Lewis acid-promoted, stereocontroiled, gram scale, diels-alder cycloadditions of electronically matched 2-pyrones and vinyl ethers: The critical importance of molecular sieves and the temperature of titanium coordination with the pyrone

Article abstract of DOI:10.1021/jo9515900

(R)-(+)-Binol-titanium coordinates with commercial methyl 2-pyrone-3-carboxylate and promotes mild, highly enantiocontrolled Diels-Alder cycloadditions with electron-rich vinyl ether CH₂=CHOCH₂-1-naphthyl and vinyl silyl ether CH

Full text of DOI:10.1021/jo9515900

J . Org. Chem. 1996, 61, 671-676
671
Lew is Acid -P r om oted , Ster eocon tr olled , Gr a m Sca le, Diels-Ald er
Cycloa d d ition s of Electr on ica lly Ma tch ed 2-P yr on es a n d Vin yl
Eth er s: Th e Cr itica l Im p or ta n ce of Molecu la r Sieves a n d th e
Tem p er a tu r e of Tita n iu m Coor d in a tion w ith th e P yr on e
Gary H. Posner,*,§ Haiyan Dai,†,§ D. Scott Bull,§ J ae-Kyoo Lee,§ Franck Eydoux,§
Yuji Ishihara,‡,§ William Welsh,| Neil Pryor,| and Stan Petr, J r.|
Department of Chemistry, The J ohns Hopkins University, Baltimore, Maryland 21218, and
Grace Davison Division, W.R. Grace & Co., Columbia, Maryland 21044
Received August 31, 1995X
(R)-(+)-Binol-titanium coordinates with commercial methyl 2-pyrone-3-carboxylate and promotes
mild, highly enantiocontrolled Diels-Alder cycloadditions with electron-rich vinyl ether CH2dCHOCH2-
1-naphthyl and vinyl silyl ether CH2dCHOSiMe2Bu-t leading to valuable 1R,25-dihydroxyvitamin
D3 (calcitriol) intermediate (-)-2. Unexpectedly, two subtle variables were found to be critical for
obtaining reproducibly over 90% enantioselectivities in gram scale cycloadditions: (1) the moisture
content (15-17% is best) of the molecular sieves used to prepare the binol-titanium complex
according to the Mikami protocol and (2) the temperature (50 °C is best) at which the pyrone ester
is mixed with the binol-titanium complex. Unsubstituted 2-pyrone undergoes ytterbium-promoted,
high-pressure, regioselective, and stereoselective Diels-Alder cycloaddition with benzyl vinyl ether
to form versatile bicyclic lactone (()-4 on gram scale.
We and others have been studying how to coax flat,
stereochemically uninteresting, semiaromatic and thus
normally unreactive 2-pyrone dienes into Diels-Alder
cycloadditions and how to control the absolute stereo-
chemistry of such cycloadditions to form synthetically
useful, versatile, bicyclic lactone adducts.1 Such bicyclic
adducts can be converted into various enantiopure and
biologically active compounds (e.g., 1R,25-dihydroxyvita-
min D3 and various analogs).2 To achieve high stereo-
chemical results, chiral auxiliaries have been incorpo-
rated separately into the dienophiles,3 into the pyrone
diene,4 and into the pyrone diene plus into the Lewis acid
activator of the (4 + 2)-cycloaddition process.5 Recently,
in preliminary fashion, we reported an even better
solution: use of a stereocontrol element on only the Lewis
acid. In this way, using titanium complexes derived from
chiral, nonracemic tartrates6 and especially binaphthol,7
Diels-Alder cycloadditions were achieved with electroni-
cally matched 2-pyrones and various dienophiles to form
bicycloadducts on milligram scale in over 90% chemical
yields and in over 95% enantiomeric purities. A major
review of vitamin D synthesis refers to this approach as
“highly attractive”.2c Because this convenient approach
uses readily available reactants and proceeds so favorably
toward important vitamin D3 intermediates, it was
desirable to scale-up the procedures to form gram quanti-
ties of the cycloadducts. In attempting to do so, it quickly
became clear that scale-up was unexpectedly complicated
and difficult. In working out the surprising subtleties
of scale up, we have discovered that these stoichiometric
binaphthol-titanium complex-promoted (4 + 2)-cycload-
ditions are very sensitive, especially to two parameters:
(1) the nature and moisture content of the molecular
sieves used to prepare the complex and (2) the temper-
ature of the initial coordination of the chiral titanium
species with the pyrone diene. We report here details
for performing these asymmetric induction Diels-Alder
cycloadditions on gram scale with reproducibly high
enantioselectivities (Scheme 1). We report also details
for gram scale, Lewis acid-promoted cycloadditions of the
unsubstituted parent 2-pyrone leading to racemic bicyclic
lactone adducts valuable as intermediates for preparation
of physiologically potent vitamin D3 analogs.2
† Current address: Procter & Gamble Pharmaceuticals, Cincinnati,
OH 45253.
‡ Current address: Takeda Chemical Industries, Osaka, J apan.
§ The J ohns Hopkins University.
| W. R. Grace & Co.
X Abstract published in Advance ACS Abstracts, December 15, 1995.
(1) For reviews: see (a) Afarinkia, K.; Vinader, V.; Nelson, T. D.;
Posner, G. H. Tetrahedron 1992, 48, 9111. (b) Kvita, V.; Fischer, W.
Chimia 1993, 47, 3. (c) Marko, I. E. Organometallic Reagents in
Organic Synthesis; Academic Press: New York, 1994; Chapter 2, p 44.
(d) Kalinin, V. N.; Shilova, O. S. Russ. Chem. Rev. 1994, 63, 661.
(2) For reviews, see: (a) Wilson, S. R.; Yasmin, A. In Studies in
Natural Products Chemistry, Vol. 10, Stereoselective Syntheses; Ur-
Raman, Ed.; Elsevier Science Publishers: Amsterdam, 1992, p 43. (b)
Dai, H.; Posner, G. H. Synthesis 1994, 1383. (c) Zhu, G.-D.; Okamura,
W. H. Chem. Rev. 1995, 95, 1877.
Resu lts a n d Discu ssion
Dienophiles CH2dCHOR were chosen so that, after
cycloaddition, facile transformation of the OR protecting
group into a free OH group could be achieved, thereby
representing the synthetic equivalent of using
CH2dCHOH as dienophile.8 Among the benzylic vinyl
ethers used (including benzyl vinyl ether, 1-naphthylm-
ethyl vinyl ether, fluorenyl vinyl ether, and diphenylm-
ethyl vinyl ether), the highest asymmetric inductions
were achieved with 1-naphthylmethyl vinyl ether; one
added advantage of this dienophile is that it is consider-
(3) (a) Posner, G. H.; Wettlaufer, D. G. Tetrahedron Lett. 1986, 27,
667. (b) Posner, G. H.; Wettlaufer, D. G. J . Am. Chem. Soc. 1986, 108,
7373. (c) Posner, G. H.; Kinter, C. M. J . Org. Chem. 1990, 55, 3967.
(4) Unpublished results of G. H. Posner and T. E. N. Anjeh.
(5) Posner, G. H.; Carry, J .-C.; Anjeh, T. E. N.; French, A. N. J . Org.
Chem. 1992, 57, 7012. For some subsequent examples involving
diastereocontrolled 2-pyrone cycloadditions, see: Marko, I. E.; Evans,
G. R. Synlett 1994, 431; Tetrahedron Lett. 1994, 35, 2767.
(6) Posner, G. H.; Carry, J .-C.; Lee, J .-K.; Bull, D. S.; Dai, H.
Tetrahedron Lett. 1994, 35, 1321.
(8) For related, highly enantiocontrolled cycloadditions using steri-
cally hindered (e.g. adamantyl) vinyl ethers, see: Marko, I. E.; Evans,
G. R. Tetrahedron Lett. 1994, 35, 2771.
(7) Posner, G. H.; Eydoux, F.; Lee, J . K.; Bull, D. S. Tetrahedron
Lett. 1994, 35, 7541.
0022-3263/96/1961-0671$12.00/0 © 1996 American Chemical Society

672 J . Org. Chem., Vol. 61, No. 2, 1996
Posner et al.
Sch em e 1
fore in titanium-promoted inverse electron-demand (4 +
2)-cycloadditions with commercial methyl 2-pyrone-3-
carboxylate. Diels-Alder cycloadditions with 1-naphth-
ylmethyl vinyl ether directed by a binol-titanium com-
plex in the presence of molecular sieves according to the
general protocols of Narasaka13 and Corey14 gave bicy-
cloadducts in no more than 60% ee. Preparation of the
binol-titanium complex according to the protocol of
Mikami,10a however, in which 4 Å molecular sieves are
used only to assemble the complex but then are removed,
gave 95-98% enantioselectivities in small scale (10 mg)
reactions.6 One troubling aspect of this process, however,
was the lack of stereochemical reproducibility. As the
Mikami group had reported,10a their enantiomeric ex-
cesses varied over a considerable range. While we
achieved very high enantioselectivity (i.e., 95% ee) oc-
casionally and good enantioselectivity (g90% ee) regu-
larly, enantiomeric excesses as low as 70% were some-
times observed. The preparation of the complex is a
multistep process involving reaction of TiCl2(OPr-i)2 with
binol in the presence of 4 Å molecular sieves, centrifuga-
tion to remove the molecular sieves, and washing of the
binol-titanium complex with a solvent capable of dis-
solving the achiral TiCl2(OPr-i)2 without dissolving the
binol-titanium complex. The washing stage was found
to be important in that unreacted TiCl2(OPr-i)2, which
can promote the cycloaddition in an achiral fashion,6
must be completely removed in order to achieve consis-
tently high enantioselectivity. Another major factor
contributing to reproducibly high enantioselectivity is the
molecular sieves. Surface chemistry/morphology and
control of the moisture content within the sieves proved
to be extremely significant variables.
The molecular sieves used in this process (4 Å, type
A) have the general composition Na12[(AlO2)12(SiO2)12]‚
nH2O.15a Mikami and co-workers proposed the action of
the sieves in this process to be as shown in eq 1, in which
the asterisk represents the binol moiety.10a
ably easier to prepare than benzyl vinyl ether. Among
the enol silyl ethers CH2dCHOSiR3 used [including
OSiMe3 (0% ee), OSiPh2Bu-t (23% ee), OSiMe2Bu-t (92%
ee)], the tert-butyldimethylsilyl enol ether CH2dCHOΣ
gave the highest asymmetric induction. Among the
solvents tried (CH2Cl2, mesitylene, toluene) for the cy-
cloadditions in Scheme I, toluene was the best.
1,1′-Bi-2-naphthol (binol) stands out as one of the most
effective C2-symmetric, chiral, nonracemic ligands useful
for a variety of metal ion-mediated asymmetric reac-
tions.9 This diol ligand is now commercially available
as the (R)- as well as the (S)-enantiomer. Binol-metal
complexes, prepared in various ways, have been reported
to promote normal electron-demand Diels-Alder
cycloadditions.10-12 We used this chiral auxiliary there-
The apparent simplicity of this process is complicated
by variables present during the manufacture of the
molecular sieves. While all 4 Å sieves have the same
general composition, differences in trace elements and
in crystallinity exist. We became aware of the effect of
these “minor” variations when a new batch of 4 Å sieves
(9) Noyori, R. Acc. Chem. Res. 1990, 23, 345. (b) Noyori, R.
CHEMTECH. 1992, 360. (c) Noyori, R. Asymmetric Catalysis in
Organic Synthesis; J ohn Wiley: New York, 1994. (d) For a general
review, see: Rosini, C.; Franzini, L.; Rafaelli, A.; Salvadori, P. Synthesis
1991, 503.
(12) For use of binol-titanium complexes in other enantiocontrolled
reactions, see: (a) Sasai, H.; Arai, T.; Shibasaki, M. J . Am. Chem. Soc.,
1994, 116, 1571; b) Sasai, H.; Kim, W.-S.; Shibasaki, M.; Mitsuda, M.;
Hasegawa, J .; Ohashi, T. Tetrahedron Lett. 1994, 35, 6123. (c) Terada,
M.; Motoyama, Y.; Mikami, K. Ibid. 1994, 35, 6693. (d) Keck, G. E.;
Krishnamurthy, D.; Chen, X. Tetrahedron 1994, 35, 8323. (e) Keck,
G. E.; Li, X.-Y.; Krishnamurthy, D. J . Org. Chem. 1995, 60, 5998.
(13) Narasaka, K.; Iwasawa, N.; Inoue, M.; Yamada, T.; Nakashima,
M.; Sugamori, J . J . Am. Chem. Soc. 1989, 111, 5340.
(14) Corey, E. J .; Matsumura, Y. Tetrahedron Lett. 1991, 32, 6289.
(15) (a) For a leading reference on molecular sieves, see: Breck, D.
W. Zeolite Molecular Sieves; J ohn Wiley: New York, 1974. (b) See
also: Thomas, J . Chem. Rev. 1993, 93, 301. (c) Davis, M. E. Acc. Chem.
Res. 1993, 26, 111. (d) An increase in diastereoselectivity with
increasing sieve hydration has been observed: Tottie, L.; Baeckstrom,
P.; Moberg, C.; Tegenfeldt, J .; Heumann, A. J . Org. Chem. 1992, 57,
6579.
(10) (a) Mikami, K.; Motoyama, Y.; Terada, M. J . Am. Chem. Soc.
1994, 116, 2812. (b) Maruoka, K.; Itoh, T.; Shirasaka, T.; Yamamoto,
H. Ibid. 1988, 110, 310. (c) Seebach, D.; Beck, A. K.; Imwinkelried, R.;
Roggo, S.; Wonnacott, A. Helv. Chim. Acta 1987, 70, 954. (d) See also:
Reetz, M. T.; Kyang, S.-H.; Bolm, C.; Zienke, T. Chem. Ind. 1986, 824.
(11) For use of other binol-lanthanide complexes in enantiocon-
trolled (4 + 2)-cycloadditions, see: (a) Koboyashi, S.; Araki, M.;
Hachiya, I. J . Org. Chem., 1994, 59, 3758; b) Koboyashi, S.; Ishitani,
H. J . Am. Chem. Soc. 1994, 116, 4083. (c) Marko, I. E.; Evans, G. R.
Tetrahedron Lett. 1994, 35, 2771. (d) Marko, I. E.; Evans, G. R.;
Declercq, J .-P.; Feneau-Dupont, J .; Tinant, B. Bull. Soc. Chim. Belg.
1994, 103, 295. Marko, I. E.; Evans, G. R.; Declercq, J .-P. Tetrahedron
1994, 50, 4557. (e) See also: Ishihara, K.; Yamamato, H. J . Am. Chem.
Soc. 1994, 46, 1561.

Diels-Alder Cycloadditions 2-Pyrones and Vinyl Ethers
J . Org. Chem., Vol. 61, No. 2, 1996 673
F igu r e 1. Enantioselectivity of cycloadditions vs hydration
of molecular sieves used to prepare the binol-titanium
complex.
F igu r e 2. Enantioselectivity of cycloadditions vs temperature
of mixing pyrone ester plus binol-titanium complex.
terion is related to hydration. As moisture content
increases, the sodium ions move away from the molecular
sieve surface toward the center of the pore. Additionally,
the hydrogen bonding of the binol with the surface may
be attenuated at higher moisture contents, thus allowing
the hydroxyl groups to be more available for complexation
with titanium. Both of these possibilities are consistent
with the Mikami model.10a
Ta ble 1. Ch a r a cter istics of Da vison 4 Å Molecu la r
Sieves
crystallinity
causticity (% as Na2O)
100%
0.57
chem compstn (%)
Al2O3
Na2O
CaO
MgO
Fe2O3
SO4
K2O
SiO2
36.112
22.815
0.061
0.010
0.044
0.035
0.208
41.81
After very considerable trial and error in scale-up over
a period of several months, it became clear that an
additional but unexpected variable is crucial: the tem-
perature at which the pyrone ester was mixed with the
binol-titanium complex. Figure 2 graphically sum-
marizes the dramatic and unanticipated sensitivity of the
cycloadduct’s enantiomeric purity to the temperature of
mixing the pyrone ester with the binol-titanium com-
plex. The data in Figure 1 are for room temperature
mixing of the pyrone ester with the binol-titanium
complex. Generally, 2-4 h of mixing at 50 °C was found
to be optimal in gram scale cycloadditions. Unfortu-
nately, 1H NMR spectroscopy was not helpful in correlat-
ing the degree of pyrone ester coordination to the binol-
titanium complex at various temperatures with the
enantioselectivity of the cycloaddition. Thus, although
this temperature effect of mixing the pyrone ester with
the binol-titanium complex was reproducible and reli-
able at 50 °C in giving highly enantiocontrolled cycload-
ditions, unfortunately it has not been possible to develop
a clear understanding of the relationship between the
structure(s) of the binol-titanium-pyrone complex(es)
and the enantioselectivity of cycloaddition.
With these two major variables now under control,
using Davison 4 Å molecular sieves of 15-17% moisture
content to prepare the binol-titanium complex and
mixing this complex, freed from molecular sieves, with
methyl 2-pyrone-3-carboxylate at 45-50 °C for 2-4 h
allowed cycloaddition to occur as in Scheme 1 leading
finally to gram quantities of the desired bicycloadducts
in 59-83% chemical yields and, most significantly,
reproducibly with 92-93% enantioselectivity (Scheme 1).
Although only four steps are required to prepare chiron
(-)-2 using the vinyl silyl ether as dienophile, this
pathway gave more variable enantioselectivity than did
the five-step pathway using the naphthylmethyl vinyl
ether as dienophile. Up to 80% of binaphthol of un-
changed rotation was recovered after the cycloadditions.
It is worth emphasizing that the absolute stereochemistry
of the cycloadducts in these gram scale Diels-Alder
particle size distribtn
90 vol % less than 22.8 µm
50 vol % less than 11.5 µm
10 vol % less than 1.7 µm
failed to produce a chiral complex capable of exceeding
60% ee in our standard Mikami-based cycloaddition. By
screening many different batches of 4 Å sieves from
different manufacturers, several batches were identified
that could produce highly efficient chiral titanium pro-
moters. Commercially available Davison 4 Å molecular
sieves, having the characteristics shown in Table 1 and
giving the most reliable results, were used in our scale-
up reactions.
Once a single batch of sieves was chosen for optimiza-
tion, conditioning became a concern. Since the initial
titanium species and the resulting chiral complex are
both moisture sensitive, it was reasonable to assume that
the sieves should contain as little water as possible.
Therefore, our initial sieve drying protocol was standard-
ized at a temperature of 175 °C for 16 h. Unfortunately,
variable enantioselectivity was still a problem. The
molecular sieve’s moisture content was then precisely
controlled through the entire range of hydration (i.e.,
2-22% by weight). Suprisingly, a maximum in enanti-
oselectivity was observed in the range of 16-17% mois-
ture content (Figure 1); a similar dependence on moisture
content (15-17% being best) was observed also for the
enantioselectivity of enol tert-butyldimethsilyl ether cy-
cloaddition with methyl 2-pyrone-3-carboxylate. As far
as we know, such an observation on the advantage of
using molecular sieves of specific moisture content for
maximum stereocontrol has been reported only once, in
palladium-catalyzed diastereocontrolled cyclizations.15d
This favorable effect of relatively high moisture content
initially seemed counterintuitive, but it might be under-
standable because the accessibility of the sodium coun-

674 J . Org. Chem., Vol. 61, No. 2, 1996
Posner et al.
P r ep a r a tion of th e Molecu la r Sieves. Davison molec-
ular sieves were hydrated fully (22% water content) by
exposing them to air in a Petri dish for 1-2 days at room
temperature. Partial dehydration was achieved by heating the
sieves for different times at 105 °C in a drying oven. Typically
about 2-3 h at 105 °C was needed to achieve 15-17% water
content. Monitoring the water content of a given batch of
molecular sieves was done by heating a weighed portion of
these sieves in a commercial electric Bunsen burner (VWR)
at 950 °C to remove all moisture and to collapse the mi-
croporous structure, thus preventing rehydration upon han-
dling in air. The weight loss after this procedure represents
the amount of water originally in this batch of sieves.
reactions (Scheme 1) is consistent with the Seebach
mechanistic generalization for titanium-promoted cy-
cloadditions of bidentate ligands such as methyl 2-py-
rone-3-carboxylate.16a When only 0.1 equiv instead of the
usual 1.3-1.4 equiv of the titanium complex was used,
surprisingly but reproducibly the mirror image cycload-
ducts were formed with ee’s values in the 50-60% range;
this stereochemical reversal may be due to a change from
bidentate to monodentate coordination of the titanium
Lewis acid to the pyrone ester when excess pyrone ester
is present.
Scale-up of our recently reported cycloaddition of
unsubstituted 2-pyrone with benzyl vinyl ether17 to
produce racemic cycloadduct (()-4 was much less trouble-
some. Multigram cycloaddition was achieved smoothly
as in eq 2, with methanolysis of the initially formed
bicyclic lactone (()-4 proceeding at -10 to -20 °C in a
substantially improved 81% yield.17 Trisubstituted cy-
clohexene (()-5 is a useful A-ring synthon for construc-
tion of 1R,25-dihydoxyvitamin D3 analogs, including
unnatural but biologically active 1â-hydroxy versions.2
P r ep a r a tion of Bin ol-Tita n iu m Com p lex. R-(+)-Bi-
naphthol (5.0 g, 17.5 mmol, Aldrich, 98% ee) and 4 Å Davison
molecular sieves (MS, 77.5 g) that had been dehydrated to
16.5% moisture were dissolved/suspended in 500 mL of CH2-
Cl2 under argon. After 1 h of stirring at room temperature,
19
Ti(OPr-i)2Cl2 (4.1 g, 17.5 mL of a 1 M solution in toluene)
was added over 10 min via argon-flushed syringe. The slurry
immediately changed from white to orange and finally to dark
red-brown. Stirring was continued for a total of 2 h, after
which the reaction mixture was cannulated into argon-filled
septum-capped 150 mL centrifuge tubes. The MS were then
sedimented via centrifugation at 5000 rpm for 1 h. The
supernatant was then cannulated into a dry 1 L round bottom
flask, and the solvent was removed under reduced pressure
(0.1 mmHg) with gentle heating. Once the volume was
reduced to ∼25 mL, the solution was cannulated into a dry 30
mL centrifuge tube and again centrifuged at 5000 rpm for 20
min to ensure complete removal of the MS. Following careful
cannulation of the supernatant into a septum-capped 250 mL
flask, the CH2Cl2 was completely removed. The flask was then
opened under argon, and the dark red-brown solid was
thoroughly scraped from the flask wall with a spatula; this
scraping process is essential for optimal enantioselectivity in
the cycloaddition step. The solid product was then suspended
via magnetic stirring in 200 mL of anhydrous pentane. After
15 min of stirring, the suspension was cannulated into a
septum-capped Buchner funnel and washed with an additional
500 mL of pentane and dried at reduced pressure (0.1 mmHg)
for 16 h. The chiral titanium species was obtained in 78% yield
(5.9 g).
Coor d in a tion of Meth yl 2-P yr on e-3-ca r boxyla te w ith
th e Bin ol-Tita n iu m Com p lex. Methyl 2-pyrone-3-carboxy-
late (500 mg, 3.2 mmol, Aldrich, sublimed, mp 74.5 °C) was
added to a dry argon-filled 250 mL flask and dissolved in 50
mL of anhydrous toluene. Separately, the chiral binol-
titanium dichloride species (1.97 g, 4.4 mmol) was dissolved
in 50 mL of toluene. Upon dissolution of the pyrone, it was
equilibrated to 50 °C and the chiral titanium species solution
was cannulated into the vigorously stirring pyrone solution.
Stirring at 50 °C was maintained for 4 h.
Cycloa d d ition w ith 1-Na p h th ylm eth yl Vin yl Eth er .
The above heterogeneous mixture was cooled to -30 °C, and
the 1-naphthylmethyl vinyl ether (3.0 g, 16.3 mmol)19 was
added via an argon-flushed syringe at a rate slow enough to
maintain -30 °C (∼5 min). The reaction was allowed to
proceed for 12 h, after which it was quenched with 100 mL of
saturated NaCl. The reaction mixture was extracted with two
additional aliquots of brine followed by 2 × 100 mL of H2O.
The aqueous extracts were extracted with 3 × 100 mL of CH2-
Cl2. The organic layers were combined, dried with MgSO4,
filtered, concentrated, and purified via column chromatogra-
phy using 0-25% EtOAc in hexanes as the eluent to yield 910
mg (83% yield) of the cycloadduct (+)-1a as a solid: mp 110.1-
111.0 °C: 1H NMR (CDCl3) δ 1.67 (dt, J ) 14.0, 1.6 Hz, 1H),
2.48 (ddd, J ) 13.6, 7.6, 3.6 Hz, 1H), 3.52 (s, 3H), 4.53 (dt, J
) 7.2, 1.6 Hz, 1H), 4.85/4.98 (AB, J )12.0 Hz, 2H), 5.22 (m,
1H), 6.60 (dd, J ) 7.6, 5.2 Hz, 1H), 6.83 (dt, J ) 7.6, 1.6 Hz,
1H), 7.37-7.44 (m, 2H), 7.48-7.55 (m, 2H), 7.83-7.88 (m, 2H),
7.98-8.00 (m, 1H); 13C NMR (CDCl3) δ 35.0, 52.6, 61.3, 70.3,
71.3, 74.1, 124.1, 125.1, 125.9, 126.3, 127.4, 128.5, 129.3, 129.7,
Thus, the art of organic synthesis has been pictured
here in some of its subtle complexity within the frame-
work of stereocontrolled, atom-economical Diels-Alder
cycloadditions. These results forcefully emphasize the
experimental nature of organic chemistry, even though
great advances in theoretical understanding and in
computer-assisted design have been made. For the
increasing number of chemists using molecular sieves in
organic reactions15 and the many now using chiral,
nonracemic Lewis acids for asymmetric Diels-Alder
cycloadditions,18 these results serve as a warning about
the importance of seemingly minor experimental vari-
ables, such as the moisture content of molecular sieves,
for optimizing chemical reactions. The unanticipated
difficulties in finding the best experimental conditions
for these gram scale cycloaddition reactions strongly
emphasize the importance of demonstrating any new
synthetic method not only on milligram scale but also
on at least gram scale as a measure of its practical utility.
Exp er im en ta l Section 3b
The purity of products was judged to be at least 95% on the
basis of their chromatographic homogeneity.
(16) (a) Seebach, D.; Dahinden, R.; Marti, R. E.; Beck, A. K.;
Plattner, D. A.; Ku¨hnle, F. N. M. J . Org. Chem. 1995, 60, 1788. (b)
For an important optimization study, see: Irurre, J .; Alonso-Alija, C.;
Fernandez-Serrat, A. Afinidad 1994, 51, 413.
(17) Posner, G. H.; Ishihara, Y. Tetrahedron Lett. 1994, 35, 7545.
(18) (a) Taschner, M. J . Asymmetric Diels-Alder Reactions in
Organic Synthesis: Theory and Applications; Hudlicky, T., Ed.; J AI
Press Inc.: Greenwich, CT, 1989; Vol. 1, pp 1-101. (b) Helmchen, G.
Asymmetric Diels-Alder Reactions with Chiral Enoates as Dienophiles
in Modern Synthetic Methods; Scheffold, R., Ed.; Springer Verlog: New
York, 1986; Chapter 2. (c) Roush, W. R. Adv. Cycloaddit. 1990, 2, 91.
(d) For recent reviews, see: (a) Oh, T.; Reilly, M. Org. Prep. Proc. Int.
1994, 26, 131. Waldmann, H. Synthesis 1994, 535.
(19) Burgstahler, W. W.; Gibbons, L. K.; Nordin, I. C. J . Chem. Soc.
1963, 24, 4986.

Diels-Alder Cycloadditions 2-Pyrones and Vinyl Ethers
J . Org. Chem., Vol. 61, No. 2, 1996 675
130.6, 131.9, 132.4, 133.7, 167.3, 168.6; HRMS calcd for
C20H18O5 338.1154, found 338.1158. Anal. Calcd for
C20H18O5: C, 71.00; H, 5.36. Found: C, 70.95; H, 5.36. The
ratio of enantiomers was determined by chiral HPLC (Daicel
Chiralpak AS, 250 mm × 4.6 mm) using 90:10 hexane:ethanol
as the eluent. The minor enantiomer eluted at 16.7 min
followed by the major enantiomer at 25.8 min. Recrystalli-
zation from ethanol gave a white solid, mp 110.9-112.4 °C.
Preparative chiral HPLC gave the major enantiomer, mp 112.1
solution. Upon completion of the reaction (∼45 min), the
reaction mixture was filtered through Celite, which was then
rinsed with ∼75 mL of EtOAc. After concentration of the
filtrate, the cyclohexenediol ester was purified by column
chromatography on silica gel (EM Science silica gel 60) using
a gradient of 80-100% EtOAc in hexane. The yield of the
white solid diol was 70% (217 mg, 1.26 mmol): 1H NMR
(CDCl3) δ 1.79 (ddd, J ) 13.2, 10.4, 4.8 Hz, 2H), 2.09-2.15
(m, 2H), 2.19 (ddd, J ) 8.4, 2.8, 1.2 Hz, 1H), 2.64 (dt, J ) 19.2,
5.2 Hz, 1H), 3.78 (s, 3H), 4.18-4.25 (m, 1H), 4.72 (m, 1H), 6.97
(dd, J ) 5.2, 2.8 Hz, 1H); 13C NMR (CDCl3) δ 34.8, 38.6, 51.9,
63.2, 64.3, 131.8, 139.8, 167.2; IR (CHCl3, cm-1) 3606, 3454,
°C; [R]30 +13.9° (c ) 1.8, CHCl3).
D
P r ep a r a tion of Cycloh exen e (-)-2 fr om Cycloa d d u ct
(+)-1a . To 910 mg (2.7 mmol) of bicyclic lactone (+)-1a in 20
mL of CH2Cl2 at 0 °C was added 5.4 mL (5.4 mmol) of a freshly
prepared 1.0 M solution of lithium allyloxide in allyl alcohol.
After completion of the reaction (∼2 h), the reaction was
quenched with 10 mL of aqueous NH4Cl and the solution was
extracted with CH2Cl2. The organic layer was dried with
MgSO4, filtered, and concentrated on a rotary evaporator. The
crude product was purified by column chromatography using
0-20% EtOAc in hexane as the eluent. The purified mixed
malonate product was obtained in 85% yield (912 mg, 2.3
mmol): 1H NMR (CDCl3) δ 1.83 (ddd, J ) 13.6, 10.0 1.6 Hz,
1H), 2.54 (dtd, J ) 13.6, 4.4, 0.8 Hz, 1H), 3.36 (s, 3H), 4.44
(m, 1H), 4.58 (m, 4H), 4.91/5.08 (AB, J ) 12.4 Hz, 2H), 5.24
(ddq, J ) 17.6, 10.0, 1.2 Hz, 2H), 5.82 (m, 1H), 6.02 (ddt, J )
26.8, 10.0, 2.0 Hz, 2H), 7.39-7.54 (m, 4H), 7.79-7.86 (m, 2H),
8.00 (dd, J ) 8.4, 1.2 Hz, 1H); 13C NMR (CDCl3) δ 33.2, 52.9,
59.5, 64.0, 66.5, 67.4, 70.1, 76.0, 118.9, 123.4, 124.4, 125.6,
126.2, 127.1, 128.9, 129.1, 131.6, 132.0, 133.5, 134.0, 135.2,
1702; mp 112.5-113.5 °C; [R]30 -88.7° (c ) 1.5, CHCl3);
D
HRMS calcd for C8H12O4 (M - H2O) 154.0630, found 154.0631.
Anal. Calcd for C8H12O4: C, 55.80, H, 7.02. Found: C, 55.57,
H, 6.98.
To the cyclohexenediol ester (217 mg, 1.26 mmol) in 20 mL
of CH2Cl2 was added 910 µL (7.8 mmol) of 2,6-lutidine. To
this stirred solution was added tert-butyldimethylsilyl triflate
(1200 µL, 5.2 mmol). The reaction was allowed to proceed for
20 h after which only a trace of starting material was evident
by TLC. The reaction was worked up by dilution in CH2Cl2,
followed by washing the organic layer with 5% HCl and
deionized water. The organic phase was then dried with
MgSO4, filtered, and concentrated. Purification was ac-
complished by column chromatography using silica eluted with
a gradient of 0-5% EtOAc in hexane, which yielded 98% (495
mg, 1.24 mmol) of the bis-protected diol (-)-2 as a yellow oil:
1H NMR (CDCl3) δ 0.05 (s,3H), 0.07 (s, 6H), 0.14 (s, 3H), 0.86
(s, 9H), 0.89 (s, 9H), 1.58 (dd, J ) 13.2, 4.0 Hz, 1H), 1.96 (dtd,
J ) 12.9, 2.5, 1.4 Hz, 1H), 2.08 (ddd, J ) 19.2, 9.2, 2.5 Hz,
1H), 2.55 (dtd, J ) 19.2, 5.5, 1.3 Hz, 1H), 3.73 (s, 3H), 4.20
(m, 1H), 4.76 (bt, J ) 3.0 Hz, 1H), 6.88 (dd, J ) 5.4, 2.8 Hz,
1H); 13C (CDCl3) δ -4.8 (2C), -4.6 (2C), 18.0, 25.8 (3C), 25.9
(3C), 36.1, 41.3, 51.5, 63.4, 64.9, 132.4, 139.8, 166.9; IR (CHCl3,
168.2, 169.3; [R]30 +97.2° (c ) 8.6 CHCl3).
D
The mixed malonate (900 mg, 2.3 mmol), formic acid (150
µL, 2.8 mmol), triethylamine (410 µL, 2.9 mmol), palladium-
(II) acetate (12.7 mg, 0.06 mmol),20 triphenylphosphine (48.7
mg, 0.19 mmol), and 10 mL of dioxane were added to a 25 mL,
three-neck flask fitted with a reflux condenser. The mixture
was heated to reflux (105 °C). The reaction mixture turned
black after ∼25 min, and heating was continued for a total of
23 h. Following cooling and concentration via rotary evapora-
tor, 10 mL of 1.0 N HCl was added and the mixture was
extracted with CH2Cl2 until the extract was colorless. The
organic layers were combined, washed with saturated NaH-
CO3, dried with MgSO4, filtered, and concentrated. The crude
conjugated cyclohexenol ester product was dissolved in 10 mL
of CH2Cl2 and cooled to -78 °C, and DBU (50 µL) was added.21
After 1 h, the reaction was allowed to warm to rt and stirring
was continued for 12 h. The reaction mixture was then
washed with aqueous NH4Cl, extracted with CH2Cl2, dried
with MgSO4, filtered, and concentrated. Column chromatog-
raphy using silica eluted with a gradient of 0-20% EtOAc in
hexane provided the purified product in an overall yield of 77%
(553 mg, 8 mmol): 1H NMR (CDCl3) δ 1.51 (ddd, J ) 25.2,
12.0, 3.6 Hz, 1H), 1.60 (bs, 1H), 2.07 (dddd, J ) 13.2, 8.4, 2.4,
1.2 Hz, 1H), 2.30 (d, J ) 13.2 Hz, 1H), 2.65 (dtd, J ) 19.2, 5.6,
1.2 Hz, 1H), 3.62 (s, 3H), 4.20 (m, 1H), 4.64 (t, J ) 2.8 Hz,
1H), 5.05/5.10 (AB, J ) 11.2 Hz, 2H), 6.97 (dd, J ) 5.6, 2.4
Hz,1H), 7.40-7.53 (m, 4H), 7.78-7.85 (m, 2H), 8.16 (d, J ) 8
Hz, 1H); 13C NMR (CDCl3) δ 35.6, 36.7, 52.1, 63.6, 71.1, 71.8,
124.8, 125.6, 126.1, 126.5, 127.4, 128.8, 129.2, 131.0, 132.4,
cm-1) 1713; [R]30 -42.3° (c ) 3.0, CHCl3). These data are
D
consistent with those reported previously.3c HRMS: calcd for
C20H40O4Si2 (M - CH3) 385.2230, found 385.2235.
Cycloaddition with En ol ter t-Bu tyldim eth ylsilyl Eth er .
To a solution of methyl 2-pyrone-3-carboxylate (0.55 g, 3.6
mmol) in 80 mL of anhydrous toluene was added the solution
of binol-Ti complex (2.1 g, weighed in air, 4.8 mmol, 1.35
equiv) in anhydrous toluene (35 mL) at 45 °C for 1 h. After
being stirred for 4 h at 45-48 °C, the resulting solution was
cooled to -30 °C. To the cooled solution was added
CH2dCHOSiMe2Bu-t (3.0 mL, 18 mmol, 5 equiv).22 The
reaction mixture was stirred at -30 °C for 21 h with monitor-
ing by TLC. After being warmed to rt, the reaction mixture
was purified by silica gel chromatography (15% EtOAc/
hexanes) without aqueous workup to give 0.66 g (59%) of
bicycloadduct (-) -1b as a white solid, mp 88-89 °C: 92% ee
by HPLC on a Daicel Chiralpak AS column, eluent ) 3% EtOH
in hexanes, flow rate ) 1 mL/min, (+)-1b 7.0 min, (-) -1b 9.2
min, with detection at 230 nm: [R]28D -4.15 ° (c ) 11.2, CHCl3);
1H NMR (400 MHz, CDCl3) δ 6.78 (dt, J ) 7.6, 1.6 Hz, 1H),
6.59 (dd, J ) 7.6, 5.2 Hz, 1H), 5.21 (dddd, J ) 5.2, 3.6, 1.6, 1.6
Hz, 1H), 4.72 (dt, J ) 7.2, 1.6 Hz, 1H), 3.89 (s, 3H), 2.61 (ddd,
J ) 14.0, 7.6, 3.6 Hz, 1H), 1.53 (dt, J ) 13.6, 1.6 Hz, 1H), 0.79
(s, 9H), 0.03 (s, 6H); 13C NMR (100 MHz, CDCl3) δ 168.86,
167.47, 130.00, 129.68, 74.11, 66.21, 62.59, 52.73, 38.11, 52.34,
17.60, -4.54, -5.44; IR (CHCl3, cm-1) 3028, 2956, 2931, 2887,
2856, 1626; MS, m/ z (70 eV, EI) 255 (M - t-Bu+, 31.30), 211
(M - t-Bu - CO2+, 100); HRMS calcd for C24H40O2Si - C4H9
255.0689, found 255.0686.
134.4, 134.3, 141.3, 167.1; [R]30 -77.2° (c ) 4.3, CHCl3); mp
D
82.4-82.7 °C. Anal. Calcd for C19H20O4: C, 73.06, H, 6.45.
Found: C, 73.01, H, 6.51.
To a vigorously stirred solution of the cyclohexenol ester (550
mg, 1.8 mmol) in 40 mL of EtOAc under H2 was added 40 mL
(0.72 mmol) of a 0.018 M solution of PdCl2(CH3CN)2 in EtOH.
This latter solution was prepared immediately prior to use by
the addition of 405 µL (1% v/v) of glacial acetic acid to a
solution of PdCl2(CH3CN)2 (207 mg, 0.09 mmol, 0.5% w/v) in
40 mL of EtOH. The reaction was closely monitored by TLC
to avoid overreduction. The reaction color proceeded from
translucent orange to a grey suspension that eventually formed
a black precipitate that, upon settling, resulted in a colorless
P r ep a r a tion of Cycloh exen e (-)-2 fr om Cycloa d d u ct
(-)-1b. To 1.5 mL of allyl alcohol was added 3 mL (3.9 mmol,
1.3 equiv) of a 1.6 M solution of n-BuLi in hexanes dropwise
at -78 °C with stirring, and then the solution was warmed
up to rt for 20 min. To the solution of 0.97 g (3.1 mmol) of
two batches of bicyclic lactone (-)-1b, 90% ee, in 4.5 mL of
CH2Cl2:allyl alcohol (2:1) was added the freshly prepared
allyloxide solution at 0 °C. After the solution was stirred for
20 min, the reaction was quenched with water, extracted with
(20) Mandai, T.; Imaji, M.; Takada, H.; Kawata, M.; Nokami, J .;
Tsuji, J . J . Org. Chem. 1989, 54, 5395.
(21) Watanabe, W. H.; Conlon, L. E. J . Am. Chem. Soc. 1957, 79,
2820.
(22) Prepared from THF according to the general procedure of J ung,
M. E.; Blum, R. B. Tetrahedron Lett. 1979, 3791.

676 J . Org. Chem., Vol. 61, No. 2, 1996
Posner et al.
EtOAc (×2), washed with brine, dried over anhydrous Na2-
SO4, and concentrated in vacuo. Purification by short path
chromatography (20% EtOAc/hexanes) gave 1.10 g (96%) of
the mixed allyl methyl malonate as a white solid: mp 45-48
The reaction mixture, purified by column chromatography
(silica gel, eluting solvent 10-20% EtOAc/hexane), gave 2.1 g
(9.12 mmol, 88%) of cycloadduct (()-4 as a white solid (mp
77-78 °C) having spectroscopic and physical properties identi-
cal to those reported previously.17
°C; [R]28 +105 ° (c ) 2.2, CHCl3); 1H NMR (400 MHz, CDCl3)
D
δ 5.99 (ddt, J ) 29.2, 10.4, 1.6 Hz, 2H), 5.85 (m, 1H), 5.28 (dq,
J ) 17.2, 1.6 Hz), 5.23 (dq, J ) 10.4, 1.2 Hz, 1H), 4.78 (d, J )
4.8 Hz, 1H), 4.60 (qdt, J ) 13.2, 5.2, 1.6 Hz, 1H), 4.41 (bs,
1H), 3.72 (s, 3H), 2.31 (m, 1H), 1.89 (ddd, J ) 13.6, 9.6, 1.6
Hz, 1H), 1.46 (d, J ) 7.2 Hz, 1H of OH), 0.82 (s, 9H), 0.12 (s,
3H), 0.10 (s, 3H); 13C NMR (100 MHz, CDCl3) δ 168.93, 168.06,
134.24, 131.25, 123.05, 118.46, 69.68, 65.96, 63.76, 59.80,
52.52, 37.31, 25.56, 17.78, -4.13, -5.56; IR (CHCl3, cm-1) 3027,
3012, 2931, 2887, 1650; MS, m/ z (70 eV, EI) 313 (M - t-Bu+,
100); HRMS m/ z (M+) calcd for C18H30O6Si - C4H9 313.1107,
found 313.1108.
Cycloh exen ol Ben zyl Eth er (()-5. To 20 mL of methanol
at 0 °C was added dropwise 3.0 mL of n-BuLi (1.6 M in hexane,
4.8 mmol, 2.2 equiv). After stirring at 0 °C for 30 min, the
mixture was diluted with 80 mL of CH2Cl2 and cooled to -78
°C, to which a precooled (-78 °C) solution of 0.5 g (2.17 mmol)
of (()-4 in 15 mL of CH2Cl2 was added dropwise over 20 min
via cannula. The resulting mixture was warmed to -10 °C,
and the progress of the reaction was closely monitored by
TLC: Rf (50% EtOAc/hexane) ) 0.50, (()-4; 0.33, â,γ-unsatur-
ated ester; 0.26, (()-5. After 5 h, just before the complete
conversion of â,γ-unsaturated ester to the product (()-5, the
mixture was cooled to -20 °C and stirred for 3 h. The reaction
was then quenched with saturated aqueous NH4Cl at -20 °C.
The combined extracts were dried over Na2SO4, concentrated,
and purified by column chromatography (silica gel, eluting
solvent 20-50% EtOAc/hexane) to yield 0.46 g (1.75 mmol,
81%) of cyclohexenol benzyl ether (()-5 as a pale yellow oil:
IR (CHCl3, cm-1) 1708; 1H NMR (CDCl3) δ 1.53 (ddd, J ) 13.2,
12.0, 3.6 Hz, 1H), 2.10 (ddd, J ) 19.2, 9.6, 2.4, 1.2 Hz, 1H),
2.31 (dm, J ) 13.2 Hz, 1H), 2.69 (dtd, J ) 19.2, 5.6, 1.2 Hz,
1H), 3.74 (s, 3H), 4.23 (m, 1H), 4.54 (t, J ) 3.2 Hz, 1H), 4.62
and 4.67 (AB, J ) 11.2 Hz, 1H), 6.99 (dd, J ) 5.6, 2.4 Hz, 1H),
7.24-7.37 (m, 5H); 13C NMR (CDCl3) δ 35.3, 36.5, 51.7, 63.2,
71.7, 72.4, 127.6, 128.0 (2C), 128.3 (2C), 130.6, 138.5, 140.6,
166.6; HRMS calcd for C15H18O4 (M - OCH3) 231.1021, found
231.1023.
A 20 mL hydrolysis tube was charged with 1.05 g (2.85
mmol) of the mixed malonate, 3.5 mL of dioxane, 0.15 mL (2.5
mmol, 1.25 equiv) of 88% formic acid, 0.55 mL (3.92 mmol,
1.4 equiv) of trimethylamine, triphenylphosphine (74 mg, 0.1
equiv) and palladium diacetate (32 mg, 0.05 equiv) with
stirring. This was sealed under argon and stirred for 16 h at
105-110 °C. The reaction mixture was cooled to rt, extracted
with EtOAc (×2), washed with 10% HCl and brine, dried over
anhydrous Na2SO4, and then concentrated in vacuo to give 0.81
g of monosilylated hydroxy R,â-unsaturated ester as a pale
yellow oil. To a solution of the crude ester (0.81 g, 2.83 mmol)
and 2,6-lutidine (1.6 mL, 14.2 mmol, 5 equiv) in 30 mL of CH2-
Cl2 was added t-BuMe2SiOTf (0.8 mL, 3.25 mmol, 1.15 equiv)
at 0 °C. After being stirred for 10 min, the reaction mixture
was diluted with ether (100 mL) and then the reaction was
quenched with water. The solution was washed with 5% HCl
(×2), and the aqueous layer was extracted with ether (100 mL).
The combined solution was washed with saturated aqueous
NaHCO3 solution, and brine, dried over NaSO4, concentrated
in vacuo, and then purified by chromatography (5% EtOAc/
hexanes) to give 0.98 g (86%) of cyclohexene (-)-2 as a colorless
oil. Spectral data are identical with those reported previously.3c
Cycloa d d ition of 2-P yr on e w ith Ben zyl Vin yl Eth er :
Bicycloa d d u ct (()-4. A piece of heat-shrinkable Teflon
tubing was sealed on one end with a glass plug by using a
heat gun. A solution of 0.95 g (1.04 mmol, 0.1 equiv) of (+)-
Yb(tfc)3 (Aldrich) in 3 mL of CH2Cl2, followed by 1.0 g (10.4
mmol) of 2-pyrone and 2.8 g (20.9 mmol, 2.0 equiv) of benzyl
vinyl ether were introduced into it. The open end of the tubing
was sealed with a second glass plug in a similar way. The
reaction vessel was then pressurized at 12 kbar at rt for 3 days.
Ack n ow led gm en t . We thank the NSF (CHE-
9321256) for financial support, the French Association
for Therapeutic Research for a postdoctoral fellowship
to Dr. Eydoux, Takeda Chemical Industries, J apan, for
supporting and granting a leave of absence to Dr.
Ishihara, and Mr. Ben Woodard (J ohns Hopkins Uni-
versity) for preparation of various vinyl ethers.
Su p p or tin g In for m a tion Ava ila ble: 1H and 13C NMR
spectra (16 pages). This material is contained in libraries on
microfiche, immediately follows this article in the microfilm
version of the journal, and can be ordered from the ACS; see
any current masthead page for ordering information.
J O9515900