Metal-catalyzed cyclization of β- and γ-allenols derived from d-glyceraldehyde-synthesis of enantiopure dihydropyrans and tetrahydrooxepines: an experimental and theoretical study

Article abstract of DOI:10.1002/chem.200901180

Regiocontrolled metal-catalyzed preparations of enantiopure dihydropyrans and tetrahydrooxepines have been synthesized starting from β- and y-allenols derived from D-glyceraldehyde. The Pd^II-catalyzed cyclizative coupling reactions of β-allenol

Full text of DOI:10.1002/chem.200901180

FULL PAPER
DOI: 10.1002/chem.200901180
Metal-Catalyzed Cyclization of b- and g-Allenols Derived from
d-Glyceraldehyde—Synthesis of Enantiopure Dihydropyrans and
Tetrahydrooxepines: An Experimental and Theoretical Study
Benito Alcaide,*[a] Pedro Almendros,*[b] Teresa Martꢀnez del Campo,[a]
Elena Soriano,[b] and Josꢁ L. Marco-Contelles[b]
Abstract: Regiocontrolled metal-cata-
lyzed preparations of enantiopure dihy-
dropyrans and tetrahydrooxepines have
been synthesized starting from b- and
g-allenols derived from d-glyceralde-
hyde. The PdII-catalyzed cyclizative
coupling reactions of b-allenols 1a and
1b with allyl bromide effectively af-
forded enantiopure tetrafunctionalized
dihydropyrans through a 6-endo oxy-
cyclization protocol, whereas the gold-,
platinum-, and palladium-mediated
heterocyclization of g-allenol 2 fur-
cyclization reactions. Moreover, density
functional calculations were performed
to predict the regioselectivity of the g-
allenol cycloetherification to tetrahy-
drooxepines on the basis of both the
tether nature and characteristics of the
metals, and to gain an insight into the
mechanism of the oxycyclization reac-
tions.
nished
tetrahydrooxepines
13–16
through regioselective 7-endo-trig oxy-
Keywords: allenes
gold palladium
reaction mechanisms
·
cyclization
· platinum
·
·
·
Introduction
and allene chemistry,[4] herein we report the metal-catalyzed
cyclization of b- and g-allenols derived from d-glyceralde-
hyde leading to enantiopure dihydropyrans and tetrahy-
drooxepines. These cyclic ethers often occur in nature as the
structural units of natural products and have been used for
the synthesis of biologically active compounds.[5]
The transition-metal-catalyzed reactions of allenes have
been extensively studied in recent years.[1] In particular, the
transition-metal-catalyzed reactions of a-allenols leading to
heterocyclization products have attracted a great deal of in-
terest.[2] However, relatively little work has been performed
on the intramolecular cyclization reactions of b- and g-alle-
nols.[3] Intramolecularization of the reactions, usually by
placing the group at such a distance that five- or six-mem-
bered rings are formed, should automatically solve the posi-
tional selectivity problems because larger rings are unfa-
vored. In connection with our previous work on heterocyclic
Results and Discussion
The starting substrates, enantiopure b-allenols 1 and g-alle-
nol 2 (Figure 1), were prepared in good overall yields start-
ing from (R)-2,3-O-isopropylideneglyceraldehyde (3) by a
regiocontrolled indium-mediated Barbier-type carbonyl–al-
lenylation reaction in aqueous media to give a-allenols 4a
and 4b (Scheme 1) followed by manipulation of the protect-
ing groups (Scheme 2). The indium-mediated allenylation of
[a] Prof. Dr. B. Alcaide, Dr. T. M. del Campo
Departamento de Quꢀmica Orgꢁnica I, Facultad de Quꢀmica
Universidad Complutense de Madrid, 28040 Madrid (Spain)
Fax : (+34)91-3944103
E-mail: alcaideb@quim.ucm.es
[b] Dr. P. Almendros, Dr. E. Soriano, Prof. Dr. J. L. Marco-Contelles
Instituto de Quꢀmica Orgꢁnica General
Consejo Superior de Investigaciones Cientꢀficas, CSIC
Juan de la Cierva 3, 28006 Madrid (Spain)
Fax : (+34)91-5644853
E-mail: Palmendros@iqog.csic.es
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200901180.
Figure 1. Structures of the starting b-allenols 1 and g-allenol 2. PMP=
4-MeOC6H4; TPS=tert-butyldiphenylsilyl; TBS=tert-butyldimethylsilyl.
Chem. Eur. J. 2009, 15, 9127 – 9138
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
9127

1
by comparison of the H NMR chemical shifts of the corre-
sponding acetylmandelates 5a and 5b;[7] the same approach
was used for a-allenol 4b. This high diastereoselectivity can
be explained by allenylation from the less hindered face of
the carbonyl group according to the model depicted in
Figure 2.[8] This model assumes the possible participation of
the b-alkoxy oxygen atom and the carbonyl group, but not
the a-alkoxy oxygen atom.
Scheme 1. Preparation of enantiopure a-allenols 4 and acetylmandelates
5. i) In, THF/NH4Cl (aq. sat.), RT, 5 h; ii) GCOOH, DCC, DMAP (cat.),
RT, 14 h. DCC=dicyclohexylcarbodiimide; DMAP=4-dimethylamino-
pyridine.
Figure 2. Model to explain the observed anti stereochemistry of the alle-
nylation reaction.
a-Allenols 4 were protected as their corresponding p-me-
thoxybenzoates 6, which were smoothly converted into the
diols 7 in quantitative yields by treatment with bismuth tri-
chloride. Selective silyl protection of the primary hydroxy
group in diols 7 afforded b-allenols 1a and 1b (Scheme 2).
g-Allenol 2 was achieved from b-allenol 1a by sequential
acetate protection to give 8 followed by cleavage of the tert-
butyldiphenylsilyl ether moiety.
Our initial goal was to evaluate the effect of metals on
the hydroalkoxylation reaction of b-allenols 1a and 1b. De-
spite the recent precedent for the gold-catalyzed cycloisome-
rization of b-allenols to dihydropyrans,[9] AuI and AuIII salts
were found not to be competent precatalysts for our sub-
strates. No trace of the oxacycle could be detected by using
AuCl or AuCl3 precatalysts in the absence or presence of
different additives. We studied the corresponding 6-endo
cyclization reaction of 1a under platinum catalysis by using
Scheme 2. Preparation of enantiopure b-allenols
1 and g-allenol 2.
i) PMPCOCl, Et3N, DMAP, CH2Cl2, RT, 18 h; ii) 20 mol% BiCl3, MeCN/
H2O, RT, 24 h; iii) TPSCl, imidazole, CH2Cl2, 08C, 1 h; or TBSCl, imida-
zole, DMF, RT, 12 h; iv) Ac2O, Et3N, CH2Cl2, RT, 0.5 h; v) TBAF, THF,
08C, 1 h. PMP=4-MeOC6H4; DMPA=4-dimethylaminopyridine; TPS=
tert-butyldiphenylsilyl; TBS=tert-butyldimethylsilyl; TBAF=tetra-n-bu-
tylammonium fluoride.
[PtCl2ACHTUNGTRNEUNG
(CH2=CH2)]2,[10] but to our surprise the reaction took
a different course. No dihydropyran was observed, but in-
stead a furan was formed (Scheme 3). This reaction can be
aldehyde 3 took place with complete diastereoselectivity.[6]
The configuration at the carbinolic chiral center of the enan-
tiopure a-allenol 4a was established by using Trostꢃs method
Abstract in Spanish: La reacciꢀn de heterociclaciꢀn de b- y
g-alenoles derivados de d-gliceraldehꢁido catalizada por me-
tales es un proceso regiocontrolado que da lugar a dihidropi-
ranos y tetrahidrooxepinas enantiopuros. La reacciꢀn de ci-
claciꢀn/acoplamiento cruzado entre los b-alenoles 1a y 1b
con bromuro de alilo fue eficaz como protocolo de oxicicla-
ciꢀn 6-endo para dar lugar a dihidropiranos tetrafuncionali-
zados, mientras que las heterociclaciones del g-alenol 2 cata-
lizadas por oro, platino y paladio generan las tetrahidrooxe-
pinas 13–16 via una oxiciclaciꢀn 7-endo-trig regioespecꢁfica.
Ademꢂs, los mecanismos de estas heterociclaciones de g-ale-
noles catalizadas por metales han sido investigados teꢀrica-
mente.
Scheme 3. Platinum-catalyzed heterocyclization reaction of b-allenol de-
rivative 1a. i) 1 mol% [PtCl2ACHTUNRGTENUNG(CH2=CH2)]2, 2 mol% TDMPP, CH2Cl2, RT,
4 h. PMP=4-MeOC6H4; TPS=tert-butyldiphenylsilyl; TDMPP=tris(2,6-
dimethoxyphenyl)phosphine.
explained by a 5-exo cyclization by attack of the hydroxy
group on the central allene carbon to give the nonisolable
dihydrofuran 9, which under the reaction conditions suffers
aromatization to furan 10. When we tested the reactivity of
our b-allenols by using a Pd–Cu bimetallic system, com-
pound 1a produced a complex mixture of products, includ-
ing bromodihydropyran 11 (10% yield; Scheme 4). Gratify-
ingly, it was found that the PdII-catalyzed cyclizative cou-
9128
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9127 – 9138

Enantiopure Dihydropyrans and Tetrahydrooxepines
FULL PAPER
metal-mediated heterocyclization reaction was tested with
substrate 2. Treatment of g-allenol 2 with AgNO3 in THF/
H2O (1:1) at reflux furnished the desired tetrahydrooxepine
13 although only in modest yield (38%), the best cycloiso-
merization result being obtained with 5 mol% AuCl3
(Scheme 6).[13] The preferential regioselective 7-endo cycliza-
Scheme 4. Palladium-promoted preparation of dihydropyrans 11 and 12.
i) 7 mol% PdACHTUNGTRENNUNG(OAc)2, LiBr, CuAHCTUTGNREN(NUGN OAc)2, K2CO3, MeCN, O2, RT, 2 h; ii) all-
yl bromide, 5 mol% PdCl2, DMF, RT, 5 h. PMP=4-MeOC6H4; TBS=
tert-butyldimethylsilyl; TPS=tert-butyldiphenylsilyl.
pling reaction of b-allenols 1a and 1b with allyl bromide
was effective as a 6-endo oxycyclization protocol, affording
enantiopure
tetrafunctionalized
dihydropyrans
12
(Scheme 4). TBS cleavage was observed to an appreciable
extent during the reaction of phenyl-b-allenol 1b with allyl
bromide in the presence of PdCl2.
We recently reported the versatile regiocontrolled metal-
catalyzed cycloetherification reactions of 2-azetidinone-teth-
ered g-allenol derivatives, which gave a variety of fused tet-
rahydrofuran-, dihydropyran-, and tetrahydrooxepine-b-lac-
tams (Scheme 5).[11] It has been shown that the control of
Scheme 6. Metal-promoted preparation of tetrahydrooxepines 13–16. i) 5
mol% AuCl3, CH2Cl2, RT, 3 h; ii) 1 mol% [PtCl
TDMPP, CH2Cl2, RT, 2 h; iii) allyl bromide, 5 mol% PdCl2, DMF, RT,
2 h; iv) 7 mol% Pd(OAc)2, LiBr, Cu(OAc)2, K2CO3, MeCN, O2, RT, 9 h.
2ACHTUNGTRNE(UNNG CH2=CH2)]2, 2 mol%
A
ACHTUNGTRENNUNG
PMP=4-MeOC6H4; TDMPP=tris(2,6-dimethoxyphenyl)phosphine.
tion reaction described herein differs markedly from that of
the reported gold-mediated oxycyclization of g-alle-
nols,[3g,11,13] namely a 5-exo cyclization leading to 2-vinylte-
trahydrofurans. Interestingly, a similar regioselectivity was
Scheme 5. Regiocontrolled metal-catalyzed preparation of fused tetrahy-
drofuran-, dihydropyran-, and tetrahydrooxepine-b-lactams.
observed with [PtCl2ACTHNUTRGNEUNG
(CH2=CH2)]2 as the catalyst.[14] Howev-
ꢀ
the regioselectivity in the O C functionalization of g-alle-
er, product 13 was not detected, its isomer 14 being the sole
reaction product. Conjugation of the double bond with the
lone pair of the oxygen atom under platinum-catalyzed con-
ditions is believed to promote the formation of 14. The
PdCl2-catalyzed reaction between allyl bromide and g-alle-
nol 2 afforded the tetrasubstituted tetrahydrooxepine 15 in
a totally regioselective fashion (Scheme 6), as was observed
for the related b-lactam precursor.[11,15] Next, we decided to
test if a related transformation could be accessible through
the palladium(II)-catalyzed oxybromination of g-allenol 2.
Indeed, bromotetrahydrooxepine 16 was obtained as a
single isomer in reasonable yield (Scheme 6). Of special in-
terest is the reversal of the regioselectivity in the nucleophil-
ic insertion of g-allenol 2 in comparison with the recently re-
ported cyclization of simple g-allenic alcohols under similar
Pd–Cu bimetallic reaction conditions.[16]
nols can be achieved through the proper choice of catalyst.
Furthermore, the regioselectivity was strongly modulated by
the status of the hydroxy group (i.e., free or protected) in
the methyl-g-allenol. To address the effects of both the
metal and the hydroxy moiety, a detailed mechanistic inves-
tigation was carried out over all the reaction pathways.[12]
For the g-allenol precursors, the results revealed an initially
kinetically favored 5-exo-trig cyclization, whereas the 7-
endo-trig cyclization mode appeared to be a less favorable
route (activation barrier 6–8 kcalmolꢀ1 higher), in part be-
cause the ring strain imposed by the lactam ring and the en-
docyclic alkene group restrains the flexibility of the tether
and the successful interaction between the reactive centers.
Therefore, attention should also be paid to the influence of
the nature of the tether between the allene moiety and the
functionality.
The cyclic structures (by DEPT, HMQC, HMBC, and
COSY) and the stereochemistries (by vicinal proton cou-
plings and qualitative homonuclear NOE difference spectra)
To address the role of the tether, the reactivity of g-alle-
nols lacking a b-lactam ring towards the regioselective
Chem. Eur. J. 2009, 15, 9127 – 9138
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9129

B. Alcaide, P. Almendros et al.
ꢀ
ꢀ
Table 1. NPA charge, bond lengths, Wiberg bond index of the M Cn and M Cn+1 interactions, respectively,
and free-energy differences between complexes of type a and b.
of furan 10, dihydropyrans 11
and 12, and tetrahydrooxepines
13–16 were assigned by 1D and
2D NMR analyses.
All of these metal-catalyzed
transformations may involve a
ꢀ
ꢀ
ꢀ
MLx
NPA charge Cn C
n+1 [ꢄ] M Cn [ꢄ] M Cn+1 [ꢄ] WBI WBI DG[a]
[kcalmolꢀ1
]
18a
18b
AuCl3
PtCl2ACHTNUGTRNEUNG(CH2CH2)
ꢀ0.335
ꢀ0.196
ꢀ0.326
ꢀ0.268
1.389
1.378
1.403
1.356
1.363
1.385
2.205
2.165
2.035
2.291
2.211
2.124
2.800
2.603
2.366
2.464
2.284
2.121
0.309 0.078
0.355 0.194
0.492 0.351
0.258 0.168
0.353 0.307
PdCl2
AuCl3
chemoselective
(7-endo-trig
ꢀ3.1
ꢀ9.5
versus 6-endo-dig versus 5-exo-
trig) g-allenol cycloetherifica-
[PtCl
2A
PdCl2
ꢀ0.304
0.480 0.428 ꢀ10.8
tion reaction.
[a] Free-energy differences relative to 18a.
With the aim of shedding
some light on the dependence
of the regioselectivity on the tether, we carried out a compu-
tational study of the possible competing cycloetherification
routes. We selected g-allenol 17 (Figure 3) as a theoretical
model, a closely related structure to the parent precursor 2,
to include the main effects and optimize computational re-
sources.
The computed energy values reveal that the 5-exo-trig
cyclization (transition-state structure TS15) takes place with
a higher activation barrier than the 6-exo-dig reaction
Table 2. Enthalpies and free energies in the gas phase and free energies
in solution for the cyclization of 17 by alternative regioisomeric 5-exo-
trig, 6-exo-dig, and 7-endo-trig heterocyclization pathways.[a]
AuCl3
G
G
PdCl2
DHgas DGgas DGsol DHgas DGgas DGsol DHgas DGgas DGsol
18b
TS15
0.0
6.6
0.0
8.6
0.0
4.9 15.5
2.9
2.5 11.0
0.0
0.0
17.6
6.4
0.0
15.2
4.2
0.0
14.6
ꢀ2.8
11.3
0.0
16.0
0.0
0.0
15.3
1.4
IN15 ꢀ7.3 ꢀ3.7 ꢀ12.5
Figure 3. Structure of g-allenol 17 as the selected theoretical model for
TS16 3.9 6.5
13.7
13.6
13.1
16.2
computational studies.
IN16 ꢀ21.9 ꢀ18.5 ꢀ26.5 ꢀ8.3 ꢀ5.5 ꢀ6.9 ꢀ17.7 ꢀ16.0 ꢀ10.4
TS17
5.1
6.6
2.2 11.3
5.8
14.2
7.8
13.5
2.7
12.2
12.3
2.3
13.9
2.6
IN17 ꢀ4.9 ꢀ0.9 ꢀ11.2
ꢀ0.4
[a] Values measured in kcalmolꢀ1
.
The coordination of an allene to a metal may lead to sev-
eral types of structures. For the model system 17, two h2
complexes involving one of the C=C bonds were found, 18a
and 18b (Figure 3), the properties of which depend on the
substitution pattern of the allene.[17] Thus, the p coordina-
tion of the proximal C=C bond leads to a partially h1-slipped
reactant complex to reduce steric repulsion with the methyl
ꢀ
substituent, as the M C distances and computed Wiberg
bond indices (WBIs) suggest (Table 1). In contrast, the coor-
dination of the distal double bond leads to an almost sym-
metric h2 structure for the three catalysts, which also show
higher WBIs, that is, more significant bonding and globally
ꢀ
shorter M C lengths. These results point to a stronger coor-
dination and more stable complexes, which is supported by
the calculated free-energy difference between both types of
coordination modes. The data summarized in Table 1 also
suggest that PdCl2 generates a more electron-deficient
allene fragment, probably due to the lack of a trans ligand
and, hence, back-bonding donation from the metal.
Alternative conformations of the starting complexes, such
as s-allylic cation structures,[17] could not be optimized as
minima but were obtained as transition states. Therefore we
can state that the ground state for the complexes is the form
18b. The similar trends observed for the catalytic systems
suggest the same activation mode of the allene moiety upon
complexation with the metal. The free-energy differences
computed for the intramolecular nucleophilic addition to
the allene following the plausible heterocyclization paths
are depicted in Table 2 and Figure 4 (for AuCl3).
Figure 4. Optimized structures of the transition states and intermediates
for the AuCl3 heterocyclization step following alternative paths.
9130
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9127 – 9138

Enantiopure Dihydropyrans and Tetrahydrooxepines
FULL PAPER
Table 3. Dihedral angle[a] O(H)-C-C-C for the first step. The values for
the AuCl3-mediated reaction of the b-lactamic structure are also shown
for comparative purposes.
(TS16), which in turn proceeds with a higher barrier than
the 7-endo-trig cyclization (TS17). Furthermore, this tenden-
cy is systematically shown by the three catalyst systems. This
kinetic preference sharply contrasts with that estimated for
the precursors bearing a b-lactam ring as the tether part be-
cause the energy of activation increases progressively with
the size of the ring and the formation of the fused tetrahy-
drofuran scaffold is clearly favored. From a thermodynamic
viewpoint, the formation of the dihydropyran intermediate
(IN16) is more exothermic than the formation of the tetrahy-
drofuran (IN15) and -oxepine (IN17) intermediates. These
divergences between the two kinds of precursors can be
easily understood on close inspection of the transition-state
AuCl3
AuCl3
G
G
PdCl2
TS15
IN15
TS16
IN16
TS17
IN17
0.3
3.7
7.3
1.0
21.2
2.5
58.8
31.3
71.0
63.7
64.9
57.8
64.2
32.2
70.2
64.0
61.2
56.9
61.3
31.4
71.3
63.4
56.0
57.6
[a] Measured in degrees.
ꢀ ꢀ ꢀ
structures. If we focus on the dihedral angle O(H) C C C
for the three alternative cyclization reactions with the three
catalysts and compare them with the b-lactam precursor, we
can observe critical differences (Table 3). The restraint im-
posed by the b-lactam forces an eclipsed conformation for
both the transition state and intermediate fused structures,
whereas for the acyclic precursor the transition state can be
reached through a lower-energy staggered conformation be-
cause of the improved tether flexibility. This effect not only
drastically reduces the energy barrier for the cycloetherifica-
tion reactions but also the 6-exo-dig and 7-endo-trig cycliza-
tions become the kinetically preferred pathways. In addition,
TS16 and TS17 evolve to the metalla–alkenyl intermediates
IN16 and IN17, respectively, in which the dihydropyran and
tetrahydrooxepine rings formed adopt the lowest-energy
conformation, that is, half-chair[18] and chair,[19] respectively.
In summary, the calculated
uct.[20] The formal 1,3-H migration is therefore assisted by
the catalyst and takes place in two steps: The formation of
ꢀ
HCl and cleavage of the M C bond by protonolysis with
HCl to liberate the cycloadduct and regenerate the active
catalyst. The first is a barrierless step when thermal correc-
tions to the energy are taken into account and yields a tran-
sient intermediate structure (IN2n; for instance, for AuCl3-
catalyzed processes, see Figure 5) in which the HCl formed
remains weakly coordinated to the metal, thus promoting
the last step.
The final step proceeds through the early four-membered-
ring transition-state structure TS2n, which still exhibits short
ꢀ
ꢀ
ꢀ
M C and Cl H breaking bonds and a large C H forming
energies indicate that the 7-
endo cycloetherification is ki-
netically favored over alterna-
tive cyclization modes, al-
though TS16 formed by the 6-
exo reaction is only slightly
less stable than TS17. Even if
these data agree with experi-
mental evidence, the small
energy difference between
TS16 and TS17 does not justify
the manifest regioselectivity.
Therefore, we examined the
full reaction profiles for all the
catalytic systems.
The alkenyl–metal inter-
mediates IN1n show the forma-
tion of a hydrogen bond be-
tween one of the chloride li-
gands and the acidic hydroxylic
hydrogen (1.75–1.85, 1.75–1.86,
and 1.76–1.77 for AuCl3,
[PtCl2ACHTUNGTRENNUNG(CH2CH2)], and PdCl2,
respectively), which promotes
the proton shift and protonoly-
Figure 5. Free-energy profile [kcalmolꢀ1] for the transformation of g-allenol 18b-Au into the complexed tetra-
hydrooxepine P7-Au (black line). Formation of the alternative five- (blue) and six-membered-ring (red) cyclo-
alkenes are shown for comparison.
ꢀ
sis of the s-carbon metal bond
to afford the cyclized prod-
Chem. Eur. J. 2009, 15, 9127 – 9138
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9131

B. Alcaide, P. Almendros et al.
Table 4. Key structural parameters[a] for the protonolysis step.
Likewise, the transition-state
structure for the rate-limiting
protonolysis step for the seven-
membered TS27 is more stable
than those for the dihydropyr-
an (TS26) and tetrahydrofuran
(TS25) cycloadducts (Table 5).
One might expect, therefore,
the favored formation of the
same oxacycle P7. However,
AuCl3
G
N
PdCl2
ꢀ
Cl H
ꢀ
M Cl
ꢀ
C H
ꢀ
Cl H
ꢀ
M Cl
ꢀ
ꢀ
ꢀ
ꢀ
C H
Cl H
M Cl
C H
TS25
P5
TS26
P6
TS27
P7
1.592
2.652
1.567
3.372
1.506
2.686
2.516
2.394
2.534
2.370
2.550
2.390
1.435
1.088
1.508
1.096
1.535
1.090
1.594
2.907
1.529
3.339
1.498
2.814
2.501
2.376
2.502
2.364
1.514
2.386
1.458
1.087
1.626
1.096
1.602
1.091
1.607
3.021
1.654
3.215
1.547
2.812
2.493
2.335
2.431
2.318
2.502
2.330
1.401
1.085
1.470
1.095
1.469
1.089
[a] Measured in ꢄ.
bond (Table 4). This step is slightly exothermic whatever the
catalyst. The free-energy barrier required to reach the tran-
sition state for the seven-membered-ring intermediate TS27
is lower than for the six-membered-ring counterpart TS26
(Table 5). That is, TS27 is more stable than TS26, probably
Table 5. Free-energy differences[a] relative to the starting complex 18b
for the protonolysis step.
AuCl3
PtCl
G
PdCl2
n=5 n=6 n=7 n=5 n=6 n=7 n=5 n=6 n=7
TS2n
12.3 9.1 8.6 22.3 18.5 17.9 19.2 15.5 14.7
Product ꢀ17.6 ꢀ27.0 ꢀ10.6 ꢀ9.4 ꢀ26.9 ꢀ6.9 ꢀ2.5 ꢀ12.1 ꢀ3.5
[a] Measured in kcalmolꢀ1
.
Scheme 7. Mechanistic explanation for the AuIII-catalyzed hydroalkoxyla-
tion reaction of g-allenol 17.
2
ꢀ
because a stronger s-Csp H bond is being formed. In this
context, the high barrier computed for the protonolysis of
the vinyl–metal intermediates IN15 (TS25, Table 5) in com-
parison with the b-lactam counterpart is remarkable. To
check whether this is due to steric repulsions with the bulky
ester, we estimated the activation barrier for its epimer
IN15’ that lacks this effect. The pertinent transition-state
structure TS25’ is even less stable (0.9 kcalmolꢀ1) than TS25
because of steric interactions with endocyclic protons in a
puckered, highly functionalized tetrahydrofuran ring.
experimental results showed that product 13 is not detected
in the platinum-catalyzed reaction of g-allenol 2 (Scheme 6),
its isomer 14 being the sole reaction product. Conjugation
of the double bond with the lone pair of the oxygen atom
only under the conditions of platinum catalysis might be
evoked to account for the formation of this structure. Thus
we question the role of the catalyst in this divergence.
The full free-energy profiles for the formation of the tet-
rahydrofuran, dihydropyran, and tetrahydrooxepine frame-
works are depicted together in Figure 5 for comparative
purposes. The data are in good agreement with experimental
observations as they indicate the kinetically favored forma-
tion of the tetrahydrooxepine skeleton.
All these data indicate that a possible pathway for the
AuCl3-catalyzed formation of the tetrahydrooxepine deriva-
tive from g-allenol 17 might initially involve the formation
of complex 18b by coordination of the gold trichloride to
the distal allenic double bond. A regioselective 7-endo oxy-
auration by nucleophilic addition to the terminal allenic
carbon atom forms the alkenyl–Au intermediate IN17. The
subsequent loss of HCl and rate-limiting protonolysis of the
Returning to our theoretical model, the transformation of
P7 into P7’ involves a formal 1,3-hydrogen migration. This
process can be viewed as two consecutive 1,2-H shift steps
rather than a direct 1,3-H shift (Figure 6). In this context we
have found a rather synchronous transition-state structure
ꢀ
ꢀ
(C H breaking bond=1.406 ꢄ; C H forming bond=
1.470 ꢄ), TS37, in which the moving atom establishes a hy-
drogen bond with the catalyst (1.849 ꢄ). This effect stabiliz-
es the transition-state structure giving rise to a moderate ac-
tivation barrier (17.3 kcalmolꢀ1) to be overcome.[21] The per-
tinent IRC calculations have verified that TS37 connects P7
with IN37 in an endothermic, reversible step (8.8 kcalmolꢀ1).
Finally, we have proposed that the last 1,2-H migration may
be a stepwise process assisted by the metal center by a b-H
elimination. Mechanistic studies on transition-metal-cata-
lyzed b-H elimination processes[22] have revealed that they
can proceed through four-membered-ring intermediate
structures[23] in which hydrogen coordinates to a metal
ꢀ
carbon gold bond provides the tetrahydrooxepine and re-
generates the gold catalyst (Scheme 7).
The results computed for the plausible PtII-catalyzed het-
erocyclization reaction of allenol 17 show the same trend as
that seen for the AuIII-catalyzed process, namely, a kinetic
center before b-C H bond dissociation takes place. Never-
ꢀ
2
ꢀ
preference for nucleophilic addition to the terminal Csp
theless, the long Pt···H distance (2.945 ꢄ) suggests that IN37
allene carbon (7-endo-trig heterocyclization mode, Table 2).
should not be a marked b-agostic structure.[24] The transi-
9132
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9127 – 9138

Enantiopure Dihydropyrans and Tetrahydrooxepines
FULL PAPER
confirmed by the IRC analysis. This transition-state struc-
ture shows the hydrogen approaching the carbon atom
ꢀ ꢀ
through the reduction of the C Pt H bond angle (from
ꢀ
83.78 in IN47 to 25.98). Note that a weak Pt O interaction
between the catalyst and the benzylic oxygen atom is detect-
ed in the stepwise 1,2-H shift (2.994, 2.723, and 2.367 ꢄ for
IN47, TS57, and P7’, respectively) due to the oxophilic nature
of the platinum, which likely stabilizes TS57 and accounts
for the stability of the complexed P7’ and, amazingly, the
chirality at the new stereocenter. This step is exothermic
and the barrier to attaining TS57 is low (4.4 kcalmolꢀ1 from
IN47). Otherwise, the uncatalyzed 1,2-shift from IN37 has
been estimated to proceed through a high-energy transition-
state structure (TS67) lying 22.2 and 28.3 kcalmolꢀ1 above
TS47 and TS57, respectively. These data support the stepwise
1,2-H migration assisted by the metal complex.
Figure 7 outlines the free-energy profile for the transfor-
mation of g-allenol 17 into the tetrahydrooxepine P7’ (cyclic
ether of type 14, Scheme 6).
According to these results, a likely pathway for the forma-
tion of tetrahydrooxepine of type 14 (Scheme 6) initially in-
volves the formation of p complex 18b through the coordi-
nation of the metal to the distal diene moiety of g-allenol.
Next, a kinetically preferred 7-endo heterocyclization to
form species IN17 is followed by a 1,3 hydrogen shift to
afford the p-complexed oxacycle P7. This tetrahydrooxepine
scaffold isomerizes to P7’ through a 1,2-hydrogen migration,
assisted by a halide ligand, a platinum-catalyzed b-hydrogen
elimination to generate the platinahydride intermediate
IN47, and a final protonolysis step (Scheme 8). The P7!P7’
isomerization is not observed under AuCl3 catalysis proba-
bly because gold shows no tendency to undergo b-hydride
elimination reactions. Indeed, gold hydrides are rare species
and difficult to access.[25] Thus, even if the formation of IN37
is possible, as long as it is reversible and the subsequent b-H
elimination is blocked, the reaction stops once P7 has been
generated.
Figure 6. Optimized structures of the stepwise hydrogen shift involved in
the isomerization of P7 to P7’. A three-step platinum-mediated mecha-
nism is favored over the alternative uncatalyzed two-step mechansim
through TS67. Most of the hydrogen atoms have been omitted for the
sake of clarity.
tion-state structure involved in this step, TS47, exhibits a
ꢀ
ꢀ
lengthening of the Pt C (2.304 ꢄ) and C H (1.201 ꢄ)
ꢀ
breaking bonds and shortening of the Pt H (2.179 ꢄ) and
ꢀ
C C (1.449 ꢄ) bonds. A detailed examination of the imagi-
nary frequency mode confirms that the most important geo-
metric change takes place on the hydrogen atom between
the carbon atom and the metal center. TS47 then evolves to
the platinahydride intermediate IN47 in which complete mi-
ꢀ
gration of the hydrogen from the methylene moiety (C H
ꢀ
3.113 ꢄ) to platinum (Pt H 1.535 ꢄ) has taken place.
The last step could therefore drive the isomer P7’ of the
IN47 intermediate through the transition state TS57, as was
Figure 7. Free-energy profile [kcalmolꢀ1] for the platinum-catalyzed transformation of g-allenol 17 (shown as the starting reactant complex 18b-Pt) into
the tetrahydrooxepine P7’ (shown as the complexed product P7’-Pt).
Chem. Eur. J. 2009, 15, 9127 – 9138
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9133

B. Alcaide, P. Almendros et al.
membered cycloadduct TS27 is about 1 and 4.5 kcalmolꢀ1
lower than the corresponding transition-state energies for
the dihydropyran (TS26) and tetrahydrofuran (TS25) scaf-
folds, respectively. However, the presence of an allyl halide
alternatively promotes a coupling reaction by trapping the
kinetically preferred alkenyl–Pd intermediate IN7-Pd. This
process is favored by the easy release of HCl from the coor-
ꢀ
dination to the metal center (Pd Cl(H)=2.592 ꢄ in IN27-
Pd vs. 2.487 ꢄ in IN17-Pd). As noted above, the cyclization
ꢀ
intermediates IN1n show the shortest H Cl distances with
palladium as the catalyst. The allyl coupling with the alken-
yl–PdII intermediate occurs through a three-step mechanism:
1) Ligand displacement from the metal coordination sphere,
ꢀ
2) insertion into the allylic halide C=C bond to give a s-C
Pd intermediate, and 3) trans b-elimination to afford the ox-
epane product (Figure 8). The palladium-coordinated HCl is
easily displaced by the incoming allyl bromide in a fast
ligand-interchange displacement mechanism to yield the h2
complex IN1Al upon p coordination of the C=C double
bond to the metal. This coordination gives rise to almost
Scheme 8. Mechanistic explanation for the PtII-catalyzed hydroalkoxyla-
tion reaction of g-allenol 17.
ꢀ
ꢀ
symmetrical Pd–alkene bonds (Pd C(H2) 2.221 ꢄ and Pd
C(H) 2.252 ꢄ) and a lengthened C=C bond (Dd=0.049 ꢄ
on going from the uncoordinated precursor to the p com-
plex IN1Al). This step is exothermic (by 11.2 kcalmolꢀ1) and
has a low activation barrier (5.3 kcalmolꢀ1).
The palladium-catalyzed heterocyclization reaction also
shows the same regioselectivity as that observed under gold
and platinum catalysis, that is, the initial 7-endo-trig cycliza-
tion takes place preferentially by nucleophilic addition of
the hydroxylic oxygen to the activated terminal allene
carbon (Table 2). In the same way, the energy of the transi-
tion-state structure for the protonolysis step for the seven-
The coordinated alkene undergoes a 2,1-insertion into the
[26]
ꢀ
Pd alkyl bond in a stepwise process that proceeds via the
2
ꢀ
formation of a h -Pd complex (Pd C 2.203 and 2.282 ꢄ),
IN2Al. This intermediate is formed via the four-membered-
ring TS1Al in which the four atoms forming the new bonds
Figure 8. Free-energy profile [kcalmolꢀ1] for the palladium-catalyzed transformation of g-allenol 17 (shown as the starting reactant complex 18b-Pd) into
the tetrahydrooxepine of type 15 (shown as the complexed product). The formation of the corresponding bicycle P7 by protonolysis of the intermediate
IN27 is shown for comparison.
9134
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9127 – 9138

Enantiopure Dihydropyrans and Tetrahydrooxepines
FULL PAPER
ꢀ
ꢀ
(Pd C 2.077 ꢄ and C C 2.133 ꢄ) are roughly planar (devia-
tion of 7.48). A cis/trans isomerization of the chloride ligand
takes place to reach the transition state, probably to reduce
the back-bonding interaction and to favor the formation of
The proposed mechanism is summarized in Scheme 9. An
initial coordination of the distal allene bond of the precursor
to the catalyst provides the allenepalladium complex 18b-
PdCl2. This starting complex undergoes an intramolecular
cycloetherification reaction to give the palladatetrahydroox-
epine IN17, which readily releases HCl to provide IN27. A
subsequent displacement from the metal coordination
sphere of HCl by allyl bromide gives intermediate IN1Al,
which, after insertion of the allylic chain and trans b-heter-
oatom elimination, generates tetrahydrooxepine of the type
15 (Scheme 6) with concomitant regeneration of the palladi-
um catalyst.
ꢀ
the Pd C bond. Likewise, a moderate activation barrier is
found for the insertion of the allylic bromide into the Pd C
ꢀ
bond (13.6 kcalmolꢀ1), the formation of IN2Al being favored
from a thermodynamic viewpoint (ꢀ10.6 kcalmolꢀ1). The h2-
Pd complex then suffers a b-heteroatom elimination[27,28] to
give the coupling product and the active catalyst PdCl2. As
has been noted, the liberated HCl plays a key role in pro-
moting the dehalopalladation and inhibiting the usual b-H
elimination.[27g,29] Lu and co-workers have postulated that
halide ions would assist the b-heteroatom elimination
through an E2-like mechanism promoted by halide ion coor-
dination to palladium.[30] Accordingly, the trans b-elimina-
tion step takes place via TS2Al, which exhibits an advanced
Conclusions
Regiocontrolled metal-catalyzed heterocyclization reactions
of b- and g-allenols derived from d-glyceraldehyde leading
to a variety of enantiopure dihydropyrans and tetrahydroox-
epines have been performed. In addition, density functional
theory (DFT) calculations were performed to obtain an in-
sight into various aspects of the reactivity of g-allenols. The
calculations predicted both a metal- and a tether-dependent
heterocyclization for g-allenols. These theoretical predic-
tions have been fully confirmed by the experimental results.
Currently we are directing our efforts towards expanding
the scope and improving the efficiency of these mild gold-,
platinum-, and palladium-catalyzed allene cycloetherifica-
tion protocols.
ꢀ
tetracycle opening (forming bond Pd Cl 2.466 ꢄ and break-
ꢀ
ꢀ ꢀ
ing bond C Br=2.097 ꢄ, opening of the bond angle C C
C=115.68). This transition-state structure finally forms the
dienic product. The b-dehalopalladation proceeds in an exo-
thermic step by overcoming
a low activation barrier
(4.8 kcalmolꢀ1).
The free-energy profile is shown in Figure 8 and reveals
features similar to those computed for 2-azetidinone-teth-
ered methyl-g-allenols. Thus, the palladium-catalyzed cycli-
zative coupling reaction between g-allenols and allyl halides
is favored kinetically and thermodynamically over the cycli-
zation and proceeds by a common 7-endo-oxypalladation
followed by a stepwise energy-downhill coupling with the
alkene.
Experimental Section
General methods: 1H and 13C NMR spectra were recorded on a Bruker
AMX-500, Bruker Avance-300, Varian VRX-300S, or Bruker AC-200
spectrometer. NMR spectra were recorded in CDCl3 solutions unless oth-
erwise stated. Chemical shifts are given in ppm relative to TMS (1H, d=
0.0 ppm) or CDCl3 (13C, d=76.9 ppm). Low- and high-resolution mass
spectra were recorded on an Agilent 6520 Accurate-Mass QTOF LC/MS
spectrometer using the electronic impact (EI) or electrospray modes
(ES) unless otherwise stated. The specific rotations ([a]D) are given in
10ꢀ1 degcm2 gꢀ1 at 208C and the concentrations (c) are expressed in g per
100 mL. All commercially available compounds were used without fur-
ther purification. IR spectra were recorded on a Perkin-Elmer 781 spec-
trometer.
PtII-catalyzed cyclization of b-allenol (ꢀ)-1a—Preparation of furan 10:
[PtCl2ACHTUNTRGNE(GNU CH2=H2)]2 (1.05 mg, 0.0018 mmol) and tris(2,6-dimethoxyphenyl)-
phosphine (0.0036 mmol) were added to a stirred solution of the b-allenol
(ꢀ)-1a (70 mg, 0.18 mmol) in dichloromethane (0.2 mL) under argon.
The resulting mixture was stirred at room temperature for 4 h. The reac-
tion was then quenched with brine (0.2 mL), the mixture was extracted
with ethyl acetate (3ꢅ4 mL), and the combined extracts were washed
twice with brine. The organic layer was dried (MgSO4) and concentrated
under reduced pressure. After flash chromatography of the residue on
silica gel eluting with a hexanes/ethyl acetate (8:1) mixture, analytically
pure compound 10 (32 mg, 47%) was obtained as a colorless oil.[37]
1H NMR (300 MHz, CDCl3, 258C): d=7.71 and 7.40 (m, each 5H), 5.90
(s, 1H), 4.55 (s, 2H), 2.17 (s, 3H), 1.90 (s, 3H), 1.06 ppm (s, 9H);
13C NMR (75 MHz, CDCl3, 258C): d=150.9, 148.5, 135.7, 133.6, 130.0,
127.6, 115.4, 110.7, 58.9, 26.8, 11.3, 9.8 ppm; MS (ES): m/z (%): 365 (100)
[M+H]+, 364 (17) [M]+; elemental analysis calcd (%) for C23H28O2Si
(364.5): C 75.78, H 7.74; found: C 75.91, H 7.67.
Scheme 9. Mechanistic explanation for the PdII-catalyzed heterocycliza-
tion reaction of g-allenol 17.
Chem. Eur. J. 2009, 15, 9127 – 9138
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9135

B. Alcaide, P. Almendros et al.
PdII-catalyzed oxybromination of b-allenols 1—Preparation of dihydro-
pyrans 11: Palladium(II) acetate (0.01 mmol), lithium bromide
(0.74 mmol), potassium carbonate (0.18 mmol), and copper(II) acetate
(0.32 mmol) were sequentially added to a stirred solution of the corre-
sponding b-allenol 1 (57 mg, 0.15 mmol) in acetonitrile (5 mL). The re-
sulting suspension was stirred at room temperature under an oxygen at-
mosphere for 2 h. The organic phase was diluted with brine (2 mL), ex-
tracted with ethyl acetate (3ꢅ5 mL), washed with brine (2 mL), dried
(MgSO4), and concentrated under reduced pressure. After flash chroma-
tography of the residue on silica gel eluting with a hexanes/ethyl acetate
(9:1) mixture, analytically pure compound 11 (10 mg, 10%) was obtained
as a colorless oil. [a]D =+21.9 (c=0.6 in CHCl3); 1H NMR (300 MHz,
CDCl3, 258C): d=7.95 and 6.93 (d, J=9.0 Hz, each 2H), 7.73 (m, 5H),
7.41 (m, 5H), 5.73 (m, 1H), 4.28 (dd, J=4.1, 2.0 Hz, 2H), 3.89 (s, 3H),
3.84 (m, 3H), 1.80 (brs, 3H), 1.08 ppm (s, 9H); IR (CHCl3): n˜ =
1725 cmꢀ1; MS (ES): m/z (%): 596 (100) [M+2+H]+, 594 (98) [M+H]+;
elemental analysis calcd (%) for C31H35BrO5Si (595.6): C 62.51, H 5.92;
found: C 62.63, H 5.96.
AuIII-catalyzed cyclization of g-allenol (ꢀ)-2—Preparation of tetrahy-
drooxepine (+)-13: AuCl3 (2.5 mg, 0.008 mmol) was added to a stirred
solution of the g-allenol (ꢀ)-2 (50 mg, 0.16 mmol) in dichloromethane
(160 mL) under argon. The resulting mixture was stirred at room temper-
ature for 3 h. The reaction was then quenched with brine (160 mL), the
mixture was extracted with ethyl acetate (3ꢅ1 mL), and the combined
extracts were washed twice with brine. The organic layer was dried
(MgSO4) and concentrated under reduced pressure. After flash chroma-
tography of the residue on silica gel eluting with a hexanes/ethyl acetate
(3:1) mixture, analytically pure compound 13 (34 mg, 67%) was obtained
as a colorless oil. [a]D =+5.4 (c=1.9 in CHCl3); 1H NMR (300 MHz,
CDCl3, 258C): d=8.05 and 6.93 (d, J=9.0 Hz, each 2H), 5.63 (m, 1H),
5.44 (m, 1H), 4.89 (m, 1H), 4.64 (m, 2H), 4.37 (m, 2H), 3.87 (s, 3H),
2.01 (s, 3H), 1.85 ppm (brs, 3H); 13C NMR (75 MHz, CDCl3, 258C): d=
170.8, 165.7, 163.6, 134.7, 131.9, 123.0, 122.3, 113.6, 87.3, 75.5, 73.1, 62.2,
55.4, 21.1, 12.7 ppm; IR (CHCl3): n˜ =1744, 1725 cmꢀ1; MS (ES): m/z (%):
321 (100) [M+H]+, 320 (9) [M]+; elemental analysis calcd (%) for
C17H20O6 (320.3): C 63.74, H 6.29; found: C 63.62, H 6.23.
PtII-catalyzed cyclization of g-allenol (ꢀ)-2—Preparation of tetrahy-
General procedure for the PdII-catalyzed coupling reaction of b-allenols
drooxepine (+)-14: [PtCl2ACHTNUGTRNE(GNU CH2=CH2)]2 (0.0015 mmol) was added to a
1
with allyl bromide—Preparation of dihydropyrans 12: PdCl2
stirred solution of the g-allenol (ꢀ)-2 (48 mg, 0.15 mmol) in dichlorome-
thane (150 mL) under argon. The resulting mixture was stirred at room
temperature for 2 h. The reaction was then quenched with brine
(150 mL), the mixture was extracted with ethyl acetate (3ꢅ1 mL), and the
combined extracts were washed twice with brine. The organic layer was
dried (MgSO4) and concentrated under reduced pressure. After flash
chromatography of the residue on silica gel eluting with a hexanes/ethyl
acetate (4:1) mixture, analytically pure compound 14 (37 mg, 77%) was
obtained as a colorless oil. [a]D =+6.2 (c=0.5 in CHCl3); 1H NMR
(300 MHz, CDCl3, 258C): d=8.02 and 6.94 (d, J=9.0 Hz, each 2H), 5.63
(m, 1H), 6.27 (m, 1H), 5.33 (dt, J=6.3, 2.9 Hz, 1H), 4.92 (t, J=2.7 Hz,
1H), 4.43 (m, 3H), 3.88 (s, 3H), 2.95 (m, 1H), 2.05 (s, 3H), 1.15 ppm (d,
J=6.8 Hz, 3H); 13C NMR (75 MHz, CDCl3, 258C): d=165.5, 163.6,
163.5, 144.5, 131.7, 122.1, 113.7, 105.9, 86.1, 72.5, 62.6, 55.5, 39.6,
20.7 ppm; IR (CHCl3): n˜ =1742, 1722 cmꢀ1; MS (ES): m/z (%): 321 (100)
[M+H]+, 320 (3) [M]+; elemental analysis calcd (%) for C17H20O6
(320.3): C 63.74, H 6.29; found: C 63.87, H 6.35.
PdII-catalyzed coupling reaction of g-allenol (ꢀ)-2 with allyl bromide—
Preparation of tetrahydrooxepine (+)-15: PdCl2 (0.0075 mmol) was
added to a stirred solution of the g-allenol (ꢀ)-2 (48 mg, 0.15 mmol) and
allyl bromide (0.75 mmol) in N,N-dimethylformamide (0.9 mL). The reac-
tion was stirred under argon for 2 h. Water (0.8 mL) was added before
being extracted with ethyl acetate (3ꢅ5 mL). The organic phase was
washed with water (2ꢅ2 mL), dried (MgSO4), and concentrated under
reduced pressure. After flash chromatography of the residue on silica gel
eluting with a hexanes/ethyl acetate (7:1) mixture, analytically pure com-
pound 15 (30 mg, 56%) was obtained as a colorless oil. [a]D =+23.9 (c=
0.8 in CHCl3); 1H NMR (300 MHz, CDCl3, 258C): d=8.05 and 6.93 (d,
J=9.0 Hz, each 2H), 5.73 (m, 1H), 5.44 (m, 1H), 5.06 (m, 1H), 4.96 (m,
1H), 4.58 (m, 2H), 4.35 (m, 2H), 3.87 (s, 3H), 2.85 (t, J=5.5 Hz, 2H),
2.01 (s, 3H), 1.75 ppm (brs, 3H); 13C NMR (75 MHz, CDCl3, 258C): d=
169.6, 166.2, 165.6, 134.1, 132.6, 131.6, 122.3, 116.3, 113.8, 113.6, 89.0,
73.4, 73.0, 62.2, 55.4, 39.1, 29.4, 20.7 ppm; IR (CHCl3): n˜ =1744,
1724 cmꢀ1; MS (ES): m/z (%): 361 (100) [M+H]+, 360 (9) [M]+; elemen-
tal analysis calcd (%) for C20H24O6 (360.4): C 66.65, H 6.71; found: C
66.78, H 6.77.
(0.005 mmol) was added to a stirred solution of the corresponding b-alle-
nol 1 (0.10 mmol) and allyl bromide (0.50 mmol) in N,N-dimethylforma-
mide (0.6 mL). The reaction was stirred under an argon atmosphere until
disappearance of the starting material (TLC). Water (0.5 mL) was added
before being extracted with ethyl acetate (3ꢅ4 mL). The organic phase
was washed with water (2ꢅ2 mL), dried (MgSO4), and concentrated
under reduced pressure. Chromatography of the residue eluting with hex-
anes/ethyl acetate mixtures gave analytically pure dihydropyrans 12.
Dihydropyran (+)-12a: From b-allenol (ꢀ)-1a (50 mg, 0.13 mmol), and
after chromatography of the residue using hexanes/ethyl acetate (3:1) as
eluent, compound (+)-12a (58 mg, 80%) was obtained as a colorless oil.
[a]D =+18.4 (c=2.5 in CHCl3); 1H NMR (300 MHz, CDCl3, 258C): d=
7.96 and 6.92 (d, J=9.0 Hz, each 2H), 7.68 (m, 5H), 7.36 (m, 5H), 5.72
(m, 1H), 5.56 (m, 1H), 5.09 (dd, J=9.0, 1.6 Hz, 1H), 5.01 (m, 1H), 4.08
(m, 2H), 3.89 (s, 3H), 3.79 (m, 3H), 2.74 (m, 2H), 1.63 (brs, 3H),
1.08 ppm (s, 9H); 13C NMR (75 MHz, CDCl3, 258C): d=166.1, 157.0,
135.6, 135.2, 134.8, 134.2, 133.4, 131.8, 129.6, 129.5, 127.7, 127.5, 125.2,
121.2, 115.9, 113.6, 76.9, 68.7, 66.9, 63.9, 55.5, 33.3, 26.6, 26.5, 19.0 ppm;
IR (CHCl3): n˜ =1744 cmꢀ1; MS (ES): m/z (%): 557 (100) [M+H]+, 556
(7) [M]+; elemental analysis calcd (%) for C34H40O5Si (556.8): C 73.35, H
7.24; found: C 73.48, H 7.18.
Preparation of dihydropyrans (+)-12b and (ꢀ)-12c: From b-allenol (+)-
1b (75 mg, 0.16 mmol), and after chromatography of the residue using
hexanes/ethyl acetate (8:1) as eluent, 25 mg (32%) of the less polar com-
pound (+)-12b and 19 mg (31%) of the more polar compound (ꢀ)-12c
were obtained.
Dihydropyran (+)-12b: [a]D =+15.7 (c=1.1 in CHCl3); 1H NMR
(300 MHz, CDCl3, 258C): d=7.78 and 6.82 (d, J=9.0 Hz, each 2H), 7.20
(m, 5H), 5.80 (m, 1H), 5.67 (m, 1H), 5.07 (m, 1H), 5.00 (m, 1H), 4.32
(d, J=2.2 Hz, 1H), 4.28 (d, J=1.7 Hz, 1H), 3.90 (m, 3H), 3.83 (s, 3H),
2.69 (d, J=6.1 Hz, 2H), 0.89 (s, 9H), 0.08 and 0.05 ppm (s, each 3H);
13C NMR (75 MHz, CDCl3, 258C): d=165.5, 163.2, 137.3, 135.4, 135.1,
131.8, 131.5, 128.9, 128.0, 127.1, 122.4, 116.5, 113.4, 78.1, 68.0, 66.8, 63.6,
55.3, 34.6, 25.9, 18.4, ꢀ5.2, ꢀ5.3 ppm; IR (CHCl3): n˜ =1743 cmꢀ1; MS
(ES): m/z (%): 495 (100) [M+H]+, 494 (9) [M]+; elemental analysis
calcd (%) for C29H38O5Si (494.3): C 70.41, H 7.74; found: C 70.28, H
7.81.
Dihydropyran (ꢀ)-12c: [a]D =ꢀ10.3 (c=0.4 in CHCl3); 1H NMR
(300 MHz, CDCl3, 258C): d=7.69 and 6.81 (d, J=9.0 Hz, each 2H), 7.22
(m, 5H), 5.93 (m, 1H), 5.69 (m, 1H), 5.08 (m, 1H), 5.01 (m, 1H), 4.41
(dd, J=16.3, 2.6 Hz, 1H), 4.27 (dd, J=16.4, 2.0 Hz, 1H), 3.83 (m, 3H),
3.82 (s, 3H), 2.72 ppm (d, J=7.1 Hz, 2H); 13C NMR (75 MHz, CDCl3,
258C): d=163.5, 160.3, 136.8, 134.9, 134.8, 131.8, 131.6, 128.7, 128.1,
127.2, 121.8, 116.6, 113.7, 113.5, 77.4, 67.7, 67.5, 62.3, 55.4, 34.5 ppm; IR
(CHCl3): n˜ =3432, 1743 cmꢀ1; MS (ES): m/z (%): 381 (100) [M+H]+, 380
(17) [M]+; elemental analysis calcd (%) for C23H24O5 (380.4): C 72.61, H
6.36; found: C 72.74, H 6.31.
PdII-catalyzed oxybromination of g-allenol (ꢀ)-2—Preparation of tetra-
hydrooxepine (+)-16: Palladium(II) acetate (0.01 mmol), lithium bro-
mide (0.74 mmol), potassium carbonate (0.18 mmol), and copper(II) ace-
tate (0.32 mmol) were sequentially added to a stirred solution of the g-al-
lenol (ꢀ)-2 (48 mg, 0.15 mmol) in acetonitrile (5 mL). The resulting sus-
pension was stirred at room temperature under an oxygen atmosphere
for 9 h. The organic phase was diluted with brine (2 mL), extracted with
ethyl acetate (3ꢅ5 mL), washed with brine (2 mL), dried (MgSO4), and
concentrated under reduced pressure. After flash chromatography of the
residue on silica gel eluting with a hexanes/ethyl acetate (3:1) mixture,
analytically pure compound 16 (31 mg, 51%) was obtained as a colorless
oil. [a]D =+1.3 (c=1.2 in CHCl3); 1H NMR (300 MHz, CDCl3, 258C):
9136
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9127 – 9138

Enantiopure Dihydropyrans and Tetrahydrooxepines
FULL PAPER
d=8.03 and 6.94 (d, J=9.0 Hz, each 2H), 5.43 (m, 1H), 4.92 (m, 1H),
4.64 (m, 2H), 4.37 (m, 2H), 3.88 (s, 3H), 2.03 (s, 3H), 1.84 ppm (m, 3H);
13C NMR (75 MHz, CDCl3, 258C): d=170.8, 163.7, 163.6, 132.5, 131.9,
131.7, 121.9, 113.7, 87.3, 78.1, 72.4, 61.8, 55.5, 29.7, 20.7 ppm; IR (CHCl3):
n˜ =1741, 1723 cmꢀ1; MS (ES): m/z (%): 400 (100) [M+2+H]+, 398 (98)
[M+H]+; elemental analysis calcd (%) for C17H19BrO6 (399.2): C 51.14,
H 4.80; found: C 51.02, H 4.85.
catalyzed cyclizative coupling reactions of b-allenols, see: d) S. Ma,
W. Gao, Synlett 2002, 0065; for ruthenium-catalyzed cyclocarbonyla-
tion of b-allenols, see: e) E. Yoneda, S.-W. Zhang, D.-Y. Zhou, K.
Onitsuka, S. Takahashi, J. Org. Chem. 2003, 68, 8571; for gold-cata-
lyzed cyclizations of g-allenols, see: f) Z. Zhang, R. A. Widenhoefer,
Angew. Chem. 2007, 119, 287; Angew. Chem. Int. Ed. 2007, 46, 283;
for gold- and platinum-mediated oxycyclization of g-allenols, see:
g) Z. Zhang, C. Liu, R. E. Kinder, X. Han, H. Qian, R. A. Widen-
hoefer, J. Am. Chem. Soc. 2006, 128, 9066.
Computational methods: All stationary points were located and charac-
terized by DFT calculations with the B3LYP hybrid functional using the
Gaussian 03 software package.[31] The gold, platinum, and palladium
atoms were described by a double-z basis set with the effective core po-
tential of Hay and Wadt (LANL2DZ).[32] The 6-31G(d) basis set[33] was
used for the other elements. The optimized geometries were character-
ized by harmonic analysis and the nature of the stationary points was de-
termined according to the number of negative eigenvalues in the Hessian
matrix. In several cases, the intrinsic reaction coordinate (IRC) pathways
from the transition-state structures were followed by using a second-
order integration method to verify the proper connections with reactants
and products.[34] The reported enthalpies and free energies include ther-
mal corrections. The relative enthalpies and free energies (298 K) are re-
ported in the manuscript. Solvent effects were allowed for through
single-point calculations on the gas-phase optimized geometries. The con-
ductor polarizable continuum model (CPCM)[35] as implemented in the
Gaussian 03 package was used with the parameters chosen by default.
CH2Cl2 and DMF were selected as model solvents with dielectric con-
stants e=8.93 and 39.0, respectively. Natural bond orbital (NBO) analy-
ses[36] were performed with the module NBO v.3.1 implemented in Gaus-
sian 03 to evaluate the NPA charges and Wiberg bond indices at the opti-
mization level.
[4] See, for instance: a) B. Alcaide, P. Almendros, R. Carrascosa, T.
Martꢀnez del Campo, Chem. Eur. J. 2009, 15, 2496; b) B. Alcaide, P.
Almendros, T. Martꢀnez del Campo, Chem. Eur. J. 2008, 14, 7756;
c) B. Alcaide, P. Almendros, R. Carrascosa, M. C. Redondo, Chem.
Eur. J. 2008, 14, 637; d) B. Alcaide, P. Almendros, T. Martꢀnez del
Campo, Angew. Chem. 2007, 119, 6804; Angew. Chem. Int. Ed. 2007,
46, 6684; e) B. Alcaide, P. Almendros, T. Martꢀnez del Campo, R.
Rodrꢀguez-Acebes, Adv. Synth. Catal. 2007, 349, 749; f) B. Alcaide,
P. Almendros, T. Martꢀnez del Campo, Angew. Chem. 2006, 118,
4613; Angew. Chem. Int. Ed. 2006, 45, 4501.
[5] For recent reviews on pyran rings, see: a) I. Larrosa, P. Romea, F.
Urpꢀ, Tetrahedron 2008, 64, 2683; b) P. A. Clarke, S. Santos, Eur. J.
Org. Chem. 2006, 2045; c) J. D. Hepworth, B. M. Heron in Progress
in Heterocyclic Chemistry, Vol. 17 (Eds.: G. W. Gribble, J. A. Joule),
Elsevier, Oxford, 2005, pp. 362–388; for recent reviews on oxepine
derivatives, see: d) N. L. Snyder, H. M. Haines, M. W. Peczuh, Tetra-
hedron 2006, 62, 9301; e) J. B. Bremner in Progress in Heterocyclic
Chemistry, Vol. 16 (Eds.: G. W. Gribble, J. A. Joule), Elsevier,
Oxford, 2004, pp. 431–450; for a review on marine natural products
containing medium-sized cyclic ethers, see: f) D. J. Faulkner, Nat.
Prod. Rep. 2002, 19, 1.
[6] a-Allenol (ꢀ)-4a has previously been prepared as a mixture of
regio- and diastereoisomers by using stannane chemistry: A.
McCluskey, I. W. Muderawan, Muntari, D. J. Young, J. Org. Chem.
2001, 66, 7811.
[7] a) B. M. Trost, J. L. Belletire, S. Godleski, P. G. McDougal, J. M. Bal-
kovec, J. J. Baldwin, M. E. Christy, G. S. Ponticello, S. L. Varga, J. P.
Springer, J. Org. Chem. 1986, 51, 2370; b) K. M. Sureshan, T. Miya-
sou, S. Miyamori, Y. Watanabe, Tetrahedron: Asymmetry 2004, 15,
3357. For a review, see: c) J. M. Seco, E. QuiÇoꢁ, R. Riguera, Chem.
Rev. 2004, 104, 17.
Acknowledgements
Support for this work by DGI-MEC (Grant CTQ2006-10292) and UCM-
BSCH (Grant GR58/08) is gratefully acknowledged. T.M.C. thanks the
MEC for a predoctoral grant. The authors are grateful to the Centro de
Supercomputaciꢆn de Galicia (CESGA) for generous allocation of com-
puting resources.
[8] J. Jurczak, S. Pikul, T. Barer, Tetrahedron 1986, 42, 447.
[9] a) B. Gockel, N. Krause, Org. Lett. 2006, 8, 4485; b) R. Zriba, V.
Gandon, C. Aubert, L. Fensterbank, M. Malacria, Chem. Eur. J.
2008, 14, 1482; c) Y. Sawama, Y. Sawama, N. Krause, Org. Biomol.
Chem. 2008, 6, 3573.
[10] As far as we know, the platinum-mediated cycloisomerization of a
b-allenol has not been previously reported. The only oxycyclization
of a b-allenol is the PtCl4-catalyzed reaction of indoles with b-alle-
nols, which affords indole derivatives containing a six-membered
ether ring at the 3-position. See: W. Kong, J. Cui, Y. Yu, G. Chen, C.
Fu, S. Ma, Org. Lett. 2009, 11, 1213.
[11] B. Alcaide, P. Almendros, T. Martꢀnez del Campo, E. Soriano, J. L.
Marco-Contelles, Chem. Eur. J. 2009, 15, 1901.
[12] B. Alcaide, P. Almendros, T. Martꢀnez del Campo, E. Soriano, J. L.
Marco-Contelles, Chem. Eur. J. 2009, 15, 1909.
[13] For gold-catalyzed enantioselective cyclizations of g-allenols, see: Z.
Zhang, R. A. Widenhoefer, Angew. Chem. 2007, 119, 287; Angew.
Chem. Int. Ed. 2007, 46, 283.
[14] To the best of our knowledge, the only available platinum-mediated
oxycyclization of a g-allenol is the 6-exo cyclization of 2,2-diphenyl-
hexa-4,5-dien-1-ol leading to 6-methyl-3,3-diphenyl-3,4-dihydro-2H-
pyran. See: Z. Zhang, C. Liu, R. E. Kinder, X. Han, H. Qian, R. A.
Widenhoefer, J. Am. Chem. Soc. 2006, 128, 9066.
[1] For reviews, see: a) R. A. Widenhoefer, Chem. Eur. J. 2008, 14,
5382; b) S. Ma, Chem. Rev. 2005, 105, 2829; c) Modern Allene
Chemistry (Eds.: N. Krause, A. S. K. Hashmi), Wiley-VCH, Wein-
heim, 2004; d) R. W. Bates, V. Satcharoen, Chem. Soc. Rev. 2002, 31,
12; e) R. Zimmer, C. U. Dinesh, E. Nandanan, F. A. Khan, Chem.
Rev. 2000, 100, 3067; f) A. S. K. Hashmi, Angew. Chem. 2000, 112,
3737; Angew. Chem. Int. Ed. 2000, 39, 3590.
[2] For a review, see: a) M. Brasholz, H.-U. Reissig, R. Zimmer, Acc.
Chem. Res. 2009, 42, 45; for selected examples of silver-mediated
heterocyclizations of a-allenols, see: b) J. A. Marshall, R. H. Yu, J. F.
Perkins, J. Org. Chem. 1995, 60, 5550; c) O. Flçgel, H.-U. Reissig,
Eur. J. Org. Chem. 2004, 2797; ror selected gold-catalyzed cycliza-
tions of a-allenols, see: d) A. Hoffmann-Rçder, N. Krause, Org.
Lett. 2001, 3, 2537; e) N. Morita, N. Krause, Eur. J. Org. Chem. 2006,
4634; A. S. K. Hashmi, M. C. Blanco, D. Fischer, J. W. Bats, Eur. J.
Org. Chem. 2006, 1387; f) F. Volz, N. Krause, Org. Biomol. Chem.
2007, 5, 1519; g) ꢇ. Aksn, N. Krause, Adv. Synth. Catal. 2008, 350,
1106; for selected pallaadium-catalyzed cyclizative coupling reac-
tions of a-allenols, see: h) S. Ma, W. Gao, J. Org. Chem. 2002, 67,
6104; i) F. Yu, X. Lian, S. Ma, Org. Lett. 2007, 9, 1703; for silver-
and palladium-promoted heterocyclizations of a-allenols, see: j) D.
Xu, Z. Li, S. Ma, Chem. Eur. J. 2002, 8, 5012; k) B. Alcaide, P. Al-
mendros, R. Rodrꢀguez-Acebes, J. Org. Chem. 2006, 71, 2346.
[3] For silver-mediated heterocyclizations of b-allenols, see: a) V. K.
Brel, V. K. Belsky, A. I. Stash, V. E. Zavodnik, P. J. Stang, Eur. J.
Org. Chem. 2005, 512; b) B. D. Sherry, F. D. Toste, J. Am. Chem.
Soc. 2004, 126, 15978; for gold-catalyzed cyclizations of b-allenols,
see: c) B. Gockel, N. Krause, Org. Lett. 2006, 8, 4485; for palladium-
[15] As far as we know, the palladium-catalyzed cyclizative coupling re-
action of acyclic-tethered g-allenols with allyl halides has not yet
been reported. For its pioneering use in a- and b-allenols, see: S.
Ma, W. Gao, J. Org. Chem. 2002, 67, 6104.
Chem. Eur. J. 2009, 15, 9127 – 9138
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9137

B. Alcaide, P. Almendros et al.
[16] C. Jonasson, A. Horvꢁth, J.-E. Bꢈckvall, J. Am. Chem. Soc. 2000,
122, 9600. These authors obtained bromoalkenyltetrahydrofurans,
whereas we obtained a bromotetrahydrooxepine.
[17] V. Gandon, G. Lemiꢉre, A. Hours, L. Fensterbank, M. Malacria,
Angew. Chem. 2008, 120, 7644; Angew. Chem. Int. Ed. 2008, 47,
7534.
[18] a) P. S. Kalsi, Stereochemistry Conformation and Mechanism, 6th
ed., New Age International, New Dehli, Chapter 4, 2005; b) V. S.
Mastryukov, K.-H. Chen, N. L. Allinger, J. Phys. Chem. A 2001, 105,
8562.
[19] a) L. I. Ermolaeva, V. S. Mastryukov, N. L. Allinger, A. Almennin-
gen, J. Mol. Struct. 1989, 196, 151; b) M. K. Leong, V. S. Mastryukov,
J. E. Boggs, J. Mol. Struct. 1998, 445, 149.
[20] A plausible direct 1,3-hydrogen shift was previously ruled out on
the basis of the high energy barrier (see ref. [12]).
[21] We have located an alternative transition-state structure that lacks
this hydrogen bond between the ligand and the moving hydrogen
and lies 16.2 kcalmolꢀ1 above TS37.
[22] a) J. S. M. Samec, J.-E. Bꢈckvall, P. G. Andersson, P. Brandt, Chem.
Soc. Rev. 2006, 35, 237; b) R. H. Crabtree, The Organometallic
Chemistry of the Transition Metals, Wiley, New York, 2001; c) S.
Niu, M. B. Hall, Chem. Rev. 2000, 100, 353.
[23] a) L. J. Ackerman, M. L. H. Green, J. C. Green, J. E. Bercaw, Orga-
nometallics 2003, 22, 188; b) L. H. Shultz, M. Brookhart, Organome-
tallics 2001, 20, 3975.
[28] For the stereoselectivity of b-heteroatom elimination, see: a) C. G.
Frost, J. Howarth, J. M. J. Williams, Tetrahedron: Asymmetry 1990,
1, 201; b) G. D. Daves, Jr., Acc. Chem. Res. 1990, 23, 201; c) G. Zhu,
X. Lu, Organometallics 1995, 14, 4899; d) see ref. [4f].
[29] X. Lu, G. Zhu, Z. Wang, Synlett 1998, 115, and references therein.
[30] Z. Zhang, X. Lu, Z. Xu, Q. Zhang, X. Han, Organometallics 2001,
20, 3724.
[31] Gaussian 03, Revision E.01, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery,
Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar,
J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega,
G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R.
Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross,
V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y.
Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg,
V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O.
Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Fores-
man, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski,
B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L.
Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Na-
nayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wallingford
CT, 2004.
[32] P. J. Hay, W. R. Wadt, J. Chem. Phys. 1985, 82, 299.
[24] M. Brookhart, M. L. H. Green, G. Parkin, Proc. Natl. Acad. Sci.
USA 2007, 104, 6908.
[25] a) A. Hoffmann-Rçder, N. Krause, Org. Biomol. Chem. 2005, 3, 387;
b) A. S. K. Hashmi, G. J. Hutchings, Angew. Chem. 2006, 118, 8064;
Angew. Chem. Int. Ed. 2006, 45, 7896.
[33] a) R. Ditchfield, W. J. Hehre, J. A. Pople, J. Chem. Phys. 1971, 54,
724; b) W. J. Hehre, R. Ditchfield, J. A. Pople, J. Chem. Phys. 1972,
56, 2257; c) P. C. Hariharan, J. A. Pople, Theo. Chim. Acta 1973, 28,
213; d) P. C. Hariharan, J. A. Pople, Mol. Phys. 1974, 27, 209;
e) M. S. Gordon, Chem. Phys. Lett. 1980, 76, 163.
[26] The regioselectivity found here is quite similar to those found in the
insertion reactions of alkenes with many neutral PdII complexes,
see: a) A. Michalak, T. Ziegler, Organometallics 2000, 19, 1850;
b) W. Cabri, I. Candiani, Acc. Chem. Res. 1995, 28, 2.
[27] a) P. J. Harrington, L. S. Hegedus, K. F. McDaniel, J. Am. Chem.
Soc. 1987, 109, 4335; b) J. W. Francis, P. M. Henry, Organometallics
1991, 10, 3498; c) M. Kimura, Y. Horino, R. Mukai, S. Tanaka, Y.
Tamaru, J. Am. Chem. Soc. 2001, 123, 10401; d) F. Ozawa, H. Oka-
moto, S. Kawagishi, S. Yamamoto, T. Minami, M. Yoshifuji, J. Am.
Chem. Soc. 2002, 124, 10968; e) K. Manabe, S. Kobayashi, Org. Lett.
2003, 5, 3241; f) G. W. Kabalka, G. Dong, B. Venkataiah, Org. Lett.
2003, 5, 893; g) M. Yoshida, T. Gotou, M. Ihara, Chem. Commun.
2004, 1124; h) G. Liu, X. Lu, Org. Lett. 2001, 3, 3879.
[34] a) K. Fukui, Acc. Chem. Res. 1981, 14, 363; b) C. Gonzalez, H. B.
Schlegel, J. Phys. Chem. 1990, 94, 5523.
[35] V. Barone, M. Cossi, J. Phys. Chem. A 1998, 102, 1995.
[36] NBO, Version 3.1, E. D. Glendening, A. E. Reed, J. E. Carpenter, F.
Weinhold; for original literature, see: a) A. E. Reed, F. Weinhold, J.
Chem. Phys. 1983, 78, 4066; b) A. E. Reed, L. A. Curtiss, F. Wein-
hold, Chem. Rev. 1988, 88, 899.
[37] Experimental procedures as well as full spectroscopic and analytical
data for compounds not included in this Experimental Section are
described in the Supporting Information. It also contains copies of
NMR spectra for all new compounds.
Received: May 5, 2009
Published online: July 27, 2009
9138
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9127 – 9138