Generation of cycloheptyne during the solvolysis of cyclohexylidenemethyliodonium salt in the presence of base

Article abstract of DOI:10.1246/bcsj.76.1849

Solvolysis of 4-methylcyclohexylidenemethyl(phenyl)iodonium tetrafluoroborate (1) and its R isomer (69% ee) was carried out in 2,2,2-trifluoroethanol (TFE) and 1,1,1,3,3,3-hexafluoropropan-2-ol (HFP) in the presence of bases such as acetate, pyridine, triethylamine, and alkoxide. The reaction is much faster in TFE than in HFP. Products in TFE include solely un-rearranged (racemized) enol ether 2 together with iodobenzene, while the main product in HFP is ring-expanded (partially racemized) 1-alkoxycycloheptene 3. Results show that 2 is formed via α-elimination with alkylidenecarbene as an intermediate, while the reaction in HFP to give 3 involves a cycloheptyne intermediate that is mostly derived from an intermediate cyclohept-1-enyl cation via the E1-type pathway.

Full text of DOI:10.1246/bcsj.76.1849

ꢀ 2003 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 76, 1849–1855 (2003) 1849
Generation of Cycloheptyne during the Solvolysis of
Cyclohexylidenemethyliodonium Salt in the Presence of Base
ꢀ
ꢀ
Morifumi Fujita, Kenji Ihara, Wan Hyeok Kim, and Tadashi Okuyama
Graduate School of Science, Himeji Institute of Technology, Kamigori, Hyogo 678-1297
Received April 17, 2003; E-mail: okuyama@sci.himeji-tech.ac.jp
Solvolysis of 4-methylcyclohexylidenemethyl(phenyl)iodonium tetrafluoroborate (1) and its R isomer (69% ee)
was carried out in 2,2,2-trifluoroethanol (TFE) and 1,1,1,3,3,3-hexafluoropropan-2-ol (HFP) in the presence of bases
such as acetate, pyridine, triethylamine, and alkoxide. The reaction is much faster in TFE than in HFP. Products in
TFE include solely un-rearranged (racemized) enol ether 2 together with iodobenzene, while the main product in
HFP is ring-expanded (partially racemized) 1-alkoxycycloheptene 3. Results show that 2 is formed via ꢀ-elimination
with alkylidenecarbene as an intermediate, while the reaction in HFP to give 3 involves a cycloheptyne intermediate that
is mostly derived from an intermediate cyclohept-1-enyl cation via the E1-type pathway.
Solvolysis of primary alkenyl iodonium salts takes place via
the rearrangement with ꢁ-alkyl participation as well as via the
termediary formation of cycloheptyne I2 was not usually ob-
served under the solvolysis conditions, but one set of delicately
balanced basic/nucleophilic conditions involving methanesul-
fonate in chloroform led to the cycloheptyne intermediate
(I2).5 There is some hope that the basic reactivity of the reac-
tion medium can be controlled to induce the E1-type elimina-
tion (rather than nucleophilic SN1-type trapping) of the inter-
mediate cycloheptenyl cation I1. The basicity/nucleophilicity
must be mild enough to allow rearrangement to I1, but not so
strong as to allow the ꢀ-elimination or the SN2-type substitu-
tions of the substrate 1 itself. The present paper describes suc-
cessful manipulations of solvolysis conditions to obtain cyclo-
heptyne I2 from cyclohexylidenemethyliodonium salt 1. There
is also some possibility of access to cycloheptyne I2 from 1 via
base-induced ꢀ-elimination–carbene rearrangement.
ꢀ
ꢀ
SN2-type substitutions with ꢂ and ꢃ attacks to avoid the un-
stable primary vinyl cation (Scheme 1).1–10 The intermediate
secondary vinyl cations formed by the 1,2-alkyl shift provide
both substitution and elimination products.2,8 In the presence
of an added base, the ꢀ-elimination predominates to give an
alkylidenecarbene which leads to intramolecular or intermo-
lecular insertion depending on the reaction conditions.2–4,6,11–15
The intramolecular insertion involves 1,2-shift of hydrogen,
phenyl, or alkyl group to afford an alkyne.2–4,6,11–16 This type
of rearrangement of cycloalkylidenecarbenes has been used as
one of the methods of generation of cycloalkynes.17
The chirality probe approach using an optically active 4-
methylcyclohexylidenemethyl(phenyl)iodonium tetrafluorobo-
rate (1) showed that a symmetric primary vinyl cation is not
involved during the reaction (Scheme 2).4,10 In this case, the
cycloheptenyl cation I1 formed by rearrangement leads only
to a substitution product in nucleophilic alcoholic solvents. In-
Results
ꢁ
Reactions of 1 with bases were carried out at 55–60 C in
poorly nucleophilic solvolytic media, i.e., fluoro alcohols,
Scheme 1.
Published on the web September 15, 2003; DOI 10.1246/bcsj.76.1849

1850 Bull. Chem. Soc. Jpn., 76, No. 9 (2003)
Cycloheptyne from Vinyl Iodonium Salt
Scheme 2.
Table 1. Reactions of 1 in Fluoro Alcohols Containing Basea)
Yield/%
3c)
75 (69)e)
58 (66)e)
0
0
0
0
Entry
1d)
2
3
4
5
6
7f)
8d)
9
10
11
12
13
14f)
15h)
Solvent
Baseb)
Time/h
2c)
4
PhI
Other
TFE
TFE
TFE
TFE
TFE
TFE
TFE
HFP
HFP
HFP
HFP
HFP
HFP
HFP
HFP
none
CF3CO2Na
CH3CO2Na
pyridine
Et3N
RONa
RONa
none
CH3CO2Na
Et3N
RONa (0.01)
RONa
RONa (0.2)
RONa
RONa
9
4
1
1
1
0
0
0
0
0
0
0
0
0
7
20
15
15
18
20
<1
84
90
100
100
68
92
82
60
66
74
67
70
65
62
65
—
—
—
—
—
—
<1 (5)
19 (0)
99
93
84 (0)
99
76
0
4
6
7
1
1
0
20
20
8
23
20
10
22
20
41 (69)e)
30
43
2g)
38
—
—
—
<1 (5)
25 (6)
7 (0)
6
5
6 (0)
43 (13)
36
28
26 (16)
ꢁ
a) The initial concentration of 1 was 5 ꢂ 10ꢃ3 mol dmꢃ3, and reactions were carried out at 55–60 C. b) The concen-
tration was 0.1 mol dmꢃ3 unless noted otherwise in parentheses, where the concentration are given in mol dmꢃ3. c) The
values in parentheses are the enantiomeric excess (ee) of the product obtained from 69% ee of (R)-1. The R isomer of 3
is in excess. d) The data has been reported previously.4 e) 4-Methylcycloheptanone. f) Cyclohexene (0.1 mol dmꢃ3
was added. g) Yield of 5-methylcyclohept-1-enyl acetate. h) ꢀ-Pyrone (0.1 mol dmꢃ3) was added.
)
2,2,2-trifluoroethanol (TFE) and 1,1,1,3,3,3-hexafluoropropan-
2-ol (HFP), and the mixed solvents. Products include both un-
rearranged and rearranged enol derivatives, 2 and 3, and 5-
methylcycloheptene (4) as well as iodobenzene (Eq. 1). For
some runs, optically active (R)-1 (69% ee) was used as the
substrate. Effects of added cyclohexene and ꢀ-pyrone were al-
so examined for possible intermediates (Eqs. 2 and 3). Product
yields were determined by gas chromatography and are sum-
marized in Tables 1 and 2. Trapping experiments were further
carried out on a preparative scale for cycloheptyne I2 with tet-
raphenylcyclopentadienone (TPCD) (Eq. 4); the results are
given in Table 3.
ð1Þ

M. Fujita et al.
Bull. Chem. Soc. Jpn., 76, No. 9 (2003) 1851
Table 2. Reactions of 1 in Mixed Alcohols Containing Sodium Alkoxidea)
Yield/%
2b
Entry
TFE/HFP
Time/h
2a
3a
3b
4
PhI
b)
6
16
17
18
19
12
100/0
70/30
50/50
30/70
10/90
0/100
1
1
3
7
20
20
99
94
76
34
9
—
—
—
—
—
—
0
0
3
<1
6
7
0
0
4
16
36
43
0
92
97
100
72
93
70
b)
b)
b)
b)
b)
6
16
28
20
15
0
a) [1] = 5 ꢂ 10ꢃ3 mol dmꢃ3, [alkoxide] = 0.1 mol dmꢃ3, and at 55–60 ꢁC. b) 3a is not observed as an
appreciable product.
Table 3. Isolated Yields of 7 in the Reaction of 1 in the Presence of TPCDa)
Entry
Solvent
b)
Base
[Base]/mol dmꢃ3
Yield of 7/%
20
21
22
23
24
25
26
27
28
CHCl3
Et3N
Et3N
RONa
Et3N
Et3N
RONa
RONa
RONa
RONa
0.05
0.1
0.1
0.02
0.1
0.1
0.1
0.1
0.1
8
2
3
58
62
67
26
45
46
TFE/CHCl3 (9/1)c)
TFE/CHCl3 (9/1)c)
HFP
HFP
HFP
HFP/TFE (1/1)
HFP/TFE (9/1)
HFP/CHCl3 (1/9)
a) [1] = 1 ꢂ 10ꢃ2 mol dmꢃ3, [TPCD] = 2 ꢂ 10ꢃ2 mol dmꢃ3, and at 55–60 ꢁC. b) At room
temperature. c) Chloroform (10% (v/v)) was added to dissolve TPCD in solution.
cal purity when the optically active substrate was used (Entries
1 and 8). These and other observations show that the symmet-
ric primary vinyl cation intermediate is avoided because of its
instability.4,10 In contrast, reactions of 1 with base gave differ-
ent results between the two fluoro solvents, TFE and HFP. On-
ly the un-rearranged product 2a was obtained in TFE, while
the rearranged ones were the major products in HFP.
Reactions in TFE. The un-rearranged substitution product
2a was obtained with accompanying iodobenzene in TFE in
the presence of a range of bases from acetate to alkoxide (En-
tries 3–6), and it lost completely the optical purity of (R)-1
(Entry 5). The reaction of 1 in a deuterium solvent TFE-O-
d containing sodium alkoxide afforded the ꢀ-deuterated prod-
uct 2a-d (Eq. 6).
ð2Þ
ð3Þ
ð4Þ
ð6Þ
The trapping experiment with TPCD was also applied to the
reaction of cyclopentylidenemethyl(phenyl)iodonium tetra-
fluoroborate (8) in HFP in the presence of alkoxide; adduct 9
was isolated in 13% yield (Eq. 5).
These results are consistent with the intermediary formation
of alkylidenecarbene I3 via ꢀ-elimination. However, the at-
tempted trapping of I3 with cyclohexene only gave a trace
amount of the cyclopropylidene adduct 5 (Eq. 2) (Entry 7).
Alcohol TFE must more effectively react with carbene I3 to
give the OH insertion product.
A poorly basic trifluoroacetate (pKa = 0.23 in aqueous solu-
tion) gave 19% of 2a and 58% of 3 (only the hydrolysis prod-
uct, 4-methylcycloheptanone (3C), was isolated here) in TFE;
the former was completely racemized while the latter kept
mostly the original ee of (R)-1 (Entry 2). This result shows
that the rearrangement still occurs via participation which is
followed by trapping of the resulting chiral cation I1 to give
ð5Þ
Discussion
Solvolyses of 1 in un-buffered TFE and HFP gave solely the
rearranged ring-expanded product 3, keeping the original opti-

1852 Bull. Chem. Soc. Jpn., 76, No. 9 (2003)
Cycloheptyne from Vinyl Iodonium Salt
optically active 3, while the competing ꢀ-elimination gives the
un-rearranged and racemized 2a.
Although the ring expansion of cycloalkylidenecarbene has
been used as one of the general routes to cycloalkynes,17 such
rearrangement does not seem to take place in TFE. So, the re-
actions of 1 with bases were examined in still less reactive al-
cohol HFP.
Reactions in HFP. The base reaction in HFP is much
slower than that in TFE, and preferentially gives the rear-
ranged products, 3b and 4, together with some un-rearranged
enol ether 2b (Entries 9–13). The products obtained from
(R)-1 of 69% ee with alkoxide as a base include extensively
racemized but still optically active 3b (13% ee) and complete-
ly racemized un-rearranged product 2b (Entry 12). The race-
mization of 2b must occur via the intermediate carbene I3,
while that of 3b must be due to formation of cycloheptyne
I2. Formation of I2 is in fact confirmed by trapping experi-
ments as described below. Nonetheless, the remaining optical
purity of 3b cannot be ascribed to the cycloheptyne route. The
rearranged chiral cation I1 is a most likely intermediate to af-
ford the optically active 3b. Thus, cycloheptyne I2 can in prin-
ciple be derived from cation I1 as well as from carbene I3. An
attempted trapping with cyclohexene of I3 was not successful
in HFP (Entry 14), and the carbene route to I2 could not be a
major one if it operated as a minor route leading to 2b. If cy-
cloheptyne I2 were formed from carbene I3 in HFP, cyclohex-
ene could have trapped I3; I3 was trapped by cyclohexene in
chloroform before the rearrangement to I2 occurred.4,5
The reaction of optically active (R)-1 was also examined in
the presence of ꢀ-pyrone as a trapping agent (Entry 15). Al-
though the trapping was not complete (25% adduct 6, Eq. 3),
the ee of the remaining 3b increased to 16% (26% yield). This
modest increase in ee corresponds to the increased fraction of
optically active 3b, which is formed independently of the cy-
cloheptyne intermediate I2, i.e., via the direct reaction with I1.
The reaction of 1 with alkoxide was monitored by 1H NMR
in a deuterium solvent, HFP-O-d/CDCl3 (1/10, v/v). Progress
of the reaction can be monitored by formation of iodobenzene
as well as by the substrate decrease (Fig. 1). Loss of the vinyl-
ic proton (ꢄ ¼ 6:49 ppm) of 1 was somewhat faster than the
decrease in the other signals of 1. That is, a deuterium is in-
corporated at the vinylic position to give deuterated 1-d. The
hydrogen–deuterium isotope exchange takes place during the
reaction. This implies that iodonium ylide I4 is formed rever-
sibly as a precursor for carbene I3 and that the rate-determin-
ing step for the ꢀ-elimination is departure of iodobenzene
(Scheme 3). This is rather unexpected in view of the high
leaving ability of the iodonio group,18–20 but it can be ration-
alized by reprotonation with acidic HFP (pKa = 9.3).21 The
reversibility of the deprotonation retards the formation of al-
kylidenecarbene, and makes formation of the rearranged cation
I1 via participation of a more favorable reaction.
Fig. 1. Reaction of 1 (0.007 mol dmꢃ3) in CDCl3/HFP-O-d
(10/1 v/v) containing sodium alkoxide (0.02 mol dmꢃ3) at
297 K, monitored by 1H NMR: 1 + 1-d ( ), PhI ( ), and
1-d/(1 + 1-d) ( ).
Scheme 3.
reported of reduction of a seven-membered cycloalkyne with
methanol.22 A higher yield of 4 was obtained in a mixed
TFE–HFP solvent at the ratio 30/70 or 10/90 than in pure
HFP (Entries 18 and 19), and TFE is a better reductant. An
alternative possibility of hydride transfer from hexafluoro-2-
propoxide to cation I1 is less likely. If this were the case, a
better yield of 4 would not be found in the mixed solvent.
A probable mechanism is the concerted one as proposed for
the methanol reduction22 and depicted in Eq. 7.
ð7Þ
Formation of Cycloalkyne. Formation of cycloheptyne I2
was examined by trapping with a very efficient trapping agent
TPCD (Eq. 4 and Table 3). Higher yields of adduct 7 were ob-
tained in HFP, and the highest yield was achieved with alkox-
ide in HFP (Entry 25). The attempted TPCD trapping in TFE
gave very poor results (Entries 21 and 22) as expected from the
product studies. That in chloroform gave a small amount (8%)
of the TPCD adduct 7 (Entry 20). Under these conditions, al-
Finally, unexpected formation of a considerable amount of
the reduction product 4 was observed in basic HFP, accompa-
nied with formation of 3b (Table 1). It is noted that 4 was not
detected in the reactions in neutral HFP or TFE where the cor-
responding enol ether 3 is formed directly from cation I1 but
not via I2. This product 4 may be derived actually from cyclo-
heptyne I2 via reduction. Very suggestive results have been

M. Fujita et al.
Bull. Chem. Soc. Jpn., 76, No. 9 (2003) 1853
kylidenecarbene I3 is effectively generated to give 5 in the
presence of cyclohexene.4,5 Partial rearrangement of carbene
I3 to cycloheptyne I2 seems to occur in basic chloroform,
but it is obviously not efficient. When 10% (v/v) of HFP
was added to the chloroform reaction, the yield of adduct 7
was considerably improved (Entry 28). The cationic elimina-
tion route to I2 must show up owing to the retardation of the
reaction with base by acidic HFP.
Generation of a smaller ring alkyne, cyclohexyne, via ring
expansion of cyclopentylidenemethyliodonium salt 8 was ex-
amined in the presence of TPCD (Eq. 5) under the best condi-
tions for cycloheptyne formation. However, the isolated yield
of the adduct 9 was only 13%. The rearrangement of 8 to cy-
clohex-1-enyl cation must occur less readily than that of 1 to I1
owing to the higher strain of the smaller ring vinyl cation. So,
the E1-type route to cyclohexyne from 8 should be less
efficient. Although the rearrangement of cyclopentylidenecar-
bene is known as one of the well-studied routes for generation
of cyclohexyne,17,23 the carbene route to cycloalkyne via ꢀ-
elimination is again not effective owing to inefficient 1,2-alkyl
shift of alkylidenecarbene.24 It is noteworthy that ꢁ-elimina-
tion of cyclohex-1-enyliodonium salt with a mild base affords
cyclohexyne in an excellent yield.25
ꢁ-alkyl participation leading to cation I1 which now can be de-
protonated with the mild base to give I2. The 1,2-alkyl shift
within alkylidenecarbene is not efficient.
Experimental
Proton and 13C NMR spectra were measured on a JEOL ECA-
600 spectrometer; the samples were solutions in CDCl3. 1H NMR
spectra were recorded using the residual CHCl3 as an internal ref-
erence (7.24 ppm) and 13C NMR using CDCl3 as an internal ref-
erence (77.00 ppm). Melting points were measured on a Yanaco
micro-melting-point apparatus and are uncorrected. Mass spec-
trometers JEOL JMS-AX505HA, JEOL JMS-T100LC, and JEOL
automass systembwere used for MS and GC-MS. GC was con-
ducted on a gas chromatograph with DB-1 (i.d. 0.25 mm ꢂ 30 m)
or Chirasil-DEX-CB (i.d. 0.25 mm ꢂ 25 m). 4-Methylcyclohex-
ylidenemethyl(phenyl)iodonium tetrafluoroborate (1) and the opti-
cally active (R)-1 were prepared in the same way as before.4 Tet-
raphenylcyclopentadienone (TCI), ꢀ-pyrone (Aldrich), and cyclo-
hexene (TCI) were used without further purification. HFP and
TFE were distilled over molecular sieves 4A just before use for
the reaction.
Cyclopentylidenemethyl(phenyl)iodonium Tetrafluorobo-
rate (8). To a solution of trimethylsilylmethylenecyclopentane
(0.86 g, 5.6 mmol), and PhIO (2.46 g, 11 mmol) in dichloro-
Reactions in Mixed Solvents. Rates of the reaction in TFE
and HFP are contrastingly different. Bases accelerate very
much the reaction in TFE, while they affect only slightly the
rate of reaction in HFP, as the reaction times given in
Table 1 imply. In the mixed solvents of TFE and HFP, the rate
of reaction in the presence of alkoxide decreases with increas-
ing fraction of HFP as is reflected in the reaction times of 1–20
h (Table 2). The proportion of the rearranged product 3 to the
un-rearranged product 2 also increases with fraction of HFP.
Acetate (pKa = 4.76 in aqueous solution) works as a perfect
base for ꢀ-elimination in TFE, while alkoxide of HFP (pKa
= 9.3) does not. The effective solvation by the acidic solvent
may strongly reduce the basicity in HFP, but a more important
factor to retard ꢀ-elimination in HFP would be the acidity it-
self of the solvent. Protonation of the intermediate ylide I4
must be greatly accelerated in HFP, and this makes the depro-
tonation reversible as observed in HFP–CDCl3. Thus, the
overall rate of reaction in HFP is diminished to make the ꢁ-al-
kyl participation leading to the rearranged cation I1 the main
reaction. The fraction of rearrangement increases sharply
above 50% content of HFP. It is noteworthy that TFE adduct
3a was not clearly observed even in the presence of a large
amount of TFE in a mixed solvent when a considerable forma-
tion of 3b was apparent. Although TFE (pKa = 12.4) is obvi-
ously more basic and nucleophilic than HFP, predominant alk-
oxide present in a mixed solvent should be that of HFP, and it
must also be the more reactive nucleophile toward I1 or I2.
methane (20 mL) was added Et2O BF3 (1.45 mL, 11 mmol) drop-
ꢄ
wise at 0 ꢁC. After stirring for 80 min at 0 ꢁC, a saturated aqueous
sodium tetrafluoroborate solution was added to the reaction
mixture. The mixture was stirred vigorously for 30 min, and ex-
tracted with dichloromethane. The organic layer was concentrated
in vacuo to give an oil. Crystallization of the crude mixture from
dichloromethane–ether–hexane gave the title cꢁompound (0.12 g,
1
0.3 mmol, 6%) as a white solid: mp 130–131 C; H NMR (600
MHz, CDCl3) ꢄ 7.91 (2H, d, J ¼ 7:9 Hz), 7.62 (1H, t, J ¼ 7:9
Hz), 7.48 (2H, t, J ¼ 7:9 Hz), 6.70 (1H, s), 2.71 (2H, t, J ¼ 6:9
Hz), 2.68 (2H, t, J ¼ 6:9 Hz), 1.93 (2H, quint, J ¼ 6:9 Hz),
1.85 (2H, quint, J ¼ 6:9 Hz); 13C NMR (CDCl3) ꢄ 171.64,
134.62, 132.49, 132.46, 109.68, 88.26, 37.77, 36.12, 27.94,
25.74; HRMS (ESI) Calcd for C12H14I (M ꢃ BF4) 285.0140,
Found 285.0177.
Standard Procedure for Reaction of 1 in Fluoro Alcohol. A
sodium alkoxide solution was prepared by dissolving a piece of
sodium metal in the alcohol. The tetrafluoroborate salt of 1 (2
mg) was dissolved in 1 mL of fluoro alcohol containing a required
amount of a base and kept in a sealed tube at 55–60 ꢁC for a speci-
fied time. To the mixture were added water and 5 mmol of tetra-
decane as an ether solution. The products were extracted with
ether and washed with water. The yields of the products were de-
termined by gas chromatography with tetradecane as an internal
standard. The GC retention times of 4, 2a, 2b, 3b, PhI, and 3C
were 2.8, 7.5, 6.3, 6.9, 6.4, andꢁ7.9 min, respectively, at the col-
umn (DB-1) temperature of 64 C. The enantiomeric excess (ee)
of the product from (R)-1 of 69% ee was determined with a chiral
GC column (DEX-CB). Preparation of the authentic samples of
2a, (R)-3C, and 5-methylcyclohept-1-enyl acetate were described
in the previous report,4 and they were used for the identification of
these products. Selected data for 2b: 1H NMR (CDCl3) ꢄ 5.90
(1H, s), 4.15 (1H, sept, J ¼ 5:8 Hz), 2.77 (1H, m), 2.02 (1H,
m), 1.88 (3H, td, J ¼ 13:1, 4.1 Hz), 1.75–1.63 (4H, m), 1.51–
1.46 (2H, m), 0.88 (3H, d, J ¼ 6:2 Hz); MS (EI) m=z (relative in-
tensity, %) 276 (Mþ, 17), 220 (11), 93 (100); HRMS (EI) Calcd
for C11H14F6O (M) 276.0949, Found 276.0969; GC (DB-1, 64
Conclusion
In HFP solution containing alkoxide, cycloheptyne I2 is ef-
fectively generated from cyclohexylidenemethyliodonium salt
1 mainly via cyloheptenyl cation I1, if the route via carbene I3
is not completely excluded. Poor nucleophilicity and mild ba-
sicity of the HFP alkoxide and high acidity of the HFP solvent
retard generation of alkylidenecarbene I3 by way of efficient
reprotonation of intermediate iodonium ylide I4. This allows

1854 Bull. Chem. Soc. Jpn., 76, No. 9 (2003)
Cycloheptyne from Vinyl Iodonium Salt
(EI) Calcd for C34H26 (M) 436.2191, Found 436.2189.
ꢁC) 6.3 min, Chiral GC (DEX-CB, 80 ꢁC) 14.7 min and 16.6 min.
Selected data for 3b: 1H NMR (CDCl3) ꢄ 5.10 (1H, m), 4.57 (1H,
sept, J ¼ 5:8 Hz), 2.38 (1H, m), 2.26 (1H, m), 2.12 (1H, m), 1.97
(1H, m), 1.76–1.60 (3H, m), 1.14 (1H, qd, J ¼ 11:7, 2.1 Hz), 1.04
(1H, qd, J ¼ 11:7, 1.4 Hz), 0.90 (3H, d, J ¼ 6:9 Hz); MS (EI) m=z
(relative intensity, %) 276 (Mþ, 2), 233 (20), 220 (13), 93 (33), 68
(100); HRMS (EI) Calcd for C11H14F6O (M) 276.0949, Found
276.0907; GC (DB-1, 64 ꢁC) 6.9 min, Chiral GC (DEX-CB, 80
ꢁC), 14.4 min (R-isomer) and 15.3 min (S-isomer). The absolute
stereochemistry of 3b was confirmed by the conversion to 3C4 un-
der acidic aqueous conditions. 5-Methylcycloheptene (4) has
identical spectroscopic properties to those reported in the litera-
ture:26 MS (EI) m=z (relative intensity, %) 110 (Mþ, 20), 95
(50), 82 (83), 67 (100); the reported value 110 (Mþ, 40), 95
(60), 82 (90), 67 (100).
Reaction of 1 in Chloroform Containing (CF3)2CHOD and
(CF3)2CHONa. To a solution of 1 (8 ꢂ 10ꢃ3 mol dmꢃ3) in
CDCl3 (0.5 mL) was added the (CF3)2CHOD solution (0.05
mL) containing sodium alkoxide (0.2 mol dmꢃ3), which was pre-
pared by adding sodium to the alcohol. The reaction was moni-
tored at 297 K by 1H NMR. On addition of alkoxide, an immedi-
ate shift of the NMR signals of 1 was apparent: before the addition
of alkoxide; ꢄ 7.90 (2H, d), 7.61 (1H, t), 7.47 (2H, t), 6.65 (1H, s),
2.75–2.66 (2H, m), 2.47–2.34 (2H, m), 1.90 (1H, m), 1.84 (m,
1H), 1.62 (1H, m), 1.15–1.00 (2H, m), 0.91 (3H, d); after the ad-
dition, ꢄ 7.80 (2H, d), 7.70 (1H, t), 7.51 (2H, t), 6.49 (1H, s), 2.70–
2.60 (2H, m), 2.48–2.36 (2H, m), 1.94 (1H, m), 1.87 (m, 1H), 1.65
(1H, m), 1.15–1.00 (2H, m), 0.93 (3H, d). The peak area at 6.49
ppm due to the vinylic proton of 1 decreased more rapidly than the
other peak areas due to 1. The yield of iodobenzene was deter-
mined by the comparison of the peak area due to the phenyl group
(7.09 ppm) and the residual CHCl3 as an internal standard. The
total yield of 1 and deuterated 1 (1-d) was determined by the peak
area at 7.80 ppm, and the protium content at the vinylic position of
1 was determined by the peak area at 6.49 ppm.
Reaction in the Presence of ꢀ-Pyrone. The standard proce-
dure for the reaction was applied to 1 (2 mg) in the presence of ꢀ-
pyrone (8 mL) in HFP (1 mL) containing sodium alkoxide (0.1
mol dmꢃ3). The mixture was analyzed by GC (DB-1) after ether
extraction. The retention time of the benzo adduct 6 wꢁas 16.2
min, when the column temperature was maintained at 64 C dur-
ing the initial 10 min and then raised up at the rate of 10 ꢁC minꢃ1
.
Selected data for 6: 1H NMR (CDCl3) ꢄ 7.07 (s, 4H), 2.81 (td,
J ¼ 14:4, 1.4 Hz, 2H), 2.72 (ddd, J ¼ 14:4, 7.6, 1.4 Hz, 2H),
1.90–1.85 (m, 3H), 1.80–1.74 (m, 2H), 0.91 (d, J ¼ 6:9 Hz,
3H); 13C NMR (CDCl3) ꢄ 143.19, 128.74, 125.97, 36.27, 36.19,
35.04, 23.71; MS (EI) m=z (relative intensity, %) 160 (Mþ, 62),
104 (100); HRMS (EI) Calcd for C12H16 (M) 160.1252, Found
160.1260.
References
1
(1997).
2
T. Okuyama and M. Ochiai, J. Am. Chem. Soc., 119, 4785
a) T. Okuyama, K. Sato, and M. Ochiai, Chem. Lett., 1998,
1177. b) T. Okuyama, H. Yamataka, and M. Ochiai, Bull. Chem.
Soc. Jpn., 72, 2761 (1999).
Reaction in the Presence of Cyclohexene. The reaction of 1
in fluoro alcohol containing base was carried out in the presence
of cyclohexene (0.1 mol dmꢃ3), and the products were analyzed
by GC. The authentic sample of cyclopropylidene product 5
was prepared according to the method reported previously.4,5
Reaction in Fluoro Alcohol-O-d. The reaction of 1 was car-
ried out in TFE-O-d or HFP-O-d containing sodium alkoxide.
The deuterium incorporation of 2a at the olefinic position was de-
termined as 95% D by comparison of the peak areas due to the
olefinic proton at ꢄ ¼ 5:80 and the methylene proton at
ꢄ ¼ 2:76. In GC-MS analysis, 2a has peaks at m=z ¼ 209 and
208 in 64 and 6% relative intensities in comparison with 8 and
61% for those of the normal product. The HFP product 2b has
peaks at m=z ¼ 277 and 276 in 33 and 4% relative intensities in
comparison with 1 and 17% for those of the normal product.
The rearranged product 3b has a very small molecular peak: the
normal product 3b has a peak at m=z ¼ 276 only in a 2% intensity
while that obtained in HFP-O-d has a peak at m=z ¼ 277 in a 2%
intensity.
3
Jpn., 72, 163 (1999).
T. Okuyama, Y. Ishida, and M. Ochiai, Bull. Chem. Soc.
4
a) M. Fujita, Y. Sakanishi, and T. Okuyama, J. Am. Chem.
Soc., 122, 8787 (2000). b) M. Fujita, Y. Sakanishi, M. Nishii, H.
Yamataka, and T. Okuyama, J. Org. Chem., 67, 8130 (2002).
5
a) M. Fujita, Y. Sakanishi, and T. Okuyama, J. Am. Chem.
Soc., 123, 9190 (2001). b) M. Fujita, Y. Sakanishi, M. Nishii, and
T. Okuyama, J. Org. Chem., 67, 8138 (2002).
6 T. Okuyama, S. Imamura, and Y. Ishida, Bull. Chem. Soc.
Jpn., 74, 543 (2001).
7 R. Gronheid, G. Lodder, M. Ochiai, T. Sueda, and T.
Okuyama, J. Am. Chem. Soc., 123, 8760 (2001).
8
9
T. Okuyama, Acc. Chem. Res., 35, 12 (2002).
T. Okuyama and G. Lodder, Adv. Phys. Org. Chem., 37, 1
(2002).
10 T. Okuyama and M. Fujita, Proc. Jpn. Acad., Ser. B, 78,
167 (2002).
11 For a review on unsaturated carbenes, see: P. J. Stang,
Chem. Rev., 78, 383 (1978).
A Typical Procedure for Reaction of 1 in the Presence of
TPCD.
12 M. Ochiai, Y. Takaoka, and Y. Nagao, J. Am. Chem. Soc.,
110, 6565 (1988).
13 M. Ochiai, M. Kunishima, S. Tani, and Y. Nagao, J. Am.
Chem. Soc., 113, 3135 (1991).
14 M. Ochiai, K. Uemura, and Y. Masaki, J. Am. Chem. Soc.,
115, 2528 (1993).
15 W. G. von der Schulenburg, H. Hopf, and R. Walsh, An-
gew. Chem., Int. Ed., 38, 1128 (1999).
16 For rearrangement of carbenoids, see, e.g.: M. Braum, An-
gew. Chem., Int. Ed., 37, 430 (1998); H. Rezaei, S. Yamanoi, F.
Chemla, and J. F. Normant, Org. Lett., 2, 149 (2000).
17 For reviews on cycloalkynes, see: a) R. W. Hoffmann,
‘‘Dehydrobenzene and Cycloalkynes,’’ Academic Press, New
York (1967). b) A. Krebs and J. Wilke, Top. Curr. Chem., 109,
A solution containing 1 (100 mg, 2:5 ꢂ 10ꢃ4 mol),
TPCD (192 mg, 5 ꢂ 10ꢃ4 mol), and triethylamine (0.35 mL, 2:5 ꢂ
10ꢃ3 mol) in HFP (25 mL) was refluxed till the TLC spot of 1 dis-
appeared, and then the solvent was removed by distillation. The
residue was purified by chromatography (SiO2, eluent: 40%
chloroform in hexane) to give the adduct 7 (72 mg, 62% yield),
which had identical spectroscopic properties to those reported
previously.5 A similar procedure was also applied to 8 to give
9. 9: mp 264–266 ꢁC; 1H NMR (CDCl3) ꢄ 7.16–7.13 (m, 4H),
7.08–7.06 (m, 6H), 6.80–6.74 (m, 10H), 2.52 (m, 4H), 1.70 (m,
4H); 13C NMR (CDCl3) ꢄ 140.78, 140.62, 140.54, 138.44,
134.55, 131.26, 130.28, 127.45, 126.38, 125.89, 124.97, 29.66,
23.11; MS (EI) m=z (relative intensity, %) 436 (Mþ, 100); HRMS

M. Fujita et al.
Bull. Chem. Soc. Jpn., 76, No. 9 (2003) 1855
189 (1983). c) R. Gleiter and R. Merger, in ‘‘Modern Acetylenic
Chemistry,’’ ed by P. J. Stang and F. Diederich, VCH, Weinheim
(1995), p. 285. d) H. Hopf, ‘‘Classics in Hydrocarbon Chemistry,’’
Wiley-VCH, Weinheim (2000).
18 T. Okuyama, T. Takino, T. Sueda, and M. Ochiai, J. Am.
Chem. Soc., 117, 3360 (1995).
19 The H–D exchange via iodonium ylide is reported for a sta-
bilized zwitterionic vinyliodonium salt carrying ꢁ-phosphate
group: P. J. Stang, H. Wingert, and A. M. Arif, J. Am. Chem.
Soc., 109, 7235 (1987).
20 A ꢁ-alkoxyvinyliodonium ylide was generated by desilyla-
tion of the appropriate precursor: M. Ochiai, M. Kunishima, K.
Fuji, M. Shiro, and Y. Nagao, J. Chem. Soc., Chem. Commun.,
1988, 1076.
21 R. Stewart, ‘‘The Proton: Applications to Organic Chemis-
try,’’ Academic Press, Orlando (1985).
22 A. Krebs and H. Colberg, Chem. Ber., 113, 2007 (1980).
23 K. L. Erickson and J. Wolinsky, J. Am. Chem. Soc., 87,
1142 (1965); J. Tseng, M. L. McKee, and P. B. Shevlin, J. Am.
Chem. Soc., 109, 5474 (1987).
24 M. Ochiai, T. Sueda, K. Uemura, and Y. Masaki, J. Org.
Chem., 60, 2624 (1995).
25 M. Fujita, Y. Sakanishi, W. H. Kim, and T. Okuyama,
Chem. Lett., 2002, 908.
26 F. MacCorquodale and J. C. Walton, J. Chem. Soc., Perkin
Trans. 1, 1989, 347.