104
Chemistry Letters 2000
Radical Cyclization Reaction Using a Combination of Phosphinic Acid
and a Base in Aqueous Ethanol
Hideki Yorimitsu, Hiroshi Shinokubo, and Koichiro Oshima*
Department of Material Chemistry, Graduate School of Engineering, Kyoto University,
Sakyo-ku, Yoshida, Kyoto 606-8501
(Received October 29, 1999; CL-990924)
Treatment of allylic ether of 2-iodophenol or 2-haloethanal
allylic acetal with phosphinic acid, a base and a radical initiator
(AIBN or triethylborane) in aqueous ethanol provided the cor-
responding radical cyclization product in excellent yield. An
addition of a base is critical to employ phosphinic acid as a
radical mediator.
During the recent dramatic development of radical
methodology in organic synthesis, tributyltin hydride has
played a leading role1 despite its toxicity and the difficulty
incurred in residue removal. For this reason, new radical medi-
ators to replace tin compounds have been actively
investigated.2,3 Among them, Barton's4 and other5 groups
reported phosphinic acid (H3PO2) as a cheap, much less toxic
and easily removable chain carrier in various reduction reac-
tions although toxic solvents such as dioxane and benzene were
employed.6 To reduce total toxicity in radical reaction, it is
important to pay attention to the solvent employed where the
radical reaction is carried out. Here, we wish to report the radi-
cal cyclization with phosphinic acid in aqueous ethanol. The
indispensable addition of a base is also described.
These facts clearly show that a base is essential to carry out the
phosphinic-acid-mediated radical reaction smoothly and sug-
gest the actual chain carrier would be a phosphinate anion.7
The results of radical cyclization using an aq H3PO2 /
NaHCO3 / EtOH system are summarized in Scheme 2. Allyl
ether 1b afforded dihydrobenzofuran derivative 2b in moder-
ate yield due to its volatility as well as the formation of by-
product 38 (15%). Triethylborane, instead of AIBN, could act
as a radical initiator at room temperature.9 Although aryl bro-
mide 1d resisted reduction, selective cyclization of 1e could
be achieved in excellent yield.10 When allyl ether of iodon-
aphthol 1f was employed as a substrate, 6-endo cyclization
was observed11 in addition to 5-exo cyclization.
Next, we focused on radical cyclization of haloacetals
(Table 1). For example, to iodoacetal 4a was added a solution
of phosphinic acid, sodium hydrogencarbonate (1.5 eq to
H3PO2), and AIBN in ethanol12 and the resulting mixture was
heated at reflux. The reaction was completed within 30 min.
Extraction (ethyl acetate/brine) and concentration followed by
purification through short silica gel column cleanly gave
bicyclic acetal 5a in 98% yield. Some comments are worth
noting. (1) tert-Butyldimethylsilyl ether 4e and benzoate ester
4f were tolerant under this basic condition. (2) Decreasing the
amount of H3PO2 and NaHCO3 led to slightly lower yet good
yields (4c and 4f). (3) Slow addition of a solution of H3PO2,
NaHCO3 and 4,4'-azobis(4-cyanopentanoic acid) in water to
4g afforded six-membered 5g in 79% yield. (4) Bromo analog
First, we chose 1a as a substrate to examine radical
cyclization by means of phosphinic acid in ethanol. A solution
of 1a (1 mmol), phosphinic acid (50% aqueous solution, 1.1
mL, 10 mmol), and AIBN (0.3 mmol) in ethanol (5 mL) was
heated at reflux for 5 h. Contrary to our expectations, the
desired product 2a was obtained in only 3% yield and 1a
(67%) was recovered. Then, this reaction was performed in the
presence of sodium hydrogencarbonate (10 mmol).
Surprisingly, the yield improved sharply up to 87%. Potassium
hydroxide or triethylamine was also highly effective to give 2a
in 91% or 89% yield, respectively. On the other hand, addition
of hydrochloric acid suppressed the reaction completely.
Copyright © 2000 The Chemical Society of Japan
Yorimitsu, Hideki
