Carbanion vs. carbon radical in tandem 1,4-addition to two Connected units of acrylate or methacrylate

Article abstract of DOI:10.1246/bcsj.80.2011

Two units of acrylate or methacrylate were connected with 2,4-pentanediol. Michael addition of organometallic compounds to this substrate failed to cause the intended in tramolecular Michael addition of the generated enolate, but the addition of carbon ra

Full text of DOI:10.1246/bcsj.80.2011

Short Articles
Bull. Chem. Soc. Jpn. Vol. 80, No. 10, 2011–2013 (2007) 2011
R2
COOR3
R3OOC
R1
COOR3
M
R1
M
+
O
Carbanion vs. Carbon Radical in
Tandem 1,4-Addition to Two
Connected Units of Acrylate
or Methacrylate
R1
M
R2
R2 R2
OR3
first adduct
second adduct
Scheme 1. Tandem Michael addition with alkylmetal.
R
R
O
O
O
O
+
Bu
M
Shinya Nagano, Siddiki S. M. A. Hakim,
Ã
O
Bu
O
then H3O+
and Takashi Sugimura
O
O
R
R
1a. R = Me
1b. R = H
2a. R = Me
2b. R = H
Graduate School of Material Science, Himeji Institute of
Technology, University of Hyogo, 3-2-1 Kohto, Kamigori,
Ako-gun, Hyogo 678-1297
Scheme 2. Intramolecular tandem Michael addition.
Received April 25, 2007; E-mail: sugimura@sci.u-hyogo.ac.jp
Oct
O
O
O
O
Base
O
O
O
O
Two units of acrylate or methacrylate were connected
with 2,4-pentanediol. Michael addition of organometallic
compounds to this substrate failed to cause the intended in-
tramolecular Michael addition of the generated enolate, but
the addition of carbon radical induced the internal addition
to give cyclic compounds in yields up to 79%.
4
5b
Oct
OEt
M
M
O
O
O
O
O
O
OEt
X
Oct
I
II
Addition of an olefinic molecule is an often employed reac-
tion for C–C bond formation, but it becomes difficult to per-
form selectively if one wishes to add two, but not three or
more, olefin-units sequentially, because the first adduct and
the second adduct could have similar reactivity to the olefin.
In fact, addition of a Grignard reagent to acrylate analogues
gives a tandem adduct with two units in yields up to 97%
after quenching with water (Scheme 1).1 The yields are high
when the acrylate has a bulky ꢀ-substituent, but are low as
21–58% with methacrylate (R2 = Me), and no such example
has been reported for unsubstituted acrylate (R2 = H) to our
knowledge.
Scheme 3. Simple intramolecular Michael addition.
component except unreacted 1a was the double intermolecular
Michael adduct 3 with two butyl groups. Similar approaches
also failed to obtain 2b with acrylate 1b.
Inefficient intramolecular addition was also found in a sim-
pler reaction system with 4. When 4 was treated with LDA in
THF or THF/TMEDA, the intramolecular adduct 5b was not
obtained, but polymeric products were obtained after quench-
ing with water (Scheme 3). Treatment of 4 with t-BuOK in t-
BuOH or THF only resulted in partial hydrolysis after aqueous
workup. The poor reactivity is not readily understood, because
ten-membered ring formation in 1a or 1b should not generate
notable strain. However, if coordination of a counter cation of
the enolate to the intramolecular carbonyl of the unsaturated
ester is considered, the reaction sites for the Michael addition
are forced to be far away from each other, and do not seem to
be accessible, as shown I. This situation is in contrast to the
intermediate when the two unsaturated esters are connected
at the olefinic ends, e.g., intermediate II is fixed to a proper
conformation to perform intramolecular addition by coordina-
tion to the metal.4
The unwanted arrangement, shown as I, does not exist in
the corresponding radical version, and actually, the radical
reaction involving 1a has previously been found to yield effec-
tively the cyclized product.2 That is, photolysis of thiophenol
with 1a (3 molar amounts) gave ten-membered ring products
(PhS or H adduct) in 94% of the total yield based on the
amount of thiophenol. Hence, the alkyl radical addition to 1a
or 1b was expected to be followed by the cyclization.5
A mixture of octyl bromide (25 mmol dmÀ3), Bu3SnH
A simple idea to realize the tandem addition irrespective of
the olefin structure is connection of the two units of the olefin.
The connector must be flexible enough to allow intramolecular
addition, but, at the same time, should keep the reaction sites
in some restricted space to perform effective intramolecular
addition. Here, an acrylate or methacrylate that is connected
with 2,4-pentanediol could not undergoes tandem Michael ad-
dition with alkylmetals, but the radical version of substrates
gave 69 and 79%, respectively, of the tandem adduct.
Two substrates, methacrylate ester 1a2 and acrylate ester 1b
of (2R,4R)-2,4-pentanediol were prepared by reacting it with
the corresponding acid chloride (82 and 63%, respectively,
after sufficient purification). The reaction of butyl anion with
1a was studied under varied conditions: butyllithium or butyl-
magnesium bromide as a source of carbanion, in the presence
of a copper(I) reagent, such as CuBr SMe2, CuCN, CuI PBu3,
Á
Á
or CuI/BF3, in THF or ether, at À78 ꢀC, or higher temperature,
À50 ꢀC or rt, prior to the aqueous quenching (Scheme 2). In
spite of the expected effectiveness of the intramolecular addi-
tion,3 2a was not detected in the reaction mixtures. The main
Published on the web October 12, 2007; doi:10.1246/bcsj.80.2011

2012 Bull. Chem. Soc. Jpn. Vol. 80, No. 10 (2007)
ꢀ 2007 The Chemical Society of Japan
R'
R'
LDA/–78 °C
then MeI/HMPA/0°C
O
O
Oct
O
O
Oct + 1a or 1b
5a
5a
7
9
O
Oct
O
O
O
III
6
MeOOC
Oct
COOMe
1) KOH/95% EtOH
2) CH2N2/ether
7
+Bu3SnH
–Bu3Sn•
O
O
O
O
Scheme 5. Conversions of 5a.
O
Oct
O
O
Oct
O
R'
R'
R'
R'
chirality at C7 generated in the cyclization step is uncontrol-
led, but the termination step to generate the chirality at C9
of 5a is fully controlled. The stereochemistry of C9 was not
determined, but it is thought to be S by the previously reported
results with thiyl radical.2 When the reaction of 1b to give 5b
was performed at À50 ꢀC, stereo-uncontrolled cyclization was
slightly suppressed, resulting in a ratio of 54:46.
In the present paper, tandem 1,4-addition with two units of
acrylate or methacrylate was demonstrated. The results show
once again that the radical reaction is superior to the anionic
version.9 The reaction could be a carbon-chain extension
method for simple iodo compounds.
5a. R' = Me
5b. R' = H
IV
Scheme 4. Radical tandem 1,4-addition with 1a and 1b.
(equimolar amount) and 1a (equimolar amount) in benzene
was heated to reflux in the presence of AIBN (0.1 molar
amount). After consumption of 1a, cyclic product 5a was iso-
lated in 34% yield (Scheme 4). The remainder was insoluble in
ether and was a mixture of polymeric compounds of 1a, con-
taining a tributyltin group. The reaction must start with gener-
ation of octyl radical by the abstraction of bromine atom with
tributyltin radical. The octyl radical adds to one of the olefins
of 1a to generate radical III, and intramolecular addition gives
the second radical IV, which is converted to 5a, when the radi-
cal reaction is terminated by trapping with the hydrogen ab-
stracted from tributyltin hydride. Mechanistic difference with
the reaction with thiophenol is that the generation process of
III is not reversible and hydride abstraction with the carbon
radicals from Bu3SnH is much slower than that from PhSH.6
In addition, the reaction temperature and molar amounts of
1a are also different.
The yield of 5a was increased to 39% when octyl iodide
was reacted at room temperature with a catalytic amount of
Bu3SnCl (0.1 molar amount) in the presence of NaBH4 (1.5
molar amounts) and 1a (equimolar amount) initiated by irradi-
ation with a high-pressure mercury lamp through a Pyrex
filter.7 The yield was further improved to 69% by the use of
two molar amounts of 1a. Under these mild conditions, poly-
merization of 1a was minimized, and 41% (0.8 molar amount)
of 1a was recovered. Similarly, by using of 1b (5 molar
amounts), 5b was obtained in 79% yield after column chroma-
tography. The high yields indicate that the reaction of Bu3SnH
with the octyl radical and radical III are sufficiently sup-
pressed, whereas that to IV still occurs. Thus, balance of the
reactivities of the radical intermediates predominates the prod-
uct yield. As a matter of fact, the yield reduced to 0–31% for
the addition of 1a to a benzylic or arene carbon.8
Experimental
All reactions were carried out under a dry nitrogen atmosphere.
Photochemical reactions were performed with a high-pressure
mercury lamp (350 W) made by Eiko-sha, Japan. MPLC was car-
ried out using an FMI pump (10 mL minÀ1) and a Lobar column
(MERCK Si-60 type B).
Preparation of 1a. To a solution of (2R,4R)-2,4-pentanediol
(10.0 g) and triethylamine (66.1 mL, 5.0 molar amounts) in di-
chloromethane (100 mL) was added methacryloyl chloride (23.4
mL, 2.5 mol) in 10 min at 0 ꢀC, and after stirring for 2 h at rt, water
was added. Extraction and purification by using a silica-gel col-
umn (elution with 10% ethyl acetate in hexane) gave 1a as a
colorless oil (18.9 g, 82% yield). Alternatively, 1a was obtained
from methacrylic acid and DCC in the presence of 4-dimethyl-
aminopyridine (82% yield).
Preparation of 1b. To a solution of (2R,4R)-2,4-pentanediol
(1.95 g) and triethylamine (7.8 mL, 3.0 molar amounts) in di-
chloromethane (20 mL) was added acryloyl chloride (4.6 mL,
3.0 molar amounts) at 0 ꢀC in 10 min. After 7 h, the mixture was
poured into water, extracted, and separated by column chromatog-
raphy on silica gel (elution with 10% ethyl acetate in hexane) to
give 1b as a colorless oil (2.51 g, 63% yield).
Reaction of 1a with a Carbanion, BuLi/CuCN. A butyllithi-
um solution (1.6 mol dmÀ3 in hexane, 1.0 mL, 4 molar amounts)
was added to a suspension of copper(I) cyanide (75.2 mg, 2 molar
amounts) in ether (1 mL) in 10 min at À78 ꢀC. The mixture was
warmed up to 0 ꢀC for 1 h, and then cooled again to À78 ꢀC. A
solution of 1a (100 mg) in ether (0.5 mL) was added in 5 min,
and the mixture was warmed up to À50 ꢀC for 30 min and then
allowed to stand for 3 h at the same temperature. The mixture
was extracted, dried, and separated on a silica-gel column (elution
with 50% ethyl acetate in hexane) and then MPLC (elution with
6% ethyl acetate in hexane) to give 3 as a colorless oil (41.6
mg, 31%).
Two stereoisomers were detected for both 5a and 5b in
equal amounts (by 1H NMR), even though 5a has four possible
stereoisomers. After ꢀ-methylation of the isomeric mixture of
5a at C9, obtained 6 was still a mixture of the isomers in the
same ratio (59% yield), and thus, C7 of 5a was determined
to be racemic (Scheme 5). Solvolysis to remove the chiral aux-
iliary resulted in 7 (74%) also in a 1:1 isomeric mixture, indi-
cating that relative stereochemistry of C9 against C7 is differ-
ent between the two isomers of 5a, and C9 must be stereo-
chemically uniform. Thus, in the reactions of 1a and 1b, the
Preparation of 4. A mixture of (2R,4R)-2,4-pentanediol (2.00
g), undecanoic acid (3.90 g, 1.1 molar amounts) 4-dimethylamino-

S. Nagano et al.
Bull. Chem. Soc. Jpn. Vol. 80, No. 10 (2007) 2013
pyridine (235 mg) in dichloromethane (200 mL) was cooled with
an ice-water bath, and a solution of DCC (4.35 g, 1.1 molar
amounts) in dichloromethane (50 mL) was added in 1 h. After
11 h, the acid (258 mg, 0.1 molar amount) and DCC (396 mg,
0.1 molar amount) were added. The mixture was stirred for an ad-
ditional 4 h, filtered through a Celite pad, concentrated, and sepa-
rated on a silica-gel column (elution with 20% ethyl acetate in
hexane) to give a mono-ester as a colorless oil (4.83 g, 92% yield).
To a solution of the mono-ester (2.0 g), triethylamine (1.2 mL, 1.2
molar amounts) in dichloromethane (100 mL), acryloyl chloride
(0.7 mL, 1.2 molar amounts) was added at 0 ꢀC in 10 min, and
after 4 h, triethylamine (0.5 mL, 0.5 molar amount and acryloyl
chloride (0.3 mL, 0.5 molar amount) were further added. After
1.5 h the mixture was extracted, dried over sodium sulfate, and
separated on a silica-gel column (elution with 10% ethyl acetate
in hexane) to give 1.3 g of 4 as a colorless oil (54% yield).
Enolate Formation from 4. Ester 4 (100 mg) was treated with
LDA (1.2 molar amounts) in ether (5 mL) at À78 ꢀC. After 2 h, the
mixture was extracted with ether, dried over sodium sulfate, and
concentrated. The obtained mixture contained mostly polymeric
compounds in addition to unreacted 4. Use of lithium hexamethyl-
disilazane in THF at 0 ꢀC, or potassium t-butoxide in t-butyl
alcohol or in THF at rt did not afford 5b.
temperature, followed by dropwise addition of a solution of 5a
(30.0 mg) in THF (1 mL). After 1.5 h, the mixture was warmed
to À15 ꢀC, and methyl iodide (52 mL) was added to this mixture.
The mixture was allowed to stand for 21 h at 0 ꢀC. The mixture
was extracted, dried over magnesium sulfate, and purified by
MPLC (elution with 6% ethyl acetate in hexane) to give 18.5
mg of 6 as a colorless oil (59% yield as a diastereomer mixture).
Conversion of 5a to 7. A solution of 5a (27 mg) in 95% etha-
nol (3 mL) containing 3% potassium hydroxide was heated at
reflux for 10 h. After cooling, the mixture was concentrated to
one half, neutralized with hydrochloric acid (2 mol dmÀ3), and
extracted with ether. The extract was dried and concentrated to
one tenth. To this solution, a solution of diazomethane was added.
After concentration, the mixture was purified by MPLC (elution
with 10–20% ethyl acetate in hexane) to give 17.8 mg of 7 as a
colorless oil (74% yield).
The authors thank Emeritus Professor Tadashi Okuyama for
his helpful suggestions.
Supporting Information
Reaction conditions for Scheme 2, and spectral data for all new
compounds. This material is available free of charge on the web at
http://www.csj.jp/journals/bcsj/.
Reaction of 1a with OctBr/Bu3SnH/AIBN. A solution of 1a
(150 mg), octyl bromide (132 mg, 1.1 molar amounts), tributyltin
hydride (198 mg, 1.1 molar amounts), and AIBN (9.8 mg) in ben-
zene (30 mL) was heated at reflux for 25 h. After concentration,
the mixture was purified by MPLC (elution with 6% ethyl acetate
in hexane) to give 74.3 mg of a diastereomeric mixture of 5a as a
colorless oil (34% yield).
References
1
J. Tanaka, S. Kanemasa, H. Kobayashi, O. Tsuge, Chem.
Lett. 1989, 1453; K. Yamaguchi, Y. Kurasawa, K. Yokota, Chem.
Lett. 1990, 719.
Reaction of 1a with OctI/Bu3SnCl/NaBH4 under Photoly-
sis. A solution of octyl iodide (240 mg), 1a (480 mg, 2 molar
amounts), sodium borohydride (55.6 mg, 1.5 molar amounts),
and tributyltin chloride (27 mL, 0.1 molar amount) in ethanol
(300 mL) was placed in a Pyrex photo-reactor. Nitrogen gas was
passed through the solution and stirring during photolysis with a
high-pressure mercury lamp at rt. Photolysis was maintained
for 32 h, and then the mixture was purified by MPLC (6% ethyl
acetate in hexane) to give 5a (195 mg) and 1a (ca. 200 mg). The
yield of 5a based on octyl iodide was 69%.
2
38, 3547.
T. Sugimura, S. Nagano, A. Tai, Tetrahedron Lett. 1997,
3
For reactions with 2,4-pentanediol as a chiral tether of two
reaction sites, see: T. Sugimura, in Recent Research Developments
in Organic Chemistry, ed. by S. G. Pandalai, Transworld Research
Network, Trivandrum, 1998, Vol. 2, pp. 47–53; T. Sugimura, Eur.
J. Org. Chem. 2004, 1185.
4 T. Uyehara, N. Shida, Y. Yamamoto, J. Chem. Soc., Chem.
Commun. 1989, 113.
5
For related study, see: N. A. Porter, G. S. Miracle, S. M.
Reaction of 1b with OctI/Bu3SnCl/NaBH4 under Photoly-
sis. A mixture of octyl iodide (40.0 mg), 1b (170 mg, 5 molar
amounts), and sodium borohydride (8.9 mg, 1.5 molar amounts) in
ethanol (50 mL) was placed in a Pyrex tube (100 mL), and de-
aerated with nitrogen gas. Tributyltin chloride (4.3 mL, 0.1 molar
amount) was added to the tube, and the tube was irradiated with a
high-pressure mercury lamp for 4 h while bubbling with nitrogen.
After removal of the solvent under vacuum, the mixture was
purified on a short column (silica gel, elution with 30% ethyl
acetate in hexane) and then MPLC (elution with 4.5% ethyl ace-
tate in hexane) to give 41.5 mg of 5b as a colorless oil (79% based
on octyl iodide). The isomers of 5b were partly separated.
Methylation of 5a with LDA/MeI to Give 6. A solution of
LDA was prepared by mixing diisopropylamine (0.12 mL) and
butyllithium (1.56 mol dmÀ3, 0.54 mL) in THF (5 mL) at À78
ꢀC. To this solution, HMPA (0.12 mL) was added at the same
Cannizzaro, R. L. Carter, A. T. McPhail, L. Liu, J. Am. Chem.
Soc. 1994, 116, 10255.
6 J. Fossey, D. Lefort, J. Sorba, Free Radicals in Organic
Chemistry, Wiley, Masson, 1995.
7
8
D. B. Gerth, B. Giese, J. Org. Chem. 1986, 51, 3726.
For benzyl bromide, the corresponding adduct with 1a was
obtained in 13 and 31% under the stoichiometric and the catalytic
conditions, respectively. Use of phenyl iodide resulted in 13 and
0%, respectively. The yield may become better by further optimi-
zation of the reaction conditions.
9
For general reviews of radical reactions for synthesis, see:
D. P. Curran, N. A. Porter, B. Giese, Stereochemistry of Radical
Reactions, VCH, Weinheim, 1996; Radicals in Organic Synthesis,
ed. by P. Renaud, M. P. Sibi, Wiley-VCH, Verlag, 2001, Vols. 1
and 2.