2-Methoxy-Δ3-1,3,4-oxadiazoline, a multipurpose precursor for the generation of a carbene, an ylide, or a diazo compound

Article abstract of DOI:10.1002/ejoc.200800709

In the course of our investigation of the intermolecular reactions of foiled carbenes of the norborn-2-en-7-ylidene type, we have investigated the decomposition of methoxyoxadiazoline 1 in alcohols. Most of the reactions performed lead to products with an anti configuration confirming the participation of the double bond to the stabilization of the transition states. The thermal behavior of spirooxadiazoline 1 is quite different from the behavior of the parent 2-methoxy-2,5,5-trimethyl-Δ3-1,3,4-oxadiazoline. Photolysis of 1 leads to the carbene after prior formation of the diazo compound whereas thermolysis cleanly generates an extremely unstable carbonyl ylide 4 that immediately decomposes to the stabilizednucleophilic carbene 5 and methyl acetate without generation of 1-methoxyethylidene. Both conformers of 4 do not interconvert and, therefore, they have different life times. Nevertheless, we were able to trap syn ylide 4a with methanol. Calculations show that nonbonding interactions between an alkyl carbene and an ester are more significant than ylide formation. Synthetically, photolysis of oxadiazoline 1 in ethyl acetate has proven to give anti ethers in excellent yields. Moreover, we report the first reactions of a norbornenylidene derivative with O-H bonds in which no products resulting from a cationic rearrangement are formed. Wiley-VCH Verlag GmbH & Co. KGaA, 2008.

Full text of DOI:10.1002/ejoc.200800709

FULL PAPER
DOI: 10.1002/ejoc.200800709
2-Methoxy-∆3-1,3,4-oxadiazoline, a Multipurpose Precursor for the Generation
of a Carbene, an Ylide, or a Diazo Compound[‡]
Jean-Luc Mieusset,[a] Peter Billing,[a] Michael Abraham,[a] Vladimir B. Arion,[b]
Lothar Brecker,[a] and Udo H. Brinker*[a]
Keywords: Carbenes / Ylides / Oxadiazolines / Diazo compounds / Photochemical reactions
In the course of our investigation of the intermolecular reac-
tions of foiled carbenes of the norborn-2-en-7-ylidene type,
we have investigated the decomposition of methoxyoxadiaz-
oline 1 in alcohols. Most of the reactions performed lead to
products with an anti configuration confirming the participa-
tion of the double bond to the stabilization of the transition
states. The thermal behavior of spirooxadiazoline 1 is quite
different from the behavior of the parent 2-methoxy-2,5,5-
trimethyl-∆3-1,3,4-oxadiazoline. Photolysis of 1 leads to the
carbene after prior formation of the diazo compound whereas
thermolysis cleanly generates an extremely unstable car-
bonyl ylide 4 that immediately decomposes to the stabilized-
tion of 1-methoxyethylidene. Both conformers of 4 do not in-
terconvert and, therefore, they have different life times. Nev-
ertheless, we were able to trap syn ylide 4a with methanol.
Calculations show that nonbonding interactions between an
alkyl carbene and an ester are more significant than ylide
formation. Synthetically, photolysis of oxadiazoline 1 in ethyl
acetate has proven to give anti ethers in excellent yields.
Moreover, we report the first reactions of a norbornenylidene
derivative with O–H bonds in which no products resulting
from a cationic rearrangement are formed.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
nucleophilic carbene 5 and methyl acetate without genera- Germany, 2008)
pattern and their activation, their decomposition pathways
Introduction
vary considerably.[1] Especially, 2-alkoxy-∆3-1,3,4-oxadiaz-
olines are very attractive since Warkentin has shown some
20 years ago that they are suitable for the generation of
alkyl carbenes under photochemical conditions after prior
formation of a diazoalkane.[2] Indeed, they have permitted
numerous investigations.[3] In this work, we will show that
the two diastereomers of oxadiazoline 1 differ significantly
in their chemistry, that their thermal activation leads exclu-
sively to the formation of the strained alkenylidene 5 with-
out the intermediacy of a diazo species, and that the gener-
ated intermediates are very efficiently trapped by alcohols.
It is still a challenging task to find a convenient source of
alkyl carbenes: for an efficient interception of the divalent
species, the precursor has to be generated under mild condi-
tions and no other reactive intermediate should be gener-
ated. Of course, the precursor should be readily available. In
many cases, nitrogen-containing compounds are used like
tosylhydrazone salts or diazirines but most of them have
significant drawbacks. First of all, usually, a quite reactive
diazo species is initially generated which decomposes only
in a further step to a carbene. Diazirines often lead directly
to the carbene but frequently, their preparation succeeds
only with very low yields. Therefore, ∆3-1,3,4-oxadiazolines
have become very popular precursors. They are readily syn-
thesized from the corresponding ketones and have been
widely used for the generation of nucleophilic carbenes like
dimethoxycarbene. And depending on their substitution
Results and Discussion
Tricyclo[6.2.1.02,7]undec-9-en-11-ylidene (5) is a typical
foiled carbene based on a cyclopent-3-enylidene scaffold. In
these reactive intermediates, stabilization occurs mainly
through electron donation from the C=C double bond into
the antibonding lone pair (LP*) of the divalent carbon
atom, making the LUMO of carbene 5 less reactive. In
other words, the electrophilicity of the carbene is strongly
reduced and, overall, the nucleophilic character of 5 is de-
termining its chemoselectivity. This led us to define norbor-
nenylidene derivatives as stabilized-nucleophilic carbenes.[4]
These electronic processes explain why foiled carbenes
could not be trapped intermolecularly in synthetically use-
[‡] Carbene Rearrangements 74. Part 73: A. Azizoglu, M. Balci,
J.-L. Mieusset, U. H. Brinker, J. Org. Chem. 2008, 73, in press.
[a] Chair of Physical Organic and Structural Chemistry, Faculty of
Chemistry, University of Vienna
Währinger Str. 38, 1090 Vienna, Austria
Fax: +43-1427752140
E-mail: udo.brinker@univie.ac.at
[b] Institute of Inorganic Chemistry, Faculty of Chemistry, Univer-
sity of Vienna
Währinger Str. 42, 1090 Vienna, Austria
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
5336
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 5336–5345

Precursor for a Carbene, an Ylide, or a Diazo Compound
ful yields by alkenes or hydrocarbons. In fact, the only ef- nitrogen and generates the carbene.[2] The only drawback of
ficient intermolecular reactions known so far concerns the precursor 1 was that the two diastereomers, which are
insertion into the N–H bond of diethylamine[5] and the formed in about equal amounts, decomposed differently, es-
O–H bond of methanol.[6] For the latter reaction, the ease pecially when photolyzed. Whereas diastereomer 1a can be
of etherification can be related to the high proton affinity conveniently decomposed within two hours, diastereomer
(PA) of 5. Indeed, 5 still possesses the same high PA charac- 1b requires reaction times of more than one day. Therefore,
teristic of nucleophilic and alkyl carbenes.[7]
in this special case, both compounds were separated by col-
In fact, it has already been proven by means of deute- umn chromatography and the photolysis was exclusively
rium labels[6b,8] that bicyclo[2.1.1]hexan-2-ylidene is proton- performed with 1a. In control experiments, diastereomer 1b
ated by methanol and that in solvolysis experiments nor- has been shown to react in the same manner as 1a, but just
bornen-7-ylidene reacts efficiently with the O–H bond of slower.
methanol through the carbocation mechanism.[6b] Indeed,
Photolysis of 1a in pure methanol afforded two methyl
three different mechanisms can be envisaged for this inser- ethers: 48% anti,endo-11-methoxytricyclo[6.2.1.02,7]undec-
tion reaction (Scheme 1):[8] (a) protonation to a carbenium 9-ene (10me) and 43% 3-methoxytricyclo[5.4.0.02,6]undec-
ion, (b) a concerted cheletropic cycloaddition and (c) the 4-ene (12me) (Table 1, Scheme 2). The presence of the re-
OH group catalysis mechanism[9] in which alcohol (ether) arrangement product 12me suggests that syn-diazonium ion
formation occurs in only one step but in which the hydrogen 3 has been generated as an intermediate[6b] and that the
atom and the hydroxy (alkoxy) group stem from two dif- reaction mainly does not proceed through carbene 5.
ferent molecules. The often discussed ylide mechanism (d)
by which the carbene first attacks the oxygen atom and then
Table 1. Distribution (%) of the products (NMR integration) in the
rearranges by a proton transfer (Stevens rearrangement) is
only relevant for reactive-electrophilic carbenes.[4,9,10] In
this study, we have attempted to repeat this etherification
with compounds 1 under less polar conditions in order to
favor the concerted reaction pathway over the cationic
mechanism. The large difference in pKa for methanol and
more generally for alcohols depending on the solvents used
(pKa = 15.54 in H2O)[11] and (pKa = 29 in DMSO)[12] sug-
gests that the cationic mechanism may not occur in nonpo-
lar organic solvents, since under these conditions a proton-
ation of the carbene is less probable.
crude reaction mixtures obtained from the decomposition of 40 mg
of oxadiazoline 1 in methanol.[a]
9
10me 11me 12me
Photolysis with 1a:
a) MeOH, hν
b) 2 equiv. MeOH in EE, hν
c) 2 equiv. MeOH in benzene, hν
d) 2 equiv. MeOH in hexane, hν
–
–
–
–
48
82
92
85
–
–
7
–
43
11
1
15
Thermolysis:
e) 1b: MeOH, ∆
–
26
–
96
71
80
–
–
15
–
–
–
f) 1a: MeOH, ∆
g) 1b: THF/MeOH (10:1), ∆
[a] No methyl ether is obtained from the thermolysis of 1a in a
saturated solution of MeOH in hexane or with 10 equiv. of MeOH
in benzene.
Under less polar conditions, the amount of 12me is re-
duced but 12me is still produced [11% 2 equiv. MeOH in
ethyl acetate (EE) and even 15% 2 equiv. MeOH in hexane].
These results show that in solution some of the diazo com-
pound 2 decomposes to carbene 5 before it can be proton-
ated. In benzene, 7% of tetracycle 11me is also formed, a
compound whose formation is best explained by trapping
of carbocation 6. It is worth noticing that the correspond-
ing capture of the closely related 7-norbornenyl cation pro-
ceeds even in methanol under methoxylation at the former
double bond.[6b] Overall, even under the less polar condi-
tions, protonation of 2 and 5, respectively, plays a signifi-
cant role in the formation of methyl ethers 10me, 11me, and
12me. It is also worth noticing that the trapping of allyl
cation 7 by methanol leads exclusively to exo-ether 12me
because the six-membered ring seems to prevent the ap-
proach of the nucleophile from the endo-side. In contrast,
capture of the parent bicyclo[3.2.0]hept-3-en-2-yl cation by
methanol leads to a mixture of endo- and exo-ethers.[6b]
Thermally, compound 1a reacts faster than 1b. Therefore,
Scheme 1. Possible reaction mechanisms for the insertion of a car-
bene into an O–H bond.
Experimental Results
2-Methoxy-2-methyl-∆3-1,3,4-oxadiazolines are known
to react photochemically via a 1,3-dipolar cycloreversion to reactions of 1a were performed at 155 °C while reactions
methyl acetate and a diazo compound that then splits off with 1b were carried out at 165 °C. However, the thermoly-
Eur. J. Org. Chem. 2008, 5336–5345
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5337

U. H. Brinker et al.
FULL PAPER
Table 2. Distribution (%) of products in the crude reaction mixtures
obtained from the decomposition of 40 mg of oxadiazoline 1 in
presence of alcohols or water.[a]
10
11
Photolysis with 1a:
1 equiv. TrOH in 1 mL of benzene, hν
1 equiv. TrOH in 1 mL of acetonitrile, hν
1 equiv. TrOH in 1 mL of EE, hν
2 equiv. BnOH in 0.5 mL of EE, hν
2 equiv. BnOH in 0.5 mL of THF, hν
10tr: 63
10tr: 59
10tr: 66
10bn: 89
10bn: 73
11tr: 2
11tr: 6
–
–
–
Thermolysis with 1b:
Dioxane/water (5:1), ∆
[a] Tr = trityl. Bn = benzyl.
10h: 94
–
Anyhow, these reaction conditions point to a formal con-
certed insertion of 5 into the O–H bond since rearrange-
ment products 11 and 12 are not obtained. Moreover, this
reaction represents a convenient synthetic procedure for the
transformation of a ketone to an ether since both reaction
steps, the formation and the decomposition of the oxadia-
zoline, take place in a cleanly manner. Whereas many car-
benoids tend to insert into the O–H bonds as well as into
the C–H bonds of alcohols,[14] here only ether formation
has been observed.
Scheme 2. Reaction mechanisms for the decomposition of oxadia-
zoline 12 under protic conditions.
From a synthetic point of view, photolysis is a more suit-
able method than thermolysis because oxadiazolines 1a and
1b react selectively with the O–H bond of alcohols even if
only stoichiometric amounts are used. With BnOH and
TrOH, no rearrangement is observed, only the anti ether is
obtained (66% 10tr and 89% 10bn in ethyl acetate). The
moderate yield of 66% achieved with the sterically hindered
Ph3COH is caused by reaction with water traces and forma-
tion of alcohol 10h and of the corresponding symmetri-
cal ether bis(anti,endo-tricyclo[6.2.1.02,7]undec-9-en-11-yl)
ether. Slightly lower yields are obtained in acetonitrile
(59%) and tetrahydrofuran (63%).
ses in saturated solutions of methanol in hexane (20 mg of
1a and 1b in 4 mL) or with 10 equiv. of methanol in ben-
zene were not successful. No volatile product was formed
and etherification was only obtained under solvolytic con-
ditions. Probably, carbene 5 rearranges intramolecularly by
a 1,2 vinyl shift before it can be trapped and the resulting
strained alkene polymerizes under the drastic reaction con-
ditions. Thermal decomposition of oxadiazoline 1a in dry
methanol at 155 °C leads to formation of anti ether 10me
but in addition, syn alcohol 9 is formed. The presence of
this compound is best explained by the capture of carbonyl
ylide 4a by methanol. Replacement of methanol by tBuOH,
AcOH or dioxane/water as a trap does not lead to an in-
crease in the formation of 9. Only 2  of NaOMe in meth-
anol increased the production of 9 to 35%. Thermally, no
other methyl ether is formed and no rearrangement is ob-
served. This result suggests that no diazo compound is gen-
Investigation of the Reaction Mechanisms
In a next step, we took a closer look at the decomposi-
erated from 1a and that the cycloreversion solely leads to tion of precursor 1 in order to better understand in which
the formation of a carbonyl ylide. Starting from dia- reactions a carbene can be expected. As a multipurpose pre-
stereomer 1b in pure methanol, anti methyl ether 10me is cursor, oxadiazolines may decompose either to a ketone, a
formed exclusively in a particularly clean reaction. The ab- diazo compound, or a carbonyl ylide.[1] Whereas the chem-
sence of syn alcohol 9 indicates that the other carbonyl ylide istry of ketones and diazo compounds is well understood,
conformer 4b is formed and that it does not interconvert to carbonyl ylides remain fascinating species with properties
4a before reacting further. Under less polar conditions, in that are extremely dependant on their substitution.[15] For
THF/MeOH (10:1), tetracycle 11me is also formed[13] from example, ylides formally derived from the addition of an
1b by an attack of the nucleophile on C(9) of cation 6 sug- electrophilic carbene to a carbonyl group are convenient in-
gesting that the etherification proceeds through the cationic termediates to undergo 1,3-dipolar cycloadditions to
mechanism. It is worth noticing that tetracycle 11 has only double bonds.[15] Moreover, trapping of tetrakis(trifluoro-
been observed by reactions with methanol and not with methyl)cyclopentadienylidene by urea leads to a push-pull
other nucleophiles. Indeed, thermal hydrolysis of oxadiaz- ylide, which is crystalline and melts at 200–202 °C.[16] In
oline 1b in dioxane/water (5:1) exclusively leads to anti contrast, nucleophilic carbenes do not even add to the oxy-
alcohol 10h in a very good yield (Table 2).
gen atom of C=O double bonds.[17]
5338
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 5336–5345

Precursor for a Carbene, an Ylide, or a Diazo Compound
In regard to the decomposition of oxadiazoline 13, com- and 16. In 13, C(7) is about sp3-hybridized (preferred angle
putations were performed at the B3LYP/6-31G(d) level of in unstrained systems: 109.5°; actual angle from the crystal
theory. It is found that diastereomer 13b is more stable than structure of 1b[5] is 94.3°) whereas for the formation of the
13a by about 1.3 kcal/mol. This difference is observed again products a double bond has to be formed leading to sp2
from the calculations of the transition states of the nitrogen hybridization (preferred angle: 120°), a geometry requiring
liberating reactions for which the barriers are 1.2 to an even larger angle. A similar behavior is found for exam-
1.3 kcal/mol higher for 13b than for 13a (Figure 1). Experi- ple for ketals of bicyclo[3.2.1]octan-8-one derivatives which
mentally, diastereomer 13b reacts slower, necessitating an are more stable than ketals of acyclic ketones.
increase of the reaction temperature from 155 °C to 165 °C
in order to maintain a similar reaction time as for dia-
stereomer 13a.
Figure 2. Potential cycloreversions for the decomposition of oxadi-
azoline 18. Enthalpies in kcal/mol as given by B3LYP/6-31G(d).
For the decomposition of oxadiazolines 13, formation of
two different conformers of ylide 16 is predicted (Figure 1).
The lowest pathway leads to E-16a and E-16b, respectively,
in which the oxygen atoms are arranged in the same manner
than in methyl acetate in its E-conformation. However, the
Figure 1. Potential cycloreversions for the decomposition of oxadi-
azolines 13a and 13b. Enthalpies in kcal/mol as given by B3LYP/ difference in the activation energies is small and a signifi-
6-31G(d).
cant amount of Z-16 should be generated. Merely due to
dipole-dipole and also to steric interactions, esters are bet-
Figure 1 shows the results of the computations for the ter stabilized in the Z-conformation.[20] Therefore, Z-16a
decomposition of oxadiazoline 13, which speak for reac- decomposes readily to norbornenylidene 17 by overcoming
tions through carbonyl ylides 16. The barrier for ylide for- a low barrier of 0.6 kcal/mol [Figure 3, TS(Z-16a/17),
mation (30.7 to 31.9 kcal/mol) is significantly lower then +10.7 kcal/mol in comparison to norbornenylidene and
that leading to diazo species 15 (34.3 and 34.6 kcal/mol) or methyl acetate, +0.6 kcal/mol from ylide Z-16a]. Ylide Z-
ketone 14 (38.5 and 38.8 kcal/mol). The differences in the 16b was not even found; instead oxadiazoline 13b directly
decomposition rates between 13a (overnight in a pressure gives carbene 17 because the decomposition of Z-16b is fur-
tube at 155 °C; calculated barrier: ∆H‡ = 30.7 kcal/mol, ther facilitated by anchimeric assistance from the double
∆G‡ = 29.0 kcal/mol) and 13b (overnight at 165 °C, calcu- bond.
lated barrier: ∆H‡ = 31.9 kcal/mol, ∆G‡ = 29.7 kcal/mol)
In contrast, for the decomposition of Z-16a to 1-meth-
are also well reproduced. The absolute values are satisfac- oxyethylidene and norbornenone 14 10.6 kcal/mol [TS(Z-
tory but a little too low; according to the Eyring equation 16a/14)] are required. This prohibitively high energy barrier
a free energy of activation of 32–33 kcal/mol would be ex- is at first surprising because the nucleophilic 1-methoxye-
pected. However, it is worth noticing that the back reactions thylidene is a more stable carbene than usual alkyl carbenes.
16aǞ13a and 16bǞ13b, respectively, occur with very low However, the stability of the second compound formed is
enthalpic barriers.
also important: liberation of norbornenylidene 17 leads to
The decomposition of 13 requires 3–4 kcal/mol more en- the release of an ester, i.e., methyl acetate, whereas genera-
ergy than is necessary for the thermolysis of the smaller tion of 1-methoxyethylidene leads to formation of strained
oxadiazoline 18 (Figure 2).[18] Indeed, 18 usually is decom- ketone 14. As a syn ylide, no participation of the neigh-
posed under exclusive formation of ylide 21 at temperatures boring double bond is expected. The preference for the gen-
as low as 80 °C with rates (k79.5
= 5.3ϫ10–6 s–1 and eration of norbornenylidene 17 can already be seen in the
CD3OD
k79.5
= 1.4ϫ10–5 s–1)[19] in good agreement with our cal- geometry of ylide Z-16a (Figure 3): the C–O bonds are
CCl4
culated values. The difference in the stability of oxadiazol- strongly asymmetric. The C(7)–O bond is relatively long
ines 13 and 18 probably is caused by the ring strain of poly- with 1.474 Å (normal single C–O bond: 1.43 Å)[21] whereas
cyclic compound 13 and its decomposition products 14, 15, the carboxylic C=O bond (1.266 Å) of the former methyl
Eur. J. Org. Chem. 2008, 5336–5345
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5339

U. H. Brinker et al.
FULL PAPER
Figure 3. Stationary points describing ylides E-16a and Z-16a and their decomposition. Enthalpies at 298 K are in kcal/mol and distances
in Ångstroms. Norbornenylidene and methyl acetate were taken as reference (0 kcal/mol).
acetate still has about the expected length of such a bond idene or 1-methoxyethylidene are correlated with high en-
(normally 1.21 Å).[21] The carboxylic carbon is still planar ergy barriers [3.8 and 8.0 kcal/mol for TS(E-16a/17) and
whereas C(7) is sp3-hybridized, suggesting the presence of a TS(E-16a/14), respectively] (Figure 3) when compared with
negative charge on C(7). This lone pair of electrons further Z-16a. However, rotation around the C–O bond is almost
interacts (d = 2.501 Å) with the C–H bonds in α-position free of energy [TS(E-16a/Z2–16a), +0.5 kcal/mol] and leads
to the carboxylic group.
to a second Z-conformer Z2-16a which readily decomposes
In contrast, E-16a has the potential to have a more sig- to norbornenylidene 17 [TS(Z2-16a/17), +1.5 kcal/mol]
nificant lifetime. The direct decompositions to norbornenyl- (Figure 4).
Figure 4. Energy diagram for the decomposition of ylides E-16a and E-16b. Enthalpies at 298 K are in kcal/mol and distances in
Ångstroms. Norbornenylidene and methyl acetate were taken as reference (0 kcal/mol).
5340
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 5336–5345

Precursor for a Carbene, an Ylide, or a Diazo Compound
For the generation of a carbene, with 18.8 kcal/mol, this conformation)[24] and therefore should possess more diradi-
reaction is quite exothermic and can be considered as cal character and be less polar than 16.[9,24,25] The heterocy-
irreversible because 17 has the possibility to react intramo- cle 18 is predicted to almost exclusively give ylide E-21
lecularly by a 1,2-vinyl shift to bicyclo[3.2.0]hepta-1,6- which lies in a relatively deep well with reasonable barriers
diene.[7,22] Because of the low barrier toward decomposition toward formation of propan-2-ylidene [TS(E-21/24),
of ylide 7-(1-methoxyethylideneoxonio)bicyclo[2.2.1]hept-2- 12.4 kcal/mol], 1-methoxyethylidene [TS(E-21/25), 7.3 kcal/
en-7-ide (16), inversion of the configuration at C(7) of syn mol], and 2-(1-methoxyethoxy)propene (27) [TS(E-21/27),
ylide E-16a to anti ylide E-16b should not occur [TS(E-16a/ 7.1 kcal/mol] (start on the right side of Figure 5). Calcula-
E-16b): barrier 3.5 kcal/mol]. Despite its low kinetic sta- tions at the CCSD(T) level confirm the validity of the
bility, we were able to trap 16a with methanol. Interestingly, B3LYP results but also show that the stabilities of the car-
the noncovalent interactions between the stabilized nucleo- bonyl ylides 21 are slightly underestimated. These relatively
philic carbene 17 and methyl acetate are quite significant high activation energies explain why ylide E-21 could be
(4.9 kcal/mol, see structure A in Figure 4). The main inter- trapped by acetone[26] or chloroform.[27] This scanning of
action results from electron donation from the LP at the the energy surface also means that thermally 18 is predicted
carbenic center toward the slightly acidic C–H bond of not to be a source of alkyl carbene, in contrast to the exper-
methyl acetate (second order perturbation theory analysis imental results which speak for the generation of about
of the Fock matrix in the NBO basis: E = 5.4 kcal/mol) equal amounts of both carbenes.[19,28] The second relevant
assisted by electron donation from the carboxylic oxygen conformer, Z-21, can neither be obtained from E-21 nor
into an alkenic C–H antibond (1.1 kcal/mol). It all sums up from oxadiazoline 18. It is predicted to be in equilibrium
to a small charge transfer of 0.01 e from carbene 17 to with the complex between propan-2-ylidene and methyl
methyl acetate. Dipole-dipole interactions between the acetate. The barrier between both structures is quite low
carbenic center and the C=O double bond should also pro- [TS(Z-21/24), 0.9 and 1.2 kcal/mol, respectively; start on
vide an important contribution to the stabilization of the the left side of Figure 5]. The next higher transition states
structure.
represent a 1,4-H shift to 2-(1-methoxyvinyloxy)propane
Whereas these calculations show that 2-methoxy-2- (26) [TS(Z-21/26), 2.3 kcal/mol] or to 27 [TS(Z-21/27),
methyl-∆3-1,3,4-oxadiazolines generated from strained 3.1 kcal/mol, see Supporting Information], and the decom-
ketones should be a convenient thermal source of alkyl position of Z-21 to acetone and 1-methoxyethylidene
carbenes, the computational results obtained for ylide 21 [TS(Z-21/25), 2.9 kcal/mol]. These results suggest that car-
generated from the monocyclic oxadiazoline 18 lead to a bene 25 may be obtained through generation of an alkyl
more complex picture (Figure 5). Once again, two conform- carbene in the corresponding ester. This process can be
ers with different reactivities have to be considered.[23] In coined a transcarbenation. This kind of reaction could be
contrast to bicyclic ylide 16, 21 is almost planar (a 0°, 0° particularly useful for the generation of carbenes for which
Figure 5. Main stationary points on the potential energy surface of ylides E-21 and Z-21. Energies in kcal/mol as given by B3LYP/6-
31G(d); values in parentheses represent CCSD(T)/6-31G(d)//B3LYP/6-31G(d) computations. Propan-2-ylidene and methyl acetate were
taken as reference (0 kcal/mol).
Eur. J. Org. Chem. 2008, 5336–5345
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5341

U. H. Brinker et al.
FULL PAPER
no readily available precursors exist. The exclusive genera-
Next, we examined the transition states 32–35 leading to
tion of ylide E-21 from the oxadiazoline 18 is also reflected formation of methyl ether (Figure 7). For an efficient reac-
in the gas phase chemistry of 18 where for steric reasons 27 tion, the hydroxy group should be well oriented with the
is the sole 1,4 H-shift product.[28] In contrast, in a carbon hydrogen atom pointing toward the LP of the divalent car-
tetrachloride solution one half of ylide 21 decomposes to bon and the oxygen atom directed toward the LP*. We
propan-2-ylidene and the rest generates 1-methoxyethylid- found that the concerted insertion into methanol is indeed
ene.[28] There are two reasons for the different behavior in predicted to be specific, since the syn-TS 33 is 6.0 kcal/mol
the gas phase and in solution: firstly, thermal fragmentation higher in energy than the anti-TS 32. This is due to the fact
of carbonyl ylides occurs from the more zwitterionic 0°, 90° that the reaction proceeds via an early transition state, a
conformation and their transition states are more polar fact that is reflected in the relatively short C(2)–C(7) dis-
than the H-shift and should therefore be disfavored in the tance (2.141 Å). The first part of the reaction being nucleo-
gas phase and favored in solution. Secondly, the H-shift is philic, the approach of the hydrogen is accompanied by the
even more favored at higher temperatures (69% at formation of a partial positive charge on the norbornenylid-
380 °C).[28,29] This observation suggests that dynamic ef- ene unit which is efficiently stabilized by the double bond.
fects[30] may play a role in the formation of vinyl ether 27.
The energy barrier for this anti approach is calculated to be
0.5 kcal/mol. This value is increased to 11.7 kcal/mol after
consideration of the hydrogen-bridged structure 30. For the
reaction of an alkyl carbene, this is a really high value and
therefore, it is unlikely that the reaction occurs through this
concerted pathway. In fact, the insertion is better described
with two methanol molecules than with one, as is depicted
in 35.
Mechanism of the Carbene Insertion into the O–H Bond
Finally, we have also analyzed the reaction mechanism
of the formal carbene insertion into O–H bonds. For this,
methanol and 7-norbornenylidene were taken as model
compounds. We investigated the approach between the two
reactants and found that a moderately strong hydrogen
bond (structure 30, –11.2 kcal/mol) is predicted between the
lone pair (LP) of the divalent carbon C(7) and the O–H
bond of methanol (Figure 6). This value is consistent with
the distance calculated between C(7) and the hydroxylic hy-
drogen atom (1.932 Å). Of course, other complexes with
different arrangements can be found but they are all higher
in energy. Our attempts to find an ylide-like structure,[10,31]
lead only to the weakly associated structure 28 in which
four donor-acceptor interactions result in an energy gain
of 3.5 kcal/mol. The relevant interaction energies from the
carbene to methanol resulting from a NBO analysis are:
LP(1) C(7)ǞBD*(1) C(A)–H: 2.47 kcal/mol and LP*(2)
C(7)ǞBD*(1) C(A)–H: 1.27 kcal/mol and from methanol
to the carbene: BD(1) C(A)–HǞLP*(2) C(7): 0.83 kcal/
mol and BD(1) C(A)–HǞBD*(1) C(5)–H: 0.35 kcal/
mol.[32] We also found structure 29 which really presents the
characteristics of a weak ylide, with a short C–O distance
of 1.712 Å but which is high in energy (+4.8 kcal/mol) be-
cause it is based on a poorly stabilized norbornenylidene
conformer.
Norbornenylidene is predicted to be significantly better
stabilized by a methanol dimer (structure 31, –14.9 kcal/
mol). From this structure, two transition states leading to
an anti methyl ether can be found. The first one, structure
34, corresponds to a concerted insertion of the carbene into
one molecule of methanol during which the breaking of the
O–H bond is assisted by a hydrogen bond pointing
from the second methanol molecule. With 7.2 kcal/mol
(–7.7 kcal/mol if norbornenylidene and a methanol dimer
are taken as reference), the energy barrier already is sub-
stantially reduced. However, reaction through the OH
group catalysis mechanism[9] is even more favored (structure
35, energy barrier: +3.3 kcal/mol from complex 31 and
–11.6 kcal/mol from the reactants). Here, the hydrogen
atom and the methoxy group stem from two different meth-
anol molecules. This mechanism is especially efficient be-
cause it allows a good overlap of the orbitals and optimal
interactions between the carbene and the two alcohol mole-
cules. Here, the hydrogen transfer energy wise is unde-
manding because the donor, the acceptor, and the hydrogen
atom are all aligned. Concomitantly, the oxygen atom of the
Figure 6. Geometries and relative energies of the most relevant complexes between norbornen-7-ylidene and methanol. The resulting
charge on the carbene unit C7H8 as given by NPA analyses.
5342
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 5336–5345

Precursor for a Carbene, an Ylide, or a Diazo Compound
Figure 7. Geometries and relative energies of the most relevant transition states for the insertion of norbornen-7-ylidene into methanol.
[a] Energy based on norbornenylidene and the methanol dimer (–5.5 kcal/mol).
second methanol molecule is directed from the beginning of this reaction in the gas phase the insertion is best described
the reaction toward the empty p-orbital of the carbene. In by the OH group catalysis mechanism because the negative
our opinion this mechanism describes best the reaction be- charge in the alkoxylate is not delocalized and therefore, no
tween a carbene and an O–H bond. Indeed, as expected, contact ion pair can be computed. This mechanism also
the stepwise mechanism in which the carbene is protonated nicely explains why the kinetics of the reaction between
and an ion pair is formed cannot be found computationally. alcohol and carbenes is strongly dependant on the concen-
This is due to the high nucleophilicity of the methoxide ion; tration of the alcohol. From a synthetic point of view, the
in this compound, the negative charge remains mostly lo- synthesis of the oxadiazoline followed by photolysis is a
cated on the oxygen atom and this species is still very reac- very convenient method to transform norbornenone deriva-
tive as long as it is not solvated.
tives to the corresponding anti ether.
Conclusions
Computational Methods
The Gaussian 03 program[33] was used for density functional theory
calculations, employing Becke’s[34] three-parameter hybrid method,
and the exchange functional of Lee, Yang and Parr (B3LYP).[35]
Geometries were optimized at the B3LYP/6-31G(d) level of theory.
The stationary points were characterized by vibrational analysis.
The values discussed in the text represent enthalpies at 298 K
(H298). Unless otherwise stated, all values in the text refer to
B3LYP/6-31G(d) calculations. Quantification of donor/acceptor in-
teractions has been made by second-order perturbation analysis as
incorporated in the Natural Bond Orbital (NBO) method.[36] The
results concerning the reactions of ylide 21 were confirmed using
single point calculations at the CCSD(T)/6-31G(d) level of theory.
This method represents a coupled cluster calculation using single
and double substitutions augmented by a non-iterative treatment
of triple excitations.[37]
In conclusion, it has been shown in this study that car-
bonyl ylides resulting from the addition of alkyl carbenes to
esters are extremely unstable species. Their decomposition
occurs usually before rotation around a single bond takes
place and therefore, each conformer has to be investigated
on its own. This means that depending on the generation
method, different conformers with different stabilities are
produced, leading to different product distributions. For ex-
ample, decomposition of oxadiazoline 6 leads to “long-
lived” ylide E-21, whereas the other conformer Z-21 disag-
gregates rapidly and is an interesting intermediate for per-
forming transcarbenations. Overall, noncovalent interac-
tions (mainly a hydrogen bond from the acidic C–H bond
and the carbon lone pair) between an alkyl carbene and an
ester lead to a complex which has a similar stability than
the carbonyl ylide. Ylide 16, resulting from the addition of
a stabilized-nucleophilic carbene to methyl acetate is even
less stable. Nevertheless, we were able to trap syn ylide 16a
with methanol, resulting in the isolation of syn alcohol 9.
This represents an experimental evidence for the thermal
generation of an ylide from a methoxylated oxadiazoline.
Oxadiazoline 1 is shown to be a very convenient carbene
precursor due to the simplicity of its synthesis and the clean
thermal generation of alkyl carbene 4 without prior forma-
tion of a reactive diazo species. Every thermolysis led exclu-
sively to the anti ether confirming the active participation
of the double bond in the stabilization of the intermediates
and the transition states. In the same way, the experimental
results show that ether formation from oxadiazoline 1 pro-
ceeds at least partly through the intermediacy of a carben-
Experimental Section
General: Melting points are uncorrected. 1H and 13C NMR spectra
were recorded on a 400 MHz spectrometer. The chemical shifts at
δ = 7.26 ppm and 77.0 ppm of CHCl3 were used as internal stan-
dards for 1H and 13C spectra. Conventional 2D COSY, NOESY,
HMBC, and HMQC spectra were used to derive proton and car-
bon assignments.
The synthesis of oxadiazoline 1 has already been described.[5]
endo,syn-Tricyclo[6.2.1.02,7]undec-9-en-11-ol (9): Oxadiazoline 1a
(302 mg, 1.22 mmol) was thermolyzed overnight in 4 mL of meth-
anol in a pressure tube at 155 °C. After removal of the solvent, syn
alcohol 9 was separated from the main product, anti methyl ether
10me by preparative TLC with hexane/2-propanol (39:1). Yield
40 mg (0.244 mmol, 20%); m.p. 81–82 °C. 1H NMR (400 MHz,
CDCl3): δ = 6.14–6.12 (m, 2 H), 3.87 (d, J = 11.0 Hz, 1 H), 2.72
ium ion even under apolar conditions. From calculations of (q, J = 1.6 Hz, 2 H), 2.14 (d, J = 11.3 Hz, 1 H), 2.02–1.95 (m, 2
Eur. J. Org. Chem. 2008, 5336–5345
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5343

U. H. Brinker et al.
H), 1.58–1.51 (m, 2 H), 1.35–1.26 (m, 4 H), 0.90–0.77 (m, 2 H) rac-(1R,2R,3S,4R,5R,6S,7S)-5-Methoxytetracyclo[5.4.0.02,4.03,6]un-
FULL PAPER
ppm. 13C NMR (100 MHz, CDCl3): δ = 132.7, 87.7, 53.2, 36.4,
decane (11me): 1H NMR (400 MHz, CDCl3, main peaks): δ = 3.87
(dd, J = 7.2, 3.8 Hz, 1 H), 3.20 (s, 3 H), 2.29–2.21 (m, 2 H), 1.95–
1.90 (m, 1 H), 1.81–1.75 (m, 1 H) ppm. 13C NMR (100 MHz,
CDCl3, main peaks): δ = 72.3, 54.5, 44.1, 36.4, 24.9, 22.9, 22.3,
22.2, 21.7, 17.6 ppm.
22.1, 19.9 ppm. IR: ν = 3316, 3061, 2926, 1416, 1330, 1288, 1248,
˜
1194, 1040, 986, 908, 884, 836, 772 cm–1. MS (70 eV): m/z (%) =
164 (24) [M+], 146 (79), 135 (100), 133 (79), 121 (43), 107 (50), 91
(96), 79 (41), 67 (18). HRMS (70 eV): calcd. for C11H16O 164.1201,
found 164.1198.
rac-(1R,2R,3R,6S,7S)-3-Methoxytricyclo[5.4.0.02,6]undec-4-ene
(12me): Oxadiazoline 1 (300 mg, 1.21 mmol) was dissolved in 4 mL
of methanol and photolyzed overnight. The solvent was rotary-
evaporated. The crude product contained 48% of 10me and 43%
of 12me. 12me was obtained from repeated chromatography with
hexane/Et2O (90:10) as eluent. Yield 68 mg (32%, 0.382 mmol) oil.
1H NMR (400 MHz, CDCl3): δ = 6.22 (ddd, J = 5.6, 2.4, 0.8 Hz,
1 H), 6.07–6.03 (m, 1 H), 4.34 (t, J = 2.5 Hz, 1 H), 3.48–3.41 (m,
1 H), 3.26 (s, 3 H), 2.85 (dd, J = 8.9, 6.3 Hz, 1 H), 2.65–2.49 (m,
2 H), 1.55–1.44 (m, 2 H), 1.43–1.21 (m, 5 H), 1.15–1.04 (m, 1 H)
ppm. 13C NMR (100 MHz, CDCl3): δ = 140.2, 132.9, 86.8, 55.2,
anti,endo-11-Benzoxytricyclo[6.2.1.02,7]undec-9-ene (10bn): Oxadi-
azoline 1a (200 mg, 0.81 mmol) was photolyzed overnight in 2 mL
of dry ethyl acetate in the presence of benzyl alcohol (87.1 mg,
0.81 mmol). After removal of the solvent, the product was submit-
ted to preparative TLC with hexane/CH2Cl2 (95:5). Yield 102 mg
(0.402 mmol, 50%) oil. 1H NMR (600 MHz, CDCl3): δ = 7.36–
7.31 (m, 4 H), 7.30–7.26 (m, 1 H), 6.01 (t, J = 2.2 Hz, 2 H), 4.45
(s, 2 H), 3.39 (t, J = 1.9 Hz, 1 H), 2.73–2.70 (m, 2 H), 2.24–2.17
(m, 2 H), 1.63–1.58 (m, 2 H), 1.45–1.36 (m, 4 H), 0.99–0.90 (m, 2
H) ppm. 13C NMR (150 MHz, CDCl3): δ = 138.7, 133.8, 128.3,
127.5, 127.4, 89.8, 70.4, 48.6, 36.0, 22.7, 20.4 ppm. MS (70 eV): m/z
(%) = 254 (1) [M+], 225 (9), 162 (2), 145 (4), 133 (13), 91 (100), 77
(15), 65 (30). HRMS (70 eV): calcd. for C18H22O 254.1671, found
254.1667.
48.9, 42.5, 33.6, 32.7, 22.2, 21.8, 20.8, 20.7 ppm. IR: ν = 3046,
˜
2930, 2864, 2815, 1461, 1449, 1371, 1190 cm–1. MS (70 eV): m/z
(%) = 178 (19) [M+], 146 (20), 131 (11), 117 (15), 105 (11), 96
(100), 91 (27), 81 (34), 67 (15), 53 (22). HRMS (70 eV): calcd. for
C12H18O 178.1358, found 178.1354.
anti,endo-Tricyclo[6.2.1.02,7]undec-9-en-11-ol (10h): In a pressure
tube, oxadiazoline 1b (205 mg, 0.825 mmol) was thermolyzed over-
night at 165 °C in 6 mL of dioxane/water (5:1). After removal of the
solvent, 127 mg (0.774 mmol, 94%) of alcohol 10h were obtained.
Crystals suitable for structure analysis were obtained by recrystalli-
zation from hexane;[38] m.p. 85–87 °C. 1H NMR (400 MHz,
CDCl3): δ = 5.98 (t, J = 2.2 Hz, 2 H), 3.62 (t, J = 1.8 Hz, 1 H),
2.54 (q, J = 1.8 Hz, 2 H), 2.20–2.09 (m, 2 H), 1.97 (br. s, 1 H),
1.66–1.53 (m, 2 H), 1.45–1.30 (m, 4 H), 1.02–0.87 (m, 2 H) ppm.
13C NMR (100 MHz, CDCl3): δ = 134.0, 83.2, 50.7, 35.7, 22.5,
Supporting Information (see also the footnote on the first page of
this article): 1H and 13C NMR spectra for the new compounds,
Cartesian coordinates and energies for all relevant stationary
points.
Acknowledgments
Calculations were performed on the Schrödinger III cluster at the
University of Vienna.
20.4 ppm. IR: ν = 3265, 3060, 2926, 2861, 1458, 1321, 1248 cm–1.
˜
MS (70 eV): m/z (%) = 164 (23) [M+], 146 (33), 135 (87), 133 (60),
121 (26), 107 (40), 91 (100), 82 (73), 67 (33). HRMS (70 eV): calcd.
for C11H16O 164.1201, found 164.1203.
[1] J. Warkentin, J. Chem. Soc. Perkin Trans. 1 2000, 2161.
[2] M. W. Majchrzak, M. Békhazi, I. Tse-Sheepy, J. Warkentin, J.
Org. Chem. 1989, 54, 1842.
anti,endo-11-Methoxytricyclo[6.2.1.02,7]undec-9-ene (10me): In
a
pressure tube, oxadiazoline 1b (200 mg, 0.81 mmol) was thermo-
lyzed for 24 h at 165 °C in 5 mL of methanol. After removal of the
solvent, 138 mg (0.778 mmol, 96%) of oily 10me were obtained. 1H
NMR (400 MHz, CDCl3): δ = 6.00 (t, J = 2.2 Hz, 2 H), 3.26 (s, 3
H), 3.18 (t, J = 1.8 Hz, 1 H), 2.68 (sextet, J = 1.8 Hz, 2 H), 2.12–
2.03 (m, 2 H), 1.62–1.56 (m, 2 H), 1.44–1.32 (m, 4 H), 0.98–0.86
(m, 2 H) ppm. 13C NMR (100 MHz, CDCl3): δ = 133.8, 91.6, 56.0,
[3] a) E. L. Tae, Z. Zhu, M. S. Platz, J. P. Pezacki, J. Warkentin, J.
Phys. Chem. A 1999, 103, 5336; b) J. R. Snoonian, M. S. Platz,
J. Phys. Chem. A 2001, 105, 2106; c) A. S. K. Hashmi, M. A.
Grundl, A. R. Nass, F. Naumann, J. W. Bats, M. Bolte, Eur. J.
Org. Chem. 2001, 4705; d) Y. Takahashi, Y. Mori, A. Naka-
mura, H. Tomioka, Tetrahedron Lett. 2005, 46, 8415.
[4] J.-L. Mieusset, U. H. Brinker, J. Org. Chem. 2008, 73, 1553.
[5] J.-L. Mieusset, A. Bespokoev, M. Pacar, M. Abraham, V. B.
Arion, U. H. Brinker, J. Org. Chem. 2008, 73, 6551.
[6] a) J. Alberti, R. Siegfried, W. Kirmse, Justus Liebigs Ann.
Chem. 1974, 1605; b) W. Kirmse, T. Meinert, J. Chem. Soc.,
Chem. Commun. 1994, 1065.
48.2, 35.9, 22.6, 20.4 ppm. IR: ν = 3071, 2934, 2856, 1427, 1302,
˜
1112 cm–1. MS (70 eV): m/z (%) = 178 (7) [M+], 146 (75), 131 (30),
121 (44), 117 (18), 105 (27), 96 (18), 91 (100), 77 (19), 67 (13).
HRMS (70 eV): calcd. for C12H18O 178.1358, found 178.1350.
[7] J.-L. Mieusset, U. H. Brinker, J. Am. Chem. Soc. 2006, 128,
anti,endo-11-Trityloxytricyclo[6.2.1.02,7]undec-9-ene (10tr): Oxadi-
azoline 1a (200 mg, 0.81 mmol) was photolyzed overnight in 2 mL
of dry ethyl acetate in the presence of trityl alcohol (209.7 mg,
0.81 mmol). After removal of the solvent, the product was submit-
ted to column chromatography with hexane/CH2Cl2 (95:5). The
product could also be recrystallized from ethanol. Yield 199 mg
(0.49 mmol, 60%); m.p. 156–157 °C. 1H NMR (400 MHz, CDCl3):
δ = 7.50–7.43 (m, 6 H), 7.30–7.18 (m, 9 H), 5.77 (t, J = 2.2 Hz, 2
H), 3.35 (t, J = 1.6 Hz, 1 H), 2.31–2.23 (m, 2 H), 1.83–1.79 (m, 2
H), 1.59–1.53 (m, 2 H), 1.45–1.35 (m, 2 H), 1.30 (br. d, J = 13.5 Hz,
2 H), 0.88–0.75 (m, 2 H) ppm. 13C NMR (100 MHz, CDCl3): δ =
145.1, 133.5, 128.7, 127.7, 126.8, 86.5, 85.6, 49.7, 36.5, 22.7,
15843.
[8] a) W. Kirmse, in Advances in Carbene Chemistry (Eds.: U. H.
Brinker), Elsevier, Amsterdam, 2001, vol. 3, p. 1–51; b) W.
Kirmse, in Advances in Carbene Chemistry (Ed.: U. H.
Brinker), JAI, Greenwich, 1994, vol. 1, p. 1–57.
[9] J. R. Pliego Jr., W. B. De Almeida, J. Phys. Chem. A 1999, 103,
3904.
[10] C. Gonzalez, A. Restrepo-Cossio, M. Márques, K. B. Wiberg,
J. Am. Chem. Soc. 1996, 118, 5408.
[11] P. Ballinger, F. A. Long, J. Am. Chem. Soc. 1960, 82, 795.
[12] N. Olmstead, Z. Margolin, F. G. Bordwell, J. Org. Chem. 1980,
45, 3295.
[13] A plausible explanation for the fact that tetracycle 11 is formed
preferably under apolar conditions is that the tricycloundecenyl
cation is only poorly solvated and therefore more reactive and
less selective.
20.5 ppm. IR: ν = 3057, 2933, 2862, 1489, 1448 cm–1. MS (70 eV):
˜
m/z (%) = 406 (0.3) [M+], 243 (100), 165 (51), 91 (28), 77 (18), 67
(18).
5344
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 5336–5345

Precursor for a Carbene, an Ylide, or a Diazo Compound
[14] G. G. Cox, D. J. Miller, C. J. Moody, E.-R. H. B. Sie, Tetrahe-
dron 1994, 50, 3195.
[15] M. C. McMills, D. Wright, in The Chemistry of Heterocyclic
Compounds, vol. 59: Synthetic Applications of 1,3-Dipolar Cy-
cloaddition Chemistry Toward Heterocycles and Natural Prod-
ucts (Eds.: A. Padwa, W. H. Pearson), Wiley, New York, 2002,
p. 253–314.
[30]
[31]
For example: H.-G. Yu, J. T. Muckerman, J. Phys. Chem. A
2005, 109, 1890.
a) B. F. Yates, W. J. Bouma, L. Radom, J. Am. Chem. Soc.
1987, 109, 2250; b) M. Moreno, J. M. Lluch, A. Oliva, J.
Bertrán, Can. J. Chem. 1987, 65, 2774; c) L. L. Zub, J. M. Stan-
dard, THEOCHEM 1996, 368, 133; d) J. M. Tucker, J. M.
Standard, THEOCHEM 1998, 431, 193.
[16] E. P. Janulis, A. J. Arduengo III, J. Am. Chem. Soc. 1983, 105,
[32]
[33]
The same interaction exists with C(6)–H as acceptor because
of the symmetry of the structure.
5929.
[17] a) J. R. Pliego Jr., W. B. De Almeida, J. Chem. Phys. 1997, 106,
3582; b) J. R. Pliego Jr., W. B. De Almeida, J. Braz. Chem. Soc.
1998, 9, 181.
[18] See also: W. Czardybon, J. Warkentin, N. H. Werstiuk, J. Org.
Chem. 2005, 70, 8431.
[19] M. Békhazi, J. Warkentin, J. Am. Chem. Soc. 1981, 103, 2473.
[20] D. M. Pawar, A. A. Khalil, D. R. Hooks, K. Collins, T. Elliott,
J. Stafford, L. Smith, E. A. Noe, J. Am. Chem. Soc. 1998, 120,
2108.
M. J. Frisch, G. W. Trucks, H. B. Schlegel, 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, C. Adamo, J. Jaramillo, R. Gomperts,
R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pom-
elli, 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. Foresman, 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. Nanayakkara,
M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W.
Wong, C. Gonzalez, J. A. Pople, Gaussian 03, Gaussian, Inc.,
Pittsburgh PA, 2003.
[21] J. March, Advanced Organic Chemistry, Wiley, New York, 1992,
p. 21.
[22] a) R. A. Moss, U.-H. Dolling, J. R. Whittle, Tetrahedron Lett.
1971, 12, 931; b) U. H. Brinker, J. Ritzer, J. Am. Chem. Soc.
1981, 103, 2116; c) J.-L. Mieusset, U. H. Brinker, J. Org. Chem.
2007, 72, 263.
[23] For the purpose of clarity, other less relevant conformers with
similar chemistry are not discussed here, see Supporting Infor-
mation.
[24] K. N. Houk, N. G. Rondan, C. Santiago, C. J. Gallo, R. W.
Gandour, G. W. Griffin, J. Am. Chem. Soc. 1980, 102, 1504.
[25] a) E. F. Hayes, A. K. Q. Siu, J. Am. Chem. Soc. 1971, 93, 2090;
b) B. Bigot, A. Sevin, A. Devaquet, J. Am. Chem. Soc. 1979,
101, 1095; c) D. Feller, E. R. Davidson, W. T. Borden, J. Am.
Chem. Soc. 1984, 106, 2513; d) S. D. Kahn, W. J. Hehre, J. A.
Pople, J. Am. Chem. Soc. 1987, 109, 1871; e) Y. Yamaguchi,
H. F. Schaefer III, I. L. Alberts, J. Am. Chem. Soc. 1993, 115,
5790.
[34]
A. D. Becke, J. Chem. Phys. 1993, 98, 5648.
[35] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785.
[36] a) E. D. Glendening, A. E. Reed, J. E. Carpenter, F. Weinhold,
NBO, version 3.1; b) A. E. Reed, L. A. Curtiss, F. Weinhold,
Chem. Rev. 1988, 88, 899.
[37] J. A. Pople, M. Head-Gordon, K. Raghavachari, J. Chem.
Phys. 1987, 87, 5968.
[26] M. Békhazi, J. Warkentin, J. Am. Chem. Soc. 1983, 105, 1289.
[27] M. Békhazi, W. Lawrynowicz, J. Warkentin, Can. J. Chem.
1991, 69, 1507.
[28] A. P. Hitchcock, S. Zweep, T. Steel, M. Békhazi, J. Warkentin,
Can. J. Chem. 1982, 60, 2914.
[29] See also: M. K. Kindermann, K. Kowski, H. M. Muchall, P.
Rademacher, Chem. Ber. 1993, 126, 2675.
[38] CCDC-693658 (for 10h) contains the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
Received: July 17, 2008
Published Online: September 22, 2008
Eur. J. Org. Chem. 2008, 5336–5345
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5345