Formation of 2,2′-diacyl-9,9′-bifluorenylidene isomers from 2-acyl-9-bromofluorene and base-catalyzed isomerization of the formed alkenes

Article abstract of DOI:10.1246/bcsj.79.1758

We studied the formation of 2,2′-diacyl-9,9′-bifluorenylidene from 2-acyl-9-bromofluorene. The E-isomer was obtained mainly when the acyl group was a short alkyl chain, such as an acetyl group, and Z-isomer was isolated predominantly when a long acyl chai

Full text of DOI:10.1246/bcsj.79.1758

1758 Bull. Chem. Soc. Jpn. Vol. 79, No. 11, 1758–1765 (2006)
ꢀ 2006 The Chemical Society of Japan
Formation of 2,20-Diacyl-9,90-bifluorenylidene Isomers
from 2-Acyl-9-bromofluorene and Base-Catalyzed
Isomerization of the Formed Alkenes
ꢀ
Takumi Takanezawa, and Michinori Karikomi
Masahiro Minabe, Ayumi Yamazaki, Toshinobu Imai,
Department of Applied Chemistry, Faculty of Engineering, Utsunomiya University,
7-1-2 Yoto, Utsunomiya 321-8585
Received April 18, 2006; E-mail: minabe@cc.utsunomiya-u.ac.jp
We studied the formation of 2,20-diacyl-9,90-bifluorenylidene from 2-acyl-9-bromofluorene. The E-isomer was
obtained mainly when the acyl group was a short alkyl chain, such as an acetyl group, and Z-isomer was isolated pre-
dominantly when a long acyl chain such as stearoyl group was attached. 9,90-Bifluorenylidene was formed via anti elim-
ination of hydrogen bromide from the intermediate 9-bromo-9,90-bifluorenyl. This indicated that the stereochemistry of
2,20-diacyl-9,90-bifluorenylidene was determined by the configuration of 9-bromo-9,90-bifluorenyl isomers, giving Z- and
E-isomers in a ratio of 1:1. Facile base-catalyzed Z/E isomerization of 9,90-bifluorenylidenes followed the formation of
the alkenes, which showed different stereoselectivity depending upon the length of the acyl side-chain. On the other
hand, thermal treatment of 2,20-diacyl-9,90-bifluorenylidene afforded a ca. 1:1 mixture of Z- and E-isomers, suggesting
that thermal isomerization took place through a different pathway than base-catalyzed isomerization.
We have reported that the reaction of 2-acetyl- (1a) and
2-stearoyl-9-bromofluorene (1b) with base affords (Z)- and
(E)-2,20-diacyl-9,90-bifluorenylidenes, 2a and 2b, respective-
ly.1 Compound 1a gives mainly the E-isomer, while 1b yields
Z-isomer predominantly (Scheme 1). (E)-2a is a little more
stable than (Z)-2a by calculation and its side chains are posi-
tioned to decrease steric repulsion. On the other hand, (Z)-2b
is more stable than (E)-2b; we attribute it to an increase in
the amount of aggregative intra-molecular interaction between
side chains.1 The reaction of 1 proceeds via the intermediate
9-bromo-9,90-bifluorenyl (3), and the rate-determining step of
1 to 2 is involved in the conversion of 1 to 3.2 The dehydrobro-
mination of 3 takes place via an E2 pathway. The reaction of
1a between insufficient amount of base roughly gives a 1:1
mixture of threo- and erythro-3a, and the configuration of 3a
influences the diastereomeric ratio of 2a, or Z/E is 1:1. We
assume that the selectivity discussed above is caused by the
facile isomerization of 2 in the presence of base during the
reaction, based on preliminary experiments of the base-cata-
lyzed isomerization of 2.2
tional method3,4 in many procedures.5–10 Unsubstituted com-
0
pound 2 (R ¼ H) is nonplanar, twisting at the central C9–C9
double bond with an angle of about 43 degree.11–13 The Z/E
isomerization of derivatives of 2 having substituents at the
1,10-14–16 and 2,20-positions16–18 has been studied experimen-
tally and theoretically. ꢀG value for the Z/E isomerization
of 2,20-dimethyl-9,90-bifluorenylidene was determined to be
104 kJ molꢁ1 18 To the best of our knowledge there is no study
.
concerning the stereoselective synthesis of derivatives of 2, ex-
cept for 1,2-[(Z)-90,900-bifluorenylidene-10,200-diyldioxy]ethane
and its analogs.17
Selectivity, i.e., formation of a single product, is a very im-
portant in organic synthesis. For example, the Wittig reaction
is a stereo-controlled procedure for alkene synthesis,19 and the
reaction between a stable ylide and carbonyl compound gives
mainly an E-alkene, and the reaction with a reactive ylide af-
fords Z-isomer as a main product. Another example is the se-
lective formation of methyl cinnamate from both methyl threo-
and erythro-2,3-dibromo-3-phenylpropionate by microwave ir-
radiation in 1-methyl-3-pentylimidazolium fluoroborate.20
In the present paper, the stereoselective synthesis of the
alkene moiety of 2, controlled by the length of the acyl side-
chains is reported. The framework of 2 is one of the partial
structural motifs found in fullerenes, and thus, 2 and its deriv-
atives may be useful materials for the bottom-up synthesis of
elaborate fullerene derivatives.21–24
Alkene 2 is easily prepared by the reaction of 1 with meth-
anolic potassium hydroxide in acetone, which is a conven-
R
(R)
Br
a
COCH3
Br
First, the reaction of 1 having various acyl groups with po-
tassium hydroxide in methanol/acetone is discussed, and it is
shown that the observed Z/E ratio depends upon the length of
the acyl side-chain. Then, the dehydrobromination of a mix-
ture of threo- and erythro-3a and 3b is presented and suggest
that an E2 pathway is involved. The third section concerns the
COC17H35
b
R
(R)
R
R
1
2 (Z, E)
3 (threo, erythro)
Scheme 1.
Published on the web November 9, 2006; doi:10.1246/bcsj.79.1758

M. Minabe et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 11 (2006) 1759
(R)
100
75
Ratio of Isomer:
:B-
2a (R = Ac): Z <E
2b (R = St): Z >E
Br
O
50
(R)
R
R
R
25
1
2
4
0
0
Scheme 2.
2
4
6
8
10 12 14 16 18 20
Length of acyl side chain
Fig. 1. Z/E selectivity vs length of acyl side chain.
Table 1. Z/E Ratio of 2 by the Dimerization of 1 with a Non-Alkanoyl Moiety
Reaction conditions
1
2
Chemical shift (part)/ppm
Z
Run
Molar
ratio of
Ratio of
acetone/1
E
Substituent
Mp (ꢂC)
Yield
/%
Z/E
zfflfflfflfflfflfflfflfflfflfflffl}|fflfflfflfflfflfflfflfflfflfflffl{
zfflfflfflfflfflfflfflfflfflfflffl}|fflfflfflfflfflfflfflfflfflfflffl{
KOH/1
(mL/mmol)
H(8)
H(1)
H(8)
H(1)
1
2
3
4
5
6
7
8
9
10
COOCH3
COOC10H21
COOC16H33
COCH2CH2COOCH3
COCH2CH2COOC10H23
COCH2CH2COOC16H33
COC6H5
COC6H4-CH3(p)
C2H5
C18H37
103–104
45–47
78–79
112–115
69–71
82–83
135–137
129–131
51–53
0.9
0.9
0.9
0.9
0.9
0.9
2.1
2.0
1.8
1.8
30
30
30
30
30
30
30
30
30
30
91
67
72
75
32
15
82
72
68
47
46/54
50/50
51/49
38/62
44/56
48/52
36/64
44/56
58/42
58/42
8.43
8.43
8.44
8.45
8.45
8.45
8.41
8.40
8.35
8.35
8.99
9.00
9.01
8.99
8.99
8.99
8.73
8.70
8.28
8.28
8.35
8.37
8.38
8.37
8.37
8.37
8.34
8.34
8.38
8.38
9.04
9.07
9.08
9.03
9.03
9.03
8.78
8.76
8.23
8.23
66–68
base-catalyzed isomerization of 2, which only occurs for acyl
compounds, but not alkyl and halogen derivatives, indicating
that the conversion occurs through an enolate species. The last
topic of this paper is the thermal isomerization of 2, which af-
forded ca. 1:1 mixture of (Z)- and (E)-2, despite of the length
of the side chain. This transformation should be explained via
biradical species at the central double bond.
(R)
R
Br
KOH
Br
(R)
2 (Z / E ≅ 1/1)
R
R
R
3 (threo/ erythro ≅ 1/1)
1
Results and Discussion
Scheme 3.
Figure 1 shows the Z-selectivity versus the length of acyl
side-chain in the formation of 2 by the reaction of 1 with meth-
anolic potassium hydroxide in acetone.1 These findings indi-
cate that Z/E selectivity depends on the length of the acyl
side-chain, despite the reproducibility of the data being low.
Ratio of components in the reaction mixture was determined
by comparing their characteristic signals on the 1H NMR.2 Sol-
ubility of acylated 1 and 2 was fairly low compared to that of
unsubstituted 1 and 225 and the amount of solvent significantly
affected the yield and stereoselectivity of 2. A large excess
amount of solvent caused to the formation of an oxidized prod-
uct, fluorenone 426 (Scheme 2).
ꢀ 8.37 ppm is the 8-hydrogen of (E)-2a. The other signals at
ꢀ 8.95 and 8.44 ppm did not correlate to each other are the
1- and 8-hydrogens, respectively, of (Z)-2a.
In order to clarify the effect of acyl substituent, reactions of
1 having substituent other than alkanoyl moiety were exam-
ined as are summarized in Table 1. In the cases of esters of
2-fluorenecarboxylic acid and -succinic acid, the Z/E ratio of
2 changes depending upon the length of side chain, but to a
lesser degree than that of the simple acyl derivatives. The Z/
E ratio of 2,20-diethyl-9,90-bifluorenylidene (2, R ¼ Et) is sim-
ilar to that of octadecyl derivative, suggesting that the carbonyl
group attached to the 2-position of the fluorenyl ring is essen-
tial for Z/E selectivity depending on the side chain.
In our previous paper,2 the reaction from 1 to 2 was shown
to take place via the intermediate 3, and the conversion of 1 to
3 proceeds non-stereoselectively giving a 1:1 mixture of threo-
and erythro-327 (Scheme 3). Also, 3 is dehydrohalogenated via
the E2 sequence, indicating that threo- and erythro-3 afford Z-
and E-2, respectively by anti-elimination. Here, we describe
The separation of pure Z and E isomers of 2 has been unsuc-
cessful because of their similar solubility and because they iso-
merize. The structures of the derivatives of 2 were determined
1
by H NMR. The spectrum of each isomeric mixture of 2 has
two singlets (ꢀ 8.95 and 8.99 ppm) and two doublets (ꢀ 8.37
and 8.44 ppm). The peak at ꢀ 8.99 ppm of 2a correlates to
the signal at ꢀ 8.37 ppm on the NOESY spectrum, and thus,
it is assigned to the 1-hydrogen of (E)-2a, and the signal at

1760 Bull. Chem. Soc. Jpn. Vol. 79, No. 11 (2006)
Isomerization of Diacyl-9,90-bifluorenylidene
Ac
HO
Ac
Br
O
O
Me
O
Me
O
O
Ac
Ac
7a
3a
5a
6a
R
Ac
Cl
NBS
3
R
9a
Scheme 4.
8
Ac
Ac
Ac
OH
Ac
Ac
OH
H
OH
OH
H
H
H
Ac
Ac
Ac
(9R,9'R)
Ac
(9R,9'S)
(9S,9'S)
(9S,9'R)
threo-7a
erythro-7a
Scheme 5.
Table 2. Crystallographic Data for threo-7a
Empirical formula
Formula weight
Crystal dimensions/mm3
Temperature/K
Crystal system
C30H22O3
430.50
0:45 ꢃ 0:44 ꢃ 0:30
93 ꢄ 1
monoclinic
P21=n (No. 14)
12.0924(2)
14.5204(2)
12.6797(10)
102.5653(10)
2173.06(18)
4
Space group
˚
a/A
˚
b/A
˚
c/A
Fig. 2. ORTEP drawing of threo-7a with atomic labeling
scheme. Aromatic and methyl H atoms are omitted for
clarity.
ꢁ/deg
3
˚
V/A
Z value
Dcalcd/g cmꢁ3
1.316
˚
Radiation/A
Cu Kꢂ (ꢃ ¼ 1:54187)
6.68
904.00
0.0472
0.1197
1.058
threo- and erythro-2,20-diacetyl-9,90-bifluorenyl-9-ol (7a). The
stereochemistry of threo- and erythro-7a (Scheme 5) was de-
termined by comparison of the spectral and physical data with
those of authentic dl- and meso-2,20-diacetyl-9,90-bifluorenyl
(8a),2 respectively. X-ray crystallographic analysis of threo-
7a (racemic) was used to determine the molecular structure.
Table 2 shows the crystallographic data for threo-7a. An
ORTEP drawing of threo-7a (9S,90R) is shown in Fig. 2. Both
fluorene moieties are twisted from an eclipsed conformation
by ca. 60 degree in the direction of the acetyl group on the
opposite fluorene ring.
ꢄ(Cu Kꢂ)/cmꢁ1
F000
R factor
Rw2 factor
Goodness of fit indicator
both the synthesis of 3a in order to confirm the structure and
the reaction between 3a (or 3b) and base to understand the
Z/E isomeric ratio of 2.
An isomeric mixture 3a was synthesized as is shown in
Scheme 4. 2-(2-Methyl-1,3-dioxolan-2-yl)fluorene (5a) was
lithiated and reacted with the corresponding ketone 6a, giving
The treatment of threo-7a with phosphorus tribromide in the
presence of pyridine afforded erythro-3a and a small amount

M. Minabe et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 11 (2006) 1761
Table 3. Reaction of 3a and 3b with KOH
Run
Reactant
Reaction conditionsa)
Products
Recovd.
threo/
erythro
2
threo/
erythro
Wt
/mg (mmol)
Molar ratio
of KOH/3
MeOH
/mL
Acetone
/mL
4
/%
Wt
/%
3 (R)
/%
Z/E
1
2
3
4
3a (Ac)
3a (Ac)
3b (St)
3b (St)
3b (St)
50/50
87/13
50/50
50/50
51/49
24.7(50)
0.5
1
0.5
1
50
110
50
110
100
5
10
10
10
10
31
54
46
91
83
43/57
20/80
48/52
94/6
3
21
3
4
6
65
49/51
49.6(100)
94.7(101)
94.5(100)
94.4(100)
35
52/48
5b)
1.2
99/1
a) Stirred at room temperature for 1 h. b) t-BuOK was used instead of KOH.
Table 4. Z/E Isomerization of 2 in the Presence of KOH
Run
Reactant 2
Reaction conditionsa)
Products
Substituent
Ac (2a)
Ac (2a)
Ac (2a)
St (2b)
St (2b)
St (2b)
COOCH3
COOC16H33
COCH2CH2COOCH3
COCH2CH2COOC16H33
COC6H5
C2H5
C18H37
Z/E
2
Wt
/mmol
Molar ratio
of KOH/3
MeOH
/mL
Acetone
/mL
Temp.
/ꢂC
4
/%
/%
Z/E
1
2
54/46
53/47
50/50
62/38
54/46
60/40
45/55
52/48
34/66
52/48
40/60
66/34
56/44
42
50
28
50
50
25
50
25
50
25
40
48
50
None
0.5
0.5
None
0.5
2
0.5
0.5
0.5
1
25
50
25
25
25
110
50
50
50
50
25
25
25
5
5
2.8
5
5
5
5
2.5
5
2.5
4
rt
26–30
36
rt
rt
97
85
94
89
82
80
91
90
93
88
97
90
94
53/47
11/89
3/97
3
10
3b)
4
64/36
82/18
85/15
10/90
73/27
27/73
70/30
4/96
5
5
14
6c)
7d)
8
rt
23–25
22–24
21–24
23–26
22–23
rt
9
10
11
12
13
0.6
0.6
0.6
5
5
64/36
60/40
rt
a) Stirred for 1 h. b) t-BuOK in t-BuOH was used. c) Pyridine was used instead of acetone. d) Stirred for 30 min.
of threo-3a. Also, similar reaction of erythro-7a yielded main-
ly threo-3a. The 1H NMR spectra of threo- and erythro-3a
were similar to those of threo- and erythro-7a, respectively,
although perfect separation of the isomers of 3a was impossi-
ble. Compound 7a was converted into 2,20-diacetyl-9-chloro-
9,90-bifluorenyl (9a) in order to clarify the relationship be-
tween 7a and 3a. The reaction between threo-7a and thionyl
chloride in the absence of base afforded threo-9a and a small
amount of erythro-9a with a low yield (<20%). Treatment of
erythro-7a showed a similar result, yielding mainly erythro-
9a. On the other hand, treatment of threo-7a with thionyl chlo-
ride only gave erythro-9a in the presence of pyridine, and
threo-9a was obtained from erythro-7a by a similar procedure.
A mixture of 3a was also obtained by the reaction of 8a with
N-bromosuccinimide (NBS) in a ratio of ca. 1:1. A mixture of
threo- and erythro-3b was obtained by a meso and dl mixture
anti-elimination. The preferential formation of (E)-2a and
(Z)-2b (Runs 2, 4, and 5) suggests that the isomerization of
2 occurs after its formation in the presence of base, as summa-
rized in the Table 4.
Dimer 2 was treated with base under the conditions similar
to those of 3 to 2, Table 4. Compound (Z)-2a converted into
(E)-2a (Runs 2 and 3), and (E)-2b changed into (Z)-2b (Runs
5 and 6); however no change within experimental error was
observed in the absence of base (Runs 1 and 4). Similar Z/E
selectivity dependence upon the length of side chain was ob-
served with different acyl groups were used (Runs 7–11), but
not when 2-alkyl groups were used (Runs 12 and 13). The find-
ings suggest both that the base-catalyzed transformation of 2
follows its formation from 3 in the presence of base (Table 3)
and that the carbonyl group must be attached directly to the
fluorene nuclei for the base-catalyzed isomerization.
Compound 2 was also heated in refluxing toluene, Table 5.
The Z/E ratio after heating for 5 min approached 1:1 when
either acyl or alkyl groups were present. On the other hand,
the isomers of 2,20-dibromo-9,90-bifluorenylidene28 did not
thermally equilibrate within experimental error. The thermally
promoted Z/E equilibrium has been explained by a biradical
transition state (10)14 (Scheme 6) via pyramidalization17 of
the crowded molecule. The isomerization is encouraged by
both the ground-state strain13 and the stabilized transition state,
which has a relatively low barrier (84–105 kJ molꢁ1).14,16–18
The bulkier compound, (Z)-N,N0-bis(1-phenylethyl)-9,90-
1
of 8b and NBS, and its H NMR spectrum was similar to that
of 3a.
Table 3 summarizes the dehydrobromination of 3a and 3b
with base. Using less than a stoichiometric amount of base af-
forded a mixture of (Z)- and (E)-2a and 2b in a ratio roughly
corresponding to the composition of 3a and 3b (Runs 1 and 3).
On the other hand, (E)-2a or (Z)-2b was the main product, in-
dependent of the ratio of 3a and 3b, in the presence of a stoi-
chiometric amount of base. These findings are consistent with
the results obtained from the treatment of 1a and 1b with
base,2 supporting that the conversion of 3 to 2 proceeds via

1762 Bull. Chem. Soc. Jpn. Vol. 79, No. 11 (2006)
Table 5. Thermal Treatment of 2a)
Isomerization of Diacyl-9,90-bifluorenylidene
Run
Reactant 2
Toluene
/mL
Products
Substituent
Ac (2a)
St (2b)
Z/E
2
Wt
/mmol
4
/%
Yield/%
Z/E
1
2
3b)
4
5
6
7
8
10/90
85/15
65/35
30/70
9/91
60
50
20
50
50
50
50
150
20
10
20
40
10
5
88
84
84
100
83
92
44/56
53/47
55/45
61/39
40/60
43/57
56/44
45/55
12
13
16
St (2b)
COCH2CH2COOCH3
COC6H5
14
5
COC6H4-CH3(p)
C18H37
Br
12/88
1/99
41/59
5
20
100
100
a) Refluxed for 5 min. b) Cyclohexanone was used as solvent.
ment to reduce steric repulsion, giving mainly (E)-2. On the
other hand, long acyl chain in a kinetically syn arrangement
is caused by the aggregative intra-molecular interaction be-
tween the long side chains, affording (Z)-2 as a main product.
This is supported by the enthalpy of formation (ꢀHf) and rel-
ative potential energy (Erel) estimated by calculations on the
optimized structure of 2 using quantum mechanics program
PM3 and molecular mechanics program MM+, respectively.1
R
10
R
For the acetyl derivatives, (E)-2a (ꢀHf ¼ 240 kJ molꢁ1, Erel
¼
Scheme 6.
94:5) was equally or slightly more stable than (Z)-2a (ꢀHf ¼
240 kJ molꢁ1, Erel ¼ 98:0). However, for the stearoyl deriva-
tives, (E)-2b (ꢀHf ¼ ꢁ474 kJ molꢁ1, Erel ¼ 117) was clearly
less stable than (Z)-2b (ꢀHf ¼ ꢁ513 kJ molꢁ1, Erel ¼ 110).
The optimized structure of (Z)-2b showed two acyl chains
stretched along one another (the nearest H–H distance, ca.
0.25 nm). The aggregative interaction between the side chains
is tentatively assumed to be the van der Waals force between
the alkyl groups, which is observed in some liquid crystals30
and phospholipids.31
bifluorenylidene-2,20-dicarboxamide isomerized in refluxing
ethanol.17 The reaction conditions in Table 5 should not be de-
fined to be in equilibrium control; however, the findings sug-
gest that thermal conversion proceeds via an another pathway
than the base-catalyzed isomerization mentioned above.
The base-catalyzed isomerization of 2 took place in only the
compounds with a carbonyl group. The base-catalyzed trans-
formation most likely occurs through a dipolar transition state
(11), Scheme 7, which is similar to the isomerization of
heterofulvene derivatives.29 Enolate 11 should be unstable,
and the cation 110 or alcohol 1100 is more reasonable species
than 11 in the presence of KOH, even though an alcohol like
7 has not been detected in the isomeric conversion of 2. In
the case of short acyl side chains, they are in an anti arrange-
Experimental
The melting points are uncorrected. The NMR (CDCl3) and IR
(KBr pellet) spectra were recorded with a Varian VXR-300 and
JASCO FT/IR-430, respectively. The elemental analyses and
COR1
COR1
:B-
:B-
COR1 COR1
C
COR1
O
R1
(E)-2
(Z)-2
11
COR1
COR1
HO
C
C
O-K+
O-K+
R1
11"
R1
Scheme 7.
11'

M. Minabe et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 11 (2006) 1763
mass spectra were measured with an EA 1108 CHNS-O (Fison
Instruments) and JMX-AX 500 (JEOL, 70 eV), respectively.
Molecular modelings were performed using the program pack-
age HyperChem (version 5.1, HyperCube Inc.). Programs MM+
and PM3 were employed without any modification of the param-
eters provided. A structure was optimized initially by MM+, and
then the structure that was obtained was optimized again by PM3.
This was repeated until the energy of the model structure was
nearly unchanged. ꢀHf and Erel were calculated for the optimized
structure by PM3 and MM+, respectively. The solvent effects
were omitted in all calculations.
X-ray Crystallographic Analysis. The intensity data were
collected at ꢁ180 ꢄ 1 ꢂC on a Rigaku RAXIS RAPID imaging
plate diffractometer with graphite-monochromated Cu Kꢂ radia-
tion in ! scan mode up to 2ꢅmax ¼ 136:5ꢂ. Unit-cell parameters
were determined by a least-squares refinement. The crystal data
and experimental details are listed in Table 2. Crystallographic
data have been deposited with Cambridge Crystallographic Data
Centre: Deposition number CCDC-609981 for threo-7a. Copies
of the data can be obtained free of charge via http://www.ccdc.
cam.ac.uk/conts/retrieving.html (or from the Cambridge Crystal-
lographic Data Centre, 12, Union Road, Cambridge, CB2 1EZ,
UK; Fax: +44 1223 336033; e-mail: deposit@ccdc.cam.ac.uk).
Materials. 2-Acylfluorene was obtained by a procedure sim-
ilar to the synthesis of 2-acetylfluorene:32 2-Butanoylfluorene, mp
118–120 ꢂC; 2-hexanoyl-, mp 119–120 ꢂC; 2-octanoyl-, mp 102–
103 ꢂC; 2-decanyl-, mp 105–106 ꢂC; 2-dodecanoyl-, mp 96–
97 ꢂC; 2-tetradecanoyl-, mp 97–98 ꢂC; 2-hexadecanoyl-, mp 97–
99 ꢂC; 2-octadecanoyl- (or 2-stearoyl-), mp 94–95 ꢂC. 2-Acyl-9-
bromofluorene was obtained by the reaction of 2-acylfluorene with
NBS in CCl4:33 2-Acetyl-9-bromofluorene, mp 118–119 ꢂC; 2-bu-
tanoyl-, mp 121–123 ꢂC; 2-hexanoyl-, mp 93–95 ꢂC; 2-octanoyl-,
mp 70–71 ꢂC; 2-decanoyl-, mp 83–84 ꢂC; 2-dodecanoyl-, mp
86–89 ꢂC; 2-tetradecanoyl-, mp 86–87 ꢂC; 2-hexadecanoyl-, mp
86–88 ꢂC; 2-octadecanoyl-, mp 92–94 ꢂC.
9.83%.
All of experiments in Fig. 1 and Table 1 were carried out under
the following conditions:25 a solution of KOH (1.2–2.3 molar
amount to 1) in MeOH (1.4–10 mL per 1 mmol of 1) was added
to a suspension of 1 in acetone (4.5–120 mL per 1 mmol of 1)
for 5 min at room temperature (10–20 ꢂC), and the mixture was
stirred for 1 h.
2-(2-Methyl-1,3-dioxolan-2-yl)fluorene (5a) and 2-(2-Meth-
yl-1,3-dioxolan-2-yl)fluorenone (6a). A solution of 2-acetyl-
fluorene (5.20 g, 25 mmol), ethylene glycol (7.75 g, 125 mmol),
and p-toluenesulfonic acid monohydrate (190 mg, 1.0 mmol) in
benzene (250 mL) was refluxed for 6 h. After cooling to room tem-
perature, the reaction mixture was neutralized with K2CO3, the
solvent was evaporated in vacuo, and the residue was chromato-
graphed on alumina with benzene, giving 5a (4.20 g, 67%): mp
103–104 ꢂC (from hexane); 1H NMR ꢀ 1.72 (s, Me), 3.81–3.85
(2H, m), 3.90 (2H, s), 4.05–4.10 (2H, m), 7.29 (td, J ¼ 7:4, 1.4
Hz, H7), 7.37 (td, J ¼ 7:4, 0.6 Hz, H6), 7.51 (dt, J ¼ 8:1, 1.5 Hz,
H3), 7.54 (d, J ¼ 7:5 Hz, H8), 7.66 (s, H1), 7.75 (d, J ¼ 8:1 Hz,
H4), 7.78 (d, J ¼ 7:5 Hz, H5); Ms m=z 252 (Mþ), 237, 221, 193,
165. Anal. Found: C, 80.59; H, 6.46%. Calcd for C17H16O2: C,
80.92; H, 6.39%.
A solution of 2-acetylfluorene (5.20 g, 25 mmol), triton B
(1.0 mL, 40% in MeOH) in pyridine (150 mL) was stirred with
bubbling air at room temperature for 3 h. Upon dilution with
dil HCl, the precipitate was chromatographed on silica gel with
toluene, yielding 2-acetyl-9-fluorenone (3.90 g, 70%): mp 158–
160 ꢂC (from EtOH). Acetalization of 2-acetyl-9-fluorenone
(1.11 g, 5.0 mmol) by the manner similar to the case of 5a afforded
6a (1.01 g, 77%): mp 113–114 ꢂC (from hexane); IR 1712 cmꢁ1
;
1H NMR ꢀ 1.67 (s, Me), 3.78–3.82 (2H, m), 4.04–4.09 (2H, m),
7.29 (td, J ¼ 7:1, 1.8 Hz), 7.45–7.53 (3H, m), 7.61–7.67 (2H, m),
7.80 (d, J ¼ 1:5 Hz, H1); Ms m=z 266 (Mþ), 251, 207, 179. Anal.
Found: C, 76.39; H, 5.27%. Calcd for C17H14O3: C, 76.67; H,
5.30%.
Reaction of 1a with KOH. Typical Procedure. A solution
of KOH (61.5 mg, purity 83%, 0.91 mmol) in MeOH (1 mL)
was added to a mixture of 1a (144 mg, 0.50 mmol) in acetone
(30 mL) at 9–10 ꢂC for 5 min, and the mixture was stirred for
1 h. The precipitate was collected by filtration, washed with water,
and extracted with hot EtOH (5 mL) giving 2a. All filtrates were
poured into water, and the precipitate was collected and weighed.
threo- and erythro-2,20-Diacetyl-9,90-bifluorenyl-9-ol. Bu-
tyllithium (3.5 mL, 1.6 M, 5.5 mmol) (1 M = 1 mol dmꢁ3) was
added dropwise to a solution of 5a (1.26 g, 5.0 mmol) in ether
(50 mL) with stirring at ꢁ20 ꢂC over 10 min under an argon atmo-
sphere. After stirring for an additional 1 h, 6a (1.07 g, 4.0 mmol) in
ether (50 mL) was added, and the resulting mixture was refluxed
for 30 min. After quenching with 1 M NH4Cl solution (20 mL),
the solvent was evaporated to give oily residue (2.16 g).
A mixture of the oily residue (3.55 g; collected after several
runs as above) and p-toluenesulfonic acid (95 mg, 0.5 mmol) in
acetone (50 mL) was refluxed for 1 h. After neutralization of the
solution with K2CO3, the solution was poured into water, afford-
ing an oil of 7a (2.60 g, 88%). Fractional recrystallization of 7a
from ether gave threo-7a (412 mg): mp 235–236 ꢂC; Rf 0.75
1
The ratio was determined by H NMR: 2a (88.2 mg, 86%; Z/E =
18/82) and 4a (4.1 mg, 3.7%). 1H NMR of (Z)-2a: ꢀ 2.57 (s, 2Me),
0
0
7.34 (td, J ¼ 7:8, 1.2 Hz, H7;7 ), 7.42 (td, J ¼ 7:8, 1.2 Hz, H6;6 ),
0
0
0
7.78–7.82 (m, H4;5;4 ;5 ), 8.01 (dd, J ¼ 8:1, 1.5 Hz, H3;3 ), 8.44
0
0
(d, J ¼ 7:8 Hz, H8;8 ), 8.95 (d, J ¼ 1:5 Hz, H1;1 ). (E)-2a: ꢀ 2.57
0
0
(s, 2Me), 7.34 (t, J ¼ 7:5 Hz, H7;7 ), 7.43 (t, J ¼ 7:5 Hz, H6;6 ),
0
0
0
7.78–7.82 (m, H4;5;4 ;5 ), 8.01 (dd, J ¼ 8:0, 1.4 Hz, H3;3 ), 8.37
(d, J ¼ 7:5 Hz, H8;8 ), 8.99 (d, J ¼ 1:4 Hz, H1;1 ). Anal for 2a
(Z/E mixture, recrystallized from AcOEt). Found: C, 87.47; H,
4.49%. Calcd for C30H20O2: C, 87.35; H, 4.89%.
(AcOEt/toluene = 1/1); IR 3448, 1676, 1659 cmꢁ1 1H NMR ꢀ
;
0
0
2.24 (s, Me), 2.27 (s, Me), 2.61 (s, OH), 4.88 (s, H9), 6.97–7.06
(2H, m), 7.32–7.60 (8H, m), 7.66 (1H, d, J ¼ 7:5 Hz), 7.74–7.82
(3H, m). Ms m=z 430 (Mþ), 368, 208, 165. Anal. Found: C,
83.42; H, 4.94%. Calcd for C30H22O3: C, 83.70; H, 5.15%.
erythro-7a (700 mg) was isolated from the mother solution: mp
140–168 ꢂC; Rf 0.68 (AcOEt/toluene = 1/1); IR 3478, 3225,
1H NMR of (Z)-2b: ꢀ 0.88 (t, J ¼ 7:2 Hz, 2Me), 1.24 (bs), 1.68
(q, J ¼ 7:2 Hz), 2.90 (t, J ¼ 7:2 Hz), 7.30 (td, J ¼ 7:5, 1.2 Hz,
0
0
0
0
H
7;7 ), 7.40 (td, J ¼ 7:5, 1.2 Hz, H6;6 ), 7.76–7.82 (m, H4;5;4 ;5 ),
0
0
8.01 (dd, J ¼ 8:1, 1.2 Hz, H3;3 ), 8.44 (d, J ¼ 7:5 Hz, H8;8 ), 8.95
1
(d, J ¼ 1:2 Hz, H1;1 ). (E)-2b: ꢀ 0.88 (t, J ¼ 7:2 Hz, 2Me), 1.24
1672, 1659 cmꢁ1; H NMR ꢀ 2.38 (s, Me), 2.44 (s, Me), 2.61 (s,
0
(bs), 1.68 (q, J ¼ 7:2 Hz), 2.90 (t, J ¼ 7:2 Hz), 7.32 (td, J ¼ 7:5,
OH), 4.89 (s, H9), 6.98–7.05 (1H, m), 7.06–7.14 (2H, m), 7.16–
7.34 (3H, m), 7.46–7.52 (3H, m), 7.58 (1H, d, J ¼ 7:5 Hz), 7.60
(1H, d, J ¼ 8:1 Hz), 7.72–7.78 (1H, m), 7.87–7.96 (2H, m). Ms
m=z 430 (Mþ), 369, 208, 165. Anal. Found: C, 80.24; H, 5.27%.
0
0
1.2 Hz, H7;7 ), 7.42 (td, J ¼ 7:5, 1.2 Hz, H6;6 ), 7.77–7.82 (m,
0
0
0
H
H
4;5;4 ;5 ), 8.01 (dd, J ¼ 7:8, 1.5 Hz, H3;3 ), 8.37 (d, J ¼ 7:5 Hz,
0
0
8;8 ), 8.99 (d, J ¼ 1:5 Hz, H1;1 ). Anal for 2b (Z/E mixture).
Found: C, 86.43; H, 9.85%. Calcd for C62H84O2: C, 86.45; H,
Calcd for C30H22O3 H2O: C, 80.33; H, 5.39%.
ꢅ

1764 Bull. Chem. Soc. Jpn. Vol. 79, No. 11 (2006)
Isomerization of Diacyl-9,90-bifluorenylidene
2,20-Diacetyl-9-bromo-9,90-bifluorenyl (3a). A solution of
PBr3 (670 mg, 2.5 mmol) in ether (5 mL) was added dropwise to
a solution of threo-7a (215 mg, 0.5 mmol) and pyridine (395 mg,
5.0 mmol) in ether (300 mL) at ꢁ25 ꢂC under an argon atmo-
sphere, and the mixture was stirred at ꢁ10 ꢂC for 1 h and 5 ꢂC
for 3 d. Upon quenching with water, neutralizing with NaHCO3,
drying with Na2SO4, and evaporating the solvent, a mixture of
3a (173 mg) was obtained as an oil: ratio of threo- and erythro-
7.97 (m). Anal. Found: C, 77.62; H, 9.01%. Calcd for C62H85-
O2Br H2O: C, 77.55; H, 9.13%.
ꢅ
Reaction of 3a (or 3b) with KOH. Typical Procedure.
A
suspension of 3a (24.7 mg, 50 mmol, threo/erythro = 50/50) in
acetone (5 mL) was stirred with KOH/MeOH (23 mmol of KOH
in 50 mL of MeOH; 0.46 molar amount based on 3a) at 24–
26 ꢂC for 1 h. After adding to water, 2a (31.3%, Z/E = 43/57),
4a (3%), and recovered 3a (64.7%, threo/erythro = 49/51) were
1
1
3a was determined to be 26/74 by H NMR. erythro-3a: Rf 0.67
(AcOEt/toluene = 1/2); IR 1670 cmꢁ1
Me), 5.22 (1H, s, H9), 7.10–8.20 (m).
confirmed by H NMR of the precipitates.
;
1H NMR ꢀ 2.26 (6H, s,
Z/E Isomerization of 2 in the Presence of KOH. Typical
Procedure. To a suspension of 2a (20.6 mg, 50 mmol, Z/E =
53/47) in acetone (5 mL) was added KOH (25 mmol in 50 mL of
MeOH) at once, and the whole was stirred at 26–30 ꢂC for 1 h.
Upon the usual treatment of the reaction mixture, 2a (85.0%,
Z/E = 11/89) and 4a (10.3%) were determined by NMR.
Thermal Isomerization of 2. Typical Procedure. A mixture
of 2a (24.6 mg, 60 mmol, Z/E = 10/90) and toluene (20 mL)
was refluxed for 5 min. Upon evaporation of the solvent, 2a
(87.9%, Z/E = 44/56) and 4a (12.3%) were determined by
NMR to be in the residue.
erythro-7a (215 mg, 0.5 mmol) was treated in a manner similar
to above, and a mixture of threo- and erythro-3a was isolated in
a ratio of 82/18. threo-3a: Rf 0.57 (AcOEt/toluene = 1/2); IR
1
1670 cmꢁ1; H NMR ꢀ 2.44 (3H, s, Me), 2.46 (3H, s, Me), 5.23
(1H, s, H9), 7.10–7.94 (m).
A mixture of 8a2 (828 mg, 2.0 mmol; meso/dl = 77/23) and
NBS (545 mg, 3.1 mmol) in CCl4 (60 mL) was refluxed for 18 h,
giving 3a (threo/erythro = 45/55 by NMR).
2,20-Diacetyl-9-chloro-9,90-bifluorenyl (9a). A mixture of
erythro-7a (108 mg, 0.25 mmol; threo/erythro = <1=>99), SOCl2
(0.21 mL, 3.0 mmol), pyridine (800 mg, 10 mmol) in benzene
(30 mL) was stirred at room temperature for 1.5 h. Upon addition
of water, the organic layer was dried over Na2SO4, and the sol-
vent was evaporated in vacuo. Based on 1H NMR, the threo/
erythro ratio of 9a (mp 222–226 ꢂC; 78 mg, 69%) was >99=<1.
We wish to thank Mr. K. Shirouzu of Rigaku Corporation
for X-ray crystallographic analysis.
References
Threo-9a: Rf 0.57 (AcOEt/toluene = 1/2); IR 1685, 1673 cmꢁ1
;
1 M. Minabe, T. Oba, M. Tanaka, K. Kanno, M. Tsubota,
Chem. Lett. 2000, 498.
2 A. Oota, T. Imai, A. Yamazaki, T. Oba, M. Karikomi,
M. Minabe, Bull. Chem. Soc. Jpn. 2006, 79, 333.
1H NMR (50 ꢂC) ꢀ 2.43 (3H, s, Me), 2.44 (3H, s, Me), 5.08 (1H,
s, H9), 6.95–7.40 (6H, m), 7.44–7.66 (5H, m), 7.77 (1H, bs), 7.88–
7.97 (2H, m).
threo-7a (108 mg, 0.25 mmol; threo/erythro = >99=<1) was
treated by the manner similar to above, giving 9a (mp 225–227
ꢂC; 82 mg, 73%; threo/erythro = 2/98). erythro-9a: Rf 0.67
3
893.
4
586.
5
E. Bergmann, J. Hervey, Ber. Dtsch. Chem. Ges. 1929, 62,
M. Minabe, K. Suzuki, Bull. Chem. Soc. Jpn. 1975, 48,
R. Kuhn, H. Zahn, K. L. Scholler, Ann. Chim. (Paris)
1
(AcOEt/toluene = 1/2); IR 1679 cmꢁ1; H NMR (50 ꢂC) ꢀ 2.26
(6H, s, Me), 5.06 (1H, s, H9), 6.99 (1H, bs), 7.07 (1H, bs), 7.34–
7.52 (6H, m), 7.56–7.61 (1H, m), 7.65–7.70 (2H, m), 7.76–7.83
(2H, m), 7.96 (1H, bs).
A mixture of erythro-7a (43 mg, 0.1 mmol; threo/erythro =
<1=>99) and SOCl2 (2.88 mL, 40 mmol) in benzene (30 mL)
was stirred at room temperature for 3 h, yielding 35 mg of a
mixture, which is composed of 9a (yield 10%; threo/erythro =
<1=>99), 4a (18%), and 7a (72%, threo/erythro = <1=>99)
by NMR.
1953, 582, 196.
6
Isr. J. Chem. 1972, 10, 423.
7
M. Mashihara, Nippon Kagaku Kaishi 1973, 762.
8
Commun. 1987, 981.
I. Gosnay, E. D. Bergmann, M. Rabinovits, I. Agranat,
S. Kajigaeshi, T. Akahoshi, Y. Tanaka, S. Fujisaki,
L. L. Yeung, Y. C. Yip, T.-Y. Luh, J. Chem. Soc., Chem.
9
1874.
L. L. Yeung, Y. C. Yip, T.-Y. Luh, J. Org. Chem. 1990, 55,
Similar to above threo-7a with SOCl2 for 5 h afforded a mix-
ture of 9a (4%; threo/erythro = 86/14), 4a (16%), and 7a (80%,
threo/erythro = >99=<1).
10 C.-H. Kuo, M.-H. Tsau, D. T.-C. Weng, G. H. Lee, S.-M.
Peng, T.-Y. Luh, P. U. Biedermann, I. Agranat, J. Org. Chem.
1995, 60, 7380.
11 N. A. Bailey, S. E. Hull, J. Chem. Soc., Chem. Commun.
1971, 960.
12 S. C. Nyburg, J. S. Lee, Acta Crystallogr., Sect. C 1985,
41, 560.
13 G. Favini, M. Simonetta, M. Scottocornola, R. Todeschini,
J. Comput. Chem. 1982, 3, 178.
14 I. R. Gault, W. D. Ollis, I. O. Sutherland, J. Chem. Soc.,
Chem. Commun. 1970, 269.
15 I. Agranat, M. Rabinovits, I. Gosnay, A. Weltzen-Dagan,
J. Am. Chem. Soc. 1972, 94, 2889.
16 I. Agranat, M. Rabinovitz, A. Weitzen-Dagan, I. Gosnay,
J. Chem. Soc., Chem. Commun. 1972, 732.
17 Y. C. Yip, X. Wang, D. K. P. Ng, T. C. W. Mak, P. Chiang,
T.-Y. Luh, J. Org. Chem. 1990, 55, 1881.
18 P. U. Biedermann, A. Levy, M. R. Suissa, J. J. Stezowski,
2,20-Distearoyl-9-bromo-9,90-bifluorenyl (3b). To a mixture
of 1b (2.55 g, 5.0 mmol) in acetone (100 mL) was added NaI
(1.63 g, 11 mmol) in acetone (25 mL), and the resulting solution
was refluxed for 5 h, yielding a mixture of meso- and dl-8b
(1.93 g, 88%, isomeric ratio = 50/50): mp 115–117 ꢂC; IR 1680
1
cmꢁ1; H NMR ꢀ 0.88 (t, J ¼ 6:8 Hz, Me), 1.21–1.33 (bs), 1.57
(q), 2.58 (t, J ¼ 7:2 Hz, dl-COCH2–), 2.73 (t, J ¼ 7:2 Hz, meso-
COCH2–), 4.92 (s, dl-H9), 4.93 (s, meso-H9), 6.94–7.94 (m). Anal.
Found: C, 86.16; H, 10.20%. Calcd for C62H86O2: C, 86.24; H,
10.06%.
A mixture of meso- and dl-8b (860 mg, 1.0 mmol, isomeric
ratio = 50/50) was treated with NBS, giving a mixture of threo-
and erythro-3b (425 mg, 45%, isomeric ratio = 48/52): mp 72–
1
76 ꢂC; IR 1680 cmꢁ1; H NMR ꢀ 0.88 (t, J ¼ 6:7 Hz, Me), 1.26–
1.30 (bs), 1.57–1.66 (m), 2.56 (bs, erythro-COCH2–), 2.74 (bs,
threo-COCH2–), 5.22 (s, erythro-H9), 5.23 (s, threo-H9), 6.98–

M. Minabe et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 11 (2006) 1765
I. Agranat, Enantiomer 1996, 1, 75.
19 M. B. Smith, Organic Synthesis, McGraw-Hill Inc., New
York, 1994, p. 787.
20 B. C. Ranu, R. Jana, J. Org. Chem. 2005, 70, 8621.
21 H. E. Bronstein, N. Choi, L. T. Scott, J. Am. Chem. Soc.
2002, 124, 8870.
22 R. Shenhar, R. Beust, S. Hagen, H. E. Bronstein, I. Willner,
L. T. Scott, M. Rabinovitz, J. Chem. Soc., Perkin Trans. 2 2002,
449.
23 S. Pogodin, S. Cohen, I. Agranat, Eur. J. Org. Chem. 1999,
1979.
(0.6 mL/1 mmol of 1). K. Suzuki, Yuki Kagobutsu Gosei Hou,
Gihodo Inc., Tokyo, 1959, Vol. 11, p. 53.
26 J. Schmidt, H. Wagner, Ber. Dtsch. Chem. Ges. 1910, 43,
1797.
27 In this paper, threo-3a means racemate of (9S,90R)- and
(9R,90S)-3a, and erythro-3a is racemate of (9R,90R)- and
(9S,90S)-3a (see Scheme 5). Also, same nomenclature is applied
to 7a, 7b, and 9a.
28 K. Suzuki, Nippon Kagaku Zasshi 1954, 75, 714.
29 I. Belsky, H. Dodiuk, Y. Shvo, J. Org. Chem. 1977, 42,
2734.
24 S. Pogodin, P. U. Biedermann, I. Agranat, J. Org. Chem.
1997, 62, 2285.
30 F. Hoffmann, H. Hartung, W. Weissflog, P. G. Jones, A.
Chrapkowski, Mol. Cryst. Liq. Cryst. 1995, 258, 61.
31 C. Chatgilialoglu, C. Ferreri, M. Ballestri, Q. G.
Mulazzani, L. Landi, J. Am. Chem. Soc. 2000, 122, 4593.
32 F. E. Ray, G. Rieveschl, Jr., J. Am. Chem. Soc. 1943, 65,
836.
25 The best conditions to obtain 2 (R ¼ H) was reported as
follows: A solution of KOH (1.8 molar amount to 1) in MeOH
(0.36 mL per 1 mmol of 1) is added to a suspension of 1 in acetone
(1.2 mL per 1 mmol of 1) for 5 min at room temperature, the whole
is stirred for 1.5 h, and the deposit is extracted with hot EtOH
33 R. C. Fuson, H. D. Porter, J. Am. Chem. Soc. 1948, 70, 895.