Bronsted acid mediated novel rearrangements of diarylvinylidenecyclopropanes and mechanistic investigations based on DFT calculations

Article abstract of DOI:10.1002/chem.200901172

Diarylvinylidenecyclopropanes undergo a novel rearrangement in the presence of the Bronsted acid Tf₂NH (Tf: trifluoromethanesulfonyl) to give the corresponding naphthalene derivatives in good to high yields upon heating, whereas in the presence

Full text of DOI:10.1002/chem.200901172

DOI: 10.1002/chem.200901172
Brønsted Acid Mediated Novel Rearrangements of
Diarylvinylidenecyclopropanes and Mechanistic Investigations Based on
DFT Calculations
Wei Li, Min Shi,* and Yuxue Li*[a]
Abstract:
Diarylvinylidenecyclopro-
responding mechanistic studies on the
basis of density functional theory
(DFT) with the Gaussian03 program
by using the B3LYP method have re-
vealed that the pKa value of the
Brønsted acid, as well as the entropy
and solvent effects, plays a significant
role in this reaction; these factors can
discriminate the differences in the reac-
tivity and regioselectivity among the
Brønsted acids used in this reaction. In
panes undergo a novel rearrangement
in the presence of the Brønsted acid
Tf2NH (Tf: trifluoromethanesulfonyl)
to give the corresponding naphthalene
derivatives in good to high yields upon
heating, whereas in the presence of the
Brønsted acid toluene-4-sulfonic acid
(p-TSA), the corresponding triene de-
rivatives are afforded in moderate to
good yields under mild conditions. Cor-
Keywords: Brønsted acids
arylvinylidenecyclopropanes · Lewis
acids naphthalenes reaction
regioselectivity ·
·
di-
AHCTUNGTRENNUNG
the presence of Lewis acid SnACHTUNGTRENNUNG(OTf)2, a
·
·
butatrienecyclopane can produce the
corresponding ring-opened products in
moderate yields.
mechanisms
trienes
·
Introduction
Brønsted acid catalyzed/mediated ring-opening reactions of
1 with various electrophiles and have found some novel re-
action patterns of these substrates.[3,4] As an interesting ex-
ample, we have found that naphthalene derivative 2a was
Vinylidenecyclopropanes, 1, are one of the most remarkable
organic compounds that are known in the research field of
highly strained small rings. They have an allene moiety con-
nected to a cyclopropyl ring and yet they are thermally
stable and reactive substances in organic synthesis. Thus far,
thermal and photochemical skeletal conversions of vinylide-
necyclopropanes have attracted much attention from mecha-
nistic, theoretical, spectroscopic, and synthetic viewpoints.[1,2]
Many intriguing transformations with these interesting sub-
stances have been explored under mild conditions. More-
over, we have recently been investigating the Lewis acid or
formed in 50% yield in the Sn
ACHTUNGTRENNUNG
diphenylvinylidenecyclopropane 1a, bearing
4
groups on the cyclopropane ring, in 1,2-dichloroethane
(DCE) upon heating at 808C (Scheme 1).
However, because the reaction mechanism remains ob-
scure and Lewis acids such as BF3·OEt2, Zr
(OTf)2, and lanthanide triflates (Ln(OTf)3; Ln: Yb, Sc, and
ACHTUNGRTENN(UNG OTf)4, Sn-
A
ACHTUNGTRENNUNG
Y) produce naphthalene derivative 2a in only moderate
yields,[3b] we attempted to examine the Brønsted acid medi-
ated reactions of diarylvinylidenecyclopropanes 1 under sim-
ilar conditions and to determine the detailed reaction mech-
anism. In this paper, we wish to report the Brønsted acid
mediated rearrangement of diarylvinylidenecyclopropanes 1
[a] Dr. W. Li, Prof. Dr. M. Shi, Dr. Y. Li
State Key Laboratory of Organometallic Chemistry
Shanghai Institute of Organic Chemistry
Chinese Academy of Sciences
354 Fenglin Lu, Shanghai 200032 (China)
Fax : (+86)21-64166128
E-mail: Mshi@mail.sioc.ac.cn
liyuxue@mail.sioc.ac.cn
Supporting information for this article is available on the WWW
1
under http://dx.doi.org/10.1002/chem.200901172. It contains H and
13C NMR spectra for the rearranged products 2, 3a, 4, 5, and 7, the
GC–MS spectra of the propene capture, the NMR study of 7a, the
calculated NMR spectra for 7a, the optimized structures for the com-
pounds in Scheme 3, and the total energies and geometrical coordi-
nates of the structures in Scheme 3 and Figure 2.
Scheme 1. Rearrangement of diphenylvinylidenecyclopropane 1a in the
presence of the Lewis acid SnACHTNUTRGNEUNG(OTf)2 in DCE at 808C.
8852
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 8852 – 8860

FULL PAPER
to form the corresponding naphthalene derivatives 2 or
triene derivatives 4 in good to high yields under mild condi-
tions, along with an investigation of the reaction mechanism
by a DFT study, which has been supported by the detection
of the released propene molecule. In addition, we will dis-
close in this paper that the pKa value of the employed
Brønsted acid, entropy, and solvent effects can significantly
affect the reactivity and regioselectivity of the reaction.
temperature (208C), 2a was produced in 71% yield as the
sole product after 3 days (Table 2, entry 1). Solvent effects
were also examined with Tf2NH (1.0 equiv) as the promoter,
Table 2. Brønsted acid promoted rearrangement of diphenylvinylidene-
cyclopropane 1a in DCE at different temperatures.
Entry
Brønsted acid
t
T [8C]
Yield [%][a]
2a
3a
Results and Discussion
1
Tf2NH
Tf2NH
Tf2NH
Tf2NH
Tf2NH
Tf2NH
Tf2NH
Tf2NH
Tf2NH
Tf2NH
p-TSA
p-TSA
MeSO3H
CF3CO2H
3 d
4 d
2 d
7 d
20
20
20
20
20
50
84
84
84
0
71
–
–
–
–
–
–
–
–
–
–
–
16
–
–
–
2[b]
3[c]
4[d]
5[e]
6
An initial examination was performed by using diphenylvi-
nylidenecyclopropane 1a as a substrate in the presence of
the Brønsted acid trifluoromethanesulfonic acid (TfOH)
(pKa =4.2 in acetic acid at 208C)[5] in DCE, and the results
of these experiments are summarized in Table 1. We found
87
40
52
92
99
76
92
50
71
93
85
71
3 d
11 h
10min
90 min
10 min
11 d
10 min
5 h
7
8[f]
9[g]
10
11
12
13
14
Table 1. TfOH-catalyzed/mediated rearrangement of diphenylvinylidene-
cyclopropane 1a in a variety of solvents.
84
50
50
50
11 h
2 d
[h]
15
4 d
50
–
[h]
Entry Brønsted acid
Solvent
t
Yield [%][a]
16
17
4 d
4 d
50
50
–
2a
3a
[h]
–
1
2
3
4
5
6
7
HOTf (1.0 equiv) DCE
HOTf (0.1 equiv) DCE
2 min 34
complex mixture
4 d
6 h
2 h
2 min 62
4 d
4 d
26
40
40
40
42
45
33
–
[h]
HOTf (0.2 equiv) DCE
HOTf (0.4 equiv) DCE
HOTf (0.6 equiv) DCE
HOTf (0.6 equiv) THF
HOTf (0.6 equiv) Et2O
18
19
4 d
4 d
50
50
–
[h]
trace
trace
–
–
[a] Yield of isolated product.
[h]
20
4 d
50
–
that the corresponding naphthalene derivative 2a was pro-
duced under various conditions, along with enyne product
3a. With TfOH (1.0 equiv) as a promoter in DCE at room
temperature (208C), the reaction proceeded extremely
quickly to give 2a in 34% yield within 2 min, along with a
complex mixture of many unidentified products and 3a
(Table 1, entry 1). However, with less TfOH (0.1 equiv) as
the promoter under otherwise identical conditions, 2a was
obtained in 26% yield, along with 3a in 40% yield, after
4 days (Table 1, entry 2). An increase in the employed
amount of TfOH significantly accelerated the reaction rate
(Table 1, entries 3–5). In the presence of TfOH (0.6 equiv),
2a was produced in 62% yield, along with 33% yield of 3a
(Table 1, entry 5). However, the reaction did not take place
in Et2O or tetrahydrofuran (THF), presumably because the
oxygen atom in these solvents could capture the proton of
TfOH and impair the catalytic ability of TfOH (Table 1, en-
tries 6 and 7).
[a] Yield of isolated product. [b] With n-pentane as the solvent. [c] With
dichloromethane as the solvent. [d] With benzene as the solvent. [e] With
chloroform as the solvent. [f] With 0.1 equiv of Tf2NH. [g] With 0.5 equiv
of Tf2NH. [h] No reaction occurred.
and the results are listed in entries 2–5 of Table 2. In di-
chloromethane, 2a was formed in 87% yield as the sole
product (Table 2, entry 3), but in other solvents, such as n-
pentane, benzene, or chloroform, 2a was produced in lower
yields (Table 2, entries 2, 4, and 5). We next examined the
influence of the reaction temperature with Tf2NH as the
promoter. We found that 2a could be produced in 92 and
99% yields at 508C and under reflux conditions (848C)
after 11 h and 10 min, respectively (Table 2, entries 6 and 7).
With a catalytic amount of Tf2NH as the promoter, a pro-
longed reaction time was required and the yields of product
2a were lower (Table 2, entries 8 and 9). At 08C, the reac-
tion was sluggish under otherwise identical conditions to
those in entry 7 of Table 2, and 2a was produced in 50%
Interestingly, with Tf2NH (1.0 equiv; pKa =7.8 in acetic
acid at 208C)[5] as the promoter in this reaction at room
Chem. Eur. J. 2009, 15, 8852 – 8860
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8853

M. Shi, Y. Li, and W. Li
yield after 11 days (Table 2, entry 10). On the other hand,
with the Brønsted acid p-TSA as the promoter, 2a was ob-
tained in 71% yield, along with 3a in 16% yield, at 848C
after 10 min (Table 2, entry 11). At 508C, 2a was obtained
in 93% yield as the sole product after 5 h (Table 2,
entry 12). Other Brønsted acids, such as MeSO3H (pKa >7.0
in acetic acid at 208C)[5] and CF3CO2H (pKa =11.4 in acetic
acid at 208C),[5] are not as effective as Tf2NH under similar
conditions (Table 2, entries 13 and 14). A sterically more
hindered aromatic sulfonic acid provided 2a in 59% yield,
along with 3a in 4% yield, at 508C (Table 2, entry 15).
Weak Brønsted acids, such as aromatic carboxylic acids, did
not catalyze this reaction (Table 2, entries 16–20). Therefore,
the pKa value of the employed Brønsted acid significantly
affected the reaction outcome. The best conditions for the
formation of the corresponding naphthalene derivative 2a
were found to be those in DCE at 848C with 1.0 equiv of
Tf2NH (Table 2, entry 7).
2,4-dimethyl-3-(2’,2’-diphenylvinyl)-penta-1,3-diene (4a) as
the sole product in 76% yield after 5 days, rather than prod-
ucts 2a or 3a (Table 4, entry 1).[7] With hexane, toluene,
Table 4. Rearrangement of diphenylvinylidenecyclopropane 1a promot-
ed by the Brønsted acid p-TSA in a variety of solvents.
Entry
Solvent
T [8C]
t [d]
Yield [%][a]
2a
3a
4a
1
2
3
4
5
6
7
DCE
20
20
20
20
0
5
4
4
4
30
4
4
–
–
76
hexane
toluene
CH2Cl2
DCE
THF
Et2O
no reaction
no reaction
14
–
no reaction
no reaction
14
–
70
46
20
20
By using these optimized reaction conditions, we next ex-
amined the rearrangement of a variety of diarylvinylidene-
cyclopropanes 1. The results are shown in Table 3. From
[a] Yield of isolated product.
THF, or diethyl ether as the solvent, no reaction occurred
under otherwise identical conditions (Table 4, entries 2, 3, 6,
and 7). When this reaction was carried out in dichlorome-
thane, 4a was formed in 70% yield, along with 2a and 3a,
both in 14% yields (Table 4, entry 4). It should be noted
that when this reaction was carried out at 08C in DCE with
p-TSA as the promoter, the reaction became extremely slug-
gish and gave 4a in 46% yield after 30 days.
Table 3. Rearrangement of a variety of diarylvinylidenecyclopropanes 1
promoted by the Brønsted Acid Tf2NH in DCE at 848C.
We examined the rearrangement of a variety of diarylvi-
nylidenecyclopropanes 1 with p-TSA as the promoter in
DCE. The results are summarized in Table 5. From symmet-
Entry Substrate R1
R2
R3
t
Product, yield [%][a]
1
2
3
4
5
6
7
1b
1c
1d
1e
1 f
1g
1h
Me
Me
H
H
H
H
H
H
10 min 2b, 84
10 min 2c, 99
MeO MeO
Cl
F
H
H
H
Cl
F
Cl
MeO
Me
2 h
2 h
2d, 60
2e, 62
Table 5. Rearrangement of a variety of diarylvinylidenecyclopropanes 1
promoted by the Brønsted acid p-TSA in DCE at room temperature.
10 min 2 f + 2 f’, 88[b]
10 min 2g + 2g’, 99[b]
Me 20 min 2h + 2h’, 67[b]
[a] Yield of isolated product or product mixtures. [b] The ratio of 2 and 2’
was determined by 1H NMR spectroscopic data. 2 f/2 f’ 1:1, 2g/2g’ 1:1,
2h/2h’ 1:1.8.
Entry
Substrate
R1
R2
R3
t
Product, yield [%][a]
symmetrical diarylvinylidenecyclopropanes 1b and 1c, with
electron-donating groups on the benzene rings, the corre-
sponding naphthalene derivatives 2b and 2c were obtained
in high yields (Table 3, entries 1 and 2). From symmetrical
diarylvinylidenecyclopropanes 1d and 1e, with electron-
withdrawing groups on the benzene rings, the corresponding
naphthalene derivatives 2d and 2e were formed in moder-
ate yields (Table 3, entries 3 and 4). From unsymmetrical di-
arylvinylidenecyclopropanes 1 f–h, the corresponding naph-
thalene derivatives were obtained as isomeric mixtures of 2
and 2’ in good total yields (Table 3, entries 5–7).
1
2
3
4
1b
1c
1d
1e
1 f
1g
1h
Me
MeO
Cl
F
H
Me
MeO
Cl
F
Cl
H
H
H
H
H
H
Me
1 d
7 h
4b, 84
4c, 92
4d, 69
4e, 78
4 f + 4 f’, 35
4g + 4g’, 98
4h’, 35
12 d
5 d
14 d
12 h
2 d
5[b]
6[b]
7[c]
H
H
MeO
Me
[a] Yield of isolated product or product mixture. [b] 4 f/4 f’ 3:2, 4g/4g’
1:1. [c] Only 4h’ was obtained, as determined from the NOESY spec-
trum, along with 42% of 2h’.
Interestingly, with the moderate Brønsted acid p-TSA
(1.0 equiv) as the promoter at room temperature (208C),
the rearrangement of 1a proceeded slowly in DCE, to give
rical diarylvinylidenecyclopropanes 1b–e, the corresponding
penta-1,3-diene derivatives 4b–e were produced in moder-
ate to good yields, although the reaction proceeded slowly
8854
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 8852 – 8860

Brønsted Acid Mediated Rearrangements of Diarylvinylidenecyclopropanes
FULL PAPER
for diarylvinylidenecyclopropanes 1d and 1e with electron-
withdrawing groups on the benzene rings (Table 5, en-
tries 1–4). From unsymmetrical diarylvinylidenecyclopro-
panes 1 f and 1g, the corresponding penta-1,3-diene deriva-
tives were obtained as isomeric mixtures of 4 and 4’ in mod-
erate to good total yields (Table 5, entries 5 and 6). In the
case of diarylvinylidenecyclopropane 1h, the corresponding
penta-1,3-diene derivative 4h’ was formed, along with the
corresponding naphthalene derivative 2h’, both in moderate
yields (Table 5, entry 7).
Scheme 2. Rearrangement of diarylvinylidenecyclopropane 1j promoted
by the Brønsted Acid CF3CO2H in DCE.
We also synthesized two other diarylvinylidenecyclopro-
panes, 1i and 1j, which have two cycloalkyl groups on the
cyclopropane ring, to examine their rearrangement behavior
in the presence of various Brønsted acids under the standard
conditions. The results of these experiments are summarized
in Table 6. We found that, in the presence of a strong
action pathway. That is, the stronger Brønsted acids promot-
ed reactions faster, whereas the weaker Brønsted acids
could not promote reactions at all. Moreover, the excellent
selectivity of Tf2NH could be explained by the increased
Brønsted acidity of Tf2NH in a polar aprotic solvent, com-
bined with the low nucleophilicity of the trifluoromethane-
sulfonimide anion.[6] However, it is clear that the reaction
mechanism is much more complicated to understand and
that a comprehensive mechanistic explanation is difficult to
achieve.
Table 6. Optimization of the reaction conditions of the Brønsted acid
promoted rearrangement of diarylvinylidenecyclopropane 1i in DCE.
To understand the mechanism of this reaction, DFT[7]
studies have been performed with the Gaussian03 program[8]
by using the B3LYP[9] method. The reaction pathway has
been studied at the B3LYP/6-31+G** level. To obtain more
accurate energies, the 6-311++G** basis set was used in
the study of the regioselectivity. The solvent effects were es-
timated with the IEFPCM[10] (UAHF atomic radii) method
in DCE (dielectric constant e = 10.36) based on the gas-
phase fully optimized structures. For each structure, har-
monic-vibration-frequency calculations were carried out and
thermal corrections were made. All structures were shown
as transition states (with one imaginary frequency) or sta-
tionary points (with no imaginary frequency).
Entry
Acid
T [8C]
t
Yield [%][a]
1
2
3
4
5
6
Tf2NH (1.0 equiv)
p-TSA (1.0 equiv)
CF3CO2H (2.0 equiv)
CF3CO2H (2.0 equiv)
CF3CO2H (1.2 equiv)
80
20
20
20
80
20
<5 min
6 d
40 h
96 h
10 h
3 h
complex
10
52
38
65
45
SnACHTUNGTRENNUNG(OTf)2 (0.2 equiv)
[a] Yield of isolated product.
Brønsted acid, such as Tf2NH, complicated product mixtures
were formed (Table 6, entry 1). The moderately strong
Brønsted acid p-TSA produced the 1-(1-cyclohexylideneal-
lyl)cyclohexene derivative 5a in 10% yield, along with other
unidentified products (Table 6, entry 2). Fortunately, with
trifluoroacetic acid (2.0 equiv) as the promoter at room tem-
perature, 5a was obtained in 52% yield after 40 h and 38%
yield after 96 h, presumably due to the partial decomposi-
tion of 5a during the prolonged reaction time (Table 6, en-
tries 3 and 4). The use of CF3CO2H (1.2 equiv) under other-
wise identical conditions produced 5a in 65% yield at 808C
(Table 6, entry 5). It should be noted that with SnACTHUNRGTNEUNG(OTf)2 as
a Lewis acid in this reaction, 5a was produced in 45% yield
after 3 h at room temperature (Table 6, entry 6).
With diarylvinylidenecyclopropane 1j, the corresponding
1-(1-cyclopentylidene-allyl)cyclopentene derivative 5b was
obtained in 50% yield at 508C by using CF3CO2H
(1.0 equiv) as the promoter (Scheme 2).
The reaction pathway: The reaction pathway has been ex-
plored with R, the complex of HOTf and substrate 1a
(Scheme 3). All of the optimized structures are collected in
the Supporting Information, and some selected structures
are shown in Figure 1. As shown in Scheme 3, the C1 or C2
atom is protonated along paths 1 and 2, which lead to prod-
uct 3a and intermediate B, respectively. The protonation
steps are rate determining in paths 1 and 2. Therefore, the
regioselectivity is determined by the relative free energies of
TS-R-A1 and TS-R-A2. In this case, the energy difference is
2.0 kcalmolÀ1, which is somewhat large because mixed prod-
ucts have been obtained in the HOTf-promoted reactions
(Table 1). The regioselectivity of four kinds of acids was
studied at the more accurate B3LYP/6-311++G** level
and will be discussed in the next section.
À
In the cationic intermediate A1 (Figure 1), the C3 C5
bond is partially broken. The cationic C5 atom has a strong
interaction with the electron-rich C3 atom, as indicated by
the short distance of 1.810 ꢁ between them. The C1–C4
À
À
Theoretical study: A reasonable explanation for the intrigu-
ing results described above could basically be that the pKa
value of the employed Brønsted acids can dominate the re-
atoms are nearly in a straight line. The angle of C3 C4 C5
À
À
is 74.88. In TS-A1–3a, the angle of C3 C4 C5 increases to
94.88; the deprotonation of the methyl group along with the
Chem. Eur. J. 2009, 15, 8852 – 8860
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8855

M. Shi, Y. Li, and W. Li
Scheme 3. Reaction pathways and the relative free energies including the solvent effects, DGsol (in kcalmolÀ1, 298 K). These were calculated at the
B3LYP/6-31+G** level of theory.
À
cleavage of the C3 C5 bond yields the product 3a. In 3a,
rivative 2a. In the final step, the protonated propene releas-
es a proton to the OTfÀ anion.
À
À
the bond angle of C3 C4 C5 increases to 107.38, and the cy-
clopropane is entirely opened. In TS-A2-B, the cleavage of
The experiments show that 4a is not obtained under the
conditions mediated by HOTf. In addition, 4a can be trans-
formed into 2a in quantitative yield (see Scheme 4 below).
These results are consistent with the calculation results that
show 4a to be a kinetic product (Scheme 3). Low reaction
temperatures and weaker Brønsted acids are beneficial for
the formation of 4a. Our experiments show that 4a will be
obtained when the reaction is promoted by p-TSA at room
temperature (208C). These facts indicate that the properties
of the conjugated base of the Brønsted acid are important in
the reaction.
À
the C4 C5 bond leads to the formation of intermediate B.
Intermediate B may be deprotonated by OTfÀ anions to
produce product 4a (Path 3) or to form intermediate C
through an intramolecular Friedel–Crafts reaction (Path 4).
Partial aromatization of
C over a small barrier of
0.9 kcalmolÀ1 (TS-C-D) produces intermediate D, which is
more stable than B by 10.4 kcalmolÀ1. Intermediate D un-
dergoes subsequent protonation (TS-D-E), 1,2-migration of
the methyl group (TS-E-F), and release of a protonated pro-
pene (TS-F-G) to produce the aromatized naphthalene de-
8856
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 8852 – 8860

Brønsted Acid Mediated Rearrangements of Diarylvinylidenecyclopropanes
FULL PAPER
Figure 1. Selected optimized structures on the reaction pathway. The bond lengths are given in ꢁngstroms, and the bond angles are given in degrees. The
relative free energies, including the solvent effects, DGsol, (298 K) are given in kcalmolÀ1. The geometries were fully optimized at the B3LYP/6-31+G**
level.
The reactivity and the regioselectivity: The optimized reac-
tants and the transition states of the two protonation transi-
tion states (TS-R-A1 and TS-R-A2) for Tf2NH, HOTf,
MeSO3H, and p-TSA are shown in Figure 2, and the corre-
sponding data are listed in Table 7. The reaction barriers
show that the sequence of reactivity is almost the same as
that of the acidity of the employed acids: HOTf (9.4 kcal
molÀ1)>p-TSA (13.8 kcalmolÀ1)>Tf2NH (14.5 kcalmolÀ1)>
MeSO3H (17.0 kcalmolÀ1). The Tf2NH acid shows stronger
acidity than MeSO3H in this reaction.
3.4 kcalmolÀ1 (Table 7, entries 1–3). For HOTf, the differ-
ence in the DE values between HOTf-TS-R-A1 and HOTf-
TS-R-A2 is also 2.2 kcalmolÀ1. The relative entropy of
HOTf-TS-R-A2 decreased by À2.607 units, and the differ-
ence in the DGgas values decreased to 1.9 kcalmolÀ1. HOTf-
TS-R-A1 has a relatively larger solvation effect. Subse-
quently, the difference in the DGsol values decreased to
0.4 kcalmolÀ1 (Table 7, entries 7–9). Therefore, the entropy
effect and the solvation effect are important for the selectiv-
ity.
As shown in Table 7, the difference in the relative free en-
ergies, DGsol, between HOTf-TS-R-A1 and HOTf-TS-R-A2
is only 0.4 kcalmolÀ1, whereas for other acids, the energy
differences are larger than 2.6 kcalmolÀ1. These results are
consistent with the experimental observation that the use of
HOTf as the promoter always leads to mixed reaction prod-
ucts.
It is very notable that, despite similar structures, HOTf
and MeSO3H have very different behaviors in this reaction.
The data shown in Table 7 indicate that the difference in the
relative electronic energies, DE, between MeSO3H-TS-R-A1
and MeSO3H-TS-R-A2 is 2.2 kcalmolÀ1. MeSO3H-TS-R-A2
has larger entropy, and the difference in the DGgas values is
increased to 3.8 kcalmolÀ1. The difference in the solvation
effects of MeSO3H-TS-R-A1 and MeSO3H-TS-R-A2 is
0.4 kcalmolÀ1. Thus, the difference in the DGsol values is
In order to confirm that both path 3 and path 4 are in-
volved in the reaction, the control experiment shown in
Scheme 4 was carried out. We found that, under the stan-
dard conditions, triene product 4a can be smoothly trans-
formed into the corresponding naphthalene derivative 2a in
quantitative yield (Scheme 4). This result suggests that 2a
could indeed be formed from protonation of 4a through the
mechanism shown as path 4 in Scheme 3.
To identify the released propene, we used GC–MS to
detect this small molecule in the reaction system of the
Tf2NH-promoted rearrangement of 1a in DCE. On the
basis of the mass spectrum, propene was indeed detected
(see spectrum in the Supporting Information). Therefore,
we can confirm that, in the Tf2NH-promoted rearrangement
of diarylvinylidenecyclopropanes 1 bearing four methyl
groups at the cyclopropane ring, the corresponding naphtha-
Chem. Eur. J. 2009, 15, 8852 – 8860
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8857

M. Shi, Y. Li, and W. Li
Figure 2. The optimized reactants and transition states of the two protonation steps for the Tf2NH, HOTf, MeSO3H, and p-TSA acids. The geometries
were fully optimized at the B3LYP/6-311++G** level. The relative free energies, including the DGsol values, are given in kcalmolÀ1
.
Table 7. The relative electronic energies (DE, in kcalmolÀ1), relative free
energies in the gas phase (DGgas, in kcalmolÀ1), relative free energies in
solvation (DGsol, in kcalmolÀ1), and relative entropies (DS, in calmolKÀ1
)
of TS-R-A2 and TS-R-A1 for the Tf2NH, HOTf, MeSO3H, and p-TSA
acids.
Entry
Acid
DE
DGgas
DGsol
DS
1
2
3
MeSO3H-TS-R-A1
MeSO3H-TS-R-A2
MeSO3H-R
24.2
22.0
0.0
25.4
21.6
0.0
20.4
17.0
0.0
À10.671
À10.276
0.0
Scheme 4. Transformation of 4a into 2a promoted by Tf2NH.
4
5
6
Tf2NH-TS-R-A1
Tf2NH-TS-R-A2
Tf2NH-R
22.0
16.9
0.0
22.4
17.2
0.0
17.7
14.5
0.0
À14.034
À13.122
0.0
lene derivatives 2 were produced, along with the release of
a propene molecule.
We also investigated the rearrangement reaction of buta-
trienecyclopane 6[11] catalyzed by a variety of Brønsted or
Lewis acids. However, it was found that compound 6 was
extremely labile and underwent a polymerization reaction in
7
8
9
10
11
12
HOTf-TS-R-A1
HOTf-TS-R-A2
HOTf-R
p-TSA-TS-R-A1
p-TSA-TS-R-A2
p-TSA-R
17.8
15.6
0.0
24.8
22.2
0.0
15.9
14.0
0.0
22.7
18.1
0.0
9.8
9.4
0.0
16.5
13.8
0.0
À3.803
À6.410
0.0
5.456
4.451
0.0
8858
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 8852 – 8860

Brønsted Acid Mediated Rearrangements of Diarylvinylidenecyclopropanes
FULL PAPER
the presence of Brønsted acids; Lewis acids were suitable
catalysts in this case. The results of the Lewis acid catalyzed
rearrangement are shown in Table 8. With ZrACTHNURGTNEUNG(OTf)4 as the
tatriene and allene moieties connected with cyclopropane
rings can significantly affect the rearrangement pathway in
the presence of Brønsted or Lewis acid.
Table 8. Lewis acid catalyzed rearrangement of butatrienecycloproane 6
in DCE.
Conclusions
In this paper, we have disclosed the transformations of dia-
rylvinylidenecyclopropanes 1 in the presence of Brønsted
acids Tf2NH or p-TSA. In these transformations, the corre-
sponding naphthalene derivatives 2 or trienes 4, respectively,
can be obtained in moderate to good yields. A wide range of
substitutent aromatic groups on diarylvinylidenecyclopro-
panes 1 have been examined in both reactions. A plausible
reaction mechanism has been discussed on the basis of the
capture of the propene by GC–MS and the transformation
of 4a into 2a. More importantly, a mechanistic study has
been performed with DFT calculations. It was found that
the pKa value of the Brønsted acid, as well as the entropy
and solvent effects, played a significant role in this reaction;
these factors can discriminate the differences in the reactivi-
ty and regioselectivity among the Brønsted acids used in this
reaction. These processes provide an efficient route to the
synthesis of naphthalene derivatives and trienes. In addition,
the investigation of the reaction mechanism has proven our
hypothesis to a certain extent. Furthermore, we have also
disclosed the rearrangement reaction of butatrienecyclopane
6 catalyzed by Lewis acids, to provide ring-opened products
in moderate yields. Efforts are in progress to further eluci-
date the mechanistic details of this reaction and to disclose
its scope and limitations.
Entry[a]
Lewis acid
t
Yield [%][b]
7a
7b
1
2
3
Zr
Sn
U
10 min
5 min
3 h
24 h
19 d
19 d
29
54
42
20
4
11
12
6
10
4
In
Sc
La
Yb
ACHTUNGTRENNUNG
4[c]
5
G
E
6
N
10
16
[a] All reactions were carried out at room temperature with 10 mol% of
catalyst. [b] Yield of isolated product. [c] The reaction was carried out at
room temperature for 18 h, then at 508C for 6 h.
catalyst at room temperature, the rearrangement proceeded
rapidly in DCE to give the ring-opened products 7a and 7b
in 29 and 11% yields, respectively, along with some poly-
merized byproducts (Table 8, entry 1). After screening vari-
ous Lewis acids, we found that Sn
for this reaction; it gave 7a and 7b in 54 and 12% yields
under similar conditions (Table 8, entry 2). La(OTf)3 and
Yb(OTf)3 are less effective in this reaction under the stan-
ACHTUNGRTEN(NUNG OTf)2 is the best catalyst
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
dard conditions (Table 8, entries 5 and 6). The structures of
1
7a and 7b were determined by H NMR, 13C NMR, DEPT,
1H–1H NOESY (Scheme 5), HMQC, HMBC, MS–HRMS,
Experimental Section
General remarks: 1H NMR spectra were recorded on a 300 MHz or
400 MHz spectrometer in CDCl3 by using tetramethylsilane (TMS) as the
internal standard. Infrared spectra were measured on a Perkin–Elmer
983 spectrometer. Mass spectra were recorded with a HP-5989 instru-
ment, and HRMS (EI) was performed with a Finnigan MA+ mass spec-
trometer. Satisfactory CHN microanalyses were obtained with a Carlo–
Erba 1106 analyzer. Melting points are uncorrected. All reactions were
monitored by TLC with Huanghai GF254 silica gel coated plates. Flash
column chromatography was carried out with 300–400 mesh silica gel.
General procedure for the reactions of diarylvinylidenecyclopropanes in
the presence of (CF3SO2)2NH: Under an argon atmosphere, diarylvinyli-
denecyclopropanes (0.18 mmol) and DCE (3.0 mL) were added into a
Schlenk tube. The mixture was heated under reflux, and then
(CF3SO2)2NH (51 mg, 1.0 equiv) was added. The reaction was quenched
by addition of anhydrous NaHCO3, and then the product mixture was pu-
rified by flash column chromatography.
Scheme 5. 1H–1H NOE interactions of 7a and 7b.
1
Compound 2b: A white solid; m.p. 64–668C; H NMR (CDCl3, 300 MHz,
and IR spectroscopic data. Furthermore, the structure of 7a
was also confirmed on the basis of computational calcula-
tions of the 13C NMR chemical shifts by the gauge including
atomic orbital (GIAO)[12] method at the B3LYP/6-31G*
level by using the Gaussian03 program (see the Supporting
Information). Although the reaction outcomes are quite dif-
ferent from those of diarylvinylidenecyclopropanes 1 under
similar reaction conditions, this result suggests that 1,2,3-bu-
TMS): d=2.43 (s, 3H, CH3), 2.48 (s, 3H, CH3), 2.52 (s, 3H, CH3), 2.60 (s,
3H, CH3), 7.16–7.37 (m, 6H, Ar), 7.77–7.84 ppm (m, 2H, Ar); 13C NMR
(CDCl3, 75 MHz, TMS): d=14.6, 20.8, 21.2, 22.0, 123.1, 226.4, 126.6,
128.5, 128.8, 129.2, 129.8, 130.0, 132.6, 133.2, 135.1, 136.5, 137.6,
138.1 ppm; IR (CH2Cl2): n˜ =3020, 2920, 2864, 2725, 1903, 1695, 1623,
1599, 1517, 1507, 1440, 1380, 1358, 1182, 1109, 1037, 1022, 933, 883, 821,
774, 728, 577, 538 cmÀ1; MS: m/z (%): 260 [M+] (100), 259 (5.42), 245
(30.34), 243 (3.63), 230 (11.30), 229 (12.43), 228 (4.80), 215 (9.42), 202
(3.03); HRMS (EI): calcd for C20H20: 260.1565; found: 260.1551.
Chem. Eur. J. 2009, 15, 8852 – 8860
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8859

M. Shi, Y. Li, and W. Li
General procedure for the reactions of diarylvinylidenecyclopropanes in
the presence of p-TSA: Under an argon atmosphere, diarylvinylidenecy-
clopropanes 1a–g (0.18 mmol), DCE (3.0 mL), and p-TSA (1.0 equiv)
were added into a Schlenk tube. The resultant reaction mixture was
stirred for different times at room temperature. The reaction was
quenched by addition of anhydrous NaHCO3, and then the product mix-
ture was purified by flash column chromatography.
Compound 4a: A colorless liquid; 1H NMR (CDCl3, 300 MHz, TMS):
d=1.58 (s, 3H, CH3), 1.66 (s, 3H, CH3), 1.69 (s, 3H, CH3), 4.43 (d, J=
1.5 Hz, 1H), 4.48 (d, J=1.5 Hz, 1H), 6.54 (s, 1H), 7.17–7.27 ppm (m,
10H, Ar); 13C NMR (CDCl3, 75 MHz, TMS): d=21.7, 21.8, 23.0, 115.3,
126.8, 127.1, 127.5, 127.9, 128.0, 128.2, 130.4, 131.1, 134.9, 140.9, 142.8,
2007, 72, 509–516; e) J.-M. Lu, M. Shi, Org. Lett. 2007, 9, 1805–
1808; f) J.-M. Lu, M. Shi, Tetrahedron 2007, 63, 7545–7549; g) J.-M.
Lu, M. Shi, J. Org. Chem. 2008, 73, 2206–2210; h) W. Li, M. Shi, J.
Org. Chem. 2009, 74, 856–860; i) W. Li, M. Shi, J. Org. Chem. 2008,
73, 4151–4154; j) J.-M. Lu, Z.-B. Zhu, M. Shi, Chem. Eur. J. 2009,
15, 963–971; k) Z.-B. Zhu, M. Shi, Chem. Eur. J. 2008, 14, 10219–
10222; l) M. Shi, L.-F. Yao, Chem. Eur. J. 2008, 14, 8725–8731.
[4] For the isomerization of alkenylidenecyclopropanes catalyzed by
Lewis acids, see: L. Fitjer, Angew. Chem. 1975, 87, 381–382; Angew.
Chem. Int. Ed. Engl. 1975, 14, 360–361.
[5] The acidity of Tf2NH was found to be higher than that of HOTf in
the gas phase,[5a] whereas the trend is reversed in H2O or AcOH.[5b]
The acidity of the two acids in an ionic liquid is comparable.[5c]
a) I. A. Koppel, R. W. Taft, F. Anvia, S. Zhu, L. Hu, K. Sung, D. D.
DesMarteau, L. M. Yagupolskii, Y. L. Yagupolskii, N. V. Ignatꢂev,
N. V. Kondratenko, A. Y. Volkonskii, V. M. Vlasov, R. Notario, P.
Maria, J. Am. Chem. Soc. 1994, 116, 3047–3057; b) J. Foropoulos,
D. D. DesMarteau, Inorg. Chem. 1984, 23, 3720–3723; c) C. Thoma-
zeau, H. Olivier-Bourbigou, L. Magna, S. Luts, B. Gilbert, J. Am.
Chem. Soc. 2003, 125, 5264–5265. H2SO4: pKa =7.0 in acetic acid at
208C.
˜
143.7, 144.1 ppm; IR (CH2Cl2): n=3444, 3058, 3026, 2974, 2926, 2853,
1950, 1731, 1660, 1598, 1578, 1491, 1446, 1369, 1318, 1278, 1158, 1073,
1032, 1001, 942, 922, 901, 841, 757, 699, 573 cmÀ1; MS: m/z (%): 274 [M+]
(59), 259 (100), 215 (45), 115 (69), 105 (51), 91 (85), 77 (47), 41 (50);
HRMS (EI): calcd for C21H22: 274.1722; found: 274.1721.
Acknowledgements
[6] For some selected Tf2NH-promoted reactions, see: a) Referen-
ces [3h,5b]; b) N. Kuhnert, J. Peverley, J. Roberston, Tetrahedron
Lett. 1998, 39, 3215–3216; c) K. Ishihara, Y. Hiraiwa, H. Yamamoto,
Synlett 2001, 1851–1854; d) J. Cossy, F. Lutz, V. Alauze, C. Meyer,
Synlett 2002, 0045–0048; e) K. Inanaga, K. Takasu, M. Ihara, J. Am.
Chem. Soc. 2005, 127, 3668–3669; f) Y. Zhang, R. P. Hsung, X.
Zhang, J. Huang, B. W. Slafer, A. Davis, Org. Lett. 2005, 7, 1047–
1050; g) J. W. Sun, S. A. Kozmin, J. Am. Chem. Soc. 2005, 127,
13512–13513; h) N. Shindoh, H. Tokuyama, K. Takasu, Tetrahedron
Lett. 2007, 48, 4749–4753; i) K. Takasu, Y. Miyakawa, M. Ihara, H.
Tokuyama, Chem. Pharm. Bull. 2008, 56, 1205–1206; j) K. Takasu,
N. Shindoh, H. Tokuyama, M. Ihara, Tetrahedron 2006, 62, 11900–
11907.
We thank the Shanghai Municipal Committee of Science and Technology
(06XD14005, 08dj1400100-2), the National Basic Research Program of
China (ACHTUNGTRENNUNG(973)-2009CB825300), and the National Natural Science Founda-
tion of China (20472096, 20672127, 20821002, 20732008 and 20702059)
for financial support.
[1] a) M. L. Poutsma, P. A. Ibarbia, J. Am. Chem. Soc. 1971, 93, 440–
450; b) W. Smadja, Chem. Rev. 1983, 83, 263–320; c) M. E. Hen-
drick, J. A. Hardie, M. Jones Jr, J. Org. Chem. 1971, 36, 3061–3062;
d) H. Sugita, K. Mizuno, T. Saito, K. Isagawa, Y. Otsuji, Tetrahedron
Lett. 1992, 33, 2539–2542; e) K. Mizuno, H. Sugita, T. Kamada, Y.
Otsuji, Chem. Lett. 1994, 449–452, and references therein; f) L. K.
Sydnes, Chem. Rev. 2003, 103, 1133–1150; g) T. Akasaka, Y.
Misawa, W. Ando, Tetrahedron Lett. 1990, 31, 1173–1176; h) K.
Mizuno, H. Sugita, K. Isagawa, M. Goto, Y. Otsuji, Tetrahedron
Lett. 1993, 34, 5737–5738; i) K. Mizuno, K. Nire, H. Sugita, Y.
Otsuji, Tetrahedron Lett. 1993, 34, 6563–6566; j) K. Mizuno, H.
Sugita, T. Hirai, H. Maeda, Chem. Lett. 2000, 1144–1145; k) K.
Mizuno, K. Nire, H. Sugita, H. Maeda, Tetrahedron Lett. 2001, 42,
2689–2692; l) K. Mizuno, H. Maeda, H. Sugita, S. Nishioka, T.
Hirai, A. Sugimoto, Org. Lett. 2001, 3, 581–584; m) K. Mizuno, H.
Sugita, T. Hirai, H. Maeda, Y. Otsuji, M. Yasuda, M. Hashiguchi, K.
Shima, Tetrahedron Lett. 2001, 42, 3363–3366; n) H. Maeda, L.
Zhen, T. Hirai, K. Mizuno, ITE Lett. Batteries, New Technol. Med.
2002, 3, 485–488.
[2] For the synthesis of vinylidenecyclopropanes, see: a) K. Isagawa, K.
Mizuno, H. Sugita, Y. Otsuji, J. Chem. Soc. Perkin Trans. 1 1991,
2283–2285, and references therein; b) J. R. Al Dulayymi, M. S.
Baird, J. Chem. Soc. Perkin Trans. 1 1994, 1547–1548. For other
papers related to vinylidenecyclopropanes, see: c) H. Maeda, T.
Hirai, A. Sugimoto, K. Mizuno, J. Org. Chem. 2003, 68, 7700–7706;
d) D. J. Pasto, J. E. Brophy, J. Org. Chem. 1991, 56, 4556–4559;
e) D. J. Pasto, M. F. Miles, J. Org. Chem. 1976, 41, 425–432; f) D. J.
Pasto, M. F. Miles, S.-K. Chou, J. Org. Chem. 1977, 42, 3098–3101;
g) D. J. Pasto, J. K. Borchardt, T. P. Fehlper, H. F. Baney, M. E.
Schwartz, J. Am. Chem. Soc. 1976, 98, 526–529; h) D. J. Pasto, A. F.-
T. Chen, G. Ciurdaru, L. A. Paquette, J. Org. Chem. 1973, 38, 1015–
1026; i) D. J. Pasto, J. K. Borchardt, J. Am. Chem. Soc. 1974, 96,
6937–6943.
[7] a) P. Hohenberg, W. Kohn, Phys. Rev. 1964, 136, B864–B871; b) W.
Kohn, L. J. Sham, Phys. Rev. 1965, 140, A1133–A1138.
[8] Gaussian 03, Revision C.02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomer-
y, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyen-
gar, 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. Pomelli, J. W. Ochterski, P. Y.
Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannengerg,
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.
[9] a) A. D. Becke, Phys. Rev. A 1988, 38, 3098–3100; b) A. D. Becke, J.
Chem. Phys. 1993, 98, 5648–5652; c) C. Lee, W. Yang, R. G. Parr,
Phys. Rev. B 1988, 37, 785–789.
[10] a) M. T. Cances, V. Mennucci, J. Tomasi, J. Chem. Phys. 1997, 107,
3032–3041; b) M. Cossi, V. Barone, B. Mennucci, J. Tomasi, Chem.
Phys. Lett. 1998, 286, 253–263.
[11] The synthesis of butatrienecyclopane 6 has been reported: a) P. J.
Stang, T. E. Fisk, Synthesis 1979, 438–440; b) P. J. Stang, T. E. Fisk,
J. Am. Chem. Soc. 1980, 102, 6813–6816.
[12] R. Ditchfield, Mol. Phys. 1974, 27, 789–807.
[3] a) G.-C. Xu, M. Ma, L.-P. Liu, M. Shi, Synlett 2005, 1869–1872;
b) G.-C. Xu, L.-P. Liu, J.-M. Lu, M. Shi, J. Am. Chem. Soc. 2005,
127, 14552–14553; c) J.-M. Lu, M. Shi, Org. Lett. 2006, 8, 5317–
5320; d) Y.-P. Zhang, J.-M. Lu, G.-C. Xu, M. Shi, J. Org. Chem.
Received: May 4, 2009
Published online: July 23, 2009
8860
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 8852 – 8860