Regioselective directed lithiation of N-Boc 3-hydroxypyrrolidine. Synthesis of 2-substituted 4-hydroxypyrrolidines

Article abstract of DOI:10.1039/a804211j

N-Boc 3-hydroxypyrrolidine 1 undergoes directed C-lithiation at C5 and not, as previously reported, at C2; the resulting dilithiated intermediate 6 has been trapped by a range of electrophiles to give 2-substituted 4-hydroxypyrrolidines 7.

Full text of DOI:10.1039/a804211j

View Online / Journal Homepage / Table of Contents for this issue
Regioselective directed lithiation of N-Boc 3-hydroxypyrrolidine. Synthesis of
2-substituted 4-hydroxypyrrolidines
Mihiro Sunose,a Torren M. Peakman,a Jonathan P. H. Charmant,a Timothy Gallagher*a† and Simon J. F.
Macdonaldb
a School of Chemistry, University of Bristol, Bristol, UK BS8 1TS
b GlaxoWellcome Research and Development, Medicines Research Centre, Gunnels Wood Road., Stevenage, UK SG1 2NY
N-Boc 3-hydroxypyrrolidine 1 undergoes directed C-lithia-
tion at C5 and not, as previously reported, at C2; the
resulting dilithiated intermediate 6 has been trapped by a
range of electrophiles to give 2-substituted 4-hydroxypyrro-
lidines 7.
TMEDA at 278 °C and then allowing the solution to warm to
246 °C. After 2 h, the mixture is then recooled (to 278 °C) and
Me3SiCl (2.2 equiv.) was added giving, after work up, the cis-
2,4-disubstituted adduct 4 as the only observed disilylated
product (Scheme 1).
Adduct 4 was then desilylated (at oxygen) to give 5 which
proved to be a more suitable intermediate for assignment of both
Directed metallation provides a versatile method for achieving
substitution adjacent (a) to an amino moiety and this is a
process which has found widespread application within hetero-
cyclic chemistry.1 The presence of additional and appropriately
positioned heteroatom residues can, by providing additional
activation, also enable regiochemical control to be exercised in
the metallation (usually lithiation) step.2–4 For example, N-Boc
pyrrolidines3 and piperidines4 carrying a methoxy residue at C3
have been shown to undergo lithiation at C2 which is then
followed by loss of MeOLi to give the corresponding cyclic
enamine. Choosing a C3 substituent that is a poorer leaving
group should, by suppressing b-elimination, allow this residue
to be retained in the resulting product and this concept has been
successfully applied to O-based heterocycles.5
More recently, this approach has also been utilised for
N-heterocycles and Pandey and Chakrabarti6 have reported that
N-Boc 3-hydroxypyrrolidine 1 undergoes lithiation (at 278 °C
using BusLi, THF–TMEDA) followed by silylation to give
exclusively the C,O-disilylated adduct 3 (in 86% yield), an
intermediate that was then later employed in an alkaloid
synthesis (Scheme 1).‡ These authors suggest that the sense of
regiocontrol ‘was expected due to the directing effects of the
hydroxy group for the initial metallation reaction’ and the
implication of this claim is that the C,O-dilithiated species 2 is
involved.
1
regio- and stereo-chemistry. This was carried out by H NMR
(COSY and NOE difference) analysis and, in addition, the
relative stereochemistry of 5 has been established by X-ray
crystallographic analysis (Fig. 1).§
The results of our study point towards the intermediacy of the
C,O-dilithiated species 6 (and not 2) as the only detectable
intermediate (based on trapping) produced by lithiation of
N-Boc-3-hydroxypyrrolidine 1, however, the relative config-
uration of 6 has not been investigated.
The broader synthetic scope of this regiospecific lithiation
chemistry has been further developed and intermediate 6 has
been trapped with a range of electrophiles, including primary
alkyl halides and enolizable aldehydes, to give the correspond-
ing N-Boc 2-substituted 4-hydroxpyrrolidines 7 (Scheme 2,
Table 1).¶ Yields for adducts 7∑ have not been optimised
because in most cases only modest cis/trans selectivity was
observed. This likely reflects a combination of a highly reactive
intermediate, i.e. 6, with a lack of a significant conformational
bias associated with the five membered ring, but we have
established that cis/trans is kinetically controlled: attempts to
lithiate cis-7e (using the standard reaction conditions), and
thereby provide evidence for an equilibration pathway, failed.
In summary, C-lithiation of N-Boc 3-hydroxypyrrolidine (1)
takes place distal to the hydroxy (lithioalkoxy) function but
nevertheless offers a versatile and synthetically useful entry into
2-substituted 4-hydroxypyrrolidines. Our observations con-
As part of a broader programme, we have had cause to
examine the lithiation of 1 and two key issues are apparent.
Firstly, in our hands, C-lithiation of 1 does not occur at 278 °C
and, secondly, at higher temperatures when C-lithiation does
take place, the only products observed are those resulting from
lithiation distal to the C3 hydroxy, rather than proximal as
claimed by Pandey. We have achieved lithiation of N-Boc
3-hydroxypyrrolidine 1 using BusLi (2.2 equiv.) in THF–
OH
OLi
OSiMe3
i
ii
ref.6
Li
SiMe3
3
N
N
N
Boc
Boc
Boc
1
2
OSiMe3
iv
OH
iii
Me3Si
Me3Si
N
N
4
Boc
5
Boc
Scheme 1 Reagents and conditions: i, (ref. 6) BusLi, THF–TMEDA,
278 °C; ii, Me3SiCl (2.2 equiv.); iii, (this work) BusLi (2.2 equiv.), THF–
TMEDA, 278 °C to 246 °C, 2 h, cool to 278 °C, then Me3SiCl (75%); iv,
Bu4NCl, KF, MeCN (100%)
Fig. 1 Molecular structure of 5
Chem. Commun., 1998
1723

View Online
Data for 4: [a]D24 + 43.3 (c 1, CH2Cl2); dH(400 MHz, C6D6) 4.03 (1 H, m,
OH
OLi
OH
H4), 3.69 (1 H, m, H5), 3.30–3.23 (2 H, m, H2 and H5), 1.91 (1 H, m H3),
1.78 (1 H, m, H3), 1.56 (9 H, s); dC(100 MHz, CDCl3) (doubling due to
rotamer populations) 154.9/154.5, 79.6/78.6, 71.0/70.2, 54.5/54.3,
46.3/46.0, 37.3/36.8, 28.5, 20.2, 21.8. Data for 5: mp 107–110 °C
(pentane–EtOAc); [a]D24 + 57.4 (c 1, CH2Cl2); dH(300 MHz, CDCl3) 4.39
(1 H, pentet, J 5.9, H 4a), 3.78 (1 H, d d, J 11.7, 5.9, H5a), 3.30 (1 H, dd,
J 9.2, 7.2, H2a), 3.05 (1H, dd, J 11.7, 5.9, H5b), 2.26 (1 H, dddd, J 12.8, 9.2,
5.9, 0.7, H3a), 1.78 (1 H, m, H3b), 1.69 (1 H, br s, OH), 1.46 (9 H, s, But)
and 0.10 (9 H, s, Me3Si). Stereochemical assignments are based on NOE
difference studies. The crystal structure of 5 was determined from data
collected on a Siemens SMART diffractometer (l = 0.71073 Å) at 173(2)
K. The structure was solved by direct and Fourier methods and refined by
least squares against all F2 data corrected for absorption. Crystal data:
i
ii
Li
E
N
N
N
Boc
Boc
Boc
1
6
7a–f
Scheme 2 Reagents and conditions: BusLi (2.2 equiv.), THF–TMEDA,
278 °C to 246 °C, 2 h; ii, cool to 278 °C, then E+ (2.2 equiv.) (see Table
1)
Table 1 Trapping of 6 with different electrophiles
2-Substituted 4-hydroxypyrrolidine (7)
C12H25NO3Si, M
=
259.4, orthorhombic, space group P212121, a
=
=
Electrophile
% yield and trans/cis ratio
6.159(1), b = 24.957(4), c = 9.709(2) Å, U = 1492.5(5) Å3, Z = 4, Dc
1.15 g cm23, m = 0.156 mm21, 3382 unique data, q < 27.4°, R1 = 0.043.
CCDC 182/194.
OH
¶ All lithiation reactions shown in Table 1 were carried out as described in
the following example: to a solution of N-Boc (3R)-hydroxypyrrolidine 1
(189 mg, 1 mmol) in THF (5 cm3) at 278 °C was added TMEDA (0.34 cm3,
2.21 mmol) followed by BusLi (1.7 cm3, 1.3 m in cyclohexane, 2.21 mmol).
The resulting bright yellow reaction mixture was warmed to 246 °C, stirred
for 2 h, then recooled to 278 °C, and dimethyl sulfate (0.21 cm3, 2.21
mmol, dried over 4 Å MS) was added dropwise. The reaction mixture was
then allowed to warm slowly to room temperature (over 5 h) and, after this
time, water (5 cm3) and CH2Cl2 (7 cm3) were added and the organic layer
was separated. The aqueous layer was extracted with CH2Cl2 (2 3 5 cm3),
the organic extracts were combined, and dried (K2CO3). Concentration in
vacuo and purification of the residue by flash chromatography (light
petroleum–EtOAc) gave N-Boc 2-methyl-4-hydroxypyrrolidine 7e (131
mg, 65%) as a 5:1 mixture of diastereomers. Recrystallisation gave the
major (cis) diastereomer as colourless crystals. Data for cis-7e: mp 88–89
7a 50%
Bu3SnCl
1 : 1
Bu3Sn
N
Boc
OH
7b 37%
1 : 1
Br
Br
Br
Me
Me
N
Boc
OH
7c 39%
1 : 1
Me
Me
N
Boc
24
°C (light petroleum–EtOAc); [a]D + 21.1 (c 1, CH2Cl2). The cis
OH
stereochemistry of this major adduct has also been conformed by X-ray
crystallographic analysis, details of which will be described elsewhere.)
∑ Enantiomerically pure (R)-1 was used in these initial studies. Carbamate
resonance complicated interpretation of the NMR data associated with
adducts 7 but where this was an issue it was overcome by N-deprotection
(using TFA, CH2Cl2, room temp.) to give the corresponding free amine. The
formation of stereoisomers at C2, rather than a mixture of C2/C5
regioisomers, was confirmed by Swern oxidation of the cis/trans mixture to
give a single 2-substituted pyrrolidin-4-one; this process was carried out for
adducts 7b–e. Octanal gave 7f as an inseparable 1:1 mixture of two of the
four possible products, but the stereochemistry of these products has not
been assigned. In addition, all new compounds have been characterised by
spectroscopic methods and either elemental analysis or HRMS.
7d 59%
1 : 1
N
Boc
OH
7e 65%
5 : 1
(see text)
Me2SO4
Octanal
Me
N
Boc
OH
7f 39%
1 : 1
(see text)
C7H15
1 P. Beak, A. Basu, D. J. Gallagher, Y. S. Park and S. Thayumanavan, Acc.
Chem. Res., 1996, 29, 552; M. Gray, M. Tinkl and V. Snieckus, in
Comprehensive Organometallic Chemistry, ed. E. W. Abel and F. G. A.
Stone, Elsevier, New York, vol. 11, ch. 1; P. Beak, W. J. Zajdel and D.
B. Reitz, Chem. Rev., 1984, 84, 471.
N
HO
Boc
2 D. L. Comins and M. A. Weglarz, J. Org. Chem., 1988, 53, 4437; for a
recent example involving a desymmetrisation process, see M. Lautens, E.
Fillion and M. Sampat, J. Org. Chem., 1997, 62, 7080. Steric factors also
play a role in determining regiochemical control in spite of the presence
of a (sulfur-based) directing group, see D. J. Hart, J. Li, W.-L. Wu and A.
P. Kozikowski, J. Org. Chem., 1997, 62, 5023.
3 M. Giles, M. S. Hadley and T. Gallagher, J. Chem. Soc., Chem. Commun.,
1990, 831.
4 P. Beak and W. K. Lee, J. Org. Chem., 1993, 58, 1109.
5 V. Wittmann and H. Kessler, Angew. Chem., Int. Ed. Engl., 1993, 32,
1091; P. Lesimple and J.-M. Beau, Bioorg. Med. Chem., 1994, 2, 1319;
O. Frey, M. Hoffmann, V. Wittmann, H. Kessler, P. Uhlmann and A.
Vasella, Helv. Chim. Acta, 1994, 77, 2060; D. Mazeas, T. Skrydstrup and
J.-M. Beau, Angew. Chem., Int. Ed. Engl., 1994, 33, 1383; M. Hoffmann
and H. Kessler, Tetrahedron Lett., 1994, 35, 6067; D. Urban, T.
Skrydstrup, C. Riche, A. Chiaroni and J.-M. Beau, Chem. Commun.,
1996, 1883. For a review of carbohydrate-based C1 nucleophiles, see
J.-M. Beau and T. Gallagher, Top. Curr. Chem., 1997, 87, 1.
6 G. Pandey and D. Chakrabarti, Tetrahedron Lett., 1996, 37, 2285.
7 D. A. Fletcher, R. F. McMeeking and D. J. Parkin, J. Chem. Inf. Comput.
Sci., 1996, 36, 746.
cerning C-lithiation clearly conflict with those described
earlier6 by Pandey and Chakrabarti. This is a concern but until
this group publishes either detailed experimental protocols or
compound data, it is not possible to identify those factors that
might account for this apparent contradiction.
We thank The Overseas Research Students Awards Scheme
and GlaxoWellcome Research and Development for financial
support (to M. S.), and acknowledge use of the EPSRC’s
Chemical Database Service at Daresbury.7
Notes and References
† E-mail: t.gallagher@bristol.ac.uk
‡ Metallation of 1 is described6 as having taken place at 278 °C. No
spectroscopic data was reported for the C,O-disilylated adduct 3 which was
also indicated as an unspecified mixture of diastereomers.
§ When we carried out lithiation and quenching of 1 at 278 °C (and the
solution was not allowed to warm above this temperature) then only the
O-silyl ether of 1 was observed. This monosilylated product was also
observed as a minor component (in 15% yield) in the generation of 4. The
exclusive formation of cis-4 has not yet been investigated but intra-
molecular delivery involving participation by oxygen cannot be ruled out.
Received in Cambridge, UK, 4th June 1998; 8/04211J
1724
Chem. Commun., 1998