Synthesis of substituted 3-hydroxy-2-furanone derivatives via an unusual enolate wittig rearrangement/alkylative cyclization sequence

Article abstract of DOI:10.1021/ol4009188

Treatment of methyl O-(alkynylmethyl) glycolate derivatives with dialkylboron triflates and Huenig's base leads to the formation of highly substituted 3-hydroxy-2-furanone derivatives. The transformations appear to proceed via an unusual mechanism involving initial 2,3-Wittig rearrangement of a boron ester enolate followed by an alkylative cyclization reaction that leads to incorporation of an alkyl group from the boron reagent into the product.

Full text of DOI:10.1021/ol4009188

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Synthesis of Substituted 3‑Hydroxy-2-
Furanone Derivatives via an Unusual
Enolate Wittig Rearrangement/Alkylative
Cyclization Sequence
Renata K. Everett and John P. Wolfe*
Department of Chemistry, University of Michigan, 930 North University Avenue,
Ann Arbor, Michigan, 48109-1055, United States
jpwolfe@umich.edu
Received April 3, 2013
ABSTRACT
€
Treatment of methyl O-(alkynylmethyl) glycolate derivatives with dialkylboron triflates and Hunig’s base leads to the formation of highly
substituted 3-hydroxy-2-furanone derivatives. The transformations appear to proceed via an unusual mechanism involving initial 2,3-Wittig
rearrangement of a boron ester enolate followed by an alkylative cyclization reaction that leads to incorporation of an alkyl group from the boron
reagent into the product.
In recent years our group has explored the development
and applications of boron-mediated cascade Wittig
rearrangement/aldol reactions of glycolate ester deriva-
tives 1 for the stereoselective construction of R-alkyl-R,β-
dihydroxy esters 2 (eq 1).1,2 These reactions proceed with
excellent diastereoselectivity, and use of the readily available
chiral auxiliary 2-phenylcyclohexanol provides access to
enantiomerically enriched products with up to 94% ee after
auxiliary cleavage. Although these reactions have demon-
strated synthetic utility, the scope of this method is currently
limited to substrates bearing O-allyl or O-benzyl migrating
groups. Substrates that contain simple O-alkyl or O-aryl
groups fail to undergo the initial Wittig rearrangement.
In order to further expand the scope of this method, we
sought to employ substrates bearing O-propargyl groups.
Related substrates are known to undergo 2,3-Wittig re-
arrangement when treated with a strong base,3 and the
resulting products 2 where R1 is an allenyl group could be
valuable intermediates for the construction of substituted
heterocycles.4 To this end we prepared ester substrate 3a
and subjected this compound to our standard reaction
conditions whereby the ester was treated with Bu2BOTf
and iPr2NEt and stirred at rt for ca. 15 min, and then an
aldehyde was added to the resulting mixture. We were
quite surprised to discover that 3a failed to undergo the
sequential rearrangement/aldol reaction and instead was
transformed to the substituted 3-hydroxy-2-furanone 4a
(eq 2). Interestingly, although the aldehyde electrophile
apparently did not participate in the reaction, a butyl
group from the Bu2BOTf reagent was incorporated at
(3) (a) Marshall, J. A.; Robinson, E. D.; Zapata, A. J. Org. Chem.
1989, 54, 5854–5855. (b) Marshall, J. A.; Wang, X.-j. J. Org. Chem. 1991,
56, 4913–4918. (c) Marshall, J. A.; Wang, X.-j. J. Org. Chem. 1992, 57,
2747–2750.
(4) (a) Alcaide, B.; Almendros, P.; Carrascosa, R.; Martinez del
Campo, T. Chem.;Eur. J. 2010, 16, 13243–13252. (b) Alcaide, B.;
Almendros, P.; Carrascosa, R.; Martinez del Campo, T. Chem.;Eur.
J. 2009, 15, 2496–2499.
(1) (a) Bertrand, M. B.; Wolfe, J. P. Org. Lett. 2006, 8, 4661–4663. (b)
Giampietro, N. C.; Kampf, J. W.; Wolfe, J. P. J. Am. Chem. Soc. 2009,
131, 12556–12557.
(2) For related asymmetric Wittig rearrangement/Mannich reactions
that generate amino alcohol products, see: Giampietro, N. C.; Wolfe,
J. P. Angew. Chem., Int. Ed. 2010, 49, 2922–2924.
r
10.1021/ol4009188
XXXX American Chemical Society

the product C5 position. Omission of the aldehyde from
the reaction mixture led to the formation of 4a in 62%
yield.
limited to internal alkyne substrates, as subjection of
terminal alkyne substrate 3h to the standard reaction
conditions resulted in generation of substituted allylic
alcohol 7 in low yield rather than a substituted furanone
(eq 4).7
Given the potential synthetic utility and biological
relevance of substituted furanone derivatives,5 we sought
to further examine this unusual transformation. Efforts to
optimize reaction conditions by modifying temperature,
solvent, base, etc. did not lead to significant improvements
in yield. As such, we proceeded to explore the effect of
substrate structure on reactivity. As shown in Table 1, the
best results were obtained with substrates bearing aryl
groups on the alkyne moiety (entries 1À4 and 8). These
substrates were converted to substituted furanones in
We also briefly examined the influence of the dialkyl-
boron reagent structure on reactivity. As anticipated, use
of diethylboron triflate resulted in the formation of fura-
none product 4h in which an ethyl group had been
incorporated at C5 (Table 1, entry 8). However, use of
the hindered and lesselectrophilicCy2BClreagent led to no
observable reaction. Interestingly, treatment of 3a with
9-BBN-OTf/iPr2NEt led to clean formation of allene 8 in
53% yield (eq 5).
Table 1. Tandem Wittig Rearrangement/Alkylative Cyclization
Reactionsa
entry
R
R1
yieldb
1
2
3
4
5
6
7
8
Bu
Bu
Bu
Bu
Bu
Bu
Bu
Et
Ph
62
47
56
43
29
23
20
50
p-F-C6H4
2-naphthyl
p-Me-C6H4
Ph(CH2)2
iPr
During the course of our initial optimization studies, we
examined an alternative workup procedure in which the
reaction of substrate 3a with Bu2BOTf/iPr2NEt was
quenched with aqueous HCl rather than treated with
H2O2 and pH 7 buffer. Interestingly, use of this acidic
workup led to the formation of 9 (39% yield) (eq 6); the
furanone product was not observed by NMR analysis of
the crude reaction mixture.
Me
Ph
a Conditions: (i) 1.0 equiv of 3, 3.2 equiv of R2BOTf, 4.0 equiv of
iPr2NEt, CH2Cl2, 0.25 M, 0 °C f rt. (ii) H2O2, pH 7 buffer. b Isolated
yields (average of two experiments).
moderate yield. In contrast, substrates that contain alkyl-
substituted alkynes were transformed in low yields (entries
5À7).6 Attempts to employ substrate 5 bearing an ethyl
group at the propargylic position led to formation of allene
6 in 52% yield; no furanone product was observed (eq 3).
The rearrangement/cyclization reactions appear to be
Taken together, the results illustrated in eqs 3À6 provide
a considerable amount of information about the likely
mechanism of the rearrangement/cyclization reactions.
The conversions of 5 to 6, 3a to 8, and 3h to 7 suggest
the mechanism of this unusual rearrangement/alkylation
involves initial 2,3-Wittig rearrangement of the substrate
to generate an intermediate allene.8 A subsequent alkyl
(5) (a) Nicoll-Griffith, D. A.; Chauret, N.; Houle, R.; Day, S. H.;
D’Antoni, M.; Silva, J. M. Drug. Metab. Dispos. 2004, 32, 1509–1515. (b)
Leblanc, Y.; Roy, P.; Wang, Z.; Li, C. S.; Chauret, N.; Nicoll-Griffith,
D. A.; Silva, J. M.; Aubin, Y.; Yergey, J. A.; Chan, C. C.; Riendeau, D.;
Brideau, C.; Gordon, R.; Xu, L.; Webb, J.; Visco, D. M.; Prasit, P.
Bioorg. Med. Chem. Lett. 2002, 12, 3317–3320. (c) Leblanc, Y.; Roy, P.;
Boyce, S.; Brideau, C.; Chan, C. C.; Charleson, S.; Gordon, R.; Grimm,
E.; Guay, J.; Leger, S.; Li, C. S.; Riendeau, D.; Visco, D.; Wang, Z.;
Webb, J.; Xu, L. J.; Prasit, P. Bioorg. Med. Chem. Lett. 1999, 9, 2207–
2212.
(7) Compound 7 was the only isolable product obtained from this
reaction. The low yield may be due to the volatility of either the product
or of other low molecular weight side products.
(6) The modest yields appear to be due to difficulties with the
isolation of these products or the formation of volatile or water soluble
side products.
(8) For a recent synthesis of furan-2-ones via addition of Grignard
reagents to allenyl alcohols followed by trapping with CO2, see: Li, S.;
Miao, B.; Yuan, W.; Ma, S. Org. Lett. 2013, 15, 977–979.
B
Org. Lett., Vol. XX, No. XX, XXXX

transfer from the boron reagent to the allene leads to CÀC
bond formation and likely proceeds via a radical pathway
that is presumably initiated by small amounts of oxygen.9À12
Finally, the fact that the modified (acidic) workup gen-
erates 9 rather than 4 suggests that R-ketoester 9 may be an
intermediate in the alkylation/cyclization reactions and
that conversion of 9 to 4 may occur during the H2O2/ pH
7 buffer workup step.
provide alcohol 15.13 Intramolecular acylation of the
alcohol then yields the product 4.14
Alternatively, the radical reaction may also occur via a
chain mechanism (Scheme 2) whereby oxygen could lead
to generation of an alkyl radical from 11 in the initiation
step. Propagation would then involve addition of the alkyl
radical to a second molecule of 11 to yield 16, which could
be captured by oxygen to afford 18 along with another
alkyl radical. Base-mediated elimination of 18 (or the
analogous boronate ester derived from homolysis of the
OÀO bond and rearrangement) would provide 19, which
upon workup could be transformed to 14 and then 4.
Scheme 1. Radical Cage Mechanism for Conversion of 11 to 13
Scheme 2. Radical Chain Mechanism
On the basis of these results, two plausible mechanisms
for the conversion of 3 to 4 are illustrated in Schemes 1À2;
the former involves a radical cage process, whereas the
latter proceeds via a radical chain mechanism. In both
pathwaystreatment ofester 3 with the dialkylboron triflate
€
and Hunig’s base leads to formation of boron enolate 10,
Regardless of which pathway is operational, the forma-
tion of allene product 6 in the reaction of ethyl-substituted
substrate 5 may result from steric inhibition of alkyl radical
addition to the more highly substituted allene intermedi-
ate. The reason behind the failed cyclization of terminal
alkene substrate 3h is less clear. However, the radical
intermediate generated from rearrangement and alkyl
transfer of 3h (12, R1 = H) is considerably less stable than
analogous intermediates derived from internal alkynes.
In order to further probe the question of cage vs chain
mechanism, we carried out a crossover experiment in
which the rearrangement of 3a was carried out in the
which undergoes 2,3-Wittig rearrangement to afford
allene 11. In the cage mechanism (Scheme 1), the oxygen-
mediated transfer of an alkyl radical from the boron group
to the allene would occur within a solvent cage to generate
allyl radical 12.9 An intramolecular 1,5-hydrogen atom
transfer of 12 would then afford enolate 13. Upon workup,
protonation of the enolate would generate 14, which can
then undergo conjugate addition of water or hydroxide to
(9) For select recent examples of alkyl radical addition to allenes, see:
(a) Ma, Z.; Zeng, R.; Fu, C.; Ma, S. Tetrahedron 2011, 67, 8808–8818. (b)
Kippo, T.; Fukuyama, T.; Ryu, I. Org. Lett. 2011, 13, 3864–3867. (c)
Yang, D.; Cwynar, V.; Donahue, M. G.; Hart, D. J.; Mbogo, G. J. Org.
Chem. 2009, 74, 8726–8732. (d) Ma, Z.; Ma, S. Tetrahedron 2008, 64,
6500–6509. (e) Alcaide, B.; Almendros, P.; Aragoncillo, C.; Redondo,
M. C. J. Org. Chem. 2007, 72, 1604–1608. (f) Shen, L.; Hsung, R. P. Org.
Lett. 2005, 7, 775–778. For a review on free-radical addition to allenes,
see: (g) Hartung, J.; Kopf, T. in Modern Allene Chemistry; Krause, N.,
Hashimi, A. S. K., Eds.; Wiley-VCH: Weinheim, Germany, 2004; pp
701À726.
(13) β,γ-Unsaturated R-ketoesters that are structurally similar to 14
have been shown to undergo conjugate addition of water followed by
lactonization to generate 3-hydroxy-furan-2-ones analogous to 4. See:
(a) Mageswaran, S.; Ollis, W. D.; Southam, D. A.; Sutherland, I. O.;
Thebtaranonth, Y. J. Chem. Soc., Perkin Trans. 1 1981, 1969–1980. (b)
Babakhanyan, A. V.; Ovsepyan, V. S.; Kocharyan, S. T.; Panosyan,
G. A. Russ. J. Org. Chem. 2003, 39, 814–819. (c) Ovsepyan, V. S.;
Babakhanyan, A. V.; Manukyan, M. O.; Kocharyan, S. T. Russ. J. Gen.
Chem. 2004, 74, 1376–1382. For an asymmetric two-step conjugate
addition/ring-closing reaction for the conversion of β,γ-unsaturated
R-ketoesters to 3-hydroxy-furan-2-ones, see: (d) Xiong, X.; Ovens, C.;
Pilling, A. W.; Ward, J. W.; Dixon, D. J. Org. Lett. 2008, 10, 565–567.
(14) Attempts to subject 9 to the workup conditions failed to generate
significant amounts of product 4. However it is possible that cleavage of
the BÀO bond is relatively slow and that either 13 or a boron-complexed
ketone analog of 14 is the actual electrophile that participates in the
conjugate addition of water or hydroxide.
(10) For an example of furan-2-one synthesis that is initiated by radical
initiation to an allene, see: Wu, Z.; Huang, X. Synthesis 2007, 45–50.
(11) For a review on the generation and reaction of alkyl radicals
from organoboron reagents, see: Ollivier, C.; Renaud, P. Chem. Rev.
2001, 101, 3415–3434.
(12) Reactions of 3a conducted with measured amounts of added air
failed to provide improved results over the standard conditions. How-
ever, a rearrangement of 3a conducted in a nitrogen-filled glovebox
afforded a ca. 2:1 mixture of allene 8 and lactone 4a, which suggests a
stoichiometric amount of oxygen is needed to facilitate the transforma-
tion of 3 to 4.
Org. Lett., Vol. XX, No. XX, XXXX
C

presence of 3.2 equiv of Et3B (eq 7). This reaction pro-
vided a ca. 1:1 mixture of 4a and 4h, which indicates an
intermolecular (i.e., noncage) alkyl radical transfer to the
allene is a viable mechanistic pathway and suggests thatthe
mechanism illustrated in Scheme 2 may be operational
(although this does not unambiguously rule out the radical
cage mechanism shown in Scheme 1).15
O-propargyl glycolate derivatives. The reactions produce
potentially useful 3-hydroxy-furan-2-one products in
moderate yield and appear to proceed via radical alkyla-
tion of an intermediate allene. Future studies will be
directed toward improving and expanding the scope of
these transformations.
Acknowledgment. The authors thank the NSF (CHE
1111218) for financial support of this work.
Supporting Information Available. Experimental pro-
cedures, characterization data for all new compounds,
descriptions of stereochemical assignments with support-
1
ing structural data, and copies of H and 13C NMR
In conclusion, we have discovered an unusual Wittig
rearrangement/alkylative cyclization reaction of methyl
spectra for all new compounds reported in the text. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(15) Treatment of 3a with Bu2BOTf/iPr2NEt in the presence of excess
(5 equiv) TEMPO as a radical trap resulted in no reaction; substrate 3a
was unchanged. This may be due to inhibition of enolate generation due
to an undesired reaction of the Lewis acidic Bu2BOTf with TEMPO.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX