyl ylide cycloadditions

Article abstract of DOI:10.1021/jo048446b

Levulinic acid-derived 6-diazoheptane-2,5-dione (9) serves as a common precursor in a formal synthesis of frontalin 19, and in syntheses of cis-nemorensic acid 1, 4-hydroxy-cis-nemorensic acid 2, 3-hydroxy-cis-nemorensic acid 3, and nemorensic acid 4. The key step in these syntheses is the Rh ₂(OAc)₄-catalyzed tandem carbonyl ylide formation-intermolecular 1,3-dipolar cycloadditions of diazodione 9 with formaldehyde, alkynes or allene, which occur with high regioselectivity. Subsequent oxidative cleavage of the ring originally derived from the cyclic carbonyl ylide intermediate provides a straightforward access to polysubstituted tetrahydrofurans, and in particular an efficient entry to the nemorensic acids. Enantioselective cycloadditions with diazodione 9, using chiral rhodium catalysts, gave cycloadducts in up to 51% ee.

Full text of DOI:10.1021/jo048446b

Stereocontrolled Syntheses of the Nemorensic Acids Using
6-Diazoheptane-2,5-dione in Carbonyl Ylide Cycloadditions
David M. Hodgson,* Fre´de´ric Le Strat, Thomas D. Avery, Andrew C. Donohue, and
Tobias Bru¨ckl
Department of Chemistry, University of Oxford, Chemistry Research Laboratory, Mansfield Road,
Oxford, OX1 3TA, United Kingdom
david.hodgson@chem.ox.ac.uk
Received September 3, 2004
Levulinic acid-derived 6-diazoheptane-2,5-dione (9) serves as a common precursor in a formal
synthesis of frontalin 19, and in syntheses of cis-nemorensic acid 1, 4-hydroxy-cis-nemorensic acid
2, 3-hydroxy-cis-nemorensic acid 3, and nemorensic acid 4. The key step in these syntheses is the
Rh2(OAc)4-catalyzed tandem carbonyl ylide formation-intermolecular 1,3-dipolar cycloadditions
of diazodione 9 with formaldehyde, alkynes or allene, which occur with high regioselectivity.
Subsequent oxidative cleavage of the ring originally derived from the cyclic carbonyl ylide
intermediate provides a straightforward access to polysubstituted tetrahydrofurans, and in
particular an efficient entry to the nemorensic acids. Enantioselective cycloadditions with diazodione
9, using chiral rhodium catalysts, gave cycloadducts in up to 51% ee.
Ibata and Padwa developed the use of copper and
rhodium complexes as catalysts for the decomposition of
diazocarbonyl compounds to produce transient carbonyl
ylides, which undergo cycloaddition with a range of
dipolarophiles.1 This is an attractive method for synthe-
sis, because of the relative ease of access of the starting
materials combined with the rapid increase in molecular
complexity that occurs during the tandem carbonyl ylide
formation-cycloaddition process. The strategy has found
wide utility in targeted synthesis. We were attracted to
the use of 6-diazoheptane-2,5-dione (9) as a substrate for
this chemistry (Scheme 1). This precursor could poten-
tially allow concise access to nemorensic acids 1-4,
following oxidative cleavage of the ring originally derived
from the cyclic carbonyl ylide 8.
The nemorensic acids 1-4 are dicarboxylic (necic) acids
obtained from certain pyrrolizidine alkaloids 12-16
(Figure 1).2 These natural products have been the focus
of considerable attention due to their challenging struc-
ture combined with diverse biological activity ranging
from potent hepatotoxicity to antitumour activity. The
synthesis of nemorensic acid 4 has been accomplished
by several approaches;3 however, we considered our
strategy would have the benefit of brevity, and enough
flexibility to address the various challenging substitution
patterns and stereochemical issues found in all of the
nemorensic acids, shown in Scheme 1.4
Prior to our work, Padwa had prepared the diazodione
10 from levulinic acid 11 and had examined its cyclo-
addition chemistry with a range of dipolarophiles;5
importantly, among these propargyl chloride had been
shown to undergo cycloaddition in the regiochemical
sense desired for our projected nemorensic acid synthe-
ses. Although during the course of our studies other
R-methyl R-diazoketones have been shown to undergo
tandem ylide formation-cycloaddition,6 it was not clear
at the outset of our work whether the additional methyl
substituent at the diazo-bearing carbon would allow our
proposed chemistry to occur, and whether this substitu-
ent might affect the regiochemical outcome. Similarly to
diazodione 10, diazodione 9 was prepared from levulinic
acid 8, but with diazoethane7 instead of diazomethane
(Scheme 2). The best yields were obtained if diazoethane
(∼3 equiv) was generated and distilled directly into a
(3) (a) Klein, L. L. J. Am. Chem. Soc. 1985, 107, 2573-2574. (b)
Dillon, M. P.; Lee, N. C.; Stappenbeck, F.; White, J. D. J. Chem. Soc.,
Chem. Commun. 1995, 1645-1646. (c) Rodriguez, J. R.; Rumbo, A.;
Castedo, L.; Mascaren˜as, J. L. J. Org. Chem. 1985, 50, 4560-4563.
Lo´pez, F.; Castedo, L.; Mascaren˜as, J. L. Chem. Eur. J. 2002, 8, 884-
899. (d) Honda, T.; Ishikawa, F. J. Org. Chem. 1999, 64, 5542-5546.
(e) Donohoe, T. J.; Guillermin, J.-B.; Frampton, C.; Walter, D. S. Chem.
Commun. 2000, 465-466. Donohoe, T. J.; Guilermin, J.-B.; Walter,
D. S. J. Chem. Soc., Perkin Trans. 1 2002, 1369-1375. (f) Liu, B.;
Moeller, K. D. Tetrahedron Lett. 2001, 42, 7163-7165. Liu, B.; Duan,
S.; Sutterer, A. C.; Moeller, K. D. J. Am. Chem. Soc. 2002, 124, 10101-
10111.
(4) (a) Hodgson, D. M.; Avery, T. D.; Donohue, A. C. Org. Lett. 2002,
4, 1809-1811. (b) Hodgson, D. M.; Le Strat, F. Chem. Commun. 2004,
822-823.
(5) Padwa, A.; Fryxell, G. E.; Zhi, L. J. Am. Chem. Soc. 1990, 112,
3100-3109.
(1) For reviews, see: (a) Padwa, A.; Weingarten, M. D. Chem. Rev.
1996, 96, 223-269. (b) Doyle, M. P.; McKervey, M. A.; Ye, T. Modern
Catalytic Methods for Organic Synthesis with Diazo Compounds; John
Wiley & Sons: New York, 1998; Chapter 7. (c) Clark, J. S. Nitrogen,
Oxygen and Sulfur Ylide Chemistry; Oxford University Press: Oxford,
UK, 2002. (d) Mehta, G.; Muthusamy, S. Tetrahedron 2002, 58, 9477-
9504.
(2) (a) Romo de Vivar, A.; Pe´rez, A.-L.; Arciniegas, A.; Vidales, P.;
Gavin˜o, R.; Villasen˜or, J. L. Tetrahedron 1995, 51, 12521-12528. (b)
Pe´rez-Castorena, A.-L.; Arciniegas, A.; Castro, A.; Villasen˜or, J. L.;
Toscano, R. A.; Romo de Vivar, A. J. Nat. Prod. 1997, 60, 1322-1325.
(c) Liddell, J. R. Nat. Prod. Rep. 2001, 18, 441-447.
(6) (a) Muthusamy, S.; Babu, S. A.; Gunanathan, C.; Suresh, E.;
Dastidar, P.; Jasra, R. V. Tetrahedron 2000, 56, 6307-6318. (b) Wood,
J. L.; Thompson, B. D.; Yusuff, N.; Pflum, D. A.; Mattha¨us, M. S. P. J.
Am. Chem. Soc. 2001, 123, 2097-2098.
10.1021/jo048446b CCC: $27.50 © 2004 American Chemical Society
Published on Web 11/06/2004
8796
J. Org. Chem. 2004, 69, 8796-8803

Stereocontrolled Syntheses of the Nemorensic Acids
SCHEME 1. Retrosynthetic Analysis of Nemorensic Acids 1-4
solution of the mixed carboxylic carbonic anhydride of
levulinic acid and isobutyl chloroformate. It is also
important that the receiving flask containing the mixed
anhydride is kept cold (-5 °C) to reduce the loss of
diazoethane (bp ca. -20 °C) by evaporation, and that the
diazodione 9 is purified by rapid chromatography.
confirmed by X-ray crystallographic analysis.8 Ketone 18
has been previously converted (by thioketalization then
desulfurization) into frontalin 19, the aggregation phero-
mone of Dendroctonus beetles.9
With the reactivity of diazodione 9 as a carbonyl ylide
precursor established, we examined dipolarophiles that
might generate cycloadducts suitable for nemorensic acid
syntheses. Padwa found unstrained alkenes were unre-
active in intermolecular carbonyl ylide cycloadditions,5
and we similarly observed that propene was not a viable
dipolarophile in our system. However, propargyl chloride
(2 equiv) was found to generate a cycloadduct 20, albeit
in a modest 31% yield (unoptimized); nevertheless and
encouragingly, NOE studies indicated cycloaddition had
occurred with the desired regiochemistry. Propargyl
bromide (3.5 equiv) and propyne were also found to be
viable dipolarophiles, generating cycloadducts 5 (84%)
and 6 (87%), respectively (Scheme 1). Interestingly, use
of propargyltrimethylsilane gave the anticipated cycload-
duct 21, along with (chromatographically inseparable)
cycloadduct 22 (combined yield 82%, 21:22, 86:14); 22
probably arises from the ∼10% allenyltrimethylsilane
present in the commercial starting silane. The structure
of this latter cycloadduct was established on a pure
sample, available following treatment of the mixture with
TBAF, which selectively desilylated the major cycload-
duct 21 to a mixture of alkenes 6 and 7 (combined yield
72%, 6:7, 55:45). The observation that propargyl- and
allenyl-trimethylsilane showed similar reactivity in the
cycloaddition process prompted us to examine the cy-
cloaddition chemistry with allene10 itself, which success-
fully gave cycloadduct 7 (77% yield, Scheme 1). The
experiments which involved the volatile unsaturated
hydrocarbons propene, propyne, and allene were carried
out in a flask equipped with a dry ice condenser, by
addition of Rh2(OAc)4 to a cooled solution of the diazo-
dione 9 in CH2Cl2 containing an excess of the dipolaro-
phile, followed by warming to ambient temperature (0
FIGURE 1. Pyrrolizidine alkaloids from the Senecio genus.
With diazodione 9 in hand we were pleased to observe
that Rh2(OAc)4-catalyzed reaction with DMAD generated
cycloadduct 17 in good yield (75%, Scheme 2). Similarly,
reaction with formaldehyde gave dioxabicyclic ketone 18
as a single regioisomer (86% yield), whose structure was
SCHEME 2. Cycloadducts 17, 18, and 20-22 from
Diazodione 9a
(7) (a) Wilds, A. L.; Meader, A. L., Jr. J. Org. Chem. 1948, 13, 763-
779. (b) Marshall, J. A.; Partridge, J. J. J. Org. Chem. 1968, 33, 4090-
4095.
(8) CCDC 249265 available at http://www.ccdc.cam.ac.uk.
(9) (a) Taniguchi, T.; Nakamura, K.; Ogasawara, K. Synthesis 1997,
509-511. (b) Majewski, M.; Nowak, P. Tetrahedron: Asymmetry 1998,
9, 2611-2617. (c) Tyvorskii, V. I.; Astashko, D A.; Kulinkovich, O. G.
Tetrahedron 2004, 60, 1473-1479.
(10) Cripps, H. N.; Kiefer, E. F. Organic Syntheses; Wiley: New
York, 1973; Collect. Vol. V, pp 22-24.
a Reagents and conditions: (a) DMAD (2 equiv), Rh2(OAc)4 (10
mol %), CH2Cl2, 25 °C, 1 h; (b) CH2O (10 equiv), Rh2(OAc)4 (1 mol
%), CH2Cl2, -78 to 25 °C, 1 h; (c) propargyl chloride (2 equiv),
Rh2(OAc)4 (10 mol %), CH2Cl2, 25 °C, 1 h; (d) propargyltrimeth-
ylsilane/allenyltrimethylsilane (9:1, 3 equiv), Rh2(OAc)4 (5 mol %),
CH2Cl2, 25 °C, 3 h.
J. Org. Chem, Vol. 69, No. 25, 2004 8797

Hodgson et al.
°C with allene).11 In the absence of a dipolarophile very
slow evolution of nitrogen commenced at approximately
-60 °C and continued until -10 °C, when the evolution
became more vigorous and a color change (pale yellow to
green) was observed.
Having several cycloadducts now available as potential
precursors to the nemorensic acids, we focused on a
synthesis of cis-nemorensic acid 1 (Scheme 3). Exo-
selective alkene hydrogenation of cycloadducts 5 (with
concomitant hydrogenolysis of the C-Br bond), 6 (HBr
was found to be beneficial as an additive in this case),
and 7 all gave the desired saturated ketone 23 (92%, 91%,
and 74% yields, respectively). The procedure starting
with cycloadduct 5 is preferred, due to the comparative
ease of handling of the precursor dipolarophile (propargyl
bromide) and of the cycloadduct 5 (the cycloadducts from
allene and propyne are highly volatile).
conditions (Scheme 3) led us to examine the necessity of
adding H2O2 in this process. Indeed, in the absence of
H2O2 significant (though not complete) conversion to cis-
nemorensic acid 1 was observed by H NMR analysis
(1:26, 4:1).
1
Cycloadduct 5 was also examined as a substrate for
accessing 4-hydroxy-cis-nemorensic acid 2 (Scheme 4).
The hydroxyl group in this necic acid could potentially
be obtained via (formal) exo-selective anti-Markovnikov
hydration of the endocyclic double bond of 5. As a
hydroboration/H2O2 protocol would probably be compli-
cated by concomitant reactivity of the ketone, we focused
on oxygenation via epoxidation. However, when cycload-
duct 5 was treated with m-CPBA, the desired epoxide
27 was formed in only modest yield (34%) and epoxylac-
tone 28 arising from a Bayer-Villiger reaction was a
significant byproduct (23%). To improve the synthesis of
the desired epoxide 27, epoxidation with dimethyldiox-
irane (DMDO) generated in situ was attempted.18 Upon
performing this reaction an inseparable mixture of
several unidentified compounds was isolated. A possible
complication in this reaction could be dioxirane formation
at the keto group of 5. An efficient synthesis of epoxide
27 (94%) was finally achieved by using a pre-prepared
solution of DMDO,19 following precedent with nor-
bornenone.20
SCHEME 3. Synthesis of cis-Nemorensic Acid 1a
SCHEME 4. Synthesis of 4-Hydroxy-cis-nemorensic
Acid 2a
a Reagents and conditions: (a) H2 (1 atm), Pd/C (10%), MeOH,
25 °C, 48 h (92% from 5, 91% from 6, 74% from 7); (b) LDA (1.2
equiv), THF, -78 °C, 2 h then TMSCl (2 equiv), -78 to 25 °C, 1 h
(90%); (c) O2/O3, CH2Cl2, -78 °C, 5 min, then 35% H2O2 (6 equiv),
88% HCO2H (27 equiv), 100 °C, 30 min (96%).
An attempt to transform saturated ketone 23 directly
into cis-nemorensic acid 1 with use of conditions that are
known to convert cyclohexanone into adipic acid (O2,
Re2(CO)10, poly(ethylene glycol) (PEG-400), KOH, and
K2CO3 in DME, 74%)12 was not successful. Therefore, a
longer sequence proceeding via oxidative cleavage of the
corresponding silyl enol ether 2413 was examined. The
procedure of Kaneda et al. with (MoO2(acac)2, t-BuOOH)14
gave small amounts of cis-nemorensic acid 1 in a com-
plicated product mixture, whereas a sequence involving
ozonolysis in CH2Cl2, followed by Me2S workup and
oxidation of the resulting oxoacid with Jones’ reagent15
was more successful, generating cis-nemorensic acid 1
in 69% yield (structure confirmed by X-ray crystal-
lographic analysis16). A more efficient process from the
silyl enol ether 24 to cis-nemorensic acid 1 (96%) involved
reaction with HCO2H and H2O2 following ozonolysis.17
Consideration of a plausible reaction pathway for the
putative silyloxyozonide intermediate 25 under the latter
a Reagents and conditions: (a) m-CPBA (2.5 equiv), CH2Cl2, 25
°C, 48 h (34% of 27 and 23% of 28), or DMDO (0.1 M in acetone,
4 equiv), CH2Cl2, 25 °C, 5 d (94% of 27); (b) Zn (3 equiv), NaI (2.5
equiv), MeOH, 65 °C, 3.5 h (97%); (c) DMDO (0.1 M in acetone,
1.8 equiv), CH2Cl2, 25 °C, 6 d (48%); (d) H2, (1 atm), [Ir(cod)py-
(PCy3)]PF6 (0.03 equiv), CH2Cl2, 1 h (93%); (e) LDA (2.2 equiv),
THF, -78 °C, 2 h then TMSCl (3 equiv), -78 to 25 °C, 1 h (90%);
(f) O2/O3, CH2Cl2, -78 °C, 5 min, then 35% H2O2 (12 equiv), 88%
HCO2H (52 equiv), 100 °C, 30 min (97%).
It was anticipated that hydrogenation of epoxide 27
over Pd/C (10%) would proceed via initial reaction at the
C-Br bond, resulting in ring-opening of the epoxide to
give the allylic alcohol 29. Under the reaction conditions
hydrogenation was then expected to occur mainly on the
exo face of allylic alcohol 29 to give the desired alcohol
30. Indeed, formation of alcohol 30 directly from epoxide
27 with H2-Pd/C in methanol did proceed, but only in
modest yield to generate an inseparable mixture of 30
and 7-epi-30 (45% total yield; 30:7-epi-30, 3:2). Interest-
ingly, synthesis and isolation of the allylic alcohol 29
(11) See Supporting Information for details.
(12) Osowska-Pacewicka, K.; Alper, H. J. Org. Chem. 1988, 53, 808-
810.
(13) Fleming, I.; Paterson, I. Synthesis 1979, 736-738.
(14) Kaneda, K.; Kii, N.; Jitsukawa, K.; Teranishi, S. Tetrahedron
Lett. 1981, 22, 2595-2598.
(15) Shiao, M.-J.; Liang, D. Synth. Commun. 1988, 18, 1553-1563.
(16) CCDC 181801 available at http://www.ccdc.cam.ac.uk.
(17) (a) Vedejs, E.; Larsen, S. D. J. Am. Chem. Soc. 1984, 106, 3030-
3032. (b) Vedejs, E.; Ahmad, S.; Larsen, S. D.; Westwood, S. J. Org.
Chem. 1987, 52, 3937-3938. (c) Cain, C. M.; Cousins, R. P. C.;
Coumbarides, G.; Simpkins, N. S. Tetrahedron 1990, 46, 523-544.
(18) Frohn, M.; Wang, Z.-X.; Shi, Y. J. Org. Chem. 1998, 63, 6425-
6426.
(19) Adam, W.; Bialas, J.; Hadjiarapoglou, L. Chem. Ber. 1991, 124,
2377.
(20) Lusinchi, X.; Hanquet, G. Tetrahedron 1997, 53, 13727-13738.
8798 J. Org. Chem., Vol. 69, No. 25, 2004

Stereocontrolled Syntheses of the Nemorensic Acids
(using Zn and NaI in methanol,21 97%), followed by
hydrogenation gave alcohol 30 in improved stereoselec-
tivity (30:7-epi-30, 9:1) and yield (62%). The allylic alcohol
29 could also be obtained directly from allylsilane 21,
using DMDO, in fair yield (48%). To circumvent the
modest yield and stereoselectivity in the synthesis of
alcohol 30, hydroxyl-directed homogeneous hydrogena-
tion of allylic alcohol 29 in the presence of Crabtree’s
catalyst was performed;22 this protocol generated alcohol
30 in excellent yield (93%) as a single isomer whose
relative chemistry was confirmed by X-ray crystal-
lographic analysis.23 With alcohol 30 in hand, simulta-
neous protection of the alcohol and silyl enol ether
formation was effected to give disilyl ether 31 in 90%
yield. Ozonolysis of disilyl ether 31 followed by oxidative
workup with aqueous HCO2H and H2O2 (as used earlier)
gave 4-hydroxy-cis-nemorensic acid 2 in excellent yield
(97%). Spectral data for synthetic 4-hydroxy-cis-nemo-
rensic acid 2 were in accord with those of the natural
isolate.2b
SCHEME 6. Synthesis of 3-Hydroxy-cis-nemorensic
Acid 3a
Hydrogenolytic cleavage of the less-substituted epoxy
C-O bond in the epoxide 32 (derived from the propyne
cycloadduct 6) was initially considered as a potential
entry to 3-hydroxy-cis-nemorensic acid 3 (Scheme 5).
However, prolonged (2 weeks) reaction of epoxide 32 with
Pd/C and H2 mainly resulted in recovered starting
material (52%), with none of the desired tertiary alcohol
33 detected, whereas no reaction was observed with
ammonium formate as the hydrogen source.24 Alternative
methods to open the epoxide also met with no success:
no reaction was observed with a source of nucleophilic
bromide [n-Bu4NBr in the presence of Mg(NO3)2];25 Li-
AlH4 in THF26 only reduced the keto group at room
temperature after 1 h (77%), whereas a complex mixture
was observed at reflux; and LiEt3BH in THF27 at room
temperature similarly gave ketone reduction (3 h) and
prolonged exposure (18 h) led to a multitude of products.
a Reagents and conditions: (a) OsO4 (2.5 mol %), NMO (2 equiv),
acetone, H2O, 25 °C, 5 d (37%); (b) DMDO (0.07 M in acetone, 2
equiv), CH2Cl2, 25 °C, 19 h (89%), then LiAlH4, THF, 0-25 °C, 1
h (89%); (c) (for 36) ethylene glycol (40 equiv), CSA (0.1 equiv),34
CH2Cl2, 25 °C, 44 h (75%); (for 37) 1,3-propanediol (12 equiv),
(MeO)3CH (3 equiv), PTSA (0.08 equiv),35 25 °C, 1 h (83%); (d)
m-CPBA (1.3 equiv), CH2Cl2, 0-25 °C, 18 h [74% for 38 (38a:38b,
5:1); 59% for 39 (39a:39b, 1:1.5)]; (e) LiAlH4 (3 equiv), THF, 0-25
°C, 24 h (quant); (f) 2 M HCl, THF, 25 °C, 3 h (57%, 6% of 7-epi-
33 also isolated); (g) LDA (3.5 equiv), THF, -78 °C, 1 h then
TMSCl (4 equiv), 25 °C, 2 h (97%); (h) (i) O2/O3, CH2Cl2, -78 °C,
5 min, then 88% HCO2H (92 equiv), 35% H2O2 (24 equiv), reflux,
30 min, (ii) TMSCHN2 (5 equiv), hexane-MeOH (4:1), 25 °C, 3 h
(63%); (i) KOH (80 equiv), H2O, 25 °C, 3 h (37%).
Ac2O and PTSA30 led only to substrate degradation.
Dihydroxylation of cycloadduct 7 was expected to lead
to a vic-diol from which selective deoxygenation of the
primary hydroxyl might be possible. In the event, reac-
tion of cycloadduct 7 with OsO4-NMO31 proved sluggish
and led after 5 days to hemiketal 34 (37%), from which
the masked primary hydroxyl could not be coaxed into
reactivity (e.g., with thiocarbonyldiimidazole in toluene).
Finally, we envisaged the synthesis of alcohol 33 via
opening of the epoxide of cycloadduct 7. Epoxidation of
cycloadduct 7 with preformed DMDO (the use DMDO
generated in situ led to substrate degradation) gave an
inseparable epimeric mixture of epoxides (3:7), which
were directly reacted with LiAlH432 giving diols 35 (89%,
35a:35b, 3:7). The relative stereochemistry of the major
diol 35b was established by X-ray crystallographic analy-
sis33 and this indicated that epoxidation had occurred
preferentially on the undesired (for 3-hydroxy-cis-nemo-
rensic acid 3 synthesis) endo face of cycloadduct 7. This
selectivity contrasts with that observed earlier for hy-
drogenation and dihydroxylation of cycloadduct 7 (and
with epoxidation of cycloadducts 5 and 6). Thus, DMDO
SCHEME 5. Initial Strategy to
3-Hydroxy-cis-nemorensic Acid 3
A related strategy to 3-hydroxy-cis-nemorensic acid 3,
which ultimately proved successful, involved manipula-
tion of the more accessible exocyclic double bond of the
allene-derived cycloadduct 7 (Scheme 6). Direct formation
of tertiary alcohol 33 via hydration of cycloadduct 7 with
aqueous HCO2H,28 TFA,29 or AcOH in the presence of
(21) Nicolaou, K. C.; Duggan, M. E.; Ladduwahetty, T. Tetrahedron
Lett. 1984, 25, 2069-2072.
(22) (a) Crabtree, R. H.; Davis, M. W. J. Org. Chem. 1986, 51, 2655-
2661. (b) Hoveyda, A. H.; Evans, D. A.; Fu, G. C. Chem. Rev. 1993, 93,
1307-1370.
(23) CCDC 181906 available at http://www.ccdc.cam.ac.uk.
(24) Dragovich, P. S.; Prins, T. J.; Zhou, R. J. Org. Chem. 1995, 60,
4922-4924.
(25) Suh, Y.-G.; Koo, B.-A.; Ko, J.-A.; Cho, Y.-S. Chem. Lett. 1993,
1907-1910.
(29) Marchand, A. P.; Vidyasagar, V.; Watson, W. H.; Nagl, A.;
Kashyap, R. P. J. Org. Chem. 1991, 56, 282-286.
(30) Dilling, W. L.; Alford, J. A. J. Am. Chem. Soc. 1974, 96, 3615-
3623.
(31) Crimmins, M. T.; Tabet, E. A. J. Org. Chem. 2001, 66, 4012-
4018.
(32) Corey, E. J.; Liu, K. J. Am. Chem. Soc. 1997, 119, 9929-9930.
(33) CCDC 237222 available at http://www.ccdc.cam.ac.uk.
(26) Coates, R. M.; Ho, J. Z.; Klobus, M.; Zhu, L. J. Org. Chem. 1998,
63, 9166-9176.
(27) Krishnamurthy, S.; Schubert, R. M.; Brown, H. C. J. Am. Chem.
Soc. 1973, 95, 8486-8487.
(28) Reddy, N. S.; Goud, T. V.; Venkateswarlu, Y. J. Chem. Res. (S)
2000, 438-439.
J. Org. Chem, Vol. 69, No. 25, 2004 8799

Hodgson et al.
appears to be able to approach both faces of the exocyclic
double bond in cycloadduct 7. Alternatively (or in addi-
tion), intramolecular endo epoxidation might occur via
dioxirane formation at the keto group of cycloadduct 7.
To favor exo-selective epoxidation, we examined the
epoxidation of ketals 36 and 37 using m-CPBA (Scheme
6). The diethyl ketal was also considered; however, this
compound could not be synthesized despite several condi-
tions being examined.36 Pleasingly, epoxidation of ketal
36 gave mainly the desired epoxide 38a (61%, 38a:38b,
5:1). The stereochemical assignment was supported by
NOESY experiments on both isomers, and the structure
of the chromatographically separable minor epoxide
isomer 38b was confirmed by X-ray crystallographic
analysis.37 Surprisingly, epoxidation of ketal 37 gave a
mixture of epoxides 39 (59%), in favor of the undesired
endo-epimer 39b (39a:39b, 2:3). The stereochemistry of
both epoxides was assigned by NOESY experiments.
Epoxide 38a was therefore used to progress toward
3-hydroxy-cis-nemorensic acid 3. Reduction of epoxide
38a efficiently provided tertiary alcohol 40; however,
subsequent deprotection to reveal ketone 33 was found
not to be straightforward due to the propensity for
concomitant epimerization at C-7. Thus, PTSA in acetone
containing 1% water at reflux38 for 24 h, or dilute H2SO4
and SiO239 in CH2Cl2 for 24 h at room temperature, gave
mixtures of ketone 33 and 7-epi-33 (33:7-epi-33, 3:7 and
1:1, respectively). The use of FeCl3 on SiO2,40 CeCl3/NaI
in MeCN,41 or Pd(PhCN)2Cl2 in acetone42 provided com-
plex mixtures in which 33 was a minor component. An
attempt [TMSOTF (2.5 equiv), i-Pr2EtN (2.5 equiv)]43 to
convert the ketal present in tertiary alcohol 40 directly
into an enol ether suitable for oxidative cleavage only
resulted in silylation of the tertiary alcohol. Finally, the
simple use of dilute HCl44 in THF provided, after
complete disappearance of 40 (18 h), a 4:1 mixture of
33:7-epi-33 from which the desired alcohol 33 could be
isolated in 41% yield after chromatography. To minimize
the epimerization of 33, minimal exposure to aqueous
HCl was studied. After 3 h at room temperature, 92%
conversion was observed and epimerization was 10%;
ketone 33 was obtained in 57% yield (62% based on
recovered 40). Ketone 33 was then converted to disilyl
ether 41 in 97% yield. Ozonolysis of disilyl ether 41
followed by oxidative workup with aqueous HCO2H and
H2O2 gave crude 3-hydroxy-cis-nemorensic acid 3. This
crude necic acid 3 was best purified as the dimethyl ester
42 (obtained by esterification with TMSCHN2).45 Saponi-
fication of diester 42 gave 3-hydroxy-cis-nemorensic acid
9 (37%), possessing spectral data consistent with those
of the natural material.2b
The approach used above for the synthesis of cis-
nemorensic acid 1 (and its hydroxylated analogues 2 and
3) relies on efficient facial discrimination (hydrogenation
for 1 and 2) in unsaturated bicyclic cycloadducts to
establish stereochemistry, prior to oxidative cleavage. A
strategy to obtain nemorensic acid 4 (Scheme 7, relative,
not absolute, stereochemistry of 4 shown) was envisaged
in which cleavage of the bicyclic system was carried out
before hydrogenation.
SCHEME 7. Synthesis of Nemorensic Acid 4 from
Allene-Derived Cycloadduct 7a
a Reagents and conditions: (a) LDA (1.2 equiv), THF, -78 °C,
1 h, then TMSCl (2 equiv), -78 to 25 °C, 1 h (97%); (b) (i) DMDO
(1.1 equiv), acetone, CH2Cl2, 0-25 °C, 30 min, (ii) NaIO4 (1.2
equiv), THF, H2O, 25 °C, 30 min; (c) AgNO3 (1.2 equiv), NaOH
(3.4 equiv), EtOH, 25 °C, 30 min; (d) CH3CHN2 (∼3 equiv), Et2O,
0 °C, 18 h (38% from 43); (e) H2 (60 psi), [Ir(cod)py(PCy3)]PF6 (0.05
equiv), CH2Cl2, 25 °C, 18 h (84% of 47a and 1% of 47b); (f) KOH
(17 equiv), H2O, 25 °C, 18 h (92%).
Selective ozonolysis of the silyl enol ether 43 of allene-
derived cycloadduct 7 could not be achieved, and reaction
of 43 under Kaneda and co-workers conditions14 resulted
only in degradation products. However, the more electron
rich double bond of 43 could be selectively reacted with
preformed DMDO46 (1.1 equiv) to give an R-hydroxy-
ketone, which was not isolated but directly treated with
sodium periodate47 to give crude oxoacid 44. Recently,
Zhang and co-workers described carboxylate-directed
stereoselective hydrogenation of cyclic olefin-containing
carboxylic acids in the presence of Wilkinson’s catalyst
and a base such as Et3N.48 Application of this methodol-
ogy to oxoacid 44 gave a complex mixture, including
mainly olefin isomerization (to form endocyclic alkene),
and a small amount of hydrogenated product that could
(34) Flessner, T.; Ludwig, V.; Siebeneicher, H.; Winterfeldt, E.
Synthesis 2002, 1373-1378.
(35) Claeys, S.; Van Haver, D.; De Clercq, P.; Milanesio, M.; Viterbo,
D. Eur. J. Org. Chem. 2002, 1051-1062.
(36) (a) Watanuki, S.; Ochifuji, N.; Mori, M. Organometallics 1995,
14, 5062-5067. (b) Tsunoda, T.; Suzuki, M.; Noyori, R. Tetrahedron
Lett. 1980, 21, 1357.
(37) CCDC 225241 available at http://www.ccdc.cam.ac.uk.
(38) Fetizon, M.; Goulaouic, P.; Hanna, I. J. Chem. Soc., Perkin
Trans. 1 1990, 1107-1110.
(39) Krohn, K.; Frese, P.; Flo¨rke, U. Chem. Eur. J. 2000, 6, 3887-
3896.
(40) Kim, K. S.; Song, Y. H.; Lee, B. H.; Hahn, C. S. J. Org. Chem.
1986, 51, 404-407.
(45) Yuan, P.; Plourde, R.; Shoemaker, M. R.; Moore, C. L.; Hansen,
D. E. J. Org. Chem. 1995, 60, 5360-5364.
(41) Marcantoni, E.; Nobili, F. J. Org. Chem. 1997, 62, 4183-4184.
(42) (a) Lipshutz, B. H.; Pollart, D.; Monforte, J.; Kotsuki, H.
Tetrahedron Lett. 1985, 26, 705-708. (b) Schmeck, C.; Hegedus, L. S.
J. Am. Chem. Soc. 1994, 116, 9927-9934.
(46) Adam, W.; Hadjiarapoglou, L.; Ja¨ger, V.; Klicic, J.; Seidel, B.;
Wang, X. Chem. Ber. 1991, 124, 2361-2368.
(47) Floresca, R.; Kurihara, M.; Watt, D. S.; Demir, A. J. Org. Chem.
1993, 58, 2196-2200.
(43) Gassman, P. G.; Burns, S. J.; Pfister, K. B. J. Org. Chem. 1993,
58, 1449-1457.
(44) Jiang, X.; Covey, D. F. J. Org. Chem. 2002, 67, 4893-4900.
(48) Zhang, M.; Zhu, L.; Ma, X.; Dai, M.; Lowe, D. Org. Lett. 2003,
5, 1587-1589.
8800 J. Org. Chem., Vol. 69, No. 25, 2004

Stereocontrolled Syntheses of the Nemorensic Acids
TABLE 1. Cycloadditions with Diazodione 9, Using Chiral Catalysts 51-55
b
entry
dipolarophile
cycloadduct
catalyst
solvent
temp (°C)
yield (%)
ee (%)a
[R]T
D
1
2
DMAD
17
17
17
17
18
18
5
Rh2(S-DOSP)4
Rh2(R-BNP)4
CF3C6H5
CF3C6H5
CF3C6H5
CF3C6H5
CH2Cl2
toluene
hexane
CH2Cl2
CH2Cl2
hexane
CH2Cl2
toluene
CF3C6H5
CF3C6H5
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
Et2O
20
20
20
20
-78c
-78c
20
20
-10
20
20
20
20
0
61
29
63
65
26
40
61
56
42
18
11
39
44
22
30
17
27
0
4
17
22
51
14
10
23
24
34
10
8
DMAD
-71
+119
+247
+9
3
DMAD
Rh2(S-PTTL)4
Rh2(S-BPTV)4
Rh2(S-DOSP)4
Rh2(S-BPTV)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(R-DDBNP)4
Rh2(R-DDBNP)4
Rh2(S-PTTL)4
Rh2(S-PTTL)4
Rh2(S-BPTV)4
Rh2(S-DOSP)4
Rh2(R-DDBNP)4
Rh2(S-BPTV)4
Rh2(R-DDBNP)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
4
DMAD
5
H2CdO
6
H2CdO
+7
7
BrCH2CtCH
BrCH2CtCH
BrCH2CtCH
BrCH2CtCH
BrCH2CtCH
BrCH2CtCH
BrCH2CtCH
BrCH2CtCH
propyne
+80
+82
+129
+33
+22
-12
-15
-3
8
5
9
5
10
11
12
13
14
15
16
17
18
19
20
5
5
5
2
5
4
5
0
6
0
7
-0.9
+0.7
-0.9
propyne
6
0
6
propyne
6
0
7
allene
7
0
allene
7
0
76
56
45
45
+42
+42
allene
7
0
a Determined by chiral stationary phase chromatography.11 b c ) 0.5-1.0 in CHCl3, T ) 22-25 °C. c Temperature is -78 to 25 °C over
15 min.
SCHEME 8. Synthesis of Nemorensic Acid 4 from
Propyne-Derived Cycloadduct 6a
not be cleanly isolated. Tollens-type oxidation of oxoacid
44 with AgNO3 in aqueous ethanolic NaOH49 provided
the corresponding diacid 45. Attempts to hydrogenate
diacid 45 in the presence Wilkinson’s catalyst and Et3N
were unsuccessful, possibly due to catalyst coordination
at the more accessible (but also more distant from the
alkene) primary carboxylate. Diacid 45 was then esteri-
fied with diazoethane to give unsaturated diester 46,
which was isolated in 38% yield over 4 steps from silyl
enol ether 43. Hydrogenation of unsaturated diester 46
with Pd/C at atmospheric pressure gave saturated diester
47 (76%) as a separable epimeric mixture, which favored
the undesired stereochemistry for nemorensic acid syn-
thesis (47a:47b, 3:7). The stereochemistry of each isomer
was assigned by NOESY experiments. To reverse the
hydrogenation selectivity, ester-directed homogeneous
hydrogenation with Crabtree’s catalyst22 was investi-
gated. Reaction of unsaturated diester 46 with [Ir(cod)-
py(PCy3)]PF6 in CH2Cl2 under hydrogen at atmospheric
pressure for 18 h provided a mixture of starting material
(40%), the product of isomerization of the double bond
into the five-membered ring (35%), and saturated diester
47a as the single product of hydrogenation (25%). To
accelerate the hydrogenation and avoid isomerization of
the double bond, the reaction was performed under 60
psi of hydrogen for 18 h; under these conditions 47a was
obtained in good yield (84%) and with high selectivity
(47a:47b, 66:1). Finally, hydrolysis of saturated diester
47a gave nemorensic acid 4 possessing spectral data in
accord with previous syntheses.3 Nemorensic acid 4 could
also be prepared by an essentially identical reaction
sequence starting from propyne-derived cycloadduct 6
(Scheme 8). Interestingly, ester-directed hydrogenation
of the endocyclic unsaturation in diester 49 was found
to be slower (40 h) than that of the exocyclic olefin in
diester 46. However, the hydrogenation of diester 49
provided exclusively the desired saturated diester 50, in
87% yield.
a Reagents and conditions: (a) as Scheme 7 (94%); (b, c) as
Scheme 7; (d) TMSCHN2 (10 equiv), hexane-MeOH (3:1), 25 °C,
18 h (37% from 48); (e) H2 (60 psi), [Ir(cod)py(PCy3)]PF6 (0.05
equiv), CH2Cl2, 25 °C, 40 h (87%); (f) KOH (17 equiv), H2O, 25 °C,
2 h (89%).
With sterecontrolled access to all the nemorensic acids
1-4 in a racemic manner demonstrated from a common
precursor diazodione 9, we have conducted some prelimi-
nary studies (Table 1) to assess the potential for enan-
tioselective cycloadditions using this precursor with
chiral catalysts 51-55 (Figure 2). These catalysts were
selected because of their previously shown ability to
achieve high levels of asymmetric induction in other
carbonyl ylide cycloadditions of diazocarbonyl com-
pounds.50,51
It should be noted at the outset of the present studies
that asymmetric induction in such cycloadditions is very
sensitive to the substitution pattern of the dipole, as well
(50) (a) Hodgson, D. M.; Stupple, P. A.; Pierard, F. Y. T. M.; Labande,
A. H.; Johnstone, C. Chem. Eur. J. 2001, 7, 4465-4476. (b) Hodgson,
D. M.; Glen, R.; Grant, G. H.; Redgrave, A. J. J. Org. Chem. 2003, 68,
581-586. (c) Hodgson, D. M.; Labande, A. H.; Pierard, F. Y. T. M.;
Exposito Castro, M. A. J. Org. Chem. 2003, 68, 6153-6159. (d)
Hodgson, D. M.; Bru¨ckl, T.; Glen, R.; Labande, A. H.; Selden, D. A.;
Dossetter, A. G.; Redgrave, A. J. Proc. Natl. Acad. Sci. U.S.A. 2004,
101, 5450-5454.
(51) (a) Kitagaki, S.; Anada, M.; Kataoka, O.; Matsuno, K.; Umeda,
C.; Watanabe, N.; Hashimoto, S. J. Am. Chem. Soc. 1999, 121, 1417-
1418. (b) Kitagaki, S.; Yasugahira, M.; Anada, M.; Nakajima, M.;
Hashimoto, S. Tetrahedron Lett. 2000, 41, 5931-5935.
(49) McMurry, J. E.; Dushin, R. G. J. Am. Chem. Soc. 1990, 112,
6942-6949.
J. Org. Chem, Vol. 69, No. 25, 2004 8801

Hodgson et al.
FIGURE 2. Chiral catalysts 51-55.
was added dropwise to the reaction flask, until the distillate
was coming over colorless (∼1 h) [in a separate experiment
the yield of MeCHN2 was determined to be 75% with use of
the method of Wilds and Meander7a]. At this time the tem-
perature of the receiving flask was maintained between -5
and 0 °C for 5 h, then allowed to attain ambient temperature
and stirred for a further 15 h under a slow stream of argon.
The remaining volatiles were then removed in vacuo and the
residue purified by column chromatography (50% EtOAc in
pentane) to give diazodione 9 (1.11 g, 51%) as a yellow oil. Rf
0.40 (50% EtOAc in pentane); IR 2917, 2078, 1719, 1637, 1365,
as the dipolarophile used. An analysis of the limited
studies carried out to date indicate that catalysts 51-
53 work well with electron-deficient carbonyl ylides
(derived from R-diazo-â-ketoesters) and nonpolarized
alkenes and alkynes,50 whereas 54 and 55 are known to
be good catalysts for R-diazoketones and highly electron
deficient dipolarophiles (especially DMAD).51 Initial stud-
ies with diazodione 9 and DMAD reinforced the latter
generalization (Table 1, entries 1-4): the best asym-
metric induction (51%, entry 4) was observed with
Hashimoto’s catalyst 55. With formaldehyde as the
dipolarophile, the frontalin intermediate 18 was gener-
ated in a low yield and asymmetric induction (entries 5
and 6). With use of propargyl bromide as the dipolaro-
phile (entries 7-14), up to 34% ee could be achieved with
Davies’ catalyst 51 (entry 9); propyne was less effective
(entries 15-17). Within the scope of the current study,
and for the dipolarophiles investigated which are relevant
to the nemorensic acid syntheses, allene provided the
most encouraging results (entry 20) with Rh2(S-DOSP)4
51 generating cycloadduct 7 in 76% yield and 45% ee.
This limited investigation of asymmetric cycloadditions
with diazodione 9 underlines the need for a greater
understanding of the factors which influence enantiose-
lective carbonyl ylide cycloadditions,52 and especially the
development of conditions (and catalysts) which are
effective with simple (not electron deficient) carbonyl
ylides and nonpolarized alkenes and alkynes.
1281, 1165, 1082, 1032 cm-1 1H NMR (500 MHz; CDCl3) δ
;
2.72-2.68 (m, 2H), 2.17 (s, 3H), 1.91 (br s, 3H); 13C NMR (125
MHz; CDCl3) δ 207.6, 193.2, 63.6, 37.7, 31.5, 30.4, 8.4; MS
(CI+) m/z (%) 155(29) [M + H]+, 146 (65), 129 (25), 128 (100),
118 (39), 111 (67), 77 (20); HRMS calcd for C7H11N2O2
155.0820, found 155.0821 [M + H]+.
7-Bromomethyl-1,5-dimethyl-8-oxabicyclo[3.2.1]oct-6-
en-2-one (5). Rh2(OAc)4 (14 mg, 0.03 mmol) was added in one
portion to a solution of diazodione 9 (400 mg, 2.60 mmol) and
propargyl bromide (1.10 g, 9.25 mmol, 80% solution in toluene)
in degassed CH2Cl2 (8 mL) in a 500-mL flask (note: the use
of a large reaction vessel is necessary to allow for rapid
evolution of N2) at ambient temperature. The mixture was
stirred at ambient temperature for 3 h and the volatiles
removed in vacuo. The residue was then purified by column
chromatography (20% EtOAc in pentane) to give cycloadduct
5 (539 mg, 84%) as a colorless oil. Rf 0.45 (20% EtOAc in
pentane); IR 2981, 2932, 1723, 1449, 1375, 1065, 938, 912, 735
cm-1; 1H NMR (500 MHz; CDCl3) δ 6.20 (X part of ABX, br s,
1H), 4.04 and 3.91 (AB part of ABX, J ) 11.5, 1.0 Hz, 2H),
2.80 (ddd, J ) 17.6, 8.8, 8.8 Hz, 1H), 2.42 (ddd, J ) 17.6, 8.8,
1.3 Hz, 1H), 2.29 (ddd, J ) 13.8, 8.8, 8.8 Hz, 1H), 1.99 (ddd, J
) 13.8, 8.8, 1.3 Hz, 1H), 1.40 (s, 3H), 1.43 (s, 3H); 13C NMR
(125 MHz; CDCl3) δ 203.7, 142.6, 137.1, 90.6, 84.1, 35.1, 33.0,
23.6, 23.5, 15.1; MS (CI+) m/z (%) 262(19) [M + NH4]+, 184-
(100), 167(90), 77(13); HRMS calcd for C10H17BrNO2 262.0443,
found 262.0446. Anal. Calcd for C10H13BrO2 (245.1): C, 49.00;
H, 5.35. Found: C, 48.79; H, 5.37. Ee determination (Table 1)
used chiral GC (Chirasil Dex-CD capillary column, 120 °C,
tR ∼ 70 min) or (for entry 14) chiral HPLC (Chiralcel OD, 99.5%
hexane/EtOH, 0.2 mL min-1).
1,5,7-Trimethyl-8-oxabicyclo[3.2.1]octan-2-one (23). To
a solution of cycloadduct 5 (539 mg, 2.20 mmol) in MeOH (9
mL) was added 10% Pd/C (130 mg). The reaction vessel was
then evacuated and back-filled with hydrogen three times.
Hydrogen was then bubbled through the solution for 5 min,
the mixture was then allowed to stir at ambient temperature
for 24 h. After this time hydrogen was again bubbled through
the solution for 5 min and the mixture stirred at ambient
temperature for a further 24 h. Upon completion, the reaction
mixture was filtered through a pad of silica gel and the pad
washed well with CH2Cl2, then the solution was made up to
50 mL with CH2Cl2 and washed with water (1 × 30 mL) and
HCl (2 M, 1 × 30 mL). The solution was then dried (MgSO4)
and the volatiles removed by evaporation under argon, and
subsequent column chromatography (15% Et2O in pentane)
Experimental Section
General experimental details have been described.50
6-Diazoheptane-2,5-dione (9). To a solution of levulinic
acid 11 (1.63 g, 14.0 mmol) in Et2O (30 mL) at 0 °C was added
Et3N (1.43 g, 14.1 mmol). Isobutyl chloroformate (1.92 g, 14.0
mmol) was then added dropwise over 5 min to the vigorously
stirred solution. Following addition, the solution was allowed
to attain ambient temperature over 15 min and was then
stirred for a further 2 h. The solution was then filtered under
argon to obtain an ethereal solution of the mixed anhydride,
which was placed in the receiving flask of a clear-fit distillation
apparatus and cooled to -5 °C. In the reaction flask of the
clear-fit distillation apparatus was placed Et2O (33 mL) and
a solution of KOH (8.26 g, 147 mmol) in PrOH (33 mL). The
solution was heated in an oil bath at 55 °C with stirring until
the Et2O began to distil. At this time a solution of N-nitroso-
N-ethylurethane7 (8.26 g, 56.5 mmol) in Et2O (25 mL) was
added dropwise over 5 min to the solution of KOH, such that
boiling was not too vigorous, and the yellow ethereal MeCHN2
was allowed to distill under a very slow stream of argon into
the stirred solution of the mixed anhydride. Et2O (∼60 mL)
(52) Hodgson, D. M.; Pierard, F. Y. T. M.; Stupple, P. A. Chem. Soc.
Rev. 2001, 30, 50-61.
8802 J. Org. Chem., Vol. 69, No. 25, 2004

Stereocontrolled Syntheses of the Nemorensic Acids
gave saturated ketone 23 (340 mg, 92%) as a colorless oil. Rf
0.3 (15% Et2O in pentane); IR 2967, 2935, 1720, 1452, 1375,
1103, 1012, 943, 865 cm-1; 1H NMR (400 MHz; CDCl3) δ 2.51
(ddd, J ) 18.0, 7.6, 1.3 Hz, 1H), 2.33-2.15 (m, 3H), 2.08 (ddd,
J ) 13.2, 8.4, 7.6 Hz, 1H), 1.89 (ddd, J ) 13.2, 8.4, 1.3 Hz,
1H), 1.68-163 (m, 1H), 1.41 (s, 3H), 1.27 (s, 3H), 0.91 (d, J )
6.8 Hz, 3H); 13C NMR (100 MHz; CDCl3) δ 208.5, 89.6, 79.3,
43.3, 43.1, 38.3, 34.7, 26.6, 18.3, 13.4; MS (EI+) m/z (%) 168
(2) [M]+, 140 (7), 97 (13), 84 (33), 49 (76), 43 (100); HMRS calcd
for C10H16O2 (M + H) 169.1228, found 169.1229.
O3 at -78 °C for 15 min. The solution was then purged with
O2 (5 min) and Ar (5 min). The volatiles were then removed
in vacuo and the residue refluxed in the presence of HCO2H
(0.5 mL of an 88% solution in water, 11.5 mmol) and H2O2
(0.25 mL of a 35% solution in water, 2.6 mmol) for 30 min.
Upon cooling the reaction mixture was diluted with brine (1
mL), extracted with EtOAc (3 × 5 mL), and dried (MgSO4)
and the volatiles were removed in vacuo to give cis-nemorensic
acid 1 (82 mg, 96%) as a white solid. Mp 131-133 °C (lit.2a
1
mp for (+)-1 96-100 °C); IR 1708, 1465, 1375, 1130 cm-1; H
Trimethyl(1,5,7-trimethyl-8-oxabicyclo[3.2.1]oct-2-en-
2-yloxy)silane (24). Saturated ketone 23 (150 mg, 0.9 mmol)
in anhydrous THF (1 mL) was added dropwise over 5 min to
a freshly prepared solution of LDA at -78 °C [prepared from
a 1.6 M solution of BuLi in hexane (0.7 mL, 1.1 mmol) and
i-Pr2NH (0.18 mL, 1.3 mmol)] and stirred for 2 h at this
temperature. TMSCl (0.22 mL, 1.8 mmol) was added and the
mixture stirred for 5 min before being allowed to attain
ambient temperature. The solution was stirred for a further
1 h at this temperature, the volatiles were removed in vacuo,
and subsequent column chromatography of the crude material
(silica, 10% Et2O in pentane containing 2% Et3N) gave silyl
enol ether 24 as a colorless oil (194 mg, 90%). Rf 0.5 (10% Et2O
in pentane); IR 2964, 2935, 2892, 2873, 1656, 1453, 1376, 1358,
NMR (500 MHz, CDCl3) δ 2.80 (d, J ) 15.5 Hz, 1H) and 2.84
(d, J ) 15.5 Hz, 1H), 2.49 (ddq, J ) 12.4, 6.8, 6.8 Hz, 1H),
2.09 (dd, J ) 12.4, 6.8 Hz, 1H), 1.87 (dd, J ) 12.4, 6.8 Hz,
1H), 1.55 (s, 3H), 1.39 (s, 3H), 1.13 (d, J ) 6.8 Hz, 3H); 13C
NMR (100 MHz, CDCl3) δ 178.1, 176.0, 87.1, 81.7, 45.5, 44.7,
44.0, 27.7, 24.9, 14.7; MS (CI+) m/z (%) 234 (21) [M + NH4]+,
217 (4) [M + H]+, 190 (25), 171 (100), 127 (22), 52 (24); HRMS
calcd for C10H20NO5 (M + NH4) 234.1341, found 234.1341.
Anal. Calcd for C10H16O5 (206.2): C, 55.55; H, 7.46. Found:
C, 55.63; H, 7.56.
Acknowledgment. We thank the Leverhulme Trust
for support of this work, the EPSRC National Mass
Spectrometry Service Center for mass spectra, and Dr.
Romo de Vivar for providing spectra of 2 and 3.
1252, 1233, 1202, 1177, 1152, 1092, 944, 890, 845 cm-1 1H
;
NMR (400 MHz, CDCl3) δ 4.59 (dd, J ) 4.8, 2.0 Hz, 1H), 2.37
(dd, J ) 16.0, 2.0 Hz, 1H), 2.11 (ddq, J ) 10.4, 8.0, 6.8 Hz,
1H), 1.95 (ddd, J ) 12.0, 10.4, 1.2 (long range to CH3) Hz, 1H),
1.84 (dd, J ) 16.0, 4.8 Hz, 1H), 1.45 (dd, J ) 12.0, 8.0 Hz,
1H), 1.34 (s, 3H), 1.27 (s, 3H), 0.97 (d, J ) 6.8 Hz, 3H), 0.21
(s, 9H); 13C NMR (100 MHz, CDCl3) δ 153.3, 96.5, 83.6, 76.1,
48.9, 45.8, 40.2, 26.9, 18.6, 14.6, 0.2; MS (EI+) m/z (%) 240
(23) [M]+, 197 (20), 157 (17), 84 (56), 73 (83), 49 (100), 43 (84);
HRMS calcd for C13H24O2Si 240.1546, found 240.1547.
Supporting Information Available: Full experimental
details of syntheses and characterization of cycloaddition
substrates and cycloadducts not described in the Experimental
Section and 1H and 13C spectra for all compounds (apart from
those with satisfactory microanalytical data or spectra already
published4a). This material is available free of charge via the
Internet at http://pubs.acs.org.
cis-Nemorensic Acid (1). A solution of silyl enol ether 24
(100 mg, 0.42 mmol) in CH2Cl2 (18 mL) was treated with O2/
JO048446B
J. Org. Chem, Vol. 69, No. 25, 2004 8803