Cyclopentanoids by uncatalyzed intramolecular carbonyl ene (ICE) reaction of α-keto esters

Article abstract of DOI:10.1021/ol501204m

The uncatalyzed intramolecular carbonyl ene (ICE) reaction of substituted ε,ζ-unsaturated α-keto esters to terpenoid-related building blocks has been studied. We found a beneficial effect of a silyl substituent at the ene segment on the kinetics of the ICE reaction. A generalizable and scalable synthesis of ε,ζ-unsaturated α-keto esters from allylic alcohols was developed.

Full text of DOI:10.1021/ol501204m

Letter
pubs.acs.org/OrgLett
Cyclopentanoids by Uncatalyzed Intramolecular Carbonyl Ene (ICE)
Reaction of α‑Keto Esters
David Tymann,* Andre Kluppel, Wolf Hiller, and Martin Hiersemann*
́
̈
Fakultat Chemie und Chemische Biologie, Technische Universitat Dortmund, 44227 Dortmund, Germany
̈
̈
S
* Supporting Information
ABSTRACT: The uncatalyzed intramolecular carbonyl ene (ICE) reaction of
substituted ε,ζ-unsaturated α-keto esters to terpenoid-related building blocks has been
studied. We found a beneficial effect of a silyl substituent at the ene segment on the
kinetics of the ICE reaction. A generalizable and scalable synthesis of ε,ζ-unsaturated α-
keto esters from allylic alcohols was developed.
he uncatalyzed intramolecular carbonyl ene (ICE) reaction
T
of an ε,ζ-unsaturated α-keto ester (1) has recently been
exploited to access the highly substituted cyclopentanoid 2,
which was pivotal in the total synthesis of jatrophane
diterpenoids (e.g., 5) from plants of the genus Euphorbia (Figure
1).1
Notably, the catalyst-free ICE reaction to afford 2 in decane
required 180 °C, and the stereochemical outcome of the bond-
forming event was thermodynamically controlled.2 To further
probe for the scope of the ICE reaction, including an application
in a projected total synthesis of the marine sesquiterpenoid
menverin C (4),3 we have synthesized a range of ε,ζ-unsaturated
α-keto esters (1a−i) and studied the influence of substituents on
the ICE reaction. In particular, we envisioned that a SiMe3
substituent at the ene segment would increase the thermody-
namic driving force of the ICE reaction and, assuming that the
Bell−Evans−Polanyi principle is operative, favorably bias the
kinetics as well.4−6 The results of our study are summarized in
this letter.
Despite the many methods known for the preparation of the α-
oxo ester functionality,7 a general and robust synthetic excess to
substituted ε,ζ-unsaturated α-keto esters 1 that would serve the
purpose of this study was not available. Therefore, we were
prompted to develop a route starting from readily available allylic
alcohols (Table 1).8 On the basis of the work of Schlaf as well as
Wei,9,10 we subjected the secondary allylic alcohols 6b−d to a
telescoped process consisting of (bphen)Pd(tfa)2-catalyzed vinyl
ether metathesis and subsequent Hurd−Claisen rearrange-
ment11 under modified conditions to deliver the corresponding
aldehydes in 85−89% yield.12−14 Unfortunately, the telescoped
process was low-yielding (25%) for the allylic alcohol 6a, and the
vinyl ether metathesis failed for ethyl 1-propenylether; hence, a
sequence of Johnson−Claisen rearrangement15 and subsequent
DIBAH reduction was utilized to access the aldehydes 7a and 7e.
With the requisite aldehydes 7a−e in hand, we turned toward the
introduction of the α-keto ester moiety. Thus, Horner−
Wadsworth−Emmons-type olefination using ethyl 2-acetoxy-2-
Figure 1. Intramolecular carbonyl ene (ICE) reaction for the synthesis
of terpenoid building blocks.
(diethoxyphosphoryl)acetate16 (8) under previously reported
conditions provided the vinyl acetates 9a−e (Table 1, entries 1−
5).1 In our hands, this procedure proved to be generalizable and
scalable for aliphatic aldehydes.17
We turned next to elaboration of the vinyl acetates 9a−e into
the desired α-keto esters 1a−g (Table 2). Using 1b for initial
exploration and optimization, we first tested our established
transesterification conditions (0.1 equiv K2CO3, MeOH, 0 °C),1
but without measurable success. Instead of the desired α-keto
ester 1b, we detected the product of a homocondensation
process between the α-keto ester 1b and its potassium enolate,
which arises as primary product of the potassium methoxide-
triggered transesterification. As expected, shortening the reaction
time or lowering the reaction temperature suppressed homo-
condensation, albeit at the expense of an incomplete conversion.
Received: April 28, 2014
Published: August 1, 2014
© 2014 American Chemical Society
4062
dx.doi.org/10.1021/ol501204m | Org. Lett. 2014, 16, 4062−4065

Organic Letters
Letter
Table 1. Synthesis of Vinyl Acetates from Allylic Alcohols via
Aliphatic Aldehydes
Scheme 1. Structural Diversification: Synthesis of the α-Keto
Ester 1h,i
a
procedure was general and, furthermore, could be extended to
other electrophiles as well (entries 6 and 7). Notably, the
aminomethylation using Eschenmoser’s salt20 enabled elimi-
nation already during workup without the necessity for
quatenarization and, hence, formation of the β-methylidene α-
keto ester 1f. However, the methylation (MeI, Me2SO4) of the
proposed magnesium enolate 10 of 1b failed to afford the desired
β-methyl-substituted α-keto ester (1h).
In an effort to access an even greater diversity of the desired
ε,ζ-unsaturated α-keto esters (1), we sought to implement a 1,4-
reduction to convert the β-methylidene α-keto ester 1f into the
elusive β-methyl-substituted α-keto ester 1h (Scheme 1). After
much experimentation, we found that a two-step procedure
consisting of 1,4-hydrosilylation followed by TASF-catalyzed
silyl enol ether cleavage delivered 1h in an useful yield (73%) on
an explorative scale.21,22 Also on a small scale, the δ-methylidene
α-keto ester 1i was made accessible by treatment of the phenyl
selenide 1d with m-CPBA (1 equiv).23
With the requisite nine α-keto esters 1a−i in hand, the
influence of substitutents on the outcome of the ICE reaction
was explored (Table 3). All experiments were performed using
decane as the solvent in a sealable glass pressure tube with a
Teflon screw cap. Thus, upon heating (180 °C) the reference
compound 1a lacking the silyl substituent at the ene, we obtained
the cyclopentanoid 3a as a single cis-configured diastereomer in
low yield (35%) along with substantial amounts of starting
material (entry 1).
a
BPhen = 4,7-diphenyl-1,10-phenanthroline, bathophenanthroline.
b
Determined by NMR; double-bond configuration assigned on the
c
basis of 13C,1H long-range coupling constant analysis. Yield (isolated
mass) after purification by chromatography. Prepared by Johnson−
Claisen rearrangement and DIBAH reduction; see the Supporting
Information. Amine = DABCO (0.1 equiv). Amine = DABCO (0.14
equiv). Amine = 2,6-lutidine (1 equiv).
d
e
f
g
Table 2. Vinyl Acetate Elaboration: Synthesis of ε,ζ-
Unsaturated α-Keto Esters
Subjecting the α-keto ester 1b featuring the Me3Si-substituted
ene to the identical conditions markedly increased the yield
(81%), and 3b was obtained in high diastereoselectivity (dr =
95:5) and as a E:Z = 91:9 mixture of double bond isomers with
respect to the vinyl silane moiety (entry 2).24 To shed some light
on the influence on the SiMe3 substituent on the kinetics of the
ICE reaction, the conversion was tracked by 1H NMR
spectroscopy for 1a and 1b. The resulting conversion-versus-
time profiles nicely illustrate the improved kinetics of the ICE
reaction of 1b compared to 1a.25 For instance, after 1 d at 180 °C,
we obtained a 15% conversion of 1a to 3a but a 64% conversion
of 1b to 3b. The substrate-induced diastereoselectivity was
investigated next.26 The ICE reaction (180 °C, 3 d) of the δ-
methyl branched α-keto ester 1c delivered 3c in very good yield
(86%) and with a significant (dr = 89:11) induced
diastereoselectivity (entry 3).27 A comparable result (92%, dr =
91:9) was obtained for the δ-PhSeCH2-substituted α-keto ester
1d (entry 4).28 The δ-methylene α-keto ester 1i very reluctantly
underwent the ICE reaction, and extensive decomposition was
observed rendering this transformation synthetically unattrac-
tive. Hence, and considering the efficiency of the ICE reaction of
1d to 3d, we decided to access the desired cyclopentanoid 3i by a
post-ICE transformation of 3d (vide infra). In striking contrast to
the synthetically useful induced diastereoselectivities obtained
a
Isolated yield (100−900 mg) after purification by chromatography.
Next, a Brønsted acid mediated process (1 equiv PTSA, MeOH,
rt) was explored; however, we obtained low yields (45−55%)
and observed decomposition. We then opted for conditions that
would prevent the coexistence of the α-keto ester and its enol or
enolate. Thus, exposure of the vinyl acetate 9b to commercially
available MeMgBr solution (1 M in Et2O) with subsequent
aqueous workup afforded the α-keto ester 1b in a synthetically
useful yield (Table 2, entry 2).18,19 The success of this procedure
probably hinges on the intermediacy of the stable chelated
magnesium enolate 10. As outlined in Table 2, the developed
4063
dx.doi.org/10.1021/ol501204m | Org. Lett. 2014, 16, 4062−4065

Organic Letters
Letter
Table 3. Substituent Effects on the ICE Reaction
Scheme 2. Post-ICE Transformations of 3b and 3d
accompanied by HBr elimination and concomitant protodesily-
lation. The ICE reaction (180 °C, 3 d) of the β-methyl α-keto
ester 1h delivered 3h in excellent yield (94%) and good induced
diastereoselectivity (dr = 86:14). Heating (140 °C) of the β-
methylidene α-keto ester 1f led to a mixture consisting of
unconsumed 1f (14%), the expected cyclopentanoid 3f (50%), as
well as a homodimer from a hetero-Diels−Alder reaction
between 1f and 3f (30%) as a single diastereomer (entry 6).30
Having gained access to terpenoid-related building blocks (3)
by ICE reaction of ε,ζ-unsaturated α-keto esters (1), we
subsequently focused on the development of post-ICE trans-
formations of the vinyl silane moiety (Scheme 2). Using 3b (E:Z
= 91:9) for initial exploration, we found that a mixture of
TMSOTf and Ac2O mediates acetylation of the tertiary hydroxyl
group as well as protodesilylation to deliver the vinyl cyclo-
pentanoid 11 (95%).31 Iododesilylation with retention of double
bond configuration required upstream LAH reduction of the α-
hydroxy moiety to the diol 12 (94%) and acetalization to afford
the acetonide 15 (99%); subsequent treatment of 15 with NIS
delivered the vinyl iodide 16 (93%, E:Z = 91:9).32 A stepwise
bromodesilylation with inversion of the double bond config-
uration was accomplished by initial epoxidation of 3b to afford
the silyl oxirane 13 (93%) as a mixture of diastereomers;
treatment of 13 with MgBr2 delivered the crude bromohydrin
which was subjected to a telescoped process consisting of
mesylation and TBAF-mediated elimination to provide the vinyl
bromide 14 (76%, E:Z = 5:95).33 Finally, and thereby completing
the circle to our initial proposal (Figure 1), treatment of the
selenide 3d with 1 equiv of m-CPBA under carefully optimized
conditions delivered the methylene cyclopentanoid 3i (79%, dr
>95:5, E:Z = 93:7) representing a promising building block for
menverin C (4).
a
Ratios determined by NMR; relative configuration assigned by NOE
b
experiments. Yield (isolated mass) after purification by chromatog-
c
raphy. Unconsumed 1a (25%) was reisolated. 1H NMR of the
reaction mixture and TLC (baseline) indicates formation of
unidentified byproducts. For a procedure on larger scale, see the
Supporting Information. Reaction conducted at 140 °C. Along with
30% (15 mg) of a dimer from a hetero-Diels−Alder reaction between
1f and 3f as well as 14% (7 mg) of 1f; see Supporting Information.
d
e
for 1c and 1d, the ICE reaction (180 °C, 3 d) of the γ-methyl
branched α-keto ester 1e (entry 5), despite being high-yielding
(93%), proceeded without notable induced diastereoselectivity
(dr = 53:47).29 We continued our study with the β-bromo- and
β-methyl-substituted α-keto esters 1g and 1h (entries 6 and 7).
The ICE reaction (77 h, 140 °C) of 1g delivered 3g in good yield
(80%) and with a moderate induced diastereoselectivity (dr =
82:18). On one hand, the conversion of 1g to 3g slowly
proceeded already at 120 °C, indicating that the β-bromo
substituent exerts the most significant rate effect yet detected; on
the other hand, however, the ICE reaction of 1g at 180 °C was
In summary, substituent effects on the uncatalyzed intra-
molecular carbonyl ene (ICE) reaction of ε,ζ-unsaturated α-keto
esters were investigated. Most notably, the presence of a silyl
4064
dx.doi.org/10.1021/ol501204m | Org. Lett. 2014, 16, 4062−4065

Organic Letters
Letter
(12) In our hands, the optimum conditions for the Pd(II)-catalyzed
vinyl ether metathesis were dependent on the substrate structure,
particularly with respect to the nature of the amine base.
(13) Hg(OAc)2 or Hg(tfa)2 (0.1−0.2 equiv) were less effective
catalysts; see: Watanabe, W. H.; Conlon, L. E. J. Am. Chem. Soc. 1957,
79, 2828−2833.
substituent at the ene exerts an enabling beneficial effect on the
kinetics of the ICE reaction. High simple diastereoselectivities,
enforced by the concertedness of the bond reorganization
process, were observed. The extent of the induced diaster-
eoselectivity depends on the position of substituents at the tether
atoms between ene and enophile. Central to the study of the ICE
reaction was the development of a robust and generalizable
synthetic access to aliphatic α-keto esters from allylic alcohols.
(14) A catalytic cycle for the Pd(II)-catalyzed vinyl ether metathesis is
proposed in the Supporting Information.
(15) Johnson, W. S.; Werthemann, L.; Bartlett, W. R.; Brocksom, T. J.;
Li, T.-T.; Faulkner, D. J.; Petersen, M. R. J. Am. Chem. Soc. 1970, 92,
741−743.
(16) Schmidt, U.; Langner, J.; Kirschbaum, B.; Braun, C. Synthesis
1994, 11, 1138−1140.
(17) Attempts to subject a methyl ketone to this olefination reaction
were unsuccessful.
(18) The α-keto esters 1a and 1b, in our hands, possessed the
propensity to undergo homocondensation during chromatographic
purification. See also: Boehlow, T. R.; Harburn, J. J.; Spilling, C. D. J.
Org. Chem. 2001, 66, 3111−3118. In order to circumvent this problem
on larger scale, the α-keto ester 1b was subjected to the conditions of the
ICE reaction unpurified; for instance, 1.9 g of 9b delivered 75% of 3b
from 9b.
ASSOCIATED CONTENT
* Supporting Information
■
S
Procedures, characterization data, and copies of spectra. This
material is available free of charge via the Internet at http://pubs.
acs.org.
AUTHOR INFORMATION
Corresponding Authors
■
*E-mail: david.tymann@tu-dortmund.de.
*E-mail: martin.hiersemann@tu-dortmund.de.
(19) In contrast to the transesterification conditions (K2CO3, MeOH),
the ethyl instead of the methyl α-keto ester is formed under the
conditions of the nucleophilic vinyl acetate degradation (MeMgBr).
(20) Schreiber, J.; Maag, H.; Hashimoto, N.; Eschenmoser, A. Angew.
Chem., Int. Ed. Engl. 1971, 10, 330−331.
(21) Optimized reductive procedure according to: Tsuda, T.; Hayashi,
T.; Satomi, H.; Kawamoto, T.; Saegusa, T. J. Org. Chem. 1986, 51, 537−
540.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
Financial support by the TU Dortmund is gratefully acknowl-
edged.
■
(22) Attempted HF-mediated silyl enol ether cleavage led to
decomposition. We propose a TASF-catalyzed 1,3-retro-Brook
rearrangement of the silyl enol ether to deliver the β-trimethylsilyl α-
keto ester, followed by protodesilylation during aqueous workup. A
catalytic cycle is proposed in the Supporting Information.
(23) Optimized procedure according to: Reich, H. J.; Shah, S. K. J. Am.
Chem. Soc. 1975, 97, 3250−3252.
(24) The relative configuration was assigned based on NOE
experiments with 12 (Scheme 2); see Supporting Information. NOE
experiments using the corresponding diol accessible from 3a support the
assignment. The relative configuration of 3c−h was assigned in analogy.
(25) See the Supporting Information for details. We currently tend to
believe that the rate-accelerating effect of the SiMe3 group is caused by
an increased thermodynamic driving force, rather than an intrinsic
transition-state stabilizing effect of the SiMe3 substituent.
(26) Kinetic studies by NMR indicate that the reported substrate-
induced diastereoselectivities are thermodynamically controlled. A
combined experimental and computational study on this issue will be
reported in due course.
REFERENCES
■
(1) (a) Helmboldt, H.; Kohler, D.; Hiersemann, M. Org. Lett. 2006, 8,
̈
1573−1576. (b) Schnabel, C.; Hiersemann, M. Org. Lett. 2009, 11,
2555−2558. (c) Helmboldt, H.; Hiersemann, M. J. Org. Chem. 2009, 74,
1698−1708.
(2) Schnabel, C.; Sterz, K.; Muller, H.; Rehbein, J.; Wiese, M.;
̈
Hiersemann, M. J. Org. Chem. 2011, 76, 512−522.
(3) Zhang, W.; Guo, Y.-W.; Mollo, E.; Cimino, G. Helv. Chim. Acta
2004, 87, 2919−2925.
(4) For the first report on the uncatalyzed ICE of α-keto esters, see:
Hiersemann, M. Eur. J. Org. Chem. 2001, 483−491.
(5) For uncatalyzed intermolecular ICE reactions of allyl silanes, see:
(a) Abel, E. W.; Rowley, R. J. J. Organomet. Chem. 1975, 84, 199−229.
(b) Laporterie, A.; Dubac, J.; Lesbre, M. J. Organomet. Chem. 1975, 101,
187−208. (c) Beak, P.; Song, Z.; Resek, J. E. J. Org. Chem. 1992, 57,
944−947. For an uncatalyzed intramolecular ICE reaction of an allenyl
silane, see: Jin, J.; Smith, D. T.; Weinreb, S. M. J. Org. Chem. 1995, 60,
5366−5367.
(6) For catalytic asymmetric Lewis acid catalyzed ICE reactions of α-
keto esters, see: (a) Kaden, S.; Hiersemann, M. Synlett 2002, 1999−
2002. (b) Yang, D.; Yang, M.; Zhu, N. Org. Lett. 2003, 5, 3749−3752.
(c) Zhao, J.-F.; Tsui, H.-Y.; Wu, P.-J.; Lu, J.; Loh, T.-P. J. Am. Chem. Soc.
2010, 132, 10242−10244.
(7) (a) Nimitz, J. S.; Mosher, H. S. J. Org. Chem. 1981, 46, 211−213.
(b) Wassermann, H. H.; Ho, W.-B. J. Org. Chem. 1994, 59, 4364−4366.
(c) Babudri, F.; Fiandanese, V.; Marchese, G.; Punzi, A. Tetrahedron
1996, 52, 13513−13520. (d) Tatlock, J. H. J. Org. Chem. 1995, 60,
6221−6223. (e) Thasana, N.; Prachyawarakorn, V.; Tontoolarug, S.;
Ruchirawat, S. Tetrahedron Lett. 2003, 44, 1019−1021. (f) Ma, M.; Li,
C.; Peng, L.; Xie, F.; Zhang, X.; Wang, J. Tetrahedron Lett. 2005, 46,
3927−3929.
(8) The Claisen rearrangement offers opportunity for the
enantioselective synthesis of 7c−e by chirality transfer.
(9) (a) Handerson, S.; Schlaf, M. Org. Lett. 2002, 4, 407−409.
(b) Bosch, M.; Schlaf, M. J. Org. Chem. 2003, 68, 5225−5227.
(10) Wei, X.; Lorenz, J. C.; Kapadia, S.; Saha, A.; Haddad, N.; Busacca,
C. A.; Senanayake, C. H. J. Org. Chem. 2007, 72, 4250−4253.
(11) Hurd, C. D.; Pollack, M. A. J. Am. Chem. Soc. 1938, 60, 1905−
1911.
(27) Scaled to 1.3 g of 1c (86% yield).
(28) Scaled to 1.0 g of 1d (88% yield).
(29) Scaled to 0.8 g of 1e (90% yield).
(30) The homodimer was purified and characterized and its
configuration is deduced from NOE experiments; see the Supporting
Information.
(31) Acylation without protodesilylation, see: Procopiou, A. P.; Baugh,
S. P. D.; Flack, S. S.; Inglis, G. G. A. J. Org. Chem. 1998, 63, 2342−2347.
(32) Stamos, D. P.; Taylor, A. G.; Kishi, Y. Tetrahedron Lett. 1996, 37,
8647−8650.
(33) See the Supporting Information for a mechanistic proposal that
accounts for the stereochemical result of the stepwise bromodesilyla-
tion. For a two-step procedure for the conversion of α,β-epoxysilanes to
vinylbromides (HBr; F3B·OEt2), see: Hudrlik, P. F.; Hudrlik, A. M.;
Rona, R. J.; Misra, R. N.; Withers, G. P. J. Am. Chem. Soc. 1977, 99,
1993−1996.
4065
dx.doi.org/10.1021/ol501204m | Org. Lett. 2014, 16, 4062−4065