Synthesis of chiral 1,2-diamines by asymmetric lithiation-substitution

Article abstract of DOI:10.1021/ol016818m

Figure presented The imidazolidine (tetrahydroimidazole) 2, prepared in one step from N-iso-propylethylenediamine, was subjected to asymmetric lithiation and substitution using sec-butyllithium, (-)-sparteine and a range of electrophiles. Substituted imid

Full text of DOI:10.1021/ol016818m

ORGANIC
LETTERS
2001
Vol. 3, No. 23
3799-3801
Synthesis of Chiral 1,2-Diamines by
Asymmetric Lithiation−Substitution
,†
Iain Coldham,* Royston C. B. Copley,‡ Thomas F. N. Haxell,† and
Steven Howard‡
School of Chemistry, UniVersity of Exeter, Exeter EX4 4QD, U.K. and
GlaxoSmithKline, New Frontiers Science Park (North), Third AVenue,
Harlow, Essex CM19 5AW, U.K.
i.coldham@exeter.ac.uk
Received September 26, 2001
ABSTRACT
The imidazolidine (tetrahydroimidazole) 2, prepared in one step from N-iso-propylethylenediamine, was subjected to asymmetric lithiation and
substitution using sec-butyllithium, (−)-sparteine and a range of electrophiles. Substituted imidazolidines were formed with high optical purity
and could be hydrolyzed under acidic conditions to chiral, substituted ethylenediamines. Kinetic data indicate that the conformation of the
carbonyl group is crucial to the extent of deprotonation, and this has implications for the lithiation of unsymmetrical carbamates and carboxylic
amides.
Asymmetric deprotonation-substitution, typically with a 1:1
complex of sec-butyllithium and (-)-sparteine, has been
exploited widely over the past decade for the formation of
highly enantiomerically enriched products.1 A typical ex-
ample is the asymmetric deprotonation of N-Boc-pyrrolidine
(Boc ) tert-butoxycarbonyl), described by Beak and co-
workers.2 This reaction proceeds with very high selectivity
for abstraction of the pro-S hydrogen atom at C-2 and leads,
after electrophilic quench, to a variety of N-Boc-2-substituted
pyrrolidines with very high optical purity.
Despite the widespread interest and usefulness of this
procedure, there are very few extensions of this asymmetric
lithiation-substitution protocol to other nitrogen-containing
heterocycles.3 Our studies4 of the lithiation-substitution of
symmetrical imidazolidines (tetrahydroimidazoles) led to the
realization that this ring system would be amenable to an
asymmetric deprotonation using an external chiral ligand
such as (-)-sparteine. This paper reports the regio- and
stereoselective lithiation and substitution of unsymmetrical
imidazolidines, leading after acidic hydrolysis to chiral,
substituted 1,2-diamines.5
The imidazolidine 2 was prepared in one pot from N-iso-
propylethylenediamine 1, paraformaldehyde, MgSO4 and
K2CO3, followed by addition of Boc2O (Scheme 1).6 It was
anticipated that on addition of sec-butyllithium to the
Scheme 1. Preparation, Lithiation and Substitution of 2a
† University of Exeter.
‡ GlaxoSmithKline.
(1) (a) Beak, P.; Basu, A.; Gallagher, D. J.; Park, Y. S.; Thayumanavan,
S. Acc. Chem. Res. 1996, 29, 552. (b) Hoppe, D.; Hense, T. Angew. Chem.,
Int. Ed. Engl. 1997, 36, 2282.
(2) Beak, P.; Kerrick, S. T.; Wu, S.; Chu, J. J. Am. Chem. Soc. 1994,
116, 3231.
(3) (a) Beak, P.; Yum, E. K. J. Org. Chem. 1993, 58, 823. (b) Kise, N.;
Urai, T.; Yoshida, J. Tetrahedron: Asymmetry 1998, 9, 3125. (c) Gross,
K. M. B.; Jun, Y. M.; Beak, P. J. Org. Chem. 1997, 62, 7679. (d) Ariffin,
A.; Blake, A. J.; Ebden, M. R.; Li, W.-S.; Simpkins, N. S.; Fox, D. N. A.
J. Chem. Soc., Perkin Trans. 1 1999, 2439.
a Reagents: (i) CH2O, CHCl3, MgSO4, K2CO3, 95%; (ii) s-BuLi,
THF, -78 °C, 1 h then Me3SiCl, 50%; or s-BuLi, Et2O, (-)-
sparteine, -78 °C, 1 h then RX; see Table 1.
10.1021/ol016818m CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/18/2001

imidazolidine 2, complexation to the carbonyl group would
occur, leading to subsequent proton abstraction R- to the
N-Boc group.7 In contrast to the proton abstraction of the
N,N′-bis-Boc analogue, which occurs at C-2, between the
two nitrogen atoms,4 addition of sec-BuLi to the imidazoli-
dine 2 promoted exclusive deprotonation at C-5 to give, after
electrophilic quench, the 5-substituted imidazolidine 3. No
products derived from proton abstraction at C-2 were
isolated.
Table 2. Yields for the Hydrolysis of Imidazolidines 3
entry
R
yield 4 (%)
yield 5 (%)
1
2
3
4
Me3Si
Ph2MeSi
Me
84
80
99
96
90
99
82
87
CH2dCHCH2
Using the standard conditions for asymmetric deprotona-
tion, as described by Beak and co-workers,2 with sec-BuLi
in Et2O and (-)-sparteine, the products 3 were formed with
high optical purity (Table 1). A range of electrophiles could
The mechanism of asymmetric induction follows an
asymmetric lithiation rather than an asymmetric substitution
reaction, since treating the racemic stannane 3, R ) SnBu3
with n-BuLi and (-)-sparteine in Et2O, followed by quench-
ing with Me3SiCl, gave the silane 3, R ) SiMe3 as a racemic
mixture. In addition, the enantiomerically enriched stannane
3 could be transmetalated in the absence of (-)-sparteine
and quenched with Me3SiCl to give the silane 3, R ) SiMe3
with good optical purity, thus illustrating that the chiral
organolithium species derived from the imidazolidine 2 can
maintain its optical purity. These results are consistent with
those using N-Boc-pyrrolidine, and we anticipate therefore
that the major enantiomer of the imidazolidine 3 has the
absolute configuration with the (S)-stereochemistry at C-5
[(R)-stereochemistry by definition for 3, R ) C(OH)Ph2].
This was confirmed by a single-crystal X-ray diffraction
study of the product 3, R ) C(OH)Ph2 (Figure 1).8 Recrys-
Table 1. Asymmetric Lithiation-Substitution of 2
entry
RX
yield 3 (%)a
er
1
2
3
4
5
6
Bu3SnCl
Me3SiCl
Ph2MeSiCl
Ph2CdO
MeI
40 (66)
40 (67)
44 (86)
50 (86)
44 (76)
40 (66)
94:6
93:7
94:6
92:8
92:8
50:50
CH2dCHCH2Br
a Yields in brackets refer to yields based on recovered 2.
be used to trap the organolithium species, with results similar
to those for the related N-Boc-pyrrolidine substrate. Thus,
no evidence for single electron-transfer processes were
apparent using the electrophile benzophenone (entry 4),
which provided the quenched product with high optical
purity; however the electrophile allyl bromide produced
racemic product (entry 6).
The enantiomeric ratios of the products 3 were determined
either by chiral HPLC (Gilson 231 XL column, type AD)
or by NMR of the ring-opened Mosher amide derivatives.
Hydrolysis of the imidazolidine 3 (carried out for R )
SiMe3, SiMePh2, Me, allyl) using malonic acid resulted in
the selective formation of the amino-carbamate 4, without
loss of the N-Boc group (Scheme 2). Subsequent treatment
Scheme 2. Hydrolysis of 3a
Figure 1. X-ray Structure of 3, R ) C(OH)Ph2. Anisotropic atomic
displacement ellipsoids for the non-hydrogen atoms are shown at
the 50% probability level.
a Reagents: (i) CH2(CO2H)2, pyridine, EtOH, heat; (ii) CF3CO2H,
CH2Cl2, rt.
tallization of the product 3, R ) C(OH)Ph2 gave crystals
suitable for X-ray analysis which were determined (HPLC)
(4) Coldham, I.; Judkins, R. A.; Witty, D. R. Tetrahedron 1998, 54,
14255.
(5) For a review on 1,2-diamines, see Lucet, D.; Le Gall, T.; Mioskowski,
C. Angew. Chem., Int. Ed. 1998, 37, 2580.
(6) Coldham, I.; Houdayer, P. M. A.; Judkins, R. A.; Witty, D. R.
Synthesis 1998, 1463.
(7) For deprotonation and diastereoselective quench of a related imid-
azolidine, see: Pfammatter, E.; Seebach, D. Liebigs Ann. Chem. 1991, 1323.
with trifluoroacetic acid (TFA) promoted further hydrolysis
to the diamine 5. The diamine 5 can alternatively be prepared
in one step by treatment of the imidazolidine 3 with TFA.
The yields of the products 4 and 5 (yield of 5 from 4) are
shown in Table 2.
3800
Org. Lett., Vol. 3, No. 23, 2001

to be essentially enantiopure (>99.8% ee). On the basis of
the refined absolute structure parameter,9 the chiral center
at C-5 was assigned the expected (R)-configuration. The
X-ray structure shows an intramolecular hydrogen bond
between the OH group and N-3.
signals was observed; for example, coalescence of C-5 (δ
44.78 and 45.18 ppm)10 occurred at approximately 60 °C
(333 K), which corresponded to a rate constant k ) 88.6 s-1
and therefore a half-life t1/2 ≈ 7.8 × 10-3 s at this
temperature. This corresponds to a barrier to rotation ∆Gq
≈ 69.4 kJ mol-1 at 333 K. Assuming that the change in
entropy is close to zero, then this value of ∆Gq gives a half-
life for rotation t1/2 > 100 h at -78 °C. This large half-life
explains why only one conformer undergoes proton abstrac-
tion at this temperature and therefore the observed limitation
in the yield. On warming, however, rotation becomes
increasingly faster, and at -40 °C, the half-life can be
calculated to decrease to approximately t1/2 ≈ 10 min. This
value would suggest that higher yields could be obtained on
warming; however no enhancement of the yield of the
products 3 was obtained at this temperature (e.g., -40 °C,
1 h, 3, R ) SiMe3 40% + recovered 2, 34%). On warming
above -40 °C, decomposition of the imidazolidine 2 in the
presence of s-BuLi occurs. It seems that the organolithium
species derived from the imidazolidine 2 is at the limit of
stability prior to significant rotation of the N-Boc group and
that the half-life for rotation of the carbamate in THF or
Et2O/(-)-sparteine may be slightly longer than that calculated
above.
In each deprotonation reaction to give the imidazolidine
3, a substantial quantity of the starting material 2 was
recovered, and this could be separated from the product 3
by conventional column chromatography. Experiments with
excess sec-BuLi, longer reaction times, or inverse addition
of the electrophile did not improve this result, which contrasts
with the significantly higher yields obtained from the
asymmetric lithiation and substitution of N-Boc-pyrrolidine.
To gain further insight into the reasons for the low yields,
we carried out the deprotonation at low temperature in d8-
1
THF and obtained the resulting H and 13C NMR spectra.
Before addition of s-BuLi, as a result of slow rotation of the
carbamate group, both conformers of the imidazolidine 2 are
present in a 1:1 ratio.10 On addition of 1 equiv of the base
at -78 °C, the signals for one of the two conformers
disappeared and were replaced by broad signals typical of
an organolithium species. The other conformer remained
unchanged, even on warming to about -40 °C, above which
temperature or after more than about 1 h, marked decom-
position occurred. These results indicate that only one
conformer is deprotonated and support the hypothesis that
initial complex formation precedes proton abstraction.11
Presumably only the conformer in which the carbonyl oxygen
atom is cis to C-5 of the imidazolidine ring undergoes
deprotonation. This result contrasts with the few reported
deprotonations of unsymmetrical, cyclic N-Boc compounds,
in which yields over 50% have been reported.3a-c In these
cases, some rotation about the carbamate C-N bond must
be possible under the reaction conditions, or there must be
a thermodynamic preference for one of the two conformers.
To probe the rate of isomerization of the two conformers,
the imidazolidine 2 was warmed in d6-DMSO and studied
by 13C NMR spectroscopy.12 Coalescence of a number of
In summary, we have shown that it is possible to obtain
chiral 1,2-diamines with high optical purity using asymmetric
deprotonation and electrophilic quench. With the imid-
azolidine 2, the extent of deprotonation is restricted to about
50% as a result of slow rotation of the N-Boc group. The
rate of rotation about the N-CO bond is clearly of crucial
importance in deprotonation chemistry and may influence
the outcome of other examples involving unsymmetrical
carbamates or carboxylic amides.
Acknowledgment. We thank the EPSRC for funding and
GlaxoSmithKline for a CASE award (to T.F.N.H.). We thank
Rosa Klein and Ivan Prokes for NMR studies and Matthew
Sanders for chiral HPLC separations.
(8) Crystal structure determination for 3, R ) C(OH)Ph2: C24H32N2O3;
M ) 396.52; colorless block, 0.28 × 0.22 × 0.08 mm3; monoclinic, space
group P21 (No. 4); a ) 6.0081(2) Å, b ) 20.5769(9) Å, c ) 9.1980(4) Å,
Supporting Information Available: X-ray coordinates
for imidazolidine 3, R ) C(OH)Ph2; NMR spectra for
imidazolidine 2 in the absence and presence of s-BuLi in
d8-THF and on warming in d6-DMSO; experimental proce-
dure for the asymmetric deprotonation and quench and for
the hydrolysis; spectroscopic data for compounds 3, 4 and
5, R ) SiMe3. This material is available free of charge via
the Internet at http://pubs.acs.org.
â) 103.3930(10)°, V ) 1106.20(13) Å3; Z ) 2; T ) 150(2) K.; dcalc
)
1.190 g cm-3; F(000) ) 428; µ(Cu KR, λ ) 1.54178 Å) ) 0.621 mm-1
;
10583 reflections collected; absorption correction by integration; 3843
unique reflections; Rint ) 0.0205; R1 ) 0.0261; wR2 ) 0.0699; absolute
structure parameter ) 0.10(13).
(9) Flack, H. D. Acta Crystallogr. 1983, A39, 876.
(10) For plots of the NMR spectra, see Supporting Information.
(11) Gallagher, D. J.; Beak, P. J. Org. Chem. 1995, 60, 7092.
(12) The barrier to rotation in carbamates is affected little by the
solvent: Cox, C.; Lectka, T. J. Org. Chem. 1998, 63, 2426.
OL016818M
Org. Lett., Vol. 3, No. 23, 2001
3801