Synthesis, isolation, and characterization of Diels-Alder adducts between 1,4-dialkoxyanthracenes and maleic anhydride

Article abstract of DOI:10.1246/bcsj.77.1385

In the Diels-Alder reaction of 1,4-dialkoxyanthracenes and maleic anhydride, which can afford syn- and anti-cycloadducts, the bridgehead methine proton of the cycloadducts has proved to be a useful probe for determining syn/anti selectivity, as supported by isolation of diastereomers, ¹H NMR spectroscopy, and X-ray analysis. In the case of methoxy and propoxy substituents, a slight anti-preference was observed, on the other hand, the reaction of 1,4-bis(benzyloxy)anthracene gave a small syn-preference. Theoretical calculations of transition states of 1,4-dimethoxyanthracene and maleic anhydride showed no stereochemical preference. From UV-vis spectra, the formation of charge transfer complexes of anthracenes and maleic anhydride is possible.

Full text of DOI:10.1246/bcsj.77.1385

ꢀ 2004 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 77, 1385–1393 (2004) 1385
Synthesis, Isolation, and Characterization of Diels–Alder Adducts
between 1,4-Dialkoxyanthracenes and Maleic Anhydride
ꢀ
Mikio Ouchi, and Akio Yoneda
Chitoshi Kitamura, Munehiro Hasegawa, Hiroshi Ishikawa, Jun Fujimoto,
Department of Materials Science and Chemistry, Graduate School of Engineering, Himeji Institute of Technology,
University of Hyogo, 2167 Shosha, Himeji, Hyogo 671-2201
Received January 8, 2004; E-mail: kitamura@eng.u-hyogo.ac.jp
In the Diels–Alder reaction of 1,4-dialkoxyanthracenes and maleic anhydride, which can afford syn- and anti-cyclo-
adducts, the bridgehead methine proton of the cycloadducts has proved to be a useful probe for determining syn/anti se-
lectivity, as supported by isolation of diastereomers, 1H NMR spectroscopy, and X-ray analysis. In the case of methoxy
and propoxy substituents, a slight anti-preference was observed, on the other hand, the reaction of 1,4-bis(benzyloxy)an-
thracene gave a small syn-preference. Theoretical calculations of transition states of 1,4-dimethoxyanthracene and maleic
anhydride showed no stereochemical preference. From UV–vis spectra, the formation of charge transfer complexes of
anthracenes and maleic anhydride is possible.
The Diels–Alder (DA) reaction is one of the fundamental or-
ganic reactions. It is well known that anthracenes give thermal
and photochemical DA reactions with alkenes readily and can
easily be reverted to the starting anthracene by a retro-DA re-
action. Recently, some groups have been reporting diastereose-
lective DA reactions of anthracenes that possess a chiral group
at the 9-position and proposing anthracene-based templates for
new asymmetric DA/retro-DA strategies.1 Even if anthracene
does not have any chiral groups, the introduction of substituent
groups on the 1-, 2-, 3-, and/or 4-positions of the anthracene
nucleus can permit the formation of two syn- and anti-diaster-
eomers (throughout this paper, syn and anti are used in the
sense that the substituents on the same side as succinic anhy-
dride are syn, the others anti) in the DA reaction.2 In general,
the importance of steric, orbital, and electrostatic factors to
control stereoselectivity has been recognized.3 With respect
Scheme 1.
to anthracenes bearing substituents except on the 9- and/or
10-positions, there have been only a few studies on stereochem-
ical DA reactions. One of the most representative experiments
is the DA reaction of 2-substituted anthracenes (1), in which the
substituent was varied from the strongly electron-donating di-
methylamino group to the strongly electron-withdrawing nitro
group, with maleic anhydride, reported by Kaplan and Conroy
(Scheme 1).4 They reported that when the substituent had elec-
tron-donating ability, the syn-cycloadduct 2 became slightly
Chart 1.
dominant. In the case of nitro group, little anti-preference of
3 was observed. They thought that the order of reactivity could
be attributed to the difference in electrostatic effects between
the anthracene and maleic anhydride in the transition state.
The above result prompted us to investigate the behavior of
the anthracenes containing alkoxy groups at the 1- and 4-posi-
tions, as dienes, in order to confirm Conroy’s hypothesis about
the stereoselectivity of DA adducts between 1,4-dialkoxyan-
thracenes and maleic anhydride from the viewpoint of the elec-
trostatic effects and to survey other factors, except for the fore-
going hypothesis. In recent years, as a part of our study about
triptycenes and iptycenes,5 we have been synthesizing some
1,4-dialkoxy anthracenes, which contain strong electron-donat-
ing alkoxy groups and should be better dienes. In this paper, we
describe the results of our experimental investigations on the
Diels–Alder reaction of 1,4-dimethoxy-, 1,4-dipropoxy-, and
1,4-bis(benzyloxy)anthracenes 4a–c (Chart 1) with maleic an-
hydride. We also report the structural characterization of the
DA adducts accomplished by X-ray analysis and 1H NMR
spectroscopy, and theoretical investigations.
Published on the web July 9, 2004; DOI 10.1246/bcsj.77.1385

1386 Bull. Chem. Soc. Jpn., 77, No. 7 (2004)
Characterization of Diels–Alder Adducts
The two diastereomers 8a and 9a in the mixture were distin-
Results and Discussion
1
1
guishable by H NMR spectroscopy. Thus, the H NMR spec-
trum, taken of the crude mixture, distinctly showed the signals
of not only bridgehead methine Ha at ꢀ 5.30 and ꢀ 5.34 (two
broad singlets), but also benzene Hd at ꢀ 7.35 and ꢀ 7.41 (two
double doublets), though chemical shifts of another methine
Hb (at ꢀ 3.47–3.48 as two singlets), benzene Hc (at ꢀ 6.67–
6.69 as two singlets), He (at ꢀ 7.17–7.20 as multiplet), and meth-
yl groups (at ꢀ 3.81–3.82 as two singlets) were almost identical.
Several attempts to separate 8a and 9a by column chromatog-
raphy and recrystallization were unsuccessful. However, we
were able to isolate small amounts of both 8a and 9a by taking
advantage of the difference in their solubility in Et2O. Thus, by
repetition of washing the mixture of 8a and 9a with Et2O, the
residue enriched 8a, while the filtrate enriched 9a (Fig. 1). Fi-
nally, they were purified by recrystallization from toluene.
NOE studies did not lead to the determination of the syn- and
anti-structures because both compounds provide the same cor-
relations such as between Ha and Hb, between Ha and Hd, and
between Hd and He (Fig. 2). Further, the Hb in 8a did not dis-
play an NOE interaction with Hd. However, we succeeded in X-
ray analysis of both 8a and 9a. The X-ray structures established
the stereochemical relationships unambiguously (Fig. 3). Thus,
the molecular structures revealed that the configuration of the
component that was more soluble in Et2O was anti-9a (Fig.
3b),8 although the stereochemistry of another component was
syn-8a (Fig. 3a). In both molecules, methoxy groups in the solid
state lay in the benzene plane. These conformations were sim-
ilar to the minimum energy syn-planar conformation9 of 1,4-di-
The anthracenes 4a and 4b,c were prepared by the modified
methods of Klanderman6 and Lepage,7 respectively (Scheme 2).
Etherification of quinizarin (5) in the presence of K2CO3 yield-
ed anthraquinones 6a–c in 83–97% yields. Reduction of 6a–c
with NaBH4 in MeOH–THF followed by careful neutralization
to pH 7 with acetic acid gave diols 7a–c as transient intermedi-
ates. Treatment of 7a–c with 5–7 M HCl in THF at 40 ꢁC under
air afforded anthracenes 4a–c in 35–40% two-step yields. This
procedure did not need a sequence of isolation of anthrone, re-
duction, and acidification.
The first DA reaction performed was that of 4a. The reaction
of 4a with maleic anhydride (1.5 equiv) in refluxing toluene un-
der an inert atmosphere for 6 h furnished a mixture of two dia-
stereoisomers, the syn-cycloadduct 8a and anti-cycloadduct 9a
in an isolated yield of 92%, accompanied by a complete loss of
4a (Table 1).
Scheme 2. Reagents and conditions: i, (a) TosOMe, K2CO3,
o-dichlorobenzene, reflux, 1 h, 83%; (b) PrBr, K2CO3,
DMF, 100 ꢁC, 6 h, 97%; (c) PhCH2Cl, K2CO3, DMF,
ꢁ
ꢁ
100 C, 6 h, 83%; ii, NaBH4, MeOH–THF, 0 C, 30 min,
ꢁ
quantitative; iii, (a) ꢁ7 M HCl, THF, 40 C, 2 h, 40%; (b)
5 M HCl, THF, 40 C, 4 h, 37%; (c) 5 M HCl, THF, 40
ꢁC, 4 h, 35%.
Table 1. Preparation of Cycloadducts 8 and 9
O
O
Hb
Ha
OR
Hc
O
8a-c
He
Hd
O
OR
O
O
O
Fig. 1. Schematic diagram showing separation of 8a and 9a.
O
4a-c
Hb
OR
toluene, reflux
O
9a-c
He
Ha
Hd
Hc
OR
Substrate
Compound
R
8/9
Yield/%a)
4a
4b
4c
8a + 9a
8b + 9b
8c + 9c
Me
n-Pr
PhCH2
45:55
43:57
57:43
92
87
86
Fig. 2. NOE correlations of 8a and 9a.
a) Isolated yield.

C. Kitamura et al.
Bull. Chem. Soc. Jpn., 77, No. 7 (2004) 1387
Thus, the mean intramolecular distances H OMe for 8a and
aꢂꢂꢂ
ꢀ
9a in the crystals were 2.54 and 2.53 A, respectively, indicating
9a had a slightly stronger van der Waals repulsion. By compar-
ing the integral ratios of Ha, Hc, and Hd protons, 8a/9a ratio of
45:55 was obtained, exhibiting a slight anti-preference. This re-
sult is different from Conroy’s result of the anthracenes with an
electron-donating group at the 2-position (Scheme 1). At this
point we were thinking that the steric effects of methoxy sub-
stituents surpassed the electronic ones. This idea was denied
by later considerations.
We recognized the bridgehead methine Ha signals in 8 and 9
as a useful probe for not only the obvious distinctions between
syn and anti but also the determination of syn/anti selectivity.
Thus, from a 1H NMR spectrum of a mixture of 8 and 9, the two
Ha signals at lower and higher fields can be assigned to syn-8
and anti-9, respectively, and the syn/anti ratio can be obtained
from the integral ratios. In order to confirm this analytical cri-
terion, we tried to apply it to characterize 8b and 9b as well as
8c and 9c.
The DA reaction of 4b,c and maleic anhydride yielded 8b/
9b and 8c/9c mixtures in yields of 87 and 86%, respectively
(Table 1). We could not separate either mixture. This differed
1
from the case of the mixture of 8a and 9a. From the H NMR
spectra of their mixtures, the bridgehead Ha signals in 8b and
9b were assigned to ꢀ 5.35 and ꢀ 5.30, respectively, and those
of 8c and 9c were assigned to ꢀ 5.42 and ꢀ 5.29, respectively
(Table 2). In the case of the mixture of 8b and 9b, we were able
to assign Hc (8b: ꢀ 6.65 and 9b: ꢀ 6.64) and Hd (8b: ꢀ 7.40 and
9b: ꢀ 7.34) protons, by considering the comparison of chemical
shifts of 8a and 9a. We also regarded the chemical shifts of Hd
in 8b and 9b as comparable to those in 8a and 9a (8a: ꢀ 7.41 and
9a: ꢀ 7.35), respectively. In the case of the mixture of 8c and 9c,
we assumed that the signals of ꢀ 3.52 and ꢀ 3.28 were assigned
to the Hb protons of 8c and 9c, respectively, and that the signals
of ꢀ 6.70 and ꢀ 6.74 were assigned to the Hc protons of 8c and
9c, respectively. The finding that the Hb signal in 9c was shifted
higher than that in 9a or 9b would indicate shielding by the ben-
zene ring, which was attached to the substituent in 9c. From the
integral ratio of Ha, we estimated the syn/anti selectivity of 8b/
9b and 8c/9c to be 43:57 and 57:43, respectively, meaning that
the former has a slight anti-preference and the latter has, inter-
estingly, a little syn-preference.
Fig. 3. Molecular structures of (a) 8a and (b) 9a.
methoxybenzene and our X-ray analysis10 of 1,4-dimethoxyan-
thracene. When 8a and 9a were heated in toluene at reflux for a
longer period, such as more than 24 h, no isomerization was ob-
served. This showed that the DA adducts 8a and 9a were ther-
modynamically stable and that at the temperature applied no
retro-DA reaction occurred.
On the basis of the X-ray analysis, the characterization of the
syn- and anti-adducts was confirmed by 1H spectroscopy
(Table 2). Thus, the signals of ꢀ 5.30 and ꢀ 5.34 were assigned
to the methine Ha of anti-9a and syn-8a, respectively, and sim-
ilarly the signals of benzene Hd of anti-9a and syn-8a were as-
signed to ꢀ 7.35 and ꢀ 7.41, respectively. The difference in the
chemical shifts of Hd (between ꢀ 7.35 and ꢀ 7.41) can be attrib-
1
uted to the H NMR anisotropy effect arising from different
spatial arrangements of the carbonyl oxygen atom in the mole-
cules. On the other hand, the difference in the chemical shifts of
Ha (between ꢀ 5.30 and ꢀ 5.34) can be explained by the van der
Waals interaction between Ha and the oxygen atom in MeO.
Because we failed in separating the mixtures of 8b and 9b as
well as 8c and 9c, we tried to transform them into the molecules
that could be easily separated in order to establish characteriza-
tion. We succeeded in isolating their reduction products. Thus,
Table 2. 500 MHz 1H NMR Spectral Data (ꢀ in ppm) for 8a–c, 9a–c, 10b,c, and 11b,c in CDCl3
8aa)
9aa)
8bb)
9bb)
8cb)
9cb)
10ba)
11ba)
10ca)
11ca)
Ha 5.34 brs
Hb 3.48 brs
Hc 6.69 s
5.30 brs
3.47 brs
6.67 s
5.35 brs
5.30 brs
5.42 brs 5.29 brs 4.75 brs
4.63 brs
4.80 brs
4.66 brs
3.52 brs 3.28 brs 2.33–2.35 m 2.34–2.36 m 2.33–2.36 m 2.25–2.27 m
3.47–3.49 m
6.65 s
7.40 dd
6.64 s
7.34 dd
6.70 s
NAc)
6.74 s
NAc)
6.58 s
7.31 dd
(3.2, 5.2)d)
7.12 dd
6.58 s
7.23 dd
6.65 s
7.31 dd
6.65 s
7.23 dd
Hd 7.41 dd
7.35 dd
(3.2, 5.2)d) (3.2, 5.2)d) (3.2, 5.2)d) (3.2, 5.2)d)
(3.2, 5.2)d)
7.08 dd
(3.2, 5.2)d)
7.12 dd
(3.2, 5.2)d)
7.09 dd
He 7.18 dd
7.19 dd
7.16–7.20 m
7.16–7.20 m
(3.2, 5.4)d) (3.2, 5.4)d)
(3.2, 5.4)d)
(3.2, 5.4)d)
(3.2, 5.4)d)
(3.2, 5.4)d)
a) Isolated compounds. b) Assigned components in the mixture. c) NA = not assigned because of overlapping of benzyl protons.
d) J in Hz.

1388 Bull. Chem. Soc. Jpn., 77, No. 7 (2004)
Table 3. Preparation of Diols 10 and 11
Characterization of Diels–Alder Adducts
HO
HO
O
O
Hb
Ha
OR
OR
Hc
O
10b,c
11b,c
He
8b,c
Hd
OR
OR
LiAlH4
THF, reflux
OR
O
HO
HO
O
O
Fig. 5. NOE correlations of 10b,c and 11b,c.
Hb
OR
Hc
did not furnish valuable information about the identification of
syn and anti on account of the same NOE correlations because
neither NOE correlation between Hb and Hd nor between meth-
ylene groups and Hd was observed, as shown in Fig. 5. As for
10c and 11c, the assignment of proton signals were made on the
basis of chemical shifts of 10b and 11b (Table 2). Thus, we no-
ticed the identity of the chemical shifts of Hd (10b: ꢀ 7.31, 10c:
ꢀ 7.31, 11b: ꢀ 7.23, and 11c: ꢀ 7.23) and He (10b: ꢀ 7.12, 10c:
ꢀ 7.12, 11b: ꢀ 7.08, and 11c: ꢀ 7.09) protons, which were not
geographically influenced by the substituent groups at the 1-
and 4-postions. Further, 10c defined above had a higher Rf val-
ue on silica gel TLC developed by (2:1) CHCl3–AcOEt than
11c, and this propensity is consistent with the observation that
syn-10b had a higher Rf value compared with anti-11b. There-
fore, we judged our characterization of 10c and 11c valid. The
bridgehead methine Ha signals of 10c and 11c were assigned to
ꢀ 4.80 and ꢀ 4.66, respectively. From the above findings, we
concluded that the assignment method for determining syn-
and anti-structure by 1H spectroscopy was reasonable. This will
be instructive to characterization of other DA adducts between
1,4-dialkoxyanthracenes and maleic anhydride.
He
Ha
9b,c
Hd
OR
OR
Substrate
Compound
R
10/11
Yield/%a)
8b + 9b
8c + 9c
10b + 11b
10c + 11c
n-Pr
PhCH2
43:57
57:43
96
91
a) Crude yield.
Fig. 4. Molecular structure of 10b.
We found that DA reaction of 1,4-dialkyloxyanthracene 4a,b
and maleic anhydride afforded a small anti-preference, on the
other hand, DA reaction of 1,4-bis(benzyloxy)anthracene 4c
with maleic anhydride resulted in a small syn-preference. It
seems unreasonable to assume that steric effects of the substitu-
ent groups play a part in direct determination of syn/anti selec-
tivity. We thought that the substituent groups in anthracenes
4a–c were located in the place where the DA reaction was
not prevented, therefore there was hardly any steric repulsion
between the substituent groups and maleic anhydride. Then,
we carried out a conformational search for 4b and 4c using
the molecular mechanics mode and examined energies for each
of the different conformers. The lowest-energy conformers for
4b and 4c are shown in Fig. 6. It can be seen that, although 4b
has a planar conformation, in the case of 4c, two methylene
parts of the benzyl groups are not only anti but almost perpen-
dicular to anthracene, and the phenyl rings are apart from
the anthracene. The special conformation of 4c may relate to
the different result in the syn/anti selectivity. In addition,
if Conroy’s interpretation on the electrostatic interactions
(Scheme 1) is right, we had to gain the results of all syn-prefer-
ence. These situations suggest that we have to consider other
factors to understand the syn/anti selectivity.
diols 10b and 11b were prepared by reduction of a mixture of
8b and 9b with LiAlH4 in THF for 4 h in 96% yield, while 10c
and 11c were obtained from a mixture of 8c and 9c in 91% yield
(Table 3). The ratios of 10b/11b and 10c/11c, which were de-
termined from the integral ratios of Ha in the state of crude mix-
tures, were 43:57 (Ha: ꢀ 4.75 vs ꢀ 4.63, Table 2) and 57:43 (Ha:
ꢀ 4.80 vs ꢀ 4.66), respectively. These values were, of course,
the same as those of 8b/9b and 8c/9c, respectively. In the case
of 10b/11b, the ratio was supported by the integral ratio of Hd
protons. Purification by column chromatography of the former
mixture afforded 10b and 11b independently, and purification
of the latter mixture gave 10c and 11c separately. We succeed-
ed in X-ray analysis of one component, with a higher Rf value
on silica gel TLC developed by (2:1) CHCl3–AcOEt, between
10b and 11b that had the Ha signals of ꢀ 4.75. The molecule
proved to be 10b by displaying syn-configuration (Fig. 4). It
was also observed that except for the terminal methyl group
of the propoxy group, two methylene groups lay in the benzene
plane. Since the stereochemistry of one component with a high-
er Rf value turned out to be 10b, another component with a low-
er Rf value was defined as 11b unequivocally. Therefore as-
signment of 1H NMR data for both diols was established
(Table 2). Even in the case of the reduction products, it was
confirmed that the bridgehead Ha signals at lower and higher
fields can be assigned to the syn- and anti-configurations, re-
spectively, in analogy with the case of 8b and 9b. NOE studies
In order to analyze the origin of the syn/anti selectivity of
the DA adducts, we carried out a computational evaluation of
Frontier orbitals (FOs) of 4a and maleic anhydride, and the
transition structures (TSs) of 8a and 9a using B3LYP/6-

C. Kitamura et al.
Bull. Chem. Soc. Jpn., 77, No. 7 (2004) 1389
Fig. 6. Top and side views of the lowest-energy conformers for (a) 4b and (b) 4c, examined by the molecular mechanics conforma-
tion search.
Fig. 8. The HOMO–LUMO interactions between 4a and
maleic anhydride in (a) syn-orientation and (b) anti-orien-
tation. Bold dashes specify the primary interactions while
dashed lines refer to the SOIs.
trated in Fig. 8. Thus, the AO coefficients at C4a, C9a, C8a, and
C10a of the HOMO of 4a are so small that the secondary orbital
interactions11 (SOIs) of C4a/C9a (HOMO) C2/C5 (LUMO)
in the syn-orientation should be comparable to that of C8a/
ꢂꢂꢂ
C10a (HOMO) C2/C5 (LUMO) in the anti-orientation. We
ꢂꢂꢂ
have optimized the TSs as shown in Fig. 9, which were charac-
terized by a single negative frequency. The degree of bond for-
mation was synchronous and the lengths of the incipient bonds
ꢀ
ꢀ
were essential identical (2.203 A in syn-orientation and 2.207 A
in anti-orientation). The degree of bending (e.g., dihedral an-
gle) was similar in magnitude in both the syn- and anti-orienta-
tion. Further, our calculations predicted that the relative differ-
ence in activation enthalpies of two possible TSs was 0.06
kcal molꢃ1 (anti-mode was faintly higher), indicating their val-
ues are substantially identical. This result means that the sum of
contributions from steric effects, SOIs, electrostatic interac-
tions, and other interactions, which are all associated with both
TSs, are the same in the syn- and anti-orientation. Then, in or-
der to evaluate Conroy’s result at the same computational level,
we have calculated the TSs of formation of the syn- and anti-
adducts derived from 2-nitoroanthracene and 2-dimethylamino-
anthracene, respectively. Our calculation indicated that the
anti-TS for 2-nitroanthracene was 0.31 kcal molꢃ1 more stable
Fig. 7. The FOs of 4a and maleic anhydride. Numbers near
levels are the B3LYP orbital energies in eV and numbers
near the atomic orbitals (AOs) are STO3G== B3LYP co-
efficients. The AO coefficients are given in units of 10ꢃ2
.
31G(d) calculations. The FOs are depicted in Fig. 7 and show
clearly that the interaction between the HOMO of 4a and the
LUMO of maleic anhydride is primary and that the largest
atomic orbital (AO) coefficients of 4a and maleic anhydride
are the 9,10-positions and the 3,4-positions, respectively. The
FOs of 4a and maleic anhydride could not explain the remark-
able difference between the syn- and anti-orientations, as illus-

1390 Bull. Chem. Soc. Jpn., 77, No. 7 (2004)
Characterization of Diels–Alder Adducts
Fig. 10. Putative reaction mechanism.
ctivity.
In conclusion, we found that the bridgehead proton of cyclo-
adducts was an effective probe for investigating the stereo-
chemistry of the DA reaction between 1,4-dialkoxyanthracenes
and maleic anhydride, which was established by the isolation of
diastereomers, 1H NMR spectroscopy, and X-ray analysis. The
stereochemical behavior of the products showed that the syn/
anti ratios were close to 1:1 but the values fluctuated according
to the substituent groups. In the case of methoxy and propoxy
derivatives, a slight anti-preference was observed. However,
the DA reaction of 1,4-bis(benzyloxy)anthracene resulted in a
small syn-preference. From computational studies, the differ-
ence in reaction courses in the syn- and anti-orientation was
not detected. The formation of pre-reactive CT complexes
was observed, and it was considered that the syn- and anti-ar-
rangement of 1,4-dialkoxyanthracenes and maleic anhydride
in the CT complexes might determine the syn/anti ratios of
the products.
Fig. 9. Calculated transition structures associated with the
formation of (a) syn-8a and (b) anti-9a, together with se-
lected geometrical parameters.
than the syn-TS, and that the anti-TS for 2-dimethylaminoan-
thracene was 0.57 kcal molꢃ1 higher than the syn-TS. These en-
ergetic differences correspond to an anti-preference for 2-nitro-
anthracene and a syn-preference for 2-dimethylaminoanthra-
cene, and are in good agreement with Conroy’s experimental
result that reflects the contribution of the electrostatic interac-
tion. We therefore should consider another factor that influen-
ces the stereochemical course of the DA reactions of 1,4-di-
alkoxyanthracenes as well as is not involved in TS.
In regard to the DA reaction of anthracene with an electron-
deficient dienophile, a mechanism via the formation of a charge
transfer (CT) complex has been proposed.12 The importance of
the CT complex formation in the [4 þ 2] cycloaddition has
been also documented.13,14 In our case, a change in solution
color with reaction time was observed. The color was initially
reddish-orange and gradually turned colorless as the reaction
proceeded, suggesting the formation and disappearance of the
CT complex. Treatment of 4a–c with maleic anhydride in
chloroform at room temperature led to the rapid formation of
CT complexes, which could not be isolated, with new UV–
vis absorption bands at 450–500 nm. The above results clearly
indicate that the DA reactions of 1,4-dialkoxyanthracenes with
maleic anhydride proceed via the formation of CT complexes.
Furthermore, judging by the consideration of the CT complex
and the above result of the theoretical calculations, the origin
of the syn/anti selectivity of the products could be the syn/anti
ratios of orientation of 4a–c and maleic anhydride in the CT
Experimental
General. THF and DMF were distilled from LiAlH4 and CaH2,
respectively, prior to use. Commercially available reagents were
used as supplied unless otherwise stated. All reactions were carried
out under a nitrogen atmosphere unless otherwise noted. Analytical
thin-layer chromatography (TLC) was performed on Merck silica
gel 60 F254 0.25 mm aluminium plates, and components were vi-
sualized by UV light or by iodine vapor. Column chromatography
was performed on Wako silica gel C-300 (45–75 mm, 300 mesh).
Mp’s were determined on a Yanaco Melting Point apparatus and
are uncorrected. 1H and 13C NMR spectra were recorded on a
Bruker DRX500 FT spectrometer at 500 and 126 MHz, respective-
ly. Chemical shifts were referenced to TMS. IR spectra were re-
corded on a Shimadzu FTIR-8400 spectrometer as KBr pressed
pellets. Electron impact mass spectra were obtained at 70 eV on
a Shimadzu QP-1000EX. Elemental analysis was carried out on
a Yanaco MT-5 CHN corder.
´
complexes. Recently, Suarez and Sordo have reported that
General Procedure for the Preparation of 1,4-Dialkoxyan-
thracene 4a–c. A mixture of quinizarin (10.0 g, 41.6 mmol),
K2CO3 (17.3 g, 125 mmol), and an excess (5 equiv) of electrophile
was heated at reflux for 1 h in o-dichlorobenzene (70 mL) for prep-
aration of 6a or at 100 ꢁC for 6 h in dry DMF (30 mL) for prepa-
ration of 6b,c. After conventional work-up and purification by re-
crystallization, anthraquinones 6a–c were treated with NaBH4 (5
equiv) in (2:1) MeOH–THF at 0 ꢁC for 30 min. The reaction mix-
tures were neutralized with acetic acid, then extracted with AcOEt,
washed with brine, and dried over Na2SO4. After evaporation of
the solvent, 9,10-dihydroxy-9,10-dihydroanthracenes 7a–c were
the pre-reactive van der Waals complexes may play a decisive
role in determining the stereochemical outcome of DA reac-
tions,15 indicating the importance of the stereochemical compo-
sition in pre-reactive molecular complexes. Therefore, we be-
lieve the DA reaction of 1,4-dialkoxyanthracenes and maleic
anhydride proceeds by not a direct pathway (Path b, Fig. 9)
but a two-step route via a CT complex (Path a, Fig. 10). We
are currently examining the effects of varying the substituent
group of anthracene in order to gain a clearer understanding
of the relation between the CT complex and syn/anti sele-

C. Kitamura et al.
Bull. Chem. Soc. Jpn., 77, No. 7 (2004) 1391
obtained as brown oils in quantitative yields. Without purification,
7a–c were dissolved in THF (100 mL) and 7 M (for 7a) or 5 M (for
7b,c) HCl (50 mL) was added. The mixtures were then stirred at 40
ꢁC for 2 h (for 7a) or 4 h (for 7b,c) under air. After neutralization
with 5 M NaOH, the solutions were extracted with CHCl3. The ex-
tracts were washed with brine and dried over Na2SO4. Purification
by column chromatography (CHCl3–haxane) gave 4a–c as yellow
solids in yields of 40% from 7a, 37% from 7b, and 35% from 7c.
1,4-Dimethoxyanthracene (4a).16 Mp 134–136 ꢁC (lit.4 134–
136 ꢁC); 1H NMR (500 MHz, CDCl3) ꢀ 4.03 (s, 6H, 2OCH3), 6.61
(s, 2H, 2-H, 3-H), 7.48 (dd, J ¼ 3:1, 6.4 Hz, 2H, 6-H, 7-H), 8.04
(dd, J ¼ 3:1, 6.4 Hz, 2H, 5-H, 8-H), 8.77 (s, 2H, 9-H, 10-H).
1,4-Dipropoxyanthracene (4b). Mp 66–67 ꢁC; IR (KBr) 2968,
2936, 1622, 1578, 1456 cmꢃ1; 1H NMR (500 MHz, CDCl3) ꢀ 1.18
(t, J ¼ 7:4 Hz, 6H, 2CH2CH3), 1.98–2.05 (m, 4H, 2CH2CH2CH3),
4.12 (t, J ¼ 6:6 Hz, 4H, 2OCH2CH2), 6.59 (s, 2H, 2-H, 3-H), 7.47
(dd, J ¼ 3:2, 6.4 Hz, 2H, 6-H, 7-H), 8.05 (dd, J ¼ 3:2, 6.4 Hz, 2H,
5-H, 8-H), 8.80 (s, 2H, 9-H, 10-H); 13C NMR (126 MHz, CDCl3)
ꢀ 10.92, 22.78, 69.84, 101.90, 120.72, 125.33, 125.75, 128.55,
131.36, 148.68; MS (EI) m=z (relative intensity) 294 (Mþ, 93),
209 (100). Anal. Calcd for C20H22O2: C, 81.60; H, 7.53%. Found:
C, 81.33; H, 7.41%.
7.34 (dd, J ¼ 3:2, 5.2 Hz, 2H, 5-H, 8-H, 9b), 7.40 (dd, J ¼ 3:2,
5.2 Hz, 2H, 5-H, 8-H, 8b); MS (EI) m=z (relative intensity) 209
(80), 294 (100), 392 (Mþ, 57). Anal. Calcd for C24H24O5: C,
73.45; H, 6.16%. Found: C, 73.37; H, 6.26%.
ꢀ
ꢀ
(11R ,15S )-1,4-Bis(benzyloxy)-9,10,11,15-tetrahydro-9,10-
[30,40]furanoanthracene-12,14-diones (8c and 9c). 8c/9c =
57:43. Purified by column chromatography (CHCl3–hexane–
AcOEt, 10:5:1) and recrystallization from CHCl3 to give a mixture
of 8c and 9c as a white solid (418 mg, 0.86 mmol, 86%), mp 204–
206 ꢁC. The diastereoisomers were inseparable by column chroma-
tography and were characterized as a mixture. IR (KBr) 3036,
1867, 1786, 1498, 1267, 1076, 932 cmꢃ1 1H NMR (500 MHz,
;
CDCl3) ꢀ 3.28 (brs, 2H, 2CHCH(C=O), 9c), 3.52 (brs, 2H,
2CHCH(C=O), 8c), 5.00–5.05 (m, 4H, 2OCH2Ph, 8c and 9c),
5.29 (brs, 2H, 2CHCH(C=O), 9c), 5.42 (brs, 2H, 2CHCH(C=O),
8c), 6.70 (s, 2H, 2-H, 3-H, 8c), 6.74 (s, 2H, 2-H, 3-H, 9c), 7.16–
7.20 (m, 2H, 6-H, 7-H, 8c and 9c), 7.37–7.50 (m, 12H, 5-H, 8-H,
10PhH, 8c and 9c); MS (EI) m=z (relative intensity) 91 (100),
488 (Mþ, 26). Anal. Calcd for C32H24O5: C, 78.67; H, 4.95%.
Found: C, 78.36; H, 5.30%.
Separation of 8a and 9a with the Difference in Their Solubil-
ity in Et2O. A mixture of 8a and 9a (ca. 300 mg) was washed with
one portion of Et2O (ca. 30 mL). The residue was further washed
with several portions of Et2O until 9a was not observed by
1H NMR spectroscopy. Finally, 8a was purified by recrystalliza-
tion from toluene. The first filtrate was evaporated to dryness,
and the residue was washed with one portion of Et2O and then also
collected and evaporated. Until the 8a/9a ratio was constant, the
washing/evaporation technique was repeated. Finally, recrystalli-
zation from toluene produced pure 8a and 9a, respectively. Follow-
ing the above procedure, 10–20 mg of 8a and 9a were obtained.
(11R,15S)-1,4-Dimethoxy-9,10,11,15-tetrahydro-9,10[30,40]-
1,4-Bis(benzyloxy)anthracene (4c). Mp 153–158 ꢁC (lit.5
149–150 ꢁC); 1H NMR (500 MHz, CDCl3) ꢀ 5.28 (s, 4H,
2OCH2Ph), 6.67 (s, 2H, 2-H, 3-H), 7.37–7.40 (m, 2H, 2PhH),
7.44–7.59 (m, 10H, 6-H, 7-H, 8PhH), 8.04 (dd, J ¼ 3:2, 6.4 Hz,
2H, 5-H, 8-H), 8.86 (s, 2H, 9-H, 10-H).
General Procedure for Diels–Alder Reaction of 1,4-Dialkoxy-
anthracene 4 with Maleic Anhydride. A mixture of 4a–c (1.0
mmol) and maleic anhydride (1.5 mmol) in toluene (3 mL) was
heated at reflux for 6 h. After evaporation of the solvent and drying
under vacuum, the residue was well ground and observed with
1H NMR to determine diastereoselectivity.
ꢁ
furanoanthracene-12,14-diones 8a. Mp 268–270 C; IR (KBr)
2949, 1778, 1495, 1261, 1078, 928 cmꢃ1 1H NMR (500 MHz,
ꢀ
ꢀ
(11R ,15S )-1,4-Dimethoxy-9,10,11,15-tetrahydro-9,10[30,40]-
furanoanthracene-12,14-diones (8a and 9a). 8a/9a = 45:55.
Purified by recrystallization from toluene to give a mixture of 8a
and 9a as a white solid (309 mg, 0.92 mmol, 92%), mp 230–232
ꢁC. The diastereoisomers were inseparable by column chromatog-
raphy and were characterized as a mixture. IR (KBr) 2949, 1782,
1497, 1261, 1078, 928 cmꢃ1; 1H NMR (500 MHz, CDCl3) ꢀ 3.47
(brs, 2H, 2CHCH(C=O), 9a), 3.48 (brs, 2H, 2CHCH(C=O), 8a),
3.81 (s, 6H, 2OCH3, 8a), 3.82 (s, 6H, 2OCH3, 9a), 5.30 (brs,
2H, 2CHCH(C=O), 9a), 5.34 (brs, 2H, 2CHCH(C=O), 8a), 6.67
(s, 2H, 2-H, 3-H, 9a), 6.69 (s, 2H, 2-H, 3-H, 8a), 7.16–7.20 (m,
2H, 6-H, 7-H, 8a and 9a), 7.35 (dd, J ¼ 3:2, 5.2 Hz, 2H, 5-H, 8-
H, 9a), 7.41 (dd, J ¼ 3:2, 5.2 Hz, 2H, 5-H, 8-H, 8a); MS (EI)
m=z (relative intensity) 223 (95), 238 (100), 336 (Mþ, 37). Anal.
Calcd for C20H16O5: C, 71.42; H, 4.79%. Found: C, 71.78; H,
4.89%.
;
CDCl3) ꢀ 3.48 (brs, 2H, 2CHCH(C=O)), 3.81 (s, 6H, 2OCH3),
5.34 (brs, 2H, 2CHCH(C=O)), 6.69 (s, 2H, 2-H, 3-H), 7.18 (dd,
J ¼ 3:2, 5.4 Hz, 2H, 6-H, 7-H), 7.41 (dd, J ¼ 3:2, 5.2 Hz, 2H,
5-H, 8-H); 13C NMR (126 MHz, CDCl3) ꢀ 38.54, 47.65, 56.38,
110.71, 124.53, 126.92, 127.88, 140.92, 149.12, 170.46; MS (EI)
m=z (relative intensity) 223 (94), 238 (100), 336 (Mþ, 37). Anal.
Calcd for C20H16O5: C, 71.42; H, 4.79%. Found: C, 71.44; H,
5.11%.
(11S,15R)-1,4-Dimethoxy-9,10,11,15-tetrahydro-9,10[30,40]-
ꢁ
furanoanthracene-12,14-diones 9a. Mp 252–254 C; IR (KBr)
2949, 1786, 1499, 1259, 1082, 930 cmꢃ1 1H NMR (500 MHz,
;
CDCl3) ꢀ 3.47 (brs, 2H, 2CHCH(C=O)), 3.82 (s, 6H, 2OCH3),
5.30 (brs, 2H, 2CHCH(C=O)), 6.67 (s, 2H, C2,3-H), 7.19 (dd,
J ¼ 3:2, 5.4 Hz, 2H, 6-H, 7-H), 7.35 (dd, J ¼ 3:2, 5.2 Hz, 2H,
5-H, 8-H); 13C NMR (126 MHz, CDCl3) ꢀ 38.66, 47.46, 55.99,
109.58, 125.31, 127.51, 130.16, 138.37, 148.71, 170.60; MS (EI)
m=z (relative intensity) 223 (93), 238 (100), 336 (Mþ, 31). Anal.
Calcd for C20H16O5: C, 71.42; H, 4.79%. Found: C, 71.46; H,
5.11%.
General Procedure for the Reduction of a Mixture of Cyclo-
adducts 8 and 9. To a suspension of LiAlH4 in dry THF, a solu-
tion of a mixture of 8b and 9b, or 8c and 9c in dry THF were added
dropwise at r.t. The mixture was stirred at reflux for 4 h, and then
cooled to 0 ꢁC. Water and conc. HCl were then added until the re-
sulting precipitate disappeared. The products were extracted with
Et2O, and the organic phase was washed with brine and dried over
Na2SO4. Removal of the solvent and drying under vacuum gave a
crude mixture of diols 10b and 11b, or 10c and 11c. The products
ꢀ
ꢀ
(11R ,15S )-1,4-Dipropoxy-9,10,11,15-tetrahydro-9,10[30,40]-
furanoanthracene-12,14-diones (8b and 9b). 8b/9b = 43:57.
Purified by column chromatography (CHCl3–hexane–AcOEt,
10:5:1) to give a mixture of 8b and 9b as a white solid (341 mg,
0.87 mmol, 87%), mp 184–186 ꢁC. The diastereoisomers were in-
separable by column chromatography and were characterized as a
mixture. IR (KBr) 2964, 1782, 1497, 1265, 1076, 934 cmꢃ1
;
1H NMR (500 MHz, CDCl3) ꢀ 1.06–1.10 (m, 6H, 2CH2CH3, 8b
and 9b), 1.81–1.87 (m, 4H, 2CH2CH2CH3, 8b and 9b), 3.47–
3.49 (m, 2H, 2CHCH(C=O), 8b and 9b), 3.82–3.96 (m, 4H,
2OCH2CH2, 8b and 9b), 5.30 (brs, 2H, 2CHCH(C=O), 9b), 5.35
(brs, 2H, 2CHCH(C=O), 8b), 6.64 (s, 2H, 2-H, 3-H, 9b), 6.65
(s, 2H, 2-H, 3-H, 8b), 7.16–7.20 (m, 2H, 6-H, 7-H, 8b and 9b),

1392 Bull. Chem. Soc. Jpn., 77, No. 7 (2004)
Characterization of Diels–Alder Adducts
1
were subjected to H NMR measurement to determine the 10/11
2CHCHCH2OH), 5.01–5.06 (m, 4H, 2OCH2Ph), 6.65 (s, 2H, 2-
H, 3-H), 7.09 (dd, J ¼ 3:2, 5.4 Hz, 2H, 6-H, 7-H), 7.23 (dd,
J ¼ 3:2, 5.2 Hz, 2H, 5-H, 8-H), 7.34–7.44 (m, 10H, 10PhH);
13C NMR (126 MHz, CDCl3) ꢀ 41.25, 42.84, 64.86, 71.48,
111.00, 124.75, 125.72, 127.48, 127.87, 128.55, 133.89, 137.49,
140.93, 148.74; MS (EI) m=z (relative intensity) 91 (100), 478
(Mþ, 26). Anal. Calcd for C32H30O4: C, 80.31; H, 6.32%. Found:
C, 80.02; H, 6.25%.
Crystal Structure Determinations. X-ray data were collected
on a Rigaku/MSC MERCURY CCD. The structures were solved
by direct methods17 and expanded using the Fourier technique.18
All calculations were performed using the teXsan program pack-
ages.19 Full crystallographic details, excluding structure factors,
for the structures of compounds 8a (CCDC 207475), 9a (CCDC
207474), and 10b (CCDC 212776) have been deposited with the
Cambridge Crystallographic Data Centre. Copies of the data can
be obtained, free of charge, on application to CCDC, 12 Union
Road, Cambridge, CB2 1EZ, UK [fax: +44(0)-1223-336033 or
e-mail: deposit@ccdc.ca.ac.uk].
ratio, affording 10b/11b = 43:57 and 10c/11c = 57:43. The prod-
ucts were separated and purified by column chromatography
(CHCl3–AcOEt, 2:1).
(11R,12S)- and (11S,12R)-9,10-Ethano-11,12-bis(hydroxy-
methyl)-1,4-dipropoxy-9,10-dihydroanthracenes (10b and
11b). The mixture was obtained in a crude yield of 96%. After
multiple column chromatography runs, purified diols 10b and
11b were obtained in isolated yields of 33 and 43%, respectively,
using the general procedure with LiAlH4 (794 mg, 21.0 mmol) in
THF (50 mL) and a mixture of 8b and 9b (1.59 g, 4.06 mmol) in
dry THF (45 mL). 10b: a white solid, mp 142–144 ꢁC, Rf ¼
0:45 (CHCl3–AcOEt, 2:1); IR (KBr) 3275, 2964, 1497, 1259,
1
1076, 756 cmꢃ1; H NMR (500 MHz, CDCl3) ꢀ 1.07 (t, J ¼ 7:3
Hz, 6H, 2CH2CH3), 1.78–1.85 (m, 4H, 2CH2CH2CH3), 2.33–
2.35 (m, 2H, 2CHCHCH2OH), 2.58 (brs, 2H, 2CH2OH), 3.35–
3.42 (m, 4H, 2CHCH2OH), 3.84–3.93 (m, 4H, 2OCH2CH2), 4.75
(brs, 2H, 2CHCHCH2OH), 6.58 (s, 2H, 2-H, 3-H), 7.12 (dd,
J ¼ 3:2, 5.4 Hz, 2H, 6-H, 7-H), 7.31 (dd, J ¼ 3:2, 5.2 Hz, 2H,
5-H, 8-H); 13C NMR (126 MHz, CDCl3) ꢀ 10.72, 22.78, 40.47,
43.51, 63.70, 70.60, 109.79, 123.50, 125.67, 130.75, 143.45,
148.93; MS (EI) m=z (relative intensity) 294 (100), 382 (Mþ,
35). Anal. Calcd for C24H30O4: C, 75.36; H, 7.91ꢁ%. Found: C,
75.36; H, 8.09%. 11b: a white solid, mp 155–157 C, Rf ¼ 0:38
(CHCl3–AcOEt, 2:1); IR (KBr) 3327, 2966, 1495, 1257, 1070,
Crystal Data for 8a. C20H16O5, M ¼ 336:34, triclinic, space
ꢁ
ꢀ
ꢀ 3
group P1 (#2), a ¼ 7:834ð6Þ, b ¼ 9:496ð7Þ, c ¼ 11:694ð9Þ A,
ꢁ
ꢁ ¼ 105:507ð7Þ, ꢂ ¼ 98:392ð7Þ, ꢃ ¼ 105:044 , V ¼ 787ð1Þ A ,
Z ¼ 2, Dcalcd ¼ 1:418 g cmꢃ3, ꢄ(Mo Kꢁ) = 1.02 cmꢃ1, T ¼ 223
K; 29480 reflections collected, 3155 unique with I > 2ꢅðIÞ
(Rint ¼ 0:018), R ¼ 0:051, Rw ¼ 0:079.
1011 cmꢃ1
;
1H NMR (500 MHz, CDCl3) ꢀ 1.06 (t, J ¼ 7:3 Hz,
Crystal Data for 9a. 2(C20H16O5) 0.5toluene, C43:5H36O10,
ꢂ
ꢁ
6H, 2CH2CH3), 1.78–1.85 (m, 4H, 2CH2CH2CH3), 2.34–2.36
(m, 2H, 2CHCHCH2OH), 2.72 (brs, 2H, 2CH2OH), 3.28 (dd,
J ¼ 9:6, 11.6 Hz, 2H, 2CHCHAHBOH), 3.65 (dd, J ¼ 2:9, 11.6
Hz, 2H, 2CHCHAHBOH), 3.82–3.94 (m, 4H, 2OCH2CH2), 4.63
(brs, 2H, 2CHCHCH2OH), 6.58 (s, 2H, 2-H, 3-H), 7.08 (dd,
J ¼ 3:2, 5.4 Hz, 2H, 6-H, 7-H), 7.23 (dd, J ¼ 3:2, 5.2 Hz, 2H,
5-H, 8-H); 13C NMR (126 MHz, CDCl3) ꢀ 10.67, 22.76, 41.15,
42.95, 64.99, 71.30, 110.51, 124.69, 125.66, 133.54, 141.16,
147.76; MS (EI) m=z (relative intensity) 294 (100), 392 (Mþ,
32). Anal. Calcd for C24H30O4: C, 75.36; H, 7.91%. Found: C,
75.07; H, 7.82%.
M ¼ 778:96, triclinic, space group P1 (#2), a ¼ 11:208ð6Þ,
ꢀ
b ¼ 11:681ð6Þ, c ¼ 13:466ð7Þ A, ꢁ ¼ 90:137ð6Þ, ꢂ ¼ 101:857ð7Þ,
ꢃ ¼ 94:414ð5Þ , V ¼ 1719ð1Þ A , Z ¼ 4, Dcalcd ¼ 1:219 g cmꢃ3
,
ꢁ
ꢀ 3
ꢄ(Mo Kꢁ) = 1.05 cmꢃ1, T ¼ 223 K; 69320 reflections collected,
7707 unique with I > 2ꢅðIÞ (Rint ¼ 0:029), R ¼ 0:063, Rw
0:084.
¼
Crystal Data for 10b. C24H30O4, M ¼ 382:50, triclinic, space
ꢁ
ꢀ
ꢀ 3
group P1 (#2), a ¼ 9:716ð2Þ, b ¼ 10:742ð2Þ, c ¼ 11:137ð1Þ A,
ꢁ
ꢁ ¼ 73:02ð1Þ, ꢂ ¼ 70:40ð1Þ, ꢃ ¼ 87:85ð2Þ , V ¼ 1045:1ð4Þ A ,
Z ¼ 2, Dcalcd ¼ 1:215 g cmꢃ3, ꢄ(Mo Kꢁ) = 0.81 cmꢃ1, T ¼ 296
K; 11616 reflections collected, 4638 unique with I > 2ꢅðIÞ
(Rint ¼ 0:076), R ¼ 0:096, Rw ¼ 0:114.
(11R,12S)- and (11S,12R)-1,4-Bis(benzyloxy)-9,10-ethano-
11,12-bis(hydroxymethyl)-9,10-dihydroanthracenes (10c and
11c). The mixture was obtained in a crude yield of 91%. After
multiple column chromatography runs, purified diols 10c and
11c were obtained in isolated yields of 36 and 28%, respectively,
using the general procedure with LiAlH4 (133 mg, 3.50 mmol)
in THF (10 mL) and a mixture of 8c and 9c (379 mg, 0.78 mmol)
in dry THF (15 mL). 10c: a white solid, mp 203–205 ꢁC, Rf ¼ 0:42
(CHCl3–AcOEt, 2:1); IR (KBr) 3265, 2961, 1499, 1263, 1034,
We thank the Research Center for Molecular-scale Nano-
science, the Institute for Molecular Science, for assistance in
obtaining X-ray data. We also thank Prof. Yamana, Himeji In-
stitute of Technology, for determination of UV–vis spectra. We
are grateful to the Himeji Institute of Technology for financial
support of this research.
752 cmꢃ1 1H NMR (500 MHz, CDCl3) ꢀ 2.33–2.36 (m, 2H,
;
References
2CHCHCH2OH), 2.51 (brs, 2H, 2CH2OH), 3.39 (dd, J ¼ 7:7,
11.2 Hz, 2H, 2CHCHAHBOH), 3.47 (dd, J ¼ 5:4, 11.2 Hz, 2H,
2CHCHAHBOH), 4.80 (brs, 2H, 2CHCHCH2OH), 5.00–5.06 (m,
4H, 2OCH2Ph), 6.65 (s, 2H, 2-H, 3-H), 7.12 (dd, J ¼ 3:2, 5.4
Hz, 2H, 6-H, 7-H), 7.31 (dd, J ¼ 3:2, 5.2 Hz, 2H, 5-H, 8-H),
7.35–7.45 (m, 10H, 10PhH); 13C NMR (126 MHz, CDCl3)
ꢀ 40.50, 43.60, 63.70, 71.02, 110.16, 123.61, 125.78, 127.31,
127.93, 128.63, 131.19, 137.36, 143.26, 148.98; MS (EI) m=z (rel-
ative intensity) 91 (100), 478 (Mþ, 28). Anal. Calcd for C32H30O4:
C, 80.31; H, 6.32%. Found: C, 80.35; H, 6.61%. 11c: a white solid,
mp 75–77 ꢁC, Rf ¼ 0:32 (CHCl3–AcOEt, 2:1); IR (KBr) 3385,
1
a) A. Sanyal and J. K. Snyder, Org. Lett., 2, 2527 (2000). b)
S. Jones and J. C. C. Atherton, Tetrahedron: Asymmetry, 12, 1117
(2001). c) J. C. C. Atherton and S. Jones, Tetrahedron Lett., 42,
8239 (2001). d) J. C. C. Atherton and S. Jones, J. Chem. Soc.,
Perkin Trans. 1, 2002, 2166. e) M. S. Corbett, X. Liu, A. Sanyal,
and J. K. Snyder, Tetrahedron Lett., 44, 931 (2003).
2
3
G. A. Martin and R. K. Hill, Chem. Rev., 61, 537 (1961).
a) A. S. Cieplak, Chem. Rev., 99, 1265 (1999). b) T.
Ohwada, Chem. Rev., 99, 1337 (1999).
4
5
F. Kaplan and H. Conroy, J. Org. Chem., 28, 1593 (1963).
H. Hart, A. Bashir-Hashemi, J. Luo, and M. A. Meador,
1
2930, 1493, 1259, 1042, 741 cmꢃ1; H NMR (500 MHz, CDCl3)
ꢀ 2.25–2.27 (m, 2H, 2CHCHCH2OH), 2.78 (brs, 2H, 2CH2OH),
3.24 (dd, J ¼ 9:6, 11.2 Hz, 2H, 2CHCHAHBOH), 3.61 (dd,
J ¼ 2:9, 11.2 Hz, 2H, 2CHCHAHBOH), 4.66 (brs, 2H,
Tetrahedron, 42, 1641 (1986).
T. R. Criswell and B. H. Klanderman, J. Org. Chem., 39,
770 (1974).
6

C. Kitamura et al.
Bull. Chem. Soc. Jpn., 77, No. 7 (2004) 1393
´
7
8
Y. Lepage, Ann. Chim., 4, 1137 (1959).
In the case of 9a, two independent molecules were found.
15 D. Suarez and J. A. Sordo, Chem. Commun., 1998, 385.
16 Recently a new synthetic method was reported: D. H. Hua,
M. Tamura, X. Huang, H. A. Stephany, B. A. Helfrich, E. M.
Perchellet, B. J. Sperfslage, J.-P. Perchellet, S. Jiang, D. E. Kyle,
and P. K. Chiang, J. Org. Chem., 67, 2907 (2002).
17 ‘‘SIR92’’: A. Altomare, M. C. Burla, M. Camalli, M.
Cascarano, C. Giacovazzo, A. Guagliardi, and G. Polidori, J. Appl.
Crystallogr., 27, 435 (1994).
However, the two were essentially identical. In Fig. 3b, the one
molecule is illustrated.
9 P. J. Breen, E. R. Bernstein, H. V. Secor, and J. I. Seeman,
J. Am. Chem. Soc., 111, 1958 (1989).
10 C. Kitamura, J. Fujimoto, M. Ouchi, and A. Yoneda, Anal.
Sci., 19, x69 (2003).
11 R. B. Woodward and R. Hoffman, ‘‘The Conservation of
Orbital Symmetry,’’ Academic Press, New York (1970).
12 a) L. J. Andrews and R. M. Keefer, J. Am. Chem. Soc., 77,
6284 (1955). b) K. Hayakawa, N. Mibu, E. Osawa, and K.
Kanematsu, J. Am. Chem. Soc., 104, 7136 (1982).
13 a) L. Salem, J. Am. Chem. Soc., 90, 543 (1968). b) N. D.
Epiotis, J. Am. Chem. Soc., 94, 1924 (1972).
18 ‘‘DIRDIF94’’: P. T. Beurskens, G. Admiraal, G. Beurskens,
W. P. Bosman, R. de Gelder, R. Israel, and J. M. M. Smits, ‘‘The
DIRDIF94 program system, Technical Report of the Crystallo-
graphic Laboratory,’’ University of Nijmegen, The Netherlands
(1994).
19 ‘‘teXsan’’: ‘‘Crystal Structure Analysis Package,’’ Molecu-
lar Structure Corporation (1999).
14 D. V. Banthorpe, Chem. Rev., 70, 295 (1970).