The asymmetric formal hetero-Diels-Alder reaction of methyl (E)-4-Aryl-2-oxo-3-butenoates catalyzed by [Sc(OTf)3/pybox] complexes

Article abstract of DOI:10.1002/ejoc.201200056

The complex between scandium triflate and pybox is a good enantioselective catalyst for the reaction between methyl (E)-4-aryl-2-oxo-3-butenoates (1) and enol silyl ethers (2). The main products with (4'S)-isopropyl-, (4'S)-phenyl-, and (4'S,5'S)-4-TIPS-pybox are methyl (4S,4aS,8aS)-4-aryl-8a-trialkylsiloxy- hexahydro-4H-chromen-2-carboxylates (7), which are obtained in good yield and enantioselectivities of more than 95 % ee. Because the reaction gives products with a trans ring junction that cannot derive from a conventional concerted hetero-Diels-Alder pathway, the mechanism involved in the enantioselective catalytic cycle and the origin of the stereoinduction were investigated. The structures of two desilylated products, 8b and 9b, were determined by X-ray crystal analysis. Their absolute configuration was then related to that of 7 by stereospecific isomerization and/or desilylation reactions. If aryl-butenoates 1 are coordinated to the catalyst, forming reacting complexes characterized by the five-membered structure 11, these rigid reacting intermediates give a face discrimination that is determined by the configuration of the pybox 4'-substituent. The resulting face-selective attack of enol silyl ethers to coordinated 1 gives an enantioselective tandem Mukaiyama-Michael addition/intramolecular ring closure reaction, which is an enantioselective formal hetero-Diels-Alder reaction, which rationalizes the absolute configuration of the reaction products. Copyright

Full text of DOI:10.1002/ejoc.201200056

Job/Unit: O20056
/KAP1
Date: 05-04-12 16:12:07
Pages: 14
FULL PAPER
DOI: 10.1002/ejoc.201200056
The Asymmetric Formal Hetero-Diels–Alder Reaction of Methyl
(E)-4-Aryl-2-oxo-3-butenoates Catalyzed by [Sc(OTf)3/pybox] Complexes
Giovanni Desimoni,*[a] Giuseppe Faita,[a] Alessandro Livieri,[a] Mariella Mella,[a]
Laura Ponta,[a] and Massimo Boiocchi[b]
Dedicated to the memory of Professor Jürgen Sauer[‡]
Keywords: Asymmetric catalysis / Cyclization / Michael addition / N ligands
The complex between scandium triflate and pybox is a good
ray crystal analysis. Their absolute configuration was then
enantioselective catalyst for the reaction between methyl (E)-
related to that of 7 by stereospecific isomerization and/or de-
4-aryl-2-oxo-3-butenoates (1) and enol silyl ethers (2). The silylation reactions. If aryl-butenoates 1 are coordinated to
main products with (4ЈS)-isopropyl-, (4ЈS)-phenyl-, and
(4ЈS,5ЈS)-4-TIPS-pybox are methyl (4S,4aS,8aS)-4-aryl-
8a-trialkylsiloxy-hexahydro-4H-chromen-2-carboxylates (7),
which are obtained in good yield and enantioselectivities of
more than 95% ee. Because the reaction gives products with
a trans ring junction that cannot derive from a conventional
concerted hetero-Diels–Alder pathway, the mechanism in-
volved in the enantioselective catalytic cycle and the origin
of the stereoinduction were investigated. The structures of
two desilylated products, 8b and 9b, were determined by X-
the catalyst, forming reacting complexes characterized by
the five-membered structure 11, these rigid reacting interme-
diates give a face discrimination that is determined by the
configuration of the pybox 4Ј-substituent. The resulting face-
selective attack of enol silyl ethers to coordinated 1 gives an
enantioselective tandem Mukaiyama–Michael addition/in-
tramolecular ring closure reaction, which is an enantioselec-
tive formal hetero-Diels–Alder reaction, which rationalizes
the absolute configuration of the reaction products.
Introduction
α,β-Unsaturated carbonyl derivatives are suitable sub-
strates for several reactions involving either the entire sys-
tem (hetero-Diels–Alder reaction, hydrogenation), the carb-
onyl group (alkylation, aldol reaction), the double bond
(cyclopropanation, epoxidation, Diels–Alder reaction, 1,3-
dipolar cycloadditions), or simply an attack to the β-carbon
atom of the C=C–C=O fragment (Michael addition, Frie-
del–Crafts reaction). Amongst them, (E)-4-aryl-2-oxo-3-bu-
tenoates 1 have a specific advantage due to the presence of
a further carbonyl group in the αЈ-position, because the 1,2-
dicarbonyl system may coordinate to a Lewis acid (Fig-
ure 1). Such coordination increases the reactivity of the sub-
strate, and the use of asymmetric catalysts may induce
enantioselectivity in the reactions.
Figure 1. Reactivity of bicoordinating methyl (E)-4-aryl-2-oxo-3-
butenoates (1).
Several examples of enantioselective variants of the
above reactions have been reported (Figure 1). Although
certain reactions have been only occasionally investigated
(e.g., alkylation,[1] hydrogenation,[2] and the aldol reac-
tion[3]), the classical Michael reaction has been extensively
studied,[4,5] as has its oxa-, thio-, and Henry variants,[6–8]
and tandem Michael–cyclization processes.[9,10] Moreover,
the Friedel–Crafts reaction has found many applications in
both aromatic and heterocyclic systems (especially with in-
doles),[11–17] and all these reactions have been asymmetri-
cally performed by using box[18] and pybox[19] based chiral
catalysts.
[‡] For his important insights into the Diels–Alder reaction
[a] Department of Chemistry, University of Pavia,
Viale Taramelli 10, 27100 Pavia, Italy
Fax: +39-0382-987323
E-mail: desimoni@unipv.it
[b] Centro Grandi Strumenti, University of Pavia,
Via Bassi 8, 27100 Pavia, Italy
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201200056.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1

Job/Unit: O20056
/KAP1
Date: 05-04-12 16:12:07
Pages: 14
G. Desimoni et al.
FULL PAPER
The main reactions undertaken with compounds of type
1 involve cycloadditions,[20] which can also be fruitfully per-
formed under organocatalysis conditions.[21] The Diels–Al-
der reaction between methyl (E)-4-aryl-2-oxo-3-butenoates
(1) and cyclopentadiene, catalyzed by [CuII/C3-symmetric
tris(oxazolines)], gave cycloadducts with acceptable
Results and Discussion
Cyclization
To investigate the anomalous results sometimes ob-
enantioselectivities.[22] When the catalyst used was the ScIII served, the reaction between methyl (E)-4-aryl-2-oxo-3-but-
triflate complex of (4ЈS,5ЈS)-2,6-bis[4Ј-(triisopropylsilyl)- enoates 1a and 1b, and 1-trialkylsiloxy-1-cyclohexenes 2a–c
oxymethyl-5Ј-phenyl-1Ј,3Ј-oxazolin-2Ј-yl]pyridine,
then with [Sc(OTf)3/pybox] complexes as the chiral catalysts
both the Diels–Alder and hetero-DielsAlder adducts were were studied. These catalysts have already been shown to
obtained with excellent enantiomeric excess.[23–25] This be very efficient in the coordination of 1 and, therefore,
schizophrenic behaviour is possible because cyclopentadi- they can be expected to be able to induce efficient chiral
ene may behave either as a diene or a dienophile; the use of induction regardless the reaction mechanism.
vinyl ethers instead of cyclopentadiene forced the reaction
Among the different pybox ligands, the enantiomeric
towards the hetero-Diels–Alder pathway, and when the cy- couples of (4Ј-isopropyl)- and (4Ј-phenyl)-2,6-bis(4Ј-isoprop-
cloaddition was catalysed by either CuII/box or pybox com- yl-2Ј-oxazolidin-2Ј-yl)pyridines, (4ЈS)- and (4ЈR)-3a, (4ЈS)-
plexes,[26,27] endo adducts retaining the stereochemistry of and (4ЈR)-3b, are commercial products, hence their com-
the dienophile were always obtained.
plexes with ScIII triflate were selected as catalysts (Figure 2).
This predictable result has some exceptions. The reaction In addition to these, the ScIII complex of (4ЈS,5ЈS)-2,6-
of 1 and ketene diethyl acetal, again an electron-rich alkene, bis[4Ј-(triisopropylsilyl)oxymethyl-5Ј-phenyl-1Ј,3Ј-oxazolin-
occurs first through an enantioselective aldol reaction lead- 2Ј-yl]pyridine [(4ЈS, 4ЈS)-3c] was also tested in the reactions
ing to diethyl 2-hydroxy-2-styrylsuccinates. The addition between 1a, 1b and 2a, because of its excellent results as an
product may add a second equivalent of ketene diethyl enantioselective catalyst for reactions involving 1.[14,15,23,24]
acetal to give methyl tetraethoxy-2-styryl-tetrahydropyran-
The reactions were performed at –20 °C with 3 Å molec-
2-carboxylates with excellent enantioselectivity.[28] The reac- ular sieves (MS) in dichloromethane as solvent, with or
tion between 1a and 1-trialkylsiloxy-1-cyclohexenes (2a and without hexafluoroisopropanol (HFIP) as additive, and
2b) as electron-rich alkenes, when catalyzed by SnCl4 or with 10 mol-% catalyst (Scheme 1 and Table 1). In the pres-
[Eu(fod)3], gives four different methyl 4-phenyl-8a-trialkyl- ence of the additive, the reaction was complete in approxi-
siloxy-4a,5,6,7,8,8a-hexahydro-4H-chromene-2-caboxylates
mately 24 hours, whereas without HFIP the reaction was
4–7 (Scheme 1).[29] In products 4 and 5 the substituents at sluggish and, after several days, significant amounts of un-
the 4a and 8a positions retain the stereochemistry of the reacted 1 were always recovered (Table 1, entries 2, 5, 8 and
starting dienophile, but in 6 and 7 the trans configuration 14).
of the substituents at the same positions cannot derive from
The reaction between 1a and 2a with HFIP and all the
a conventional concerted hetero-Diels–Alder pathway. This pybox chiral ligands gave one single silylated adduct in ac-
apparent violation of the principle of retention of configu- ceptable yields and with excellent enantiomeric excess. A
ration in a [4+2] cycloaddition was the reason for our inter- comparison of its 1H NMR spectrum with the data re-
est in this reaction.
ported in the literature[29] confirmed that the adduct
Scheme 1. Reaction between 4-aryl-2-oxo-3-butenoates 1 and 1-trialkylsiloxy-1-cyclohexenes 2.
2
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O20056
/KAP1
Date: 05-04-12 16:12:07
Pages: 14
[Sc(OTf)3/pybox]-Catalyzed Hetero-Diels–Alder Reaction
obtained with very high enantiomeric excess, however, two
1
new adducts were also isolated. Their H NMR spectra al-
lowed the structures of 5ba (low yields, good ee), and 6ba
(trace amount) to be assigned (Table 1, entries 11 and
12).[29] All these products, which are crystalline solids as
racemates, are oily products when obtained with high enan-
tiomeric excess; all attempts to obtain crystals that were
suitable for the determination of their absolute configura-
tion failed.
In almost all the previously described reactions, two de-
silylated products were also isolated. Two of them (9a and
9b) have already been described in the enantioselective reac-
tion of 1a and 1b with cyclohexanone, and their configura-
tion will be discussed later.[30] The second couple of desil-
ylated products was isolated in high enantiomeric purity,
and the sample detailed in Table 1, entry 10 gave crystals
that were suitable for X-ray analysis, allowing the product
Figure 2. Pybox ligands tested in the reactions between 1 and 2.
was methyl (4S*,4aS*,8aS*)-4-phenyl-8a-trimethylsiloxy-
4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate (7aa)
(Table 1, entries 1, 3, 4, 6 and 7). The reaction without
HFIP became interesting when [Sc(OTf)3/3c] complex was
used as the catalyst because, under these conditions, the
only product obtained in acceptable yield was 5aa, which
was obtained with an enantiomeric excess of 96% (Table 1,
entry 8).
The reaction between 1b and 2a with catalysts based on
3a again gave a single silylated product (7ba). With catalysts
Figure 3. ORTEP view of 8b with the absolute 4R,4aR,8aR config-
uration (ellipsoids are drawn at the 50% probability level; color
derived from 3b the main product was again 7ba, which was code: grey for C, red for O, orange for Br).
Table 1. Catalytic enantioselective cycloaddition reaction of methyl (E)-4-aryl-2-oxo-3-butenoates (1a and 1b) with 1-trimethylsiloxy-1-
cyclohexene (2a) catalyzed by [Sc(OTf)3/3].
Entry Reagent Pybox 3
HFIP
Time
5x,y
6x,y
7x,y
8x
9x
equiv.[a]
Yield [%], ee [%] Yield [%], ee [%] Yield [%], ee [%]
Yield [%], ee [%] Yield [%], ee [%]
(absol. config.)
(absol. config.)
(absol. config.)
(absol. config.)
(absol. config.)
1
2
3
4
5
6
7
8
9
1a
1a
1a
1a
1a
1a
1a
1a
1b
(4ЈS)-3a
(4ЈS)-3a
(4ЈR)-3a
(4ЈS)-3b
(4ЈS)-3b
(4ЈR)-3b
(4ЈS,5ЈS)-3c
(4ЈS,5ЈS)-3c
(4ЈS)-3a
1
–
1
1
–
1
1
–
1
24 h
4 d[b]
24 h
18 h
4 d[c]
18 h
24 h
4 d[d]
24 h
–
–
–
trace
–
–
–
trace
–
trace
–
Ͻ2
–
44, 98
18, 98
35, 99
49, 98
38, 98
55, 98
31, 98
4
30, 93
37, 94
41, 95
34, 86
28, 60
40, 85
–
21
16, 99
15, 98
15, 98
Ͻ1
Ͻ1
Ͻ1
–
6
4, 67
–
Ͻ2, 80
–
26, 96
–
38, 99
39, 96
(4S,4aS,8aS)
33, 98
(4R,4aR,8aR)
54, Ͼ99
(4S,4aS,8aS)
62, 99
(4R,4aR,8aR)
31, 99
(4S,4aS,8aS)
15, 11
(4S,4aS,8aS)
48, 99
(4R,4aR,8aR)
12, 92
(4S,4aS,8aS)
15, 94
(4R,4aR,8aR)
10, 34
10
11
12
13
14
1b
1b
1b
1b
1b
(4ЈR)-3a
1
1
1
1
–
24 h
24 h
24 h
24 h
7 d[e]
–
–
Ͻ2, 50
Ͻ1
(4ЈS)-3b
8, 94
(4S,4aS,8aR)
11, 90
(4R,4aR,8aS)
–
Ͻ2
3
(4ЈR)-3b
Ͻ2
(4ЈS,5ЈS)-3c
(4ЈS,5ЈS)-3c
–
10, racemate
Ͻ2
(4S,4aS,8aS)
12
12, 50
3, 50
(4S,4aS,8aR)
(4S,4aR,8aR)
[a] Equivalents with respect to 1. [b] 30% recovered 1a. [c] 20% recovered 1a. [d] 34% recovered 1a. [e] 45% recovered 1b.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3

Job/Unit: O20056
/KAP1
Date: 05-04-12 16:12:07
Pages: 14
G. Desimoni et al.
FULL PAPER
Table 2. Catalytic enantioselective cycloaddition reaction of methyl (E)-4-aryl-2-oxo-3-butenoates (1a and 1b) with 1-tert-butyldimethylsi-
loxy-1-cyclohexene (2b) or 1-triisopropylsiloxy-1-cyclohexene (2c) catalyzed by [Sc(OTf)3/3] with HFIP[a] as additive.
Entry Reagents Pybox 3 Time
4x,y
5x,y
6x,y
7x,y
8x
9x
Yield [%], ee [%] Yield [%], ee
Yield [%], ee
Yield [%], ee [%] Yield [%], ee [%] Yield [%], ee [%]
[%]
[%]
(absol. config.)
5, ca. 80
4, ca. 80
trace
trace
–
(absol. config.) (absol. config.) (absol. config.)
(absol. config.)
9, 74
16, 68
10
15
30, 64
(4S,4aS,8aS)
26, 66
(4R,4aR,8aR)
14, 60
(4S,4aS,8aS)
15, 65
(absol. config.)
9, 74
8, 70
trace
trace
25, 62
(4S,4aR,8aS)
17, 65
(4R,4aS,8aR)
6, 48
(4S,4aR,8aS)
6, 50
1
2
3
4
5
1a + 2b
1a + 2b
1a + 2b
1a + 2b
1b + 2b
(4ЈS)-3a
(4ЈR)-3a 24 h
(4ЈS)-3b 2 d
(4ЈR)-3b 2 d
24 h
–
–
trace
trace
–
3, 45
3, 30
Ͻ2
Ͻ2
2
22, Ͼ99
22, Ͼ99
70, 99
73, 99
(4ЈS)-3a
24 h
38, Ͼ99
(4S,4aS,8aS)
37, 99
(4R,4aR,8aR)
53, 98
6
7
8
9
1b + 2b
1b + 2b
1b + 2b
1b + 2c
(4ЈR)-3a 24 h
–
–
–
–
–
Ͻ2
2
(4ЈS)-3b
24 h
trace
trace
–
(4S,4aS,8aS)
44, 98
(4ЈR)-3b 24 h
5 d[b]
3
(4R,4aR,8aR)
–
(4R,4aR,8aR)
21, 65
(4R,4aS,8aR)
trace
(4ЈS)-3a
–
(4S,4aS,8aS)
[a] 1 mol. [b] 55% recovered 1b.
methyl 4-(4-bromophenyl)-4a,5,6,7,8,8a-hexahydro-8a-hy- ence of the [ScIII triflate/3] complexes under the reaction
droxy-4H-chromene-2-carboxylate (8b) to be assigned conditions. None of the main adducts 7 isomerized after
4R,4aR,8aR absolute configuration (Figure 3).
one week, either at –20 °C or at ambient temperature.
The reactions between 1a/1b and 2b/2c were then exam-
To determine the absolute configuration of the silylated
ined under the same conditions (HFIP as additive), with the adducts, a structure correlation with known desilylated
ScIII complexes of 3a and 3b, which were the most efficient products was planned. Enantiomerically pure 7ba, obtained
catalysts investigated; the results are reported in Table 2.
from the reaction described above (Table 1, entry 9), was
The reactions with 1-tert-butyldimethylsiloxy-1-cyclohex- treated with tetrabutyammonium fluoride (TBAF) in tetra-
ene (2b) gave acceptable (sometimes good) yields of 7ab or hydrofuran (THF) at 0 °C (Scheme 2). After 10 min all
7bb, always with excellent enantiomeric excesses. In ad- starting material disappeared and two products were iso-
dition to traces of 6ab or 6bb, low yields of 4ab were ob- lated by column chromatography. The first product,
tained by using the enantiomeric catalysts derived from 3a (4S,4aS,8aS)-8b, was already obtained from the reaction
(Table 2, entries 1 and 2). In each reaction, significant (Table 1, entry 9). The second product, 9b, has already been
amounts of desilylated products 8a/8b and 9a/9b were al- described in the literature,[30] and was obtained here with
ways obtained, but the enantioselectivities were in the range 98% enantiomeric excess. Due to the discrepancy between
50–70% ee.
the [α]D value of 9b and the value reported in the literature
The reactions with 1-triisopropylsiloxy-1-cyclohexene (see Exp. Section),[30] its crystal structure was determined
(2c) probably suffered from steric congestion of the reagent, and the absolute configuration was found to be
which greatly decreased the reactivity, and only one exam- (4S,4aR,8aS)-9b (Figure 4).
ple of the resulting reaction with (4ЈS)-3a, which gave 21%
yield of (4S,4aS,8aS)-8b, is reported (Table 2, entry 9).
In conclusion, the [Sc(OTf)3/pybox] catalyzed reactions
between methyl (E)-4-aryl-2-oxo-3-butenoates (1a and 1b)
and 1-trialkylsiloxy-1-cyclohexenes (2a and 2b) are strongly
diastereo- and enantioselective processes; the best condi-
tions lead to the formation of adducts 7 in acceptable yields
and very high enantiomeric excess. Remarkably, this adduct
derives from a formal hetero-Diels–Alder reaction, hence,
the stability of the products, the mechanism involved in the
catalytic cycle, the origin of the desilylated products, and
the stereochemical outcome of the reaction are all interest-
Scheme 2. Reaction of (4S,4aS,8aS)-7ba with TBAF.
ing features to be examined.
Because the desilylation of 7ba depicted in Scheme 2 re-
sulted in the formation of (4S,4aS,8aS)-8b and
Desilylation and Isomerization of the Enantiomerically Pure
(4S,4aR,8aS)-9b, two adducts with the same (4S) configura-
Products
tion, then the absolute configuration of the starting sil-
Control experiments were performed to check the sta- ylated product 7ba obtained from the experiments in
bility of the enantiomerically enriched adducts in the pres- Table 1, entries 9, 11 and 13 must be (4S,4aS,8aS). This ab-
4
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O20056
/KAP1
Date: 05-04-12 16:12:07
Pages: 14
[Sc(OTf)3/pybox]-Catalyzed Hetero-Diels–Alder Reaction
Figure 4. ORTEP view of 9b with the absolute 4S,4aR,8aS configu-
ration (ellipsoids are drawn at the 50% probability level; color
code: grey for C, red for O, orange for Br).
solute configuration required not only the expected epi-
merization at C8a, but also a more unusual change in the
configuration of the C4a centre.
The next step was to investigate the behaviour of enan-
tiomerically pure (4R,4aR,8aR)-7ba in CH2Cl2 with
Sc(OTf)3 as Lewis acid. The reaction occurred at ambient
temperature and depended on the presence of water, which
may influence the acidic character of the reagent
Scheme 3. Stereospecific reactions of (4R,4aR,8aR)-7ba with scan-
dium triflate under different conditions.
(Scheme 3). With one equivalent of [Sc(OTf)3/H2O], after was obtained. The formation of the latter from the former
18 h, three enantiomerically pure products were obtained, was confirmed by allowing (4S,4aS,8aS)-8b to come to am-
all of which are reported in Table 1. One of the products bient temperature for 4 h. It was rather surprising that, un-
was (4R,4aR,8aR)-8b (19% yield), which retained the con- der the same conditions, (4S,4aR,8aS)-9b did not react.
figuration at C4 and C4a of the starting compound 7ba.
The reactions depicted in Schemes 2, 3 and 4 allowed the
This allowed the absolute configuration of the other reac- absolute configuration of the products derived from 2a to
tion products to be assigned: (4R,4aR,8aS)-5ba (10% be assigned. A similar approach was attempted with 7bb,
yield), in which the configuration at C4a remained, and which was the product of the reaction between 1-tert-butyl-
(4R,4aS,8aS)-6ba (20% yield), in which the configuration dimethylsiloxy-1-cyclohexene (2n) and 1b, catalyzed by
at C4a changed. When the reaction was performed without [Sc(OTf)3/(4ЈR)-3a or (4ЈR)-3b] (Table 2, entries 6 and 8).
water, (4R,4aR,8aR)-8b was formed as the main product This product is more stable against desilylation than 7ba,
(67% yield) (Scheme 3).
and with Sc(OTf)3 at ambient temperature it simply re-
These results, which full support the absolute configura- arranged into a mixture of enantiomerically pure 5bb (30%
tion given to (4S,4aS,8aS)-7ba from its reaction with TBAF yield) and 6bb (55% yield); the latter was isolated from the
reported in Scheme 2, show that the stereocentres at C4a reaction reported in Table 2, entry 8 with a low yield and
(proton) and C8a (the TMSO group) may isomerise. If without any asymmetric induction (Scheme 5).
ScIII, acting as a Lewis acid, coordinates the TMSO oxygen
Whereas the reaction of 7bb with trifluoroacetic acid at
atom, then epimerisation at C8a may occur and –20 °C selectively gave the enantiomerically pure product
(4R,4aR,8aS)-5ba can form. This epimerisation has already 5bb (Scheme 5), the same reaction performed at 0 °C for 4 h
been reported in the literature to take place in the presence provided vital information that allowed the absolute config-
of SnCl4.[29] If ScIII coordinates the chromen oxygen atom, uration of the products to be assigned. The complex mix-
a ring-opening reaction occurs with loss of the C4a proton ture was separated and, together with some unreacted start-
and formation of the assumed intermediate reported in ing material and a small amount of enantiomerically pure
Scheme 3. Ring closure of the intermediate and protonation 6bb, the desilylated products of known absolute configura-
at C-4a should give (4R,4aS,8aS)-6ba.
tions (4R,4aR,8aR)-8b (30% yield) and (4R,4aS,8aR)-9b
An alternative desilylating reagent is trifluoroacetic acid, (15% yield) were isolated. Therefore the absolute confi-
which was tested on enantiomerically pure (4S,4aS,8aS)- gurations (4R,4aR,8aR)-7bb, (4R,4aR,8aS)-5bb and
7ba in CH2Cl2. It was found that both temperature and (4R,4aS,8aS)-6bb could be assigned to the products re-
time had a significant impact on the product composition. ported in Scheme 5.
Two conditions are reported in Scheme 4: at –20 °C, within
The lack of a bromine atom on the aryl group made 7ab
a short time, a quantitative yield of enantiomerically pure less stable than 7bb towards trifluoroacetic acid. When the
(4S,4aS,8aS)-8b was obtained, however, at 0 °C an equimo- product of the reaction between 2b and 1a, catalyzed by
lecular mixture of 8b and its dehydrated product (4S)-10b [Sc(OTf)3/(4ЈR)-3a or -3b] (Table 2 entries 2 and 4), was
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5

Job/Unit: O20056
/KAP1
Date: 05-04-12 16:12:07
Pages: 14
G. Desimoni et al.
FULL PAPER
Scheme 4. Stereospecific reactions of (4S,4aS,8aS)-7ba with trifluoroacetic acid under different conditions.
Scheme 5. Stereospecific reactions of (4R,4aR,8aR)-7bb with scan-
dium triflate or trifluoroacetic acid under different conditions.
stirred for 10 min at –20 °C with CF3CO2H in CH2Cl2, all
the starting material disappeared. From the reaction mix-
ture, three products were isolated (Scheme 6): (i) pure 8a,
the same enantiomer that was obtained from the reaction
described in Table 1, entry 3, (ii) the product of its formal
dehydration, 10a, and (iii) enantiomerically pure 5ab (the
same enantiomer obtained from the reaction described in
Table 2, entry 4).
Scheme 6. Stereospecific reactions of 7ab with scandium triflate or
trifluoroacetic acid under different conditions.
When enantiomerically pure 7ab was treated with lated: the desilylated products 8a and 9a (16 and 25% yield,
Sc(OTf)3 at ambient temperature for 18 h, 8a, 9a and 10a respectively), which were enantiomerically pure and with
were obtained (the former products being the same enantio- opposite configuration to the products obtained from the
mers obtained from the reaction described in Table 1, en- reaction described in Table 1, entry 3, and the dehydration
try 3, and 10a being identical to the enantiomer obtained product 10a (15% yield), as the enantiomer of the com-
by treatment with CF3CO2H, Scheme 6). When the same pound illustrated in Scheme 6. When the reaction with tri-
reaction was performed in the presence of 3 Å molecular fluoroacetic acid was stirred at ambient temperature for
sieves (MS) a mixture of 6ab (61% yield of the same enanti- 90 min, only 9a and 10a could be isolated in 25 and 15%
omer isolated in very low yield from the reaction reported yield, respectively. A rationale could involve the specific
in Table 2, entry 2) and 5ab (22% yield) was obtained conversion of 8a into 10a. This was proved by independent
(Scheme 6).
experiments starting from 9a and 8a, which were allowed to
The last example of this section concerns the treatment react under the same severe conditions. Whereas the former
of product 5aa, derived from the reaction reported in product did not evidence any reaction, the latter compound
Table 1, entry 8, with trifluoroacetic acid in CH2Cl2 at am- was stereoselectively converted into 10a with up to 85%
bient temperature. After 15 min, three products were iso- yield (Scheme 7).
6
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O20056
/KAP1
Date: 05-04-12 16:12:07
Pages: 14
[Sc(OTf)3/pybox]-Catalyzed Hetero-Diels–Alder Reaction
complex 11, which adopts an octahedral geometry as al-
ready reported (Scheme 8).[23,31] ScIII is coordinated by the
three nitrogen atoms of (4ЈS)-3a, the ketonic CO group of
1 is in an equatorial position, with the ester moiety occupy-
ing one apical position, and a triflate anion in the residual
apical position. If the favoured attack of 1-trialkylylsiloxy-
1-cyclohexene (2) occurs on 11 from the less shielded Re-
face of bound 1, then the (4S)-configuration of stereocentre
of 7 can be rationalized.
Scheme 7. Stereospecific reactions of 5aa with scandium triflate or
trifluoroacetic acid under different conditions.
All the stereochemical relationships reported in
Schemes 6 and 7 are of paramount importance in under-
standing the univocal configuration of the stereocentre 4
for all the different products of the reactions. These consid-
erations do not allow the absolute configurations to be at-
tributed, but a reasonable analogy between the behaviour
of 1a and 1b under the same catalytic conditions justifies
the specific configurations reported in the schemes for each
product.
Conclusions
The [ScIII/pybox]-catalyzed reaction between methyl (E)-
4-aryl-2-oxo-3-butenoates (1a and 1b) and 1-trialkylsiloxy-
1-cyclohexenes (2a and 2b) enantioselectively gives methyl
(4S,4aS,8aS)-4-aryl-8a-trialkylsiloxy-4a,5,6,7,8,8a-hexahy-
dro-4H-chromene-2-carboxylates (7) as the main reaction
products when pybox are (4ЈS)-3a, (4ЈS)-3b and (4ЈS,4ЈS)-
3c. This means that the sense of induction in the reaction
derives from the configuration of the pybox substituent at
the 4Ј-position.
Scheme 8. Catalytic cycle of the enantioselective formal hetero-
Diels–Alder reaction between 1a and 2.
Control of the (4a) stereocentre derives from an “anti”
vs. “syn” approach of the silyl vinyl ether 2 to the coordi-
The question can be asked: What kind of reaction be- nated Michael acceptor 1. Formation of the second ste-
tween α,β-unsaturated carbonyl heterodienes and electron- reocentre as (4aS) to give the Michael intermediate 12 stems
rich dienophiles enantioselectively gives products in which from a minimization of the steric interaction that is feasible
the configuration of the dienophile is not retained in the with an “anti” attack of the Re-face of 2, with the trialkylsi-
product? This formal hetero-Diels–Alder reaction has been lyloxy group removed from the congested core of complex
interpreted as being the result of a stepwise sequence of two 11. Hence, this is the second participation of the pybox
reactions: a Mukaiyama–Michael addition, in which the complex in inducing the stereochemical configuration of the
C4–C4a bond is formed, and an intramolecular ring closure product because, with a different catalyst (SnCl4), the more
in which the Lewis acid coordinated oxygen atom of 1 at- favourable “syn” addition was obtained.[29]
tacks the carbon atom of the oxonium intermediate to give
The third step is the intramolecular cyclization of the
oxygen atom to the oxonium carbon atom of 12. A simple
the C1–C8a bond.[29]
The enantioselective variant of the reaction requires inspection of this model, in the correct conformation with
three steps of stereocontrol to give enantiomerically pure 7. syn hydrogen atoms, allows the advantage of an attack that
The first control drives the approach of 2 to the reacting affords the thermodynamically favoured trans ring junction
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
7

Job/Unit: O20056
/KAP1
Date: 05-04-12 16:12:07
Pages: 14
G. Desimoni et al.
FULL PAPER
lowing literature methods. (4Ј-Isopropyl)- and (4Ј-phenyl)-2,6-
bis(4Ј-isopropyl-2Ј-oxazolidin-2Ј-yl)pyridines (4ЈS)-3a and (4ЈS)-
3b, and their enantiomers (4ЈR)-3a and (4ЈR)-3b were Aldrich ACS
reagents; (4ЈS,5ЈS)-2,6-bis[4Ј-(triisopropylsilyl)oxymethyl-5Ј-phen-
yl-1Ј,3Ј-oxazolin-2Ј-yl]pyridine (3c) was prepared as described pre-
viously.[38]
with formation of the third stereocentre as (8aS) to be seen.
Thus, when the catalyst employed is [ScIII/(4ЈS)-3a], the
overall stepwise reaction gives (4S,4aS,8aS)-7 (Scheme 8).
In conclusion, the scandium triflate complexes of pybox
3 are efficient catalysts for enantioselective reactions with
methyl (E)-4-aryl-2-oxo-3-butenoates (1). These reagents,
behaving as α-dicarbonyl derivatives, operate through bi-
dentate coordination to form rigid complexes that are char-
acterized by a five-membered structure. These reacting in-
termediates give excellent face discriminations, which are
essentially determined by the configuration of the pybox 4Ј
substituent, in reactions involving several reagents: cyclo-
pentadiene in enantioselective Diels–Alder and hetero-
Diels–Alder cycloadditions,[23,24] and either electron-rich
benzenes or indoles in Friedel–Crafts reactions.[14,15] The
same reacting intermediate involved in the above reactions,
which is represented as 11 with (4ЈS)-3a as pybox in
Scheme 8, is found to induce the same face selectivity with
1-trialkylsiloxy-1-cyclohexenes in a tandem-Mukaiyama–
Michael addition/intramolecular ring closure reaction,
which is a formal hetero-Diels–Alder reaction.
General Procedure for the Reaction of Methyl (E)-4-Aryl-2-oxo-3-
butenoates (1a and 1b) and 1-Trialkylsiloxy-1-cyclohexenes 2a–c
Catalyzed by [Sc(OTf)3/pybox (3)]: Solid Sc(OTf)3 (14.8 mg,
0.03 mmol) and 3 Å molecular sieves (ca. 40 mg) were added to
a solution of methyl 4-aryl-2-oxo-3-butenoate 1 (0.30 mmol), the
selected pybox (0.03 mmol) and, when required, HFIP (53 μL,
0.50 mmol), in anhydrous CH2Cl2 (0.3 mL) in a vial sealed with a
rubber septum. The reaction mixture was stirred for 15 min at am-
bient temperature, then cooled to –20 °C and 1-trialkylsiloxy-1-cy-
clohexene 2 (0.5 mmol) was added. The reaction mixture was
stirred at –20 °C for the time reported in Table 1 and Table 2, then
quenched with a small amount of silica gel and the dispersed mix-
ture was separated by column chromatography on silica gel. When
the reaction mixture has only one silylated adduct (e.g., Table 1,
entries 4 and 6) the column required 30 cm length and 1.5 cm dia-
meter, and the eluant was first cyclohexane/ethyl acetate, 90:10 (to
separate 7), then cyclohexane/ethyl acetate, 75:25 (to separate 8 and
9). When the reaction mixture contained more than one silylated
adduct (e.g., Table 1, entries 11 and 12) the column required 60 cm
length and 1.5 cm diameter, and the eluant was first cyclohexane/
ethyl acetate, 98:2 (to separate the different silylated products), then
elution was continued with cyclohexane/ethyl acetate, 75:25 (to
separate 8 and 9 if present). HPLC analysis, column, eluant, experi-
mental conditions, and retention times, as well as the IR, 1H-,
13C NMR, the [α]D values, and the elemental analyses are given for
each specific product. Whereas the racemates of 4, 5 and 7 are
crystalline solids, the enantiomerically enriched products are always
oils.
The above reaction, in which the main product has a
trans-fused junction in a bicyclic six to six-membered rings,
can be regarded as an extension to the hetero-parent reac-
tion of the Danishefsky trans-Diels–Alder paradigm.[32]
Experimental Section
General Remarks: Melting points were determined by the capillary
method and are uncorrected. 1H and 13C NMR spectra were re-
corded at 300 and 75 MHz, respectively. IR spectra were recorded
with a Perkin–Elmer RX I spectrophotometer. Optical rotations
were measured with a Perkin–Elmer 241 polarimeter. Separation
and purification of the products was carried out by column
chromatography using Merck silica gel 60 (230–400 mesh) and the
purity of each separated adduct was checked by comparing their
1H NMR spectra with those of pure samples. The enantiomeric
excess (ee) of the products were determined by HPLC using Daicel
columns: Chiralpak AD column (hexane/iPrOH, 98:2) or Chiralcel
OD column (hexane/iPrOH, 90:10), taking into consideration that
small differences in eluant composition, particularly when the elu-
ant is less polar, may give significantly different retention times.
Reaction Between 1a and 2a Catalyzed by [Sc(OTf)3/(R)- or (S)-
3a,b]: The reactions given in Table 1, entries 1–6 were performed
by following the general method described above. For reaction con-
ditions, type of separation, yields, and ee of each product see the
general procedure and Table 1. The following products were ob-
tained from (S)-3a as pybox (Table 1, entry 1).
Methyl (4S*,4aS*,8aS*)-4-Phenyl-8a-trimethylsiloxy-4a,5,6,7,8,8a-
hexahydro-4H-chromen-2-carboxylate (7aa): Oily product. Chi-
ralpak AD column (hexane/iPrOH, 98:2; flow rate: 0.50 mL/min):
tR = 10.3 (major), 11.6 min (minor); 98% ee; [α]D = +75.5 (c = 1.2,
1
CHCl3). H NMR (300 MHz, CDCl3, 25 °C, TMS): δ = 7.28–7.21
Materials: Dichloromethane was hydrocarbon-stabilised Aldrich
ACS grade, distilled from calcium hydride and used immediately;
other solvents were dried according to standard procedures. Scan-
dium triflate was Aldrich ACS reagent grade; powdered molecular
sieves 3 Å (Aldrich reagent grade) was heated under vacuum at
300 °C for 5 h and kept in sealed vials in a dryer. Methyl (E)-2-
oxo-4-phenylbut-3-enoate (1a) was prepared following a literature
method[27,33] from its potassium salt, which was obtained from
benzaldehyde and pyruvic acid in the presence of KOH, and was
esterified with methanol [1a was obtained as yellow needles from
diisopropyl ether; m.p. 69–70 °C (Lit.[27] 70–71 °C)]. Following the
same procedure, starting from p-bromobenzaldehyde, methyl (E)-
2-oxo-4-(4-bromophenyl)-but-3-enoate (1b) was prepared in 35%
yield as yellow needles from methanol [m.p. 120 °C (Lit.[34]
122 °C)]. Hexafluoroisopropanol (HFIP) was Aldrich ACS reagent
grade, and 1-trimethylsiloxy-1-cyclohexene (2a) was Alfa Aesar rea-
gent grade (98%), 1-tert-butyldimethylsiloxy-1-cyclohexene (2b)[35]
and 1-triisopropylsiloxy-1-cyclohexene (2c)[36,37] were prepared fol-
3
(m, 5 H, ArH), 6.38 (d, JH,H = 4.8 Hz, 1 H, 3-H), 3.85 (s, 3 H,
3
3
OCH3), 3.49 (dd, JH,H = 8.1, JH,H = 4.8 Hz, 1 H, 4-H), 2.16 (m,
3
3
3
1 H), 2.15 (ddd, JH,H = 12.2, JH,H = 8.1, JH,H = 3.7 Hz, 1 H,
4a-H), 1.7–1.5 (m, 4 H), 1.3 (m, 2 H), 0.95 (m, 1 H), 0.15 (s, 9 H,
SiMe3) ppm. 13C NMR (75 MHz, CDCl3, 25 °C, TMS): δ = 163.6,
141.3, 139.3, 130.7, 127.0, 126.2, 115.3, 99.9, 52.0, 44.1, 40.3, 38.0,
27.3, 25.9, 22.8, 1.7 ppm. H and 13C NMR spectra were identical
1
to those reported in ref.[29] IR (nujol): ν = 1744 (C=O) cm–1.
˜
Methyl (4S*,4aS*,8aS*)-8a-Hydroxy-4-phenyl-4a,5,6,7,8,8a-hexa-
hydro-4H-chromen-2-carboxylate (8a): White crystals; m.p. 55 °C
(hexane). Chiralcel OD column (hexane/iPrOH, 90:10; flow rate:
0.50 mL/min): tR = 9.3 (major), 11.1 min (minor); 93% ee; [α]D
=
+48.6 (c = 1.1, CHCl3). 1H NMR (300 MHz, CDCl3, 25 °C, TMS):
δ = 7.38–7.22 (m, 5 H, ArH), 6.34 (dd, 4JH,H = 1.4, 3JH,H = 2.2, Hz
3
3
1 H, 3-H), 4.28 (dd, JH,H = 6.1, JH,H = 2.2 Hz, 1 H, 4-H), 3.84
(s, 3 H, OCH3), 3.42 (br. s, 1 H, OH), 2.29 (m, 1 H), 1.92 (dt, 3JH,H
3
3
= 8.4, JH,H = 6.1, JH,H = 6.1 Hz, 1 H, 4a-H), 1.7–0.9 (m, 7
8
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O20056
/KAP1
Date: 05-04-12 16:12:07
Pages: 14
[Sc(OTf)3/pybox]-Catalyzed Hetero-Diels–Alder Reaction
H) ppm. 13C NMR (75 MHz, CDCl3, 25 °C, TMS): δ = 163.5, Methyl
(4R,4aR,8aR)-4-(p-Bromophenyl)-8a-trimethylsiloxy-
140.8, 140.7, 128.5, 128.2, 126.4, 113.0, 98.6, 52.1, 41.7, 38.6, 38.3,
4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate (7ba): Oily
product. Chiralpak AD column (hexane/iPrOH, 98:2; flow rate:
26.8, 24.6, 22.9 ppm. IR (nujol): ν = 3446 (OH), 1732 (C=O) cm–1.
˜
C17H20O4 (288.1): calcd. C 70.81, H 6.99; found C 70.65, H 6.88.
0.50 mL/min): tR = 10.6 (minor), 13.9 min (major); 98% ee; [α]D
=
–76.8 (c = 1.2, CHCl3). 1H NMR (300 MHz, CDCl3, 25 °C, TMS):
Methyl (4S*,4aR*,8aS*)-8a-Hydroxy4-phenyl-4a,5,6,7,8,8a-hexa-
hydro-4H-chromen-2-carboxylate (9a): White crystals; m.p. 119–
120 °C (hexane/cyclohexane). Chiralcel OD column (hexane/
iPrOH, 90:10; flow rate: 0.50 mL/min): tR = 9.3 min (major),
11.1 min (minor); 99% ee; [α]D = +90.3 (c = 1.1, CHCl3) [for the
opposite enantiomer with 99% ee, literature[30] reports [α]D = –9.9
3
3
δ = 7.37 (d, JH,H = 8.4 Hz, 2 H, ArH), 7.15 (d, JH,H = 8.4 Hz, 2
3
H, ArH), 6.31 (d, JH,H = 4.8 Hz, 1 H, 3-H), 3.85 (s, 3 H, OCH3),
3.44 (dd, JH,H = 8.1, JH,H = 4.8 Hz, 1 H, 4-H), 2.19 (m, 1 H),
3
3
3
3
3
2.03 (ddd, JH,H = 12.1, JH,H = 8.1, JH,H = 3.4 Hz, 1 H, 4a-H),
1.7–0.9 (m, 7 H), 0.15 (s, 9 H, SiMe3) ppm. 13C NMR (75 MHz,
CDCl3, 25 °C, TMS): δ = 163.7, 141.5, 138.3, 132.4, 130.1, 120.2,
114.4, 99.9, 52.1, 43.9, 39.8, 37.9, 27.2, 25.8, 22.7, 1.7 ppm. IR
1
(c = 0.1, CHCl3)]. H NMR (300 MHz, CDCl3, 25 °C, TMS): δ =
3
7.37–7.22 (m, 5 H, ArH), 6.20 (d, JH,H = 2.2 Hz, 1 H, 3-H), 3.81
(nujol): ν = 1732 (C=O) cm–1. C H BrO Si (439.4): calcd. C
3
3
˜
20 27
4
(s, 3 H, OCH3), 3.35 (dd, JH,H = 11.3, JH,H = 2.2 Hz, 1 H, 4-H),
54.67, H 6.19; found C 54.75, H 6.07.
3
3
2.53 (br. s, 1 H, OH), 2.51 (m, 1 H), 1.74 (dt, JH,H = 13.2, JH,H
= 11.2, JH,H = 3.7 Hz, 1 H, 4a-H), 1.7–1.1 (m, 7 H) ppm. 13C
3
Methyl (4R,4aR,8aR)-4-(p-Bromophenyl)-8a-hydroxy-4a,5,6,7,8,8a-
hexahydro-4H-chromen-2-carboxylate (8b): White crystals; m.p. 95–
96 °C (diisopropyl ether/hexane); crystal structure is reported in
Figure 3; [α]D = –70.6 (c = 0.6, CHCl3). Chiralcel OD column (hex-
ane/iPrOH, 90:10; flow rate: 0.50 mL/min): tR = 10.1 (major),
11.0 min (minor); 99% ee. 1H NMR (300 MHz, CDCl3, 25 °C,
NMR (75 MHz, CDCl3, 25 °C, TMS): δ = 163.5, 141.8, 140.3,
128.5, 128.4, 126.8, 116.4, 98.4, 52.1, 45.1, 41.4, 38.1, 26.8, 25.3,
22.9 ppm. The 1H and 13C NMR spectra are identical to those
reported in ref.[30,39] IR (nujol): ν = 3474 (OH), 1723 (C=O) cm–1.
˜
Reaction Between 1a and 2a Catalyzed by [Sc(OTf)3/(4ЈS,5ЈS)-3c]:
The reactions in Table 1, entries 7 and 8 were performed following
the general method described above. The reaction conditions, type
of separation, yields, and ee of each product are those reported the
Table 1. When 1 mol equivalent of HFIP was used (entry 7), only
a low yield of methyl (4S*,4aS*,8aS*)-4-phenyl-8a-trimethylsiloxy-
4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate (7aa) was ob-
tained, but its ee was excellent.
3
3
TMS): δ = 7.47 (d, JH,H = 8.3 Hz, 2 H, ArH), 7.11 (d, JH,H
=
4
3
8.3 Hz, 2 H, ArH), 6.26 (d, JH,H = 1.5, JH,H = 2.3 Hz, 1 H, 3-
H), 4.23 (dd, JH,H = 6.1, JH,H = 2.3 Hz, 1 H, 4-H), 3.85 (s, 3 H,
3
3
3
OCH3), 3.00 (s, 1 H, OH), 2.2 (m, 1 H), 1.88 (ddd, JH,H = 9.9,
3JH,H = 6.1, 3JH,H = 5.5 Hz, 1 H, 4a-H), 1.7–0.9 (m, 7 H) ppm. 13
C
NMR (75 MHz, CDCl3, 25 °C, TMS): δ = 163.2, 141.0, 139.7,
131.3, 130.1, 120.3, 112.1, 98.5, 52.2, 41.4, 38.7, 37.8, 24.6, 23.6,
22.9 ppm. IR (nujol):
ν =
˜
3344 (OH), 1719 (C=O) cm–1.
When the reaction was performed in the absence of HFIP (Table 1,
entry 8), in addition to products 7aa, 8a and 9a, the following prod-
ucts were obtained.
C17H19BrO4 (367.2): calcd. C 55.60, H 5.21; found C 55.83, H 5.11.
Methyl (4R,4aS,8aR)-4-(p-Bromophenyl)-8a-hydroxy-4a,5,6,7,8,8a-
hexahydro-4H-chromen-2-carboxylate (9b): Chiralcel OD column
(hexane/iPrOH, 90:10; flow rate: 0.50 mL/min–1): tR = 9.6 (minor),
10.8 min (major). This product was obtained in very low yield,
therefore its physical characteristics will be described in a following
paragraphs.
Methyl (4S*,4aS*,8aR*)-4-Phenyl-8a-trimethylsiloxy-4a,5,6,7,8,8a-
hexahydro-4H-chromen-2-carboxylate (5aa): Oily product. Chi-
ralpak AD column (hexane/iPrOH, 98:2; flow rate: 0.50 mL/min):
tR = 7.6 (major), 10.9 min (minor); 96% ee; [α]D = +86.7 (c = 0.2,
1
CHCl3). H NMR (300 MHz, CDCl3, 25 °C, TMS): δ = 7.37–7.21
4
3
(m, 5 H, ArH), 6.31 (dd, JH,H = 1.5, JH,H = 1.8 Hz, 1 H, 3-H),
From (R)-3b as pybox (Table 1, entry 12), in addition to the above
products, methyl (4R,4aR,8aS)-4-(p-bromophenyl)-8a-trimethyl-
3
3
4.18 (dd, JH,H = 5.8, JH,H = 1.8 Hz, 1 H, 4-H), 3.85 (s, 3 H,
3
3
3
OCH3), 2.28 (m, 1 H), 1.81 (dt, JH,H = 8.6, JH,H = 6.9, JH,H
=
C
siloxy-4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate
(5ba)
5.8 Hz, 1 H, 4a-H), 1.7–0.9 (m, 7 H), 0.19 (s, 9 H, SiMe3) ppm. 13
was obtained. Oily product. Chiralpak AD column (hexane/iPrOH,
98:2; flow rate: 0.50 mL/min): tR = 11.5 (minor), 12.9 min (major);
90% ee; [α]D = –79.9 (c = 0.15, CHCl3). 1H NMR (300 MHz,
NMR (75 MHz, CDCl3, 25 °C, TMS): δ = 163.3, 141.2, 141.0,
128.4, 128.1, 126.2, 112.7, 100.2, 51.9, 43.9, 38.8, 38.3, 24.6, 23.6,
23.2, 1.5 ppm. The 1H and 13C NMR spectra are identical to those
3
CDCl3, 25 °C, TMS): δ = 7.46 (d, JH,H = 8.4 Hz, 2 H, ArH), 7.10
reported in ref.[29] IR (nujol): ν = 1735 (C=O) cm–1.
3
4
3
˜
(d, JH,H = 8.4 Hz, 2 H, ArH), 6.22 (dd, JH,H = 1.5 Hz, JH,H
=
3
3
2.2 Hz, 1 H, 3-H), 4.14 (dd, JH,H = 6.0 Hz, JH,H = 2.2 Hz, 1 H,
Methyl (4S*,4aR*,8aR*)-4-Phenyl-8a-trimethylsiloxy-4a,5,6,7,8,8a-
hexahydro-4H-chromen-2-carboxylate (6aa): Small amount of oily
product. Chiralpak AD column (hexane/iPrOH, 98:2; flow rate:
0.50 mL/min): tR = 8.5 (minor), 9.6 min (major). 1H NMR
(300 MHz, CDCl3, 25 °C, TMS): δ = 7.35–7.20 (m, 5 H, ArH), 6.15
3
4-H), 3.84 (s, 3 H, OCH3), 2.30 (m, 1 H), 1.76 (dt, JH,H = 9.1 Hz,
3JH,H = 6.2 Hz, JH,H = 6.0 Hz, 4a-H), 1.6–0.9 (m, 7 H), 0.18 (s, 9
3
H, SiMe3) ppm. 13C NMR (75 MHz, CDCl3, 25 °C, TMS): δ =
163.1, 141.2, 140.3, 131.2, 130.1, 120.1, 111.8, 100.1, 52.0, 43.7,
3
(d, JH,H = 2.2 Hz, 1 H, 3-H), 3.81 (s, 3 H, OCH3), 3.20 (dd, 3JH,H
38.8, 37.9, 24.6, 23.5, 23.1, 1.5 ppm. IR (nujol): ν = 1740
˜
(C=O) cm–1. C20H27BrO4Si (439.4): calcd. C 54.67, H 6.19; found
C 54.58, H 6.12.
3
3
= 11.0, JH,H = 2.2 Hz, 1 H, 4-H), 2.16 (m, 1 H), 1.48 (dt, JH,H
=
3
3
11.3, JH,H = 11.0, JH,H = 3.6 Hz, 1 H, 4a-H), 1.7–1.1 (m, 7 H),
0.19 (s, 9 H, SiMe3) ppm. 13C NMR (75 MHz, CDCl3, 25 °C,
TMS): δ = 163.5, 142.2, 140.4, 128.5, 128.3, 126.6, 116.3, 100.1,
51.9, 46.8, 41.4, 37.5, 26.9, 25.3, 23.0, 1.3 ppm. The 1H and 13C
NMR spectra are identical to those reported in ref.[29] IR (nujol):
From (4ЈS,5ЈS)-3c as pybox (Table 1, entry 14), in addition to
(4S,4aS,8aR)-5ba and nearly racemic 7ba, methyl (4S,4aR,8aR)-4-
(p-bromophenyl)-8a-trimethylsiloxy-4a,5,6,7,8,8a-hexahydro-4H-
chromen-2-carboxylate (6ba) was obtained. Oily product. Chi-
ralpak AD column (hexane/iPrOH, 98:2; flow rate: 0.50 mL/min–1):
tR = 13 (minor), 14.5 min (major); 50% ee. 1H NMR (300 MHz,
ν = 1740 (C=O) cm–1.
˜
Reaction Between 1b and 2a Catalyzed by [Sc(OTf)3/3]: The reac-
tions in Table 1, entries 9–14 were performed following the general
method described above. For reaction conditions, type of separa-
tion, yields, and ee of each product, see the general procedure and
Table 1. The following products were obtained from (R)-3a as py-
box (Table 1, entry 10).
3
CDCl3, 25 °C, TMS): δ = 7.45 (d, JH,H = 8.4 Hz, 2 H, ArH), 7.07
3
3
(d, JH,H = 8.4 Hz, 2 H, ArH), 6.09 (d, JH,H = 2.2 Hz, 1 H, 3-H),
3
3
3.82 (s, 3 H, OCH3), 3.17 (dd, JH,H = 10.9 Hz, JH,H = 2.2 Hz, 1
H, 4-H), 2.17 (m, 1 H), 1.48 (dt, 3JH,H = 11.3 Hz, 3JH,H = 10.9 Hz,
3JH,H = 3.6 Hz, 1 H, 4a-H), 1.7–1.1 (m, 7 H), 0.18 (s, 9 H,
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
9

Job/Unit: O20056
/KAP1
Date: 05-04-12 16:12:07
Pages: 14
G. Desimoni et al.
FULL PAPER
SiMe3) ppm. 13C NMR (75 MHz, CDCl3, 25 °C, TMS): δ = 163.3,
Methyl (4S*,4aS*,8aR*)-4-(p-Bromophenyl)-8a-tert-butyldimethyl-
siloxy-4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate (5bb): A
very small amount of oily product. 1H NMR (300 MHz, CDCl3,
25 °C, TMS): δ = 7.45 (d, 3JH,H = 8.4 Hz, 2 H, ArH), 7.07 (d, 3JH,H
141.3, 140.7, 131.4, 130.2, 120.4, 115.4, 100.0, 52.0, 46.8, 40.9, 37.4,
26.8, 25.3, 23.0, 1.5 ppm. IR (nujol): ν = 1738 (C=O) cm–1.
˜
C20H27BrO4Si (439.4): calcd. C 54.67, H 6.19; found C 54.81, H
6.27.
4
3
= 8.4 Hz, 2 H, ArH), 6.21 (dd, JH,H = 1.5 Hz, JH,H = 2.2 Hz, 1
3
3
H, 3-H), 4.17 (dd, JH,H = 5.9 Hz, JH,H = 2.2 Hz, 1 H, 4-H), 3.84
Reaction Between 1a and 2b Catalyzed by [Sc(OTf)3/3]: The reac-
tions in Table 2, entries 1–4 were performed by following the gene-
ral method described above. For reaction conditions, type of sepa-
ration, yields, and ee of each product see the general procedure and
Table 2. The following products were obtained from (S)-3a as py-
box (Table 2, entry 1).
3
3
(s, 3 H, OCH3), 2.27 (m, 1 H), 1.78 (dt, JH,H = 9.2 Hz, JH,H
=
3
6.1 Hz, JH,H = 5.9 Hz, 4a-H), 1.7–0.9 (m, 7 H), 0.89 (s, 9 H,
SitBu), 0.22 (s, 3 H, SiMe), 0.10 (s, 3 H, SiMe) ppm. 13C NMR
(75 MHz, CDCl3, 25 °C, TMS): δ = 163.1, 141.4, 140.3, 131.3,
130.1, 120.1, 111.6, 100.0, 51.9, 44.0, 38.7, 38.3, 25.6, 24.6, 23.6,
23.1, 17.9, –2.7, –3.8 ppm.
Methyl
(4S*,4aR*,8aS*)-8a-tert-Butyldimethylsiloxy-4-phenyl-
Methyl (4S*,4aR*,8aR*)-4-(p-Bromophenyl)-8a-tert-butyldimethyl-
siloxy-4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate (6bb): A
very small amount of oily product. Chiralpak AD column (hexane/
iPrOH, 98:2; flow rate: 0.50 mL/min): tR = 8.5 (major), 11.5 min
(minor); 32% ee. 1H NMR (300 MHz, CDCl3, 25 °C, TMS): δ =
4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate (4ab): Oily
product. Chiralpak AD column (hexane/iPrOH, 98:2; flow rate:
0.50 mL/min): tR = 7.5 (major), 8.1 min (minor); 76% ee. 1H NMR
(300 MHz, CDCl3, 25 °C, TMS): δ = 7.35–7.20 (m, 5 H, ArH), 6.08
3
(d, JH,H = 2.8 Hz, 1 H, 3-H), 3.82 (s, 3 H, OCH3), 3.52 (dd, 3JH,H
3
3
7.44 (d, JH,H = 8.4 Hz, 2 H, ArH), 7.06 (d, JH,H = 8.4 Hz, 2 H,
= 9.0 Hz, 3JH,H = 2.8 Hz, 1 H, 4-H), 1.94 (dt, 3JH,H = 9.0 Hz, 3JH,H
ArH), 6.08 (d, 3JH,H = 2.2 Hz, 1 H, 3-H), 3.82 (s, 3 H, OCH3), 3.21
3
= 8.4 Hz, JH,H = 4.4 Hz, 1 H, 4a-H), 1.8–1.3 (m, 8 H), 0.85 (s, 9
3
3
(dd, JH,H = 11.0 Hz, JH,H = 2.2 Hz, 1 H, 4-H), 2.16 (m, 1 H),
H, SitBu), 0.26 (s, 3 H, SiMe), 0.13 (s, 3 H, SiMe) ppm. 13C NMR
(75 MHz, CDCl3, 25 °C, TMS): δ = 163.2, 142.8, 141.2, 128.4,
128.0, 126.5, 112.4, 101.8, 51.9, 45.5, 40.6, 33.9, 25.8, 25.6, 23.3,
3
3
3
1.48 (dt, JH,H = 11.4 Hz, JH,H = 11.0 Hz, JH,H = 3.5 Hz, 1 H,
4a-H), 1.6–1.3 (m, 7 H), 0.94 (s, 9 H, SitBu), 0.25 (s, 3 H, SiMe),
0.07 (s, 3 H, SiMe) ppm. 13C NMR (75 MHz, CDCl3, 25 °C, TMS):
δ = 163.3, 141.2, 140.9, 131.4, 130.3, 120.5, 115.3, 99.9, 52.0, 47.2,
41.0, 37.4, 26.8, 25.9, 25.3, 22.9, 18.3, –2.9, –4.0 ppm.
C23H33BrO4Si (481.5): calcd. C 57.37, H 6.91; found C 57.15, H
6.84.
1
20.6, 17.9, –2.6, –3.1 ppm. The H NMR spectrum is identical to
that reported in ref.[29] IR (nujol):
ν =
˜
1735 (C=O) cm–1.
C23H34O4Si (402.6): calcd. C 68.62, H 8.51; found C 68.74, H 8.69.
Methyl (4S*,4aR*,8aR*)-8a-tert-Butyldimethylsiloxy-4-phenyl-
4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate (6ab): Oily
product. Chiralpak AD column (hexane/iPrOH, 98:2; flow rate:
0.50 mL/min): tR = 8.6 (minor), 10.9 min (major); 45% ee. 1H
NMR (300 MHz, CDCl3, 25 °C, TMS): δ = 7.36–7.17 (m, 5 H,
ArH), 6.15 (d, 3JH,H = 2.2 Hz 1 H, 3-H), 3.81 (s, 3 H, OCH3), 3.24
Methyl (4S,4aS,8aS)-4-(p-Bromophenyl)-8a-tert-butyldimethylsil-
oxy-4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate (7bb): Oily
product. Chiralpak AD column (hexane/iPrOH, 98:2; flow rate:
0.50 mL/min): tR = 10.7 (major), 16.0 min (minor); 98% ee; [α]D
=
+56.1 (c = 0.9, CHCl3). 1H NMR (300 MHz, CDCl3, 25 °C, TMS):
3
3
(dd, JH,H = 11.1 Hz, JH,H = 2.2 Hz, 1 H, 4-H), 2.17 (m, 1 H),
3
3
δ = 7.36 (d, JH,H = 8.4 Hz, 2 H, ArH), 7.18 (d, JH,H = 8.4 Hz, 2
3
3
3
1.55 (dt, JH,H = 11.3 Hz, JH,H = 11.1 Hz, JH,H = 3.4 Hz, 1 H,
4a-H), 1.7–1.1 (m, 7 H), 0.95 (s, 9 H, SitBu), 0.25 (s, 3 H, SiMe),
0.08 (s, 3 H, SiMe) ppm. 13C NMR (75 MHz, CDCl3, 25 °C, TMS):
δ = 163.4, 142.1, 140.6, 128.6, 128.3, 126.7, 116.2, 100.0, 52.0, 47.1,
3
H, ArH), 6.31 (d, JH,H = 4.8 Hz, 1 H, 3-H), 3.85 (s, 3 H, OCH3),
3.47 (dd, 3JH,H = 8.3 Hz, 3JH,H = 4.8 Hz, 1 H, 4-H), 2.18 (m, 1 H),
3
2.11 (ddd, 3JH,H = 12.1 Hz, 3JH,H = 8.4 Hz, JH,H = 3.2 Hz, 4a-H),
1.8–1.0 (m, 7 H), 0.79 (s, 9 H, SitBu), 0.29 (s, 3 H, SiMe), 0.10 (s,
3 H, SiMe) ppm. 13C NMR (75 MHz, CDCl3, 25 °C, TMS): δ =
163.5, 141.6, 138.4, 132.1, 130.5, 120.2, 114.2, 100.0, 52.1, 44.6,
39.6, 38.5, 27.3, 26.4, 26.0, 22.6, 18.3, –2.3, –2.6 ppm. IR (nujol):
1
41.5, 37.5, 26.9, 25.9, 25.4, 22.9, 18.3, –2.9, –4.0 ppm. The H and
13C NMR spectra are identical to those reported in ref.[29] IR (nu-
jol): ν = 1727 (C=O) cm–1.
˜
ν = 1724 (C=O) cm–1. C H BrO Si (481.5): calcd. C 57.37, H
Methyl
(4S*,4aS*,8aS*)-8a-tert-Butyldimethylsiloxy-4-phenyl-
˜
23 33
4
4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate (7ab): Oily
product. Chiralpak AD column (hexane/iPrOH, 98:2; flow rate:
6.91; found C 57.58, H 7.05.
In addition to these products (4S,4aS,8aS)-8b and (4S,4aR,8aS)-9b
were obtained.
0.50 mL/min): tR = 9.5 (major), 10.8 min (minor); 99% ee; [α]D
=
+50.0 (c = 0.5, CHCl3). 1H NMR (300 MHz, CDCl3, 25 °C, TMS):
3
δ = 7.41–7.21 (m, 5 H, ArH), 6.38 (d, JH,H = 4.8 Hz, 1 H, 3-H),
Reaction Between 1b and 2c Catalyzed by [Sc(OTf)3/(4ЈS)-3a]: The
3.85 (s, 3 H, OCH3), 3.53 (dd, 3JH,H = 8.4 Hz, 3JH,H = 4.8 Hz, 1 H, reaction in Table 2, entry 9 was performed by following the general
4-H), 2.14 (ddd, 3JH,H = 12.5 Hz, 3JH,H = 8.4 Hz, 3JH,H = 3.2 Hz, 1
method described above. In addition to unreacted 1b (55%), the
H, 4a-H), 1.8–1.0 (m, 8 H), 0.78 (s, 9 H, SitBu), 0.29 (s, 3 H, SiMe), reaction gave (4S,4aS,8aS)-8b (21% yield) with a trace of 9b.
0.13 (s, 3 H, SiMe) ppm. 13C NMR (75 MHz, CDCl3, 25 °C, TMS):
Desilylation of Enantiomerically Pure 7ba with Tetrabutyammonium
δ = 163.6, 141.4, 139.4, 130.4, 127.4, 126.2, 115.2, 100.0, 52.0, 44.8,
Fluoride: A sample of enantiomerically enriched 7ba (44 mg,
40.1, 38.6, 27.3, 26.4, 25.8, 22.6, 18.3, –2.3, –2.9 ppm. IR (nujol):
0.1 mmol, 99% ee; Table 1, entry 9 [(4ЈS)-3a as legand]), dissolved
ν = 1743 (C=O) cm–1. C H O Si (402.6): calcd. C 68.62, H 8.51;
found C 68.81, H 8.39.
˜
23 34
4
in THF (5.0 mL), was cooled to 0 °C and TBAF (1 mol) solution
in THF (120 μL, 0.12 mmol) was added whilst stirring. After
10 min, all starting material had disappeared, the reaction was con-
tinued for an additional 15 min at 0 °C and for 15 min at ambient
temperature. Water was added and the mixture was extracted with
CH2Cl2 (3ϫ15 mL). After drying with Na2SO4, the solution was
evaporated and the residue was purified by chromatography on a
silica gel column (60 cm length; cyclohexane/ethyl acetate, 75:25).
In addition to these products (4S*,4aS*,8aS*)-8a and
(4S*,4aR*,8aS*)-9a were obtained.
Reaction Between 1b and 2b Catalyzed by [Sc(OTf)3/3]: The reac-
tions in Table 2, entries 5–8 were performed by following the gene-
ral method described above. For reaction conditions, type of sepa-
ration, yields, and ee of each product, see the general procedure
and Table 2. The following products were obtained from (S)-3b as
pybox (Table 2, entry 7).
Enantiomerically pure (4S,4aR,8aS)-9b (7.0 mg, 19% yield) eluted
first. White crystals; m.p. 138 °C (diisopropyl ether/hexane); its
10
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O20056
/KAP1
Date: 05-04-12 16:12:07
Pages: 14
[Sc(OTf)3/pybox]-Catalyzed Hetero-Diels–Alder Reaction
crystal structure is reported in Figure 4; [α]D = +117.3 (c = 0.4,
CHCl3) {For the opposite enantiomer with 93% ee: literature[30]
reports [α]D = –7.9 (c = 0.2, CHCl3)}. Chiralcel OD column (hex-
Methyl (4R*,4aS*,8aR*)-8a-Hydroxy-4-phenyl-4a,5,6,7,8,8a-hexa-
hydro-4H-chromen-2-carboxylate (9a): Yield: 25%. Chiralcel OD
column (hexane/iPrOH, 90:10; flow rate: 0.50 mL/min): tR = 9.9
ane/iPrOH, 90:10; flow rate: 0.50 mL/min): tR = 9.5 (major), (minor), 12.3 min (major); 98% ee; [α]D = –95.2 (c = 0.15, CHCl3).
10.6 min (minor); 98% ee. 1H NMR (300 MHz, CDCl3, 25 °C,
Methyl (4R*)-4-Phenyl-5,6,7,8-tetrahydro-4H-chromen-2-carboxyl-
3
3
TMS): δ = 7.46 (d, JH,H = 8.3 Hz, 2 H, ArH), 7.08 (d, JH,H
=
ate (10a): Yield: 51%. Chiralcel AD column (hexane/iPrOH, 98:2;
flow rate: 0.50 mL/min): tR = 18.3 (major), 23.9 min (minor);
98% ee; [α]D = –71.6 (c = 0.3, CHCl3). 1H NMR (300 MHz,
8.3 Hz, 2 H, ArH), 6.13 (d, 3JH,H = 2.1 Hz 1 H, 3-H), 3.81 (s, 3 H,
3
OCH3), 3.32 (dd, 3JH,H = 11.2 Hz, JH,H = 2.1 Hz, 1 H, 4-H), 2.57
3
3
(br. s, 1 H, OH), 2.1 (m, 1 H), 1.59 (dt, JH,H = 11.6 Hz, JH,H
=
3
CDCl3, 25 °C, TMS): δ = 7.37–7.23 (m, 5 H, ArH), 6.07 (d, JH,H
11.2 Hz, JH,H = 3.8 Hz, 1 H, 4a-H), 1.8–1.2 (m, 7 H) ppm. 13C
NMR (75 MHz, CDCl3, 25 °C, TMS): δ = 163.5, 140.9, 140.5,
131.5, 130.1, 120.6, 115.5, 98.3, 52.2, 45.1, 40.9, 38.0, 26.7, 25.2,
22.9 ppm. The 1H and 13C NMR spectra are identical to those
3
3
= 4.4 Hz, 1 H, 3-H), 3.92 (d, JH,H = 4.4 Hz, 1 H, 4-H), 3.81 (s, 3
H, OCH3), 2.28 (m, 2 H), 1.8–1.5 (m, 6 H) ppm. 13C NMR
(75 MHz, CDCl3, 25 °C, TMS): δ = 162.5, 144.6, 143.7, 139.8,
128.5, 128.1, 126.8, 113.6, 107.0, 52.1, 42.7, 26.4, 26.3, 22.6,
reported in ref.[30] IR (nujol): ν = 3462 (OH), 1724 (C=O) cm–1.
˜
22.4 ppm. IR (nujol): ν = 1734 (C=O) cm–1. C H O (270.3):
˜
17 18
3
Enantiomerically pure (4S,4aS,8aS)-8b (8.5 mg, 24% yield) eluted
second. Chiralcel OD column (hexane/iPrOH, 90:10; flow rate:
0.50 mL/min): tR = 10.2 (major), 11.1 min (minor); 99% ee. The
spectroscopic data are reported above.
calcd. C 75.53, H 6.71; found C 75.37, H 6.85.
The same reaction was performed with enantiomer 7ab (from the
reaction detailed in Table 2 entry 3; 99% ee) and the enantiomers
of the same products were obtained.
Isomerization and Desilylation of Enantiomerically Enriched Ad-
ducts with Sc(OTf)3; General Procedure: A sample of enantiomer-
ically enriched 7 (ca. 0.1 mmol) was dissolved in CH2Cl2 (ca. 5 mL)
and, whilst stirring at ambient temperature, Sc(OTf)3 (ca. 10 mg,
0.025 mmol) and 3 Å MS (when required, ca. 30 mg) was added
and stirring was continued for the time reported in Schemes 5, 7
and 8, when the starting product had disappeared and the transfor-
mations were complete. The reaction mixture was quenched with a
small amount of silica gel and the products were separated by col-
umn chromatography on silica gel.
Reaction of (4R,4aR,8aR)-7ba with Sc(OTf)3: A solution of 7ba
(35 mg, 99% ee) from the reaction detailed in Table 1, entry 12, in
CH2Cl2 (3 mL) was treated with Sc(OTf)3 (8 mg) and water (5 μL)
whilst stirring at ambient temperature. After 1 h the reaction was
unchanged and MS (in different portions for a total amount of ca.
40 mg) were added. Slow development of different products was
noted and the starting material disappeared within approximately
18 h. Extensive separation by column chromatography on silica gel
gave the following products.
Methyl
(4R,4aR,8aS)-4-(p-Bromophenyl)-8a-trimethylsiloxy-
Reaction of (4R*,4aR*,8aS*)-7ab with Sc(OTf)3: A sample of 7ab
(99% ee) from the reaction detailed in Table 2, entry 4, after 1 h
with Sc(OTf)3 and MS at ambient temperature gave the following
products.
4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate (5ba): Yield:
10%; oily product. Chiralpak AD column (hexane/iPrOH, 98:2;
flow rate: 0.50 mL/min): tR = 11.7 (minor), 12.9 min (major);
99% ee. The 1H NMR spectrum is identical to that reported above.
Methyl
(4R*,4aR*,8aS*)-8a-tert-Butyldimethylsiloxy-4-phenyl-
Methyl
(4R,4aS,8aS)-4-(p-Bromophenyl)-8a-trimethylsiloxy-
4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate (5ab): Yield:
22%; oily product; chiralpak AD column (hexane/iPrOH, 98:2;
flow rate: 0.50 mL/min): tR = 8.1 (major), 8.4 min (minor); 99% ee;
[α]D = –94.1 (c = 0.26, CHCl3). 1H NMR (300 MHz, CDCl3, 25 °C,
TMS): δ = 7.37–7.20 (m, 5 H, AH), 6.30 (dd, 4JH,H = 1.5 Hz, 3JH,H
4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate (6ba): Yield:
20%; oily product. Chiralpak AD column (hexane/iPrOH, 98:2;
flow rate: 0.50 mL/min): tR = 11.7 (major), 13.4 min (minor);
97% ee; [α]D = –55.0 (c = 0.3, CHCl3). The H NMR spectrum is
identical to that reported above.
1
3
3
= 2.3 Hz, 1 H, 3-H), 4.23 (dd, JH,H = 5.9 Hz, JH,H = 2.3 Hz, 1
3
H, 4-H), 3.84 (s, 3 H, OCH3), 2.27 (m, 1 H), 1.82 (dt, JH,H
=
Methyl (4R,4aR,8aR)-4-(p-Bromophenyl)-8a-hydroxy-4a,5,6,7,8,8a-
hexahydro-4H-chromen-2-carboxylate (8b): Yield: 19%; Chiralcel
OD column (hexane/iPrOH, 90:10; flow rate: 0.50 mL/min): tR
10.1 (minor), 11.0 min (major); 98% ee.
3
3
8.5 Hz, JH,H = 6.1 Hz, JH,H = 5.9 Hz, 1 H, 4a-H), 1.7–1.0 (m, 7
H), 0.91 (s, 9 H, SitBu), 0.24 (s, 3 H, SiMe), 0.12 (s, 3 H,
SiMe) ppm. 13C NMR (75 MHz, CDCl3, 25 °C, TMS): δ = 163.2,
141.2, 128.4, 128.3, 128.2, 126.3, 112.6, 100.0, 51.9, 44.1, 38.8, 38.7,
25.6, 24.6, 23.7, 23.2, 17.9, –2.6, –3.8 ppm; The 1H NMR spectrum
=
When the same reaction was performed without water and with
MS, the same products were obtained with 3, 9 and 67% yields,
respectively.
is identical to that previously reported.[29] IR (nujol): ν = 1742
˜
(C=O) cm–1. C23H34O4Si (402.6): calcd. C 68.62, H 8.51; found C
68.39, H 8.40.
Reaction of (4R,4aR,8aR)-7bb with Sc(OTf)3: A sample of 7bb
(99% ee) from the reaction detailed in Table 2, entry 8 after 3 h
at ambient temperature with Sc(OTf)3 and MS gave the following
products.
Methyl
(4R*,4aS*,8aS*)-8a-tert-Butyldimethylsiloxy-4-phenyl-
4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate (6ab): Yield:
61%; oily product; chiralpak AD column (hexane/iPrOH, 98:2;
flow rate: 0.50 mL/min): tR = 8.2 (minor), 10.5 min (major);
Methyl (4R,4aR,8aS)-4-(p-Bromophenyl)-8a-tert-butyldimethylsil-
oxy-4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate
Yield: 30%; oily product. Chiralpak AD column (hexane/iPrOH,
98:2; flow rate: 0.50 mL/min): tR = 8.2 (minor), 12.8 min (major);
98% ee; [α]D = –90.0 (c = 0.2, CHCl3). The H NMR spectrum is
1
Ͼ99% ee; [α]D = –139.5 (c = 0.69, CHCl3). The 13C and H NMR
(5bb):
spectra are identical to those reported above.
The same reaction from 7ab after 18 h without MS gave the follow-
ing products.
1
identical to that reported above.
Methyl (4R*,4aR*,8aR*)-8a-Hydroxy-4-phenyl-4a,5,6,7,8,8a-hexa-
hydro-4H-chromen-2-carboxylate (8a): Yield: 16%. Chiralcel OD
column (hexane/iPrOH, 90:10; flow rate: 0.50 mL/min): tR = 10.0
(major), 10.8 min (minor); 99% ee; [α]D = –51.7 (c = 1.0, CHCl3).
Methyl (4R,4aS,8aS)-4-(p-Bromophenyl)-8a-tert-butyldimethylsil-
oxy-4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate
Yield: 55%; oily product. Chiralpak AD column (hexane/iPrOH,
98:2; flow rate: 0.50 mL/min): tR = 8.6 (minor), 13.7 min (major);
(6bb):
1
The H NMR spectrum is identical to that reported above.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
11

Job/Unit: O20056
/KAP1
Date: 05-04-12 16:12:07
Pages: 14
G. Desimoni et al.
FULL PAPER
98% ee; [α]D = –55.0 (c = 0.3, CHCl3). The H NMR spectrum is
1
44% yield) as an oil. Chiralcel AD column (hexane/iPrOH, 98:2;
flow rate: 1.00 mL/min): tR = 12.1 (minor), 16.9 min (major);
Ͼ99% ee; [α]D = +69.4 (c = 0.8, CHCl3). 1H NMR (300 MHz,
CDCl3, 25 °C, TMS): δ = 7.46 (d, JH,H = 8.4 Hz, 2 H, ArH), 7.12
(d, JH,H = 8.4 Hz, 2 H, ArH), 6.02 (d, JH,H = 4.4 Hz, 1 H, 3-H),
identical to that reported above.
Isomerization and Desilylation of Enantiomerically Enriched Ad-
ducts with Trifluoroacetic Acid; General Procedure: A sample of
enantiomerically enriched 4 (or 7) (ca. 0.05 mmol) was dissolved in
CH2Cl2 (ca. 3 mL) and, whilst stirring at the temperature reported
in Scheme 4, Scheme 5, Scheme 6 or Scheme 7, CF3CO2H (ca.
100 μL) was added. The reaction times (when the starting product
disappeared and the transformations were complete) are reported
in the above Schemes. The reaction mixture was quenched with a
small amount of silica gel and the products were separated by col-
umn chromatography on silica gel.
3
3
3
3
3.89 (d, JH,H = 4.4 Hz, 1 H, 4-H), 3.82 (s, 3 H, OCH3), 2.26 (m,
2 H), 1.8–1.5 (m, 6 H) ppm. 13C NMR (75 MHz, CDCl3, 25 °C,
TMS): δ = 162.4, 145.0, 142.7, 140.0, 131.6, 129.8, 120.7, 112.8,
106.7, 52.2, 42.2, 26.4, 26.3, 22.6, 22.4 ppm. IR (nujol): ν = 1732
˜
(C=O) cm–1. C17H17BrO3 (349.2): calcd. C 58.47, H 4.91; found C
58.64, H 5.03.
The same (4S)-10b was quantitatively obtained from (4S,4aS,8aS)-
8b after treatment with CF3CO2H at ambient temperature for 4 h.
Reaction of (4S*,4aS*,8aR*)-5aa with CF3CO2H: A sample of 5aa
(96% ee) from the reaction detailed in Table 1, entry 8, after 15 min
at ambient temperature with CF3CO2H, gave the following prod-
ucts.
The same reaction, performed on (4R,4aR,8aR)-7ba with
CF3CO2H, gave the opposite enantiomers of the same products.
Reaction of (4R,4aR,8aS)-7bb with CF3CO2H: A sample of 7bb
(99% ee) from the reaction detailed in Table 2, entry 6, after 10 h
at –20 °C with CF3CO2H gave methyl (4R,4aR,8aS)-4-(p-bromo-
phenyl)-8a-tert-butyldimethylsiloxy-4a,5,6,7,8,8a-hexahydro-4H-
chromen-2-carboxylate (5bb; 76% yield) as an oily product. Chi-
ralpak AD column (hexane/iPrOH, 98:2; flow rate: 0.50 mL/min):
tR = 8.2 (minor), 12.8 min (major); 98% ee. The 1H NMR spectrum
is identical to that reported above.
Methyl (4S*,4aS*,8aS*)-8a-Hydroxy-4-phenyl-4a,5,6,7,8,8a-hexa-
hydro-4H-chromen-2-carboxylate (8a): Yield: 16%. Chiralcel OD
column (hexane/iPrOH, 90:10; flow rate: 0.50 mL/min): tR = 9.5
(minor), 11.0 min (major); 96% ee. The 1H NMR spectrum is iden-
tical to that reported above.
Methyl (4S*,4aR*,8aS*)-8a-Hydroxy-4-phenyl-4a,5,6,7,8,8a-hexa-
hydro-4H-chromen-2-carboxylate (9a): Yield: 25%. Chiralcel OD
column (hexane/iPrOH, 90:10; flow rate: 0.50 mL/min): tR = 9.3
(major), 11.1 min (minor); 94% ee. The 1H NMR spectrum is iden-
tical to that reported above.
When the same reaction was performed at 0 °C for 4 h, a complex
mixture was obtained, which was separated to give, besides unre-
acted 7bb (12%), the following products.
Methyl (4S*)-4-Phenyl-5,6,7,8-tetrahydro-4H-chromen-2-carboxyl-
ate (10a): Yield: 15%. Chiralcel AD column (hexane/iPrOH, 98:2;
flow rate: 0.50 mL/min): tR = 19 (minor), 23 min (major); 94% ee.
Methyl (4R,4aS,8aS)-4-(p-Bromophenyl)-8a-tert-butyldimethylsil-
oxy-4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate
(6bb):
Yield: 8%. Chiralpak AD column (hexane/iPrOH, 98:2; flow rate:
1
The H NMR spectrum is identical to that reported above.
1
0.50 mL/min): tR = 8.5 (minor), 13.6 min (major); 97% ee. The H
When the same reaction was performed at ambient temperature for
90 min, 9a and 10a were obtained with 25 and 15% yields, respec-
tively. Under the same experimental conditions, 8a gave the above
enantiomer of 10a in 85% yield, and 9a did not react.
NMR spectrum is identical to that reported above.
(4R,4aR,8aR)-8b (30% yield, 99% ee) identical to the sample re-
ported above.
(4R,4aS,8aR)-9b (15% yield, 99% ee) identical to the sample re-
ported above.
Reaction of (4R*,4aR*,8aR*)-7ab with CF3CO2H: A sample of 7ab
(99% ee) from the reaction detailed in Table 2, entry 4, after 10 min
at –20 °C with CF3CO2H, gave the following products.
X-ray Diffraction Studies of 8b and 9b: X-ray diffraction data of a
hemihydrate (4R,4aR,8aR)-8b crystal were collected with an Enraf–
Nonius CAD4 diffractometer, those of (4S,4aR,8aS)-9b were col-
lected with a Bruker-Axs CCD-based diffractometer. Both data
collections were performed at ambient temperature using graphite-
monochromatized Mo-Kα radiation (λ = 0.7107 Å). Both crystal
structures were solved by direct methods (SIR97)[40] and refined
with full-matrix least-square procedures on F2 (SHELXL-97)[41]
using all reflections. Lorentz, polarisation and absorption effect
corrections (psi-scan method[42] for (4R,4aR,8aR)-8b crystal and
semi-empirical method[43] for (4S,4aR,8aS)-9b crystal) were ap-
plied. Hydrogen atoms of organic moieties were placed at the calcu-
lated positions; positions of hydrogen atoms of water molecules
were identified in the final ΔF map. The absolute configuration was
established by anomalous dispersion effects in the diffraction data.
Methyl
(4R*,4aR*,8aS*)-8a-tert-Butyldimethylsiloxy-4-phenyl-
4a,5,6,7,8,8a-hexahydro-4H-chromen-2-carboxylate (5ab): Yield:
26%. Chiralpak AD column (hexane/iPrOH, 98:2; flow rate:
0.50 mL/min): tR = 8.1 (major), 8.4 min (minor); 99% ee. 1H NMR
spectrum is identical to that reported above.
Methyl (4R*,4aR*,8aR*)-8a-Hydroxy-4-phenyl-4a,5,6,7,8,8a-hexa-
hydro-4H-chromen-2-carboxylate (8a): Yield: 43%. Chiralcel OD
column (hexane/iPrOH, 90:10; flow rate: 0.50 mL/min): tR = 10.0
(major), 10.8 min (minor); 99% ee. The 1H NMR spectrum is iden-
tical to that reported above.
Methyl (4R*)-4-Phenyl-5,6,7,8-tetrahydro-4H-chromen-2-carboxyl-
ate (10a): Yield: 15%. Chiralcel AD column (hexane/iPrOH, 98:2;
flow rate: 0.50 mL/min): tR = 19.6 (major), 23.2 min (minor);
96% ee. The 1H NMR spectrum is identical to that reported above.
Crystal Data for Hemihydrate (4R,4aR,8aR)-8b: C17H20BrO4.5; Mr
= 376.23; monoclinic C2 (No. 5); a = 23.089(4), b = 5.559(2), c
Reaction of (4S,4aS,8aS)-7ba with CF3CO2H: A sample of 7ba
(99% ee) from the reaction detailed in Table 1, entry 9, after 10 min
at –20 °C with CF3CO2H, quantitatively gave (4S,4aS,8aS)-8b
(99% ee), which was identical in every respect to the sample re-
ported above.
= 13.083(3) Å, β = 93.09(2)°; V = 1676.9(7) Å3; Z = 4; σcalcd.
=
1.490 gcm–3; μ(Mo-Kα) = 2.470 mm–1; 2θmax = 50°; min./max.
transmission 0.512/0.609, 3900 reflections collected, 2979 indepen-
dent reflections (Rint = 0.014), 2107 strong reflections [FOϾ2σ(FO)],
218 refined parameters, R1/wR2 = 0.0364/0.0534 (strong reflec-
tions) and 0.0699/0.0607 (all reflections), goodness of fit = 1.038;
max/min residuals 0.32/–0.29 e–3, Flack’s parameter –0.01(1).
The same reaction, when performed at 0 °C for 15 min, gave the
above reported (4S,4aS,8aS)-8b (41% yield) and methyl (4S)-4-(p-
bromophenyl)-5,6,7,8-tetrahydro-4H-chromen-2-carboxylate (10b;
12
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O20056
/KAP1
Date: 05-04-12 16:12:07
Pages: 14
[Sc(OTf)3/pybox]-Catalyzed Hetero-Diels–Alder Reaction
[15] G. Faita, M. Mella, M. Toscanini, G. Desimoni, Tetrahedron
2010, 66, 3024–3029.
[16] J. Lv, X. Li, L. Zhong, S. Luo, J.-P. Cheng, Org. Lett. 2010, 12,
1096–1099.
[17] Y. Liu, D. Shang, X. Xhou, Y. Zhu, L. Lin, X. Liu, X. Feng,
Org. Lett. 2010, 12, 180–183.
[18] a) G. Desimoni, G. Faita, K. A. Jørgensen, Chem. Rev. 2006,
106, 3561–3651; b) G. Desimoni, G. Faita, K. A. Jørgensen,
Update 1: Chem. Rev. 2011, 111, PR284–PR437.
[19] G. Desimoni, G. Faita, P. Quadrelli, Chem. Rev. 2003, 103,
3119–3154.
Crystal Data for (4S,4aR,8aS)-9b: C17H19BrO4; Mr = 367.22; or-
thorhombic P212121 (No. 19); a = 7.762(1), b = 9.875(1), c =
21.092(1) Å; V = 1616.7(2) Å3; Z = 4; σcalcd. = 1.509 gcm–3; μ(Mo-
Kα) = 2.558 mm–1; 2θmax = 50°; min./max. transmission 0.682/
0.920, 8598 reflections collected, 2832 independent reflections (Rint
= 0.059), 1699 strong reflections [FOϾ2σ(FO)], 200 refined param-
eters, R1/wR2 = 0.0466/0.0890 (strong reflections) and 0.0963/
0.1080 (all reflections), goodness of fit = 0.955; max/min residuals
0.35/–0.27 e–3, Flack’s parameter –0.04(2).
[20] D. A. Evans, J. S. Johnson, E. J. Olhava, J. Am. Chem. Soc.
2000, 122, 1635–1649.
[21] A. Moyano, R. Rios, Chem. Rev. 2011, 111, 4703–4832.
[22] J. Zhou, Y. Tang, Org. Biomol. Chem. 2004, 2, 429–433.
[23] G. Desimoni, G. Faita, M. Mella, F. Piccinini, M. Toscanini,
Eur. J. Org. Chem. 2007, 1529–1534.
CCDC-847016 (for 8b) and 847017 (for 9b) contain the supplemen-
tary crystallographic data for this paper. These data can be ob-
tained free of charge from The Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
Supporting Information (see footnote on the first page of this arti-
cle): HPLC chromatograms, 1H NMR and 13C NMR spectra of all
products.
[24] G. Desimoni, G. Faita, M. Toscanini, M. Boiocchi, Chem. Eur.
J. 2007, 13, 9478–9485.
[25] Y. Zhu, X. Chen, M. Xie, S. Dong, Z. Qiao, L. Lin, X. Liu,
X. Feng, Chem. Eur. J. 2010, 16, 11963–11968.
[26] J. Thorhauge, M. Johannsen, K. A. Jørgensen, Angew. Chem.
1998, 110, 2543; Angew. Chem. Int. Ed. 1998, 37, 2404–2406.
[27] H. Audrain, J. Thorhauge, R. G. Hazel, K. A. Jørgensen, J.
Org. Chem. 2000, 65, 4487–4497.
Acknowledgments
This work was supported by the Ministero dell’Università e della
Ricerca (MiUR) (PRIN 2008) and by the University of Pavia.
[28] H. Audrain, K. A. Jørgensen, J. Am. Chem. Soc. 2000, 122,
11543–11544.
[29] A. Martel, S. Leconte, G. Dujardin, E. Brown, V. Maison-
neuve, R. Retoux, Eur. J. Org. Chem. 2002, 514–525; erratum:
Eur. J. Org. Chem. 2012, to be published (ejoc 201200157).
[30] Z. Xu, L. Liu, K. Wheeler, H. Wang, Angew. Chem. Int. Ed.
2011, 50, 3484–3488.
[1] M. W. Fennie, E. F. DiMauro, E. M. O’Brien, V. Annamalai,
M. Kozlowski, Tetrahedron 2005, 61, 6249–6265.
[2] Q. Meng, L. Zhu, Z. Zhang, J. Org. Chem. 2008, 73, 7209–
7212.
[31] D. A. Evans, C. E. Masse, J. Wu, Org. Lett. 2002, 4, 3375–3378.
[3] A. Yanagisawa, Y. Terajima, K. Sugita, K. Yoshida, Adv. [32] W. H. Kim, J. H. Lee, S. J. Danishefsky, J. Am. Chem. Soc.
Synth. Catal. 2009, 351, 1757–1762.
[4] L. Zhou, L. Lin, W. Wang, J. Ji, X. Liu, X. Feng, Chem. Com-
mun. 2010, 46, 3601–3603.
2009, 131, 12576–12578.
[33] E. D. Secher, H. F. Ryder, J. Am. Chem. Soc. 1952, 74, 4392–
4395.
[5] Z. Dong, J. Feng, W. Cao, X. Liu, L. Lin, X. Feng, Tetrahedron
Lett. 2011, 52, 3433–3436.
[34] M. Reimer, E. Tobin, J. Am. Chem. Soc. 1940, 62, 2515–2520.
[35] F. C. E. Sarabèr, S. Dratch, G. Bosselaar, B. J. M. Jansen, A.
de Groot, Tetrahedron 2006, 62, 1717–1725.
[36] P. Magnus, J. Lacour, I. Coldham, B. Murgage, W. B. Bauta,
Tetrahedron 1995, 41, 11087–11110.
[6] H. L. van Lingen, W. Zhuang, T. Hansen, F. P. J. T. Rutjes,
K. A. Jørgensen, Org. Biomol. Chem. 2003, 1, 1953–1958.
[7] Y. Hui, J. Jiang, W. Wang, W. Chen, Y. Cai, L. Lin, X. Liu, X.
Feng, Angew. Chem. Int. Ed. 2010, 49, 4290–4293.
[8] C. Christensen, K. Juhl, R. G. Hazell, K. A. Jørgensen, J. Org.
Chem. 2002, 67, 4875–4881.
[37] J.-Q. Yu, H.-C. Wu, E. J. Corey, Org. Lett. 2005, 7, 1415–1417.
[38] G. Desimoni, G. Faita, M. Guala, A. Laurenti, M. Mella,
Chem. Eur. J. 2005, 11, 3816–3824.
[9] N. Halland, T. Velgaard, K. A. Jørgensen, J. Org. Chem. 2003,
[39] C.-L. Cao, X.-L. Sun, Y.-B. Kang, Y. Tang, Org. Lett. 2007, 9,
68, 5067–5074.
4151–4154.
[10] D. T. Cohen, B. Cardinal-David, K. A. Scheidt, Angew. Chem.
Int. Ed. 2011, 50, 1678–1682.
[11] K. B. Jensen, J. Thorhauge, R. G. Hazel, K. A. Jørgensen, An-
gew. Chem. 2001, 113, 164; Angew. Chem. Int. Ed. 2001, 40,
160–163.
[12] K. A. Jørgensen, Synthesis 2003, 1117–1125.
[13] M. P. A. Lyle, N. D. Draper, P. D. Wilson, Org. Lett. 2005, 7,
901–904.
[14] G. Desimoni, G. Faita, M. Toscanini, M. Boiocchi, Chem. Eur.
J. 2008, 14, 3630–3636.
[40] A. Altomare, M. C. Burla, M. Camalli, G. L. Cascarano, C.
Giacovazzo, A. Guagliardi, A. G. G. Moliterni, G. Polidori, R.
Spagna, J. Appl. Crystallogr. 1999, 32, 115–119.
[41] G. M. Sheldrick, Acta Crystallogr., Sect. A 2008, 64, 112–122.
[42] A. C. T. North, D. C. Phillips, F. S. Mathews, Acta Crystallogr.,
Sect. A 1968, 24, 351–359.
[43] G. M. Sheldrick, SADABS Siemens Area Detector Absorption
Correction Program, University of Göttingen, Germany, 1996.
Received: January 16, 2012
Published Online: ᭿
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
13

Job/Unit: O20056
/KAP1
Date: 05-04-12 16:12:07
Pages: 14
G. Desimoni et al.
FULL PAPER
Asymmetric Catalysis
G. Desimoni,* G. Faita, A. Livieri,
M. Mella, L. Ponta, M. Boiocchi ... 1–14
The Asymmetric Formal Hetero-Diels–
Alder Reaction of Methyl (E)-4-Aryl-2-
oxo-3-butenoates Catalyzed by [Sc(OTf)3/
pybox] Complexes
Why does [Sc(OTf)3/(S)-pybox] catalyze the
enantioselective formal hetero-Diels–Alder
reaction between enol silyl ethers and
methyl 4-aryl-2-oxo-3-butenoates to give
bicyclic products that do not retain the di-
enophile configuration? Does the reaction
follow Danishefsky’s trans-Diels–Alder
paradigm or Mukaiyama–Michael ad-
dition followed by intramolecular ring clos-
ure? Here, we address these questions.
Keywords: Asymmetric catalysis / Cycli-
zation / Michael addition / N ligands
14
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0