A case of anti carbolithiation of alkynes resulting from intramolecular lithium coordination

Article abstract of DOI:10.1002/chem.200701834

The mechanism of the intramolecular carbolithiation of lithiated propargylic ether 2 has been investigated both experimentally and theoretically. The results show that the action of one equivalent of n-butyllithium on 1 is sufficient to trigger halogen-li

Full text of DOI:10.1002/chem.200701834

FULL PAPER
DOI: 10.1002/chem.200701834
A Case of anti Carbolithiation of Alkynes Resulting from Intramolecular
Lithium Coordination
Catherine FressignØ,*[a, b] Anne-Lise Girard,[a] Muriel Durandetti,[a] and
Jacques Maddaluno*[a]
Abstract: The mechanism of the intra-
molecular carbolithiation of lithiated
propargylic ether 2 has been investigat-
ed both experimentally and theoretical-
ly. The results show that the action of
one equivalent of n-butyllithium on 1 is
sufficient to trigger halogen–lithium ex-
change and the subsequent heterocycli-
zation step. Interestingly, the reaction
stops at the stage of dihydrobenzofuran
6; no spontaneous elimination of lithi-
um ethylate was observed. The fact
that the E configuration of this adduct
was exclusively produced suggests that
the reaction proceeds by following an
unprecedented anti addition on the
alkyne. According to DFT calculations,
this unexpected outcome is related to
the intramolecular coordination of the
lithium by one oxygen atom of the ter-
minal acetal appendage: the O–Li in-
teraction, which persists all along the
ring-closure process, drives the cation
to the E site of the final olefin. The cal-
culations also show that in the absence
of this coordination (as in conformers
B and C of acetal 2), the Z olefin that
results from a classical syn addition of
the aryllithium should be obtained.
The experiments were repeated with
allene 1d. In this case, one equivalent
of n-butyllithium suffices to trigger not
only the exchange and the cyclization,
but also the final elimination of lithium
ethoxide. The DFT results indicate that
the intramolecular addition of the orig-
inal aryllithium on the central carbon
atom of the allene 2b yields the ex-
pected benzofuran skeleton 3b, which
bears a lithiated lateral chain at the 3-
position. Both cyclizations go through
low-lying transition states, as is expect-
ed for rapid reactions at low tempera-
ture.
Keywords: alkynes
·
carbolithia-
tion · cyclization · density functional
calculations · stereoselectivity
Introduction
and with good-to-total E control of the lateral double
bond.[1] It was proposed in the case of 1 that this reaction
relies on an anionic cascade that begins with a 5-exo-dig in-
tramolecular nucleophilic addition of aryllithium 2 on the
triple bond. The vinyllithium 3, which is expected to result
from a syn addition,[2] would then undergo a b-elimination
of lithium ethylate to provide the exocyclic allene 4. Finally,
this latter product can isomerize into the 3-vinylbenzofurane
5 (Scheme 1). This rapid reaction prohibits the characteriza-
tion of the intermediates and it explains the speculative
character of the proposed mechanism. The details of the
cyclization process for allene 2b were not discussed. An in-
tramolecular addition of the aryllithium on the central
carbon of the allenic appendage, directly followed by an
elimination of lithium ethoxide was expected (Scheme 1).
We thus decided to undertake an experimental and theo-
retical study to separately detail the cyclization and elimina-
tion steps, in view of the determination of the structure of
the different intermediates, the activation energies, and the
origin of the E control over the double bond formation.
Beyond the particular case of this heterocyclization, the car-
A previous experimental study showed that the halogen–
lithium exchange that takes place at low temperature on
propargylic acetal 1, or its allenyl isomer 1b, triggers a het-
erocyclization and yields 3-vinylbenzofurane 5 in good yield
[a] Dr. C. FressignØ, Dr. A.-L. Girard, Dr. M. Durandetti,
Dr. J. Maddaluno
Laboratoire des Fonctions AzotØes &
OxygØnØes Complexes de l’IRCOF
UMR CNRS 6014, UniversitØ et INSA de Rouen
76821Mont St Aignan Cedex (France)
Fax : (+33)235-522-971
E-mail: cfressig@crihan.fr
jmaddalu@crihan.fr
[b] Dr. C. FressignØ
Laboratoire de Chimie ThØorique
UMR 7616 CNRS, UniversitØ P. & M. Curie
Case Courrier 137, 4, place Jussieu
75252 Paris Cedex 05 (France)
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
Chem. Eur. J. 2008, 14, 5159 – 5167
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5159

Our own study was first con-
ducted on the aryllithium 2,
which was unsolvated and then
solvated by two discrete solvent
molecules (diethyl ether or
THF). The case of solvated al-
lenyllithium 2b was considered
separately. In this first paper,
we focus on the details of the
carbolithiation step, the elimi-
nation is fully described in a
paper that is currently in prepa-
ration.
Scheme 1. Putative mechanisms for the anionic heterocyclization of propargylic 1 or allenic 1b acetal in benzo-
furan 5.
Computational details: The size
bometallation reaction is the object of a sustained interest,
and important papers and reviews that describe its applica-
tions in synthesis have been published recently.[3]
Let us first discuss the case of alkyne 1. Intramolecular
carbolithiations of triple bonds that are associated with b-
eliminations have been observed in other situations. In par-
ticular, Bailey and Aspris have reported a comparable phe-
nomenon with propargylic ethers (Scheme 2).[4] In this case,
of the systems that are considered led us to restrict our com-
putations to the DFT level. The B3P86 functional and 6–
31G** basis set, which behave satisfactorily in related situa-
tions[10] were retained. All the calculations were performed
by using the Jaguar 5.0 software.[11] The energy given for the
minima of the solvated systems include the zero-point
energy corrections (ZPE). The transition states were ob-
tained by relaxed potential energy surface (PES) scans, then
fully optimized and characterized by frequency calculations.
Our previous results in the field[10a] suggest that the basis set
superposition error (BSSE) can be ignored in these cases.
The electron localization function (ELF) relies on a topolog-
ical approach of the chemical bond that is described in origi-
nal articles by Savin, Silvi et al.[12] We[13] and others[14] have
shown in previous works that this tool can help to determine
the bonding scheme in systems that present ill-defined va-
lences, such as noncovalent organolithium aggregates.
Therefore, we thought that such an electron distribution,
which implicitly takes into account the superposition of the
resonance forms could provide useful information on the
electron reorganization that is induced by the rearrange-
ment of the intermediates. In the ELF figures presented
hereafter, the color code characterizes the cores (magenta),
the monosynaptic (red) and disynaptic (green) valence
basins, and the hydrogens (blue).
Scheme 2. Anionic heterocyclization of propargylic ethers in exocyclic al-
lenes.[4] KHMDS=potassium hexamethyl disilazide.
however, no migration of the double bond was observed,
and exocyclic allenes were efficiently recovered. This intri-
guing difference in behavior, which might be related to the
mobility of the a proton, to the syn/anti character of the
lithium methoxide elimination, or to the recovery of the aro-
maticity upon rearrangement was also worth investigating.
The analysis of the cascade that is reported in Scheme 1
has been restricted to the cyclization, elimination, and
proton-transfer steps. The theory underlying the halogen–
lithium exchange process has been extensively examined re-
cently,[5] and is not considered in this paper. In contrast, the-
oretical studies that are devoted to the carbometallation re-
action are relatively scarce. Two early ab initio Hartree–
Fock (HF) studies that were published by Nakamura and
colleagues underlined the importance of the metal–multiple
bond interaction in the course of the reaction between
methyllithium and cyclopropene[6] or acetylene.[7] In parallel,
papers by Bailey and co-workers described the key inter-
mediates along the intramolecular cyclization of 5-hexen-1-
yllithium (at the HF level)[8] and of a styrene derivative (at
the DFT-B3LYP level).[9] In both cases, a double bond–lithi-
um interaction is shown to play an important role in the or-
ganization of the transition state, at least in the absence of
explicit solvent molecules.
Results and Discussion
Experimental investigation: We first checked experimentally
if a single equivalent of n-butyllithium could prompt the
heterocyclization and give access to the putative vinyllithi-
um intermediate. The original experiments in THF had
clearly shown that a significant excess of n-butyllithium
(3.3 equiv) was required to obtain benzofuran 5 in decent
yields. With respect to the previous procedure, the iodoaryl
1 was conveniently replaced by the more stable bromoaryl
1c. This latter compound was treated with exactly one
equivalent of n-butyllithium at ꢀ788C in THF (Scheme 3).
To our delight, quenching the medium with water led to the
ethylidenic dihydrobenzofuran
6
in 88% yield.[15] The
1H NMR spectrum of this compound, when recorded in
5160
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 5159 – 5167

Carbolithiation of Alkynes
FULL PAPER
however, the product was con-
taminated by several unidenti-
fied side products, even when
quenching the reaction after a
few minutes at ꢀ508C.
The fact that 6 was recovered
Scheme 3. Cyclization of propargylic ether 1c under the influence of nBuLi. Conditions A: i) 1.0 equiv nBuLi/
THF, ꢀ788C; ii) H2O or EtOD. Conditions B: i) 1.0 equiv nBuLi/Et2O, ꢀ788C; ii) EtOD.
as the geometrically pure
E
isomer remained very surprising
because the acetal group, in
contrast to a silyl or phenyl
CDCl3, was relatively puzzling. The fortuitous superposition
of the vinylic signals and that of the acetal explains that 6
could easily be mistaken for aromatic 7, and its unambigu-
ous identification required the recording of the spectrum in
[D6]benzene. NOE experiments that were run in this solvent
showed that 6 was obtained as the E isomer exclusively,
which is a result that was completely unexpected for the car-
bolithiation of an alkyne.[16] Isolating 6 also demonstrated
that the intermediate 3 (E, Scheme 1) was not necessarily
undergoing a b-elimination step in THF. This contrasts with
our previous results that were obtained with lithioallenes[17]
and is in complete discrepancy with the mechanism that is
proposed in Scheme 1.
moiety,[21] can hardly be considered to be a stabilizing entity
that is capable of triggering the post-cyclization isomeriza-
tion of a double bond. Furthermore, the isomerizations that
have been reported in literature occurred at room tempera-
ture, whereas the experiment that is reported here was
maintained at ꢀ788C until the final quench. The origin of
this puzzling phenomenon, which suggests a possible anti
carbolithiation of the alkyne is difficult to determine on ex-
perimental bases. We thus decided to pursue the investiga-
tion by theoretical means and we were nicely rewarded.
Theoretical examination of the unsolvated propargylic
acetal 2: Because of the relatively large size of the system
and the long computational runs that are often associated
with the shallow potential surfaces of the organolithium spe-
cies, the study was first conducted without taking the solvent
into account, either implicitly by using a continuum tech-
nique[22] or explicitly by taking into account a discrete
number of solvent molecules to simulate the solvation first
shell.[23] The optimization of the starting aryllithium deriva-
tive was performed first. Two local minima were identified,
one of which was more stable by 12.8 kcalmolꢀ1 (Figure 1).
The interconversion pathway between these conformers was
not studied because of this large energy difference. In the
more-stable situation, the lithium cation is tricoordinated by
Attempts to treat 3 with various electrophiles (PhCHO,
MeI, PhCH2Br, TMSCl) have remained unsuccessful, where-
as quenching the medium with EtOD led to 34% deuterat-
ed 6. We thought that this mediocre reactivity could be due
to an early protonation of 3 by an unexpected source of pro-
tons in the medium. We thus tried to replace one equivalent
of n-butyllithium by two equivalents of tert-butyllithium to
avoid the in-situ formation of n-butylbromide;[18] however,
the chemical yield and the deuteration level remained prac-
tically unchanged. Repeating the experiment in freshly dis-
tilled [D8]THF instead of regular THF did not trigger a
deuteration; this excludes the participation of the solvent.
This low reactivity could be related to the formation of
bulky and relatively inert aggregates of 3 in THF. Vinyllithi-
um is known to form tetramers and dimers in an 8:1ratio at
ꢀ908C in this solvent,[19] and this compact organization is
most probably reinforced in the case of 3 by the intramolec-
ular coordination that is imposed by the oxygen atoms of
the acetal appendage. However, the literature suggests that
vinyllithiums that result from the carbolithiation of a triple
bond could be efficiently trapped by electrophiles.[20] We
thus tried to adapt our conditions to those that are described
in the corresponding papers, but disappointing results were
obtained when warming the reaction medium and when
adding one equivalent (with respect to n-butyllithium) of
hexamethylphosphoramide (HMPA) or lithium bromide to
the reagents (good yields, low deuteration). Repeating the
deuteration experiment in diethyl ether at ꢀ788C led to di-
hydrobenzofuran 6 in only 12% yield (which was not im-
proved by using two equivalents of n-butyllithium), but with
78% label incorporation (Scheme 3). The rest consisted of
ether 7, of which the labeling ratio was not determined.
Note that the latter yield could be enhanced by increasing
the temperature (up to ꢀ508C or even room temperature);
ꢀ
CAr, one of the oxygen atoms of the acetal moiety (d(Li
1
ꢁ
ꢀ
Oacetal)=2.0 Š), and by the C C bond (d
G
2
ꢀ
d
U
ium atom that is dicoordinated by CAr and the oxygen atom
of the ether. Note that the more stable arrangement (E=
ꢀ700.2021u.a.) also brings the nucleophilic carbon atom,
CAr to within a favorably short distance of the acetylenic
carbon atom (3.02 Š). When the distance between CAr and
C1, which is taken as the reaction coordinate, was shortened,
the energy increased, and the system reached a transition
state in which the five-membered heterocycle is preformed.
At this stage, the lithium cation is coordinated by the aro-
matic carbon, C1 and the original oxygen atom. The corre-
sponding activation energy was found to be 3.8 kcalmolꢀ1.
ꢀ
Pursuing the C C bond formation leads to the expected
benzofuranic structure that exhibits an exocyclic E double
1
2
ꢀ
bond (d(C C )=1.36 Š), in accord with the experimental
A
data. Interestingly, the lithium cation lies out of the aromat-
ic plane and remains coordinated to the acetal oxygen atom.
It was confirmed that this corresponded to a true local mini-
mum (no negative frequency) by performing the calculations
with Jaguar 5.0 as well as Gaussian 98.[24] Additionally, a
Chem. Eur. J. 2008, 14, 5159 – 5167
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5161

J. Maddaluno, C. FressignØ et al.
Figure 1. Energy profile for the cyclization of the unsolvated propargylic acetal 2. The localization domains of the ELF representation corresponds to
h(r)=0.75.
lone pair that was localized on C2 and oriented within the
cation is relatively variable, the tri- and tetracoordinations
are most frequently encountered. We privileged a tetracoor-
dination and chose to add two solvent molecules to the
system, the other ligands were CAr or C2, and either one of
the oxygen atoms of the acetal appendage, the oxygen atom
of the ether tether, or the triple bond. This led to two fami-
lies of conformers (folded/unfolded) in each solvent that
were independently reoptimized. The results that are pre-
sented in Table 1show that reasoning on the sole conforma-
tions cannot account for the observed results: the more
stable folded conformer is preferred over the stretched one
both in THF and diethyl ether. However, the energy gap be-
tween the conformers is five times smaller in diethyl ether,
which suggests that the hard-to-cyclize stretched arrange-
ment becomes somewhat competitive in this solvent. This is
likely to slow the reaction, as was observed experimentally.
ꢀ
aromatic plane anti to the newly formed C C bond could be
visualized by resorting to an ELF (see Computational De-
tails) analysis (Figure 1). The highly ionic character of the
[10c]
ꢀ
C Li bond in vinyllithium
explains this spectacular lack
of directionality.[25] Thus, it seems that the O–Li intramolec-
ular coordination keeps the cation out of the vinyl plane,
and is responsible for the final E configuration of the
double bond in 6Li.
Justifying such surprising results required pursuing our in-
vestigation on a model that is as realistic as possible. We
thus included the solvent in our calculations because: 1) no
transition state that links unsolvated 6Li to more advanced
intermediates along the elimination process could be located
and 2) the above results suggest that the lithium coordina-
tion can exert a strong influence on the course of the reac-
tion. To avoid possible artifacts, we preferred to incorporate
discrete solvent molecules (THF or diethyl ether).[1] In con-
sideration of the experimental results that show that the cyc-
lization is difficult in diethyl ether, we considered the solva-
tion by both THF and diethyl ether to better understand the
origin of the detrimental influence of the latter solvent. We
did not replace diethyl ether by the smaller dimethyl ether
to keep the full steric effect of the ethyl groups.[23a]
Conformers of propargylic acetal 2 in THF: On the basis of
the experimental data, THF was the sole solvent to be con-
sidered in the following. The full optimization of the solvat-
ed folded conformer of 2 indicated that three local minima
that correspond to the three staggered arrangements of the
acetal group lie within less than 3 kcalmolꢀ1 of each other
(Figure 2).
Interestingly, in the overall more stable C conformer that
is considered above, the lithium atom does not lie in the
phenyl plane but remains connected to the CAr atom. The
coordination sphere is completed by the triple bond and the
two THF molecules, and leaves the acetal oxygen atoms in
an “anti” arrangement. A similar chelation pattern is found
for conformer B. In these two situations, the lithium atom is
coordinated to the two oxygen atoms of the THF molecules,
and undergoes p-binding[27] with the triple bond. Thus, the
Solvated system: effect of the solvent on the orientation of
the lateral chain of 2: Our previous experience with the
computational treatment of solvated organolithium species
prompted us to incorporate explicit solvent molecules rather
than to apply the polarizable dielectric field that is imposed
by a continuum model.[23b,26] We first considered the influ-
ence of the two solvents on the two local minima of 2 that
are described above (Figure 1). The solvation of the lithium
5162
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 5159 – 5167

Carbolithiation of Alkynes
FULL PAPER
Table 1. The DFT-optimized minima of the two conformers of disolvated 2, their absolute (E), and ZPE-cor-
rected (Ecorr) energies. The relative energies are given in parenthesis.
energy surfaces (PES). The
highest barrier that was located
by this technique lies at about
4.4 kcalmolꢀ1 above the ground
state. This relatively low value
suggests that the acetal group
rotates almost freely around the
propargyl axis, and the inter-
conversion between the con-
formers cannot be regarded as
the limiting step of the reaction.
Therefore, the three conformers
had to be considered for the
rest of the work.
THF
energy
Diethyl ether
energy
conformer
conformer
E=ꢀ1166.6050 a.u.
E=ꢀ1169.0979 a.u.
ZPE=336.4 kcalmolꢀ1
corr =ꢀ1168.5618 a.u.
(1.1 kcalmolꢀ1
ZPE=310.5 kcalmolꢀ1
Ecorr =ꢀ1166.1102 a.u
E
(5.5 kcalmolꢀ1
)
)
unfolded
unfolded
E=ꢀ1169.0965 a.u.
E=ꢀ1166.6130 a.u.
ZPE=334.4 kcalmolꢀ1
ZPE=310.0 kcalmolꢀ1
Cyclization
of
propargylic
E
corr =ꢀ1168.5636 a.u.
E
corr =ꢀ1166.1190 a.u. (0.0 kcalmolꢀ1
)
acetal 2: The study of the cycli-
zation was undertaken accord-
ing to the approach that is de-
scribed above for conformers
A–C. When the CAr
length is decreased, the system
goes through a transition state
(0.0 kcalmolꢀ1
)
folded
folded
1
ꢀ
C
bond
in the three cases (checked by frequency calculations) in
which the coordination of the lithium atom is only slightly
altered: a strong Li–Oacetal coordination is observed in situa-
tion A, whereas a Li–C2 “bonding” appears in the transition
1
ꢀ
state of B and C (Figure 3). Comparing the CAr
C
bond
Figure 2. Three conformers of acetal 2 that are solvated by two THF,
their absolute (E), ZPE-corrected (Ecorr) and relative energies (in paren-
thesis).
Li–Oacetal distance is significantly increased with respect to
the nonsolvated situation by the introduction of the two
ꢀ
THF molecules (d(Li Oacetal)=3.19 and 3.84 Š in conformer
B and C, respectively). In contrast, for conformer A, the Li–
Figure 3. Transition states between conformers of acetal 2 that are solvat-
ed by two THF, and cyclized product 3, the absolute (E) and corrected
(Ecorr) energies, the activation barriers (E°), and the difference between
the activation barriers (dE°; in parentheses).
ꢀ
Oacetal coordination is unchanged (d(Li Oacetal)=2.16 Š) and
the lithium atom is surrounded by three oxygen atoms, and
1
ꢀ
CAr (d
(Li C )=2.82 Š cannot be considered to be a coordi-
nation anymore). The measurement of the dihedral angles
shows, however, that the lithium atom is not in a regular tet-
rahedral surrounding. The zero-point energy (ZPE) calcula-
tions that were run for the three conformers A–C led to rel-
atively small values; this reflects the somewhat rigid organi-
zation of the molecule. After ZPE correction, the relative
order of stability is unaltered, but the energy differences be-
tween A, B, and C decrease (Figure 2).
lengths in the three TSs suggests that a relatively early tran-
sition state is reached in A, whereas the transition state is
1
ꢀ
later in C and B (d
A
ly). The activation barriers (ca. 8 to 12 kcalmolꢀ1) are signif-
icantly higher than in the unsolvated situation (3.8 kcal
molꢀ1). The lowest one is associated with the less-stable con-
former A, whereas the barriers for B and C are of compara-
ble heights. This particularity confines the three transitions
states in a narrow energy range, and thus the three arrange-
ments are considered in the next step of the study. The ZPE
corrections are comparable for the three transition states,
and do not change the relative stability order.
Before considering the reaction profile, we examined the
energy barriers that separate the three conformers in an at-
tempt to identify the proper limiting step of the transforma-
2
ꢀ
tion. Thus, a rotation of the C Cacetal bond has been im-
posed on conformer A to scan the corresponding potential
Chem. Eur. J. 2008, 14, 5159 – 5167
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5163

J. Maddaluno, C. FressignØ et al.
ꢀ
Pursuing the potential energy scan to completion led to
three structures, which were then fully optimized. As ex-
pected, a benzofuran nucleus that featured an exocyclic
double bond was obtained in the three cases (Figure 4).
Note that A features an E double bond (in full agreement
with the experimental data, vide supra) whereas conformers
B and C exhibit a Z double bond. In A, the flexible interac-
tions between the lithium atom and C2 plus the Oacetal atoms
keep the cation in a tetracoordinated state and ꢂ308 outside
the plane of the heterocycle and of the double bond (CAr
1
2
1
2
ꢀ
C
ꢀ
ꢀ ꢀ
C
Li=2118 and C
C
Li=1488). Running a ELF
analysis shows that the electronic doublet, which is still con-
sidered to be a lone pair (red),[10c] is located between the C2
and the lithium atoms. In both conformers B and C, a simi-
larly atypical vinyllithium moiety is obtained: the metal re-
1
2
ꢀ
ꢀ ꢀ
mains out of the heterocyclic plane (CAr
C
C
Li=346
C Li=118 and 1128, and
1
2
ꢀ
ꢀ
and 3378, respectively), with C
the lithium atom is tricoordinated. The ELF analysis for B
and C confirms these points. Further, the three products lie
in a 6.3 kcalmolꢀ1 range, which is decreased to 2.7 kcalmolꢀ1
when the ZPE is taken into account. Interestingly, this cor-
rection changes the relative stability order: it renders the
two lowest conformers A and C quasi-isoenergetic. In the
three situations A, B, and C, the reaction is similarly exo-
thermic (in the 50 kcalmolꢀ1 range).
The data that are obtained by the theoretical treatment of
the cyclization can be summarized, see Figure 5. This graphi-
cal representation demonstrates that the three possibilities
remain competitive. However, as mentioned above, only
process A affords an E vinyllithium, which is in agreement
with the experimental observations that are presented
below.
The intriguing feature of the cyclization is the E selectivi-
ty that is associated with the cyclization of A. To get a
deeper insight into this process, snapshots of the system
during its evolution along the PES, which results from the
1
ꢀ
CAr C shortening are provided in Figure 6. Step I shows the
Figure 4. Optimized products 3, the absolute (E), ZPE-corrected (Ecorr
)
situation about 0.1Š after the TS. It indicates that the
and condensation energies (Econd =E°ꢀEcorr), as well as the energy differ-
ences between conformers/isomers (dE, in parenthesis). The localization
domains of the ELF representation (bottom) correspond to h(r)=0.75.
dCAr increases and the lithium “slides” more or less parallel
ꢀ
Li
to the phenyl plane. The resulting acetal coordination forces
Figure 5. Overall energy characteristics of the three cyclization routes (A left, B middle, C right).
5164
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 5159 – 5167

Carbolithiation of Alkynes
FULL PAPER
of THF. Interestingly, the allen-
yl lateral chain adopts mainly
one conformation (except for
meaningless local minima that
are due to rotations of the
acetal) in which the large dis-
tances between the lithium
atom and the oxygen atoms of
the acetal group forbid all Li–
ꢀ
OMe interactions (Li O=6.1
and 5.8 Š, Figure 7). When the
distance between CAr and C1,
which was taken as before as
the reaction coordinate to scan
the PES was shortened, the
Figure 6. Four snapshots along the PES of the A cyclization process.
system reached
a
transition
1
ꢀ
state for CAr C =3.0 Š. At this
the double bond to adopt an E configuration, and the lone
stage, the allene undergoes a twisting of the double bonds
pair is already localized on the C2 atom, whereas a valence
(the three sp2 carbon atoms are misaligned, C
C C =
2
1
3
ꢀ
ꢀ
1
basin appears between CAr and C1. In step II, the CAr
1678). The corresponding activation energy was found to be
2.5 kcalmolꢀ1, compared to the 8–12 kcalmolꢀ1 that was cal-
culated for the cyclization of 2 (Figure 6). This low barrier
can explain the good yields that were obtained experimen-
ꢀ
C
bond is reinforced and the lithium atom continues its move-
ment parallel to the newly formed benzofuran nucleus. This
goes on during step III, the acetal group pivots to accompa-
ny the lithium atom. At step IV, an intermediate that is
ꢀ
tally. Pursuing the C C bond shortening leads to the expect-
close to the cyclization product is obtained in which the E
ed benzofuran 3b, which bears a lithiated lateral chain at
the 3-position (Figure 7). This ring closing is an exothermic
process (ꢀ48.8 kcalmolꢀ1), which is slightly larger than that
calculated for 2!3 (ꢀ38.7 to ꢀ42.8 kcalmolꢀ1).
2
ꢀ
double bond is formed, and the C Li bond length has
reached its final value.
Thus, it seems that the robust Li–Oacetal coordination in A
controls the configuration of the double bond. Therefore,
our calculations suggest that the unprecedented anti charac-
ter of this carbolithiation[2c] can
Note that the lateral chain of benzofuran 3b is an alkyl-
lithium, whereas it is a vinyllithium in 3. One can thus
be explained by the strong in-
tramolecular
coordination,
which derails the reaction from
its usual course.
Cyclization of allenyl isomer
1d: The previous study had evi-
denced the facile cyclization of
allene 1b (Scheme 1) in reac-
Scheme 4. Cyclization of allene 1d under the influence of 1.0 equiv of nBuLi.
tion conditions that were comparable to those that were em-
ployed for 1 (Scheme 1). The yield of this step was good
(85% instead of 62% for 1) but the selectivity was slightly
diminished (E/Z=86:14 instead of 100:0). A complementary
experiment was thus performed to check that the cyclization
takes place if only one equivalent of n-butyllithium is added
to the allene. Bromoallene 1d, which was prepared from 1c
according to a simple and quantitative isomerization proce-
dure was used instead of iodoallene 1b. When performed at
ꢀ788C in THF, the cyclization reaction led to a quantitative
transformation of 1d into 5 as a 86:14 E/Z mixture
(Scheme 4).
It is noteworthy that 1d does not require an excess of n-
butyllithium to trigger the final elimination; this indicates
that this step follows a different mechanism.
We first carried out a full optimization of the structure of
2b (Scheme 1) that was solvated by two explicit molecules
Figure 7. Energy diagram for the cyclization of allene 2b. Ecorr refers to
the absolute energy of the compounds after ZPE corrections (in a.u.).
The relative energies on the diagram are in kcalmolꢀ1
.
Chem. Eur. J. 2008, 14, 5159 – 5167
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5165

J. Maddaluno, C. FressignØ et al.
warmed to room temperature, stirred for 3 h and washed by adding 0.4m
NaOH (100 mL). The mixture was diluted by Et2O (150 mL) and washed
again by 0.4m NaOH solution (2”100 mL). The combined organic layers
were dried over anhydrous MgSO4 and concentrated. Pentane (300 mL)
was added to the residue to precipitate triphenylphosphine oxide. After
filtration, the resulting brown liquid was purified by column chromatog-
raphy (30% EtOAc in cyclohexane) to give the pure product 1c (14.97 g,
92%) as a yellow liquid. 1H NMR (CDCl3): d=1.20 (t, J=7.2 Hz, 6H),
3.55 (qd, J=7.2, 9.4 Hz, 2H), 3.68 (qd, J=7.2, 9.4 Hz, 2H), 4.83 (d, J=
1.1 Hz, 2H), 5.28 (t, J=1.1 Hz, 1H), 6.88 (td, J=7.5, 1.5 Hz, 1H), 7.07
(dd, J=7.5, 1.5 Hz, 1H), 7.27 (td, J=7.5, 1.5 Hz, 1H), 7.55 ppm (dd, J=
7.5, 1.5 Hz, 1H); 13C NMR (CDCl3): d=15.2, 54.1, 61.0, 79.7, 83.7, 91.4,
112.5, 114.4, 123.1, 128.7, 133.8, 154.2 ppm; EIMS: m/z (%): 313 (21)
[M]+, 311 (21) [M]+, 268 (34), 188 (27) [MꢀOEt]+, 160 (68) [Mꢀ
expect that the mechanism of the final b-elimination will be
quite different from that involving 3. These two routes will
be detailed in a forthcoming article.
Conclusion
This paper details the mechanism of the intramolecular car-
bolithiation of lithiated propargylic ether 2. The experimen-
tal results show that one equivalent of n-butyllithium trig-
gers the halogen–lithium exchange and the subsequent het-
erocyclization step to afford a dihydrobenzofuran nucleus
that bears an exocyclic vinyllithium moiety. The NMR spec-
troscopic study on the adduct has shown that this reaction
results from an unprecedented anti addition on the alkyne.
DFT calculations show that this unexpected characteristic is
related to the intramolecular coordination of the lithium
atom by one oxygen atom of the terminal acetal appendage.
This persisting O–Li interaction is observed all along the
cyclization pathway, and drives the cation to the E site of
the olefin. The calculations show that in absence of this co-
ordination (as in conformers B and C), the Z olefin that
would result from a classical syn addition should be ob-
tained. The experiments were repeated on the allene 1d. In
this case, one equivalent of n-butyllithium suffices to trigger
the exchange, the cyclization, and the final elimination of
lithium ethoxide. The DFT results suggest that the intramo-
lecular addition of the original aryllithium on the central
carbon atom of the allene 2b yields the expected benzofur-
an skeleton, which bears a lithiated lateral chain at the 3-po-
sition that is ready for a b-elimination process. This cycliza-
tion goes through a low-lying transition state, as is expected
for a rapid reaction at low temperature.
(OEt)2]+, 131 (62) [MꢀCH
(OEt)2]+, 85 (85), 68 (100).
1-Bromo-2-(4,4-diethoxybuta-1,2-dienyloxy)benzene (1d): Potassium tert-
butoxide (0.179 g, 1.6 mmol, 1.6 equiv) was added portionwise to a solu-
tion of acetylene 1c (0.313 g, 1.0 mmol) in THF (10 mL) at 08C under
N2. The mixture was stirred for 20 min at 08C and H2O (10 mL) was
added. The organic layers were washed with a saturated solution of NaCl
(5 mL), dried over anhydrous MgSO4, and concentrated to give the
allene 1d (0.312 g, 99%) as a brown oil. 1H NMR (CDCl3): d=1.16 (t,
J=7.1Hz, 3H), 1.21 (t, J=7.1Hz, 3H), 3.50 (m, 4H), 4.91 (dd, J=5.6,
1.1 Hz, 1H), 5.79 (t, J=5.6 Hz, 1H), 6.96 (m, 2H), 7.19 (dd, J=1.5,
7.9 Hz, 1H), 7.26 (td, J=1.5, 7.9 Hz, 1H), 7.55 ppm (dd, J=1.5, 7.9 Hz,
1H); 13C NMR (CDCl3): d=15.4, 15.5, 61.4, 61.9, 100.1, 105.9, 114.3,
119.6, 120.9, 125.1, 128.7, 133.8, 153.7, 197.3 ppm.
3-(2-Ethoxyvinyl)benzofuran (5): A 2.05m solution of n-butyllithium in
hexane (0.95 mL, 0.51mmol, 1.03 equiv) was added to a solution of the
allene 1d (0.156 g, 0.50 mmol) in anhydrous THF (2.5 mL) at ꢀ788C
under an argon atmosphere. After 15 min of stirring, the mixture was hy-
drolyzed with H2O (4 mL). The aqueous phase was separated and ex-
tracted with Et2O (3”4 mL). The combined organic phases were dried
over anhydrous MgSO4, and concentrated to provide a mixture two iso-
mers of benzofuran 5[1] (0.094 g, 100%, E/Z=86:14) as a brown oil.
E isomer: 1H NMR (CDCl3): d=1.27 (t, J=7.1Hz, 3H), 3.83 (q, J=
7.1Hz, 2H), 5.76 (d, J=13.2 Hz, 1H), 6.98 (d, J=13.2 Hz, 1H), 7.18 (m,
2H), 7.37 (dd, J=1.7,.2 Hz, 1H), 7.42 (s, 1H), 7.55 ppm (dd, J=1.5,
6.0 Hz, 1H); 13C NMR (CDCl3): d=15.2, 65.9, 95.6, 112.0, 117.3, 120.8,
123.0, 124.8, 126.8, 140.7, 148.50, 156.0 ppm.
The complete description of the reaction requires a de-
tailed study of the mechanism of the final elimination step.
This work is under way on the E and Z olefins that result
from the cyclization of conformers A–C. The results will be
reported in due time.
Z isomer: 1H NMR (CDCl3): d=1.21 (t, J=7.1Hz, 3H), 3.86 (q, J=
7.1Hz, 2H), 5.24 (d, J=6.4 Hz, 1H), 6.20 (d, J=6.4 Hz, 1H), 7.05–7–45
(m, 4H), 7.87 ppm (s, 1H); 13C NMR (CDCl3): d=15.7, 69.0, 93.8, 111.4,
115.4, 119.5, 122.4, 124.2, 127.2, 143.6, 146.8, 154.6 ppm; MS (CI, CH4):
m/z (%): 189 (100) [M+H]+, 161 (29); IR: n˜ =2926, 1380, 1127 cmꢀ1
(E)-3-(2,2-Diethoxyethylidene)-2,3-dihydrobenzofuran (6 and 6D):
.
A
2.37m solution of n-butyllithium in hexane (0.51 mL, 1.2 mmol, 1.2 equiv)
was added to a solution of the acetal 1c (0.313 g, 1.0 mmol) in anhydrous
THF (5 mL) at ꢀ788C under argon atmosphere. After 15 min of stirring,
the mixture was deuterolyzed with EtOD (0.6 mL). After 15 min of stir-
ring, H2O (5 mL) was added. The aqueous phase was separated and ex-
tracted with Et2O (3”5 mL). The combined organic phases were dried
over anhydrous MgSO4 and concentrated. The residue was purified by
column chromatography (10% EtOAc in cyclohexane) to provide a mix-
ture of 6 and 6D (0.197 g, 84%, H/D=66:34) as an orange liquid.
Experimental Section
General consideration: THF and Et2O were dried from Na/benzophe-
none. All reagents were of reagent grade and were used as such or dis-
tilled prior to use. Reactions were monitored by TLC, which was carried
out on 0.25 mm E. silica-gel-coated aluminium plates (60 F254) by using
UV light as a visualizing agent, and 7% ethanolic phosphomolybdic acid
and heat were used as a developing agent, or by GC with a 24 m HP-
methyl silicon capillary column. E. silica gel (60, particle size 0.04–
0.063 mm) was used for flash chromatography. 1H and 13C NMR spectra
were recorded at room temperature at 300 and 75 MHz respectively, and
were calibrated by using residual undeuterated solvent as an internal ref-
erence. The solvent was CDCl3 or [D6]benzene. The mass spectra were
obtained under electron impact conditions (EI) at 70 eV ionizing poten-
tial.
Compound 6: 1H NMR (CDCl3): d=1.23 (t, J=7.2 Hz, 6H), 3.58 (qd,
J=7.2, 9.4 Hz, 2H), 3.71(qd, J=7.2, 9.4 Hz, 2H), 5.10 (d, J=1.1 Hz,
2H), 5.55 (m, 2H), 6.87 (d, J=7.9 Hz, 1H), 6.94 (t, J=7.9 Hz, 1H), 7.22
(td, J=1.2, 7.9 Hz, 1H), 7.62 ppm (d, J=7.9 Hz, 1H); 1H NMR
([D6]benzene): d=1.12 (t, J=7.2 Hz, 6H), 3.45 (qd, J=7.2, 9.4 Hz, 2H),
3.59 (qd, J=7.2, 9.4 Hz, 2H), 4.63 (dd, J=1.5, 2,3 Hz, 2H), 5.43 (dt, J=
2.3, 5.6 Hz, 1H), 5.56 (dt, J=1.5, 5.6 Hz, 1H), 6.77 (td, J=1.5, 7.6 Hz,
1H), 6.82 (d, J=7.6 Hz, 1H), 6.95 (td, J=1.5, 7.6 Hz, 1H), 7.86 ppm (d,
J=7.6 Hz, 1H); 13C NMR (CDCl3): d=15.7, 60.7, 75.6, 98.1, 111.0, 116.8,
121.2, 124.3, 126.2, 131.3, 139.8, 165.4 ppm; EIMS: m/z (%): 234 (1)
1-Bromo-2-(4,4-diethoxybut-2-ynyloxy)benzene (1c): Diisopropylazodi-
carboxylate (10.30 mL, 52 mmol, 1.0 equiv) was added dropwise to a so-
lution of 4,4-diethoxy-but-2-yn-1-ol[1a] (8.21g, 52 mmol), 2-bromophenol
(6.18 mL, 52 mmol, 1.0 equiv), and triphenylphosphine (13.63 g, 52 mmol,
1.0 equiv) in THF (300 mL) at 08C under argon. The solution was
[M+], 188 (80) [MꢀOEt]+, 160 (21) [Mꢀ(OEt)2]+, 131 (100) [MꢀCH-
A
A
5166
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 5159 – 5167

Carbolithiation of Alkynes
FULL PAPER
1
Compound 6D: H NMR ([D6]benzene): d=1.12 (t, J=7.2 Hz, 6H), 3.45
4283; c) C. FressignØ, A. Lautrette, J. Maddaluno, J. Org. Chem.
2005, 70, 7816–7828.
(qd, J=7.2, 9.4 Hz, 2H), 3.61(qd, J=7.2, 9.4 Hz, 2H), 4.63 (d, J=1.5 Hz,
2H), 5.56 (s, 1H), 6.77 (td, J =1.5, 7.6 Hz, 1H), 6.83 (d, J=7.6 Hz, 1H),
6.95 (td, J=1.5, 7.6 Hz, 1H), 7.88 ppm (d, J=7.6 Hz, 1H); EIMS: m/z
[11] Jaguar 5.0, release 22, Schrçdinger, LLC., Portland, OR, 2002.
[12] a) B. Silvi, A. Savin, Nature 1994, 371, 683–686; b) A. Savin, B.
Silvi, F. Colonna, Can. J. Chem. 1996, 74, 1088–1096. For a review,
see: c) A. Savin, R. Nesper, S. Wengert, T. F. Fässler, Angew. Chem.
1997, 109, 1892–1918; Angew. Chem. Int. Ed. Eng. 1997, 36, 1808–
1832.
[13] C. FressignØ, J. Maddaluno, C. Giessner-Prettre, B. Silvi, J. Org.
Chem. 2001, 66, 6476–6479.
[14] a) D. B. Chesnut, J. Phys. Chem. A 2000, 104, 11644–11650; b) D. B.
Chesnut, J. Phys. Chem. A 2003, 107, 4307–4313; c) E. Matito, J.
Poater, F. M. Bickelhaupt, M. Sola, J. Phys. Chem. B 2006, 110,
7189–7198.
[15] C. FressignØ, A.-L. Girard, M. Durandetti, J. Maddaluno, Angew.
Chem. 2008, 120, 905–907; Angew. Chem. Int. Ed. 2008, 47, 891–
893.
[16] Note however that the anti isomers that result from the isomeriza-
tion of the original syn product were observed when the resulting vi-
nyllithium was stabilized. See reference [2] and a) W. Bauer, M.
Feigel, G. Müller, P. von R. Schleyer, J. Am. Chem. Soc. 1988, 110,
6033–6046; b) N. F. McKinley, D. F. OꢁShea, J. Org. Chem. 2006, 71,
9552–9555.
[17] F. Le Strat, J. Maddaluno, Tetrahedron Lett. 2000, 41, 5367–5371.
[18] W. F. Bailey, J. J. Patricia, J. Organomet. Chem. 1988, 352, 1–46.
[19] W. Bauer, C. Griesinger, J. Am. Chem. Soc. 1993, 115, 10871–10882.
[20] See, for instance reference [2a] and G. Gralla, B. Wibbeling, D.
Hoppe, Org. Lett. 2002, 4, 2193–2195.
[21] See, for instance: a) W. F. Bailey, T. V. Ovaska, Tetrahedron Lett.
1990, 31, 627–630; b) W. F. Bailey, T. V. Ovaska, J. Am. Chem. Soc.
1993, 115, 3080–3090; c) X. Wei, R. J. K. Taylor, Tetrahedron Lett.
2003, 44, 7143–7146.
(%): 235 (1) [M]+, 189 (80) [MꢀOEt]+, 161 (21) [Mꢀ
(OEt)2]+, 1 32
(100) [MꢀCH
(OEt)2]+.
(4,4-Diethoxybut-2-ynyloxy)benzene (7): A 2.0m solution of n-butyllithi-
um in hexane (1.0 mL, 2.0 mmol, 2.0 equiv) was added to a solution of
the acetal 1c (0.313 g, 1.0 mmol) in anhydrous Et2O (5 mL) at ꢀ788C
under an argon atmosphere. After 15 min of stirring, the mixture was hy-
drolyzed with H2O (5 mL). The aqueous phase was separated and ex-
tracted with Et2O (3”5 mL). The combined organic phases were dried
over anhydrous MgSO4 and concentrated. The residue was purified by
column chromatography (10% EtOAc in cyclohexane) to provide 7
(0.624 g, 62%) as a yellow liquid and a mixture of 6 (0.120 g, 12%,) as a
1
yellow liquid. H NMR (CDCl3): d=1.21 (t, J=6.8 Hz, 6H), 3.56 (qd, J=
7.2, 9.4 Hz, 2H), 3.70 (qd, J=7.2, 9.4 Hz, 2H), 4.75 (s, 2H), 5.29 (s, 1H),
6.98, 7.29 ppm (m, 5H); MS (EI): m/z (%): 234 (5) [M]+, 189 (60)
[MꢀOEt]+, 161 (47), 131 (22), 85 (100), 68 (95).
Acknowledgements
Computations have been carried out at the Centre de Ressources Infor-
matiques de Haute-Normandie (CRIHAN, St Etienne-du-Rouvray) and
the Centre Informatique National de l’Enseignement SupØrieur (CINES,
Montpellier). ALG acknowledges the PUNCHorga interregional net-
work for a PhD fellowship. We warmly thank Prof. H. Oulyadi (Univ.
Rouen) for his continuous help and spectroscopic contribution to this
program. We would also like to express our gratitude to Dr. C. Giessner-
Prettre (LCT, University Paris VI) for her important input to these re-
sults.
[22] D. J. Tannor, B. Marten, R. Murphy, R. A. Friesner, D. Sitkoff, A.
Nicholls, N. Ringnalda, W. A. Goddard III, B. Honig, J. Am. Chem.
Soc. 1994, 116, 11875–11882.
[23] See, for instance: a) L. M. Pratt, Bull. Chem. Soc. Jpn. 2005, 78,
890–898; b) Y. Yuan, S. Desjardins, A. Harrison-Marchand, H. Ou-
lyadi, C. FressignØ, C. Giessner-Prettre, J. Maddaluno, Tetrahedron
2005, 61, 3325–3334; c) L. M. Pratt, D. G. Truhlar, C. J. Cramer,
S. R. Kass, J. D. Thompson, J. D. Xidos, J. Org. Chem. 2007, 72,
2962–2966.
[24] Gaussian 98 (Revision A5), M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G. Zakrzewfki,
J. A. Montgomery, R. E. Stratmann, J. C. Burant, S. Dapprich, J. M.
Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J.
Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli,
C. Adamo, F. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala, Q.
Cui, K. Morokuma, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. Cioslowski, J. V. Ortiz, B. B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R. L. Martin,
D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,
C. Gonzalez, M. Challacombe, P. M. W. Gill, B. G. Johnson, W.
Chen, M. W. Wong, J. L. Andres, M. Head-Gordon, E. S. Replogle,
J. A. Pople, Gaussian Inc., Pittsburgh PA, 1998.
[25] It is well-known that the lone pair of a free pyramidal carbanion
does not necessarily face its counterion. For a spectacular example
that involves a monomeric organolithium compound, see: F. Breher,
J. Grunenberg, S. C. Lawrence, P. Mountford, H. Rüegger, Angew.
Chem. 2004, 116, 2575–2578; Angew. Chem. Int. Ed. 2004, 43, 2521–
2524.
[26] A continuum solvent effect can be used to describe reactivity prob-
lems in organolithium chemistry very well: T. Balle, M. Begtrup,
J. W. Jaroszewski, T. Liljefors, P.-O. Norrby, Org. Biomol. Chem.
2006, 4, 1261–1267.
[1] a) F. Le Strat, J. Maddaluno, Org. Lett. 2002, 4, 2791–2793; b) F.
Le Strat, D. C. Harrowven, J. Maddaluno, J. Org. Chem. 2005, 70,
489–498.
[2] a) W. F. Bailey, T. V. Ovaska, T. K. Leipert, Tetrahedron Lett. 1989,
30, 3901–3904; b) W. F. Bailey, T. V. Ovaska, J. Am. Chem. Soc.
1993, 115, 3080–3090; c) W. F. Bailey, N. M. Wachter-Jurcsak, M. R.
Pineau, T. V. Ovaska, R. R. Warren, C. E. Lewis, J. Org. Chem.
1996, 61, 8216–8228.
[3] See, for instance: a) M. J. Mealy, W. F. Bailey, J. Organomet. Chem.
2002, 646, 59–67; b) I. Marek, N. Chinkov, D. Banon-Tenne in
Metal-Catalyzed Cross-Coupling Reactions (Eds.: F. Diederich, A.
de Meijere) Wiley-VCH, Stuttgart, 2004, Chapter 7, pp. 395–478;
c) F. J. FaÇanµs, R. Sanz in The Chemistry of Organolithium Com-
pounds (Eds.: Z. Rappoport, I. Marek), Vol. 2, Wiley, Chichester,
2006, Chapter 4, pp. 295–379; d) R. Sanz, J. M. Ignacio, M. A. Ro-
driguez, F. J. FaÇanµs, J. Barluenga, Chem. Eur. J. 2007, 13, 4998–
5008.
[4] W. F. Bailey, P. H. Aspris, J. Org. Chem. 1995, 60, 754–757.
[5] See, for instance: a) G. Boche, M. Schimeczek, J. Cioslowski, P. Pis-
korz, Eur. J. Org. Chem. 1998, 1851–1860; b) K. B. Wilberg, S. Skle-
nac, W. F. Bailey, J. Org. Chem. 2000, 65, 2014–2021; c) K. Ando, J.
Org. Chem. 2006, 71, 1837–1850. For a recent in-depth experimental
study see also: d) W. F. Bailey, M. R. Luderer, K. P. Jordan, J. Org.
Chem. 2006, 71, 2825–2828.
[6] E. Nakamura, M. Nakamura, Y. Miyachi, N. Koga, K. Morokuma, J.
Am. Chem. Soc. 1993, 115, 99–106.
[7] E. Nakamura, Y. Miyachi, N. Koga, K. Morokuma, J. Am. Chem.
Soc. 1992, 114, 6686–6692.
[8] W. F. Bailey, A. D. Khanolkar, K. Gavaskar, T. V. Ovaska, K. Rossi,
Y. Thiel, K. B. Wilberg, J. Am. Chem. Soc. 1991, 113, 5720–5727.
[9] W. F. Bailey, M. J. Mealy, K. B. Wilberg, Org. Lett. 2002, 4, 791–794.
[10] See, for instance: a) C. FressignØ, J. Maddaluno, A. Marquez, C.
Giessner-Prettre, J. Org. Chem. 2000, 65, 8899–8907; b) L. M. Pratt,
N. V. NguyÞn, B. Ramachandran, J. Org. Chem. 2005, 70, 4279–
[27] Y. Yamamoto, J. Org. Chem. 2007, 72, 7817–7831.
Received: November 21, 2007
Published online: April 25, 2008
Chem. Eur. J. 2008, 14, 5159 – 5167
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5167