A stereoselective synthesis of anti-gamma,delta-alkynyl- and -alkenyl-beta-hydroxy-alpha-amino esters from tin(II) enolates of glycinate.

Article abstract of DOI:10.1039/b306291k

Condensations between the tin(II) enolate 11 of ethyl N-tosylglycinate and conjugated ynals 12 and ynones 14 are highly diastereoselective, in favour of the anti-isomers 13 and 15; similar reactions of enals and enones 17 show lower but still useful levels of anti-stereoselectivity.

Full text of DOI:10.1039/b306291k

A stereoselective synthesis of anti-g,d-alkynyl- and
-alkenyl-b-hydroxy-a-amino esters from tin(II) enolates of glycinate†
Jonathan J. Gridley, Michael P. Coogan, David W. Knight,* K. M. Abdul Malik, Christopher M.
Sharland, Jirada Singkhonrat and Siân Williams
Chemistry Department, Cardiff University, P.O. Box 912, Cardiff, UK CF10 3TB.
E-mail: knightdw@cf.ac.uk; Fax: 44(0) 2920 874210; Tel: 44(0) 2920 874210
Received (in Cambridge, UK) 3rd June 2003, Accepted 18th July 2003
First published as an Advance Article on the web 17th September 2003
Condensations between the tin(II) enolate 11 of ethyl N-
tosylglycinate and conjugated ynals 12 and ynones 14 are
highly diastereoselective, in favour of the anti-isomers 13
and 15; similar reactions of enals and enones 17 show lower
but still useful levels of anti-stereoselectivity.
especially at the new b-hydroxyl centre. For example, such a
reaction between N-tosyl alaninate 5 and aldehyde 6 gave
essentially only the expected products 7 and 8, in an 80 : 20 ratio
(Scheme 2). Unfortunately, similar condensations involving
glycinates gave products [e.g. 9] which were epimeric at the a-
centre. This was explained by post-condensation epimerisation,
rather than a lack of stereocontrol during the condensation, or its
reversibility.4
Despite a plethora of recent advances in the synthesis of non-
natural and more highly functionalised a-amino acid deriva-
tives,1 there remains a need for the definition of practical and
general approaches to such compounds. Not only are these of
considerable importance as components of novel pharmaceuti-
cals, but such densely functionalised compounds can find many
applications in general synthesis. Doubtless with these ideas in
mind, the Kazmaier group have carried out extensive studies of
condensations between aldehydes and dianionic species derived
from simple N-protected a-amino esters. Variations in both the
amine protecting group and the counter cations, unsurprisingly,
were found to have profound effects on both the efficiency and
stereoselectivity of such condensations. In general, these favour
formation of an anti arrangement between the amino group and
the new hydroxyl group. Of a number of alternatives, one of the
best turned out to be a combination of the dilithio dianion 1 of
N-Z-valine tert-butyl ester with 2 equiv. of TiCl(OPri)3 which
provided, almost exclusively, the anti-diastereoisomer 2 when
isobutanal was the electrophile, in 60% isolated yield (Scheme
1).2 These conditions were less successful with similar alanine
derivatives but it was subsequently discovered that replacement
of the titanium alkoxides with 2.5 equiv. tin(II) chloride restored
this excellent level of stereoselection, as well as giving
improved yields.3 In addition, both N-tosyl and benzyl esters of
alanine could be used to obtain the anti-diastereoisomers 3, but
this generality was restricted to condensations with aliphatic
aldehydes.
The structures provided by this methodology were ideal for
projected developments of our 5-endo cyclisation methodology
aimed at the construction of pyrrolidines and pyrroles,5 given
that it could be applied successfully to conjugated ynals and
enals, types of electrophiles not examined by Kazmaier. Herein,
we report that such condensations are indeed successful and
also show synthetically very useful levels of stereoselection.
Our initial requirements were for a series of alkynyl-b-
hydroxy-a-amino ester derivatives 10. In view of the foregoing
chemistry, it seemed possible that these could be obtained from
the corresponding ynals and the tin(II) enolate of a glycinate.
Such an intermediate has been formulated as structure 11; the
requirement of a second equivalent of SnCl2 is ascribed to a role
in electrophile activation.2–4 The required ynals 12 were
obtained using the excellent method developed by Journet et al.
by 1-alkyne formylation (BuLi, THF; DMF; inverse quench) in
around 90% yields.6 We were pleased to find that the desired
condensations proceeded smoothly in THF, following mixing
the reactants at 270 °C and warming to ambient temperature.2–4
Isolated yields of the hoped-for adducts 13 were between 60 and
83%. Further, these were isolated as largely the anti-isomers, as
shown (Scheme 3). This was a surprise, in view of the results
To achieve similarly high levels of stereoselection (98 : 2)
with aromatic aldehydes, it was found essential to use a
combination of SnCl2 and N-tosyl protection to obtain the anti-
isomers 4. The more easily removed N-SES group could be used
equally effectively.
A combination of the tin(II) enolate derived from N-tosyl
alanine tert-butyl ester and a chiral a-hydroxyaldehyde deriva-
tive resulted in excellent levels of asymmetric induction,
Scheme 2
Scheme 1
† Electronic supplementary information (ESI) available: crystal data. See
http://www.rsc.org/suppdata/cc/b3/b306291k/
Scheme 3
2550
CHEM. COMMUN., 2003, 2550–2551
This journal is © The Royal Society of Chemistry 2003

obtained from previous condensations with glycinates 11 (cf.
9).4 The anti : syn ratios were determined by integration of the
1H NMR spectra of the crude products. These data also revealed
that the remaining material balance from the condensations was
largely N-tosyl glycinate, probably formed in most cases by
competing deprotonation of the ynals 12. In many cases, the
major anti-isomers 13 could be separated by fractional
crystallisation in 60–70% yields. Subsequent chemistry re-
vealed that the adducts 13 were relatively prone to a-
epimerization and so we were somewhat nervous about
structural determination by, for example, manipulation through
to a cyclic structure. Fortunately, the anti-phenyl derivative [13;
R = Ph] provided crystals suitable for X-ray analysis,7 which
revealed the anti-stereoselection.
Thus encouraged, we carried out similar condensations using
conjugated ynones 14 as the electrophiles. These too were
successful and showed very useful levels of, again, anti-
stereoselectivity (Scheme 4). 3-Butyn-2-one also condensed
well with enolate 11, without the need to protect the potentially
labile alkynyl proton, to give adduct 16 in similar yield and
selectivity. Again, the pure anti-isomers could be separated by
careful crystallization in 60–70% yields and the relative
stereochemistry confirmed by X-ray analysis [15; R = Pri].7
well with Kazmaier’s results and with his explanation based on
a Felkin–Anh model.2–4 However, these results are not
consistent with his conclusion that the mixture of adducts 9,
obtained from glycinate 11, was due to epimerization.
Uncertainties regarding the true nature of glycinate enolate
11 preclude much speculation. Indeed, the picture appears more
complicated, because the condensations with ynones 14 gave
adducts 15 with the same anti-selection, despite the larger
substituent now being the “R” group rather than the alkyne
residue, in contrast to the related ynals 12, especially when “R”
is branched [14; R = Pri]. Currently, one suggestion is that, if
structure 11 is correct, the anti-stereoselection requires the
alkyne group to be positioned axially, in a typical chair-like
transition state 19. Could this be due to donation from the
alkyne bond into vacant tin orbitals? The same effect could be
responsible for the similar but lower anti-stereoselection of
condensations with enals and enones 17. Of relevance is our
observation, consistent with Kazmaier’s, that condensations
between enolate 11 and saturated aldehydes show almost no
stereoselection. Hence, lack of complexation and not later
epimerization may be responsible for this. Further studies aimed
at shedding light on this along with synthetic applications of the
various products reported herein are in progress.
We are grateful to the Government of Thailand, Millennium
Pharmaceuticals and the EPSRC for financial support and the
EPSRC Mass Spectrometry Service, University College, Swan-
sea for high resolution mass spectrometric data
Scheme 4
Notes and references
1 For recent contributions, see A. E. Taggi, A. M. Hafez and T. Lectka, Acc.
Chem. Res., 2003, 36, 10; V. A. Soloshonok, Curr. Org. Chem., 2002, 6,
341.
Finally, we briefly examined the outcome of such condensa-
tions when applied to enals and enones 17. In general, these
showed lower, but still useful, levels of anti-stereoselection in
the expected products 18. These results are collected in Scheme
5.8
The high anti-stereoselectivity of the condensations between
the glycinate enolate 11 and ynals 12 (Scheme 3) certainly fits
2 R. Grandel, U. Kazmaier and B. Nuber, Annalen, 1996, 1143.
3 R. Grandel and U. Kazmaier, Eur. J. Org. Chem., 1998, 407.
4 U. Kazmaier and R. Grandel, Eur. J. Org. Chem., 1998, 1833. See also R.
Grandel, U. Kazmaier and F. Rominer, J. Org. Chem., 1998, 63, 4524.
For the use of related zinc enolates in Pd-catalysed allylations, see U.
Kazmaier and F. L. Zumpe, Angew. Chem., Int. Ed., 1999, 38, 1468; U.
Kazmaier and F. L. Zumpe, Eur. J. Org. Chem., 2001, 4067; U.
Kazmaier, S. Maier and F. L. Zumpe, Synlett., 2000, 1523.
5 D. W. Knight, A. L. Redfern and J. Gilmore, J. Chem. Soc., Perkin Trans.
1, 2001, 2874; 2002, 622.
6 M. Journet, D. W. Cai, L. M. DiMichele and R. D. Larsen, Tetrahedron
Lett., 1998, 39, 6427.
7 See ESI† for full details of these determinations which have been
deposited at the Cambridge Crystallographic Data Centre, CCDC 210386
[13; R = Ph] and 210604 [15; R = Pri].
8 The anti-stereochemistry of the major diastereoisomer was determined
by X-ray analysis of anti-18 [R1–3 = Me] [CCDC 210387, ESI†],
together with comparisons of spectroscopic data.
Scheme 5
CHEM. COMMUN., 2003, 2550–2551
2551