A synthesis of tamiflu by using a barium-catalyzed asymmetric diels-alder-type reaction

Article abstract of DOI:10.1002/anie.200804777

(Chemical Equation Presented) In pursuit of a better route: A new catalytic asymmetric synthesis of Tamiflu was developed. The key transformation was an asymmetric Diels-Alder-type reaction of 1 and 2 catalyzed by a barium/F ₂-FujiCAPO complex in the presence of a CsF co-catalyst to construct the core of Tamiflu (see scheme; TMS=trimethylsilyl). The product was converted into Tamiflu in 11 steps on a gram scale.

Full text of DOI:10.1002/anie.200804777

Communications
DOI: 10.1002/anie.200804777
Asymmetric Synthesis
A Synthesis of Tamiflu by Using a Barium-Catalyzed
Asymmetric Diels–Alder-Type Reaction**
Kenzo Yamatsugu, Liang Yin, Shin Kamijo, Yasuaki Kimura, Motomu Kanai,*
and Masakatsu Shibasaki*
Angewandte
Chemie
1070
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 1070 –1076

Angewandte
Chemie
Influenza viruses pose a serious threat to world public health.
Two of the drugs currently used to treat influenza patients are
Tamiflu (1; (À)-oseltamivir phosphate)[1] and Relenza
(zanamivir),[2] both of which inhibit viral neuraminidase.
Currently, Tamiflu is produced and supplied by Roche using
(À)-shikimic acid as the starting material.[3] The production of
(À)-shikimic acid with consistent purity, however, requires a
lot of time and is costly. In addition, the dependence on a
single synthetic route for the supply of such an important drug
is unwise. Therefore there is an urgent demand for the
development of alternative practical syntheses of Tamiflu,
starting from easily available starting materials.[4,5] We report
herein a new synthesis of Tamiflu which features a novel
asymmetric Diels–Alder-type reaction catalyzed by a barium/
F2-FujiCAPO complex.
Our retrosynthetic analysis is shown in Scheme 1. The
3-pentyloxy group at C3 should be introduced at a late stage
by a ring-opening reaction of an N-acetyl aziridine[4a]
produced from b-alcohol 2 under Mitsunobu conditions. In
our previous reports,[4b,d,e] we utilized a cyanide group as a
precursor for the ester group at C1, however, its introduction
and conversion into an ethoxycarbonyl group afforded only
moderate yields of the desired product. Therefore, we sought
an alternative ester surrogate which led to the identification
of protected hydroxy malononitrile anion 4, developed by
Nemoto and Yamamoto et al.,[6] as a promising candidate.
The introduction of 4 to cyclic carbamate 5 by a palladium-
catalyzed allylic substitution would produce cyclohexene
derivative 3, which in turn would be converted into b-alcohol
2 by epoxidation and subsequent unveiling of the masked
ester equivalent. Cyclic carbamate 5 would be obtained from
hydroxy dicarboxylic acid 6 by using the Curtius rearrange-
ment. A catalytic asymmetric Diels–Alder-type reaction
between diene 7 and dienophile 8[7] should produce a
precursor of 6. Thus, the first step was to develop such a
key reaction.
Although a number of Lewis acid catalyzed asymmetric
Diels–Alder reactions have been reported to date,[8] including
those using ketone-derived siloxy dienes (such as Danishef-
skyꢀs diene),[9] none of them utilize 7 because of the lability of
7 under acidic conditions. Indeed, 7 readily polymerized in the
presence of representative chiral Lewis acid catalysts. There-
fore, we examined a conceptually distinct catalytic asymmet-
ric Diels–Alder-type reaction that was not dependent on acid
catalysis. We envisioned that metal alkoxides (or phenoxides)
might activate the siloxy diene through the formation of a
hypervalent silicate or transmetalation (HOMO-raising
mechanism; HOMO = highest occupied molecular orbital).[10]
Table 1 shows the initial optimization results. First we used
(R)-binol (11; binol = 2,2’-dihydroxy-1,1’-binaphthyl) as the
chiral ligand, and examined several metal isopropoxides as
catalytic metal sources (30 mol%) in CH2Cl2 at room temper-
ature (Table 1, entries 1–4). Among the metal isopropoxides
examined, Ba(OiPr)2 afforded the desired products in mod-
erate yield (Table 1, entry 2; isomers 9 and 10 were insepa-
rable by silica gel column chromatography),[11] however a
meaningful enantioselectivity was not induced. The effects of
several chiral ligands on the enantioselectivity were examined
at À208C (Table 1, entries 5–10). Whereas the reaction did
not proceed at all using (R)-binol at this low temperature,
products were obtained in 97% yield within 30 minutes by
using taddol (12; Table 1, entry 5). Unfortunately, enantioin-
duction was also not observed in this case. Careful analysis of
the reaction mixture revealed that the ligand was partially
silylated (see below), suggesting that the chiral ligand was
partially dissociated from the barium metal. This observation
may explain why meaningful enantioinduction was not
realized, even with the use of privileged chiral ligands. We
therefore used multidentate ligands 13–15 which were
developed in our laboratory,[12] anticipating that an effective
chiral environment would be retained even after partial
ligand silylation.[13] Initial promising enantioselectivity of
desired product 9 was attained using F2-GluCAPO (13); the
products were produced in 72% combined yield (9/10 3:1)
and with 77% ee for 9 (Table 1, entry 6). When the amount of
catalyst used was reduced to 20 mol%, however, the reaction
was sluggish and the enantiomeric excess of 9 decreased to
61% ee (Table 1, entry 7). In contrast, when the barium/
FujiCAPO (14) complex was used, the reaction proceeded
quickly and desired product 9 was obtained with 73% ee.
However, several unidentified products were also produced,
resulting in a moderate combined yield of 9 and 10 (34%;
Scheme 1. Retrosynthetic analysis.
[*] K. Yamatsugu, L. Yin, Dr. S. Kamijo, Y. Kimura, Dr. M. Kanai,
Prof. Dr. M. Shibasaki
Graduate School of Pharmaceutical Sciences
The University of Tokyo, Hongo, Bunkyo-ku, Tokyo 113-0013 (Japan)
Fax: (+81)3-5684-5206
E-mail: mshibasa@mol.f.u-tokyo.ac.jp
[**] This work was supported by Grant-in-Aid for Specially Promoted
Research of MEXT and Grant-in-Aid for Scientific Research (S) and
(B) from JSPS. Akihiro Sato and Sanae Furusho of JASCO Interna-
tional Co., Ltd. are acknowledged for ESI-QFT-MS measurements of
the barium complexes. K.Y. thanks the JSPS for research fellowships.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200804777.
Angew. Chem. Int. Ed. 2009, 48, 1070 –1076
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1071

Communications
Table 1: Initial screening of reaction conditions.
Table 2: Additive metal fluoride effects.
Entry
Additive
t [h]
Combined
yield [%]
d.r.[a] (9:10)
ee [%] of 9[a]
Entry
M
Ligand t [h] Combined d.r.[a] (9:10) ee [%] of 9[a]
yield [%]
1[b]
–
KF
23
38
64
79
74
3:1
3:1
3:1
3:1
2:1
2:1
4:1
4:1
5:1
5:1
88
87
85
19
96
93
97
94
95
91
2[b]
1[b,c,e]
2[b,c,e]
3[b,c,e]
4[b,c,e]
5[c,f]
Mg 11
17
17
17
17
0
45
0
–
3:1
–
–
4
–
–
1
77
61
73
73
88
3[b]
LaF3
ZnF2
CsF
CsF
CsF
CsF
CsF
CsF[h]
18.5
12
0.5
0.5
36
36
96
24
Ba
Sc
11
11
4[b]
67
5[b]
>99
>99
91
88
91
Gd 11
0
–
6[c]
Ba
Ba
Ba
Ba
Ba
12
13
13
14
15
15
0.5 97
2.5 72
74
1.5 34
4:1
3:1
3:1
5:1
6:1
3:1
7[d]
6[c,f]
8[d,e]
9[d,f]
10[g]
7[d,f]
42
8[d,f]
9[d,f]
17
23
62
64
>99
10[d,f,g] Ba
[a] Determined by chiral GC analysis. [b] Ba(OiPr)2 (20 mol%), 15
(20 mol%). [c] Ba(OiPr)2 (10 mol%), 15 (10 mol%). [d] Ba(OiPr)2
(2.5 mol%), 15 (2.5 mol%), 0.67m of 8. [e] 12 g scale. [f] 58 g scale.
[g] Ba(OiPr)2 (1 mol%), 15 (1 mol%). [h] CsF=5 mol%. In other
entries, the same mol% of additives as those of Ba and 15 was used.
[a] Determined by chiral GC analysis. [b] Added 4 ꢀ molecular sieves
(500 mg per mmol 8). [c] Metal (30 mol%), ligand (30 mol%). [d] Metal
(20 mol%), ligand (20 mol%). [e] RT. [f] À208C. [g] THF was used as the
solvent. TMS=trimethylsilyl.
diastereomeric ratio was less satisfactory (2:1). The catalyst
loading was reduced to 2.5 mol% under more concentrated
conditions (Table 2, entry 7),[15] and the products were
afforded in 91% combined yield with a higher diastereomeric
ratio of 4:1. The enantiomeric excess of 9 was maintained
(97% ee). The reaction is practical, and a 12 gram scale
reaction was conducted without altering its efficiency
(Table 2, entry 8). On a 58 gram scale, both the enantiomeric
excess and the diastereomeric ratio were almost constant
(Table 2, entry 9), though the reaction was slower, possibly
because of the less satisfactory stirring efficiency. The reaction
can be performed in the presence of 1 mol% of barium/15
and 5 mol% of CsF to afford the products (9 and 10) in
greater than 99% yield (d.r. 5:1) with 91% ee of 9 (Table 2,
entry 10). We have established a practical catalytic asymmet-
ric Diels–Alder-type reaction, which is the key methodology
used for our Tamiflu synthesis.
Having established the key catalytic asymmetric Diels–
Alder-type reaction, we investigated the synthesis of Tamiflu.
Our optimized route is summarized in Scheme 2. The
hydrolysis of the Diels–Alder products (5:1 mixture of 9
and 10) proceeded uneventfully to afford dicarboxylic acids
16, and treatment of 16 with DPPA[16] and Et3N cleanly
produced the corresponding hydroxy diacyl azide derived
from 9. In this step, the reaction temperature had to be strictly
maintained below 48C to prevent spontaneous Curtius
rearrangement, which would diminish the yield of the desired
products. Products derived from the minor isomer (exo-10)
decomposed during this transformation. The desired diacyl
azide was roughly isolated, after silica gel column chroma-
tography in approximately 95% yield in two steps from 9.[17]
Next, the Curtius rearrangement was performed to construct
the two neighboring nitrogen functionalities of Tamiflu. A
solution of the diacyl azide in anhydrous tBuOH was heated
Table 1, entry 8). The catalyst derived from F2-FujiCAPO
(15) was the most effective in terms of reactivity, enantiose-
lectivity, and diastereoselectivity, affording 9 with 73% ee
(Table 1, entry 9). The enantiomeric excess of 9 was improved
to 88% ee when THF was used as the solvent (Table 1,
entry 10).
The reactivity and enantioselectivity of the Diels–Alder-
type reaction were additionally improved, such that the
reaction could be run on large scale, by optimizing the
additives (Table 2). Thin layer chromatography analysis of the
progress of the reaction (Table 1) revealed that the reaction
mixture contained significant amounts of unsilylated prod-
ucts, even under strictly anhydrous conditions. This result
suggested that the silicon transfer step from 7 to the
secondary alkoxide products intervened in the catalytic
cycle (see below), and that this step was catalyst turnover-
limiting. The silicon transfer should be accelerated by the
formation of a pentavalent silicate with Lewis base addi-
tives,[14] therefore we examined the additive effects of
catalytic metal fluorides (20 mol%, Table 2).
Although the addition of KF had no effect on the reaction
(Table 2, entry 2), the addition of LaF3, ZnF2, and CsF
facilitated the reaction (Table 2, entries 3–5). Specifically, CsF
had a remarkable effect (Table 2, entry 5); the reaction was
complete within 30 minutes and the desired 9 was obtained
with significantly higher enantioselectivity (96% ee), but the
1072
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 1070 –1076

Angewandte
Chemie
Scheme 2. Catalytic asymmetric synthesis of Tamiflu. Reagents and conditions: a) Ba(OiPr)2 (2.5 mol%), F2-FujiCAPO (15, 2.5 mol%), CsF
(2.5 mol%), THF, À208C, 36–96 h; aq. 1m HCl (91%, 5:1, 95% ee for 9); b) aq. 2m NaOH (10 equiv), MeOH, 608C, 10 h; c) DPPA (3 equiv),
Et3N (3 equiv), THF, 08C, 21 h (95% (2 steps from 9)); d) tBuOH, 808C, 13 h; e) Ac2O (2 equiv), Et3N (4 equiv), DMAP (10 mol%), CH2Cl2, RT,
2.5 h (ca. 98% (2 steps)); f) Recrystallization from CH2Cl2/cyclopentyl methyl ether (1:2; 80%); g) 4 (1.2 equiv), [Pd2(dba)3]·CHCl3 (2 mol%), dppf
(4 mol%), toluene, 608C, 30 min (85%); h) TFAA (10 equiv), urea/H2O2 (20 equiv), Na2HPO4 (15 equiv), CH2Cl2, 48C, 2 h; i) K2CO3 (5 equiv),
EtOH, RT, 5 h; j) DEAD (2 equiv), PPh3 (2 equiv), p-nitrobenzoic acid (2 equiv), THF, À208C, 1.5 h; LiOH (3 equiv), EtOH, À208C, 15 min (65%
(3 steps); k) DIAD (2.1 equiv), Me2PPh (2.1 equiv), Et3N (21 mol%), CH2Cl2, 48C, 10 min (76%); l) BF3·OEt2 (1.5 equiv), 3-pentanol, À208C,
15 min (75%); m) TFA; H3PO4 (73%). DPPA=diphenylphosphoryl azide; DMAP=4-dimethylaminopyridine; dba=dibenzylideneacetone;
dppf=bis(diphenylphosphanyl)ferrocene; TFAA=trifluoroacetic anhydride; DEAD=diethyl azodicarboxylate; DIAD=diisopropylazodicarboxylate;
TFA=trifluoroacetic acid; Boc=tert-butoxycarbonyl; Ac=acetyl.
at 808C for 13 hours. The Curtius rearrangement and
subsequent intramolecular trapping of the resulting C4
isocyanate group by the cis-C3 hydroxy group, as well as the
intermolecular addition of tBuOH to the C5 isocyanate group
proceeded in one pot to afford cyclic carbamate 5. Thus, the
two amino functionalities were successfully differentiated by
this sequence. Treatment of 5 with Ac2O and Et3N selectively
installed an acetyl group on the cyclic carbamate nitrogen
atom, producing 17 in 98% yield in two steps from the diacyl
azide. Enantiomerically pure 17 was obtained in 80% yield
after recrystallization of the crude product from CH2Cl2/
cyclopentyl methyl ether (1:2).
Although the conversions of 16 into 17 were high yielding,
it was desirable to develop more practical procedures such
that the isolation of the potentially explosive azide compound
could be avoided. Such a protocol was established: after
completion of the diacyl azide formation from 16 was
confirmed by TLC analysis, the reaction was quenched with
1m HCl and the excess Et3N and the diphenylphosphate
byproduct were removed in an aqueous workup,[17] which was
essential for clean conversion in the next Curtius rearrange-
ment.[18] Neither additional purification of the diacyl azide nor
complete removal of the extraction solvent (AcOEt) was
necessary in this protocol, which was very advantageous
considering the potentially explosive characteristics of neat
forms of azide compounds.[19] After the Curtius rearrange-
ment, 5 was isolated in 72% yield from the desired Diels–
Alder-type product 9 (ca. 95% yield of 5 was isolated from
the treatment of 9 by the above-described procedure using the
isolated diacyl azide).
Among many reported C1 acyl anion equivalents,[20]
protected hydroxy malononitrile 4, developed by Nemoto
and Yamamoto et al.,[6] was effective. Thus, regioselective
allylic substitution proceeded when cyclic carbamate 17 was
heated with 4 in the presence of 2 mol% of [Pd2(dba)3]·CHCl3
and 4 mol% of dppf in toluene at 608C, and product 3 was
obtained in 85% yield. Epoxidation of 3 with in situ
generated trifluoroperacetic acid[21] afforded a-epoxide 18
as the sole product.[22] This exclusive a-face selectivity is a
result of the directing effect of the neighboring acetamide
moiety at C4.[23] Treatment of 18 with K2CO3 in EtOH
converted the acetoxydicyanomethyl unit into an ethoxycar-
bonyl group,[6] and subsequent E2 epoxide opening pro-
ceeded concomitantly to produce a-allyl alcohol 19 in one
pot.
The final task was to install the 3-pentyloxy group at C3.
After conversion of the C3 hydroxy group of 19 into a better
leaving groups (such as triflate, mesylate, chloride, and cyclic
sulfamidate), SN2 inversion with the 3-pentyloxy anion was
unsuccessfully attempted.[24] Therefore, we focused on the
formation of N-acyl aziridine and the subsequent aziridine-
opening process.[1,3,4a,b,d–g,j] Mitsunobu esterification of
a-alcohol 19 with p-nitrobenzoic acid, and one-pot ethanol-
ysis of the resulting ester produced C3 b-alcohol 2, which was
Angew. Chem. Int. Ed. 2009, 48, 1070 –1076
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1073

Communications
isolated in 65% yield over 3 steps from 3.[25] A Mitsunobu
aziridine synthesis from 2 was successfully performed using
Me2PPh and DIAD in the presence of 21 mol% of Et3N,
producing 20 in 76% yield. Phosphines had a significant
influence in this reaction, and oxazoline 21 was the major
byproduct when using other phosphines. The ring-opening
reaction of 20 with 3-pentanol was best performed using
BF3·OEt2 to afford Boc-protected (À)-oseltamivir in 75%
yield. Cleavage of the Boc group with TFA and salt formation
with phosphoric acid produced Tamiflu (1) in 73% yield. All
the analytical data matched those reported in the literature.[3a]
By using the synthetic route described above, we achieved a
gram-scale synthesis of Tamiflu.
Success in developing this new synthesis of Tamiflu
depended on the novel barium-catalyzed asymmetric Diels–
Alder-type reaction. Preliminary mechanistic studies indi-
cated that a chiral barium dienolate is generated through
transmetalation from siloxy diene 7, and the barium dienolate
is the active species in this reaction. Two experimental results
support this mechanism.[26] 1) when the barium/15 complex
was treated with 30 equivalents of 7 in the absence of
dienophile 8, trimethylsilylated ligand 22 was observed in
the ESI-MS analysis.[27] This result suggests that transmeta-
lation between 7 and the barium/15 complex occurred. 2) A
competitive experiment using 7 (3 equiv), 1-methoxy-1,3-
butadiene (23; 3 equiv), and 8 (1 equiv) in the presence of
10 mol% barium/15 complex and CsF afforded products 9
and 10 derived from siloxy diene 7 in 99% combined yield.
Products 24 derived from 23 were not detected, therefore 7 is
much more reactive than 23 under the current catalysis. A
LUMO-lowering (LUMO = lowest occupied molecular orbi-
tal) mechanism does not explain this result. Together, these
two results suggest that the barium-catalyzed asymmetric
Diels–Alder-type reaction is likely promoted through a
HOMO-raising mechanism, possibly via transmetalation.[28,29]
Next, catalyst constitution was determined based on ESI-
MS studies. Two MS peaks corresponding to complexes of
barium/15 = 3:3 and 3:4 were observed in the catalyst
solution.[26] The 3:4 complex, however, was labile and
converted into the 3:3 complex by a tandem MS/MS experi-
ment. The isotope distribution completely matched with the
calculated pattern. Therefore, the 3:3 complex 25 should be
the active catalyst for the asymmetric Diels–Alder-type
reaction.
On the basis of the above results, we propose a catalytic
cycle for the asymmetric Diels–Alder-type reaction
(Figure 1). First, the active barium dienolate 27 is generated
through transmetalation between catalyst 25 and siloxy diene
7. In this step, the chiral ligand is partially silylated (27). The
co-catalyst, CsF, would facilitate the generation of 27 through
the formation of pentavalent silicate 26.[30] Barium dienolate
27 should be reactive enough, and cyclization with 8 occurs in
either a concerted manner (Diels–Alder pathway) or a
stepwise manner (Michael-aldol pathway), producing inter-
mediate barium alkoxide 29.[31] Because of the existence of
multiple barium metals of different electronic characteristics
in a catalyst molecule, it is possible that the catalyst promotes
the reaction through an intramolecular transfer of the barium
dienolate to an activated dienophile by a Lewis acidic barium
Figure 1. Postulated catalytic cycle of the asymmetric Diels–Alder-type
reaction.
(28: dual activation mechanism[32]). Finally, the product
barium alkoxide in 29 attacks the trimethylsilyl group
attached to the ligand, and catalyst 25 is regenerated while
silylated products 9 and 10 are liberated from the catalytic
cycle. Alternatively, 29 reacts with another molecule of siloxy
diene 26, liberating the products while regenerating active
barium dieonolate 27. Given the synthetic usefulness of siloxy
dienes[33] but their lability to strongly Lewis acidic conditions,
the HOMO-raising activation mode of siloxy dienes induced
by the barium complex will have great potential for the
development of new asymmetric catalyses.
In summary, we developed a novel barium-catalyzed
asymmetric Diels–Alder-type reaction between 1-trimethyl-
siloxy-1,3-butadiene (7) and dimethyl fumarate (8) to access
the cyclohexene framework of Tamiflu. This methodology led
to the establishment of a new synthetic route of Tamiflu by
using the Curtius rearrangement and a palladium-catalyzed
allylic substitution reaction using the C1 acyl anion equiv-
alent, as key steps. Each conversion in the synthetic scheme is
easily conducted, and almost all the intermediates are
crystalline compounds, which could minimize the need for
column chromatography purification. Additional improve-
ment of the synthetic efficiency, especially by the develop-
ment of a novel methodology for the epoxide opening
reaction of 18 with 3-pentanol, is currently ongoing in our
laboratory.
Received: September 30, 2008
Published online: November 17, 2008
1074
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 1070 –1076

Angewandte
Chemie
[10] For examples of nucleophile activation through transmetalation
Keywords: barium · cycloaddition · dienes ·
homogeneous catalysis · tamiflu
.
from silicon enolates to chiral catalytic metal enolates, see: a) M.
Sodeoka, K. Ohrai, M. Shibasaki, J. Org. Chem. 1995, 60, 2648;
b) J. Krꢃger, E. M. Carreira, J. Am. Chem. Soc. 1998, 120, 837;
c) K. Oisaki, D. Zhao, M. Kanai, M. Shibasaki, J. Am. Chem.
Soc. 2006, 128, 7164.
[1] C. U. Kim, W. Lew, M. A. Williams, H. Liu, L. Zhang, S.
Swaminathan, N. Bischofberger, M. S. Chen, D. B. Mendel, C. Y.
Tai, G. Laver, R. C. Stevens, J. Am. Chem. Soc. 1997, 119, 681.
[2] M. von Itzstein, W.-Y. Wu, G. B. Kok, M. S. Pegg, J. C. Dyason,
B. Jin, T. V. Phan, M. L. Smythe, H. F. White, S. W. Oliver, P. M.
Colman, J. N. Varghese, D. M. Ryan, J. M. Woods, R. C. Bethell,
V. J. Hotham, J. M. Cameron, C. R. Penn, Nature 1993, 363, 418.
[3] a) J. C. Rohloff, K. M. Kent, M. J. Postich, M. W. Becker, H. H.
Chapman, D. E. Kelly, W. Lew, M. S. Louie, L. R. McGee, L. M.
Prisbe, R. H. Yu, L. Zhang, J. Org. Chem. 1998, 63, 4545; b) M.
Federspiel, R. Fischer, M. Hennig, H.-J. Mair, T. Oberhauser, G.
Rimmler, T. Albiez, J. Bruhin, H. Estermann, C. Gandert, V.
Gꢁckel, S. Gꢁtzꢁ, U. Hoffmann, G. Huber, G. Janatsch, S.
Lauper, O. Rꢁckel-Stꢂbler, R. Trussardi, A. G. Zwahlen, Org.
Process Res. Dev. 1999, 3, 266; c) S. Abrecht, P. Harrington, H.
Iding, M. Karpf, R. Trussardi, B. Wirz, U. Zutter, Chimia 2004,
58, 621; d) S. Abrecht, M. C. Federspiel, H. Estermann, R.
Fisher, M. Karpf, H.-J. Mair, T. Oberhauser, G. Rimmler, R.
Trussardi, U. Zutter, Chimia 2007, 61, 93.
[4] a) Y.-Y. Yeung, S. Hong, E. J. Corey, J. Am. Chem. Soc. 2006,
128, 6310; b) Y. Fukuta, T. Mita, N. Fukuda, M. Kanai, M.
Shibasaki, J. Am. Chem. Soc. 2006, 128, 6312; c) X. Cong, Z.-J.
Yao, J. Org. Chem. 2006, 71, 5365; d) T. Mita, N. Fukuda, F. X.
Roca, M. Kanai, M. Shibasaki, Org. Lett. 2007, 9, 259; e) K.
Yamatsugu, S. Kamijo, Y. Suto, M. Kanai, M. Shibasaki,
Tetrahedron Lett. 2007, 48, 1403; f) K. M. Bromfield, H.
Graden, D. P. Hagberg, T. Olsson, N. Kann, Chem. Commun.
2007, 3183; g) N. Satoh, T. Akiba, S. Yokoshima, T. Fukuyama,
Angew. Chem. 2007, 119, 5836; Angew. Chem. Int. Ed. 2007, 46,
5734; h) J.-J. Shie, J.-M. Fang, S.-Y. Wang, K.-C. Tsai, Y.-S. E.
Cheng, A.-S. Yang, S.-C. Hsiao, C.-Y. Su, C.-H. Wong, J. Am.
Chem. Soc. 2007, 129, 11892; i) N. T. Kipassa, H. Okamura, K.
Kina, T. Hamada, T. Iwagawa, Org. Lett. 2008, 10, 815; j) B. M.
Trost, T. Zhang, Angew. Chem. 2008, 120, 3819; Angew. Chem.
Int. Ed. 2008, 47, 3759; k) U. Zutter, H. Iding, P. Spurr, B. Wirz, J.
Org. Chem. 2008, 73, 4895; l) J.-J. Shie, J.-M. Fang, C.-H. Wong,
Angew. Chem. 2008, 120, 5872; Angew. Chem. Int. Ed. 2008, 47,
5788.
[5] For reviews, see: a) V. Farina, J. D. Brown, Angew. Chem. 2006,
118, 7488; Angew. Chem. Int. Ed. 2006, 45, 7330; b) M.
Shibasaki, M. Kanai, Eur. J. Org. Chem. 2008, 1839.
[6] a) H. Nemoto, Y. Kubota, Y. Yamamoto, J. Org. Chem. 1990, 55,
4515; b) H. Nemoto, X. Li, R. Ma, I. Suzuki, M. Shibuya,
Tetrahedron Lett. 2003, 44, 73.
[7] Diene 7 is commercially available, or it can be prepared on a
large scale from inexpensive crotonaldehyde ($ 40/500 mL) by
treatment with TMSCl, Et3N, and ZnCl2 (0.9 mol%) in Et2O (O.
Gaonacꢀh, J. Maddaluno, J. Chauvin, L. Duhamel, J. Org. Chem.
1991, 56, 4045. Also see the Supporting Information). Dimethyl
fumarate (8) is also readily available ($ 80/500 g).
[8] For a review of the catalytic asymmetric Diels–Alder reaction,
see: E. J. Corey, Angew. Chem. 2002, 114, 1724; Angew. Chem.
Int. Ed. 2002, 41, 1650.
[9] For examples of the use of ketone-derived siloxy dienes in
asymmetric Diels–Alder reactions, see: a) A. N. Thadani, A. R.
Stankovic, V. H. Rawal, Proc. Natl. Acad. Sci. USA 2004, 101,
5846; b) D. E. Ward, J. Shen, Org. Lett. 2007, 9, 2843; c) H.
Usuda, A. Kuramochi, M. Kanai, M. Shibasaki, Org. Lett. 2004,
6, 4387; d) D. Nakashima, H. Yamamoto, J. Am. Chem. Soc.
2006, 128, 9626; e) D. Liu, E. Canales, E. J. Corey, J. Am. Chem.
Soc. 2007, 129, 1498; f) Y. Sudo, D. Shirasaki, S. Harada, A.
Nishida, J. Am. Chem. Soc. 2008, 130, 12588.
[11] For chiral barium aryloxide catalysis, see: a) Y. M. A. Yamada,
M. Shibasaki, Tetrahedron Lett. 1998, 39, 5561; b) S. Saito, S.
Kobayashi, J. Am. Chem. Soc. 2006, 128, 8704; c) A. Yamaguchi,
N. Aoyama, S. Matsunaga, M. Shibasaki, Org. Lett. 2007, 9, 3387.
[12] Ligands 13 and 15 are commercially available from Junsei
Chemical Co., Ltd. (shiyaku-t@junsei.co.jp).
[13] Rare-earth metal complexes of 13–15 are versatile asymmetric
catalysts for cyanation and azidation. The catalysts activate the
nucleophile (TMSCN or TMSN3) through transmetalation.
Therefore, an effective chiral environment is still constructed
after partial silylation (or protonation). For reviews of the utility
and mechanism of the catalysts derived from those chiral ligands,
see: a) M. Shibasaki, M. Kanai, Org. Biomol. Chem. 2007, 5,
2027; b) M. Kanai, N. Kato, E. Ichikawa, M. Shibasaki, Synlett
2005, 1491.
[14] a) V. Gutmann, The Donor-Acceptor Approach to Molecular
Interactions; Plenum, New York, 1978; b) S. E. Denmark, G. L.
Beutner, T. Wynn, M. D. Eastgate, J. Am. Chem. Soc. 2005, 127,
3774; c) Y. Zhao, J. Rodrigo, A. H. Hoveyda, M. L. Snapper,
Nature 2006, 443, 67. Also see reference [10c].
[15] 0.67m is the saturating concentration of dimethyl fumarate in
THF at room temperature. The reaction mixture was
a
suspension at À208C.
[16] T. Shioiri, K. Ninomiya, S. Yamada, J. Am. Chem. Soc. 1972, 94,
6203.
[17] Although the reason is unclear, the diacyl azide is more stable
during aqueous workup or purification by silica gel column
chromatography rather than under the reaction conditions of the
acyl azide formation (in the presence of DPPA and Et3N).
[18] The Curtius rearrangement of the hydroxy diacyl azide in the
presence of Et3N afforded a complex reaction mixture.
[19] Detailed characterization of the pyrolytic properties of the
diacyl azide is currently ongoing.
[20] For reviews, see: a) T. A. Hase, J. K. Koskimies, Aldrichimica
Acta 1982, 15, 35; b) T. A. Hase, Umpoled synthons, Wiley, USA,
1987; c) H. Nemoto, J. Synth. Org. Chem. Jpn. 2004, 62, 347.
[21] D. Wiktelius W. Berts, A. J. Jensen, J. Gullbo, S. Saitton, I.
Csꢁreghd, K. Luthman, Tetrahedron 2006, 62, 3600.
=
[22] The C C bond of 3 is poorly reactive, and other oxidation
methods using mCPBA, CH3CO3H, mCPBA, and a radical
inhibitor at 808C (Y. Kishi, M. Aratani, H. Tanino, T. Fukuyama,
T. Goto, S. Inoue, S. Sugiura, H. Kakoi, Chem. Commun. 1972,
64), MMPP (P. Brougham, M. S. Cooper, D. A. Cummerson, H.
Heaney, N. Thompson, Synthesis 1987, 1015), and oxone did not
afford the desired product.
[23] For a comprehensive review of substrate-directed chemical
reactions, see: A. H. Hoveyda, D. A. Evans, G. C. Fu, Chem.
Rev. 1993, 93, 1307.
[24] Preliminary attempts at epoxide-opening reactions of 18 using
3-pentanol were not successful.
[25] To eliminate this Mitsunobu inversion step, it is desirable for the
oxidation of 3 to proceed from the b face. Such conditions were
identified: dihydroxylation of 3 with NaIO4 in the presence of
catalytic RuCl3 and H2SO4 (B. Plietker, M. Niggemann, Org.
Lett. 2003, 5, 3353) afforded the corresponding cis-b-diol.
However, the chemical yield was only approximately 50%.
[26] See the Supporting Information for the details.
[27] Isolation of 22 was not successful, possibly because of the lability
of the phenol-derived trimethylsilyl ether. On the basis of our
previous studies of cyanosilylation using a gadolinium/Glu-
CAPO catalyst (GluCAPO = 13; TMSCN as a stoichiometric
Angew. Chem. Int. Ed. 2009, 48, 1070 –1076
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1075

Communications
nucleophile: K. Yabu, S. Masumoto, S. Yamasaki, Y. Hama-
reaction. Preliminary application of the asymmetric barium
catalyst to reaction between and chalcone produced
compound 30 through conjugate addition of the barium dien-
olate at the a position and subsequent trapping of the enolate
oxygen atom by the aldehyde. This result suggests that the
reaction proceeds through a stepwise mechanism.
shima, M. Kanai, W. Du, D. P. Curran, M. Shibasaki, J. Am.
Chem. Soc. 2001, 123, 9908), and the fact that the electron
density of catechol significantly influenced the catalyst activity
(Table 1, entry 8 versus 9), we assume that the phenolic oxygen
atom is the silylation site (22).
a
7
[28] A HOMO-raising mechanism going through barium silicate
formation cannot be completely excluded. In the presence of
10 mol% TBAT (tetrabutylammonium difluorotriphenylsili-
cate), however, the reaction between 7 and 8 produced only
trace amounts of the cyclized products in THF at À208C.
Therefore, we assume that a barium dienolate generated through
transmetalation is the actual active species in the barium-
catalyzed asymmetric Diels–Alder-type reaction.
[29] Bienaymꢄ and Longeau utilized 7 in a racemic Diels–Alder
reaction via transmetalation to the corresponding aluminum
enolate by the successive treatment of 7 with stoichiometric
amounts of MeLi and Me2AlCl. See: H. Bienaymꢄ, A. Longeau,
Tetrahedron 1997, 53, 9637.
[30] For related co-catalyst effects on the facilitation of the trans-
metalation step, see: a) R. Wada, T. Shibuguchi, S. Makino, K.
Oisaki, M. Kanai, M. Shibasaki, J. Am. Chem. Soc. 2006, 128,
7687; b) R. Wada, K. Oisaki, M. Kanai, M. Shibasaki, J. Am.
Chem. Soc. 2004, 126, 6536.
[32] M. Shibasaki, H. Sasai, T. Arai, Angew. Chem. 1997, 109, 1290;
Angew. Chem. Int. Ed. Engl. 1997, 36, 1236. See also refer-
ence [13].
[33] For selected recent examples, see: a) V. B. Birman, S. J. Dani-
shefsky, J. Am. Chem. Soc. 2002, 124, 2080; b) K. Tiefenbacher,
V. B. Arion, J. Mulzer, Angew. Chem. 2007, 119, 2744; Angew.
Chem. Int. Ed. 2007, 46, 2690; c) J. Yamaguchi, I. B. Seiple, I. S.
Young, D. P. OꢀMalley, M. Maue, P. S. Baran, Angew. Chem.
2008, 120, 3634; Angew. Chem. Int. Ed. 2008, 47, 3578.
[31] At present, we cannot determine which pathway is operative.
Therefore, we call the present reaction a Diels–Alder-type
1076
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 1070 –1076