Communication
with dodecanoic acid and desilylation delivered aldol 21. Fi- tion) was also synthesized in enantiopure forms. Given that the
nally, a Dess–Martin oxidation[13] afforded pure (2S,2′′′S)-5.
Enantiopure 5 was also much more stable than one would
presume. A pure sample of (2S,2′′′S)-5 recovered from CDCl3
after recording NMR could be kept in a freezer for 7 days with-
out any discernible racemization (on NMR), although standing
of a neat sample at r.t. for 8 h indeed resulted in partial racemi-
zation.
substrate scope of the present approach is apparently defined
by Dess–Martin oxidation, a reaction that has generally proved
broad functional compatibility, the so far hardly accessible opti-
cally active α-alkyl-ꢀ-keto esters[15] may thus become readily
attainable in general.
Acknowledgments
The 1H NMR of (2S,2′′′S)-5 and a C-2′′′ fully racemized sample
indeed contained significant differences (cf. the Supporting In- This work was supported by the National Natural Science Foun-
formation). In 13C NMR, racemization of C-2′′′ were verified by dation of China (21532002, 21672244) and the Strategic Priority
“doublets” at 205.9/205.8 (C-3′′′), 172.80/172.77 (C-1′′), 68.66/ Research Program of the Chinese Academy of Sciences
68.65 (C-2), 63.2/63.1 (C-3), 52.3/52.2 (C-2′′′), 12.9/12.8 (C-6′′′) (XDB20020200). Prof. Taglialatela-Scafati is gratefully thanked
ppm.
for a scanned HMBC full spectrum of santinol C.
We also synthesized (2S,2′′′R)-5 (Scheme 4) and thus con-
firmed that the “extra” signals in the 1H and 13C NMR of the
racemized sample indeed stemmed from this enantiomer (the
C-2′′′ epimer). Although the NMR data reported for the natural
sample appeared to be for a single enantiomer, its [α]D (–8.5)
was consistent with that for the C-2′′′ racemized synthetic sam-
ple (–8.5).[14]
Keywords: Racemization · Oxidation · Kinetics · Natural
products · Epimers
[1] a) R. M. DiPardo, M. G. Bock, Tetrahedron Lett. 1983, 24, 4805–4808;
b) D. A. Evans, M. D. Ennis, T. Le, N. Mandel, G. Mandel, J. Am. Chem. Soc.
1984, 106, 1154–1156.
[2] a) T. Hashimoto, H. Miyamoto, Y. Naganawa, K. Maruoka, J. Am. Chem.
Soc. 2009, 131, 11280–11281; b) W. Li, J. Wang, X. Hu, K. Shen, W. Wang,
Y. Chu, L. Lin, X. Liu, X. Feng, J. Am. Chem. Soc. 2010, 132, 8532–8533; c)
L. Gao, B. C. Kang, G.-S. Hwang, D. H. Ryu, Angew. Chem. Int. Ed. 2012,
51, 8322–8325; Angew. Chem. 2012, 124, 8447; cf. also d) W. Li, F. Tan, X.
Hao, G. Wang, Y. Tang, X. Liu, X. Feng, Angew. Chem. Int. Ed. 2015, 54,
1608–1611; Angew. Chem. 2015, 127, 1628; e) F. Tan, X. Liu, Y. Wang, S.
Dong, H. Yu, X. Feng, Angew. Chem. Int. Ed. 2018, 57, 16176–16179;
Angew. Chem. 2018, 130, 16408–16411.
[3] Records for (S)- and (R)-2-(tert-butoxycarbonyl)propanoic acid in Reaxys
were wrong (no such compounds exist); cf. the originals cited therein:
a) WO2008/54599 A2, 2008 (on p. 117); b) WO2010/104851 A1, 2010
(on p. 112); c) X. Yu, Q. Liu, J. Wu, M. Zhang, X. Cao, S. Zhang, Q. Wang,
L. Chen, T. Yi, Chem. Eur. J. 2010, 16, 9099–9106.
[4] Ideally, the last step of reaction gives no side products so that the prod-
uct only needs to be separated from the oxidant-related species.
[5] a) G. Berkhan, F. Hahn, Angew. Chem. Int. Ed. 2014, 53, 14240–14244;
Angew. Chem. 2014, 126, 14464; for the data of the antipode, cf b) B.
Loubinoux, J.-L. Sinnes, A. C. O′Sullivan, T. Winkler, Helv. Chim. Acta 1995,
78, 122–128.
Scheme 4. Reagents and conditions: cf. corresponding steps in Scheme 3.
[6] A chromatography through a short column of silica gel gave the clean
product, with 1H and 13C NMR and [α]D consistent with that in ref.[2c]
[7] For further details, cf. Supporting Information.
Conclusion
α-Monoalkyl-ꢀ-keto esters are broadly believed to racemize
quickly. As a consequence, most investigators would not even
consider enantioselective synthesis of such species in the first
place. Although Feng and Hwang-Ryu's elegant works on cata-
lytic Roskamp reaction created the first entry to such com-
pounds, acquisition of optically active α-monoalkyl-ꢀ-keto es-
ters remained difficult in general. Now, a facile alternative ap-
proach (illustrated through synthesis of several representative
compounds) is found. Racemization rates of typical α-alkyl-ꢀ-
keto esters were also made known for the first time, which
could help to reduce the uncertainty/anxiety (which probably
scared away many investigators) about the racemization en-
countered in related studies. The present results show that un-
der neutral or slightly acidic conditions the racemization of α-
alkyl-ꢀ-keto esters is not as fast as broadly presumed, although
such species do racemize very rapidly under basic conditions.
In light of the newly gained knowledge, a much more complex
α-monoalkyl-ꢀ-keto ester santinol C (probably considered to be
unrealistic to date by most investigators due to the racemiza-
[8] a) NaBH4/MeOH reduction of the 11 obtained by the same filtration
technique led to a ca. 1:1:1:1 mixture of four isomers as shown by HPLC,
proving that the 11 was nearly fully racemized. In comparison, 12 re-
mained unracemized, not even after an additional slow column chroma-
tography; b) the [α]D of 12 in EtOAc reduced quickly from +20 to +6.7
within first 7 min but remained unchanged in the following 50 min
(probably formed certain supramolecular structures in solution). The
sample recovered from this EtOAc solution still showed [α]D value (in
CHCl3) consistent with that of an unracemized sample.
[9] a) O. Taglialatela-Scafati, F. Pollastro, G. Chianese, A. Minassi, S. Gibbons,
W. Arunotayanun, B. Mabebie, M. Ballero, G. Appendino, J. Nat. Prod.
2013, 76, 346–353; b) “Doublet” for a few signals in 13C NMR were ex-
pected for a mixture of two epimers when two chiral centers are close
enough to one another: see e.g., Y.-C. Yu, B. Liu, Y. Wu, ChemistrySelect
2016, 1, 3120–3123.
[10] W.-J. Wu, H.-J. Chen, J. You, Y. Wu, B. Liu, Eur. J. Org. Chem. 2015, 5817–
5825.
[11] Prolong reaction led to over acylation at the C-2 (cf. also ref.[10]).
[12] Prolonged reaction led to a mixture of C-2′′/C-3′′ (E) and (Z) isomers.
[13] Extension of the reaction time to 2 h did not lead to any racemization.
[14] After completion of this work we finally succeeded in contact with au-
thors of ref.[9a] and were informed that the 13C shifts of natural 5 were
Eur. J. Org. Chem. 0000, 0–0
4
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim