Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
1732
A. KODAMA et al.
tion, UV 260 or 400 nm. Under these conditions, 3 and 4
were eluted at tR ¼ 23:4 and 37.5 min, respectively.
compared to using radical species.12,13,19) Changing the
oxidant from the DPPH radical to o-chloranil could
simplify the chromatogram of the reaction mixture,
probably due to the mild oxidative property of the non-
radical chlorinated quinone compared to the DPPH
radical. Hence, a distinctive new peak (tR ¼ 43:5 min)
was observed in the equimolar reaction mixture of 4 and
o-chloranil in acetonitrile at 5 min, as well as a broad
peak (tR ¼ 23:6 min) detected at 400 nm. The former
peak increased in size during 70 min after the reaction
had started. These two peaks were also observed in the
corresponding reaction mixture in methanol. In addition,
a dark brown precipitate appeared in the acetonitrile
reaction solution after several hours. The precipitated
product was filtered off and directly subjected to in-
strumental analyses, which unveiled that it was a novel
benzocoumarin-type coupling product (5) derived from
two molecules of 4. The mass spectral analysis gave
a molecular peak at m=z 334, and the high-resolution
analysis indicated the molecular formula of C15H10O9,
which corresponded to a loss of CH3OH and 2H from
two molecules of 4. The 1H-NMR spectrum of 5 gave a
relatively simple peak pattern. Only two singlet signals
of aromatic protons at ꢀ 7.09 and 7.86, as well as a
methoxyl signal at ꢀ 3.94 were observed, apart from five
exchangeable hydroxyl protons in the low-field region of
ꢀ 9–11. In the 13C-NMR spectrum, 15 carbon signals,
including two ester carbonyls at ꢀ 164.0 and 168.7 and
one methoxyl at ꢀ 52.7, were observed. In the HMBC
spectrum, the lower-field aromatic proton of ꢀ 7.86
showed correlation peaks with two oxycarbons at ꢀ 142.5
and 149.4, as well as one of the ester carbonyls at 168.7
(C-7), and was thus assigned to H-6 of one unit of 4. On
the other hand, the other aromatic proton at ꢀ 7.09 had a
correlation peak with one upfield-shifted oxycarbon at ꢀ
132.7, being assignable to the central oxygenated carbon
(C-30) of a pyrogallol structure. This proton and the other
ester carbonyl of ꢀ 164.0 (C-70) gave a weak cross peak,
suggesting their W-shaped long-range relationship.
Hence, the higher-field proton could be reasonably
assigned as H-50 of the second unit of 4. The signal of
H-6 showed an additional strong three-bond cross peak
with a quaternary carbon at ꢀ 126.5, and similarly, H-50
also had a cross peak with another quaternary carbon at ꢀ
110.8. The former quaternary carbon was reasonably
assigned as C-60 (ꢀꢀ +13.2), and the latter as C-5 (ꢀꢀ
+11.0), since the chemical shifts of those carbons were
equally shifted downfield compared to unsubstituted C-6
(ꢀ 113.3) and C-50 (ꢀ 99.8), respectively, by their aryl
substitution. These HMBC correlations well support the
connectivity between C-5 of the first unit and C-60 of the
second unit. The connection of C-5 and C-60 enabled ꢀ-
lactone formation between OH-4 of the former unit and
the carboxyl of the latter. A detailed analysis of the
HMBC correlations unambiguously confirmed the total
structure of 5 (Fig. 2).
Isolation of benzocoumarin dimer 5 from the reaction
of methyl 2,3,4-trihydroxybenzoate (4) with o-chloranil.
To a solution of 4 (23 mg, 0.13 mmol) in acetonitrile
(1 ml) was added o-chloranil (31 mg, 0.13 mmol) in
acetonitrile (1 ml). After 16 hr at room temperature, a
dark brown precipitate appearing in the reaction mixture
was separated and washed with acetonitrile to give a
pure dimer (5) as dark brown powder (4 mg, 18%). 5.
Mp >300 ꢁC; FD-MS m=z (%): 334 (100); EI-HR-MS
m=z: 334.0334 ([M]þ, calcd for C15H10O9, 334.0325);
1H-NMR ꢀ (DMSO-d6) ppm: 3.94 (3H, s, H–OCH3),
7.09 (1H, s, H-50), 7.86 (1H, s, H-6), 9.17 (1H, s, OH-30),
9.86 (1H, s, OH-3), 10.57 (1H, s, OH-2), 10.58 (1H, s,
OH-40), 10.98 (1H, s, OH-20); 13C-NMR ꢀ (DMSO-d6)
ppm: 52.7 (C–OCH3), 98.0 (C-10), 99.8 (C-50), 110.7 (C-
1), 110.8 (C-5), 113.3 (C-6), 126.5 (C-60), 132.7 (C-30),
133.6 (C-3), 142.5 (C-4), 149.4 (C-2), 150.0 (C-20),
154.6 (C-40), 164.0 (C-70), 168.7 (C-7); HMBC corre-
lation: H–OCH3/C-7, H-50/C-5, 10, 30, 40 and 70, H-6/C-
2, 4, 7 and 60, OH-30/C-20, OH-3/C-2 and 4, OH-2/C-1,
2 and 3, OH-40/C-30 and 50, OH-20/C-10, 20 and 30.
Results and Discussion
The DPPH radical-scavenging abilities of 2,3,4-
trihydroxybenzoic acid (3) and methyl 2,3,4-trihydrox-
ybenzoate (4) were determined by the colorimetric
method. After 30 min, the relative radical-scavenging
equivalence of each compound, when that of ꢁ-toco-
pherol in ethanol as standard was designated as 2,17,18)
was as follows: 3, 4.3 and 4, 3.7 in acetonitrile; 3, 3.1
and 4, 2.1 in acetone; 3, 2.9 and 4, 4.7 in methanol; and
3, 2.7 and 4, 3.2 in ethanol. The radical-scavenging
ability of 3 was highest in acetonitrile and decreased in
the order of acetonitrile > acetone > methanol > etha-
nol, whereas methyl ester 4 showed the order of meth-
anol > acetonitrile > ethanol > acetone. Unlike the
protocatechuic esters, the DPPH radical-scavenging
equivalence of 2,3,4-trihydroxybenzoic acid (3) and its
ester (4) did not show any marked solvent dependence,
without any difference in alcoholic and non-alcoholic
solvents.9) The radical-scavenging equivalence of 3 and
4 in acetonitrile exceeded three, whereas nearly three or
fewer radicals were consumed in other solvents, except
with 4 in methanol. These results indicate that a further
oxidative reaction might have occurred in acetonitrile,
since only three radicals were expected to be scavenged
during the simple quinone-oxy radical formation from
pyrogallols.
An HPLC analysis of the reaction mixture of 3 or 4
with the DPPH radical in acetonitrile, however, gave
no significant peak of an oxidation product. Oxidation
mechanism studies on polyphenols have frequently used
o-chloranil as an oxidant from a catechol to the cor-
responding quinone, since the reaction proceeds cleanly
Benzocoumarin 5 comprises two molecules of 4 by
connecting between C-5 of one unit and C-6 of the other.