Tetrahedron Letters
Re-visiting the diastereoselectivity of organocatalytic conjugate addition
of 2-trimethylsiloxyfuran to trans-crotonaldehyde
⇑
Liubo Li, Anton El Khoury, Brennan O. Clement, Patrick G. Harran
Department of Chemistry and Biochemistry, University of CaliforniaÀLos Angeles, 607 Charles E. Young Drive East, Los Angeles, CA 90095-1569, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
We describe the re-assignment of configuration previously ascribed to product diastereomers resultant
from imidazolidinone-catalyzed conjugate addition of 2-trimethylsiloxyfuran to trans-crotonaldehyde.
A modified procedure that uses a diphenylprolinol catalyst was subsequently developed to selectively
provide the ‘syn’ diastereomeric product in high enantiomeric excess on decagram scales.
Ó 2021 Elsevier Ltd. All rights reserved.
Received 18 February 2021
Revised 25 March 2021
Accepted 30 March 2021
Available online 6 April 2021
Keywords:
Mukaiyama-Michael reaction
Butenolide
The catalyzed conjugate addition of 2-siloxyfurans to Michael
acceptors is a valuable method to prepare -substituted buteno-
We synthesized racemic 1 using a procedure developed by
Yadav and co-workers for the addition of 2-trimethylsiloxyfuran
to enones [6]. Stirring 2-trimethylsiloxyfuran with crotonalde-
c
lides – a structural motif commonly observed in natural products
and synthetic drug substances [1]. Studies by Brown and co-work-
ers [2] demonstrated that chiral imidazolidinones could catalyze
asymmetric conjugate addition of siloxyfurans to enal acceptors.
Following that report, additional organocatalysts were shown
effective for the transformation [3]. As part of studies aimed at syn-
thesizing the pro-apoptotic natural product portimine [4], we
required scalable access to the (3S, 5R) enantiomer of butenolide
2
hyde in the presence of catalytic I gave (±)-1 with exquisite
stereocontrol (76%, dr > 20:1). When that material was reduced
under Luche conditions and the incipient alcohol treated with
NaH (THF, rt), one isomer of bicyclic tetrahydropyran 3 was iso-
lated, albeit in low yield. J-couplings and NOE data (see ESI for
details) obtained for 3 supported the stereochemistry drawn in
Scheme 1, thereby implying precursor (±)-1 was ‘anti’ (as
drawn). Further support for this assignment came from hydro-
genating (±)-1, oxidizing the product under Pinnick conditions
and coupling the resultant acid with 2-oxazolidone to afford
imide (±)-4. NMR spectral data for (±)-4 were identical to those
reported by Katsuki [5a].
1
2
, a ‘syn’ diastereomer resultant from adding 2(5H)-furanone (or
-trimethylsiloxyfuran) to trans-crotonaldehyde (Fig. 1A).
Of the known syntheses of 1 [3a-c], only Brown and co-workers
reported [2] isolating the molecule in high diastereomeric and
enantiomeric excess. In our hands, however, imidazolidinone cat-
alyzed addition of 2-trimethylsiloxyfuran to trans-crotonaldehyde
gave two isomeric butenolides in a ratio of ~2:1. Moreover, data
originally reported for those products were inconsistent with that
expected when using trans-crotonaldehyde. Instead, it appeared to
reflect products derived from using 4-methyl-2-pentenal in the
Having identified the anti diastereoisomer of 1, we confirmed
the identity of the syn diastereomer through regioselective
Wacker–Tsuji oxidation [7] of (3S, 5S)-c-butyrolactone 5 [8] to
1
afford aldehyde 6. The H NMR spectrum of 6 was identical to that
obtained by hydrogenating the minor isomer of 1 obtained from
the Brown catalysis (Table 1, entry 1). Thus, the 2:1 mixture of
isomers generated using the Brown procedure favors the
anti-diastereomer, not the syn as reported. Our assignments
are in agreement with those made by Yanai and co-workers [3a]
and contrary to those made by Luo and co-workers. The major
diastereomer isolated by Luo is syn, not anti as reported [3b].
Confident in our relative stereochemical assignments, we next
sought to optimize the catalytic asymmetric synthesis of (3S,
5R)-1. We screened diastereoselectivity in the addition of 2-
trimethylsiloxyfuran to trans-crotonaldehyde using several
1
reaction [9]. Adding to this puzzle was the fact that H NMR data
for isomers 1 prepared by others were conflicting (Fig. 1B) [3a-c].
Lastly, related reactions of 2(5H)-furanone derivatives with (E)-b-
methyl Michael acceptors gave mainly anti diastereomers [3a,5]
or were non-selective [3b,3c]. For our own studies to progress,
we needed to clarify these findings and develop an efficient,
selective and scalable synthesis of (3S, 5R)-1.
⇑
040-4039/Ó 2021 Elsevier Ltd. All rights reserved.
0