Allyltrichlorostannane additions to chiral aldehydes

Article abstract of DOI:10.1016/S0040-4039(03)01713-1

Chiral and achiral allyltrichlorostannanes reacted with chiral β-alkoxy and syn and anti α-methyl-β-alkoxy aldehydes to give the corresponding homoallylic alcohols with moderate to high diastereoselectivities.

Full text of DOI:10.1016/S0040-4039(03)01713-1

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 6861–6866
Allyltrichlorostannane additions to chiral aldehydes
Luiz Carlos Dias,* De´bora Ribeiro dos Santos and Leonardo Jose´ Steil
Instituto de Qu´ımica, Universidade Estadual de Campinas, UNICAMP C.P. 6154, 13084-971, Campinas, SP, Brazil
Received 23 May 2003; revised 7 July 2003; accepted 12 July 2003
This paper is dedicated to Professor Albert James Kasheres on the occasion of his 60th birthday and also to the Brazilian Chemical Society
(SBQ)
Abstract—Chiral and achiral allyltrichlorostannanes reacted with chiral b-alkoxy and syn and anti a-methyl-b-alkoxy aldehydes
to give the corresponding homoallylic alcohols with moderate to high diastereoselectivities.
© 2003 Elsevier Ltd. All rights reserved.
The Lewis-acid mediated reaction of allylsilanes and
allylstannanes with aldehydes is a well-known proce-
dure for the preparation of homoallylic alcohols.1 Chi-
ral allylmetal reagents may be thought of as
acetate–enolate equivalents for diastereoselective con-
struction of stereochemically well-defined homoallylic
alcohols. Because these reactions complement the aldol
reactions, allylsilanes and allylstannanes are among the
most important groups of organometallic reagents
available for the control of acyclic stereochemistry.2
We recently communicated that in situ prepared chiral
allyltrichlorostannanes react with chiral a-methyl alde-
hydes to give 1,2-syn homoallylic alcohols that are key
intermediates for the preparation of polyacetate and
polypropionate-derived natural products.3–5 We have
described also that chiral and achiral allyltrichlorostan-
nanes react with N-Boc-a-aminoaldehydes to give 1,2-
syn N-Boc-a-aminoalcohols that are key intermediates
for the preparation of hydroxyethylene dipeptide
isosteres.6–9
We wish to describe here a divergently stereocontrolled
reaction between chiral b-alkoxy and a-methyl-b-
alkoxy aldehydes with achiral and chiral allyl-
trichlorostannanes to give homoallylic alcohols with
moderate to high diastereoselectivities.10 This study
details our efforts to understand the double stereodif-
ferentiating stereocontrol elements involved in chiral
allyltrichlorostannane additions to chiral b- and a,b-
disubstituted aldehydes.3–9 In this part of the investiga-
tion, we have examined the interplay between 1,2-
(Felkin–Anh), 1,3- and 1,4-asymmetric induction in
allyltrichlorostannane reactions with a-methyl-b-alkoxy
aldehydes under conditions that preclude internal chela-
tion with the aldehyde b-alkoxy substituent.
Scheme 1.
Scheme 2.
Achiral allylsilane 2 was prepared from phenylacetic
acid methyl ester 1, while chiral allylsilanes (R)-5 and
(S)-5 were prepared from methyl 3-hydroxy-2-methyl-
* Corresponding author. Tel.: +55-019-3788-3021; fax: +55-019-3788-
3023; e-mail: ldias@iqm.unicamp.br
0040-4039/$ - see front matter © 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0040-4039(03)01713-1

6862
L. C. Dias et al. / Tetrahedron Letters 44 (2003) 6861–6866
propionate 4, both enantiomers of which are commer-
cially available (Scheme 1).9,11 According to previously
established experimental procedures, allylsilanes 2 and
5 and SnCl4 were mixed before the addition of a
solution of the aldehyde in order to promote the ligand
exchange reaction leading to the corresponding allyl-
trichlorostannanes 3 and 6, respectively (Scheme 1).4,5
Aldehyde (S)-8 was prepared in excellent yields from
(3S)-1,3-butanediol 7 by a sequence that involved full
protection as its bis-TBS ether, selective removal of the
primary TBS protecting group and Swern oxidation
under standard conditions (Scheme 2).
The 1,2-syn and 1,2-anti aldehydes 11 and 13 were
easily prepared from oxazolidinone 9 by using syn12
and anti13 selective aldol reactions, respectively, as the
key steps (Scheme 3).
Scheme 4.
These substrates have been selected to be representative
of the complex fragments that might be coupled in
polyacetate and polypropionate-derived aldol-type
reactions.14 For these aldehydes, internal chelation is
presumably prevented by use of bulky silyl protecting
groups since, with few exceptions, silyl ethers are recog-
nized generally for their poor coordinating and chelat-
ing abilities.15
In order to check the facial selectivity of aldehyde (S)-8
we reacted it with achiral allyltrichlorostannane 3
(Scheme 4).16
Achiral allyltrichlorostannane 3 reacted with chiral b-
alkoxy aldehyde (S)-8 in CH2Cl2 at −78°C to give the
corresponding 1,3-anti product 14 as the major product
in good yield and with 78:22 diastereoselectivity favor-
ing the 1,3-anti isomer (Scheme 4).17 The stereoinduc-
tion observed in this reaction indicates that the intrinsic
facial bias imposed by the resident b-OTBS substituent
results in preferential formation of the 1,3-anti
diastereomer, with a preference for aldehyde Si-face
attack.16,17
Scheme 5.
In the proposed transition state conformation (A),
steric interactions are expected to be reduced when the
b-Me substituent in this aldehyde is placed anti to the
Ca-CꢀO bond (Scheme 4).16 In this conformation we
might also expect minimization of destabilizing dipole
interactions since the b-OTBS group and CꢀO are
oriented in opposite directions. A chair-like transition
state (A) with the alkyl group of the aldehyde in an
equatorial position and with attack to the Si-face of the
aldehyde, explains the observed sense of induction.
The relative stereochemistry for homoallylic alcohol 14
was ascertained on the basis of the 13C NMR analysis
of the corresponding acetonide 15 (Scheme 4).18 Treat-
ment of 14 with TBAF at rt effected smooth deprotec-
tion of the silyl ether to provide the corresponding diol
that was treated with Me2C(OMe)2 in the presence of
p-TsOH to give acetonide 15 in 73% overall yield. 13C
NMR resonances at 25.0, 25.2 and 100.1 are character-
istic of a 1,3-anti-acetonide.19
Under the same conditions allyltrichlorostannane (S)-6
reacted with aldehyde (S)-8 to give 1,3-anti-1,4-syn
product 16 as the major product (85:15 diastereoselec-
tivity) (Scheme 5).17 As we know from previous
Scheme 3.

L. C. Dias et al. / Tetrahedron Letters 44 (2003) 6861–6866
6863
work,3–9 the facial bias of this chiral allyltrichlorostan-
nane is dominated by the a-methyl stereocenter and
tends to give the 1,4-syn isomer with Si-face attack.
This is an example of a matched reaction.
Addition of the enantiomeric allyltrichlorostannane
(R)-6 to aldehyde (S)-8 led to a 67:33 mixture favoring
the 1,4-syn-1,3-syn product 18, in a mismatched case
(Scheme 5).17
The stereoselectivity of these reactions is consistent with
an intermediate allyltin trichloride which is stabilized
by tin–oxygen interaction (boat-like), and which then
reacts with the aldehyde via a chair-like six-membered
ring transition state (B) in which the aldehyde
approaches the boat-like complex opposite to the
methyl group (Scheme 5). A boat-like arrangement is
proposed, as it avoids steric interactions between the
aldehyde substituents and the axial methyl group a to
the double bond in the chair structure. The preference
of the alkyl group of the aldehyde to adopt an equato-
rial position controls the aldehyde facial selectivity,
resulting in the favored 1,4-syn stereochemistry in the
adduct.
Scheme 7.
The relative stereochemistry for homoallylic alcohols 16
and 18 was unambiguously established on the basis of
the 13C NMR analysis of their respective acetonides 17
and 19 (Scheme 5).18,19 Treatment of 16 and 18 with
TBAF at rt followed by treatment of the corresponding
diols under acidic conditions with 2,2-dimethoxy-
propane gave acetonides 17 (99%) and 19 (83%), respec-
tively. Observed 13C NMR resonances at 25.0, 25.2 and
100.1 for 17 are characteristic of an anti-1,3-diol-ace-
tonide and 13C NMR resonances at 19.9, 30.4 and 98.4
for 19 are consistent with a syn-1,3-diol-acetonide.19
This example shows that a 1,2-syn aldehyde has a
preference to give the product with Felkin addition as
well as 1,3-syn addition.20 In the presence of an a-
methyl stereocenter, 1,3-asymmetric induction imposes
an intrinsic facial bias on the carbonyl that results in
the formation of the 1,3-syn-dioxygen relationship.
In the proposed transition state conformation (C),
steric interactions are expected to be reduced when the
b-OTBS substituent is placed anti to the Ca-CꢀO bond
(Scheme 6).16 In this conformation we might also expect
minimization of destabilizing dipole interactions since
the b-OTBS group and CꢀO are oriented in opposite
directions. A chair-like transition state (C) with attack
to the Re-face of the aldehyde (Felkin addition),
explains the observed induction direction.20
We next examined the stereochemical impact of both a
and b-aldehyde substituents with chiral-syn and anti
disubstituted a-methyl-b-alkoxy aldehydes.
Achiral allyltrichlorostannane 3 reacted with chiral syn-
a,b-disubstituted aldehyde 11 to give the corresponding
1,2-syn-1,3-syn product 20 in 92% yield and with 96:04
diastereoselectivity (Scheme 6).17
The stereochemical assignment of compound 20 was
determined by analysis of the 13C NMR spectra of
acetonide 21. 13C NMR resonances at 19.7, 30.1 and
98.7 are characteristic of a syn-1,3-diol-acetonide.18,19
The reaction of chiral allyltrichlorostannane (R)-6 with
aldehyde 11 gives homoallylic alcohol 22 (all-syn
product) as the major isomer (Felkin addition, matched
case)17,20 (Scheme 7).
The stereoselectivity of this reaction is consistent with a
chair-like six-membered ring transition state (D) in
which the aldehyde approaches the boat-like allyltin
complex opposite to the methyl group (Scheme 7). The
preference of the alkyl group of the aldehyde to adopt
an equatorial position controls the aldehyde facial
selectivity, resulting in the favored 1,4-syn stereochem-
istry in the adduct.
Allyltrichlorostannane (S)-6 reacted with aldehyde 11
to give a 70:30 ratio favoring isomer 24, in a mis-
Scheme 6.

6864
L. C. Dias et al. / Tetrahedron Letters 44 (2003) 6861–6866
matched case (Scheme 7).17,20 In this latter case, the
a-methyl stereocenter in allyltrichlorostannane (propen-
sity for 1,4-syn addition) exerts a dominant influence on
aldehyde facial selectivity, by overriding the intrinsic
bias imposed by the a and b stereocenters in the
aldehyde, to give the 1,3-syn product.
The stereochemical assignment of compounds 22 and
24 was again determined by 13C NMR analysis of
acetonides 23 and 25, respectively (Scheme 7). 13C
NMR resonances at 19.8, 30.1 and 98.7 for 23 are
characteristic of a syn-1,3-diol-acetonide and 13C NMR
resonances at 24.8, 25.7 and 100.5 observed for 25 are
consistent with an anti-1,3-diol-acetonide.18,19
Before starting the study described in Scheme 8, we
expected that under conditions that preclude internal
chelation, the carbonyl facial bias of anti-disubstituted
aldehyde 13 should be highly predictable, since the
factors which favors both 1,2- and 1,3-asymmetric
induction mutually reinforce nucleophilic addition to
give 1,2-syn-1,3-anti diastereomer.16 We have observed
that this is not the case under the reaction conditions
described here.
Scheme 9.
Achiral allyltrichlorostannane 3 additions to chiral
anti-a,b-disubstituted aldehyde 13 gave the correspond-
ing 1,2-syn-1,3-anti-product 26 as the major product in
good yields, although with only 55:45 diastereoselectiv-
ity (Scheme 8).17,20
Under the same conditions described before allyl-
trichlorostannane (S)-6 reacted with aldehyde 13 to
give 1,2-syn-1,3-anti isomer 29 with 92:08 diastereo-
selectivity (Scheme 9).17,20
This example shows that an anti aldehyde has no facial
preference under these conditions, since the Felkin
addition to give 1,2-syn isomer competes with the b-
alkoxy stereocenter to give the 1,3-syn isomer. We
believe that in this case, the corresponding transition
states should be very similar in energy, with conformer
E (which gives the 1,2-syn isomer) being destabilized by
the gauche interaction iPr/Me while conformer F (which
gives the 1,2-anti isomer) being destabilized by
the iPr/CꢀO and Me/OTBS gauche interactions
(Scheme 8).
The reaction of allyltrichlorostannane (R)-6 with alde-
hyde 13 gave homoallylic alcohol 31 as the major
isomer in 88:12 diastereoselectivity (anti-Felkin addi-
tion, partially matched case).17,20
The results described in Scheme 9 can be rationalized
with dominant acyclic 1,4-asymmetric induction from
the chiral allyltrichlorostannane. These are examples of
partially matched reactions, with the chiral allyl-
trichlorostannanes (S)-6 and (R)-8 being responsible
for the control of the observed diastereoselectivities,
through transition states analogous to G and H, respec-
tively (Scheme 9). This reaction with 1,2-anti b-OTBS
aldehydes is characterized by poor levels of diastereo-
selectivity only when an achiral allyltrichlorostannane is
used. This attenuated 1,3-anti selectivity for 1,2-anti
aldehydes with the TBS protecting group appears to be
general, as similar trends were observed for titanium
and boron aldol reaction variants.16 One might project
that the transition states of these reactions exhibit less
charge separation than the aldol processes, and are
accordingly less subject to the electrostatic influence of
the b-OTBS function.
As before, the relative stereochemistry for compounds
29 and 31 was determined by analysis of the 13C NMR
of the corresponding acetonides 29 and 31, respectively
(Scheme 9). 13C NMR resonances at 23.8, 25.8 and
100.1 observed for 30 are characteristic of an anti-1,3-
diol-acetonide and 13C NMR resonances at 19.7, 30.0
and 98.8 for 32 are characteristic of a syn-1,3-diol-
acetonide.18,19
Scheme 8.

L. C. Dias et al. / Tetrahedron Letters 44 (2003) 6861–6866
6865
The examples presented in this work show that the
levels of p-facial selection are dependent on the abso-
lute stereochemistries of the aldehydes as well as of the
allyltrichlorostannane.
5. Dias, L. C.; Meira, P. R. R.; Ferreira, E. Org. Lett. 1999,
1, 1335.
6. Dias, L. C.; Meira, P. R. R. Synlett 2000, 37.
7. Dias, L. C.; Ferreira, E. Tetrahedron Lett. 2001, 42, 7159.
8. Dias, L. C.; Ferreira, A. A.; Diaz, G. Synlett 2002, 1845.
9. Dias, L. C.; Diaz, G.; Ferreira, A. A.; Meira, P. R. R.;
Ferreira, E. Synthesis 2003, 4, 603.
10. We have recently described a very efficient and syntheti-
cally useful 1,4-anti-1,5-anti boron-mediated aldol reac-
tion of chiral a-methyl-b-alkoxy methyl ketone with
achiral aldehydes: Dias, L. C.; Bau´, R. Z.; de Sousa, M.
A.; Zukerman-Schpector, J. Org. Lett. 2002, 4, 4325.
11. (a) Bunnelle, W. H.; Narayanan, B. A. Tetrahedron Lett.
1987, 28, 6261; (b) Mickelson, T. J.; Koviach, J. L.;
Forsyth, C. J. J. Org. Chem. 1996, 61, 9617.
The results from these experiments suggest that the
stereochemical relationships between the a and b alde-
hyde substituents may confer either a reinforcing
(matched) or opposing (mismatched) facial bias on the
carbonyl moiety. In this complex scenario, the chiral
allyltrichlorostannane may adopt either a reinforcing or
nonreinforcing relationship. One possible reason for
this result could be attributed to the involvement of
energetically similar chair and twist-boat pericyclic
transition states which lead to diastereomeric product
formation. Another possibility to consider in these reac-
tions is that nonbonded interactions between the allyl-
trichlorostannane and aldehyde a substituents may not
be significant in pericyclic transition states leading to
either Felkin or anti-Felkin addition products.20 We
believe that this chemistry is truly significant in the
context of acyclic diastereoselection and will prove to
be useful in the synthesis of more complex molecules
like polyacetate and polypropionate-derived natural
products.21,22
12. (a) Gage, J. R.; Evans, D. A. Org. Synth. 1990, 68, 83; (b)
Evans, D. A.; Bartroli, J.; Shih, T. L. J. Am. Chem. Soc.
1981, 103, 2127; (c) Evans, D. A.; Taber, T. R. Tetra-
hedron Lett. 1980, 21, 4675.
13. Evans, D. A.; Downey, C. W.; Shaw, J. T.; Tedrow, J. S.
Org. Lett. 2002, 4, 1127.
14. Since the diastereoselectivity of the reactions of these
aldehydes with allyltrichlorostannanes 3 and 6 depends
on their enantiomeric purity, crude aldehydes were used
in all of the studies described in the text.
15. Schreiber, S. L.; Shambayati, S.; Blake, J. F.; Wierschke,
S. G.; Jorgensen, W. L. J. Am. Chem. Soc. 1990, 112,
697.
16. Evans, D. A.; Dart, M. J.; Duffy, J. L.; Yang, M. G. J.
Am. Chem. Soc. 1996, 118, 4322.
Acknowledgements
17. (a) The ratios were determined by 1H and 13C NMR
spectroscopic analysis of the purified product mixture; (b)
The syn and anti-products could not be separated and
were characterized as mixtures. We have been able to
separate both syn and anti diols originating from homoal-
lylic alcohols 18 and 24 and they were characterized
individually; (c) All of the percentage values represent
data obtained from three individual trials.
We are grateful to FAEP-UNICAMP, FAPESP and
CNPq (Brazil) for financial support. We also thank
Professor Carol H. Collins, from IQ-UNICAMP, for
helpful suggestions about English grammar and style.
References
18. Having confirmed the relative (syn or anti ) relationship
between allylsilane derived stereogenic centers, the abso-
lute stereochemistry of the newly formed hydroxyl sub-
stituent was determined by ascertaining its relationship to
the stereocenter originating from the aldehydes, which
are of known configuration.
19. (a) Rychnovsky, S. D.; Skalitzky, D. J. Tetrahedron Lett.
1990, 31, 945; (b) Evans, D. A.; Rieger, D. L.; Gage, J.
R. Tetrahedron Lett. 1990, 31, 7099; (c) Rychnovsky, S.
D.; Rogers, B. N.; Richardson, T. I. Acc. Chem. Res.
1998, 31, 9.
20. (a) Che´rest, M.; Felkin, H.; Prudent, N. Tetrahedron
Lett. 1968, 18, 2199; (b) Anh, N. T.; Eisenstein, O. Nouv.
J. Chem. 1977, 1, 61; (c) We use the ‘Felkin’ descriptor to
refer to the diastereomer predicted by the Felkin–Ahn
paradigm. The ‘anti-Felkin’ descriptor refers to
diastereomers not predicted by this transition state
model.
1. (a) Fleming, I.; Barbero, A.; Walter, D. Chem. Rev. 1997,
97, 2063; (b) Nishigaichi, Y.; Takuwa, A.; Naruta, Y.;
Maruyama, K. Tetrahedron 1993, 49, 7395; (c) Panek, J.
S.; Xu, F.; Rondon, A. C. J. Am. Chem. Soc. 1998, 120,
4113; (d) Zhu, B.; Panek, J. S. Eur. J. Org. Chem. 2001,
9, 1701; (e) Huang, H. B.; Spande, T. F.; Panek, J. S. J.
Am. Chem. Soc. 2003, 125, 626; (f) Keck, G. E.; Abbott,
D. E. Tetrahedron Lett. 1984, 25, 1883.
2. (a) Trost, B. M.; Urabe, H. J. Org. Chem. 1990, 55, 3982;
(b) Nishigaishi, Y.; Takuwa, A.; Jodai, A. Tetrahedron
Lett. 1991, 32, 2383; (c) Almendros, P.; Gruttadauria,
M.; Helliwell, M.; Thomas, E. J. J. Chem. Soc., Perkin
Trans. 1 1997, 2549; (d) Deka, D. C.; Helliwell, M.;
Thomas, E. J. Tetrahedron 2001, 57, 10017; (e) Martin,
N.; Thomas, E. J. Tetrahedron Lett. 2001, 42, 8373; (f)
Kumar, P.; Thomas, E. J.; Tray, D. R. J. Braz. Chem.
Soc. 2001, 12, 623.
3. (a) Dias, L. C.; Giacomini, R. J. Braz. Chem. Soc. 1998,
9, 357; (b) Evans, D. A.; Coleman, P. J.; Dias, L. C.
Angew Chem., Int. Ed. Engl. 1997, 36, 2738; (c) Evans, D.
A.; Trotter, B. W.; Cote, B.; Coleman, P. J.; Dias, L. C.;
Tyler, A. N. Angew Chem., Int. Ed. Engl. 1997, 36, 2744.
4. Dias, L. C.; Giacomini, R. Tetrahedron Lett. 1998, 39,
5343.
21. All new compounds were isolated as chromatographically
pure materials and exhibited acceptable 1H, 13C NMR,
IR, MS, and HRMS spectral data.
22. General procedure for allyltrichlorostannane coupling reac-
tions: To a solution of 2.5 mmol of allylsilane 6 in 7 mL
of dry CH2Cl2 at −78°C was added 2.5 mmol of SnCl4.
The resulting solution was stirred at −78°C for 30 min

6866
L. C. Dias et al. / Tetrahedron Letters 44 (2003) 6861–6866
when 2.7 mmol of aldehyde in 2 mL of CH2Cl2 was
extracted with CH2Cl2 (2×5 mL). The combined organic
layer was dried (MgSO4), filtered, and concentrated in
vacuo. Purification by flash chromatography on silica gel
(30% EtOAc/hexanes) gave the corresponding homoal-
lylic alcohols.
added. This mixture was stirred at −78°C for 1 h and
quenched by the slow addition of 0.2 mL of Et3N,
followed by 10 mL of saturated NH4Cl solution. The
layers were separated and the aqueous layer was