Use of lithium hexafluoroisopropoxide as a mild base for Horner-Wadsworth-Emmons olefination of epimerizable aldehydes

Article abstract of DOI:10.1021/ol051785m

(Chemical Equation Presented) The weak base lithium 1,1,1,3,3,3- hexafluoroisopropoxide (LiHFI) is shown to be highly effective as a reagent for intermolecular Horner-Wadsworth-Emmons (HWE) olefination of epimerizable aldehydes with trimethyl phosphonoace

Full text of DOI:10.1021/ol051785m

ORGANIC
LETTERS
2005
Vol. 7, No. 19
4281-4283
Use of Lithium Hexafluoroisopropoxide
as a Mild Base for
Horner−Wadsworth−Emmons Olefination
of Epimerizable Aldehydes
Landy K. Blasdel and Andrew G. Myers*
Department of Chemistry and Chemical Biology, HarVard UniVersity,
Cambridge, Massachusetts 02138
myers@chemistry.harVard.edu
Received July 27, 2005
ABSTRACT
The weak base lithium 1,1,1,3,3,3-hexafluoroisopropoxide (LiHFI) is shown to be highly effective as a reagent for intermolecular Horner
−
Wadsworth Emmons (HWE) olefination of epimerizable aldehydes with trimethyl phosphonoacetate, affording products with little or no
−
epimerization and notably high E-selectivity.
The Horner-Wadsworth-Emmons (HWE) olefination of
aldehydes is one of the more reliable carbon-carbon (double)
bond-forming reactions in synthesis.1 Its utility has been
extended for applications involving base-sensitive substrates
by the introduction of modified protocols.2 Recently, during
the development of a synthetic route to natural products of
the cytochalasin family, the intramolecular HWE reaction
depicted in Scheme 1 provided a challenging problem whose
solution required the development of novel conditions to
minimize epimerization of the stereocenter adjacent to the
aldehyde group of substrate 1.3 As shown in Scheme 1, use
of the unusual base sodium trifluoroethoxide (slightly less
than 1 equiv) in the presence of trifluoroethanol (TFE, 100
equiv) in 1,2-dimethoxyethane (DME) at 80 °C (16 h) proved
Scheme 1. Optimal Conditions for Intramolecular HWE
Cyclization of Cytochalasin Precursor 1.3
(1) (a) Horner, L.; Hoffmann, H.; Wippel, H. G.; Klahre, G. Chem. Ber.
1959, 92, 2499. (b) Wadsworth, W. S., Jr.; Emmons, W. D. J. Am. Chem.
Soc. 1961, 83, 1733. (c) Wadsworth, W. S., Jr. Org. React. 1977, 25, 73.
(d) Maryanoff, B. E.; Reitz, A. B. Chem. ReV. 1989, 89, 863.
(2) (a) Blanchette, M. A.; Choy, W.; Davis, J. T.; Essenfeld, A. P.;
Masamune, S.; Roush, W. R.; Sakai, T. Tetrahedron Lett. 1984, 25, 2183.
(b) Rathke, M. W.; Nowak, M. J. Org. Chem. 1985, 50, 2624. (c) Paterson,
I.; Yeung, K.-S.; Smaill, J. B. Synlett 1993, 774.
optimal in that case to maximize the yield of the cyclized
product while minimizing epimerization.3 This finding
prompted us to explore the potential broader utility of
fluorinated alkoxides as bases for intermolecular HWE
(3) Haidle, A. M.; Myers, A. G. Proc. Natl. Acad. Sci. U.S.A. 2004,
101, 12048.
10.1021/ol051785m CCC: $30.25
© 2005 American Chemical Society
Published on Web 08/23/2005

reactions of base-sensitive substrates, leading to the discovery
that lithium 1,1,1,3,3,3-hexafluoroisopropoxide (LiHFI) is
a nearly ideal reagent in such contexts,4 as detailed herein.
Table 1. Comparison of Results of HWE Olefination of
Epimerizable Aldehydes
We first conducted a screen to determine the most effective
fluorinated alkoxide for optimal efficiency and E-selectivity
in the intermolecular HWE reaction of cyclohexanecarbox-
aldehyde (1 equiv) with trimethyl phosphonoacetate (1.2
equiv). Lithium, sodium, potassium, and cesium salts of the
alcohols TFE (pKa 12.4 in water4) and HFI (pKa 9.3 in water,4
1.1 equiv in each case) were investigated as bases, initially
using acetonitrile as solvent. LiHFI emerged as the superior
reagent, leading to complete conversion of the aldehyde
substrate and a high yield of product, with nearly perfect
E-selectivity. Counterions other than lithium gave reduced
E-selectivity, and reactions employing 3,3,3-trifluoroethoxide
led to partial transesterification of the product with the base.
Although LiHFI could be prepared and used as a discrete
reagent (a hygroscopic, white solid), invariably we opted for
a more convenient procedure involving generation of the base
in situ using solutions of n-butyllithium and HFI.5 In addition
to acetonitrile, tetrahydrofuran (THF) and DME were also
found to be effective solvents. Because DME proved to be
slightly advantageous in minimizing epimerization among
the most base-sensitive substrates and because it afforded
highly reproducible results, we adopted this solvent for
routine use. Compiled in Table 1 are our optimized results
from a series of experiments using LiHFI as a base for the
HWE reaction of trimethyl phosphonoacetate with aldehydes
containing an epimerizable R-stereocenter. Also provided for
comparison are results with the same substrates using other
methods developed for HWE olefination of base-sensitive
aldehydes: the Masamune-Roush protocol (LiCl, DBU)2a
and, in the case of the substrates of the first and final entries,
the method of Paterson and co-workers (employing barium
hydroxide as base).2c
The results of Table 1 show that HWE olefination of a
number of epimerizable aldehydes with trimethyl phospho-
noacetate (1.2 equiv) using LiHFI (1.18 equiv) as base and
DME as solvent is highly efficient and provides products
with high E-selectivity and little or no epimerization. Of the
substrates we investigated, the peptidyl aldehyde of the final
entry proved to be the most susceptible to epimerization and
provided the greatest distinction among the three protocols
examined. With other substrates, existing protocols were
quite effective,2,6 but generally afforded products with
a LiOCH(CF3)2/HOCH(CF3)2 (1.24 equiv), n-BuLi (1.18 equiv), (CH3O)2-
POCH2CO2CH3 (1.20 equiv), DME, 0.24 M in substrate, -14 °C; LiCl,
DBU: see ref 2a; Ba(OH)2: see ref 2c. b E/Z ratios reported that exceed
20:1 were determined by HPLC analysis; all others were determined by 1H
NMR analysis (maximal value: >19:1). c Isolated yield of the E-olefin,
unless otherwise noted. d HPLC yield, determined using biphenyl as an
internal standard. e Isolated yield of E and Z isomers combined.
(4) For a discussion of the properties of fluorinated alcohols and their
use in synthesis, see: Be´gue´, J.-P.; Bonnet-Delpon, D.; Crousse, B. Synlett
2004, 18.
(5) Deprotonation of HFI with n-butyllithium produces a modest exo-
therm that is readily controlled under typical laboratory conditions. For
example, addition of a solution of n-butyllithium in hexanes (2.48 M, 2.28
mL, 5.65 mmol, 1 equiv) via syringe over 6 min to a solution of HFI (0.627
mL, 5.95 mmol, 1.05 equiv) in DME (20 mL) at -14 °C produced a
maximum internal temperature of -6.5 °C.
somewhat diminished E-selectivities. The superior E-selec-
tivity of the new protocol for HWE olefination, though not
anticipated, proved to be one of its primary advantages. This,
plus the extraordinary mildness of the conditions may
recommend the new method for use even in cases where
epimerization is not an issue.
(6) We did encounter some difficulty in reproducing our results from
run to run in small-scale experiments using the Masamune-Roush
protocol.2a This we traced to variations in the concentration of lithium
chloride, introduced by errors in weighing small amounts of this hygroscopic
reagent. In this regard, the in situ protocol for the generation of LiHFI was
found to be both convenient and highly reproducible in small-scale
experiments. A similar protocol for the generation of lithium chloride in
situ might also be effective, but was not explored.
4282
Org. Lett., Vol. 7, No. 19, 2005

We investigated the degree to which (adventitious) mois-
ture might influence HWE olefination by the new procedure
and found that as much as 5 equiv of water only slightly
increased the amount of epimerized product (examining the
transformation of entry 1, Table 1) and had little effect on
the extent of conversion and E-selectivity of the reaction.
By contrast, the presence of excess HFI proved to be
deleterious by all measures (conversion, degree of epimer-
ization, and E-selectivity). This is a distinction from the
intramolecular HWE reaction of Scheme 1, where a large
excess of TFE (100 equiv) was required for optimum results.
It is evident that the intramolecular HWE olefination of
Scheme 1 differs considerably from the intermolecular HWE
reactions reported here, not only from the varying influences
of the respective fluorinated alcohols in the two reactions
but also from the much higher reaction temperature required
(+80 °C vs -14 °C). Further evidence in this regard is the
fact that when substrate 1 was subjected to LiHFI-mediated
olefination conditions (CH3CN, 92 °C), the macrocycles 2
and 18-epi-2 were formed in a ratio of 2.6:1, with poor
conversion (16%).7
To determine the degree to which trimethyl phosphono-
acetate (pKa ∼18-19 in DMSO,8 ∼12 in H2O9) is depro-
tonated by LiHFI (pKa HFI ) 18.2 in DMSO,10 9.3 in H2O4)
in DME we conducted a 31P NMR study (Figure 1), much
like that conducted by Masamune, Roush, and co-workers
in their original report.2a The data suggest that the equilibrium
between trimethyl phosphonoacetate and its lithium salt using
0.98 equiv of LiHFI as base greatly favors the former species
(cf. Figure 1b,c), but the breadth of the 31P resonance in
Figure 1c is indicative of rapid exchange under the conditions
of the NMR experiment. Thus, we speculate that the primary
anionic species in the reaction is LiHFI, and further, that
LiHFI does not induce epimerization of the aldehyde
substrates we have examined at a rate that is competitive
with HWE olefination. It must be the case that lithium
trimethyl phosphonoacetate, as formed under the conditions
described, also does not induce any appreciable epimerization
of the aldehyde substrates examined.
Figure 1. 31P NMR spectra of (a) (CH3O)2POCH2CO2CH3 in DME
at 23 °C, (b) as in (a) plus 1.00 equiv t-BuOLi, (c) as in (a) plus
0.98 equiv of LiOCH(CF3)2. Each spectrum is referenced to an
external standard of 85% phosphoric acid in water.
In conclusion, we have found that the exceedingly weak
base LiHFI is notably effective in intermolecular HWE
olefinations of epimerizable aldehydes. We anticipate that
this reagent may be of utility in other base-mediated
transformations as well.
Acknowledgment. We are grateful to the National
Science Foundation and the Martin and Mary Kilpatrick
Fellowship for funding.
(7) Haidle, A. M. Harvard University, Cambridge, MA. Unpublished
results.
(8) The reported pKa of triethyl phosphonoacetate in DMSO is 18.6:
Evans, D. A. Evans pKa Table, http://daecr1.harvard.edu/ (accessed Jul
2005).
(9) The reported pKa of triethyl phosphonoacetate in water is 11.9:
Sokolov, M. P.; Gazizov, I. G.; Mavrin, G. V. Zh. Obshch. Khim. 1989,
59, 1043.
Supporting Information Available: Experimental pro-
cedures and characterization data for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(10) Arnett, E. M.; Small, L. E. J. Am. Chem. Soc. 1977, 99, 808.
OL051785M
Org. Lett., Vol. 7, No. 19, 2005
4283