Radical-polar crossover domino reactions involving organozinc reagents and β-(allyloxy)-enoates

Article abstract of DOI:10.1016/j.tetlet.2008.04.108

Organozinc reagents (organozinc halides, diorganozincs and mixed copper-zinc reagents) react with β-(allyloxy)-enoates via a radical-polar crossover process to afford substituted furans in one single synthetic step following a domino reaction involving Michael addition and carbocyclisation. Reversal of diastereoselectivity can be obtained varying the organometallic and/or the reaction conditions.

Full text of DOI:10.1016/j.tetlet.2008.04.108

Tetrahedron Letters 49 (2008) 3963–3966
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: http://www.elsevier.com/locate/tetlet
Radical–polar crossover domino reactions involving organozinc
reagents and b-(allyloxy)-enoates
*
*
Steven Giboulot, Alejandro Pérez-Luna , Candice Botuha, Franck Ferreira, Fabrice Chemla
UPMC Univ Paris 06, UMR 7611, Laboratoire de Chimie Organique, Institut de Chimie Moléculaire (FR 2769), Case 183, 4 Place Jussieu, F-75005 Paris, France
CNRS, UMR 7611, Laboratoire de Chimie Organique, Institut de Chimie Moléculaire (FR 2769), Case 183, 4 Place Jussieu, F-75005 Paris, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Organozinc reagents (organozinc halides, diorganozincs and mixed copper–zinc reagents) react with
b-(allyloxy)-enoates via a radical–polar crossover process to afford substituted furans in one single
synthetic step following a domino reaction involving Michael addition and carbocyclisation. Reversal
of diastereoselectivity can be obtained varying the organometallic and/or the reaction conditions.
Ó 2008 Elsevier Ltd. All rights reserved.
Received 26 March 2008
Revised 17 April 2008
Accepted 18 April 2008
Available online 23 April 2008
Keywords:
Domino reactions
Radical–polar crossover
Organozinc reagents
Radicals
Furans
Organozinc chemistry has witnessed over the last decades a
remarkable regain of interest from synthetic chemists.1 Being less
reactive than the corresponding organolithium and Grignard
reagents it is possible to prepare organozinc reagents bearing func-
tional groups such as esters or nitriles. However, this reduced reac-
tivity also implies the use of metallic or organic catalysts or
promoters to provoke reactions with electrophiles.2,3 Thus, the
mastering of these activating methods remains essential to take
full advantage of the potential of such unique functionalised orga-
nometallic reagents. Of particular interest is the use of organozinc
reagents as radical precursors following oxidation or reaction with
carbon or heteroatom-centred radicals. Known since their dis-
covery,4 this possibility has only been started to be exploited for
synthetic purposes rather recently.5
likely lead to b-elimination, should also be suitable for substrates
bearing an oxygen in b position.12b,d Reaction with b-(allyloxy)-
enoates, readily available by reaction of the corresponding allylic
alcohol with methyl 2-(bromomethyl)acrylate, should thus
offer a new multi-component approach to substituted furans
(Scheme 1).
We initiated our work with a model study on substrate 113 de-
rived from allyl alcohol (Table 1). We were very glad to see that
reaction with nBu2Zn under our previously reported conditions
(Et2O, 16 h, room temperature)8 led, after acidic quench, to furan
2 in 64% yield and 75:25 (cis:trans) diastereoselectivity (entry 1).
Using the dialkylzinc reagent obtained by mixing ZnBr2 and nBuLi
(written as nBu2Zn–2LiBr) resulted in a slight increase in selectiv-
ity, albeit with a loss of yield (entry 2).
As part of our ongoing interest in carbometallation of unactivat-
ed alkenes by zinc enolates,6 we have recently shown that dialkyl-
zincs (Bu2Zn),7,8 organozinc reagents (RZnX)8 and mixed copper–
zinc reagents (R(CuCN)ZnX)9,10 react with (N-allyl)-aminoenoates
to give substituted (pyrrolidylmethyl)zinc derivatives through a
domino 1,4-addition/carbocyclisation sequence involving the
transformation of a simple readily available organozinc into a more
elaborated one via a radical chain transfer mechanism (vide in-
fra).11,12 We reasoned that this radical–polar crossover transforma-
tion, unlike reactions involving metal enolates that would most
Other organometallic reagents were also surveyed. Organozinc
(nBuZnX) reagents also perform the 1,4-addition/cyclisation reac-
tion (Table 1, entries 3–6). Without additives, nBuZnBr–LiBr
obtained by transmetallation of nBuLi with ZnBr2 gave the
corresponding furan 2-cis in good yield (65%) but quite low selec-
tivity (cis:trans = 58:42). In presence of two extra equivalents of
ZnY
CH2R
E
CH2R
RZnY
E+
X
X
X
CO2Me
CO2Me
CO2Me
RZnY :
R2Zn, RZnX,
RCuCNZnX
X = NR' : refs 7-10
X = O : this paper
* Corresponding authors. Tel.: +33 (1) 44 27 64 36; fax: +33 (1) 44 27 75 67 (F.C.).
E-mail addresses: alejandro.perez_luna@upmc.fr (A. Pérez-Luna), fabrice.
chemla@upmc.fr (F. Chemla).
Scheme 1.
0040-4039/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2008.04.108

3964
S. Giboulot et al. / Tetrahedron Letters 49 (2008) 3963–3966
Table 1
Table 2
Domino reaction between alkylzincs and b-(allyloxy)-enoate 1
Domino reaction between alkylzincs and b-(allyloxy)-enoate 5
Ph
Ph
(i) BuM
O
CO2Me +
Bu
Bu
CO2Me
O
O
(i) BuM
Ph
O
(ii) H2O
O
Bu
+
(ii) H3O+
CO2Me
OH
CO2Me
1
2-cis
2-trans
CO2Me
5
6a:6b:6c:6d
7
Entry
nBuM (equiv)
Additive (equiv)
dra (cis:trans)
2b (%)
Entry
nBuM (equiv)
Additive (equiv)
6 (a:b:c:d)a,b (%)
7c (%)
1
2
3
4
5
6
7
8
9
ZnBu2 (2)
—
—
—
75:25
80:20
58:42
19:81
38:62
62:38
50:50
31:69
22:78
64
41
65
48
45
62
48
28
10
ZnBu2–2LiBrc (2.5)
BuZnBr–LiBrc (2.4)
BuZnBr–LiBrc (2.3)
BuZnBr–LiClc (2.2)
BuZnBr (2.7)
1
2
3
ZnBu2 (2)
BuZnBr–LiBr (2)
ZnBu2–2LiBr (2)
—
68 (62:14:17:7)
51 (66:34:0:0)
0
0
0
85
ZnBr2 (2)
ZnBr2 (2.3)
ZnCl2 (2.2)
ZnBr2 (2)
ZnBr2 (0.5)
ZnBr2 (1)
ZnBr2 (3)
Conditions: (i) BuM (equiv), Et2O, rt, 16 h; (ii) HCl (1 M).
a
Bu(CuCN)ZnBr–LiBrd (1)
Bu(CuCN)ZnBr–LiBrd (1)
Bu(CuCN)ZnBr–LiBrd (1)
Determined by NMR analysis of the crude material.
Combined yield of isolated diastereomers after chromatography.
Isolated yield.
b
c
Conditions: (i) BuM (equiv), Et2O, rt, 16 h; (ii) HCl (1 M) or (ii) NH4Cl (2)/NH3 (1) for
entries 7–9.
a
Determined by NMR analysis of the crude material.
Combined yield of isolated diastereomers after chromatography.
Prepared from salt-free nBuLi and ZnX2.
Prepared from salt-free nBuLi, CuCN and ZnBr2.
9 %
3 %
b
H
c
9 %
H
H
d
Ph
H
H
CH3
12 %
CH3
O
H
O
H
CO2Me
Bu
H
H
H CO2Me
ZnBr2, reversal of selectivity was observed and 2-trans was ob-
tained with reasonable selectivity (cis:trans = 19:81) but moderate
yield. The use of ZnCl2 as zinc salt also led to 2-trans but with low-
er diastereoselectivity. Salt free nBuZnBr was prepared by equili-
bration between ZnBu2 and ZnBr2 via the Schlenk equilibrium.
Reaction with 1 led quite unexpectedly to 2-cis furan even with ex-
cess ZnBr2 (entry 6), in a behaviour that matches rather the dialkyl-
zinc case and contrasts with our observations with aminoenoates.8
Copper–zinc mixed reagents were also found to react with 1 to
afford 2 but with very disappointing results (Table 1, entries 7–9).
As for BuZnX, addition of ZnBr2 was necessary to obtain some lev-
els of diastereoselectivity in favour of the trans diastereomer but it
6a
6b
Scheme 3.
2LiBr only resulted in b-elimination. Unlike the non-substituted
case, the same major isomer was observed with the dialkylzinc
and the organozinc reagents. The structure of the two diastereo-
mers obtained using BuZnBr was determined by NOE experiments
(Scheme 3).
The observed transformations are believed to follow the radi-
cal–polar mechanism that we evidenced for the related reactions
involving (N-allyl)-aminoenoates (Scheme 4, path a).8 Initial oxida-
tion of the organometallic species by traces of oxygen produces a
radical that undergoes 1,4-addition onto the Michael acceptor to
afford an enoxy radical 8. Subsequent 5-exo-trig cyclisation and
reduction of the alkyl radical 9 by the organometallic species give
the (tetrahydofuranylmethyl)zinc intermediate and a new radical
that propagates the chain.
The formation of allylic alcohol 7 from 5 in the presence of
formed LiBr in contrast with the salt free reaction is unexpected,
as it arises most likely from anionic b-elimination of zinc enolate
10. Formation of 10 cannot arise from the polar conjugate addition
of ZnBu2 alone,8 but could possibly be the result of a polar conju-
gate addition of a putative ate-complex such as ZnBu2BrꢀLi+
formed from ZnBr2 and LiBr (Scheme 4, path b). Alternatively, on
the basis of a radical mechanism, it could also arise from homolytic
substitution at zinc of radical 8 (Scheme 4, path a0). Even though
tertiary alkyl-substituted a-alkoxycarbonyl radicals are believed
not to undergo homolytic substitution at zinc,15 the increased
SH2 rate might again be related to the formation of an ate-complex
such as ZnBu2BrꢀLi+, as observed for the reduction of similar a-alk-
oxycarbonyl radicals by alkylmercury derivatives,16 or it can also
result from an increase of the electrophilicity of radical 8 by coor-
dination of the Lewis-acid to the carbonyl moiety.17 Additionally, it
has also been suggested that the ease for a-alkoxycarbonyl radicals
to undergo SH2 at zinc is dependent on the possibility to form
chelated enolates.5
also resulted in
degradation.
a dramatic loss of yield due to extensive
The presence of the intermediate (furanylmethyl)zinc resulting
from the domino sequence was confirmed by an iodine quench fol-
lowing the reaction of 1 with nBu2Zn which afforded iodides 3-cis
and 3-trans in a 75:25 ratio (Scheme 2). The structure of the major
diastereomer was determined after refluxing the mixture of insep-
arable isomers in toluene. The major 3-cis compound afforded
quantitatively bicyclic lactone 4 while the minor 3-trans one
remained unchanged.14
The domino reaction was next investigated with substrate 5
bearing a Ph substituent in allylic position (Table 2). nBu2Zn affor-
ded furan 6 in good yield (68%) albeit as a mixture of 4 diastereo-
mers (62:14:17:7). The stereoselectivity was enhanced using
nBuZnBr–LiBr with excess ZnBr2, as only two diastereomers
(66:34:0:0) were obtained. By contrast, salt containing nBu2Zn–
(i) Bu2Zn
I
Bu
I
CO2Me
Bu
O
O
+
O
(ii) I2
88 %
CO2Me
CO2Me
1
3-cis (75)
3-trans (25)
I
Toluene
quantitative
O
O +
O
Bu
CO2Me
3-trans
Whatever its origin, this observation sheds some light on other
salt effects observed in reactions with substrate 1 as indeed addi-
tion of Lewis acids generally results in significant yield decreases
(Table 1, entry 1 vs 2, entry 3 vs 4,5 and entry 7 vs 8,9). Even
though neither of the two expected b-elimination products (11
O
Bu
4
Scheme 2. Reagents and conditions: (i) Bu2Zn (3 equiv), Et2O, rt, 16 h; (ii) I2, THF, rt,
3 h.

S. Giboulot et al. / Tetrahedron Letters 49 (2008) 3963–3966
3965
RM
O2
R
R'
R'
R'
RM
R
M
CH2R
CO2Me
path a
R'
O
O
O
R
R
SH
O
5-exo-trig
2
radical
addition
CO2Me
CO2Me
9
8
CO2Me
R'
RM
R
1, 5
RM
SH
path a'
O
oligomerisation
2
n
path c
polar
addition
path b
CO2Me
MeO2C
R'
M
R
CO2Me
O
R'
CO2Me
R
+
n
R
O
anionic
OM
CO2Me
R'
R'
O
β-elimination
11
12
10
13
Scheme 4. Mechanism for the radical–polar crossover domino reaction.
(R = Bu) and 12 (R0 = H)) were detected (allyl alcohol obtained from
12 is too volatile and 11 presumably polymerises in the reaction
media), we strongly suspect the formation of zinc enolate 10 lead-
ing to fragmentation to be the reason for the observed yield loss.
Concerning the diastereoselectivity, the observed results can be
understood using the model we proposed previously for reactions
with aminoenoates.8 The diastereoselectivity of the domino pro-
cess results from the diastereoselectivity of the 5-exo-cyclisation
(Scheme 5). The cis isomer results from cyclisation via I where, fol-
lowing the Beckwith–Houk model,18,19 for A1,3 strain20 and intra-
molecular dipole–dipole effect minimisation,21 the carbomethoxy
moiety adopts à pseudo-equatorial position. This is the case for
dialkylzincs and, unlike for aminoenoates and to a lesser extent,
also for organozinc and copper–zinc mixed reagents. Addition of
extra Lewis acids (ZnX2) makes the cyclisation occur preferentially
via II where chelation between the oxygen atom, the carbonyl
group and the metal salt counterbalances A1,3-strain minimisation,
thus leading to the trans isomer. However, the impact of LiBr on
CO2Me
MeO2C
R
A)
Ph
B)
O
n
CO2Me
11
CO2Me
1
O
O
13
Scheme 6. Reagents and conditions: (A) (i) PhZn(CuCN)Br–LiBr, Et2O, 16 h, rt; (ii)
NH4Cl (2)/NH3 (1); (B) (i) PhZnBr–LiBr, Et2O, 16 h, rt; (ii) HCl (1 M).
With the former, only oligomeric material 13, spectroscopically
(
13C NMR) identical to the previously reported radical polymerisa-
tion product of 1,13 was isolated, evidencing that alkyl radical 9
does not undergo homolytic substitution at zinc with the aryl zinc
compound (Scheme 4, path c). With the latter, only product 11
(R = Ph) was obtained.
In conclusion, we have shown that the radical–polar domino
1,4-addition/carbocyclisation process that we described with (N-
allyl)-aminoenoates can be extended to b-(allyloxy)-enoates thus
providing polysubstituted furans. The general features of the se-
quence remain the same in both cases: the use of dialkylzincs leads
to 2,3-cis furans, while the use of organozinc and copper–zinc
mixed reagents (though the latter in much lower yields) leads to
2,3-trans furans. Nevertheless, some limitations have been evi-
denced. In first place, competitive 1,4-addition/fragmentation of-
ten hampers the efficiency of the reaction, especially in the
presence of Lewis acids. In particular, it precludes the use of cop-
per–zinc mixed reagents. Further studies are currently underway
to determine whether the formation of the intermediate zinc eno-
late leading to b-elimination results from an initial polar addition
or from reduction of the intermediate enoxy radical as a conse-
quence of a less favourable 5-exo-cyclisation step than for the
nitrogen containing substrates.22,23 The latter would imply that
tertiary alkylsubstituted a-alkoxycarbonyl radicals can undergo
homolytic substitution at zinc, a transformation that is, to the best
of our knowledge, unprecedented. In second place, for reactions
leading to the trans diastereomers, both as a result of its lower Le-
wis basicity and for geometrical reasons, chelation of zinc salts by
the oxygen atom and the carbonyl group is less favourable, giving
lower levels of selectivity.
the diastereoselectivity (Table 1, entry
unexplained.
4 vs 6) remains
For substrate 5 bearing an allylic substituent, cyclisation start-
ing from Bu2Zn occurs mainly via the non-chelated intermediate
III where the phenyl group adopts a pseudo-equatorial position.
It is also the case using BuZnBr–LiBr despite the presence of added
ZnBr2, presumably because steric hindrance prevents coordination
to the allylic oxygen atom, hence chelation of the zinc salt.
Finally, we explored the possibility to use arylzinc reagents
(Scheme 6). Neither PhZnBr–LiBr nor PhZn(CuCN)Br–LiBr gave
the domino addition/cyclisation products after reaction with 1.
O
O
2-cis
O
OMe
CO2Me
Bu
Bu
I
9-cis (R' = H)
O
X2Zn
Bu
O
2-trans
CO2Me
Bu
O
OMe
II
9-trans
Ph
Acknowledgement
O
Ph
O
The authors thank E. Caytan for help with NOE experiments.
References and notes
6a
O
OMe
Bu
III
CO2Me
Bu
9-cis (R' = Ph)
1. The Chemistry of Organozinc Compounds; Rappoport, Z., Marek, I., Eds.; Wiley:
Scheme 5.
Chichester, 2007.

3966
S. Giboulot et al. / Tetrahedron Letters 49 (2008) 3963–3966
2. Knochel, P.; Millot, N.; Rodrigues, A. L. Org. React. 2001, 58, 417–731.
3. Knochel, P.; Calaza, M. I.; Hupe, E. In Metal-Catalyzed Cross-Coupling Reactions;
Meijere, de A., Diederich, F., Eds., 2nd ed.; Wiley-VCH: Weinheim, 2004; pp
619–670.
4. For a historical account of alkylzinc chemistry, see: Seyferth, D. Organometallics
2001, 20, 2940–2955.
5. For a review of the use of dialkylzinc in radical reactions see: Bazin, S.; Feray,
L.; Bertrand, M. P. Chimia 2006, 60, 260–265.
6. Pérez-Luna, A.; Botuha, C.; Ferreira, F.; Chemla, F. New J. Chem. 2008, 32, 594–
606.
12. For selected related preparations of (cyclopentylmethyl)zinc or
(tetrahydrofuranylmethyl)zinc derivatives involving a radical–polar crossover
see: (a) Stadtmüller, H.; Lentz, R.; Tucker, C.; Stüdemann, T.; Dörner, W.;
Knochel, P. J. Am. Chem. Soc. 1993, 115, 7027–7028; (b) Vaupel, A.; Knochel, P.
Tetrahedron Lett. 1994, 35, 8349–8352; (c) Vaupel, A.; Knochel, P. J. Org. Chem.
1996, 61, 5743–5753; (d) Riguet, E.; Klement, I.; Kishan Reddy, Ch.; Cahiez, G.;
Knochel, P. Tetrahedron Lett. 1996, 37, 5865–5868; (e) Cohen, T.; Gibney, H.;
Ivanov, R.; Yeh, E. A.-H.; Marek, I.; Curran, D. P. J. Am. Chem. Soc. 2007, 129,
15405–15409.
13. For preparation and radical polymerisation of 1: Thompson, R. D.; Jarrett, W. L.;
Mathias, L. J. Macromolecules 1992, 25, 6455–6459.
14. Curran, D. P.; Chang, C.-T. J. Org. Chem. 1989, 54, 3140–3157.
15. For a detailed discussion see Ref. 11g and references cited therein.
16. Russell, G. A.; Shi, B. Z.; Jiang, W.; Hu, S.; Kim, B. H.; Baik, W. J. Am. Chem. Soc.
1995, 117, 3952–3962.
7. Denes, F.; Chemla, F.; Normant, J.-F. Angew. Chem., Int. Ed. 2003, 42, 4043–
4046.
8. Denes, F.; Pérez-Luna, A.; Cutri, S.; Chemla, F. Chem. Eur. J. 2006, 12, 6506–6513.
9. Denes, F.; Chemla, F.; Normant, J.-F. Eur. J. Org. Chem. 2002, 3536–3542.
10. Denes, F.; Pérez-Luna, A.; Chemla, F. J. Org. Chem. 2007, 72, 398–406.
11. For other selected radical 1,4-additions of dialkylzincs see: (a) Bertrand, M. P.;
Feray, L.; Nouguier, R.; Perfetti, P. J. Org. Chem. 1999, 64, 9189–9193; (b) Van
der Deen, H.; Kellog, R. M.; Feringa, B. L. Org. Lett. 2000, 2, 1593–1595; (c)
Bazin, S.; Feray, L.; Siri, D.; Naubron, J.-V.; Bertrand, M. P. Chem. Commun. 2002,
2506–2507; (d) Miyabe, H.; Asada, R.; Yoshida, K.; Takemoto, Y. Synlett 2004,
540–542; (e) Miyabe, H.; Asada, R.; Takemoto, Y. Tetrahedron 2005, 61, 385–
393; (f) Bazin, S.; Feray, L.; Vanthuyne, N.; Bertrand, M. P. Tetrahedron 2005, 61,
4261–4274; (g) Bazin, S.; Feray, L.; Vanthuyne, N.; Siri, D.; Bertrand, M. P.
Tetrahedron 2007, 63, 77–85.
17. Guindon, Y.; Rancourt, J. J. Org. Chem. 1998, 63, 6554–6565.
18. Curran, D. P. In Comprehensive Organic Chemistry; Trost, B. M., Flemming, I.,
Eds.; Pergamon Press, 1991; Vol. 4, pp 779–831.
19. Rajanbabu, T. V. Acc. Chem. Res. 1991, 24, 139–145.
20. Bulliard, M.; Giese, B.; Zeitz, H. G. Synlett 1991, 425–427.
21. Guindon, Y.; Yoakim, C.; Lemieux, R.; Boisvert, L.; Delorme, D.; Lavallée, J.-F.
Tetrahedron Lett. 1990, 31, 2845–2848.
22. Beckwith, A. L. J. Tetrahedron 1981, 37, 3073–3100.
23. Della, E. W.; Knill, A. M. Aust. J. Chem. 1995, 48, 2047–2051.