Radical zinc-atom transfer based multicomponent approaches to 3-alkylidene-substituted tetrahydrofurans

Article abstract of DOI:10.1055/s-0030-1259993

A domino 1,4-addition/alkyne carbozincation sequence based on a radical zinc-atom transfer process is disclosed. Two efficient multicomponent approaches to 3-alkylidenetetrahydrofurans from -(propargyloxy)enoates bearing pendant alkynes (including ynamides) have been established: one involving the direct addition of dialkylzincs, and the second involving the dimethylzinc-mediated addition of alkyl iodides. Both sequences utilize the stereoselective formation of intermediate alkylidenezincs well suited for in situ functionalization with electrophiles. 1 Introduction 2 1,4-Addition/Cyclization of Dialkylzincs on -(Propargyloxy)enoates 2.1 -(Propargyloxy)enoates with a Pendant Terminal Alkyne 2.2 -(Propargyloxy)enoates with a Pendant Substituted Alkyne 2.3 -(Propargyloxy)enoates with a Pendant Ynamide 3 1,4-Addition/Cyclization of Alkylzinc Halides on -(Propargyloxy)enoates 4 Dialkylzinc-Mediated 1,4-Addition/Cyclization of Alkyl Iodides on -(Propargyloxy)enoates 5 Conclusion; Current and Future Work. Georg Thieme Verlag Stuttgart New York.

Full text of DOI:10.1055/s-0030-1259993

FEATURE ARTICLE
1347
Radical Zinc-Atom Transfer Based Multicomponent Approaches to
3-Alkylidene-Substituted Tetrahydrofurans
Radical
Z
inc-
A
a
tom Trans
b
fer Based D
o
r
mino 1,
i
4-Addi
c
tion/Cycliz
e
Institut Parisien de Chimie Moléculaire (UMR 7201), FR 2769, UPMC Univ Paris 06, CNRS, Bâtiment F 2ème et., Case 183,
4 place Jussieu, 75005 Paris, France
Fax +33(1)44277567; E-mail: fabrice.chemla@upmc.fr; E-mail: alejandro.perez_luna@upmc.fr
Received 22 December 2010; revised 14 February 2011
have then been engaged in situ in 1,2-additions to alde-
hydes, ketones, imines, or esters or in allylation reactions.
Some examples where the resulting species is an alkylzinc
that can be further elaborated have also been disclosed.15
Abstract: A domino 1,4-addition/alkyne carbozincation sequence
based on a radical zinc-atom transfer process is disclosed. Two ef-
ficient multicomponent approaches to 3-alkylidenetetrahydrofurans
from b-(propargyloxy)enoates bearing pendant alkynes (including
ynamides) have been established: one involving the direct addition They involve the formation of a primary alkylzinc species
by SH2 reaction of primary radicals with dialkylzincs,16–19
of dialkylzincs, and the second involving the dimethylzinc-mediat-
primary16b and secondary9 organozinc halides, and cop-
ed addition of alkyl iodides. Both sequences utilize the stereoselec-
tive formation of intermediate alkylidenezincs well suited for in situ
per–zinc mixed reagents.16b,c Conversely, radical-polar re-
functionalization with electrophiles.
actions involving the formation of a vinylzinc species are
barely known.20–23
1
2
Introduction
1,4-Addition/Cyclization of Dialkylzincs on b-(Propargyl-
oxy)enoates
In this context, some time ago, we reported preliminary
results wherein we showed that dialkylzinc derivatives re-
acted with b-(propargyloxy)enoates to afford the corre-
sponding zincated alkylidenetetrahydrofuran adducts that
2.1
2.2
b-(Propargyloxy)enoates with a Pendant Terminal Alkyne
b-(Propargyloxy)enoates with
a Pendant Substituted
Alkyne
could then undergo a subsequent in situ reaction.24
A
2.3
3
b-(Propargyloxy)enoates with a Pendant Ynamide
1,4-Addition/Cyclization of Alkylzinc Halides on b-(Prop- mechanism based on a radical zinc-atom transfer process
argyloxy)enoates
was proposed (Scheme 1). In an initiation step, oxidation
of the dialkylzinc by traces of oxygen produces a radical
4
5
Dialkylzinc-Mediated 1,4-Addition/Cyclization of Alkyl
Iodides on b-(Propargyloxy)enoates
that undergoes 1,4-addition onto the a,b-unsaturated ester
Conclusion; Current and Future Work
1 to afford an enoxy radical 2. Following 5-exo-dig cy-
Key words: zinc, radicals, tandem reaction, alkynes, metalation
clization on the alkyne moiety, a vinyl radical 3 is formed
and it reacts subsequently with the dialkylzinc via ho-
molytic substitution to give an alkenylzinc species 4 with
concomitant formation of a radical that propagates the
chain.
1
Introduction
The use of organozinc reagents as nontoxic radical precur-
sors or mediators is a field of growing interest.1 The fast
reaction of alkylzincs with oxygen has been known since
their discovery2 and even though its exact mechanism is
still under investigation,3–6 the fact that this oxidation
leads to the formation of alkyl radicals has been docu-
mented since the late 1960s.7
R1
R1
alkyne
ZnR2
carbozincation
CO2Me
O
CO2Me
CH2R2
R22Zn
[O]
O
1
4
R2
R22Zn
R1
SH2
1,4-addition
While diethyl- and dimethylzinc are established reagents
that initiate or mediate in the presence of oxygen a range
of radical reactions,1,8 radical zinc-atom transfer
reactions9 wherein an alkylzinc is transformed into a more
elaborated one via a radical chain transfer mechanism al-
lowing for a so-called radical-polar process are far more
rare. Most strategies have been based on the reduction of
enoxy radicals by diethyl- or dimethylzinc via homolytic
substitution at zinc (SH2). The zinc enolates produced fol-
lowing radical 1,4-additions10–13 or from a-iodo esters14
R1
O
O
CO2Me
CH2R2
CO2Me
R2
3
5-exo-dig
2
Scheme 1
The 1,4-addition/cyclization radical-polar crossover reac-
tion of zinc reagents with enoates bearing pendant alkynes
provides a potential multicomponent approach to poly-
functional five-membered cyclic molecules. In addition to
its potential synthetic value, we felt that it could also rep-
SYNTHESIS 2011, No. 9, pp 1347–1360
x
x
.
x
x
.2
0
1
1
Advanced online publication: 07.04.2011
DOI: 10.1055/s-0030-1259993; Art ID: Z55710SS
© Georg Thieme Verlag Stuttgart · New York

1348
F. Chemla et al.
FEATURE ARTICLE
Biographical Sketches
Prof. Fabrice Chemla was Prof. M. Julia. After a one- research interests are fo-
born in Paris, France, in year postdoctoral fellowship cused on the design and the
1963. After his diploma with Prof. R. W. Hoffmann development of new func-
degree from the Ecole (Marburg, Germany), he tionalized carbenoids, as
Supérieure de Physique et joined Prof. J. F. Normant’s well as on carbometalation
Chimie Industrielles de group (Paris) as an assistant reactions and the design of
Paris in 1987, he received professor in 1992. He was new metal-mediated reac-
his Ph.D. in 1990 from the appointed as a full professor tions for the total synthesis
Ecole Normale Supérieure in Université Pierre et Marie of biologically useful com-
under the supervision of Curie (Paris) in 2001. His pounds.
Florian Dulong was born in Curie under the guidance of et aux Energies Alterna-
Bernay, France, in 1986. He Prof. F. Chemla and Dr A. tives, France, where he is
studied chemistry at the Pérez-Luna. He then joined currently undertaking his
Ecole Normale Supérieure Dr M. Ephritikhine’s group Ph.D. studies working on
of Cachan (France) from at the Laboratoire de Chimie CO and CO2 activation sup-
2006 to 2010. In 2010, he de Coordination des Elé- ported by new redox struc-
carried out his M.Sc. thesis ments f, in the Commissari- tures for lanthanides.
at Université Pierre et Marie at aux Energies Atomiques
Dr Franck Ferreira was 1996. After a postdoctoral 2001. His current work is
born in France in 1968. He stay with Prof. J.-Y. focused on the use of 3-het-
obtained an engineer diplo- Lallemand at the Ecole ero-substituted propargyl-
ma from the Ecole Nation- Polytechnique (Palaiseau, metals for the asymmetric
ale Supérieure de Chimie de France), he was appointed synthesis of acetylenic azir-
Paris in 1992 and received assistant professor in J. F. idines, 1,2-amino alcohols,
his Ph.D. from the Univer- Normant’s group (Univer- and biologically interesting
sité Pierre et Marie Curie sité Pierre et Marie Curie, compounds.
Paris VI under the supervi- Paris VI) in 1998. He joined
sion of Prof. J.-P. Genêt in Prof. F. Chemla’s group in
Max Peter Nüllen was born Witulski. In 2007 he joined catalyzed Ullmann–Ziegler
in Düsseldorf, Germany, in the group of Prof. R. Göttli- cross-coupling reaction. In
1981. He studied chemistry ch at the Justus Liebig-Uni- 2009 he was a visiting stu-
at the Westfälische Wil- versität in Gieben, Germany dent for six months in Prof.
helms-Universität of Mün- where he is currently carry- F. Chemla’s group at Uni-
ster (Germany) from 2000 ing out his Ph.D. studies versité Pierre et Marie
to 2005 and received his di- working on the synthesis of Curie.
ploma thesis in 2005 under sterically hindered biaryls
the guidance of Prof. B. by means of a palladium-
Dr Alejandro Pérez-Luna University, Paris V) under (Université Pierre et Marie
was born in London, United the guidance of Dr L. Curie, Paris VI) as Chargé
Kingdom, in 1977. He stud- Micouin in 2003. After a de Recherche in Prof. F.
ied at the Ecole Nationale postdoctoral stay at the Uni- Chemla’s group. His scien-
Supérieure de Chimie de versity of Geneva (Switzer- tific interests include the
Paris, where he obtained an land) as a Lavoisier Fellow fields of metal-mediated
engineer diploma in 2000. in Prof. E. P. Kündig’s synthesis, organozinc chem-
He obtained his Ph.D. in the group, he obtained a perma- istry, and asymmetric syn-
laboratory of Prof. H.-P. nent position in CNRS in thesis.
Husson (Paris Descartes 2004 and returned to Paris
Synthesis 2011, No. 9, 1347–1360 © Thieme Stuttgart · New York

FEATURE ARTICLE
Radical Zinc-Atom Transfer Based Domino 1,4-Addition/Cyclizations
1349
resent a good working tool that would advance the estab- and 59%, respectively), albeit as a mixture of diastereo-
lishment of a new general entry to the carbozincation of mers (dr 75:25 and 70:30, respectively) (entries 3 and 4).
alkynes based on radical zinc-atom transfer. From this The modest level of diastereoselectivity obtained was in
twofold perspective, we have investigated the scope and line with that observed for the related 1,4-addition/cy-
limitations of the reaction, including substrate and nucleo- clization reactions of b-(allyloxy)enoates.16d As a conse-
phile variations. We disclose hereafter a full account of quence, the relative configuration between the
our work combining our recent advances and, when nec- propargylic substituent and the CO2Me group of the major
essary for the sake of clarity, a reminder of the most sig- isomers was assumed to be the same as for b-(allyl-
nificant initial results.
oxy)enoates.
In order to check for the possibility to functionalize the
expected vinylzinc species arising from the tandem se-
quence, we conducted deuterium labeling experiments
(Table 2). Much to our surprise, when the intermediate
obtained from dibutylzinc and enoate 1a was quenched
with deuterium oxide (entry 1), even if 5aa-D was ob-
tained in 33% yield as a mixture of diastereomers in a ratio
E/Z 51:49, reduced 5aa was also produced in 33% yield.
In addition, product 6, wherein the terminal hydrogen
atom of the alkyne had been exchanged by deuterium, was
also recovered in 10% yield. A similar trend was observed
with other electrophiles (entries 2 and 3). In particular,
treatment with iodine led to the formation of vinyl iodide
5aa-I (ratio E/Z 55:45) in 24% yield, but also reduced 5aa
was obtained in 37% yield and iodinated starting material
7 in 9% yield. Increasing the number of equivalents of
dibutylzinc had a limited (if not negative) impact since
similar amounts of deuterium incorporation were ob-
tained (entry 4). Conversely, using diethylzinc, the ratio
of zincation versus reduction was increased, but non-met-
alated 5ab was still produced in 23% yield. Worthy of
note, the E/Z ratio of the deuterated compounds varied
2
1,4-Addition/Cyclization of Dialkylzincs on
b-(Propargyloxy)enoates
b-(Propargyloxy)enoates are readily available compounds
that are efficiently prepared by reaction of the parent pro-
pargylic alcohol and methyl 2-(bromomethyl)acrylate.24
2.1
b-(Propargyloxy)enoates with a Pendant
Terminal Alkyne
We initiated our studies by considering substrates having
a pendant terminal alkyne (Table 1). We were glad to find
that dibutyl- and diethylzinc reacted smoothly in diethyl
ether at room temperature with enoate 1a to afford, after
acidic workup, the corresponding 1,4-addition/cyclization
products 5aa and 5ab in 54% and 55% yields (entries 1
and 2). Optimal conditions involved the use of three
equivalents dialkylzinc; the use of an equimolar quantity
of dialkylzinc generally resulted in lower yields.
Table 1 Reaction of Dialkylzinc Reagents with b-(Propargyl-
oxy)enoates 1a–c with Pendant Terminal Alkynesa
Table 2 In Situ Functionalization of Vinylzinc Reaction Product of
Dialkylzincs with Enoate 1aa
R1
H
R1
1) R22Zn
2) H3O+
H
H
H
E
CO2Me
O
1) R2Zn
(equiv)
CO2Me
H
E
O
CO2Me
+
+
1a
O
R2
CO2Me
R
O
2) EX
O
CO2Me
1
5
R
Entry
1d
Enoate R1
R2
Bu
Et
Product Yieldb (%) [dr]c
5-D (E = D)
5-I
5
6 (E = D)
7 (E = I)
(E = I)
1a
1a
1b
1c
H
5aa
54 [–]
Entry R
Equiv E
Products [yieldb (%) (ratio)]
Dd 5aa-D [33 (E/Ze 51:49)], 5aa [33], 6 [10]
2d
H
5ab
5ba
5ca
55 [–]
1c
2c
3
Bu
3
3
3
6
6
3e
pentyl
Ph
Bu
Bu
60 [75:25]
59 [70:30]
Bu
Bu
Bu
Et
Df
Ig
5aa-D [32 (E/Ze 51:49)], 5aa [35], 6 [7]
5aa-I [24 (E/Ze 55:45)], 5aa [37], 7 [9]
4
a Reaction conditions: (1) R2 Zn (3 equiv), Et2O, r.t., 20 h; (2) aq 1 M
2
HCl.
4
Dd 5aa-D [26 (E/Ze 65:35)], 5aa [32], –
Dd 5ab-D [39 (E/Ze 38:62)], 5ab [23], –
b Combined isolated yield after chromatography.
c Determined by 1H NMR analysis of the crude material. Relative con-
figuration as indicated.
5
a Reaction conditions: (1) R2Zn, Et2O, r.t., 20 h; (2) D2O, DCl, or I2,
d Ref. 24.
e Bu2Zn (2 equiv) was used.
THF.
b Determined by 1H NMR spectroscopy based on analysis of the crude
mixture with biphenyl as internal standard.
c Ref. 24.
Substrates 1b and 1c substituted at the propargylic posi-
tion by a pentyl and a phenyl group respectively, also re-
acted with dibutylzinc under the same conditions and
provided exomethylene-substituted tetrahydrofurans 5ba
(from 1b) and 5ca (from 1c) in good chemical yields (60%
d EX = D2O.
e Determined by 1H NMR analysis of the crude material. Assignment
was based on NOE experiments on 5ag (R = t-Bu).
f EX = DCl.
g EX = I2 in THF.
Synthesis 2011, No. 9, 1347–1360 © Thieme Stuttgart · New York

1350
F. Chemla et al.
FEATURE ARTICLE
both with the amount and the nature of the dialkylzinc mixture (entry 1). In contrast, and to our pleasure, in the
used, even though low diastereoselectivities were record- case of TMS-substituted enoate 1e, reaction with dibutyl-
ed in all cases.
and diethylzinc provided the corresponding vinylsilanes
(Z)-5ea and (Z)-5eb in good to excellent yields (67% and
91%, respectively) and in diastereomerically pure form
(entries 2 and 3). Similarly, the addition of dibutylzinc
onto 1f led to the formation of two diastereomeric (Z)-5fa
vinylsilanes in 61% combined yield in a 71:29 ratio (entry
4). Thus, in spite of the fact that, as previously, the levels
of diastereocontrol of the propargylic center on the newly
created stereogenic center were only moderate, the pres-
ence of a substituent in the propargylic position was well
tolerated and had no influence on the Z/E selectivity.
The formation of non-deuterated 5aa and 5ab was unfore-
seen. On the basis of the presence of products 6 and 7 that
provided evidence for the formation of an alkynyl anion
10 in the reaction medium, we first considered deprotona-
tion of the starting terminal alkyne 1a by the formed vinyl
species 9 as a possible reaction pathway (Scheme 2).25
However, to be consistent with the observed yields higher
than 50%, this possibility implies the formation of a gem-
Zn,Zn-bimetalated species but this was not detected after
either deuterium or iodine quench.26 In fact, recovery of 6
and 7 rather seems to indicate that 10 does not undergo the
1,4-addition/cyclization sequence readily. Therefore, it
seems reasonable to consider that, while in situ protona-
tion of 9 to some extent cannot be ruled out,27 compounds
5aa and 5ab arise mainly from reduction of the interme-
diate vinyl radical 8 by hydrogen-atom abstraction. The
possibility of an intramolecular hydrogen shift can be
eliminated based on the fact that no deuterium incorpora-
tion has been detected by 2H NMR elsewhere in the mol-
ecule.26 H-Abstraction must, consequently, occur
intermolecularly. Even though they remain unidentified,
possible hydrogen-atom donors might include molecules
leading to a-alkoxyalkyl radicals1b (i.e., the solvent, the
starting propargylic ethers or the final tetrahydrofurans)
but also alkylzinc or alkoxyzinc species.17 Consistent with
this picture, it can be argued that the higher metalation ra-
tio obtained with diethylzinc over dibutylzinc is related to
an easier zinc-atom transfer as a result of a weaker C–Zn
bond.6 Finally, concerning the diastereoselectivity of vi-
nylzinc formation, no clear-cut conclusion can be ad-
vanced (besides the fact that it is low in all cases), as we
have no means to estimate the possible consumption of 9
by protonation that might be stereoselective and substrate
dependent.
Table 3 Reaction of Dialkylzincs with b-(Propargyloxy)enoates
Having Ethyl- and TMS-Substituted Alkynesa
R1
R2
R2
R1
1) R32Zn
2) H3O+
R2
R1
+
O
O
CO2Me
R3
CO2Me
O
R3
CO2Me
(Z)-5
(E)-5
Entry Enoate R1
R2
R3
Product Yieldb Ratioc
(%)
(Z/E)
64:36
>95:5
>95:5
>95:5
>95:5
1d
2d
3d
4
1d
1e
1e
1f
Et
H
H
H
Ph
H
Bu
Bu
Et
5da
5ea
5eb
5fae
5ec
64
TMS
TMS
TMS
TMS
67
91
Bu
i-Prg
61f
49h
5d
1e
a Reaction conditions: (1) R3 Zn (3 equiv), Et2O, r.t., 20 h; (2) aq 1 M
2
HCl.
b Isolated yield after chromatography.
c Determined by 1H NMR analysis of the crude material; Z/E assign-
ment was on the basis of NOE experiments, see ref. 24 and Supporting
Information.
d Ref. 24.
e A 71:29 mixture of two diastereomeric (Z)-vinylsilanes was ob-
tained.
H
H
f Determined on the crude 1H NMR with biphenyl as internal standard.
g Prepared salt-free from i-PrMgCl.28
H-transfer
H
O
CO2Me
CH2R
O
CO2Me
h 40% of 3-(trimethylsilyl)prop-2-yn-1-ol was isolated.
CH2R
5aa, 5ab
8
1a
ZnR
ZnR
As illustrated by the reaction of diisopropylzinc with 1e,
the use of secondary dialkylzinc reagents was also possi-
ble (entry 5). Diastereomerically pure (Z)-5ec was ob-
tained in 49% yield. However, 3-(trimethylsilyl)prop-2-
yn-1-ol was also recovered in 40% yield. While it seems
likely that the propargylic alcohol arises from a competi-
tive process involving the b-fragmentation of a zinc eno-
late,16d it is not clear why enolate formation occurs with
secondary dialkylzincs and not with primary ones. One
might put forward that since homolytic substitution at zinc
is easier in the case of diisopropylzinc (the C–Zn bond is
weaker29 and a secondary i-Pr radical is produced) reduc-
tion to a zinc enolate of the intermediate enoxy radical
arising from radical 1,4-addition might occur.
H
R2Zn
CO2Me
O
O
CO2Me
CH2R
10
9
Scheme 2
2.1
b-(Propargyloxy)enoates with a Pendant
Substituted Alkyne
Substrates possessing ethyl- or trimethylsilyl-substituted
alkynes were next examined (Table 3). Reaction of dibu-
tylzinc with ethyl-substituted enoate 1d afforded, after hy-
drolysis, alkylidenefuran 5da in 64% yield as a 64:36 Z/E
Synthesis 2011, No. 9, 1347–1360 © Thieme Stuttgart · New York

FEATURE ARTICLE
Radical Zinc-Atom Transfer Based Domino 1,4-Addition/Cyclizations
1351
Prompted by a report on the formation of alkenylzinc in- ethyl ether afforded the corresponding styrenes 5 in good
termediates by reaction of tetrahydrofuran with phenyl- yields (71–81%). However, the diastereoselectivity varied
acetylenes in the presence of dimethylzinc/oxygen,20 we with the aryl substitution. While a 90:10 Z/E ratio was ob-
also became interested in the possibility of using dimeth- tained with acceptor 1g having no substituent on the aryl
ylzinc for our 1,4-addition/cyclization process. Not sur- moiety (entries 1 and 2), barely any selectivity was ob-
prisingly, owing to the fact that the oxidation of served with 1h having an ortho methoxycarbonyl substit-
dimethylzinc is slower than that of other dialkylzincs,6,30 uent (entries 4 and 5). This drop in selectivity occurred
no reaction occurred between dimethylzinc and 1e under regardless of the solvent used (Et2O or CH2Cl231). In con-
our previously optimized conditions involving the use of trast, dimethylzinc was found to react both with 1g and 1h
diethyl ether as solvent under an argon atmosphere con- to afford the corresponding styrenes 5gd and 5hd with ex-
taining only traces of oxygen. We, thus, examined the cellent diastereoselectivity (>95:5 Z/E ratio), albeit in
conditions reported for radical 1,4-additions of dimethylz- somewhat lower yields (entries 3 and 6).
inc wherein dichloromethane is used as the solvent and air
is bubbled into the reaction medium.10e After some exper-
Table 5 Reaction of Dialkylzincs with b-(Propargyloxy)enoates
Having Aryl-Substituted Alkynesa,b
imentation, we found that 1e reacted with dimethylzinc (5
equiv) completely within two hours at 0 °C in dichlo-
R1
romethane under bubbling air (Table 4, entry 1). Follow-
ing acidic hydrolysis, vinylsilane (Z)-5ed was obtained in
73% yield and in diastereomerically pure form. Further-
more, ethyl-substituted enoate 1d also reacted smoothly
and afforded vinylsilane (Z)-5dd in a somewhat lower
44% yield (entry 2), but interestingly as a single isomer,
in sharp contrast with the addition of dibutylzinc
(Table 3). Both (Z)-5ed and (Z)-5dd were, however, con-
taminated with around 10% of (Z)-5ee and (Z)-5de, re-
spectively, arising from the addition of the dichloromethyl
radical to 1d. 1,2-Dichloroethane was thus used as the sol-
vent and (Z)-5ed and (Z)-5dd were obtained pure in 76%
and 56% yields, respectively. In 1,2-dichloroethane, how-
ever, six equivalents dimethylzinc were required to reach
full conversion of the starting acceptors.
R1
R1
1) R22Zn
2) H3O+
+
O
O
CO2Me
CO2Me
R2
CO2Me
O
R2
(Z)-5
(E)-5
1
Entry R1
Enoate R2 Zn
Solvent Product Yieldc (%)
2
(equiv)
[ratio Z/E]d
1e
2e
3
H
H
H
1g
1g
1g
Bu2Zna (5) Et2O
Et2Zna (3)
Et2O
5ga
5gb
81 [88:12]
79 [90:10]
57 [>95:5]
71 [59:41]
Me2Znb (5) CH2Cl2 5gd
4
CO2Me 1h
Bu2Zna
(3)
Et2O
5ha
Table 4 Reaction of Dimethylzinc with b-(Propargyloxy)enoates
5
6
CO2Me 1h
CO2Me 1h
Bu2Zna (3) CH2Cl2 5ha
71 [40:60]
27f [>95:5]
Having Ethyl- and TMS-Substituted Alkynesa
Me2Znb
(5)
CH2Cl2 5hd
R
R
R
H
1) Me2Zn, O2
2) H3O+
H
+
CO2Me
a Reaction conditions: (1) R2 Zn, solvent, r.t., 20 h; (2) aq 1 M HCl.
O
O
CO2Me
2
Cl
b Reaction conditions: (1) Me2Zn, CH2Cl2, dry air (20 mL) bubbled
through at 0 °C for 1 h, then 0 °C, 1 h (2) aq 1 M HCl.
c Isolated yield after chromatography.
O
Me
CO2Me
Cl
1d
1e
(Z)-5dd
(Z)-5ed
(Z)-5de
(Z)-5ee
d Determined by 1H NMR analysis of the crude material. Z/E assign-
ment was on the basis of NOE experiments, see ref. 24 and supporting
information.
Entry Enoate
R
Me2Zn Solvent Product
Yieldb (%)
[ratio]c
(equiv)
(s)
e Ref. 24.
f The product was contaminated with ~10% of product resulting from
the addition of the dichloromethyl radical (R2 = CHCl2; Z/E >95:5).
1
2
3
4
5
1e
1d
1e
1e
1d
TMS
Et
5
5
5
6
6
CH2Cl2 5ed/5ee
CH2Cl2 5dd/5de
73 [90:10]
44 [89:11]
TMS
TMS
Et
DCE
DCE
DCE
5ed
5ed
5dd
53
76
56
In order to determine if the 1,4-addition/cyclization reac-
tion leads to the formation of an alkylidenezinc species in
the case of substrates having substituted alkynes, deuteri-
um labeling experiments were carried out (Table 6). In the
preliminary account of this work, we had shown that fully
metalated adducts were obtained from reactions involving
TMS- or phenyl-substituted enoates 1e and 1g that were
highly Z selective (see, for instance, entry 1). The interme-
diate alkenylzinc adducts could thus be reacted with vari-
ous electrophiles in a second in situ reaction. Conversely,
for adduct 5da arising from ethyl-substituted enoate 1d as
a mixture of diastereomers, whereas complete deuterium
a Reaction conditions: (1) Me2Zn, solvent, dry air (20 mL) bubbled
through at 0 °C for 1 h, then 0 °C, 1 h; (2) aq 1 M HCl.
b Combined isolated yield after chromatography.
c Determined by 1H NMR analysis of the crude material.
Enoates with aryl-substituted pendant alkynes also under-
went the addition/cyclization process smoothly (Table 5).
Reaction of dibutyl- or diethylzinc with 1g and 1h in di-
Synthesis 2011, No. 9, 1347–1360 © Thieme Stuttgart · New York

1352
F. Chemla et al.
FEATURE ARTICLE
ZnR2
labeling was observed for the (Z)-5da-D isomer, only par-
tial labeling was observed for the (E)-5da-D isomer (entry
2). This trend was further confirmed with the reaction of
dibutylzinc with acceptor 1g (entry 3). Even though the
pendant alkyne was substituted by an aryl group, full deu-
terium incorporation was observed for the (Z)-5ga-D iso-
mer, but only 17% for the (E)-5ga-D isomer.
2
ZnR2
R1
R1
ZnR2
R1
O
ZnR2
O
O
2
O
syn
O
anti
O
R2
R2
R2
OMe
OMe
OMe
(Z)-4
3
(E)-4
Zn-transfer
Table 6 Deuterium Labeling of Vinylzinc Reaction Product of Di-
alkylzincs with Enoates Having Substituted Pendant Alkynesa,b
D-H
R1
(D / H)
R1
R1
H
R1
H
R1
H
O
1) R22Zn
2) D2O
D
+
O
R1
CO2Me
H
O
CO2Me
O
O
CO2Me
R2
O
syn
R2
anti
D-H
CO2Me
CO2Me
CO2Me
1
(Z)-5-D
(E)-5-D
R2
R2
R2
3
(E)-5
(Z)-5
Entry Enoate R1
R2
Product
Yieldc (%)
[ratio Z/E]d
83 [>95:5]f
78 [66:34]g
78 [59:41]h
73i [>95:5]j
H-transfer
Scheme 3
1e
2e
3
1e
1d
1g
1e
TMS
Et
Bua
Bua
5ea-D
5da-D
5ga-D
5ed-D
For linear sp hybridized radicals (3, R1 = aryl), the com-
petition between metalation and reduction reflects the rel-
ative transfer rate of zinc atom (mainly from the syn side)
versus hydrogen atom (only from the anti side)
(Scheme 4). Seemingly, the electronic nature of the styryl
radical being reduced influences this relative rate. Where-
as in the case of substrate 1g leading to a phenyl-substitut-
ed styryl radical (3, R1 = Ph), Zn-transfer (syn) overrides
H-transfer (anti), in the case of substrate 1h having an
electron-withdrawing ortho-substituted phenyl ring (3,
R1 = 2-MeO2CC6H4) H-abstraction competes with ho-
molytic substitution. We cannot, however, provide a con-
vincing explanation for this trend.
2-MeO2CC6H4 Bua
TMS
Med
4
a Reaction conditions: (1) Bu2Zn (3 equiv), Et2O, r.t., 20 h; (2) D2O.
b Reaction conditions: (1) Me2Zn (5 equiv), CH2Cl2, dry air (20 mL)
bubbled through at 0 °C for 1 h, then 0 °C, 1 h; (2) D2O.
c Combined isolated yield after chromatography.
d Determined by 1H NMR analysis of the crude material.
e Ref. 24.
f Z-Isomer 99% D.
g Z-Isomer 92% D, E-isomer 47% D.
h Z-Isomer 99% D, E-isomer 17% D.
i The product was contaminated with 8% of the product arising from
the addition of the dichloromethyl radical (R2 = CHCl2; Z/E >95:5 Z-
isomer 98% D).
j Z-Isomer 98% D.
D-H
R1
(E)-5
anti
It was also important to check vinylzinc formation in the
case of reactions involving dimethylzinc as this dialkyl-
zinc is less prone to homolytic substitution at zinc (the C–
Zn bond is stronger29 and it leads to the production of a
high energy methyl radical). Gratifyingly, 98% D-incor-
poration was obtained for the (Z)-5ed-D adduct resulting
from dimethylzinc and 1e (entry 4), indicating that H-ab-
straction did not interfere with the SH2 step.
O
O
ZnR2
2
intramolecular SH2
syn
R2
(E)-4
OMe
3
Scheme 4
Bent sp2 hybridized radicals (R1 = TMS, Et) are consid-
ered to exist in two diastereomeric forms and to invert
with a very low barrier. Thus, the final reaction outcome
depends both on the relative ease of each form to undergo
zinc- or hydrogen-transfer and on the Z/E ratio
(Scheme 5). With ‘big’ R1 substituents (i.e., R1 = TMS35),
the Z/E equilibrium is largely in favor of the Z form that
can only react from the syn side. Thus, solely zinc transfer
is observed. With ‘smaller’ R1 substituents (i.e., R1 = Et),
the concentration of (E)-3 increases. Since it reacts exclu-
sively from the anti side, H-transfer becomes competitive.
The diastereoselectivity of the 1,4-addition/cyclization
process arises from the stereoselectivity of the zinc-atom
radical transfer, since vinylzincs are configurationally sta-
ble.32 To rationalize it, we propose an interpretation based
on a kinetically controlled SH2 step depending both on a
directing effect brought about by the methoxycarbonyl
group and on the structure (linear or bent)33,34 of the vinyl
radical being reduced. As a result of pre-coordination to
the ester group of the dialkylzinc undergoing homolytic
substitution, zinc-atom transfer occurs preferentially from
the syn side (Scheme 3). Conversely, competitive inter-
molecular H-atom transfer takes place only from the anti
side, less sterically hindered (Scheme 3).
The outcome of the reactions where dimethylzinc is used
can also be interpreted according to this model but they
Synthesis 2011, No. 9, 1347–1360 © Thieme Stuttgart · New York

FEATURE ARTICLE
Radical Zinc-Atom Transfer Based Domino 1,4-Addition/Cyclizations
1353
R1
Table 7 Reaction of Dialkylzincs with b-(Propargyloxy)enoates
Having Pendant Ynamidesa
R1
O
ZnR2
ZnR2
2
2
NR1R2
H
O
O
H
O
NR1R2
NR1R2
R3
1) R32Zn
2) H3O+
OMe
OMe
+
R2
O
R2
O
R3
CO2Me
(Z)-5
O
(Z)-3
(E)-3
CO2Me
CO2Me
H- and Zn-
transfer
(E)-5
1
(syn) Zn-transfer
(anti)
Entry Enoate NR1R2
R3
Product Yieldb dr (Z/E)c
(E)-4
(E)-5 and (Z)-4
(%)
Scheme 5
1
2
3
1i
1j
1j
NTsBn
Et
5ib
5jb
5ja
47
28
27
54:46
18:82
10:90
O
merit further comment. For all the substrates studied, the
exclusive formation of (E)-alkylidenezincs (E)-4 was ob-
served. Thus, zinc transfer from the anti side did not oc-
cur, consistent with the fact that as dimethylzinc is less
prone to homolytic substitution at zinc, activation by pre-
coordination to the Lewis basic methyl ester was neces-
sary. Worthy of note, in spite of a less favorable Zn-trans-
fer, no competitive H-transfer took place.
Et
N
N
O
Bu
O
4
1k
Et
5kb
52
2:98 (55:45)d
O
Bn
a Reaction conditions: (1) R3 Zn (3 equiv), Et2O, r.t., 20 h; (2) aq 1 M
2.3
b-(Propargyloxy)enoate with a Pendant
2
HCl.
Ynamide
b Combined isolated yield after chromatography.
c Determined by 1H NMR analysis of the crude material.
Seeking to produce even more functionalized alkenylzinc
derivatives, we next became interested in the use of pen- d A 55:45 mixture of two diastereomeric (E)-enamides was obtained.
dant ynamides since ynamides stabilized by an electron-
withdrawing group on nitrogen have been recently shown
to be excellent substrates for regioselective carbometala-
tion.36–38 Furthermore, as ynamides are electron-rich radi-
cal acceptors39–41 we felt that they would be well suited to
our radical-polar crossover sequence involving the addi-
tion of an electron-deficient enoxy radical.
(Scheme 6). High levels of deuterium incorporation (86%
and 74%) were obtained for both diastereomers of (E)-
enamide 5kb, indicating that, unlike for other substituted
alkynes, even though zincation of the intermediate vinylic
radical by homolytic substitution at zinc occurred from
the side opposite to the methoxycarbonyl group, compet-
itive H-transfer was only a minor side reaction. As a con-
sequence, this reaction corresponds to an anti
carbometalation process of the ynamide. This is of signif-
icant interest since previously reported anionic carbomet-
alation of ynamides were found to be syn.36–38
The required enoates were readily prepared by reaction of
methyl 2-(bromomethyl)acrylate and propargylic alco-
hols containing ynamide functionalities.42 With them in
hand, we investigated the 1,4-addition/cyclization process
(Table 7). Acyclic N-tosyl ynamide 1i reacted with dieth-
ylzinc to afford, after acidic hydrolysis, enamide 5ib in a
reasonable 47% yield, but with very poor diastereoselec-
tivity. Pleasingly, a much better E/Z ratio was obtained for
enamides 5jb (E/Z 82:18) and 5ja (E/Z 90:10) arising
from the reaction between diethyl- and dibutylzinc and
cyclic oxazolidin-2-one-derived ynamide-containing
enoate 1j (the E configuration of 5ja and 5jb was assigned
by NOE experiments42). However, product purification
proved problematic as a consequence of their low solubil-
ity, and only low isolated yields were obtained in these
two cases. Finally, in an attempt to prepare non-racemic
enamides, we investigated the reaction of diethylzinc with
ynamide 1k derived from an enantiopure oxazolidin-2-
one. Disappointingly, though not unsurprisingly, in spite
of a good 51% chemical yield, a 55:45 mixture of two
diastereomeric (E)-enamides was obtained. Thus, the se-
lectivity of the alkenylzinc formation was high, but barely
any induction was obtained from the chiral inductor.
O
D
O
N
1) Et2Zn
1k
O
CO2Me
2) D2O
60%
Bn
Et
5kb dr = 56 (86% D):44 (74% D)
Scheme 6
Following our above-described model, the results ob-
tained with oxazolidinone-derived ynamides can be ratio-
nalized as following (Scheme 7). According to previous
reports on radical additions to ynamides stabilized by
electron-withdrawing groups on nitrogen, 5-exo-dig cy-
clization of enoxy radical 11 resulting from the 1,4-addi-
tion step should lead to a vinyl radical 12 of E geometry
(two in the case of substituted oxazolidinone 1k since no
chiral induction is observed). As a consequence of the
electronegative nitrogen, interconversion of (E)-12 into
Deuteration of the product mixture arising from the reac-
tion between diethylzinc and 1k was investigated
Synthesis 2011, No. 9, 1347–1360 © Thieme Stuttgart · New York

1354
F. Chemla et al.
FEATURE ARTICLE
(Z)-12 (not represented in the scheme) should not be
fast.40 Complexation to the carbonyl group of the oxazoli-
dinone moiety activates the dialkylzinc towards homolyt-
ic substitution and thus directs the zinc-atom transfer to
produce stereoselectively vinylzinc (Z)-13.
Table 8 Reaction of Alkylzinc Reagents with b-(Propargyl-
oxy)enoates 1a and 1ea
R1
R1
1) R2M
2) H3O+
O
CO2Me
R2
O
CO2Me
R12Zn
O
O
1
5
Entry Enoate R1
R2M (equiv)a
Product Yield (%)
O
O
N
N
5-exo-dig
O
O
CO2Me
R2
CO2Me
1
2
1a
1a
1a
1a
1a
1a
1a
1a
1e
1e
1e
H
H
H
H
H
H
H
H
BuZnBr·LiBr (2)b
BuZnCl·LiCl (2)b
BuZnI·LiI (2)b
5aa
5aa
5aa
5aa
5aa
46c
42d
28d,e
46d
53d
40d
41c
41c
24c
76d
22c,g
2
R1
(E)-12
R
R1
11
intramolecular
Zn-transfer
3
4
BuZnBr·LiBr (3)b
BuZnBr·LiBr (6)b
O
ZnR1
5
O
N
BuZn(CuCN)Br·LiBr (2)f 5aa
O
6
CO2Me
R2
R1
7
s-BuZnBr·LiBr (2)b
t-BuZnBr·LiBr (2)b
5af
5ag
5ea
(Z)-13
8
Scheme 7
9
TMS BuZnBr·LiBr (3)b
10
11
TMS BuZnBr (5) + ZnBr2 (2.5) 5ea
TMS Bu2Zn·2 LiBr (3)b
5ea
3
1,4-Addition/Cyclization of Alkylzinc
Halides on b-(Propargyloxy)enoates
a Reaction conditions: (1) BuM, Et2O, r.t., 20 h; (2) aq 1 M HCl.
b Prepared from salt-free R2Li and ZnX2.
Seeking an additional synthetic value for our 1,4-addition/
cyclization protocol, we envisaged using organozinc ha-
lides (RZnX) as nucleophiles.43 These species show a
high functional group tolerance, are more readily avail-
able than dialkylzincs, and are suitable for zinc-atom
transfer based processes.16b
c Isolated yield after chromatography.
d Yield determined by 1H NMR spectroscopy based on analysis of the
crude mixture with biphenyl as internal standard.
e 41% yield was obtained based on recovered starting material.
f Prepared from salt-free BuLi, CuCN, and ZnBr2.
g 70% of 3-(trimethylsilyl)prop-2-yn-1-ol was isolated.
Gratifyingly, two equivalents of butylzinc bromide–lithi-
um bromide obtained by transmetalation of butyllithium
with zinc bromide reacted in diethyl ether with enoate 1a
having a pendant terminal alkyne to afford, after hydroly-
sis, methylenefuran 5aa in 46% yield (Table 8, entry 1).
Unfortunately, this yield could not be significantly im-
proved by modification of the reaction conditions. While
no difference was observed when zinc chloride was used
instead of zinc bromide to prepare the organozinc reagent
(entry 2), a decrease in conversion (with the same yield
based on recovered starting material though) was noted
when zinc iodide was used (entry 3). Copper–zinc mixed
reagent Bu(CuCN)ZnBr·LiBr also afforded 5aa, but again
in a comparable 40% yield (entry 6). Finally, the only
slight progress was achieved by increasing the amount of
organozinc reagent. Using six equivalents of butylzinc
bromide–lithium bromide, 1a yielded 53% of 5aa (entry
5). Interestingly however, the procedure was also well
suited to secondary and tertiary organozinc halides (en-
tries 7 and 8). Indeed, addition of sec-butyl- and tert-bu-
tylzinc bromide–lithium bromide (2 equiv) led in 41%
yield to 5af and 5ag, respectively.
subsequent treatment with deuterium oxide or iodine, only
the reduced compound 5aa was isolated in 48% and 51%
yields, respectively. Thus, in the case of organozinc ha-
lides, hydrogen-atom transfer seems to override homolyt-
ic substitution (but still enables propagation of a radical
chain) (Scheme 8).
Moreover, the 1,4-addition/cyclization process with orga-
nozinc halides could not be generalized to other sub-
strates. Reaction between TMS-substituted enoate 1e only
yielded 5ea in a very poor 24% yield (entry 9). Suspecting
H
R1
R1
H
O
O
CO2Me
R2
CO2Me
O
CO2Me
5-endo-dig
H-abstraction
R2
R2
2
3
5
R2ZnBr⋅LiBr
R2
path a'
path a
OZnX
R1
R1
R1
path b
R2ZnBr⋅LiBr
O
ZnR2
O
+
CO2Me
CO2Me
β-elimination
CO2Me
R2
Disappointingly, attempts to functionalize a putative vi-
nylzinc intermediate were unsuccessful. Following reac-
tion of butylzinc bromide–lithium bromide with 1a and
R2
1a, 1e
14
Scheme 8
Synthesis 2011, No. 9, 1347–1360 © Thieme Stuttgart · New York

FEATURE ARTICLE
Radical Zinc-Atom Transfer Based Domino 1,4-Addition/Cyclizations
1355
a deleterious influence of the lithium bromide present in ‘nucleophilic’ fragment being added would come from
the reaction mixture, we conducted two control experi- easy-to-access alkyl iodides.
ments (entries 10 and 11). In the first one (entry 10), bu-
R1
tylzinc bromide was prepared via the equilibration
between dibutylzinc and zinc bromide via the Schlenk
equilibrium.44 Even in the presence of excess zinc bro-
mide, 5ea was produced in 76% yield, a similar result to
that of the reaction with dibutylzinc (see Table 3, entry 2).
In the second one, non-salt-free dibutylzinc prepared by
mixing butyllithium (2 equiv) and zinc bromide (1 equiv)
was reacted with 1e. In sharp contrast with the results ob-
tained using salt-free dibutylzinc, 5ea was produced in
only 22% yield and 3-(trimethylsilyl)prop-2-yn-1-ol was
isolated in 70% yield.
R32Zn
R2
I
[O]
ZnR3
O
R3
I
O
R3
I-atom
transfer
OMe
(E)-4
R2
R2
Zn-atom
transfer
R1
radical
addition/
R32Zn
R1
cyclization
O
CO2Me
O
CO2Me
CH2R2
1
I-atom transfer
I
3
The formation of the latter is indicative of the role played
by lithium halides as it most likely implies the intermedi-
ate formation of a zinc enolate and its subsequent anionic
b-elimination (Scheme 8). A similar observation had been
made in our previous studies regarding additions on b-(al-
lyloxy)enoates.16d At the time, we suggested that preferen-
tial formation of zinc enolate 14 in the presence of lithium
bromide could be related to the formation of an ate com-
R2
I
R2
R1
15
O
O
OMe
R2
Scheme 9
plex such as (BuZnBr2 , Li+) that could either make pos-
–
Success of such an approach relies upon two significant
features. First, iodine-atom transfer between radical R3
arising from the dialkylzinc and the alkyl iodide has to
override competitive 1,4-addition on the Michael accep-
tor. Second, vinyl radical 3 resulting from the 1,4-addi-
tion/cyclization has to undergo the SH2 reaction faster
than competitive iodine-atom transfer with the alkyl io-
dide.
sible an initial polar conjugate addition or undergo SH2
with radical 2 more readily than butylzinc bromide. In
light of both these new results and our previous observa-
tions involving cyclization on alkenes, it seems that the
extent of ‘anionic’ b-fragmentation increases when the
rate of the radical cyclization decreases (it is known that
hex-5-ynyl radical undergoes 5-exo-cyclization with a
slower rate than hex-5-enyl radical).45 While we cannot
put forward a definitive answer, this trend is rather consis-
tent with an enhanced reduction rate of a-alkoxycarbonyl
radical 2 by SH2 in the presence of lithium halides, that
can thus eventually compete with radical cyclization de-
pending on the ease of cyclization of the radical.
Looking for a proof of concept, we first investigated the
addition of isopropyl iodide to several substrates in the
presence of dimethylzinc. This reagent was chosen, first
because in this case the chain would be propagated by a
poorly nucleophilic methyl radical49 (ensuring low 1,4-
addition rates) and second, because complete diastereose-
lective metalation had been observed for all the acceptors.
Using our previously established reaction conditions,
TMS-substituted enoate 1e was treated in dichloro-
methane at 0 °C with five equivalents of isopropyl iodide
in the presence of three equivalents of dimethylzinc
(Table 9, entry 1). Following acidic quench, the formation
in a good 60% yield of only two (diastereomerically pure)
compounds 5ec and 5ed in 87:13 ratio was observed. This
ratio could be increased to 93:7 (in 68% overall yield)
when ten equivalents of isopropyl iodide were used (entry
3). These results indicated first that, in spite of some com-
petitive addition of a methyl group, addition of the isopro-
pyl moiety was the major reaction pathway and second,
that vinyl iodide formation was not taking place. Further-
more, deuterium-labeling studies indicated full deuterium
incorporation in 5ec-D and 5ed-D (entry 2). In conse-
quence and as anticipated, an iodine-atom transfer could
be efficiently included into our radical 1,4-addition/cy-
clization zinc-atom transfer process. Interestingly, the use
of dichloromethane was not a problem for these reactions
since only traces of products resulting from the addition of
the dichloromethyl radical were observed.
4
Dialkylzinc-Mediated 1,4-Addition/Cycliza-
tion of Alkyl Iodides on b-(Propargyl-
oxy)enoates
Given the modest results obtained for the addition of or-
ganozinc halides, to try to broaden the synthetic scope of
our domino sequence, we next considered an alternative
approach depicted in Scheme 9. Building on the estab-
lished dialkylzinc-mediated radical additions of alkyl io-
dides,10,17,46–48 we reasoned that an iodine-atom transfer
step could be advantageously combined with the zinc-
atom radical transfer to provide a new radical-polar pro-
cess. The radical arising from the oxidation of the dialkyl-
zinc could react with an alkyl iodide to give a new radical
that would then undergo the 1,4-addition/cyclization se-
quence. Reduction of the resulting radical by the dialkyl-
zinc reagent would then produce a vinylzinc and a radical
to propagate the radical chain. Thus, in the end, the zinc
reagent would ensure the metalation step and could, thus,
be simple and readily (commercially) available, while the
Synthesis 2011, No. 9, 1347–1360 © Thieme Stuttgart · New York

1356
F. Chemla et al.
FEATURE ARTICLE
A similar reaction outcome was obtained with aryl-substi- The possibility to use other alkyl iodides was surveyed.
tuted acceptors (entries 4–7). In the presence of five The reaction between 1e and another secondary iodide,
equivalents of isopropyl iodide, enoate 1g or 1h led to a cyclohexyl iodide, proceeded well and afforded adducts
mixture of 5gc and 5gd or 5hc and 5hd in 85:15 or 84:16 5eh and 5ed in 56% yield but only in a 76:24 ratio (entry
and 58% or 50% overall yield, respectively. Again, pro- 10), somewhat lower than in the case of isopropyl iodide.
viding evidence for the intermediate formation of a vi- As a result of a slower iodine-atom transfer between the
nylzinc species, full deuterium incorporation was methyl radical and ethyl iodide, a large excess was re-
observed for adducts 5gc-D, 5gd-D, 5hc-D, and 5hd-D quired to override methyl addition to enoate 1e and reach
when the reactions were quenched with deuterium oxide.
reasonable levels of ethyl incorporation (entries 11 and
12). Using 10 equivalents, a mixture of 5eb and 5ed in a
55:45 ratio was obtained, and 20 equivalents were re-
quired to attain 77% ethyl transfer.
In contrast, following reaction under the same conditions,
enoate 1d having a pendant ethyl-substituted alkyne led to
the formation not only of alkylidenetetrahydrofurans 5dc
and 5dd, but also of vinyl iodide (Z)-15dc [5dc/5dd/(Z)- By contrast, no methyl addition was observed when
15dc 71:15:14] (entry 8). Quenching with deuterium ox- enoates 1e and 1g were reacted with five equivalents of
ide afforded 5dc-D and 5dd-D, thereby proving the inter- tert-butyl iodide in the presence of dimethylzinc (entries
mediate formation of two parent (E)-vinylzinc (entry 9).
13 and 14). However, this time, in addition to alkylidenes
Table 9 Dimethylzinc-Mediated 1,4-Addition/Cyclization of Alkyl Iodides on b-(Propargyloxy)enoatesa
R1
R1
I
R2I (5 equiv)
Me2Zn, O2
1)
2)
R1
E
E
+
+
O
1
O
O
CO2Me
R2
CO2Me
R2
CO2Me
E-X
Me
5dc
5dd
(Z)-15dc
5eb, 5ec, 5eh, 5eg
5hc
5ed
5hd
5gd
(Z)-15eg
5gc, 5gg
(Z)-15gg
Entry
1
Enoate
1e
R1
R2
E
Products
Yieldb (%) [ratio]c
60 [87:13]
TMS
TMS
TMS
Ph
i-Pr
i-Pr
i-Pre
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
Cy
H
D
H
H
D
H
D
H
D
H
H
H
H
H
5ec/5ed
2
1e
5ec-D/5ed-D
5ec/5ed
64 [87:13]d
68 [93:7]
3
1e
4
1g
5gc/5gd
58 [85:15]
5
1g
Ph
5gc-D/5gd-D
5hc/5hd
54 [85:15]f
50 [84:16]
6
1h
1h
1d
1d
1e
2-MeO2CC6H4
7
2-MeO2CC6H4
5hc-D/5hd-D
5dc/5dd/15dc
5dc-D/5dd-D/15dc
5eh/5ed
65 [84:16]g
72 [71:15:14]
75 [71:15:14]h
56 [76:24]
8
Et
9
Et
10
11
12
13
14
TMS
TMS
TMS
TMS
Ph
1e
Ete
5eb/5ed
66 [55:45]
1e
Eti
5eb/5ed
61 [77:23]
1e
t-Bu
t-Bu
5eg/5ed/15eg
5gg/5gd/15gg
77 [91:0:9]
80 [81:0:19]
1g
a Reaction conditions: (1) Me2Zn (3 equiv), R2I, CH2Cl2, dry air (20 mL) bubbled through at 0 °C for 1 h, then 0 °C, 1 h; (2) EX = HCl (aq 1
M), or D2O.
b Combined isolated yield after chromatography.
c Determined by 1H NMR analysis of the crude material.
d 5ec-D 95% D, 5ed-D 95% D.
e 10 equiv R2I were used.
f 5gc-D 95% D, 5gd-D 95% D.
g 5hc-D 95% D, 5hd-D 95% D.
h 5dc-D 95% D, 5dd-D 95% D.
i 20 equiv EtI were used.
Synthesis 2011, No. 9, 1347–1360 © Thieme Stuttgart · New York

FEATURE ARTICLE
Radical Zinc-Atom Transfer Based Domino 1,4-Addition/Cyclizations
1357
I
5eg and 5gg, formation of vinyl iodides (Z)-15eg and (Z)-
15gg was observed as a result of a more favorable iodine-
atom transfer between the tertiary iodide and the interme-
diate vinyl radical arising from the cyclization step. The
ratio of alkylidene versus iodide was dependent on the
substrate considered and varied from 91:9 for TMS-sub-
stituted 1e to 81:19 for phenyl-substituted 1g. In agree-
ment with this picture, starting from 1e, vinyl iodide
formation could be suppressed using diethylzinc as medi-
ator, since the zinc-atom transfer step was accelerated
(Scheme 10). However, this was done to the expense of a
diminished ratio of tert-butyl incorporation related to a
less favorable atom transfer between the ethyl radical and
tert-butyl iodide. As a consequence, a mixture of 5eg and
5eb in a 92:8 ratio was isolated in 66% yield (Scheme 10).
Et
Et
i-PrI (5 equiv)
1)
2)
Et
I
I
Me2Zn, O2
+
+
O
1d
O
O
CO2Me
CO2Me
CO2Me
I2
i-Pr
i-Pr
Me
(E)-15dc
(E)-15dd
(Z)-15dc
86% [(E)-15dc(+ 5dc)/(E)-15dd/(Z)-15dc = 62(+12):12:14]
I
Ph
t-BuI (5 equiv)
Me2Zn, O2
1)
Ph
I
1g
+
O
O
CO2Me
CO2Me
2)
I2
t-Bu
t-Bu
(E)-15gg
(Z)-15gg
85% [(E)-15gg/(Z)-15gg = 73:27]
Scheme 11
TMS
TMS
1)
2)
t-BuI (5 equiv)
Et2Zn (2 equiv)
1e
overrides iodine transfer from the alkyl iodide, and even
in the case of tert-butyl iodide only minor amounts of vi-
nyl iodide are formed.
+
O
O
CO2Me
t-Bu
CO2Me
Et
H3O+
5eg
5eb
66% (92:8)
R2
I
R1
O
Scheme 10
ZnR3
(anti)
(syn)
2
O
15
(E)-4
Zn-transfer
Worthy of note, vinyl iodides 15 arising from iodine-atom
transfer were found to be of Z-configuration, thus evi-
dencing a stereoselective process. This conclusion was
reached by carrying out iodine quench experiments
(Scheme 11). Following reaction between 1d and isopro-
pyl iodide, iodine quench led to the formation of three vi-
nyl iodides (E)-15dc, (E)-15dd, and (Z)-15dc in a
62:12:14 ratio (12% of 5dc were also present), (E)-15dc
and (Z)-15dc having incorporated an isopropyl group [the
ratio of isopropyl to methyl incorporation was the same as
for other substrates (88:12)]. Since (E)-15dc was formed
by iodination of the intermediate vinylzinc, it should be of
E configuration, thus implying the Z configuration of (Z)-
15dc. Similarly, treatment with iodine of the reaction mix-
ture resulting from the addition of tert-butyl iodide to 1g
afforded a mixture of diastereomeric iodides (E)-15gg and
(Z)-15gg in a 73:27 ratio (Scheme 11). The first one aris-
ing from the parent vinyl zinc it should be of E configura-
tion, hence the Z-configuration of the second one.
I-transfer
OMe
R2
Scheme 12 (R1 = aryl)
In the case of rapidly interconverting bent radicals
(Scheme 13), while the radical of Z-configuration under-
goes exclusive zinc-atom transfer, the radical of E config-
uration only reacts with the alkyl iodide. The final
metalation versus iodination depends again on the relative
rate of this elementary steps (i.e., from TMS-substituted
1e, iodide formation is observed from tert-butyl iodide but
not with isopropyl iodide, and it is observed using di-
methylzinc but not with diethylzinc), but also on the Z/E
ratio of the reacting vinylic radical [i.e. for the same alkyl
iodide (i-PrI), vinyl iodide formation is observed from
ethyl-substituted radical 3 (R1 = Et, R2 = i-Pr), but not
from TMS-substituted radical 3 (R1 = TMS, R2 = i-Pr) for
which the amount of Z/E ratio should be higher]. General-
ly however, as for linear radicals, zinc-atom transfer pre-
dominates over iodine-atom transfer for bent radicals.
The model proposed above to account for the selectivity
of Zn- versus H-atom transfer in the case of direct addition
of dialkylzincs also provides an adequate explanation for
the results of the alkyl iodide additions. As previously
stated, both for linear and bent radicals, zinc-atom transfer
from dimethylzinc only occurs from the syn side. Con-
versely, as evidenced by the observed Z-configurations of
the vinyl iodides formed, iodine-atom transfer only takes
place from the anti side, less sterically hindered.50
R2
I
R1
R1
O
ZnR3
ZnR3
2
2
O
O
O
OMe
OMe
R2
(Z)-3
(syn)
R2
(E)-3
In the case of linear radicals (Scheme 12), the competition
between metalation and iodination reflects the relative
ease of the latter to undergo zinc-atom transfer versus io-
dine-atom transfer. As illustrated in the case of phenyl-
substituted enoate 1g, SH2 on the complexed dimethylzinc
Zn-transfer
(anti)
I-transfer
(E)-4
15
Scheme 13 (R1 = TMS, alkyl)
Synthesis 2011, No. 9, 1347–1360 © Thieme Stuttgart · New York

1358
5
F. Chemla et al.
FEATURE ARTICLE
mations.51 It is remarkable by the fact that an alkylzinc de-
rivative is transformed into a vinylzinc in what represents
Conclusion; Current and Future Work
In conclusion, we have disclosed a new stereoselective one of a few examples of a radical zinc-atom transfer be-
domino 1,4-addition/alkyne carbozincation sequence tween a dialkylzinc and a vinyl radical. Moreover, this
based on a radical zinc-atom transfer reaction. In terms of work has brought about valuable information concerning
synthetic value, two efficient multicomponent approaches the key homolytic substitution at zinc (SH2) step and we
to polysubstituted alkylidene tetrahydrofurans from b- hope that our findings will foster future studies to gener-
(propargyloxy)enoates have been established: one involv- alize this new entry to alkyne carbozincation. Amongst
ing the direct addition of dialkylzincs, and the second in- other, it provides further evidence that adjacent Lewis ba-
volving the dimethylzinc-mediated addition of alkyl sic functions such as esters or oxazolidinones, presumably
iodides. Both sequences involve the stereoselective for- by activating by coordination the dialkylzinc towards ho-
mation of an alkylidenezinc intermediate that is well suit- molytic substitution,3,14b,17 are able to direct efficiently the
ed for subsequent functionalization by in situ reaction zinc transfer and thus ensure highly diastereoselective
with electrophiles. Having shown that the processes toler- transformations. Significantly, in terms of reactivity, this
ate well a range of alkyne substituents, future work in our activation by coordination provides zinc-atom transfer re-
group will be directed to extending the approach to the actions so efficient that they can override processes as fast
synthesis of other hetero- or carbocycles by using diverse- as iodine-atom transfer between a vinyl radical and tert-
ly linked 1,6-enynes. Preliminary studies involving the butyl iodide.
enoates with a malonate linkage have shown a behavior
very similar to that of b-(propargyloxy)enoates
Experiments involving organometallic compounds were carried out
(Table 10).
in dried glassware under a positive pressure of dry argon. All sol-
vents were distilled to remove stabilizers and dried with an Mbraun
Solvent Purification System SPS-800. ZnBr2 (98%) was melted un-
der dry N2 and, immediately after cooling to r.t., was dissolved in
Et2O. Bu2Zn (Fluka, ~1 M in heptane), Et2Zn (Aldrich, 1.0 M in
hexanes), Me2Zn (Aldrich, 1.0 M in heptane) and all other reagents
were of commercial quality and were used without purification. 1H
and 13C NMR spectra were recorded with a Bruker Avance 400
spectrometer fitted with BBFO probe with Z gradient with an inter-
nal standard of residual CHCl3 (d = 7.27 for 1H NMR and 77.16 for
13C NMR). IR spectra were recorded with an ATR diamond spec-
trophotometer. HRMS were obtained on a Finnigan MAT 95.
Table 10 Reaction of Dialkylzinc Reagents with Malonate-Derived
Enoates 16a and 16ba,b
R1
R2
CO2Me
R1
R2
CO2Me
R1
CO2Me
1) R22Zn
2) H3O+
+
CO2Me
CO2Me
CO2Me
MeO2C
MeO2C
MeO2C
16
(Z)-17
(E)-17
Entry
Enoate R1
R2
Product Yieldc
(%)
drd
(Z/E)
Methyl 4-Methylene-3-pentyltetrahydrofuran-3-carboxylate
(5aa) by Reaction of Bu2Zn or Et2Zn with b-(Propargyl-
oxy)enoates (Table 1, Entry 1);24 Typical Procedure
1
2
3
16a
16a
16b
TMS
Eta
17ab
17ad
17bb
70
>95:5
>95:5
70:30
Under argon, to a stirred soln of b-(propargyloxy)enoate 1a (0.5
mmol) in Et2O (5 mL) at r.t. was added ~1 M Bu2Zn in heptane (0.6
mL, 0.6 mmol). The mixture was stirred at r.t. for 20 h. The reaction
was hydrolyzed with aq 1 M HCl (10 mL). The layers were separat-
ed and the aqueous layer was extracted with Et2O (2 × 15 mL). The
combined organic layers were washed with brine and dried
(MgSO4), and the solvents were evaporated under reduced pressure.
Purification by flash chromatography (silica gel, pentane–Et2O) af-
forded 5aa (57 mg, 54%).
TMS
Et
Meb
Eta
53e
50f
a Reaction conditions: (1) Et2Zn (3 equiv), Et2O, r.t., 20 h; (2) aq 1 M
HCl.
b Reaction conditions: (1) Me2Zn (5 equiv), CH2Cl2, dry air (20 mL)
bubbled through at 0 °C for 1 h, then 0 °C, 1 h; (2) aq 1 M HCl.
c Isolated yield after chromatography.
d Determined by 1H NMR analysis of the crude material.
e The product was contaminated with 8% of the product resulting
from the addition of the bichloromethyl radical (R2 = CHCl2).
f Determined by 1H NMR spectroscopy based on analysis of the crude
mixture with biphenyl as internal standard.
IR (neat): 2954, 2928, 2859, 1731, 1379 cm–1.
1H NMR (400 MHz, CDCl3): d = 0.89 (t, J = 7.0 Hz, 3 H), 1.20–
1.32 (m, 6 H), 1.61 (m, 1 H), 1.96 (m, 1 H), 3.73 (s, 3 H), 3.74 (d,
J = 9.0 Hz, 1 H), 4.34 (dt, J = 13.1, 2.2 Hz, 1 H), 4.37–4.43 (m, 2
H), 5.07 (t, J = 2.1 Hz, 1 H), 5.22 (t, J = 2.4 Hz, 1 H).
13C NMR (100 MHz, CDCl3): d = 13.9, 22.4, 25.2, 32.0, 37.1, 52.3,
Diethylzinc addition on enoates 16a and 16b having re-
spectively a TMS- and ethyl-substituted pendant alkyne
led to the corresponding alkylidene cyclopentanes 17ab
and 17bb in yields and selectivities similar to those of the
analogous b-(propargyloxy)enoates. Dimethylzinc addi-
tion/cyclization on 16a provided 17ad in 53% yield (entry
2).
56.8, 71.7, 75.0, 106.2, 150.4, 173.5.
HRMS: m/z [M + Na]+ calcd for C12H20O3Na: 235.13047; found:
235.13043.
Methyl (Z)-3-Ethyl-4-[(trimethylsilyl)methylene]tetrahydrofu-
ran-3-carboxylate (5ed) by Reaction of Me2Zn with b-(Propar-
gyloxy)enoates (Table 4, Entry 4); Typical Procedure
Under argon, to a stirred soln of b-(propargyloxy)enoate 1e (0.2
mmol) in DCE (1 mL) at 0 °C was added 1.0 M Me2Zn in heptane
(1.2 mL, 1.2 mmol). Air (20 mL) was slowly introduced over 1 h
into the soln via a syringe pump using a syringe fitted with a CaCl2
guard. The mixture was then stirred at 0 °C for 1 h. CH2Cl2 (5 mL)
From a mechanistic perspective, the above described
transformations are to be related to carbozincations of
non-activated alkynes that are highly valuable transfor-
Synthesis 2011, No. 9, 1347–1360 © Thieme Stuttgart · New York

FEATURE ARTICLE
Radical Zinc-Atom Transfer Based Domino 1,4-Addition/Cyclizations
1359
was then added and the reaction was hydrolyzed with aq 1 M HCl
(5 mL). The layers were separated and the aqueous layer was ex-
tracted with CH2Cl2 (3 × 10 mL). The combined organic layers were
washed with brine and dried (MgSO4), and the solvents were evap-
orated under reduced pressure. Purification by flash chromatogra-
phy (silica gel, pentane–Et2O) afforded 5ed (37 mg, 76%).
Acknowledgment
Prof. R. Göttlich and Justus Liebig University Giessen are acknow-
ledged for a visiting grant (MPN). The authors thank Dr. E. Caytan
for help with NMR and N. Ghézali for synthesis of some starting
materials.
IR (neat): 2953, 2360, 1731, 1627, 837 cm–1.
1H NMR (400 MHz, CDCl3): d = 0.08 (s, 9 H), 0.86 (t, J = 7.4 Hz,
3 H), 1.64 (m, 1 H), 1.98 (m, 1 H), 3.70 (s, 3 H), 3.74 (d, J = 7.2 Hz,
1 H), 4.32 (dd, J = 13.6, 2.3 Hz, 1 H), 4.38 (dd, J = 13.6, 2.3 Hz, 1
H), 4.38 (d, J = 7.2 Hz, 1 H), 5.67 (t, J = 2.3 Hz, 1 H).
13C NMR (100 MHz, CDCl3): d = 173.5, 158.0, 120.1, 73.8, 71.1,
60.0, 52.4, 30.2, 10.0, 0.7.
References
(1) Reviews: (a) Bazin, S.; Feray, L.; Bertrand, M. P. Chimia
2006, 60, 260. (b) Akindele, T.; Yamada, K.-i.; Tomioka, K.
Acc. Chem. Res. 2009, 42, 345.
(2) Historical account, see: Seyferth, D. Organometallics 2001,
20, 2940.
(3) (a) Lewinski, J.; Marciniac, W.; Lipkowski, J.; Justyniak, I.
J. Am. Chem. Soc. 2003, 125, 12698. (b) Lewinski, J.;
Sliwinski, W.; Dranka, M.; Justyniak, I.; Lipkowski, J.
Angew. Chem. Int. Ed. 2006, 45, 4826. (c) Lewinski, J.;
Suwala, K.; Kubisiak, M.; Ochal, Z.; Justyniak, I.;
Lipkowski, J. Angew. Chem. Int. Ed. 2008, 47, 7888.
(d) Lewinski, J.; Bury, W.; Dutkiewicz, M.; Maurin, M.;
Justyniak, I.; Lipkowski, J. Angew. Chem. Int. Ed. 2008, 47,
573. (e) Lewinski, J.; Koscielski, M.; Suwala, K.; Justyniak,
I. Angew. Chem. Int. Ed. 2009, 48, 7017. (f) Lewinski, J.;
Suwala, K.; Kaczorowski, T.; Galezowski, M.; Gryko, D. T.;
Justyniak, I.; Lipkowski, J. Chem. Commun. 2009, 215.
(4) Jana, S.; Berger, R. J. F.; Fröhlich, R.; Pape, T.; Mitzel, N.
W. Inorg. Chem. 2007, 46, 4293.
HRMS: m/z [M + Na]+ calcd for C12H22O3NaSi: 265.12304; found:
265.12251.
Methyl 4-Methylene-3-pentyltetrahydrofuran-3-carboxylate
(5aa) by Reaction of RZnBr·LiBr with b-(Propargyloxy)enoates
(Table 8, Entry 1); Typical Procedure
Under argon, to stirred 1 M ZnBr2 in Et2O (1.0 mL, 1.0 mmol) was
added dropwise 2.2 M BuLi in hexane (0.45 mL, 0.99 mmol) at –60
°C. After 5 min, the initial slurry gave a colorless soln. The mixture
was further stirred at r.t. for 15 min before it was cooled again to
–30 °C. Et2O (2 mL), and a soln of b-(propargyloxy)enoate 1a (0.5
mmol) in Et2O (2 mL) were added. The mixture was stirred at r.t.
for 20 h. The reaction was hydrolyzed with aq 1 M HCl (10 mL).
The layers were separated, the aqueous layer was extracted with
Et2O (2 × 15 mL). The combined organic layers were washed with
brine and dried (MgSO4), and the solvents were evaporated under
reduced pressure. Purification by flash chromatography (silica gel,
pentane–Et2O) afforded 5aa (49 mg, 46%).
(5) Mileo, E.; Benfatti, F.; Cozzi, P. G.; Lucarini, M. Chem.
Commun. 2009, 469.
(6) Maury, J.; Feray, L.; Bazin, S.; Clément, J.-L.; Marque, S. R.
A.; Siri, D.; Bertrand, M. P. Chem. Eur. J. 2011, 17, 1586.
(7) Davies, A. G.; Roberts, B. P. Acc. Chem. Res. 1972, 387; and
references cited therein.
(8) For the first use of Et2Zn as radical mediator in the presence
of oxygen in the context of synthesis, see: Ryu, I.; Araki, F.;
Minakata, S.; Komatsu, M. Tetrahedron Lett. 1998, 39,
6335.
Methyl (Z)-3-Isobutyl-4-[(trimethylsilyl)methylene]tetrahydro-
furan-3-carboxylate (5ec) by Me2Zn-Mediated Reaction of
Alkyl Iodides with b-(Propargyloxy)enoates having Pendant
Alkynes (Table 9, Entry 3); Typical Procedure
Under argon, to a stirred soln of b-(propargyloxy)enoate 1e (0.2
mmol) and i-PrI (2 mmol) in CH2Cl2 (1 mL) at 0 °C was added 1.0
M Me2Zn in heptane (0.6 mL, 0.6 mmol). Air (20 mL) was slowly
introduced over 1 h into the soln via a syringe pump using a syringe
fitted with a CaCl2 guard. The mixture was then stirred at 0 °C for
1 h. CH2Cl2 (5 mL) was the added and the reaction was hydrolyzed
with aq 1 M HCl (5 mL). The layers were separated, the aqueous
layer was extracted with CH2Cl2 (3 × 10 mL). The combined organ-
ic layers were washed with brine and dried (MgSO4), and the sol-
vents were evaporated under reduced pressure. Purification by flash
chromatography (silica gel, pentane–Et2O) afforded 5ec (36 mg,
68%).
(9) Cohen, T.; Gibney, H.; Ivanov, R.; Yeh, E. A.-H.; Marek, I.;
Curran, D. P. J. Am. Chem. Soc. 2007, 129, 15405.
(10) (a) Bertrand, M. P.; Feray, L.; Nouguier, R.; Perfetti, P.
J. Org. Chem. 1999, 64, 9189. (b) Bazin, S.; Feray, L.; Siri,
D.; Naubron, J.-V.; Bertrand, M. P. Chem. Commun. 2002,
2506. (c) Bazin, S.; Feray, L.; Vanthuyne, N.; Bertrand, M.
P. Tetrahedron 2005, 61, 4261. (d) Bazin, S.; Feray, L.;
Vanthuyne, N.; Siri, D.; Bertrand, M. P. Tetrahedron 2007,
63, 77. (e) Maury, J.; Feray, L.; Perfetti, P.; Bertrand, M. P.
Org. Lett. 2010, 12, 3590.
(11) Van der Deen, H.; Kellog, R. M.; Feringa, B. L. Org. Lett.
2000, 2, 1593.
(12) (a) Miyabe, H.; Asada, R.; Yoshida, K.; Takemoto, Y.
Synlett 2004, 540. (b) Miyabe, H.; Asada, R.; Takemoto, Y.
Tetrahedron 2005, 61, 385.
IR (neat): 2954, 1731, 1626, 837 cm–1.
1H NMR (400 MHz, CDCl3): d = 0.09 (s, 9 H), 0.87 (d, J = 6.6 Hz,
3 H), 0.89 (d, J = 6.6 Hz, 3 H), 1.46 (dd, J = 13.9, 6.2 Hz, 1 H), 1.60
(m, 1 H), 2.02 (dd, J = 13.9, 7.1 Hz, 1 H), 3.66 (d, J = 9.0 Hz, 1 H),
3.69 (s, 3 H), 4.28 (dd, J = 13.6, 2.4 Hz, 1 H), 4.40 (dd, J = 13.6, 2.4
Hz, 1 H), 4.53 (d, J = 9.0 Hz, 1 H), 5.74 (t, J = 2.4 Hz, 1 H).
(13) Yamada, K.-i.; Umeki, M.; Maekawa, M.; Yamamoto, Y.;
Akindele, Y.; Nakano, M.; Tomioka, K. Tetrahedron 2008,
64, 7258.
(14) (a) Cozzi, P. G. Angew. Chem. Int. Ed. 2006, 45, 2951.
(b) Cozzi, P. G. Adv. Synth. Catal. 2006, 348, 2075.
(c) Cozzi, P. G.; Mignogna, A.; Vicennati, P. Adv. Synth.
Catal. 2008, 350, 975. (d) Fernández-Ibañez, M. A.; Maciá,
B.; Minnaard, A. J.; Feringa, B. L. Angew. Chem. Int. Ed.
2008, 47, 1317. (e) Cozzi, P. G.; Benfatti, F.;
13C NMR (100 MHz, CDCl3): d = 0.3, 23.2, 24.4, 26.5, 46.4, 52.8,
58.4, 71.0, 74.5, 120.8, 158.7, 174.1.
HRMS m/z [M + Na]+ calcd for C14H26O3NaSi: 293.15434; found:
293.15470.
Guiteras Capdevila, M.; Mignogna, A. Chem. Commun.
2008, 3317.
Supporting Information for this article is available online at
http://www.thieme-connect.com/ejournals/toc/synthesis.
Synthesis 2011, No. 9, 1347–1360 © Thieme Stuttgart · New York

1360
F. Chemla et al.
FEATURE ARTICLE
(15) Furthermore, it has been proposed that transition-metal-
promoted zinc/iodine exchange involves a radical
mechanism, Cu(I): (a) Rozema, M. J.; AchyuthaRao, S.;
Knochel, P. J. Org. Chem. 1992, 57, 1956. Pd(0):
(b) Stadtmüller, H.; Vaupel, A.; Tucker, C. E.; Stüdemann,
T.; Knochel, P. Chem. Eur. J. 1996, 2, 1204.
(31) When CH2Cl2 was used as solvent, the Z/E ratio of product
5ha might not fully represent the diastereoselectivity of the
zinc-atom transfer since in addition to these two
diastereomers, a third tetrahydrofuran side product having
incorporated a butyl residue was detected (~20% yield in the
crude) but could not be fully identified. In any case the
diastereoselectivity of the reaction between 1h and Bu2Zn in
CH2Cl2 should at best be mediocre.
(32) Guijarro, A. In The Chemistry of Organozinc Compounds;
Rappoport, Z.; Marek, I., Eds.; Wiley: Chichester, 2007,
193–236.
(33) Jourmet, M.; Malacria, M. J. Org. Chem. 1992, 57, 3085.
(34) Curran, D. P.; Chen, M.-H.; Kim, D. J. J. Am. Chem. Soc.
1989, 111, 6265.
(35) Even though a-silyl stabilization of vinyl radicals leads to a
tendency towards linearization: (a) Lalitha, S.;
Chandrasekhar, J. Proc.–Indian Acad. Sci., Chem. Sci. 1994,
106, 259; they are better represented as very rapidly Z/E
interconverting bent radicals:. (b) Rubin, H.; Fischer, H.
Helv. Chim. Acta 1996, 79, 1670. (c) Bucher, G.; Mahajan,
A. A.; Schmittel, M. J. Org. Chem. 2009, 74, 5850.
(36) (a) Chechik-Lankin, H.; Livshin, S.; Marek, I. Synlett 2005,
2098. (b) Prakash Das, J.; Chechik, H.; Marek, I. Nature
Chem. 2009, 1, 128.
(37) (a) Gourdet, B.; Lam, H. W. J. Am. Chem. Soc. 2009, 131,
3802. (b) Gourdet, B.; Rudkin, M. E.; Watts, C. A.; Lam, H.
W. J. Org. Chem. 2009, 74, 7849.
(38) (a) Yasui, H.; Yorimitsu, H.; Oshima, K. Chem. Lett. 2007,
36, 32. (b) Yasui, H.; Yorimitsu, H.; Oshima, K. Bull. Chem.
Soc. Jpn. 2008, 81, 373.
(c) Stadtmüller, H.; Lentz, R.; Tucker, C.; Stüdemann, T.;
Dörner, W.; Knochel, P. J. Am. Chem. Soc. 1993, 115, 7027.
Ni(0): (d) Vaupel, A.; Knochel, P. Tetrahedron Lett. 1994,
35, 8349. (e) Vaupel, A.; Knochel, P. J. Org. Chem. 1996,
61, 5743. Mn(II)/Cu(I): (f) Riguet, E.; Klement, I.;
Kishan Reddy, Ch.; Cahiez, G.; Knochel, P. Tetrahedron
Lett. 1996, 37, 5865.
(16) (a) Denes, F.; Chemla, F.; Normant, J.-F. Angew. Chem. Int.
Ed. 2003, 42, 4043. (b) Denes, F.; Pérez-Luna, A.; Cutri, S.;
Chemla, F. Chem. Eur. J. 2006, 12, 6506. (c) Denes, F.;
Pérez-Luna, A.; Chemla, F. J. Org. Chem. 2007, 72, 398.
(d) Giboulot, S.; Pérez-Luna, A.; Botuha, C.; Ferreira, F.;
Chemla, F. Tetrahedron Lett. 2008, 49, 3963.
(17) Feray, L.; Bertrand, M. P. Eur. J. Org. Chem. 2008, 3164.
(18) For a related photoinduced reaction, see: Charette, A. B.;
Beauchemin, A.; Marcoux, J.-F. J. Am. Chem. Soc. 1998,
120, 5114.
(19) For important studies on radical formation by SET from
dialkylzinc coordination complexes, see: (a) Wissing, E.;
Rijnberg, E.; van der Schaaf, P. A.; van Gorp, K.; Boersma,
J.; van Koten, G. Organometallics 1994, 13, 2609.
(b) Wissing, E.; van der Linden, S.; Rijnberg, E.; Boersma,
J.; Smeets, W. J. J.; Spek, A. L.; van Koten, G.
Organometallics 1994, 13, 2602; and references cited
therein.
(20) Chen, Z.; Zhang, Y.-X.; An, Y.; Song, X.-L.; Wang, Y.-H.;
Zhu, L.-L.; Guo, L. Eur. J. Org. Chem. 2009, 5146.
(21) Reductive zincation of a vinyl radical has also been
suggested to account for the formation of allenoates
following zinc-mediated radical addition to propiolates, see
ref. 17.
(39) Marion, F.; Courillon, C.; Malacria, M. Org. Lett. 2003, 5,
5095.
(40) (a) Sato, A.; Yorimitsu, H.; Oshima, K. Synlett 2009, 28.
(b) Sato, A.; Yorimitsu, H.; Oshima, K. Bull. Korean Chem.
Soc. 2010, 31, 570.
(41) Banerjee, B.; Litvinov, D. N.; Kang, J.; Bettale, J. D.; Castle,
S. L. Org. Lett. 2010, 12, 2650.
(22) Vinylcyclopentyl zinc has been prepared by reductive
cyclization: Crandall, J. K.; Ayers, T. A. Organometallics
1992, 11, 473.
(23) Ni(II)-catalyzed intermolecular carbozincation, see:
Stüdemann, T.; Knochel, P. Angew. Chem. Int. Ed. 1997, 36,
93.
(42) See the Supporting Information.
(43) (a) The Chemistry of Organozinc Compounds; Rappoport,
Z.; Marek, I., Eds.; Wiley: Chichester, 2007. (b) Knochel,
P.; Millot, N.; Rodrigues, A. L. Org. React. 2001, 58, 417.
(44) Blake, A. J.; Shannon, J.; Stephens, J. C.; Woodward, S.
Chem. Eur. J. 2007, 13, 2462.
(24) Pérez-Luna, A.; Botuha, C.; Ferreira, F.; Chemla, F. Chem.
Eur. J. 2008, 14, 8784.
(45) Beckwith, A. L. J. Tetrahedron 1981, 37, 3073.
(46) (a) Bertrand, M. P.; Feray, L.; Nouguier, R.; Perfetti, P.
Synlett 1999, 1148. (b) Bertrand, M. P.; Coantic, S.; Feray,
L.; Nouguier, R.; Perfetti, P. Tetrahedron 2000, 56, 3951.
(47) (a) Miyabe, H.; Ushiro, C.; Ueda, M.; Yamakawa, K.; Naito,
T. J. Org. Chem. 2000, 65, 176. (b) Miyabe, H.; Konishi,
C.; Naito, T. Org. Lett. 2000, 2, 1443.
(48) (a) Yamada, K.-i.; Yamamoto, Y.; Maekawa, M.; Akindele,
T.; Umeki, H.; Tomioka, K. Org. Lett. 2006, 8, 87.
(b) Yamada, K.-i.; Nakano, Y.; Maekawa, M.; Akindele, T.;
Tomioka, K. Org. Lett. 2008, 10, 3805.
(49) (a) Vleeschouwer, F. D.; Van Speybroeck, V.; Waroquier,
M.; Geerlings, P.; De Proft, F. Org. Lett. 2007, 9, 2721.
(b) Fischer, H.; Radom, L. Angew. Chem. Int. Ed. 2001, 40,
1340.
(50) Similar stereoselectivities have been reported for other
related iodine-atom transfer radical cascades, see for
example: Miyabe, H.; Asada, R.; Toyoda, A.; Takemoto, Y.
Angew. Chem. Int. Ed. 2006, 45, 5863.
(25) Selected examples: (a) Bertrand, M. T.; Courtois, G.;
Migniniac, L. Tetrahedron Lett. 1974, 15, 3147.
(b) Nakamura, M.; Fujimoto, T.; Endo, K.; Nakamura, E.
Org. Lett. 2004, 6, 4837. (c) Cantagrel, F.; Pinet, S.;
Gimbert, Y.; Chavant, P. Y. Eur. J. Org. Chem. 2005, 2694.
(26) See the supporting information in ref. 24.
(27) Deprotonation of 1a responsible for the formation of 10
could also be effected by Bu2Zn, even though deprotonation
of terminal alkynes by dialkylzincs in the absence of ligands
is not a fast process, see: (a) Okhlobystin, O. Y.; Zakharkin,
L. I. J. Organomet. Chem. 1965, 3, 257. (b) De Koning, A.
J.; Van Rijn, P. E.; Boersma, J.; Van der Kerk, G. J. M.
J. Organomet. Chem. 1979, 174, 129. (c) Pinet, S.; Pandya,
S. U.; Chavant, P. Y.; Ayling, A.; Vallée, Y. Org. Lett. 2002,
4, 1463.
(28) Cote, A.; Charette, A. B. J. Am. Chem. Soc. 2008, 130, 2771.
(29) Haaland, A.; Green, J. C.; McGrady, G. S.; Downs, A. J.;
Gullo, E.; Lyall, M. J.; Timberlake, J.; Tutukin, A. V.;
Volden, H. V.; Ostby, K.-A. Dalton Trans. 2003, 4356.
(30) Bamford, C. H.; Newitt, D. M. J. Chem. Soc., Abstr. 1946,
688.
(51) See, (a) Chinkov, N.; Tene, D.; Marek, I. In Metal-
Catalyzed Cross Coupling Reactions, 2nd ed.; Diederich,
D.; de Meijere, A., Eds.; Wiley-VCH: New York, 2004,
395. (b) Denès, F.; Pérez-Luna, A.; Chemla, F. Chem. Rev.
2010, 110, 2366.
Synthesis 2011, No. 9, 1347–1360 © Thieme Stuttgart · New York