Key words
isoxazoles - pyrrolidinones - reductive rearrangement - heterocyle–heterocyle strategy
- (3+2) cycloaddition - MCR
Access to polysubstituted nitrogen heterocycles are crucial to the discovery of novel
biologically active compounds (Figure 1).[1] Pyrrolidinones, including 3-hydroxy-2-pyrrolidinones, are one such class of nitrogen
heterocycles being actively pursued for their biological activity ranging from HIV-1
inhibitors,[2] antitumor oral drugs,[3] antimicrobial,[4] and antibacterial applications.[5] Recently, they have been accessed via a multicomponent reaction (MCR) approach requiring
an aldehyde, substituted aniline, and either acetylene dicarboxylates or 2-oxo-1,4-dicarboxylates.
The MCR provides a convenient approach to 2-arylated 4-hydroxy-5-pyrrolidinones (Scheme
1).[6]
Figure 1 Naturally occurring leopoilic acid A (i) and cytochalasin B (ii)
Scheme 1 Literature-known multicomponent reaction approach to 2-pyrrolidinones and the current
planned H–H approach to 2-pyrrolidinones
With our ongoing interest in heterocycle–heterocycle (H–H) interconversion strategy[7] we became interested to extend the H–H strategy to access 5-aryl/alkyl-substituted
3-hydroxy-2-pyrrolidinones via re-organization of the corresponding isoxazoles under
reducing conditions (Scheme 1).
The isoxazoles needed for the study were prepared via the (3+2) cycloaddition of the
corresponding nitrile oxides with dimethyl acetylene dicarboxylate (symmetrically
substituted acetylene) at room temperature in moderate to good yields (50–84%, Scheme
2).[8]
Scheme 2 Synthesis of isoxazole derivatives 6a–l via a (3+2) cycloaddition on dimethyl acetylenedicarboxyalte (DMAD)
In general, the yield for isoxazole formation was relatively higher for electron-withdrawing
substituents on the aryl rings than in the presence of electron-donating substituents
on the ring. The low yield for the (3+2) cycloaddition in electron-donating nitrile
oxide can be attributed to self-dimerization of nitrile oxides.[9]
Attempt to carry out reductive reorganization of isoxazoles 6a–l
[10] to 2-pyrrolidinones 1a–l was initially optimized with 6a as model substrate with iron as choice of reductant (Scheme 3).[11] Reductive rearrangement of 5a to 1a was not observed when the reaction was carried out with ammonium chloride as additive
and ethanol as solvent, even after prolonged heating at 80 °C (5 equiv.; Table 1,
entry 1).
Scheme 3 Reductive rearrangement of 6a–l to 1a–l in the presence of iron as reductant and acetic acid as solvent
Table 1 Optimization of the Reductive Rearrangement of 6 to 1
|
Entry
|
Solvent
|
Reductant (equiv.)
|
Temp (°C)
|
Time (min)
|
Conversion (yield, %)
|
|
1
|
EtOH
|
Fe/NH4Cl (5)
|
80
|
300
|
no reaction
|
|
2
|
EtOH
|
Zn/NH4Cl (5)
|
80
|
300
|
no reaction
|
|
3
|
EtOH
|
Fe/HCl (5)
|
80
|
300
|
multiple spots
|
|
4
|
AcOH
|
Fe (5)
|
110
|
300
|
50
|
|
5
|
AcOH
|
Fe (10)
|
110
|
60
|
100 (71)
|
|
6
|
AcOH
|
Fe (10)
|
110
|
120
|
100 (65)
|
|
7
|
EtOH
|
Pd/C, H2 (5 bar)
|
110
|
120
|
multiple spots
|
Replacement of either the reductant, i.e., iron with zinc, or the additive, i.e.,
ammonium chloride with hydrochloric acid, did not yield the desired product (entry
2, Table 1). With hydrochloric acid as additive the starting material was consumed
albeit with extensive degradation of 6a within 5 h at 80 °C (entry 3, Table 1).
To our surprise, replacing ethanol with acetic acid as reaction solvent, formation
of 1a was observed (no additive) at reflux, however, with conversion of only 50% even after
5 h (entry 4, Table 1). To accelerate the reaction, further optimization was carried
out by doubling the reductant quantity from 5 equiv. to 10 equiv., and to our delight
it gave 1a in 71% yield within 1 h of the reaction time (entry 5, Table 1). When the reaction
time was extended to 2 h, a 5–10% drop in yield was observed suggesting decomposition
of product under these conditions (entry 6, Table 1). Attempts to carry out reductive
reorganization under hydrogenating conditions did not yield 1a despite consumption of staring material (entry 7, Table 1).
The optimized conditions for reductive reorganization (entry 5, Table 1), i.e., iron
as reductant (10.0 equiv.) in acetic acid under reflux conditions for 1 h, were used
for general applicability of the method on other isoxazoles 6b–l. Indeed, formation of 1b–l was observed in all the cases in moderate to good yields (50–80%, Table 2). Interestingly,
trends in the yield after isolation of 1a–l were similar to those observed in the synthesis of isoxazoles 6a–l, i.e., electron-withdrawing groups on aryl ring at 2-position in 6a–l gave higher yields of rearranged product (1d–f,h, Table 2) than alkyl/electron-donating substituents on the aryl ring (1b,g,i,l, Table 2).
This could be due to stabilization of the developing charge during the reduction step.
It was also observed that the yield of alkyl-substituted isoxazole gave reasonable
to good yields of 1 (1c,k, Table 2) under the reaction conditions.[12]
A plausible mechanism for the reductive rearrangement might be attributed to an initial
SET between the reductant and 6 followed by protonation to form an intermediate A. Intermediate A could lead to the desired product following either path 1 or path 2 characterized
by tautomerization, cyclization, and reduction. Path 1 involves an initial tautomerization
(B), cyclization (C), reduction (1′), or path 2 involves further reduction (D), tautomerization (E), cyclisation (1′) and tautomerization to yield 1 (Scheme 4).
Scheme 4 Plausible mechanism for the reductive rearrangement of isoxazoles 5 to 2-pyrrolidinone 1 using iron as reductant in acetic acid as solvent
In conclusion the present work demonstrates the conversion of isoxazoles 6a–l
[10] into polysubstituted 2-pyrrolidinones (1a–l, 50–80% yield) under reductive rearrangement conditions with iron as reductant in
acetic acid as solvent. The approach has a distinct advantage in accessing unsubstituted
2-pyrrolidinones at the nitrogen center allowing further scope of derivatization.
The work further demonstrates the usefulness of heterocycle–heterocycle interconversion
approach to access polysubstituted 2-pyrrolidinones from their corresponding isoxazoles.
Further work is necessary to understand the overall mechanism and to exploit the full
potential of this methodology.
a Oximes 7a–l were prepared using the literature protocol.
