Enantioselective cascade formal reductive insertion of allylic alcohols into the C(O)-C bond of 1,3-diketones: Ready access to synthetically valuable 3-alkylpentanol units

Article abstract of DOI:10.1021/ol500821c

An unprecedented cascade reaction combining dual iron-amine-catalyzed enantioselective functionalization of allylic alcohols and chemoselective acyl transfer is presented. It allows, from diketones and allylic alcohols, preparation of efficiently function

Full text of DOI:10.1021/ol500821c

Letter
pubs.acs.org/OrgLett
Enantioselective Cascade Formal Reductive Insertion of Allylic
Alcohols into the C(O)−C Bond of 1,3-Diketones: Ready Access to
Synthetically Valuable 3‑Alkylpentanol Units
Mylene Roudier, Thierry Constantieux, Adrien Quintard,* and Jean Rodriguez*
̀
́
Aix Marseille Universite, Centrale Marseille, CNRS, iSm2 UMR 7313, 13397 Marseille, France
S
* Supporting Information
ABSTRACT: An unprecedented cascade reaction combining dual
iron−amine-catalyzed enantioselective functionalization of allylic
alcohols and chemoselective acyl transfer is presented. It allows,
from diketones and allylic alcohols, preparation of efficiently
functionalized γ-chiral alcohols in up to 96% yield and 96:4 er.
The interest of this redox-, atom-, and step-economomical approach
was further demonstrated in the short synthesis of several key
fragments of biologically active natural products or odorant
molecules.
Scheme 1. Some Examples of the Occurrence of 3-
Alkylpentanol Units and Projected Cascade
nantioselective access to functionalized chiral building
E
blocks in a minimum number of operations is key for the
development of sustainable fine chemistry. To orient researcher’s
efforts, several concepts such as the atom economy from Trost,
the step economy from Wender, or the redox economy from
Baran have recently been introduced.1 These guidelines all tend
to reach the same goal of a limited economical and thus
environmental impact of newly developed synthetic strategies
and methodologies. Although tremendous progress was achieved
in the last decades, some crucial chiral building blocks still require
lengthy and costly sequences for their preparation. This is the
case for optically active functionalized 3-alkylpentanol units
found in important biologically active natural products such as
apratoxin A and lyconadin A or in Doremox, a constituent of the
rose scent commercialized by Firmenich. All these chiral units
should be readily available from key protected γ-alkyl alcohols 3
(Scheme 1a, R3 = Me).2
The two main difficulties to access this important 3-
alkylpentanol motif 3 rely, first, on the selective generation of
the appropriate defined function (ketone and protected alcohol),
often requiring functional group interconversion, namely redox
transformations and protecting group manipulations that
inevitably lengthen the synthetic strategy. Second, the control
of the absolute configuration of these γ-chiral alcohols
constitutes a real synthetic challenge that requires innovative
approaches.3 Stereoselective cascade reactions able to generate
molecular complexity and functional diversity in a limited
number of operations can bring a unique answer to these two
synthetic problems.4 Here, we propose a conceptually new
shortcut solution based on the combination of our recent
enantioselective functionalization of allylic alcohols5 with an
intramolecular product-driven acyl transfer.6 The initial allylic
alcohol transformation proceeds through an iron-catalyzed
hydrogen transfer generating the corresponding α,β-unsaturated
aldehyde that upon iminium activation gives the expected
enantioselective Michael addition. The resulting chiral transient
saturated aldehyde is then reduced by the in situ generated iron
hydride to the key hydroxy diketone A, triggering the final retro-
Claisen step.
The strategy for an efficient acyl transfer relies on the high
reactivity of 1,3-diketones toward Michael addition.7 These pro-
nucleophiles bear a noninnocent activating acyl group (R2CO)
Received: March 19, 2014
Published: May 21, 2014
© 2014 American Chemical Society
2802
dx.doi.org/10.1021/ol500821c | Org. Lett. 2014, 16, 2802−2805

Organic Letters
Letter
that should be transferred afterward via a retro-Claisen reaction
from lactol B, cyclic isomer of adduct A resulting from the dual
catalysis (Scheme 1b). This unprecedented, enantioselective
overall quadruple cascade, involving dual borrowing hydrogen/
iminium catalysis/retro-Claisen fragmentation, would result in a
formal fully atom-, step- and redox-economic chemoselective
reductive insertion of a 3-substituted allylic alcohol into a C(O)−
C bond of the 1,3-diketone substrates (Scheme 1b).8
Scheme 3. Scope of the Cascade Dual Iron/Amine Catalysis
and Acyl Transfer
To validate the proof of concept for this domino process, dual
iron−amine catalysis/acyl transfer, we subjected different pro-
nucleophiles 1 and crotyl alcohol (2a) to activation by iron
cyclopentadienone [Fe] as borrowing hydrogen complex,9
Me3NO to generate the active iron catalyst (liberating CO2
and Me3N)9d and cat1 in toluene (Scheme 2).
Scheme 2. Influence of the Electron-Withdrawing Group over
the Reactivity
Activation by strong electron-withdrawing groups (EWG =
NO2, 1a or COCF3, 1b) resulted in no product formation. While
supposedly more reactive than ketoesters, the fact that these two
molecules do not react in the dual catalytic system suggests that
they probably inhibit the initial hydrogen transfer by
coordinating to the iron catalyst. Gratifyingly, turning to 1,3-
diketone 1c (EWG = COPh), the designed process cleanly
occurred directly providing, under mild conditions, the protected
alcohol 3a in 59% yield and a promising 78:22 er (Scheme 2).
Encouraged by the observed promising cascade using
symmetrical 1,3-diketone 1c and given the synthetic potential
of this process, reaction conditions were carefully optimized.
Extensive screening (see the Supporting Information) revealed
that the use of 13 mol % of silylated diarylprolinols either cat1 or
cat210 depending on the diketone used, xylenes as solvent and
Me3NO·2H2O as initiator gave the best results (Scheme 3).
Under these conditions, using cat2, 3a was obtained in 96% yield
and 90:10 er.
With these optimized conditions in hand, systematic variation
of the substrates was then carefully examined. Gratifyingly,
structurally different diketones 1c−l and allyl alcohols 2a−f
provided the expected rearranged products under mild
conditions (Scheme 3). It must be pointed out that while for
1c the fragmentation was spontaneous at room temperature,
using other diketones, conversion of the intermediate lactol B
was sometimes incomplete at this temperature. Fortunately,
upon evaporation of the solvent with the rotavapor bath adjusted
at 40 °C, clean evolution to the rearranged form was observed
after 5−15 min. Using crotyl alcohol (2a) (R4 = Me), the nature
of the diketone could be modified efficiently giving rise to the
cascade products in yields ranging from 34 to 96%. 2,4-
Pentanedione furnished the expected protected product 3b in
90:10 er and 94% yield. Electron-withdrawing or donating
groups (pF, or pOMe) could be placed on symmetrical aryl
diketones giving the cascade products 3c and 3d in 86:14 and
90:10 er, respectively. When starting from unsymmetrical cyclic
α-acetyl ketones, only the chemoselective migration of the acetyl
group was observed in this transformation (3e−g). A moderate
diastereoselectivity was obtained, while the er ranged from 80:20
to 87.5:12.5 for the formation of 3e−g. The moderate
diastereocontrol might be explained by the nature of the
diastereocontrolling step, namely the protonation of the product
after the fragmentation.
Other acyclic aliphatic-1,3-diketones could also be efficiently
applied in the cascade giving the products 3h−l in 89:11 to 96:4
er. Use of unsymmetrical monoacetyl diketones proceeded with
variable chemoselectivity, always in favor of the acetyl migration,
leading to 3i−l with good er. The more substituted the R1 group,
the higher the selectivity.11 For example, with tert-butyl methyl
1,3-diketone, the acetyl migration is total leading to compounds
3i,l, while a separable 1.6/1 mixture of 3j and 3k is obtained in an
excellent enantiocontrol (92.5:7.5 to 96:4 er) from the
corresponding cyclohexyl methyl 1,3-diketone.12 Finally, the
2803
dx.doi.org/10.1021/ol500821c | Org. Lett. 2014, 16, 2802−2805

Organic Letters
Letter
nature of the allylic alcohol could be varied leading to structurally
different products 3l−q, in 32 to 84% yield and 70:30 to 96:4 er.
The combination of the dual catalytic process with the acyl
transfer is interesting in terms of its mechanism. As already
mentioned, the presence of the electron-withdrawing ketone is
absolutely required for the spontaneous fragmentation since
ketoesters do not rearrange.5a But if one reduces too strongly the
electron density on the pro-nucleophile, no more reaction occurs
(Scheme 2). This suggests that electron-poor diketones or nitro-
ketone coordinate to the metal, inhibiting its catalytic proper-
ties.13
Scheme 5. Synthetic Applications of the Cascade
Interestingly, when Me3NO is removed from the mixture and
replaced by UV activation (350 nm to remove one CO from the
initial iron complex),14 the overall dual-catalysis/acyl transfer
cascade does take place in 47% yield (Scheme 4a). In addition,
Scheme 4. Cascade Process under UV Activation and
Difference in Reactivity between (E)-2c and (Z)-2c
by diastereoselective acid promoted cyclization18 efficiently
gives, as the major product (10:1 dr), the expected cyclic pyran 4.
We next turned to the synthesis of key fragments of two
distinct bioactive natural products with both remarkable
cytotoxic activities, namely lyconadin A and apratoxin A.19
From 3b, obtained in 90:10 er by the cascade from 2,4-
pentanedione and crotyl alcohol (2a), protecting group exchange
provides 5 in a straightforward three-step sequence (Scheme 5b).
Compound 5 was previously applied to the synthesis of
lyconadin A, but its preparation required six steps from expensive
(R)-1-methyl hydrogen 3-methylglutarate.20
Finally, the C1−C5 fragment of apratoxin A was then prepared
according to the synthetic sequence presented in Scheme 5c.
Following the disclosed cascade, (+)-3i, obtained in 95:5 er using
(R)-cat1 to obtain the appropriate configuration at C3, was
reduced with (R)-CBS directly, giving upon hydrolysis in one
single operation the corresponding diol 6 (99:1 dr and 99.5:0.5
er). Compound 6 is a known precursor of apratoxin A but
previously required 5−11 steps depending on the strategy used
for its preparation.2g−i
In conclusion, we have developed the first example of
enantioselective dual iron−amine catalysis acyl-transfer cascade
for the synthesis of enantioenriched γ-chiral alcohols directly
from simple achiral diketones and allylic alcohols. The efficient
combination of the different activation modes, an iron-catalyzed
borrowing hydrogen and an iminium catalysis together with the
final retro-Claisen fragmentation, allow the preparation of these
valuable building blocks in up to 96% yield and 96:4 er. The
designed process occurs under mild conditions, fulfills the
principles of redox, step- and atom-economies, and in addition
avoids the handling of toxic α,β-unsaturated aldehydes. This
strategy gives access in one single operation to synthetically
useful functionalized building blocks possessing distinct
functions directly from commercially or easily prepared
substrates. The efficiency of this approach was demonstrated in
the rapid synthesis of several natural product fragments or
odorant molecules. Given the considerable synthetic shortcuts
provided by this methodology, we are convinced that it will in the
near future find widespread applications in natural product
synthesis.
84.5:15.5 er was observed under these conditions, a result close
to the one obtained using Me3NO as initiator. This suggests that
beside from generating the active iron complex, Me3NO and the
Me3N generated in this iron activation do not take part in the
different steps of the catalytic cycle. This probably indicates that
it is the enol form of the diketone (observed by NMR) and not a
deprotonated form, which adds to the iminium ion intermediate.
Similarly, the acyl transfer occurs simply by thermal activation,
without the involvement of either Me3N or cat2. Control
experiments comparing the reactivity profile of both E and Z
starting allylic alcohol 2c gave the same enantiomer of the
product 3n in, respectively, 83.5:16.5 and 70:30 er (Scheme 4b).
This indicates that in the case of (Z)-allyl alcohols, even though
the lifetime of the α,β-unsaturated aldehyde is short, isomer-
ization to the parent (E)-α,β-unsaturated aldehyde via dienamine
formation predominantly occurs prior to nucleophilic Michael
addition.15
Given the apparent interest of the series of functionalized γ-
chiral protected alcohols readily obtained from raw materials
through this cascade, wide arrays of potential synthetic
applications seem possible. To fully demonstrate this potential,
we decided to apply it to the synthesis of structurally different
attractive natural products fragments or odorant molecule
(Scheme 5).
As a first application of the discovered cascade, we synthesized
the (2S,4R) isomer of Doremox, the most odorant isomer
(Scheme 5a).16 This odorant molecule was previously prepared
in an enantioselective manner in six- to nine-step synthetic
sequences.2a From 3a, obtained in 90:10 er from easily available
starting materials, reduction of both ketone and ester17 followed
2804
dx.doi.org/10.1021/ol500821c | Org. Lett. 2014, 16, 2802−2805

Organic Letters
Letter
Kwong, F. Y. Org. Biomol. Chem. 2011, 9, 7997. (f) Liu, Y.; Wang, Y.;
Song, H.; Zhou, Z.; Tang, C. Adv. Synth. Catal. 2013, 355, 2544.
(g) Grenning, A. J.; Van Allen, C. K.; Maji, T.; Lang, S. B.; Tunge, J. A. J.
Org. Chem. 2013, 78, 7281.
(7) For examples of iminium-catalyzed Michael additions of 1,3-
diketones to α,β-unsaturated aldehydes, see: (a) Franke, P. T.; Richter,
B.; Jørgensen, K. A. Chem.Eur. J. 2008, 14, 6317. (b) Rueping, M.;
Sugiono, E.; Merino, E. Chem.Eur. J. 2008, 14, 6329. (c) Yu, C.;
Zheng, F.; Ye, H.; Zhong, W. Tetrahedron 2009, 65, 10016. (d) Lathrop,
S. P.; Rovis, T. J. Am. Chem. Soc. 2009, 131, 13628. (e) Wu, X.; Fang, H.;
Liu, Q.; Nie, L.; Chen, J.; Cao, W.; Zhao, G. Tetrahedron 2011, 67, 7251.
ASSOCIATED CONTENT
* Supporting Information
Reaction optimization, experimental procedures, character-
ization of compounds, and spectra. This material is available
free of charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Authors
■
*E-mail: adrien.quintard-inv@univ-amu.fr.
*E-mail: jean.rodriguez@univ-amu.fr.
Notes
(f) Polat, M. F.; Hettmanczyk, L.; Zhang, W.; Szabo, Z.; Franzen
́
, J.
ChemCatChem 2013, 5, 1334.
(8) For recent reviews on dual organocatalysis, see: (a) Allen, A. E.;
MacMillan, D. W. C. Chem. Sci. 2012, 3, 633. (b) Du, Z.; Shao, Z. Chem.
Soc. Rev. 2013, 42, 1337. For selected recent reviews on borrowing
hydrogen, see: (c) Hamid, M. H. S. A.; Slatford, P. A.; Williams, J. M. J.
Adv. Synth. Catal. 2007, 349, 1555. (d) Dobereiner, G. E.; Crabtree, R.
H. Chem. Rev. 2010, 110, 681. (e) Gunanathan, C.; Milstein, D. Science
2013, 341, 6143.
(9) For a review, see: (a) Quintard, A.; Rodriguez, J. Angew. Chem., Int.
Ed. 2014, 53, 4044. For selected examples, see: (b) Casey, C. P.; Guan,
H. J. Am. Chem. Soc. 2007, 129, 5816. (c) Thorson, M. K.; Klinkel, K. L.;
Wang, J.; Williams, T. J. Eur. J. Inorg. Chem. 2009, 295. (d) Moyer, S. A.;
Funk, T. Tetrahedron Lett. 2010, 51, 5430. (e) Coleman, M. G.; Brown,
A. N.; Bolton, B. A.; Guan, H. Adv. Synth. Catal. 2010, 352, 967.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The Agence Nationale pour la Recherche (ANR-13-PDOC-
0007-01), the COST ORCA (CM0905), the Centre National de
la Recherche Scientifique (CNRS) and the Aix-Marseille
Universite
We warmly thank Marion Jean and Nicolas Vanthuyne (Aix-
Marseille Universite) for chiral-phase HPLC analysis.
́
are gratefully acknowledged for financial support.
́
REFERENCES
■
(f) Knolker, H.-J.; Heber, J.; Mahler, C. H. Synlett 1992, 1002.
̈
(1) For reviews on economy in synthesis: (a) Trost, B. M. Science 1991,
254, 1471. (b) Wender, P. A.; Bi, F. C.; Gamber, G. G.; Gosselin, F.;
Hubbard, R. D.; Scanio, M. J. C.; Sun, R.; Williams, T. J. Pure Appl. Chem.
2002, 74, 25. (c) Burns, N. Z.; Baran, P. S.; Hoffmann, R. W. Angew.
Chem., Int. Ed. 2009, 48, 2854.
(2) For Doremox syntheses, see: (a) Brenna, E.; Fuganti, C.; Ronzani,
S.; Serra, S. Can. J. Chem. 2002, 80, 714. (b) Kimura, M.; Ezoe, A.; Mori,
M.; Iwata, K.; Tamaru, Y. J. Am. Chem. Soc. 2006, 128, 8559.
(g) Knolker, H.-J.; Baum, E.; Goesmann, H.; Klauss, R. Angew. Chem.,
̈
Int. Ed. 1999, 38, 2064.
(10) (a) Marigo, M.; Wabnitz, T. C.; Fielenbach, D.; Jørgensen, K. A.
Angew. Chem., Int. Ed. 2005, 44, 794. (b) Hayashi, Y.; Gotoh, H.;
Hayashi, T.; Shoji, M. Angew. Chem., Int. Ed. 2005, 44, 4212.
(11) Use of 1-phenyl-1,3-butanedione (R1 = Ph, R2 = Me, R3 = H) gave
the product in a 1/1 mixture of Bz and Ac products.
(12) The difference in er between 3j and 3k might reflect a different
enantiocontrol between the two diastereomers formed prior to the
rearrangement.
(c) Coulombel, L.; Weïver, M.; Dunach, E. Eur. J. Org. Chem. 2009,
̃
5788. For the first lyconadin A synthesis, see: (d) Beshore, D. C.; Smith,
A. B., III. J. Am. Chem. Soc. 2007, 129, 4148. For the most notable other
syntheses of this natural product, see: (e) Bisai, A.; West, S. P.; Sarpong,
R. J. Am. Chem. Soc. 2008, 130, 7222. (f) Nishimura, T.; Unni, A. K.;
Yokoshima, S.; Fukuyama, T. J. Am. Chem. Soc. 2011, 133, 418. For the
first synthesis of apratoxin A, see: (g) Chen, J.; Forsyth, C. J. J. Am. Chem.
Soc. 2003, 125, 8734. (h) Chen, J.; Forsyth, C. J. Proc. Nat. Acad. Sci.
2004, 101, 12067. For the most notable other syntheses of this natural
product, see: (i) Ma, D.; Zou, B.; Cai, G.; Hu, X.; Liu, J. O. Chem.Eur.
J. 2006, 12, 7615. (j) Zou, B.; Wei, J.; Cai, G.; Ma, D. Org. Lett. 2003, 5,
3503. (k) Xu, Z.; Chen, Z.; Ye, T. Tetrahedron: Asymmetry 2004, 15, 355.
(l) Doi, T.; Numajiri, Y.; Munakata, A.; Takahashi, T. Org. Lett. 2006, 8,
531. (m) Gilles, A.; Martinez, J.; Cavelier, F. C. R. Chim. 2011, 14, 437.
(3) For the limited direct addition of ketones to α,β-unsaturated
aldehydes: (a) Li, W.; Wu, W.; Yang, J.; Liang, X.; Ye, J. Synthesis 2011, 7,
1085. For an example of addition using silyl enol ethers, see: (b) Wang,
W.; Li, H.; Wang, J. Org. Lett. 2005, 7, 1637.
(4) For selected reviews on cascade and organocascade reactions, see:
(a) Nicolaou, K. C.; Edmonds, D. J.; Bulger, P. G. Angew. Chem., Int. Ed.
2006, 45, 7134. (b) Nicolaou, K. C.; Chen, J. S. Chem. Soc. Rev. 2009, 38,
2993. (c) Grondal, C.; Jeanty, M.; Enders, D. Nat. Chem. 2010, 2, 167.
(d) Bonne, D.; Constantieux, T.; Coquerel, Y.; Rodriguez, J. Chem.
Eur. J. 2013, 19, 2218. (e) Volla, C. M. R.; Atodiresei, I.; Rueping, M.
Chem. Rev. 2014, 114, 2390.
(13) Unfortunately, attempts at observing such coordination between
a 1,3-diketone and the activated iron complex have remained
inconclusive.
(14) Berkessel, A.; Reichau, S.; von der Hoh, A.; Leconte, N.; Neudorfl,
̈
̈
J.-M. Organometallics 2011, 30, 3880.
(15) For the observation of aldehyde isomerization, see: Ouellet, S. G.;
Tuttle, J. B.; MacMillan, D. W. C. J. Am. Chem. Soc. 2005, 127, 32. Given
all these observations, the mechanism of the cascade is described in
detail in the Supporting Information.
(16) Doremox is sold as a racemic mixture. For the initial patent, see:
Firmenich. US Patent 005 219 836A, June 15, 1993.
(17) Under these conditions, reduction performed at rt gives 1.3:1 dr
while reducing the temperature to 0 °C gives 1.55:1 dr of the
corresponding alcohol.
(18) For the origin of the diastereocontrol in this type of cationic
cyclization, see ref 1c.
(19) For the isolation of lyconadin A, see: (a) Kobayashi, J.; Hirasawa,
Y.; Yoshida, N.; Morita, H. J. Org. Chem. 2001, 66, 5901. For the
isolation of apratoxin A, see: (b) Leusch, H.; Yoshida, W. Y.; Moore, R.
E.; Paul, V. J.; Corbett, T. H. J. Am. Chem. Soc. 2001, 123, 5418.
(20) See ref 2d. The opposite enantiomer of 5 was also prepared in six
steps and applied to the synthesis of amphidinolide B: Gopalarathnam,
A.; Nelson, S. G. Org. Lett. 2006, 8, 7.
(5) (a) Quintard, A.; Constantieux, T.; Rodriguez, J. Angew. Chem., Int.
Ed. 2013, 52, 12883. For another example of allylic alcohol
functionalization by hydrogen autotransfer, see: (b) Black, P. J.;
Harris, W.; Williams, J. M. J. Angew. Chem., Int. Ed. 2001, 40, 4475.
(6) For recent successful examples of such acyl transfer cascades, see:
(a) Grenning, A. J.; Tunge, J. A. Angew. Chem., Int. Ed. 2011, 50, 1688.
(b) Grenning, A. J.; Tunge, J. A. J. Am. Chem. Soc. 2011, 133, 14785.
(c) Lu, R.-J.; Yan, Y.-Y.; Wang, J.-J.; Du, Q.-S.; Nie, S.-Z.; Yan, M. J. Org.
Chem. 2011, 76, 6230. (d) Gao, Y.; Ren, Q.; Siau, W.-Y.; Wang, J. Chem.
Commun. 2011, 47, 5819. (e) Li, P.; Hing Chan, S.; Chan, A. S. C.;
2805
dx.doi.org/10.1021/ol500821c | Org. Lett. 2014, 16, 2802−2805