Evaluation of a sparteine-like diamine for asymmetric synthesis

Article abstract of DOI:10.1039/b103220h

Evaluation of a sparteine-like diamine indicates that only the ABC rings of sparteine are required for high enantioselectivity in the lithiation-substitution of N-Boc pyrrolidine.

Full text of DOI:10.1039/b103220h

Evaluation of a sparteine-like diamine for asymmetric synthesis
a
a
a
b
Justin R. Harrison, Peter O'Brien,* David W. Porter and Neil M. Smith
a
Department of Chemistry, University of York, Heslington, York, UK YO10 5DD.
E-mail: paob1@york.ac.uk
GlaxoSmithKline, New Frontiers Science Park (North), Third Avenue, Harlow, Essex, UK CM19 5AW
b
Received (in Cambridge, UK) 10th April 2001, Accepted 21st May 2001
First published as an Advance Article on the web 13th June 2001
Evaluation of a sparteine-like diamine indicates that only
the ABC rings of sparteine are required for high enantiose-
lectivity in the lithiation–substitution of N-Boc pyrrolidine.
Sparteine 1 is a naturally occurring alkaloid extracted from
plants such as Scotch Broom. It is commercially available and
has been widely used as a chiral diamine ligand in asymmetric
1
synthesis over the last 30 years. For example, asymmetric
lithiation–electrophilic quench using the combination of spar-
teine and alkyllithiums on a wide range of substrates occurs
2
routinely with > 90% enantioselectivity. The groups of Hoppe
3
and Beak have led the way with pioneering contributions in the
applications of sparteine in synthesis. More recently, Hoppe et
al. and Wiberg and Bailey5 have carried out theoretical
calculations of transition state energies aimed at elucidating
how sparteine exerts such high levels of enantiodifferentia-
tion.
4
Scheme 1 Reagents and conditions: i, MeNH , (CH O) , AcOH, MeOH,
2
2
n
2 4 2
reflux, 16 h; ii, N H , H O, KOH, diethylene glycol, reflux, 2 h; iii, acetone,
evaporation over 64 h.†
as 5. Since diamine 2 is structurally similar to sparteine, we
speculated that enantiomerically pure 5 could be used for
resolution.
Thus, Toda's method was employed to synthesise alcohol
(
2
R)-5 of 98% ee (by chiral HPLC on an Astec Cyclobond I
000-RSP column) and it was used in turn to partially resolve
diamine 2. In this way, we obtained a 23% yield of
enantiomerically enriched diamine 2† ([a] 215.7 (c 0.5 in
D
One of the main limitations of using sparteine in synthesis is
that it is only commercially available in one enantiomeric form.
Attempts to find other chiral diamine ligands capable of
matching the enantioselectivity of sparteine have been moder-
ately successful.6 With the long term aim of developing a
ligand that will function as the enantiomer of sparteine, we have
investigated whether diamine 2, which lacks the D-ring of
sparteine as well as one of the chiral centres, mimicks sparteine
sufficiently to give high enantioselectivity. Structural compar-
isons of diamines 1 and 2 complexed to lithium together with a
EtOH); ~ 60% ee by chiral shift NMR). This resolution was
reproducible in the 50–60% ee range and after repeating it a few
times, we obtained a sufficient quantity of diamine 2 of ~ 55%
ee. Since alcohol (R)-5 had been obtained via crystal formation
with sparteine 1, it seemed likely that resolved diamine 2
(isolated via crystal formation with (R)-5) would have the same
absolute stereochemistry as the ABC rings of sparteine.
With diamine 2 of ~ 55% ee in hand, we elected to directly
compare it with sparteine 1 using lithiation of N-Boc pyrrolidine
,7
6,12
6 and subsequent trapping with Me SiCl.
3
The results are
5
recent calculated transition state for reaction suggested that the
presented in Table 1. Using sparteine 1, we obtained a 73%
isolated yield of silylated pyrrolidine (S)-7 of 95% ee (by chiral
GC). Under the same conditions, use of diamine 2 of ~ 55% ee
gave an unoptimised 41% yield of silylated pyrrolidine (S)-7 of
53% ee (by chiral GC).‡ The sense of asymmetric induction was
D-ring of sparteine was not a key element in the enantiodiscri-
minating process. In this communication, we provide experi-
mental evidence in support of this conjecture.
8
Although racemic diamine 2 is a known compound and has
found recent application in the functionalisation of terminal
9
epoxides, there have been no reports on the preparation of
Table 1 Lithiation–substitution of N-Boc pyrrolidine using diamines 1 and
2
enantiomerically enriched diamines like 2. An approach from
amino acids investigated in our laboratory10 was unsuccessful
due to unavoidable racemisation in one of the steps. Thus, we
resorted to resolution as a means of preparing non-racemic
diamine 2 as outlined in Scheme 1.
Racemic diamine 2 was prepared using a published route:8
double Mannich reaction of ketone 3 gave a single diastereoi-
somer of 4 (58% yield) which was converted into the required
diamine 2 (68% yield) using Wolff–Kishner reduction. Un-
fortunately, we were unable to develop a resolution protocol for
racemic 2 using commercial chiral acids (e.g. tartaric acid and
derivatives, malic acid and camphorsulfonic acid). However,
we had more success with resolution by inclusion complex
formation using a method previously described by Toda et al.11
Toda had discovered that it was possible to use naturally
ocurring sparteine 1 to resolve racemic acetylinic alcohols such
Yield of
Diamine
(S)-7 (%)a
ee (%)b
Sparteine 1
73
41
95
53
(1R,2R,9R)-2 ( ~ 55% ee)
a
Isolated yield after column chromatography; b Enantiomeric excess
determinined by chiral GC on a Chiraldex G-PN 20 m 3 0.25 mm id (g-
cyclodextrin, propionyl derivative in the 3-position) column.
1202
Chem. Commun., 2001, 1202–1203
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b103220h

View Article Online
the same in both reactions indicating that resolved diamine 2
does indeed possess the same absolute stereochemistry as the
ABC rings of sparteine. More importantly, however, the results
presented in Table 1 provide the first experimental evidence that
only the ABC rings of sparteine 1 are required for high
enantioselectivity in the lithiation of N-Boc pyrrolidine 6.
In summary, our results indicate that diamine 2 exhibits
essentially the same level of enantioselectivity as sparteine 1 in
the lithiation of N-Boc pyrrolidine 6. Thus, we conclude that the
sparteine D-ring is superfluous and is not a prerequisite for high
enantioselectivity.
D
2,2,2-trifluoro-1-(9-anthryl)ethanol) as a colourless oil, [a] 215.7 (c 0.5 in
EtOH). Treatment of the petrol washings in the same way as described
above for the crystals gave diamine (1S,2S,9S)-2 (503 mg, 68%; ~ 30% ee
1
by H NMR spectroscopy in the presence of (R)-2,2,2-trifluoro-1-(9-
D
anthryl)ethanol) as a colourless oil, [a] +6.1 (c 0.6 in EtOH).
‡
In theory, silylated pyrrolidine (S)-7 of > 55% ee could have been
generated if a non-linear effect was occurring. Our results clearly indicate
the absence of a non-linear effect and are in line with the calculated
5
transition state for reaction
complex.
which invokes deprotonation by a monomeric
1 For a review, see: D. Hoppe and T. Hense, Angew. Chem., Int. Ed. Engl.,
1997, 36, 2282.
2 For recent developments, see: A. Deiters, R. Frölich and D. Hoppe,
Angew. Chem., Int. Ed., 2000, 39, 2105.
We thank the EPSRC and GlaxoSmithKline for the award of
an Industrial CASE award (to D. W. P.), The Leverhulme Trust
for the award of a fellowship (to J. R. H.) and Dr D. B. Matthews
for chiral HPLC and GC analyses.
3
For recent developments, see: T. A. Johnson, M. D. Curtis and P. Beak,
J. Am. Chem. Soc., 2001, 123, 1004.
4
E.-U. Würthwein, K. Behrens and D. Hoppe, Chem. Eur. J., 1999, 5,
Notes and references
†
3
459.
Resolution of diamine rac-2: A solution of freshly distilled diamine rac-28
5 K. B. Wiberg and W. F. Bailey, Angew. Chem., Int. Ed., 2000, 39, 2127;
K. B. Wiberg and W. F. Bailey, Tetrahedron Lett., 2000, 41, 9365.
6 D. J. Gallagher, S. Wu, N. A. Nikolic and P. Beak, J. Org. Chem., 1995,
60, 8148.
3
(742 mg, 3.8 mmol) in acetone (3 cm ) was added dropwise to a solution of
alcohol (R)-5 (929 mg, 3.8 mmol; 98% ee by chiral HPLC on an Astec
Cyclobond I 2000-RSP column) in acetone (3 cm ) at rt. The solvent was
3
allowed to evaporate slowly by standing at rt for 64 h and pale yellow
7 X. Li, L. B. Schenkel and M. C. Kozlowski, Org. Lett., 2000, 2, 875.
8 P. Scheiber and P. Nemes, Liebigs Ann. Chem., 1994, 1033.
9 D. M. Hodgson and S. L. M. Norsikian, Org. Lett., 2001, 3, 461.
10 J. R. Harrison, P. O'Brien, D. W. Porter and N. M. Smith, J. Chem. Soc.,
Perkin Trans. 1, 1999, 3623.
11 F. Toda, K. Tanaka, H. Ueda and T. Oshima, Chem. Commun., 1983,
743.
12 P. Beak, S. T. Kerrick, S. Wu and J. Chu, J. Am. Chem. Soc., 1994, 116,
3231.
3
crystals formed. Petrol (5 cm ) was added and the crystals were collected by
3
filtration and washed well with petrol (3 3 5 cm ). The crystals were
3
3
2
dissolved in Et O (15 cm ) and 2 M HCl(aq) (10 cm ), the layers were
separated and the organic layer was extracted with 2 M HCl(aq) (2 3 10
3
cm ). Then, 20% NaOH(aq) was added to the combined aqueous layers until
3
pH 9 and the solution was extracted with Et
2
O (3 3 10 cm ), dried (K
2 3
CO )
and evaporated under reduced pressure to give diamine (1R,2R,9R)-2 (171
mg, 23%; ~ 60% ee by H NMR spectroscopy in the presence of (R)-
1
Chem. Commun., 2001, 1202–1203
1203