Stereoselective allylations of erythrulose derivatives under anhydrous conditions

Article abstract of DOI:10.1016/S0957-4166(01)00262-2

We have investigated a number of nucleophilic additions of allylating reagents to several α,α′,β-trioxygenated ketones (?-protected erythrulose derivatives). Reagents based on lithium, magnesium, copper and titanium gave low to medium stereoselectivities and did not display any recognizable trend in the sense of stereoselection. In contrast, reactions involving silicon and tin derivatives were highly stereoselective and gave rise to essentially a single diastereoisomer, the structure of which depended on the type of protecting group. Thus, α,β-di-?-benzylated derivatives experienced almost exclusive addition to the carbonyl Si side, whereas α,β-?,?-alkylidene derivatives (dioxolane acetals) yielded the opposite diastereoisomers as a result of addition to the Re side. These results suggest the intermediacy of α-chelates in the additions of silicon and tin reagents to the di-?-benzylated derivatives. In contrast, the opposite stereoisomers, formed in the reactions of dioxolanes, are believed to be formed through Felkin - Anh transition states, pointing again to the reluctance of acetal oxygens to participate in chelated intermediates.

Full text of DOI:10.1016/S0957-4166(01)00262-2

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 12 (2001) 1417–1429
Stereoselective allylations of erythrulose derivatives under
anhydrous conditions
Miguel Carda,a,* Encarna Castillo,b Santiago Rodr´ıguez,a Florenci Gonza´leza and
J. Alberto Marcoc,*
aDepartamento de Qu´ımica Inorga´nica y Orga´nica, Univ. Jaume I, E-12080 Castello´n, Spain
bDep. de Farm. y Tecn. Farmac., Fac. de C. Exp. y de la Salud, Univ. Cardenal Herrera-CEU, E-46115 Moncada,
Valencia, Spain
cDepartamento de Qu´ımica Orga´nica, Univ. de Valencia, E-46100 Burjassot, Valencia, Spain
Received 9 April 2001; accepted 7 June 2001
Abstract—We have investigated a number of nucleophilic additions of allylating reagents to several a,a%,b-trioxygenated ketones
(O-protected erythrulose derivatives). Reagents based on lithium, magnesium, copper and titanium gave low to medium
stereoselectivities and did not display any recognizable trend in the sense of stereoselection. In contrast, reactions involving silicon
and tin derivatives were highly stereoselective and gave rise to essentially a single diastereoisomer, the structure of which depended
on the type of protecting group. Thus, a,b-di-O-benzylated derivatives experienced almost exclusive addition to the carbonyl Si
side, whereas a,b-O,O-alkylidene derivatives (dioxolane acetals) yielded the opposite diastereoisomers as a result of addition to the
Re side. These results suggest the intermediacy of a-chelates in the additions of silicon and tin reagents to the di-O-benzylated
derivatives. In contrast, the opposite stereoisomers, formed in the reactions of dioxolanes, are believed to be formed through
Felkin–Anh transition states, pointing again to the reluctance of acetal oxygens to participate in chelated intermediates. © 2001
Published by Elsevier Science Ltd.
1. Introduction
these reactions proved strongly dependent on the type
of hydroxyl protecting group. In fact, addition pro-
cesses involving Grignard reagents and some specifi-
Carbohydrates, particularly monosaccharides, represent
one of the most convenient chiral sources in the synthe-
cally protected ketones
1
were particularly
sis of enantiopure compounds.1 The ketotetrose
L-(S)-
diastereoselective as well as extremely fast. This was
attributed to the intermediacy of reactive five-mem-
bered a-chelates involving the metal, the carbonyl oxy-
gen and the a-oxygen atoms, a conclusion supported by
quantum-mechanical calculations.7 In contrast, no con-
clusive evidence could be gathered to substantiate the
intermediacy of six-membered b-chelates in these addi-
tions, in spite of the demonstrated existence of such
chelates.8 Since the stereoisomer predicted in this case
was the same as that predicted by the Felkin–Anh
model, this was chosen as the preferred explanation of
the stereochemical outcome in those cases where the
intermediacy of an a-chelate was excluded.
erythrulose and its derivatives2 of general formula 1
(R1, R2, R3=protecting groups) are useful additions to
the list of chiral precursors of this type. The carbonyl
group of 1 is a prominent site for the appendage of
additional carbon fragments through nucleophilic addi-
tion. The selection of suitable protecting groups is
important, as they exert a control on the steric course
of nucleophilic additions.3 The literature contains a
great deal of studies on stereoselective additions of
C-nucleophiles to polyoxygenated aldehydes.4 How-
ever, the corresponding behavior of highly functional-
ized ketones such as
1
(Scheme 1) is less
documented.2d,5 We have recently described the stereo-
chemical outcome of many organometallic additions to
the carbonyl groups of several protected erythrulose
derivatives of the type 1.6 The diastereoselectivity of
It is worth mentioning here that, among the many
organometallic additions to derivatives 1 investigated,
those of allylating reagents gave unsatisfactory stereo-
selectivities in many cases.6a Still, nucleophilic addition
of allyl appendages to aldehydes, ketones or their imino
derivatives often constitutes a convenient feature in
* Corresponding authors. Fax: +34-96-3864328; e-mail: mcarda@
qio.uji.es; alberto.marco@uv.es
0957-4166/01/$ - see front matter © 2001 Published by Elsevier Science Ltd.
PII: S0957-4166(01)00262-2

1418
M. Carda et al. / Tetrahedron: Asymmetry 12 (2001) 1417–1429
synthetic designs; the CꢀC bond of the allyl moiety may
be later manipulated in various ways, for example, via
oxidative cleavage to a functionalized fragment bearing
alcohol or carbonyl groups (equivalent of aldol addi-
tion), via intramolecular ring-closing metathesis (RCM)
to heterocyclic systems,9 etc. (Scheme 1). Such strategies
have proven useful in our syntheses of some naturally
occurring molecules.10 We have thus investigated in
detail the influence of the allylation conditions (het-
eroelement and protecting group type, addition of
Lewis acids, etc.) in the stereoselectivity of the reaction.
As in our previous publication,6a we wished to deter-
mine the extent to which the stereoselectivity of nucle-
ophilic additions would be controlled by either
Cram-type chelation mechanisms involving the various
oxygen atoms or by transition states of the Felkin–Anh
type.3,13 In the case of Lewis acidic allyl heteroatomic
derivatives, the transfer of the allyl moiety takes place
through a cyclic, six-membered transition state of the
metallo-ene type (the concomitant inversion of the allyl
residue is not relevant in the case of the unsubstituted
allyl moiety). As depicted in Scheme 3, the five-mem-
bered a-chelate is predicted to react from the less
hindered Si side of the carbonyl group (TS-1), leading
predominantly to stereoisomers 8 or 10. The alternative
stereoisomers 9 or 11, derived from attack on the Re
side, are expected to be the major products when the
non-chelated Felkin–Anh transition state TS-2 is oper-
ating (on the basis of our previous experiences with
protected erythruloses,6a we have discarded here the
intermediacy of b-chelates).
The substrates selected for this purpose were L-erythru-
lose derivatives 2–6 (Scheme 1), the preparation of
which has recently been reported.6a,11 These ketones
were subjected to a range of nucleophilic allylating
reagents 712 under various reaction conditions. Our
purpose was to find reaction conditions suitable for the
preparation of allylation products of either configura-
tion in the newly generated stereogenic center (Scheme
2). Herein, we describe these results in detail.
(Y = O, NR)
+
Y
M
O
4
2
3
carbonyl or
imino derivative
allylating
reagent
1
1
R1O
OR2 OR3
O
[O]
O
O
YH
YH
Aldol or β-amino
O
BnO
O
OR
OBn OR
carbonyl compound
2
3
4
R = H
R = TPS
R = Tr
5
6
R = TPS
R = Tr
RCM
Y
Y
Z
Z
Heterocyclic
derivative
(Z = atom chain)
Scheme 1.
O
HO
O
HO
M
7
+
O
O
O
(M = heteroatom)
O
OR
OR
8a-8c
O
OR
9a-9c
a, R = TPS
b, R = Tr
c, R = H
2-4
(for 8-11)
HO
HO
O
M
7
+
BnO
BnO
BnO
(M = heteroatom)
OBn OR
OBn OR
10a-10b
OBn OR
5-6
11a-11b
Scheme 2. Addition of nucleophilic allylating reagents 7 to erythruloses 2–6.

M. Carda et al. / Tetrahedron: Asymmetry 12 (2001) 1417–1429
1419
hindered approach
to the Re face
(Ln = further
Ln
R1O
H
metal ligands)
M
OR1
OR3
O
Si approach
R2
O
R2O
8/10 (major)
9/11 (major)
O
M
Ln
H
Re approach
CH2OR3
TS-1 (α-chelate)
TS-2 (Felkin-Anh)
Scheme 3. Transition states during addition of Lewis acidic allyl metal reagents to erythrulose derivatives.
2. Results and discussion
latter result was particularly unexpected as non-allyl
Grignard reagents had previously given very high
diastereoisomeric ratios in their reactions with 5 and
6,6a most likely via TS-1 (Scheme 3, M=Mg). This
behavior does not extend to allylmagnesium bromide,
perhaps because the non-chelated TS-2 has a decreased
energy barrier and becomes thus more competitive in
this case.
As in the compounds described in our previous work,
the oxygen atom of the primary hydroxyl group at C(1)
in the erythrulose derivatives (e.g. 3–6) was blocked
using various protecting groups. We later found, how-
ever, that some allylating reagents could actually be
used in the presence of the free OH group at C(1). For
this reason, acetonide 211 was also included in the list.
In relation to one of our research projects, we also
investigated the aldol-type reactions of 3 and 5 with
equivalents of acetate enolate. Even though an acetate
appendage can also be obtained through oxidative
cleavage of the double bond of an allyl moiety (Scheme
1), a stereoselective aldol reaction of this type would
constitute a shorter and thus more convenient route.
Enolates of t-butyl acetate were first selected for this
purpose (Scheme 4). After investigating the influence of
a number of factors (solvent, temperature, presence of
additives), the best conditions in terms of yield and
stereoselectivity were those indicated in the last three
entries of Table 1. Additives such as HMPA or DMPU
were not beneficial as they markedly reduced the reac-
tion rate without changing the diastereoisomeric ratio
to any relevant extent. Transmetalation of lithium to
magnesium, zinc or titanium enolates was similarly not
useful, either, as this led mainly to recovery of the
starting material at −78°C or to decomposition at
higher temperatures. Equally unsuccessful was the use
The additions of Li and Mg allyl reagents to ketones 3
and 5 were briefly mentioned in our previous report.6a
A more detailed account is presented in Table 1. These
allyl metal reagents did not give very high
diastereomeric ratios in their reactions with ketones
3–6. The erratic behavior of the lithium reagents
toward erythrulose derivatives, either of the acetonide
or of the dibenzyl type, was already noted. It seems that
the lithium atom is not particularly prone to participate
in a-chelates in the reactions under study; in fact, the
Felkin–Anh stereoisomers 9 and 11 were the major
products formed, albeit with low diastereoisomeric
ratios (entries 1–4). More surprising, however, were the
similarly low diastereoisomeric ratios found with Grig-
nard reagents. Furthermore, while acetonides 3 and 4
displayed low diastereoisomeric ratios with predomi-
nance of the Felkin–Anh stereoisomer 9, the benzylated
derivatives 5 and 6 gave a slightly higher percentage of
Cram’s a-chelation stereoisomer 10 (entries 5–8). The
Table 1. Stereoselectivity in the additions of allyl lithium, allylmagnesium bromide and acetic acid enolate equivalents to
erythrulose derivatives 3–6a
Entry
Ketone
Reagent
Solvent
% Yield
8:9 or 10:11b
1
2
3
4
5
6
7
8
3
4
5
6
3
4
5
6
3
5
5
AllylLic
THF/Et2O
THF/Et2O
THF/Et2O
THF/Et2O
Et2O
Et2O
Et2O
Et2O
Et2O
85
88
87
90
90
85
86
95
78
95
70
22:78
1:3
3:7
2:3
1:4
AllylLic
AllylLic
AllylLic
AllylMgBr
AllylMgBr
AllylMgBr
AllylMgBr
LiCH2CO2tBu
LiCH2CO2tBu
LiCH2CN
32:68
67:33
3:2
9
10
11
1:9d
1:4d
24:76d
Et2O
THF
a In all cases, 2.5 equiv. of the nucleophile were added at −78°C to the appropriate ketone in the indicated solvent.
b Determined by means of 1H and 13C NMR.
c Prepared according to Ref. 25.
d Ratio 12:13, see Scheme 4 and comments in the text.

1420
M. Carda et al. / Tetrahedron: Asymmetry 12 (2001) 1417–1429
HO
HO
O
G
G
MCH2G
a
b
R,R = CMe2
R = Bn
+
RO
(M = metal,
G = CN or
CO2tBu)
RO
RO
OR OTPS
OR OTPS
OR OTPS
(for 12 and 13)
3 or 5
12a-12b
13a-13b
Scheme 4. Addition of acetic acid enolate equivalents to erythruloses 3 and 5.
acids4g,18 (Table 3). Reactions with allyl trimethylsilane
were unsuccessful with ketones 2–4: strong Lewis acids
such as TiCl4 or SnCl4 caused the decomposition of the
acetonide moiety, whereas less strong Lewis acids such
as the etherates of BF3 or MgBr2 were unable to
promote the reaction. With TiCl2(O-iso-Pr)2, the only
result observed was electrophilic attack at the acetonide
moiety with concomitant ring opening.19 Basic catalysis
met also with failure (entry 1): partial decomposition
was the only result in the reaction of 2 with allyl
trimethylsilane in the presence of tetra-n-butylammo-
nium fluoride.20 In contrast with these failures, ketone 5
displayed an excellent stereoselectivity in its reaction
with allyl trimethylsilane in the presence of SnCl4 (TiCl4
was similarly effective) as the minor isomer 11a was not
detected by high-field NMR (entry 2). The stereoisomer
which was formed, 10a, was that predicted by Cram’s
cyclic chelation model. It is thus likely that the biden-
tate Lewis acid SnCl4 first forms a five-membered
chelate with ketone 5 (Scheme 5), and the chelate is
then intermolecularly attacked by the allyl silane. The
closely related ketone 6 was expected to behave in the
same way, but its somewhat labile protecting group did
not survive the Lewis acidic reaction conditions (entry
3).
of the dianion of acetic acid. The lithium anion of
acetonitrile gave results similar to those observed with
the anion of tert-butyl acetate. In all cases where a
reaction was observed, the Felkin–Anh stereoisomers
13a or 13b were the major products. However, the
observed diastereoisomeric ratios were not high.
We then turned our attention to other allyl metal
derivatives. Among these, titanium, copper and zinc
seemed to be the metals of choice for our purposes.
Four different types of titanium reagents14 were assayed
on ketones 3 and 5 (Table 2). Good yields were
obtained in all cases but the degrees of diastereoselec-
tivity were too variable (with diastereoisomeric ratios of
between 1:1 and 12:88 and the Felkin–Anh isomers
predominating). The allyl copper reagent15 displayed a
somewhat better degree of diastereoselection, particu-
larly in the case of acetonide 3. Here again, the Felkin–
Anh isomer was the major compound formed. This
contrasts with the reactions of the same ketones with
non-allylic cuprates, in which the major products were
the isomers predicted from chelation control.6a Finally,
both ketones proved unreactive toward allyl zinc
derivatives prepared either by insertion of zinc into the
C−Br bond of allyl bromide,16a,b or by transmetalation
of allylmagnesium bromide with zinc chloride in
THF.16c
The more reactive allyl tri-n-butylstannane was able to
allylate either type of ketone under mild conditions.
Even with the mild Lewis acid MgBr2·Et2O as the
promoter, clean and highly stereoselective reactions
were observed at −40°C. In the case of 5 and 6, single
The best results in terms of stereoselectivity were even-
tually obtained with allyl trimethylsilane12,17 or allyl
tri-n-butylstannane12 in the presence of suitable Lewis
Table 2. Stereoselectivity in the additions of allyl copper, titanium and zinc reagents to erythrulose derivatives 3 and 5
Entry
Ketone
Reagent
Solvent
T (°C)
% Yield
8:9 or 10:11a
b
1
2
3
4
5
6
7
8
3
3
3
3
3
3
5
5
5
5
5
5
AllylTiCl3
Toluene/Et2O
Et2O
THF
THF
THF
THF
Toluene/Et2O
Et2O
THF
THF
rt
81
73
79
84
92
3:7
c
AllylTiCl(OiPr)2
AllylTi(OiPr)3
−40“rt
rt
−10
27:73
36:64
32:68
6:94
–
12:88
27:73
ꢀ1:1
27:73
19:81
–
d
−e
AllylTi(OiPr)4
AllylCuLif
−78“−40
AllylZnBrg
–
–
g
g
b
AllylTiCl3
rt
79
91
89
88
94
c
AllylTiCl(OiPr)2
AllylTi(OiPr)3
−40“rt
rt
−10
d
9
−e
10
11
12
AllylTi(OiPr)4
AllylCuLif
AllylZnBrg
THF
THF
−78“−40
g
g
–
–
a Determined by means of 1H and 13C NMR.
b Prepared according to a modification of the procedure described in Ref. 14c.
c Prepared according to Ref. 14d.
d Prepared according to Ref. 14e.
e Prepared according to Ref. 14b.
f Prepared according to Ref. 15b.
g Prepared under various conditions according to Refs. 16a–c, but no reaction took place.

M. Carda et al. / Tetrahedron: Asymmetry 12 (2001) 1417–1429
1421
Table 3. Stereoselective additions of allyl silicon and tin reagents to erythrulose derivatives 2–6a
Entry
Ketone
Reagent
T (°C)/t (h)
% Yield
8:9 or 10:11b
1
2
3
4
5
6
7
8
9
2
5
6
2
3
3
4
5
6
AllylSiMe3/TBAFc
AllylSiMe3/SnCl4
AllylSiMe3/SnCl4
−78“−10/5
−78“−40/6
−78“−40/6
−78“−40/4
−78“−40/4
25/72
−78“−40/4
−78“−40/4
−78“−40/4
Dec.
80
–
60
80
86
85
80
85
–
\19:1
–
B1:19
B1:19
3:7
B1:19
\19:1
\19:1
d
e
AllylSnBu3/MgBr2
AllylSnBu3/MgBr2
e
f
AllylSnBu3/LiClO4
e
AllylSnBu3/MgBr2
AllylSnBu3/MgBr2
AllylSnBu3/MgBr2
e
e
a In all cases except entry 1, 1.2 equiv. of the organometallic reagent was added to a mixture of the appropriate ketone and the Lewis acid (1.2
equiv.) in dry CH2Cl2.
b Determined by means of 1H and 13C NMR (d.r.>19:1 or <1:19 means that NMR signals from the minor isomer were not visible).
c TBAF and allylSiMe3 (1.4 equiv. each) were added in dry THF to ketone 2 at the indicated temperature.
d Cleavage of the trityl group took place.
e The monodentate Lewis acid BF3·Et2O not only gave inferior yields and d.r.s, but also caused cleavage of the trityl protecting group in ketones
4 and 6. Strong Lewis acids (TiCl4, SnCl4) caused decomposition of the acetonide moiety in ketones 2–4.
f Solvent was either Et2O or CH2Cl2.
L
L
O
OBn
Bn
O
O
O
Si approach
SnBu3
10 (major)
9 (major)
R3M
O
Re approach
H
H
(M = Si or Sn)
CH2OP
CH2OP
L = Lewis acid
P = protecting group
α-Chelated TS
(ketones 5-6)
Felkin-Anh TS
(ketones 2-4)
Scheme 5. Transition states in Lewis acid promoted addition of allyl silanes/stannanes to erythrulose derivatives.
(Scheme 6). Compound 9a, obtained from 3 and allyl
tri-n-butylstannane (Table 3, entry 5) was converted in
two steps into lactone 15 via b-hydroxy ester 14. The
NMR data of 15 were similar, although not identical,
to those of lactone 19, obtained as described below.
The known compound 16,6a prepared by reaction of
ketone 5 with vinyl magnesium bromide, was trans-
formed in four steps into lactone 18 via cyanohydrin
17. The absolute configuration of 18 was confirmed by
means of X-ray diffraction analysis.21 Lactone 18 was
also obtained from allylation product 10a, a fact which
establishes the configuration of the new stereogenic
center formed in the allylations of 5 (Table 3, entries 2
and 8). Two simple reactions on 18 transformed it into
lactone 19, not identical with 15, as commented above.
Finally, the configurations of the other products of the
allylation of ketones 2–4 and 5–6 were established
through cleavage of the protecting group to yield in all
cases the same diols, 9c and 10c, respectively. The
configurations of the products formed in the aldol
reaction of 5 with the carbanion of acetonitrile (Scheme
4) were established by comparison of their NMR data
with those of 17, which turned out to be identical with
the minor component of the mixture. Likewise, the
mixture of stereoisomers formed in the aldol reactions
with acetic ester enolates were converted by reaction
with TBAF into a mixture of the stereoisomeric lac-
tones 15 (major) and 19 (minor) in the case of 3, and
lactone 18 (minor) and its stereoisomer 21 (major) in
the case of 5.
stereoisomers were obtained (10a and 10b, respectively)
with the configuration corresponding to that predicted
by Cram’s cyclic chelation model (entries 8 and 9). Here
again, the mechanistic model of Scheme 5 may be
applied. It is worth mentioning that the monodentate
Lewis acid BF3·Et2O, which cannot form chelates, gave
poorer results in terms of both yield and
stereoselectivity.
On the basis of these findings, we were surprised by the
results observed in the reactions of the allylstannane
with acetonides 2–4 (entries 4, 5 and 7). Ketones 3 and
4, which have a very bulky protecting group at the OH
at C(1), gave rise to only one diastereoisomer in each
case, 9a and 9b, which are those predicted by the
Felkin–Anh model (Scheme 5). The same result was
observed when the Lewis acid MgBr2·Et2O was
replaced by TiCl2(O-iso-Pr)2.19 Interestingly, ketone 2,
bearing an unprotected C(1) hydroxyl group, gave rise
to the Felkin–Anh diastereoisomer 9c. In an attempt to
invert the stereochemical bias of the acetonide-pro-
tected erythruloses, we also used LiClO4 as the Lewis
acid18a–c in either Et2O or CH2Cl2 as the solvent.
Unfortunately, the reactions were poorly selective and
very slow even at room temperature.
The configurations of the diastereoisomers formed in
the aforementioned reactions have been established
with the aid of unambiguous chemical correlations

1422
M. Carda et al. / Tetrahedron: Asymmetry 12 (2001) 1417–1429
HO
HO
O
HO
O
CO2Me
O3, NaOH, MeOH
deprot.
O
(for 9a)
O
O
O
OH
OR
OTPS
9c
9a R=TPS
9b R=Tr
14
TBAF/THF
O
O
O
O
HO
O
HO
HO
O
O
13b
13a
O
O
BnO
O
OBn
19
21
15
1: H2, Pd(OH)2
2: Me2C(OMe)2, cat. H+
1: O3, NaBH4
2: MsCl, NEt3
3: NaCN, DMSO
O
O
HO
HO
HO
CN
HCl, MeOH, ∆
BnO
BnO
BnO
BnO
OBn OTPS
OBn OTPS
OBn
18
16
17
TBAF/THF
HO
HO
HO
deprot.
CO2Me
O3, NaOH, MeOH
BnO
BnO
(for 10a)
OBn OH
OBn OR
OBn OTPS
10c
10a R=TPS
10b R=Tr
20
Scheme 6. Chemical correlations between allylation products.
one seems more likely to us in view of existing experi-
mental evidence.23 It is also worth mentioning that even
ketone 2 with an unprotected OH at C(1) reacts very
stereoselectively under these conditions. This reinforces
our view of the absence of chelation with acetonides, as
it has been shown that a,a%-oxygenated ketones often
display low diastereoselectivities in cases where chela-
tion at both sides of the carbonyl group is possible.24
Consideration of the overall results in Tables 1–3 leads
to some general conclusions. The chelation isomers
(Cram) 8/10 are formed with a very high diastereoiso-
meric ratio with allyl silane or stannane and the O-ben-
zylated ketones 5 and 6 in the presence of a bidentate
Lewis acid.4g,18 It is thus likely that a-chelates are
formed in these cases and there is a great deal of
experimental evidence3,13 for their intermediacy in car-
bonyl additions. In contrast, the Felkin–Anh isomers
9/11 were formed with a very high diastereoisomeric
ratio only in acetonides 2–4, where the intermediacy of
an a-chelate would not have been otherwise excluded in
principle. These results should be now viewed in light of
previous observations on nucleophilic additions to 2,3-
O-isopropylidene glyceraldehyde and its 2,3-di-O-ben-
zyl analogue.22 In all of the cases examined, the
acetonide was clearly the less diastereoselective com-
pound, with the major stereoisomer being the opposite
of that predicted by the a-chelation model. In order to
explain this, it has been proposed22a that glyceraldehyde
acetonide reacts mainly under non-chelation control
because of: (a) a too high chelate energy content due to
appreciable ring strain and to non-bonded interactions
of the acetonide geminal methyl groups with metal
ligands (cf. TS-1 in Scheme 3), and/or (b) a depressed
donor ability of the acetonide a-oxygen atom owing to
the mutual electron-withdrawing inductive effect by the
other oxygen atom. Of these explanations, the second
3. Conclusions
In summary, we have been able to perform highly
stereoselective nucleophilic allylations of a,a%,b-trioxy-
genated ketones of the type described above. Either of
the two possible diastereoisomeric tertiary alcohols
were obtained by means of a simple change in the type
of protecting group. When the a-oxygen atom bears
substituents (e.g. benzyl) which do not prevent the
formation of Cram-type five-membered chelates, nucle-
ophilic attack will take place from the less hindered face
of the carbonyl group, leading in most cases to very
high stereoselectivities. If these conditions are not met
(e.g. insufficient Lewis basicity of the a-oxygen atom on
steric and/or electronic grounds, absence or insufficient
acidity of bidentate Lewis acids, including the het-
eroatom in the allylating reagent), the Felkin–Anh
stereoisomer will normally predominate albeit to an
extremely variable and non-predictable extent.

M. Carda et al. / Tetrahedron: Asymmetry 12 (2001) 1417–1429
1423
4. Experimental
4.1. General methods
15–20 min at room temp., the solution turned dark red.
The mixture was then re-cooled to −78°C and allowed
to stand. An aliquot of this solution (3.5 ml) was
carefully taken with a syringe and added dropwise to a
solution of the appropriate ketone (0.5 mmol) in THF
(5 ml) at the same temperature. The reaction mixture
was then stirred for 1 h. Work-up (CH2Cl2) and column
chromatography (hexane/EtOAc mixtures) yielded the
desired allylation product.
1H NMR spectra (400 or 500 MHz) and 13C NMR
spectra (100 or 125 MHz) were measured at 22°C. The
signals of the deuterated solvent (CDCl31) were taken as
the reference (the singlet at l7.25 for H NMR data
and the triplet centered at 77.00 ppm for 13C NMR
data). Unambiguous assignments of H and 13C NMR
1
signals were made with a combination of spin decou-
pling, DEPT and HMQC experiments. Mass spectra
were run in a VG AutoSpec mass spectrometer using
either the electron impact mode (EIMS, 70 eV), the
chemical ionization mode (CIMS, with CH4 as the
carrier gas) or the fast atom bombardment mode
(FABMS, m-nitrobenzyl alcohol matrix). Samples for
IR spectral measurements were prepared as oily films
on NaCl plates (oils) or KBr pellets (solids). Optical
rotations were measured at 22°C. The diastereoisomeric
4.4. Addition of allyl magnesium bromide
A solution of the appropriate ketone (1 mmol) in Et2O
(10 ml) was cooled to −78°C. A solution of allyl
magnesium bromide (1 M commercial solution in Et2O,
2.5 ml) was then added dropwise at the same temp.,
and the reaction mixture was stirred for 1 h. Work-up
(Et2O) and column chromatography (hexane/EtOAc
mixtures) yielded the desired allylation product.
1
composition was determined by means of high-field H
and 13C NMR of the crude reaction mixture just prior
to chromatography.
4.5. Addition of allyl titanium reagents14
Column chromatography (CC) was performed on silica
gel Su¨d-Chemie AG (60–200 mm). Experiments which
required an inert atmosphere were carried out under
dry argon (Ar) in a flame-dried glassware. THF and
Et2O were freshly distilled from sodium/benzophenone
ketyl and were transferred via syringe. Toluene was
freshly distilled from sodium wire. Methylene chloride
4.5.1. AllylTiCl3. A solution of titanium tetrachloride
in toluene (1 M, 2.7 ml, 2.7 mmol) was added dropwise
at 0°C to a solution of allyl magnesium chloride in
Et2O (1 M, 13.5 ml, 13.5 mmol). The solution was then
diluted with dry toluene (30 ml) and stirred, wherein it
turned dark blue. An aliquot of this solution (0.6 ml)
was added dropwise at 0°C to a solution of the appro-
priate ketone (0.2 mmol) in dry toluene (2 ml). After
removal of the cooling bath, the mixture was stirred
overnight at room temp. Work-up (Et2O) and column
chromatography (hexane/EtOAc mixtures) yielded the
desired allylation product, together with 70% of recov-
ered starting material. Chemical yields are based on
recovered material.
,
was distilled from P2O5 and stored over 4 A molecular
sieves. Commercially available reagents (Aldrich or
Fluka) were used as received. If not detailed otherwise,
the work-up of the reactions was consistently per-
formed in the following manner: the reaction mixture
was poured into brine and extracted twice with solvent
(Et2O, CH2Cl2 or EtOAc), the organic layer was
washed with diluted aq. NH4Cl and then washed again
with brine, the organic layer was dried over anhydrous
MgSO4 or Na2SO4, and the solvent was eliminated with
a rotary evaporator at aspirator pressure. The starting
ketones 2–6 were prepared according to described
procedures.6a,11
4.5.2. AllylTiCl(O-iso-Pr)2. Titanium tetra-iso-propox-
ide (4.43 ml, 15 mmol) and allyl bromide (1.3 ml, 15
mmol) were dissolved in Et2O (10 ml), cooled to −40°C
and treated at this temp. with a solution of iso-propyl
magnesium chloride in Et2O (2 M, 15 ml, 30 mmol).
After stirring at −40°C for 1 h, a solution of the
appropriate ketone (1 mmol) in Et2O was added. The
stirring was continued for 30 min at −40°C and then
overnight at room temp. Work-up (Et2O) and column
chromatography (hexane/EtOAc mixtures) yielded the
desired allylation product.
4.2. General experimental procedures for allylations of
ketones 2–6
All synthetic operations were performed under careful
exclusion of oxygen and moisture. All solvents used in
synthetic operations were dried according to the afore-
mentioned procedures. Yields and diastereoisomeric
compositions are indicated in Tables 1–3.
4.5.3. AllylTi(O-iso-Pr)3. A solution of chlorotitanium
tri-iso-propoxide in hexanes (1 M, 10 ml, 10 mmol) was
treated at −78°C with allyl magnesium bromide (1 M in
THF, 10 ml, 10 mmol). After stirring for 30 min at the
same temp., the appropriate ketone (1 mmol) dissolved
in THF (10 ml) was added dropwise. After stirring for
30 min, the cooling bath was removed and the mixture
was stirred overnight at room temp. Work-up (Et2O)
and column chromatography (hexane/EtOAc mixtures)
afforded the desired allylation product.
4.3. Addition of allyl lithium25
Lithium wire (125 mg, 18 mmol) was cut and washed
with hexane and THF. Further THF (2 ml) was added
followed by rapid stirring and cooling to −15°C. A
solution of allyl phenyl ether (206 ml, 1.5 mmol) in Et2O
(2 ml) was then added via syringe. Stirring was then
continued for 15 min at the same temp., wherein a pale
bluish-green color appeared. After further stirring for

1424
M. Carda et al. / Tetrahedron: Asymmetry 12 (2001) 1417–1429
−
4.5.4. AllylTi(O-iso-Pr)4 . A solution of allyl magne-
sium chloride in THF (1 M, 18 ml, 18 mmol) was
treated at −78°C with titanium tetra-iso-propoxide (5.9
ml, 20 mmol). The formation of the orange ‘ate’ com-
plex was complete after stirring for 30 min at the same
temp. The appropriate ketone (1 mmol) dissolved in
THF (10 ml) was then added dropwise, the temperature
was increased to −10°C and the mixture was stirred for
5 h at this temperature. Work-up (Et2O) and column
chromatography (hexane/EtOAc mixtures) afforded the
desired allylation product.
With TiCl2(OiPr)2 as the Lewis acid, the reaction was
conducted for 5 h at −10°C. With LiClO4, the reaction
was conducted for 72 h at room temperature (in this
case, Et2O and CH2Cl2 gave essentially the same result).
4.9. Addition of acetonitrile and tert-butyl acetate
enolates
This was completed as previously described.6a
4.10. Desilylation of 9a and 10a to 9c and 10c,
respectively
4.6. Addition of allyl copper reagent15
A solution of 9a (440 mg, 1 mmol) in dry THF (5 ml)
was treated with solid tetra-n-butylammonium fluoride
hydrate (288 mg, 1.1 mmol) and stirred under Ar at
room temp. for 30 min. After adding water (1 ml), the
volatiles were totally eliminated in vacuo. CC of the
residue (hexane/EtOAc 9:1, then 7:3) provided 9c (170
mg, 84%). The same procedure was used to convert 10a
to 10c (90%).
A mixture of CuI (400 mg, 2.10 mmol) and anhydrous
LiCl (89 mg, 2.10 mmol) was gently heated under Ar.
After cooling at room temp., THF (2 ml) was added.
The mixture was stirred for 5 min and cooled to −78°C.
A solution of allyl lithium, freshly prepared by reaction
of MeLi (1.6 M in hexanes, 1.25 ml, 2 mmol) with allyl
tri-n-butyltin (0.62 ml, 2 mmol) at −78°C, was added to
the CuCl/LiCl solution at the same temp. The mixture
was then stirred for 30 min, wherein it turned dark
brown. The appropriate ketone (0.7 mmol) dissolved in
THF (2 ml) was then added dropwise, and the temper-
ature was slowly increased to −40°C. The reaction
mixture was stirred overnight at this temperature.
Work-up (Et2O) and column chromatography (hexane/
EtOAc mixtures) yielded the desired allylation product.
4.11. Detritylation of 9b and 10b to 9c and 10c,
respectively
Method A. A 1.8 M solution of trifluoroacetic acid/
trifluoroacetic anhydride was prepared by dissolving
these reagents in the appropriate amount of dry
CH2Cl2.26 Compound 9b (178 mg, 0.4 mmol) was then
dissolved in dry CH2Cl2 (2 ml) and treated dropwise at
room temperature under Ar with 0.65 ml of the afore-
mentioned solution (ca. 1.2 mmol). The reaction mix-
ture turned yellow and was then cooled to 0°C,
followed by addition of triethylamine (0.5 ml, 3.6
mmol). After stirring for 5 min, the reaction mixture
was poured into MeOH (10 ml). Stirring was continued
for 30 min at room temp. After removal of all solvents
in vacuo, the residue was chromatographed (hexane/
EtOAc 9:1, then 7:3) to yield 9c (65 mg, 80%). Method
B. Compound 10b (292 mg, 0.5 mmol) was dissolved in
a 1:1 MeOH/THF mixture (5 ml) and treated with
conc. HCl (0.1 ml). The mixture was then stirred at
room temp. for 18 h. After addition of solid Na2CO3
(210 mg, 2 mmol), the mixture was stirred for further 5
min and evaporated in vacuo. CC of the residue on
silica gel (hexane–EtOAc 9:1, then 7:3) furnished 10c
(140 mg, 82%).
4.7. Addition of allyl zinc reagents
The allylzinc reagents were prepared as described in the
literature by: (a) insertion of Rieke Zn into the CꢁBr
bond of allyl bromide,16a followed by reaction in anhy-
drous THF at −78°C (4 h). (b) in situ generation by
reaction of Zn with allyl bromide under Barbier condi-
tions,16b including sonication, followed by reaction in
anhydrous THF at room temp. (6 h). (c) Transmetalla-
tion of zinc chloride with allyl magnesium chloride,16c
followed by reaction in anhydrous THF at room temp.
(6 h). In none of these conditions could a reaction be
observed.
4.8. Addition of allyl silane and allyl stannane in the
presence of Lewis acids
,
Molecular sieves (4 A) were added at room temp. to a
solution of the appropriate ketone (1 mmol) in CH2Cl2
(2 ml). The solution was then stirred for 10 min at
room temp. and cooled to −78°C. The appropriate
Lewis acid (1.2 mmol) dissolved in CH2Cl2 (2 ml) was
then added dropwise at the same temperature. The
stirring was continued for 10 min, followed by addition
of a solution of allyl trimethyl silane or allyl tri-n-butyl
stannane (1.2 mmol) in CH2Cl2 (2 ml). The reaction
mixture was then stirred for about 6 h (allyl silane) or
4 h (allyl stannane) (TLC monitoring). During this time
the temperature was slowly increased to −40°C. Work-
up (CH2Cl2) and column chromatography (hexane/
EtOAc mixtures) yielded the desired allylation product.
4.12. Ozonolysis of 9a in basic medium to yield methyl
ester 14
A solution of alcohol 9a (220 mg, 0.5 mmol) in CH2Cl2
(5 ml) was cooled to −78°C and treated with methanolic
NaOH (2.5 M, 1 ml). Ozone (O2/O3) was then bubbled
through the solution for 30 min. After eliminating the
residual ozone by bubbling Ar through the solution, the
reaction mixture was worked up (CH2Cl2) and chro-
matographed on silica gel (hexane–EtOAc 8:2) to
afford ester 14 (196 mg, 83%). The same procedure was
used to convert 10a to 20 (80%).

M. Carda et al. / Tetrahedron: Asymmetry 12 (2001) 1417–1429
1425
4.13. Desilylation of 14 with concomitant ring closure to
lactone 15
two times with EtOH, and the organic layers were
concentrated to dryness at reduced pressure. This gave
a crude triol which was used in the next reaction.
A solution of ester 14 (472 mg, 1 mmol) in dry THF (5
ml) was treated with solid tetra-n-butylammonium
fluoride hydrate (288 mg, 1.1 mmol) and stirred under
Ar at room temp. for 45 min. After eliminating the
volatiles at reduced pressure, CC of the residue on silica
gel (hexane/EtOAc 8:2, then 7:3) provided lactone 15
(136 mg, 67%). The same procedure was used to con-
vert ester 20 to lactone 18 (64%).
The triol obtained above was dissolved in dry acetone
(2 ml) and treated with anhydrous CuSO4 (32 mg, 0.2
mmol).27 After adding p-toluenesulphonic acid (5 mg),
the solution was stirred under Ar for 24 h at room
temp. (TLC monitoring). Work-up (CH2Cl2) and CC
on silica gel (hexane–EtOAc 8.2 and then 1:1) furnished
lactone 19 (36 mg, 60% overall from 18).
4.17. Conversion of the stereoisomeric mixtures 12/13a
and 12/13b (G=COOtBu) into lactones 15/19 and 18/
21, respectively
4.14. Conversion of vinyl carbinol 16 to nitrile 17
A solution of vinyl carbinol 166a (567 mg, 1 mmol) in
dry CH2Cl2 (10 ml) was cooled to −78°C. Ozone (O2/
O3) was then bubbled through the solution for 45 min.
After eliminating the residual ozone by bubbling Ar
through the reaction mixture, a solution of NaBH4 (190
mg, 5 mmol) in EtOH (5 ml) was added dropwise with
subsequent removal of the cooling bath. After reaching
room temperature, the reaction mixture was stirred for
further 2 h and worked up (CH2Cl2) to furnish a crude
diol, which was directly subjected to mesylation.
The mixtures of esters 12/13a and 12/13b (1 mmol in
each case) was treated with TBAF (1.1 mmol) in dry
THF as described above for the conversion 14“15. In
the case of 12/13a, this yielded a mixture of lactones 15
(major) and 19 (minor). In the case of 12/13b, this
yielded a mixture of lactones 18 and 21, from which the
major lactone 21 could be isolated in 65% yield by
crystallization from hexane–EtOAc.
4.18. (2S)-1-O-(tert-Butyldiphenylsilyl)-2-[(4S)-2,2-
The diol obtained above was dissolved in dry CH2Cl2 (5
ml), cooled to −30°C and treated under Ar with mesyl
chloride (78 ml, 1 mmol), triethyl amine (210 ml, 1.5
mmol) and DMAP (122 mg, 1 mmol). The reaction
mixture was stirred for 2 h at −30°C, then for 5 h at
0°C. Work-up (CH2Cl2) gave a crude mesylate which
was used in the next reaction.
dimethyl-1,3-dioxolan-4-yl]-4-pentene-1,2-diol 9a
Oil, [h]D=+2.8 (c 4; CHCl3); IR wmax (NaCl) cm−1: 3550
(br, OH), 3065, 2936, 2853, 2846, 1464, 1432, 1375,
1
1216, 1106, 1065, 920, 820, 740; H NMR (400 MHz) l
7.65 (4H, m, aromatic), 7.40 (6H, m, aromatic), 5.84
(1H, ddt, J=17, 10.5, 7.5 Hz, H-4), 5.08 (1H, br d,
J=17 Hz, H-5t), 5.07 (1H, br d, J=10.5 Hz, H-5c), 4.23
A solution of the crude mesylate and NaCN (122 mg,
2.5 mmol) in dry DMSO (5 ml) was heated at 55° for
18 h. After cooling to room temp., the reaction mixture
was poured onto brine and extracted with EtOAc. The
organic layer was back-washed with brine and dried on
anhydrous Na2SO4. After solvent removal, the residue
was chromatographed on silica gel (hexane–EtOAc 7:3)
to yield nitrile 17 (348 mg, 60% overall from 16).
(1H, t, J=7 Hz, H-4%), 3.90 (2H, m, H-5%, H-5% ), 3.63,
a
3.49 (2H, AB system, J=10.5 Hz, H-1a/H-1bb), 2.42
(2H, m, H-3a, H-3b), 1.40, 1.36 (2×s, 3H each, acetonide
methyls), 1.07 (9H, s, tBuSi); 13C NMR (100 MHz) l
135.6, 135.5, 132.8, 132.7, 129.9, 129.8, 127.8, 127.7,
127.6 (aromatic), 133.3 (C-4), 118.4 (C-5), 108.9 (C-2%),
78.2 (C-4%), 73.5 (C-2), 66.0 (C-1), 64.6 (C-5%), 38.8
(C-3), 26.9 (SiCMe3), 26.3, 25.1 (acetonide methyls),
19.3 (SiCMe3); CIMS, m/z (% rel. int.) 441. 2467
[M+H]+ (16), 383 [M−tBu]+ (74), 247 (100), 101 (90).
Calcd for C26H37O4Si, M=441.2461. Anal. calcd for
C26H36O4Si: C, 70.87; H, 8.23. Found, C, 70.60; H,
8.42%.
4.15. Hydrolysis of 17 to lactone 18
Nitrile 17 (290 mg, 0.5 mmol) was dissolved in a
MeOH/conc. HCl 5:1 mixture (8 ml) and heated under
reflux for 2 h. After cooling to room temp. and removal
of most volatiles in vacuo, the reaction mixture was
worked up (EtOAc). CC of the residue on silica gel
(hexane–EtOAc 9:1, then 7:3) provided lactone 18 (154
mg, 90%).
4.19. (2S)-1-O-Trityl-2-[(4S)-2,2-dimethyl-1,3-dioxolan-
4-yl]-4-pentene-1,2-diol 9b
Oil, [h]D=−7.1 (c 6.6; CHCl3); IR wmax (NaCl) cm−1:
3400 (br, OH), 2985, 2932, 1595, 1472, 1221, 1069, 753,
1
700; H NMR (400 MHz) l 7.50–7.20 (15H, m, aro-
4.16. Conversion of lactone 18 to lactone 19
matic), 5.61 (1H, ddt, J=17.2, 10.5, 7.5 Hz, H-4), 5.02
(1H, br dd, J=17.2, 2 Hz, H-5t), 4.96 (1H, br dd,
J=10.5, 2 Hz, H-5c), 4.38 (1H, dd, J=8, 6.3 Hz, H-4%),
The hydrogenation catalyst 20% Pd(OH)2/C (50 mg)
was suspended in EtOH (5 ml) and stirred for 10 min
under an H2 atmosphere. The substrate (103 mg, 0.3
mmol) was dissolved in EtOH (5 ml) and added via
syringe to the catalyst suspension. The reaction mixture
was then stirred for 24 h at room temp. After this time
(TLC monitoring), the mixture was filtered through
Celite, the reaction flask and the Celite were washed
4.00 (1H, dd, J=8, 6.3 Hz, H-5%), 3.91 (1H, t, J=8 Hz,
a
H-5% ), 3.19, 3.00 (2H, AB system, J=8.8 Hz, H-1a/H-
1b), b2.35 (1H, br dd, J=13.8, 7.7 Hz, H-3a), 2.19 (1H,
dd, J=13.8, 7 Hz, H-3b), 1.41, 1.37 (2×s, 3H each,
acetonide methyls); 13C NMR (100 MHz) l 143.7,
128.7, 127.7, 127.1 (aromatic), 132.3 (C-4), 118.7 (C-5),
108.9 (C-2%), 86.7 (CPh3), 78.1 (C-4%), 73.1 (C-2), 65.3

1426
M. Carda et al. / Tetrahedron: Asymmetry 12 (2001) 1417–1429
(C-1), 64.5 (C-5%), 38.3 (C-3), 26.4, 25.6 (acetonide
methyls); CIMS, m/z (% rel. int.) 445.2387 [M+H]+
(0.5), 271 (65), 244 (100). Calcd for C29H33O4, M=
445.2379. Anal. calcd for C29H32O4: C, 78.35; H, 7.26.
Found, C, 78.60; H, 7.40%.
system, J=9.5 Hz, H-1a/H-1b), 3.00 (1H, br s, OH),
2.50 (1H, br m, H-3a, H-3b); 13C NMR (125 MHz) l
143.7, 138.7, 138.1, 128.8, 128.3, 128.2, 127.7, 127.6,
127.4, 127.0 (aromatic), 133.6 (C-4), 117.8 (C-5), 86.8
(CPh3), 80.8 (C-1%), 75.5 (C-2), 73.4 (2×OCH2Ph), 70.8
(C-1), 65.5 (C-2%), 38.8 (C-3); EIMS, m/z (% rel. int.)
493.2377 [M−Bn]+ (1), 333 (12), 244 (100). Calcd for
C40H40O4-Bn, M=493.2378. Anal. calcd for C40H40O4:
C, 82.16; H, 6.89. Found, C, 82.00; H, 7.01%.
4.20. (2S)-2-[(4S)-2,2-Dimethyl-1,3-dioxolan-4-yl]-4-
pentene-1,2-diol 9c
Oil, [h]D=+7.1 (c 4.8; CHCl3); IR wmax (NaCl) cm−1:
3450 (br, OH), 3076, 2985, 2935, 1376, 1257, 1217,
4.23. (2R)-2-[(1S)-1,2-Bis(benzyloxy)ethyl]-4-pentene-
1,2-diol 10c
1
1157, 1064, 920, 858; H NMR (500 MHz) l 5.82 (1H,
ddt, J=17, 10.3, 7 Hz, H-4), 5.12 (1H, br d, J=10.3
Hz, 1H, H-5c), 5.11 (1H, br d, J=17 Hz, H-5t), 4.10
Oil, [h]D=+24.7 (c 2.2; CHCl3); IR wmax (NaCl) cm−1:
1
(1H, t, 7 Hz, H-4%), 4.00 (2H, d, J=7 Hz, H-5%, H-5% ),
3430 (br, OH), 2928, 2873, 1720, 1453, 1096, 1068; H
a
3.72, 3.52 (2H, AB system, J=11.2 Hz, H-1a/H-1bb),
2.70 (1H, br s, OH), 2.50 (1H, br s, OH), 2.28 (1H, dd,
J=14, 7 Hz, H-3a), 2.13 (1H, dd, J=14, 7 Hz, H-3b),
1.42, 1.35 (2×s, 3H each, acetonide methyls); 13C NMR
(125 MHz) l 132.3 (C-4), 119.0 (C-5), 109.4 (C-2%), 80.2
(C-4%), 72.3 (C-2), 67.3 (C-1), 64.7 (C-5%), 38.9 (C-3),
26.2, 25.4 (acetonide methyls); EIMS, m/z (% rel. int.)
187.0960 [M−CH3]+ (6), 171 [M−CH2OH]+ (4), 101
(100). Calcd for C10H18O4−CH3, M=187.0970. Anal.
calcd for C10H18O4: C, 59.39; H, 8.97. Found, C, 59.60;
H, 9.06%.
NMR (500 MHz) l 7.40–7.25 (10H, m, aromatic), 5.90
(1H, ddt, J=17.5, 10, 7 Hz, H-4), 5.12 (1H, br d, J=10
Hz, H-5c), 5.10 (1H, br d, J=17.5 Hz, H-5t), 4.78, 4.62
(2H, AB system, J=11.5 Hz, OCH2Ph), 4.58 (2H, s,
OCH2Ph), 3.84 (1H, dd, J=10, 5 Hz, H-2%), 3.74 (1H,
a
t, J=5 Hz, H-1%), 3.70–3.65 (2H, m, H-2%b, H-1a), 3.53
(1H, dd, J=11.5, 5 Hz, H-1b), 3.10 (1H, br s, 2-OH),
2.50 (1H, t, J=6 Hz, 1-OH), 2.41 (1H, dd, J=14, 7.5
Hz, H-3a), 2.32 (1H, dd, J=14, 7.5 Hz, H-3b); 13C
NMR (125 MHz) l 138.1, 137.4, 128.6, 128.5, 128.0,
127.9, 127.8 (aromatic), 133.4 (C-5), 118.3 (C-6), 80.6
(C-2), 75.2 (C-3), 73.8, 73.5 (2×OCH2Ph), 69.1
(CH2OH), 65.9 (C-1), 38.6 (C-4); FAB MS, m/z
343.1923 [M+H+]. Calcd for C21H27O4, M=343.1909.
Anal. calcd for C21H26O4: C, 73.66; H, 7.65. Found, C,
73.58; H, 7.80%.
4.21. (2R)-1-O-(tert-Butyldiphenylsilyl)-2-[(1S)-1,2-bis-
(benzyloxy)ethyl]-4-pentene-1,2-diol 10a
Oil, [h]D=−2.4 (c 11.8; CHCl3); IR wmax (NaCl) cm−1:
3550 (br, OH), 3069, 2931, 2859, 1457, 1428, 1391,
1
1364, 1266, 1108, 918, 821, 740; H NMR (500 MHz) l
4.24. Methyl (3S)-4-(tert-butyldiphenylsilyloxy)-3-[(4S)-
2,2-dimethyl-1,3-dioxolan-4-yl]-3-hydroxybutanoate 14
7.65 (4H, m, aromatic), 7.40–7.20 (16H, br m, aro-
matic), 5.85 (1H, ddt, J=16.5, 11, 6 Hz, H-4), 5.03 (1H,
br d, J=11 Hz, H-5c), 5.02 (1H, br d, J=16.5 Hz,
H-5t), 4.90, 4.58 (2H, AB system, J=11.5 Hz,
OCH2Ph), 4.51, 4.48 (2H, AB system, J=12 Hz,
OCH2Ph), 3.90 (1H, dd, J=10, 2 Hz, H-2%), 3.82 (1H,
Oil, [h]D=−9.2 (c 0.35; CHCl3); IR wmax (NaCl) cm−1:
3490 (br, OH), 3072, 3049, 2933, 2891, 2858, 1739 (ester
CꢀO), 1473, 1428, 1371, 1210, 1100, 1070, 1008, 858,
1
825, 742, 703; H NMR (500 MHz) l 7.70 (4H, br m,
dd, J=6, 2 Hz, H-1%), 3.73 (1H, dd, J=10, a6 Hz, H-2% ),
aromatic), 7.50–7.40 (6H, br m, aromatic), 4.35 (1H,
dd, J=7.5, 6.5 Hz, H-4%), 4.05 (1H, dd, J=8, 6.5 Hz,
b
3.69, 3.61 (2H, AB system, J=10 Hz, H-1a/H-1b), 2.90
(1H, br s, OH), 2.40 (2H, m, H-3a, H-3b), 1.07 (9H, s,
tBuSi); 13C NMR (125 MHz) l 138.8, 138.2, 135.7,
133.0, 129.8, 128.4, 128.2, 127.7, 127.6 (aromatic), 133.6
(C-4), 117.8 (C-5), 81.0 (C-1%), 75.6 (C-2), 73.6, 73.5
(2×OCH2Ph), 71.3 (C-1), 66.1 (C-2%), 38.1 (C-3), 27.0
(SiCMe3), 19.3 (SiCMe3); CIMS, m/z (% rel. int.)
581.3079 [M+H+] (1), 107 (100). Calcd for C37H45O4Si,
M=581.3087. Anal. calcd for C37H44O4Si: C, 76.51; H,
7.64. Found, C, 76.77; H, 7.78%.
H-5%), 3.98 (1H, dd, J=8, 7.5 Hz, H-5% ), 3.74, 3.72 (2H,
b
AB asystem, J=10 Hz, H-4a/H-4b), 3.70 (3H, s, OMe),
3.60 (1H, br s, OH), 2.75, 2.66 (2H, AB system, J=15
Hz, H-2a/H-2b), 1.41, 1.40 (2×3H, 2×s, acetonide Me),
1.10 (9H, s, Me3CSi); 13C NMR (125 MHz) l 172.1
(CꢀO), 135.6, 132.8, 132.7, 129.8, 127.9, 127.7 (aro-
matic), 109.1 (C-2%), 78.4 (C-4%), 73.2 (C-3), 66.5 (C-4),
64.6 (C-5%), 51.8 (OMe), 36.9 (C-3), 26.7 (Me3CSi), 26.1,
25.1 (acetonide Me), 19.2 (Me3CSi); FAB MS, m/z (%
rel. int.) 473.2368 [M+H+] (1), 415 (100). Calcd for
C26H37O6Si, M=473.2359. Anal. calcd for C26H36O6Si:
C, 66.07; H, 7.68. Found, C, 66.20; H, 7.81%.
4.22. (2R)-1-O-Trityl-2-[(1S)-1,2-bis(benzyloxy)ethyl]-4-
pentene-1,2-diol 10b
Oil, [h]D=−2.4 (c 11.4; CHCl3); IR wmax (NaCl) cm−1:
3550 (br, OH), 3062, 3030, 2873, 1491, 1449, 1364,
1214, 1078, 915, 743, 700; 1H NMR (500 MHz) l
7.50–7.25 (25H, br m, aromatic), 5.82 (1H, ddt, J=
17.2, 10.2, 7 Hz, H-4), 5.05 (2H, br m, H-5c, H-5t), 4.90,
4.60 (2H, AB system, J=11.5 Hz, OCH2Ph), 4.47, 4.45
(2H, AB system, J=12 Hz, OCH2Ph), 3.87 (1H, dd,
4.25. (4S)-4-[(4S)-2,2-Dimethyl-1,3-dioxolan-4-yl]-4-
hydroxytetrahydrofuran-2-one 15
White needles, mp 69–70°C (from hexane–Et2O), [h]D=
−35.4 (c 1.4; CHCl3); IR wmax (NaCl) cm−1: 3450 (br,
OH), 2988, 2933, 1775 (lactone CꢀO), 1462, 1377, 1263,
1
1210, 1157, 1111, 1067, 1033, 852; H NMR (400 MHz)
J=6, 3 Hz, H-1%), 3.82 (1H, dd, J=10.4, 3 Hz, H-2%),
l 4.33, 4.22 (2H, AB system, J=10 Hz, H-5a/H-5b),
4.18 (1H, dd, J=7, 6 Hz, H-4%), 4.03 (1H, dd, J=8.8, 7
a
3.63 (1H, dd, J=10.4, 6 Hz, H-2% ), 3.30, 3.17 (2H, AB
b

M. Carda et al. / Tetrahedron: Asymmetry 12 (2001) 1417–1429
1427
Hz, H-5%), 3.84 (1H, dd, J=8.8, 6 Hz, H-5% ), 3.20 (1H,
6.8 Hz, H-5%), 3.84 (1H, dd, J=8.5, 6 Hz, H-5% ), 2.79,
a
b
a
br s, OH), 2.64, 2.47 (2H, AB system, J=17.5 Hz,
H-3a/H-3b), 1.42, 1.32 (2×s, 3H each, acetonide
methyls); 13C NMR (100 MHz) l 175.4 (C-2), 110.3
(C-2%), 77.7 (C-4%), 76.6 (C-4+C-5), 64.8 (C-5%), 38.6
(C-3), 26.2, 24.5 (acetonide methyls); CIMS, m/z (% rel.
int.) 203.0919 [M+H+] (97), 187 [M+−Me] (52), 145
(100), 127 (82). Calcd for C9H15O5, M=203.0919. Anal.
calcd for C9H14O5: C, 53.46; H, 6.98. Found, C, 53.69;
H, 7.00%.
2.57 (2H, AB system, J=17.8 Hz, H-3a/H-3bb), 2.60
(1H, br s, OH), 1.45, 1.36 (2×s, 3H each, acetonide
methyls); 13C NMR (100 MHz) l 171.9 (C-2), 110.3
(C-2%), 77.5 (C-4%), 75.3 (C-4+C-5), 65.0 (C-5%), 39.2
(C-3), 26.2, 24.5 (acetonide methyls); CIMS, m/z (% rel.
int.) 203.0922 [M+H+] (54), 185 [M+H+−H2O] (100).
Calcd for C9H15O5, M=203.0919. Anal. calcd for
C9H14O5: C, 53.46; H, 6.98. Found, C, 53.40; H, 7.05%.
4.29. Methyl (3R,4S)-4,5-bis(benzyloxy)-3-(tert-
4.26. (3R,4S)-4,5-Bis(benzyloxy)-3-(tert-butyldiphenylsi-
butyldiphenylsilyloxymethyl)-3-hydroxypentanoate 20
lyloxymethyl)-3-hydroxypentanenitrile 17
Oil, [h]D=+7.5 (c 2.2; CHCl3); IR wmax (NaCl) cm−1:
3470 (br, OH), 3065, 3031, 2932, 2859, 1732 (ester
CꢀO), 1453, 1430, 1392, 1342, 1259, 1205, 1109, 1024,
Oil, [h]D=+6.7 (c 4.2; CHCl3); IR wmax (NaCl) cm−1:
3450 (br, OH), 3066, 3033, 2932, 2860, 2250 (CꢂN),
1459, 1426, 1364, 1265, 1208, 1107, 1027, 820, 740, 701;
1H NMR (400 MHz) l 7.60 (4H, br m, aromatic),
7.50–7.20 (16H, br m, aromatic), 4.74, 4.48 (2H, AB
system, J=11.5 Hz, OCH2Ph), 4.42, 4.40 (2H, AB
system, J=12 Hz, OCH2Ph), 3.80–3.70 (2H, m, H-4/H-
5a), 3.76, 3.66 (2H, AB system, J=10.5 Hz, TPSOCH2),
3.60 (1H, dd, J=10.5, 4 Hz, H-5b), 3.45 (1H, br s, OH),
2.82, 2.64 (2H, AB system, J=16.5 Hz, H-2a/2b), 1.06
(9H, s, Me3CSi); 13C NMR (100 MHz) l 143.6, 138.7,
138.0, 128.8, 128.3, 128.2, 127.7, 127.6, 127.4, 127.0
(aromatic), 133.5 (C-4), 117.8 (C-5), 86.8 (CPh3), 80.8
(C-1%), 75.5 (C-2), 73.4 (2×OCH2Ph), 70.8 (C-1), 65.4
(C-2%), 38.7 (C-3); EIMS, m/z (% rel. int.) 579.2811 [M]+
(1), 536 (10), 522 [M+−tBu] (100). Calcd for
1
822, 740, 702; H NMR (500 MHz) l 7.65 (4H, br m,
aromatic), 7.45–7.20 (16H, br m, aromatic), 4.86, 4.56
(2H, AB system, J=11.5 Hz, OCH2Ph), 4.51, 4.49 (2H,
AB system, J=12 Hz, OCH2Ph), 4.00–3.70 (5H, br m,
H-4, H-5a, H-5b, TPSOCH2), 3.49 (3H, s, OMe), 2.74,
2.60 (2H, AB system, J=15 Hz, H-2a/H-2b); 13C NMR
(125 MHz) l 172.8 (CꢀO), 138.7, 138.0, 135.7, 129.8,
128.3, 127.9, 127.7, 127.6, 127.4 (aromatic), 80.3 (C-4),
75.6 (C-3), 73.5 (2×OCH2Ph), 70.7 (C-5), 67.2
(TPSOCH2), 51.5 (OMe), 38.0 (C-2), 27.0 (Me3CSi),
19.3 (Me3CSi); CIMS, m/z (% rel. int.) 613.2990 [M+
H+] (7), 337 (82), 199 (100). Calcd for C37H45O6Si,
M=613.2985. Anal. calcd for C37H44O6Si: C, 72.52; H,
7.24. Found, C, 72.63; H, 7.39%.
C36H41NO4Si,
M=579.2804.
Anal.
calcd
for
C36H41NO4Si: C, 74.57; H, 7.13. Found, C, 74.33; H,
7.17%.
4.30. (4S)-4-[(1S)-1,2-Di(benzyloxy)ethyl]-4-hydroxy-
tetrahydrofuran-2-one 21
4.27. (4R)-4-[(1S)-1,2-Bis(benzyloxy)ethyl]-4-hydroxy-
White needles, mp 77–78°C (from hexane–EtOAc),
[h]D=+26.7 (c 3.4; CHCl3); IR wmax (NaCl) cm−1: 3450
(br, OH), 3031, 2923, 2869, 1779 (lactone CꢀO), 1454,
1368, 1264, 1180, 1094, 1023, 740; 1H NMR (400 MHz)
l 7.40–7.25 (10H, m, aromatic), 4.70, 4.49 (2H, AB
system, J=11.7 Hz, OCH2Ph), 4.53 (2H, s, OCH2Ph),
4.21, 4.19 (2H, AB system, J=10.2 Hz, H-5a/H-5b),
3.75–3.60 (3H, m, H-1%, H-2%, H-2% ), 3.30 (1H, d,
J=1Hz, OH), 2.71 (1H, dd, J=17.5, 1 Hz, H-3a), 2.41
(1H, d, J=17.5 Hz, H-3b); 13C NMR (100 MHz) l
175.4 (C-2), 137.0, 136.9, 128.6, 128.2, 128.1, 127.9
(aromatic), 78.8 (C-4), 78.2 (C-1%), 77.2 (C-5), 73.9, 72.5
(2×OCH2Ph), 68.2 (C-2%), 39.2 (C-3); CIMS, m/z (% rel.
int.) 343.1537 [M+H]+ (21), 91 (100). Calcd for
C20H23O5, M=343.1545. Anal. calcd for C20H22O5: C,
70.16; H, 6.48. Found, C, 70.04; H, 6.29%.
tetrahydrofuran-2-one 18
White needles, mp 90–91°C (from hexane–EtOAc),
[h]D=+46.5 (c 2; CHCl3); IR wmax (NaCl) cm−1: 3450
(br, OH), 3064, 3031, 2870, 1780 (lactone CꢀO), 1496,
1454, 1398, 1367, 1253, 1179, 1095, 1027, 911, 817, 736;
1H NMR (400 MHz) l 7.40–7.20 (10H, m, aromatic),
4.70, 4.49 (2H, AB system, J=12 Hz, OCH2Ph), 4.52
(2H, s, OCH2Ph), 4.26 (1H, dd, J=9.9, 1.5 Hz, H-5a),
4.11 (1H, d, J=9.9 Hz, H-5b), 3.72 (1H, dd, J=10, 4
Hz, H-2%), 3.65 (1H, dd, J=10, 6.8 Hz, H-2% ), 3.58 (1H,
a
b
a
b
dd, J=6.8, 4 Hz, H-1%), 3.30 (1H, d, J=1.5 Hz, OH),
2.70, 2.49 (2H, AB system, J=17.5 Hz, H-3a/H-3b); 13C
NMR (100 MHz) l 175.3 (C-2), 136.9, 136.8, 128.7,
128.4, 128.2, 127.8 (aromatic), 79.0 (C-4), 77.7 (C-1%),
75.9 (C-5), 73.9, 72.7 (2×OCH2Ph), 68.2 (C-2%), 39.8
(C-3); FAB MS, m/z 343.1557 [M+H]+. Calcd for
C20H23O5, M=343.1545. Anal. calcd for C20H22O5: C,
70.16; H, 6.48. Found, C, 70.34; H, 6.30%.
Acknowledgements
4.28. (4R)-4-[(4S)-2,2-Dimethyl-1,3-dioxolan-4-yl]-4-
The authors wish to acknowledge the financial support
of the DGICYT (project PB98-1438), BANCAJA (pro-
ject P1B99-18), and the Conseller´ıa de Cultura de la
Generalitat Valenciana (project GV-99-77-1-02). E.C.
thanks the latter institution for a pre-doctoral fellow-
ship. The authors are further indebted to Dr. Harald
Ro¨per, from Eridania Be´ghin-Say, R&D Centre, Vil-
voorde, Belgium, for his very generous gift of erythrulose.
hydroxytetrahydrofuran-2-one 19
Oil, [h]D=+9.1 (c 1.5; CHCl3); IR wmax (NaCl) cm−1:
3450 (br, OH), 2989, 2938, 2901, 1776 (lactone CꢀO),
1384, 1374, 1209, 1159, 1033, 851; 1H NMR (400 MHz)
l 4.26, 4.18 (2H, AB system, J=10 Hz, H-5a/H-5b),
4.17 (1H, dd, J=6.8, 6 Hz, H-4%), 4.07 (1H, dd, J=8.5,

1428
M. Carda et al. / Tetrahedron: Asymmetry 12 (2001) 1417–1429
structurally related erythrulose 1,3-O-ethylidene acetals
References
have also been reported: Carda, M.; Casabo´, P.; Gonza´-
lez, F.; Rodr´ıguez, S.; Domingo, L. R.; Marco, J. A.
Tetrahedron: Asymmetry 1997, 8, 559–577.
1. (a) Vasella, A. In Modern Synthetic Methods 1980; Schef-
fold, R., Ed.; Salle & Sauerla¨nder Verlag: Frankfurt,
1980; pp. 173–267; (b) Fraser-Reid, B.; Anderson, R. C.
Progr. Chem. Org. Nat. Prod. 1980, 39, 1–61; (c) Hanes-
sian, S. In Total Synthesis of Natural Products: The
Chiron Approach; Pergamon Press: Oxford, 1983; (d)
Inch, T. D. Tetrahedron 1984, 40, 3161–3213; (e) Licht-
enthaler, F. W. In Modern Synthetic Methods 1992;
Scheffold, R., Ed.; VHCA-VCH: Basel, 1992; pp. 273–
376; (f) Ferrier, R. J.; Middleton, S. Chem. Rev. 1993, 93,
2779–2831; (g) Monneret, C.; Florent, J.-C. Synlett 1994,
305–318; (h) Collins, P.; Ferrier, R. J. In Monosaccha-
rides. Their Chemistry and Their Roles in Natural Prod-
ucts; John Wiley and Sons: New York, 1995; (i) Bols, M.
In Carbohydrate Building Blocks; John Wiley and Sons:
New York, 1996.
2. (a) De Wilde, H.; De Clercq, P.; Vandewalle, M.; Ro¨per,
H. Tetrahedron Lett. 1987, 4757–4758; (b) Van der
Eycken, E.; De Wilde, H.; Deprez, L.; Vandewalle, M.
Tetrahedron Lett. 1987, 1, 4759–4760; (c) Marco, J. L. J.
Chem. Res. (S) 1988, 276, (M) 2013–2016; (d) Nagano,
H.; Ohno, M.; Miyamae, Y.; Kuno, Y. Bull. Chem. Soc.
Jpn. 1992, 65, 2814–2820; (e) Dequeker, E.; Compernolle,
F.; Toppet, S.; Hoornaert, G. Tetrahedron 1995, 51,
5877–5890.
7. Safont, V. S.; Moliner, V.; Oliva, M.; Castillo, R.;
Andre´s, J.; Gonza´lez, F.; Carda, M. J. Org. Chem. 1996,
61, 3467–3475.
8. (a) Keck, G.; Castellino, S. J. Am. Chem. Soc. 1986, 108,
3847–3849; (b) Keck, G.; Castellino, S.; Wiley, M. R. J.
Org. Chem. 1986, 51, 5478–5480. For reports where
b-chelation has been invoked to explain the stereochemi-
cal outcome of reactions involving 1,3-difunctional com-
pounds, see citations in Ref. 6a.
9. Phillips, A. J.; Abell, A. D. Aldrichimica Acta 1999, 32,
75–90.
10. For recent contributions of our group where allyl groups
were usefully manipulated in this sense, see: (a) Carda,
M.; Castillo, E.; Rodr´ıguez, S.; Falomir, E.; Marco, J. A.
Tetrahedron Lett. 1998, 8895–8896; (b) Carda, M.;
Castillo, E.; Rodr´ıguez, S.; Uriel, S.; Marco, J. A. Synlett
1999, 1639–1641; (c) Carda, M.; Castillo, E.; Rodr´ıguez,
S.; Marco, J. A. Tetrahedron Lett. 2000, 5511–5513.
11. (a) Marco, J. A.; Carda, M.; Gonza´lez, F.; Rodr´ıguez, S.;
Murga, J. Liebigs Ann. Chem. 1996, 1801–1810; (b)
Carda, M.; Rodr´ıguez, S.; Murga, J.; Falomir, E.; Marco,
J. A.; Ro¨per, H. Synth. Commun. 1999, 29, 2601–2610.
12. (a) Roush, W. R. In Comprehensive Organic Synthesis;
Trost, B. M.; Fleming, I.; Schreiber, S. L., Eds.; Perga-
mon Press: Oxford, 1991; Vol. 2, pp. 1–53; (b) Fleming, I.
In Comprehensive Organic Synthesis; Trost, B. M.; Flem-
ing, I.; Schreiber, S. L., Eds.; Pergamon Press: Oxford,
1991; Vol. 2, pp. 563–593; (c) Yamamoto, Y.; Asao, N.
Chem. Rev. 1993, 93, 2207–2293; (d) Yamamoto, Y.;
Shida, N. In Advances in Detailed Reaction Mechanisms;
Coxon, J. M., Ed.; JAI Press: Greenwich, CT, 1994; pp.
1–44; (e) Hoppe, D. In Houben-Weyl’s Methods of
Organic Chemistry, Stereoselective Synthesis; Helmchen,
G.; Hoffmann, R. W.; Mulzer, J.; Schaumann, E., Eds.;
Georg Thieme Verlag: Stuttgart, 1996; Vol. 3, pp. 1357–
1602.
13. (a) Anh, N. T. Top. Curr. Chem. 1980, 88, 145–162; (b)
Reetz, M. Top. Curr. Chem. 1982, 106, 1–54; (c) Reetz,
M. T. Angew. Chem., Int. Ed. Engl. 1984, 23, 556–569; (d)
Reetz, M. T. Acc. Chem. Res. 1993, 26, 462–468. For the
importance of chelation in ketone reductions, see, for
example: Oishi, T.; Nakata, T. Acc. Chem. Res. 1984, 17,
338–344; Iida, H.; Yamazaki, N.; Kibayashi, C. J. Org.
Chem. 1986, 51, 3769–3771; Taniguchi, M.; Fujii, H.;
Oshima, K.; Utimoto, K. Tetrahedron 1993, 49, 11169–
11182; Sarko, C. R.; Guch, I. C.; DiMare, M. J. Org.
Chem. 1994, 59, 705–706; Sarko, C. R.; Collibee, S. E.;
Knorr, A. L.; DiMare, M. J. Org. Chem. 1996, 61,
868–873. For the role of chelation in other synthetically
important processes, see: Kauffmann, T. Synthesis 1995,
745–755.
3. Chen, X.; Hortelano, E. R.; Eliel, E. L.; Frye, S. V. J.
Am. Chem. Soc. 1992, 114, 1778–1784.
4. (a) Huryn, D. M. In Comprehensive Organic Synthesis;
Trost, B. M.; Fleming, I.; Schreiber, S. L., Eds.; Perga-
mon Press: Oxford, 1991; Vol. 1, pp. 49–75; (b) Lipshutz,
B. H. In Comprehensive Organic Synthesis; Trost, B. M.;
Fleming, I.; Schreiber, S. L., Eds.; Pergamon Press:
Oxford, 1991; Vol. 1, pp. 107–138; (c) Ferreri, C.;
Palumbo, G.; Caputo, R. In Comprehensive Organic Syn-
thesis; Trost, B. M.; Fleming, I.; Schreiber, S. L., Eds.;
Pergamon Press: Oxford, 1991; Vol. 1, pp. 139–172; (d)
Saccomano, N. A. In Comprehensive Organic Synthesis;
Trost, B. M.; Fleming, I.; Schreiber, S. L., Eds.; Perga-
mon Press: Oxford, 1991; Vol. 1, pp. 173–209; (e)
Knochel, P. In Comprehensive Organic Synthesis; Trost,
B. M.; Fleming, I.; Schreiber, S. L., Eds.; Pergamon
Press: Oxford, 1991; Vol. 1, pp. 211–229; (f) Imamoto, T.
In Comprehensive Organic Synthesis; Trost, B. M.; Flem-
ing, I.; Schreiber, S. L., Eds.; Pergamon Press: Oxford,
1991; Vol. 1, pp. 231–250; (g) Shambayati, S.; Schreiber,
S. L. In Comprehensive Organic Synthesis; Trost, B. M.;
Fleming, I.; Schreiber, S. L., Eds.; Pergamon Press:
Oxford, 1991; Vol. 1, pp. 283–324; (h) Yamaguchi, M. In
Comprehensive Organic Synthesis; Trost, B. M.; Fleming,
I.; Schreiber, S. L., Eds.; Pergamon Press: Oxford, 1991;
Vol. 1, pp. 325–353.
5. (a) Chikashita, H.; Nakamura, Y.; Uemura, H.; Itoh, K.
Chem. Lett. 1992, 439–442; (b) Evans, D. A.; Dart, M. J.;
Duffy, J. L.; Yang, M. G. J. Am. Chem. Soc. 1996, 118,
4322–4343; (c) Mulzer, J.; Pietschmann, C.; Buschmann;
Luger, P. J. Org. Chem. 1997, 62, 3938–3943.
6. (a) Marco, J. A.; Carda, M.; Gonza´lez, F.; Rodr´ıguez, S.;
Castillo, E.; Murga, J. J. Org. Chem. 1998, 63, 698–707;
(b) Gonza´lez, F. Ph.D. Thesis, University Jaume I,
Castello´n, 1995; (c) Castillo, E., Ph.D. Thesis, University
Jaume I, Castello´n, 2000. Diastereoselective additions to
14. (a) Reetz, M. T. Organotitanium Reagents in Organic
Synthesis; Springer Verlag: Berlin, 1986; (b) Reetz, M. T.
In Organometallics in Synthesis; Schlosser, M., Ed.; John
Wiley and Sons: New York, 1994; Chapter 3; the prepa-
−
ration of the ‘ate’ complex allylTi(OiPr)4 from allylmag-
nesium bromide and Ti(O-iso-Pr)4 is described on p. 260;
(c) AllylTiCl3 was generated from allylmagnesium bro-
mide and TiCl4 according to a modification of the proce-

M. Carda et al. / Tetrahedron: Asymmetry 12 (2001) 1417–1429
dure described in: Thiele, K.-H.; Jacob, K. Z. Anorg.
1429
Allg. Chem. 1968, 356, 195–201; (d) AllylTiCl(O-iso-Pr)2
was generated from iso-propylmagnesium bromide, Ti(O-
iso-Pr)4 and allyl bromide as described in: Gao, Y.; Sato,
F. J. Org. Chem. 1995, 60, 8136–8137; (e) AllylTi(O-iso-
Pr)3 was generated from allylmagnesium bromide and
ClTi(O-iso-Pr)3 as described in: Reetz, M. T.; Steinbach,
R.; Westermann, J.; Peter, R.; Wenderoth, B. Chem. Ber.
1985, 118, 1441–1454.
20. Hosomi, A.; Shirahata, A.; Sakurai, H. Tetrahedron Lett.
1978, 3043–3046. Ketones 3 and 5, which bear the TBAF-
cleavable TPS group, were not subjected to these reaction
conditions.
21. Crystallographic data of lactone 18 have been deposited
in the Cambridge Crystallographic Data Centre as sup-
plementary publication no. CCDC 156077. Copies of the
data can be obtained free of charge on application to
CCDC, 12 Union Road, Cambridge CB2 1EZ, UK [Fax:
int. code+44(1223)336-033; e-mail: deposit@ccdc.cam.
ac.uk].
22. (a) Mead, K.; Macdonald, T. L. J. Org. Chem. 1985, 50,
422–424; (b) Jurczak, J.; Pikul, S.; Bauer, T. Tetrahedron
1986, 42, 447–488. Steric crowding may also channel a
process towards a non-chelation pathway: Ley, S. V.;
Woods, M.; Zanotti-Gerosa, A. Synthesis 1992, 52–54.
23. In the reactions of Grignard reagents with 2-acyl tetra-
hydrofurans, which only have an oxygen atom in the
ring, high chelation-controlled stereoselectivities are often
observed: Eliel, E. L. In Asymmetric Synthesis; Morrison,
J. D., Ed.; Academic Press: New York, 1983; Vol. 2,
Chapter 5. For a further example structurally very close
to 4, see: Rao, A. V. R.; Gurjar, M. K.; Devi, T. R.;
Kumar, K. R. Tetrahedron Lett. 1993, 1653–1656.
24. (a) Annunziata, R.; Cinquini, M.; Cozzi, F.; Fuchicello,
A. Tetrahedron 1991, 47, 3853–3868; (b) Frye, S. V.; Eliel,
E. L. J. Am. Chem. Soc. 1988, 110, 484–489; (c) Bai, X.;
Eliel, E. L. J. Org. Chem. 1992, 57, 5166–5172.
25. Schlosser, M. In Organometallics in Synthesis; Schlosser,
M., Ed.; John Wiley and Sons: New York, 1994; pp.
120–121.
15. (a) Lipshutz, B. H.; Sengupta, S. Org. React. 1992, 41,
135–631; (b) Lipshutz, B. H. In Organometallics in Syn-
thesis; Schlosser, M., Ed.; John Wiley and Sons: New
York, 1994; Chapter 4. The allylcopper reagent used in
our study (Table 2, entries 5 and 11) was prepared
according to the procedure described on pp. 325–326.
16. (a) Hanson, M. V.; Brown, J. D.; Rieke, R. D.; Niu, Q.
J. Tetrahedron Lett. 1994, 7205–7208; (b) Ranu, B. C.;
Majee, A.; Das, A. R. Tetrahedron Lett. 1995, 4885–4888;
(c) McGarvey, G. J.; Mathys, J. A.; Wilson, K. J.;
Overly, K. R.; Buonora, P. T.; Spoors, P. G. J. Org.
Chem. 1995, 60, 7778–7790.
17. Weber, W. P. Silicon Reagents for Organic Synthesis;
Springer Verlag: Berlin, 1983; pp. 173–205.
18. (a) Ipaktschi, J.; Heydari, A.; Kalinowski, H.-O. Chem.
Ber. 1994, 127, 905–909; (b) Schmid, U.; Waldmann, H.
In Organic Synthesis Highlights III; Mulzer, J.; Wald-
mann, H., Eds.; Wiley-VCH: Weinheim, 1998; pp. 205–
210; (c) Saito, S. In Lewis Acids in Organic Synthesis;
Yamamoto, H., Ed.; Wiley-VCH: Weinheim, 2000; Chap-
ter 2; (d) Motoyama, Y., Nishiyama, H. Ref. 18c, Chap-
ter 3; (e) Ishihara, K. Ref. 18c, Chapter 4; (f) Ishihara, K.
Ref. 18c, Chapter 9; (g) Urabe, H., Sato, F. Ref. 18c,
Chapter 15.
19. This Lewis acid was prepared according to the procedure
described in Ref. 14b, p. 258. With ketone 3 and allyl
trimethylsilane, the only product isolated after 5 h in
CH2Cl2 at −10°C was alcohol i (63%). With ketone 3 and
allyl tri-n-butylstannane, the only product isolated after 5
h in CH2Cl2 at −10°C was 9a (77%).
26. Krainer, E.; Naider, F.; Becker, J. Tetrahedron Lett.
1993, 1713–1716.
27. Kocienski, P. K.; Yeates, C.; Street, S. D. A.; Campbell,
S. F. J. Chem. Soc., Perkin Trans. 1 1987, 2183–2187.
.