Stereoselectivity in the double reductive alkylation of pyrroles: Synthesis of cis-3,4-disubstituted pyrrolidines

Article abstract of DOI:10.1039/a809193e

The preparation and Birch reduction of a 1,3,4-tri-substituted pyrrole is described: the heterocycle is loaded with electron-withdrawing groups and undergoes a double reductive alkylation reaction to yield cis-3,4-disubstituted pyrrolidines.

Full text of DOI:10.1039/a809193e

Stereoselectivity in the double reductive alkylation of pyrroles: synthesis of
cis-3,4-disubstituted pyrrolidines
Timothy J. Donohoe,*a Rakesh R. Harjia and Rick P. C. Cousinsb
a Department of Chemistry, The University of Manchester, Oxford Road, Manchester, UK M13 9PL.
E-mail: t.j.donohoe@man.ac.uk
b GlaxoWellcome Research and Development, Medicines Research Centre, Gunnels Wood Road, Stevenage,
UK SG1 2NY
Received (in Liverpool, UK) 23 November 1998, Accepted 26th November 1998
The preparation and Birch reduction of a 1,3,4-tri-substi-
tuted pyrrole is described: the heterocycle is loaded with
electron-withdrawing groups and undergoes a double re-
ductive alkylation reaction to yield cis-3,4-disubstituted
pyrrolidines.
Subjection of pyrrole 3 to reductive alkylation conditions (6
equiv. Li in NH3) is described in Scheme 3. We routinely add
(MeOCH2CH2)2NH (10 equiv.) to Birch reductions as this
amine appears to ‘mop-up’ lithium amide formed during the
reaction and reduces the presence of undesirable by-products.5†
Isoprene ‘quenches’ excess electrons in the reaction after
reduction is complete (addition of isoprene caused the reaction
to turn from deep blue to yellow—this colour was dissipated
upon addition of an alkyl halide). The first reaction was
quenched with MeI and produced a single product in good yield.
The Birch reduction is a particularly useful synthetic reaction
capable of transforming aromatic substrates into partially
unsaturated products.1 We have recently investigated and
developed the partial reduction of heterocycles and found that
Birch reductive alkylation of electron-deficient pyrroles takes
place readily (Scheme 1).2 For example, the transformation of
pyrroles 1 and 2 into either 2- or 3-pyrroline isomers is a high
yielding process that enables introduction of an alkyl group
adjacent to the activating ester.3–5 In each case, we believe that
addition of electrons to the pyrrole forms a dianion which
deprotonates ammonia to furnish an enolate; this reacts with an
external electrophile to produce the observed products.
1
We identified the pyrrolidine 4 (R = Me) from H NMR
spectroscopic data (Table 1).‡ We were then able to repeat the
reaction and quench with a variety of electrophiles to generate
compounds 5–7 in good yields (Table 1). In every case 1H NMR
spectroscopy showed that an alkyl group had been introduced a
to each of the two esters, and also revealed the presence of the
four protons adjacent to the pyrrolidine nitrogen.
Unfortunately, the NMR spectra of 4–7 did not allow us to
assign relative stereochemistry and this was subsequently
proved by conversion to the corresponding quarternary ammo-
nium salts via a two step sequence (Scheme 3). Assignment of
cis relative stereochemistry to compounds 4–7 would mean that
the two N-methyl groups on (meso) salts 8–11 are diaster-
eotopic, and could therefore resonate at different chemical
shifts. On the other hand, trans stereochemistry would make the
N-methyl groups of 8–11 (now C2 symmetrical) homotopic and
they would therefore have identical chemical shifts. Gratify-
i
CO2Pri
CO2Pri
CO2Cy
N
N
87%
Me
Boc
Boc
1
Me
CO2Cy
i
N
N
72%
1
ingly, salts 8–11 each displayed two (3H) singlets in the H
Adoc
Adoc
2
NMR spectrum,‡ thus allowing us to assign with confidence the
stereochemistry of 4–7 as shown (Table 2).
Adoc = adamantyloxycarbonyl
R
R
Scheme 1 Reagents and conditions: i, Na (3 equiv.), NH3/THF, then MeI
(excess), 278 °C.
EtO2C
CO2Et
EtO2C
CO2Et
i
See Table 1
Recently we have attempted to expand this work by using
more highly-substituted pyrroles as substrates for the Birch
reduction. In these cases diastereoisomeric products can be
formed and we now present data that shows a high level of
stereochemical control is possible. In addition, we also describe
a double reductive alkylation reaction whereby two alkyl groups
can be introduced onto the heterocycle in one step.
Studies began with the preparation of electron-deficient
pyrrole 3 using standard methodology (Scheme 2). Reaction of
TsCH2NC,6 KOBut and diethyl fumarate gave a 3,4-di-
substituted pyrrole which was protected, on nitrogen, as a
adamantyloxycarbonyl (Adoc) derivative using adamantyl
fluoroformate (45% overall yield from diethyl fumarate).
N
N
Adoc
4–7
Adoc
3
ii, iii
R
See Table 2
R
EtO2C
CO2Et
I–
+
N
8–11
Me Me
Scheme 3 Reagents and conditions: i, Li (6 equiv.), NH3, THF,
(MeOCH2CH2)2NH, 278 °C, then isoprene (3 drops), then RI (excess); ii,
TFA, CH2Cl2; iii, MeI, KHCO3, MeOH.
Table 1 Birch reduction of 5
EtO2C
CO2Et
i, ii
Entry
RX
cis/trans
Yield (%)
Compound
EtO2C
45%
N
1
2
3
4
MeI
EtI
BuiI
Allyl-I
!20:1
!10:1
!10:1
!10:1
77
82
79
70
4
5
6
7
CO2Et
3
Adoc
Scheme 2 Reagents and conditions: i, TsCH2NC, ButOK, THF; ii, Adoc-F,
MeCN, Et3N.
Chem. Commun., 1999, 141–142
141

O–
O–
Table 2 Quaternary ammonium salts
EtO2C
CO2Et
EtO2C
OEt
Li+
EtO2C
OEt
2Li+
Entry
R
dH(NMe) (CDCl3)
Yield (%)
Compound
+ 2e
NH3
–
1
2
3
4
Me
Et
3.52, 3.77
3.67, 3.81
3.69, 3.77a
3.65, 3.83
52
61
55
58
8
9
10
11
N
N
N
3
B
Adoc
Adoc
A
Adoc
Bui
Allyl
+ 2e
+ H+ (NH3)
a NMR run in acetone-d6.
–O
–O
O–
O
O
O
EtO
OEt
OEt
EtO
OEt
EtO
R
Compounds 4–7 were formed with high levels of ster-
eoselectivity; such ratios were formulated by examination of the
NMR spectra of the crude reaction mixtures. As far as the
formation of 4 is concerned we were able to prepare an authentic
sample of the trans isomer by another route and (in comparison
with this standard) could not observe the trans isomer in the
crude reduction reaction. Although each of the other cases was
assigned !10:1 selectivity without comparison to a standard,
we believe this is a conservative estimate. Not only were the
Birch reduced products 5–7 free from detectable impurities but
also the amines formed by Adoc deprotection and the salts 8–11
appeared as single isomers by NMR spectroscopy.
We noticed that the alkylation step of the Birch reduction
proceeded at different rates with the electrophiles that were
used. Not surprisingly, reaction with BuiI was much slower than
reaction with MeI. Using this information we developed a
protocol for the sequential dialkylation of the pyrrole with two
different electrophiles (Scheme 4). So, reaction of 3 with
RX
RX
R
R
N
N
N
2Li+
Adoc Li+
C
Adoc
Adoc
D
Fig. 2
ments show that both potassium and sodium metals give
identical selectivity to lithium, thus dampening arguments
based on chelation. However, we have preliminary results
which show that, remarkably, the ammonia solvent is essential
in order to achieve high stereoselectivity. We cannot comment
on the exact role of the ammonia at this point, but note that
Schultz has previously observed a similar relationship between
solvent and the stereoselectivity displayed by enolates gen-
erated in the Birch reduction.8
We believe this type of reaction will be of use in both natural
product synthesis and medicinal chemistry and that, as this
reaction results in the formation of two adjacent quarternary
chiral centres with control of relative stereochemistry, it is
worthy of further study.
R
EtO2C
CO2Et
EtO2C
CO2Et
We wish to thank GlaxoWellcome (CASE award to R. R. H.)
and Zeneca Pharmaceuticals (Strategic Research Fund) for
financial support.
i
N
N
12 R = Me (76%)
13 R = Bn (84%)
Adoc
3
Adoc
Notes and references
Scheme 4 Reagents and conditions: i, Li (6 equiv.), NH3, THF,
(MeOCH2CH2)2NH, 278 °C, then isoprene (3 drops), then BuiI (excess) ,
then RX (excess).
† We assume that (MeOCH2CH2)2NH is acidic enough to become
deprotonated by lithium amide. Presumably, the anion derived from the
amine additive is chelated and is a relatively unreactive species compared to
lithium amide itself. We also note that (MeOCH2CH2)2NH is not acidic
enough to protonate any enolate formed in the reaction, hence allowing us
to add 10 equiv. without complication.
‡ Selected data for 4: dH(140 °C, 1,2-Cl2C6H4) 4.28 (2H, AB, CH2N), 4.24
(4H, q, J 7.5, CH2O), 3.38 (2H, AB, CH2N), 2.32 (6H, br s, Adoc), 2.22 (3H,
br s, Adoc), 1.75 (6H, br s, Adoc), 1.50 (6H, s, CH3), 1.33 (6H, t, J 7.5 ,
CH3CH2); HRMS (CI): C23H35NO6 requies 422.2542, found 422.2534. For
8: dH(CDCl3) 4.40 (AB, 2H, CH2N), 4.20–4.08 (4H, m, CH2O), 4.05 (2H,
AB, CH2N), 3.77 (3H, s, NCH3), 3.52 (3H, s, NCH3), 1.58 (s, 6H, CH3),
1.23 (6H, t, J 7.1, CH3CH2); Calc. for C14H26NO4I: C, 42.12; H, 6.56; N,
3.51. Found C, 42.39; H, 6.40; N, 3.40%.
lithium metal as before but quenching with excess BuiI and then
(after 2 min) excess MeI (or BnBr) enabled the synthesis of 12
and 13 in good yields. In both cases the reaction gave a single
1
product as judged by H NMR spectroscopy. Further proof of
the identity of 12 was obtained by conversion to 14 under
standard conditions (14 appeared as one isomer) and subsequent
NOE studies (Fig. 1). The NOE experiment described shows
that 14 is the cis isomer. With all of this evidence for cis
stereoselectivity in the dialkylation reaction, compound 13 was
assigned as cis by analogy.
Percent NOE observed
14
1 L. N. Mander, in Comprehensive Organic Synthesis, ed. B. M. Trost and
I. Fleming, Pergamon, New York, 1991, vol. 8; P. W. Rabideau and Z.
Marcinow, Org. React., 1992, 42, 1; P. W. Rabideau, Tetrahedron, 1989,
45, 1579.
Irradiate
1
X
1
2
1
X
3
0
4
3'
3
3, 3′
Me2
CO2Et
EtO2C
1
2
5
+
2 T. J. Donohoe, R. Garg and C. A. Stevenson, Tetrahedron: Asymmetry,
1996, 7, 317.
3 T. J. Donohoe and P. M. Guyo, J. Org. Chem., 1996, 61, 7664.
4 T. J. Donohoe, P. M. Guyo, R. L. Beddoes and M. Helliwell, J. Chem
Soc., Perkin Trans. 1, 1998, 667.
N
I–
3
1
3
4
6
X
8
9
1Me Me
3'
X
14
Fig. 1 NOE studies on 14
5 T. J. Donohoe, P. M. Guyo, R. R. Harji, M. Helliwell and R. P. C.
Cousins, Tetrahedron Lett., 1998, 39, 3075.
In terms of mechanism, we suggest that 3 accepts two
electrons and forms dianion A (Fig. 2). Dianion A is then basic
enough to deprotonate ammonia and form enolate B. Pre-
sumably, the presence of an ester group at C-4 means that the C-
4,5 alkene in B is susceptible to further reduction by addition of
two electrons and protonation at C-5 (by ammonia) to give C,7
which is then alkylated twice. Presumably, the relative
stereochemistry is determined by the facial selectivity of the
second alkylation step (reaction of D) and it is surprising that
such high levels of control are observed. Additional experi-
6 A. M. van Leusen, H. Siderius, B. E. Hoogenboom and D. van Leusen,
Tetrahedron Lett., 1972, 5337; D. P. Arnold, L. J. Nitschinsk, C. H. L.
Kennard and G. Smith, Aust. J. Chem., 1991, 44, 323.
7 S. Li, X. Fang, Z. Wang, Y. Yang and Y. Li, Synth. Commun., 1993, 23,
2051; A. G. Schultz, M. Macielag, D. E. Podhorez, J. C. Suhadolnik and
R. K. Kullnig, J. Org. Chem., 1988, 53, 2456.
8 See A. G. Schultz, M. Macielag, P. Sundararaman, A. G. Taveras and M.
Welch, J. Am. Chem. Soc., 1988, 110, 7228.
Communication 8/09193E
142
Chem. Commun., 1999, 141–142