Correction of a reported xanthone synthesis: The preparation of a Benzo[c]coumarin

Article abstract of DOI:10.1002/hlca.200900289

Reinvestigation of a reported synthesis of 5,8-dihydroxy-1,3- dimethoxyxanthone from the reaction of 3,5-dimethoxyphenol with 2-(methoxycarbonyl)-1,4-benzoquinone resulted in the identification of the product as the isomer of benzo[c]coumarin, i.e., 7,10-

Full text of DOI:10.1002/hlca.200900289

Helvetica Chimica Acta – Vol. 93 (2010)
389
Correction of a Reported Xanthone Synthesis: The Preparation of a
Benzo[c]coumarin
by Michael H. Benna), Masood Parveza), Nigel B. Perryb), Arvi Rauka), and John W. van Klink*b)
a) Chemistry Department University of Calgary, Calgary, Alberta, T2N1N4, Canada
b) Plant Extracts Research Unit, New Zealand Institute for Plant and Food Research Ltd, Chemistry
Department, University of Otago, P. O. Box 56, Dunedin, New Zealand
(phone: þ 64-3-479-8966; e-mail: john.vanklink@plantandfood.co.nz)
Reinvestigation of a reported synthesis of 5,8-dihydroxy-1,3-dimethoxyxanthone from the reaction
of 3,5-dimethoxyphenol with 2-(methoxycarbonyl)-1,4-benzoquinone resulted in the identification of
the product as the isomer of benzo[c]coumarin, i.e., 7,10-dihydroxy-1,3-dimethoxy-6H-dibenzo[b,d]pyr-
an-6-one, established by X-ray crystallography. This requires the revision of the structures of the
derivatives that were reported.
Introduction. – While searching the literature for 1H- and 13C-NMR data for 1,3,5,8-
tetrahydroxyxanthone (¼ bellidin; 1; Scheme) in order to compare them with those of a
compound which we had isolated from a New Zealand gentian, Gentianella antarctica
(Hook.f.) T. N. Ho and S. W. Liu, we noticed an account of a synthesis [1].
According to a literature procedure [2], 3,5-dimethoxyphenol (2) and 2-methoxy-
pyridine were added to a solution of 2-(methoxycarbonyl)-1,4-benzoquinone (3) in
benzene [1]. This was anticipated to yield a diphenyl ether, 4, from which a xanthone
could be produced via reductive methylation, hydrolysis of the ester, and cyclization
with polyphosphoric acid [2]. Instead, two products were obtained, the major product
being identified as the xanthone 5, i.e., 4 was thought to cyclize spontaneously under the
reaction conditions to provide 5. The minor product was ascribed the structure 6.
It was reported that selective demethylation of the presumed 5 was surprisingly
more difficult than expected for a 1-methoxyxanthone, and that, when one MeO was
finally cleaved under forcing conditions, the product was not the known natural product
bellidifolin (7). As only one chelated OH was revealed by the 1H-NMR spectrum, and,
as the 13C-NMR spectrum lacked the resonances expected for a 1-hydroxyxanthone, it
was concluded that this product was 3,5,8-trihydroxy-1-methoxyxanthone (8) [1].
Complete demethylation gave what was considered to be the 1,3,5,8-tetrahydroxy-
xanthone, bellidin (1) [1].
However, while the melting point of 318 – 3208 of the synthetic product was in
agreement with the literature values for bellidin (1; see [1]), the NMR data were
significantly different from what we (unpublished data) and others [3] subsequently
observed. This induced us to reexamine the synthesis.
Results and Discussion. – In our hands, repetition of the reaction of 2 and 3
afforded, after chromatographic purification, a major product with the same melting
ꢀ 2010 Verlag Helvetica Chimica Acta AG, Zꢁrich

390
Helvetica Chimica Acta – Vol. 93 (2010)
Scheme. Reported (r) and Corrected (c) Reaction Pathways
i) MgSO4, 2-Methoxypyridine/258. ii) – iv) See [1]: for monodemethylation: AlCl3 in benzene/608, or
pyridinium hydrochloride/1408, or ZnCl2 and POCl3/608, and for complete demethylation: HI/Ac2O, or
AlCl3 in benzene/808.

Helvetica Chimica Acta – Vol. 93 (2010)
391
1
point and H-NMR spectrum as those reported in [1], and an EI-MS showing an
apparent molecular-ion peak at m/z 288 consistent with the required C15H12O6
composition for 5. However, there were very peculiar discrepancies in the 13C-NMR
spectra (see Table): 14 of the 15 resonances were as reported, but absent was one at
d(C) 178.5 ppm attributed to the C¼O group of a xanthone, and, instead, there was one
for a fully substituted sp2-C-atom at d(C) 101.1 ppm. The lowest-field 13C-NMR signal
which we observed was at 164.8 ppm.
Table. 13C-NMR Data for the Major Product (Rf 0.6) of the Reaction of 2 with 3 (in (D6)DMSO; ref. d(C)
39.5 ppm)
From [1] (20 MHz)
This work (100 MHz)
178.5 (s, CO)
–
164.8 (s, C(3))
161.0 (s, C (1))
155.9 (s, C(4a) or C(8))
155.1 (s, C(8) or C(4a))
151.5 (s, C(10a))
145.0 (s, C(5))
128.5 (d, C(6))
118.5 (s, C(8a))
116.8 (d, C(7))
105.2 (s, C(9a))
–
97.4 (d, C(2))
164.8 (s)
161.0 (s)
155.9 (s)
155.0 (s)
151.5 (s)
145.0 (s)
128.4 (d)
118.5 (s)
116.7 (d)
105.2 (s)
101.1 (s)
97.3 (d)
95.0 (d)
57.4 (q)
55.9 (q)
95.2 (d, C(4))
57.6 (q, MeO at C(1) or C(3))
56.0 (q, MeO at C(3) or C(1))
These data were compatible with the formation of a product isomeric with the
xanthone, the benzo[c]coumarin, 6H-dibenzo[b,d]pyran-6-one 10, which we visualized
as formed via Michael-type addition of 2 to 3 yielding 9, i.e., initial C- rather than O-
addition of the ambident 2 to 3, with a subsequent base-catalyzed intramolecular
lactonization transforming 9 to 10 (see Scheme). From a mechanistic viewpoint this
seemed much more plausible than the original suggestion [1] that xanthone formation
occurred via a spontaneous cyclization of the product of O-addition: while O-adducts
had been obtained by the base-catalyzed addition of phenols to 2, their conversion to
xanthones required acid catalysis [2]. Also, although the C,C coupling of 2 with
phenols is normally associated with acid catalysis [4], it does occur with highly
nucleophilic species such as enols [5] without such catalysis.
Consistent with our hypothesis, 13C¼O signals around 165 ppm have been reported
for a series of 7-hydroxy-6H-dibenzo[b,d]pyran-6-ones [6].
The matter was solved by X-ray crystallography which revealed our product to have
the structure shown in the Figure: 7,10-dihydroxy-1,3-dimethoxy-6H-dibenzo[b,d]pyr-
an-6-one (10). As shown, there were two intermolecularly H-bonded molecules per

392
Helvetica Chimica Acta – Vol. 93 (2010)
unit cell, exhibiting the strong intramolecular H-bonds expected. Full crystallographic
details have been deposited with the Cambridge Crystallographic Data Centre1).
Figure. X-Ray crystal structures of the two crystallographically independent molecules 10
In short, a crucial error in misidentifying the product of the condensation of 3,5-
dimethoxyphenol (2) and 2-(methoxycarbonyl)-1,4-benzoquinone (3) as the xanthone
5, rather than the benzocoumarin 9, has resulted in the need to revise the structures of
all of the derivatives of it described in the original publication [1]: 23 other compounds,
the true structures of 7 of which can be deduced with reasonable certainty (the
diacetate, dibenzoate, dimethyl, and dibenzyl derivatives of 10; and the completely
demethylated product, 11, and its tetraacetate and tetrabenzoate). It also appears
probable that the minor product assigned the structure 6 is actually 12.
Some 6H-dibenzo[b,d]pyran-6-ones have biological activities (see those noted in
[6]) so it would be interesting to examine the activities of the derivatives synthesized by
Vermes et al. [1].
We thank Dorothy Fox, Johnson Li, Loeke Janzen, Qiao Wu, and Ragav Yamdagni for assistance in
collecting the spectrometric data. Financial support was provided from funds provided by Sigma Aldrich
(to M. H. B.).
Experimental Part
General. Anal. and prep. TLC: SiO2 60 (0.2 mm) on glass plates (5 ꢀ 20 and 20 ꢀ 20 cm, resp.) with
EtOH/toluene 1:9 (v/v) as eluent, visualization under short- and long-wavelength UV light. UV Spectra:
Cary 300 spectrometer, absorptions reported as lmax in nm (log e). Fluorescence spectrum: Aminco
Bowman Series 2 luminescence spectrometer. NMR Spectra: Bruker Instruments DRX 400 spectrometer
using CDCl3 or (D6)DMSO (chemical shifts in ppm referred to solvent signals d(H) 7.25 or 2.51, d(C) 77.0
or 39.5, resp.), coupling constants J in Hz. EI-MS (70 eV): Finnigan MAT instrument; only ions with an
abundance greater than 20% of the most abundant ion are reported; in m/z.
For the X-ray crystallographic study, a colorless plate crystal of the compound was coated with
Paratone 8277 oil (Exxon) and mounted on a glass fiber. All measurements were made on a Bruker
APEX2 CCD installed on a Nonius Kappa Goniometer diffractometer with graphite monochromated
MoKa radiation. The data were collected [7] using w and f scans. The data were corrected for Lorentz
1
)
CCDC-729937 contains the supplementary crystallographic data for this article. These data can be
obtained free of charge from the Cambridge Crystallographic Data Centre, via www.ccdc.cam.ac.uk/
data_request/cif.

Helvetica Chimica Acta – Vol. 93 (2010)
393
and polarization effects, and for absorption using multi-scan method [8]. The structure was solved by
direct methods [9] and expanded using Fourier techniques [10]. The asymmetric unit contains two
independent molecules. The non-H-atoms were refined anisotropically. All of the H-atoms were located
from a difference Fourier map and were allowed to refine with isotropic thermal parameters. The final
cycle of full-matrix least-squares refinement using SHELXL97 [11] converged with unweighted and
weighted agreement factors, R ¼ 0.0415 and wR ¼ 0.1071 (all data), resp., and goodness-of-fit, S ¼ 1.081.
The weighting scheme was based on counting statistics, and the final difference Fourier map was
essentially featureless. The figures were plotted with the aid of ORTEP-3 for Windows [12].
7,10-Dihydroxy-1,3-dimethoxy-6H-dibenzo[b,d]pyran-6-one (10). According to the procedure
described in [1], a soln. of 2-(methoxycarbonyl)-1,4-benzoquinone (3) in anh. benzene (17 ml) was
prepared in situ by the oxidation of methyl 2,5-dihydroxybenzoate (1.68 g) with a suspension of Ag2O
(5.1 g) and anh. K2CO3 (850 mg) at 508. After filtration, MgSO4 (1.7 g) was added, and the suspension
was stirred, while a soln. of 3,5-dimethoxyphenol (1.07 g) and 2-methoxypyridine (2.28 g) in dry benzene
(10 ml) was added dropwise over the course of 10 min. After having been stirred for a further 2 h at r.t.,
the mixture was filtered, and the filter cake was washed with a little dry benzene. The combined filtrate
and washings were evaporated under reduced pressure (17 mm, rotovap, 658) to yield a dark red-brown
oil. This was dissolved in MeOH (50 ml) and stored at 08 overnight. The bronze-colored solid which
separated was collected by filtration, washed with a little ice-cold MeOH, and air-dried to yield a copper-
colored solid (540 mg, ca. 19%). TLC (toluene/EtOH 9 :1 (v/v)) revealed two components as yellow
spots with Rf 0.6 and 0.42. A portion of this mixture (15 mg) was separated by prep. TLC in the same
solvent system, and the material with Rf 0.6 was recrystallized from EtOH to afford 10 (8 mg, 0.4%). Pale
yellow laths. M.p. 168 – 169 ([1]: 167 – 1688). UV and fluorescence spectra (in 95% EtOH): 213 (4.57),
246 (4.48), 276 (4.02), 379 (4.16); lex 379, lem 464. 1H-NMR (CDCl3): 11.21 (s, 1 H); 8.77 (s, 1 H); 7.33 (d,
J ¼ 9.0, 1 H); 7.01 (d, J ¼ 9.0, 1 H); 6.62 (d, J ¼ 2.5, 1 H); 6.53 (d, J ¼ 2.5, 1 H); 4.06 (s, 3 H); 3.88 (s, 3 H).
1H-NMR from [1] (100 MHz): 11.22 (s, 1 H); 8.77 (s, 1 H); 7.31 (d, J ¼ 9, 1 H); 7.00 (d, J ¼ 9, 1 H); 6.57 (d,
J ¼ 2.5, 1 H); 6.49 (d, J ¼ 2.5, 1 H); 4.05 (s, 3 H); 3.86 (s, 3 H). 13C-NMR: see Table. EI-MS: 288 (100),
273 (92), 245 (33), 213 (47).
The minor product, 7,10-dihydroxy-8-(2-hydroxy-4,6-dimethoxyphenyl)-1,3-dimethoxy-6H-diben-
zo[b,d]pyran-6-one (12) with Rf 0.42 was similarly obtained as bronze prisms (2 mg, 0.06%). M.p.
1
259 – 2628 ([1]: 259 – 2628). H-NMR ((D6)DMSO): 11.11 (s, 1 H); 9.26 (s, 1 H); 8.80 (s, 1 H); 7.08 (s,
1 H); 6.77 (d, J ¼ 2.5, 1 H); 6.74 (d, J ¼ 2.5, 1 H); 6.15 (s, 2 H); 4.05 (s, 3 H); 3.87 (s, 3 H); 3.75 (s, 3 H);
3.63 (s, 3 H). 1H-NMR from [1] (100 MHz): 11.11 (s, 1 H); 9.25 (s, 1 H); 8.80 (s, 1 H); 7.10 (s, 1 H); 6.77 (2
arom. H); 6.2 (s, 2 H); 4.10 (s, 3 H); 3,90 (s, 3 H); 3.77 (s, 3 H); 3.65 (s, 3 H). 13C-NMR ((D6)DMSO):
165.6 (s); 160.8 (s); 160.4 (s); 158.6 (s); 156.3 (s); 155.9 (s); 154.2 (s); 151.4 (s); 144.0 (s); 131.5 (d); 123.9
(s); 117.0 (s); 104.8 (s); 104.6 (s); 101.3 (s); 97.3 (d); 95.3 (d); 94.1 (d); 90.1 (d); 57.6 (q); 56.0 (q); 55.6 (q);
55.1 (q). 1C-NMR from [1] (20.15 MHz; only 22 resonances listed): 165.7 (s); 160.9 (s); 160.6 (s); 158.7
(s); 156.4 (s); 155.8 (s); 154.5 (s); 151.5 (s); 144.0 (s); 131.7 (d); 124.1 (s); 117.1 (s); 105.2 (s); 104.7 (s);
97.5 (d); 95.3 (d); 94.1 (d); 90.1 (d); 57.6 (q); 56.0 (q); 55.6 (q); 55.1 (q). EI-MS: 440 (100), 422 (30), 407
(22).
REFERENCES
[1] B. Vermes, O. Seligmann, H. Wagner, Helv. Chim. Acta 1985, 92, 375.
[2] P. Mꢁller, T. Venakis, C. H. Eugster, Helv. Chim. Acta 1979, 62, 2350.
[3] a) H. Kanamori, I. Sakumoto, M. Mizuta, K. Hashimoto, O. Tanaka, Chem. Pharm. Bull. 1984, 32,
2290; b) H.-M. Zhou, Y.-L. Liu, G. Blasko, G. A. Cordell, Phytochemistry 1989, 28, 3569.
[4] P. Mꢁller, T. Venakis, C. H. Eugster, Helv. Chim. Acta 1979, 62, 2833.
[5] S. E. Fumagalli, C. H. Eugster, Helv. Chim. Acta 1971, 54, 959.
[6] B. Appel, N. N. R. Saleh, P. Langer, Chem. – Eur. J. 2006, 12, 1221.
[7] Z. Otwinowski, W. Minor, Methods Enzymol. 1997, 276, 307.
[8] R. Hooft, COLLECT, Nonius BV, Delft, The Netherlands, 1998.
[9] A. Altomare, M. Cascarano, C. Giacovazzo, A. Guagliardi, J. Appl. Crystallogr. 1993, 26, 343.

394
Helvetica Chimica Acta – Vol. 93 (2010)
[10] P. T. Beurskens, G. Admiraal, G. Beurskens, W. P. Bosman, R. de Gelder, R. Israel, J. M. N. Smits,
The DIRDIF-94 program system, Technical Report of the Crystallography Laboratory, University
of Nijmegen, The Netherlands, 1994.
[11] G. M. Sheldrick, Acta Crystallogr., Ser. A 2008, A64, 112.
[12] L. J. Farrugia, J. Appl. Crystallogr. 1997, 30, 565.
Received August 4, 2009