Rapid preparation of 3-deoxyanthocyanidins and novel dicationic derivatives: New insight into an old procedure

Article abstract of DOI:10.1002/ejoc.200601071

Common 3-deoxyanthocyanidins and original dicationic flavylium- benzopyrylium derivatives are easily and efficiently synthesized through reactions between the corresponding phenols and aryl ethynyl ketones in the presence of aqueous hexafluorophosphoric acid. The mechanism of the reaction is discussed and two competitive pathways are consistent with our results. Wiley-VCH Verlag GmbH & Co. KGaA, 2007.

Full text of DOI:10.1002/ejoc.200601071

FULL PAPER
DOI: 10.1002/ejoc.200601071
Rapid Preparation of 3-Deoxyanthocyanidins and Novel Dicationic Derivatives:
New Insight into an Old Procedure
Stefan Chassaing,*[a] Marie Kueny-Stotz,[a] Géraldine Isorez,[a] and Raymond Brouillard[a]
Keywords: Oxygen heterocycles / Ketones / Dyes / Pigments / Natural products / Synthetic methods
Common 3-deoxyanthocyanidins and original dicationic flav-
ylium-benzopyrylium derivatives are easily and efficiently
synthesized through reactions between the corresponding
phenols and aryl ethynyl ketones in the presence of aqueous
hexafluorophosphoric acid. The mechanism of the reaction is
discussed and two competitive pathways are consistent with
our results.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2007)
structures (Figure 2).[2] At a structural level, it is also
worthwhile to distinguish 3-oxygenated aglycones 2 (antho-
cyanidins) from their 3-deoxy analogues 3 (3-deoxyantho-
cyanidins) (Figure 1). Unlike the red-coloured anthocyanid-
ins 2, which are colour-unstable and therefore seldom found
naturally in their free forms, 3-deoxyanthocyanins 3 are yel-
low-orange pigments quite stable in acidic media, and such
aglycones 3 – in particular apigeninidin (4), luteolinidin (5)
and tricetanidin (6) – have been isolated from food plants
such as corn, sorghum and black tea leaves (Figure 1).[4]
Introduction
The flavylium (2-phenylbenzopyrylium) skeleton 1 is the
chromophoric core of one of the most widely encountered
family of natural pigments found in the plant kingdom: the
anthocyanins (Figure 1). Flavylium-derived pigments are
indeed largely responsible for cyanic colours going from
salmon pink, through red and violet, to dark blue in most
flowers, fruits and leaves of angiosperms.[1]
Figure 1. The flavylium chromophore (1: R3 = R5 = R7 = R3Ј
=
R4Ј = R5Ј = H), the two flavylium-derived families of pigments
(anthocyanidin 2: R3 = OH and 3-deoxyanthocyanidin 3: R3 = H)
and the most common 3-deoxyanthocyanidin patterns: apigenini-
din (4: R3 = R3Ј = R5Ј = H and R5 = R7 = R4Ј = OH), luteolinidin
(5: R3 = R5Ј = H and R5 = R7 = R3Ј = R4Ј = OH) and tricetanidin
(6: R3 = H and R5 = R7 = R3Ј = R4Ј = R5Ј = OH).
In these water-soluble natural pigments, the flavylium
chromophore is usually substituted and/or functionalized in
various and numerous ways. Nature has used sophisticated
tools such as hydroxylation, alkylation, acylation and/or
glycosylation to adjust the properties of these pigments.
Also, because of the impressive number of possible combi-
nations, it is not surprising that new anthocyanins are regu-
larly isolated from natural sources so that, at the present
time, about 580 pigments from this family are already listed
in the literature, ranging from simple to very complicated
Figure 2. Illustration of the exceptional diversity existing around
the flavylium nucleus with two characteristic malvidin-derived pig-
ments.
Anthocyanins, and especially 3-deoxyanthocyanidins,
have long been used as food colorants,[5] but during the last
decade many other attractive applications have been devel-
oped. They have been advanced as potential hair dyes,[6]
[a] Laboratoire de Chimie des Polyphénols, Université Strasbourg
I, CNRS: LC3 UMR 7177,
4, rue Blaise Pascal, 67008 Strasbourg, France
E-mail: stefanchassaing@mac.com
2438
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2438–2448

3-Deoxyanthocyanidins and Novel Dicationic Derivatives
FULL PAPER
laser dyes,[7] or sensitizers for solar cells.[8] Molecular-level venient and efficient strategy allowing precisely the prepara-
memory systems[9] have been based on such structures and tion of these monocationic flavylium derivatives but also
their use in the area of self-assembled and interface systems the preparation of novel dicationic flavylium-benzopyryl-
has been reported.[10] Moreover, unlike most of the other ium derivatives.
classes of flavonoids, anthocyanins were only recently dem-
Despite the longstanding interest in flavylium-derived
onstrated to possess biological properties beneficial to hu- pigments, only a few synthetic approaches towards 3-deoxy-
man health: for illustration, regular consumption of antho- anthocyanidins 3 have been reported in the literature until
cyanins (and also other polyphenols) from fruits, vegetables, now.[5a] The standard synthetic method, pioneered in the
wines, jams and preserves is closely associated with reduced early 1930s by Pratt and Robinson, is based on an acid-
risks of chronic diseases such as cancer, cardiovascular dis- mediated condensation between a protected salicylaldehyde
eases and Alzheimer’s disease.[11] In this pharmacological derivative 7 and a substituted acetophenone 8, this strategy
context, anthocyanins are thus regarded as important po- being analogized to a “C6–C1 + C2–C6” synthetic ap-
tential nutraceuticals for humans.
proach (Scheme 1).[14] Although this method was the first
For all these reasons and because such compounds con- to allow the preparation of both apigeninidin (4) and luteo-
stitute good models for studying the structural[12] and the linidin (5) (in moderate yields), the Robinson’s condensa-
redox[13] behaviours of anthocyanins in aqueous medium, tion presents severe limitations and cannot accommodate
we decided to elaborate synthetic methodologies to prepare any kind of salicylaldehyde reagent. Indeed, when the sali-
3-deoxyanthocyanidin derivatives 3. Here we report a con- cylaldehyde is a phloroglucinol-type derivative, it is well
Scheme 1. Reported retrosynthetic analyses of the 3-deoxyanthocyanidin skeleton.
Scheme 2. Synthetic routes to 3-deoxyanthocyanidins through the chemical modification of a flavonoid precursor.[17]
Eur. J. Org. Chem. 2007, 2438–2448
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2439

S. Chassaing, M. Kueny-Stotz, G. Isorez, R. Brouillard
FULL PAPER
known that Robinson’s conditions give rise to a hard-to- Table 1. Synthesis of aryl ethynyl ketone patterns.
purify mixture of organic salts, among which the targeted
flavylium salt 3 is generally isolated in very limited amounts
(5–20% yields).[14] However, two alternatives, circumventing
this structural restriction, have recently been proposed:
Kuhnert et al. developed an efficient BF3·Et2O-mediated
version,[15] while T. Mas optimized the condensation by
using the peracetylated derivative of phloroglucinaldehyde
as reagent and by performing the reaction in acidic meth-
anol.[16]
Another classic route to 3-deoxyanthocyanidins 3 is
based on the chemical modification of a flavonoid precur-
sor already possessing the appropriate carbon skeleton.
Sweeny and Iacobucci, for example, described two re-
duction/oxidation sequences that, when applied to 2Ј-hy-
droxychalcones and flavanones, allowed access to phloro-
glucinol-type 3-deoxyanthocyanidins (Scheme 2).[17] This
approach, however, is inherently limited by the oxidation
step, which generally proceeded in moderate yields (30–
70%). The other drawback of this approach is that it re-
quires the native C6–C3–C6 flavonoid skeleton, which is
not always easily accessible.
In view of these different synthetic limitations, we looked
for a more general approach towards 3-deoxyanthocyanid-
ins 3 and, more particularly, phloroglucinol-type deriva-
tives. On the basis of a report by Johnson and Melhuish
dating back to 1947, we investigated the acid-mediated con-
densation between a phenolic derivative 9 and an aryl eth-
ynyl ketone 12, corresponding to a “C6 + C3–C6” synthetic
strategy (Scheme 1).[18] Surprisingly, we noted that this ap-
proach had never been optimized or even used since its de-
scription sixty years ago. Our goals were therefore on one
hand to study this neglected procedure further, and on the
other to explore its scope.
Results and Discussion
In order to prepare the required aryl ethynyl ketone syn-
thons 12, we have developed a simple and efficient two-step
nucleophilic addition/oxidation sequence (Table 1). The nu-
cleophilic addition of ethynylmagnesium bromide to benz-
aldehyde derivatives 10a–g resulted in 1-arylprop-2-yn-1-ols
11a–g (10a–d are commercially available materials and 10e–
g were obtained from the corresponding commercially
[a] Isolated yield after purification by column chromatography.
We then investigated the acid-mediated key-step conden-
available phenols by conventional O-silylation methods). sation involving a phenolic derivative 9 as the nucleophilic
We then considered the oxidation of these 1-arylprop-2-yn- entity and an aryl ethynyl ketone 12 as its electrophilic part-
1-ols 11 and although various conditions had already been ner (Table 2). In order to improve the condensation pro-
reported (chromium salts, Swern conditions, oxovanadium cedure adopted by Johnson and Melhuish (concentrated
complexes, Dess–Martin reagent),[19] use of the IBX/ethyl H2SO4/AcOH/room temp.; then recrystallization from di-
acetate system for the oxidation of propargylic alcohols had lute HCl),[18] we decided to substitute concentrated sulfuric
so far never been described.[20] Interestingly, this easy-to- acid with dilute hexafluorophosphoric acid. This choice was
use oxidative system proved to be very efficient and thus motivated by the fact that we had recently examined various
represents a good alternative for the preparation of ynones acidic conditions for the formation of flavylium salts (i.e.,
(Table 1). By this two-step sequence, all desired synthons aqueous HCl, HBr, HBF4 or HPF6) and had observed that
12a–g were synthesized in good to excellent overall yields flavylium hexafluorophosphates exhibited the lowest solu-
(74–90%).
bility products in acetic acid and so could easily be reco-
2440
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2438–2448

3-Deoxyanthocyanidins and Novel Dicationic Derivatives
FULL PAPER
Table 2. Formation of 3-deoxyanthocyanidin derivatives from vari- vered in pure form. In addition, whereas flavylium chlo-
rides,[16] tetrafluoroborates,[15,21] or triflates[21] have already
ous phenol/aryl ethynyl ketone combinations.
been described, flavylium hexafluorophosphates had never
previously been disclosed in the literature.
In this context, we initially studied the effect of HPF6 on
some 9/12 combinations already reported by Johnson and
Melhuish so that we could check the potential of these
modified conditions (Entries 1–7). On one hand, in terms
of workability of the process, the comparison between the
two conditions showed strong analogies. Although the con-
densation is still inefficient in the case of phenol (9a), m-
cresol (9b) and p-cresol (9c) (Entries 1–3), it is nevertheless
efficient in the case of resorcinol (9d), phloroglucinol (9e)
and phloroglucinol dimethyl ether (9f) (Entries 4–6). It is
worth noting that the interaction between resorcinol (9d)
and phenyl ethynyl ketone (12a) resulted in the selective for-
mation of the single compound 7-hydroxyflavylium (13a),
with a distinctive AMX-type system of spins on its A ring
(no traces of its isomer 13b – i.e., 5-hydroxyflavylium com-
pound presenting a different but also distinctive AMX sys-
tem – could be detected by NMR). The origin of this re-
gioselectivity is explained later when the mechanism of the
condensation is discussed (Scheme 3). Anyway, all these
preliminary results confirmed that the success of the con-
densation is directly dependent on the nucleophilic charac-
ter of the phenolic derivative. Indeed, this condensation re-
quires a minimal nucleophilicity in the phenolic partner;
otherwise it does not work at all and the starting materials
are recovered even after stirring for 72 hours. On the other
hand, in terms of efficiency of the process, it clearly ap-
peared that the use of HPF6 instead of H2SO4 induced a
significant increase in the condensation process, providing
the corresponding 3-deoxyanthocyanidin derivatives in
good to excellent yields (Entries 4–7). The most significant
improvement involved the preparation of chrysinidin (14)
through the interaction between phloroglucinol (9e) and
phenyl ethynyl ketone (12a), the yield increasing from 37%
with H2SO4 to 91% with HPF6 (Entry 5).
We next checked whether this method could be general-
ized to other 3-deoxyanthocyanidin derivatives. We first
corroborated the efficiency of the procedure for the prepa-
ration of polymethylated 3-deoxyanthocyanidin derivatives
[i.e., chrysinidin dimethyl ether (17) and tricetanidin penta-
methyl ether (18); Entries 8 and 9]. Permethylated deriva-
tives were often used as synthetic precursors of their pheno-
lic analogues but it is noteworthy that the demethylation
step is generally effective in only moderate yields.[15,22] We
therefore focused on a one-step synthesis of polyphenolic
3-deoxyanthocyanidins by the procedure investigated here.
Further to the promising result obtained with the use of
phloroglucinol (9e) as the nucleophilic entity (Entry 5), we
decided to subject 9e to the condensation process with aryl
ethynyl ketones possessing phenolic functions protected
[a] Reference conditions: concentrated H2SO4/AcOH/room temp.,
with an appropriate group. We chose the acid-labile tert-
then recrystallization from dilute HCl.[18] [b] Studied conditions:
butyldimethylsilyl group, and the reaction between phloro-
HPF6 (50% in water)/AcOH/room temp./24 to 48 h. [c] No reac-
tion. [d] Successful reaction, but corresponding yield not given in
the reference. [e] Only regioisomer observed. [f] Not tested under
the reference conditions.
glucinol (9e) and silyl-protected 12e–g resulted in the one-
step formation of polyphenolic 4, 5 and 6 (Entries 10–12).
As a consequence, here we have performed total syntheses
Eur. J. Org. Chem. 2007, 2438–2448
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2441

S. Chassaing, M. Kueny-Stotz, G. Isorez, R. Brouillard
FULL PAPER
Scheme 3. Plausible mechanistic pathways for the acid-mediated condensation between an aryl ethynyl ketone and a phenolic derivative.
of the natural apigeninidin (4), luteolinidin (5) and tricetan-
Until now, no mechanism has been proposed for this
idin (6) in four steps with high overall yields starting from acid-mediated condensation, so it seemed constructive to us
the corresponding unprotected benzaldehydes (69%, 72% to discuss our mechanistic hypothesis here. What is sure is
and 61%, respectively).
that such a condensation would involve a cascade of suc-
Moreover, by analogy with the interaction between resor- cessive events, resulting in the overall formation of one C–
cinol (9d) and phenyl ethynyl ketone (12a) (Entry 4), we in- C bond and one C–O bond, together with the establishment
vestigated the selectivity of the condensation with unsym- of the benzopyrylium aromatic character. The question is
metrical 1,2,4-trihydroxybenzene (9g) as the nucleophilic to determine how these events would occur and, in fact,
partner (Entry 13). Although four regioisomers could be ex- two competitive pathways could be envisaged, especially
pected (i.e., 5,6-, 5,8-, 6,7- and 6,8-dihydroxyflavylium de- with regard to the way in which the C–C bond could be
rivatives), we observed the exclusive formation of one iso- formed (Scheme 3). It is noteworthy that both would give
mer, which was shown to be 6,7-dihydroxyflavylium hexa- the same retrochalcone-type intermediate C1. On one hand,
fluorophosphate (19). Indeed, the NMR spectrum of the Pathway A involves the formation of the C–C bond through
obtained product showed two characteristic signals in the the nucleophilic 1,4-addition of a carbon atom of 9 to acid-
aromatic region (i.e., one singlet at δ = 6.81 ppm and one activated 12 to generate intermediate A1, in which proton
doublet at δ = 7.67 ppm) and 19 is precisely the only re- migration produces the expected intermediate C1. On the
gioisomer that is in agreement with this observation.
other hand, Pathway B involves a more complex pathway
2442
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2438–2448

3-Deoxyanthocyanidins and Novel Dicationic Derivatives
FULL PAPER
to C1, starting with the formation of the hemiketal interme- could nevertheless exhibits interesting properties, especially
diate B1 through the nucleophilic addition of 9 through its thanks to its dicationic character. Until now, the only syn-
phenolic position to acid-activated 12. The intermediate B1 thetic approach to such derivatives has been that described
thus presents a suitable propargyl vinyl ether pattern pre- in 1969 by Reynolds and Van Allan.[23] In their work, the
cursor to a Claisen-type rearrangement to afford the inter- authors mentioned the formation of the derivative 23
mediate B2, which could progress towards C1 through a through a double Robinson condensation between two
simple tautomerization process. Once the intermediate C1 equivalents of salicylaldehyde 21 and one equivalent of di-
is produced, we then join the common pathway involving ketone 22 in a poor yield of 8% (see 3b in Figure 3). As a
the establishment of the C–O bond through a 6-exo-trig new alternative, we imagined that, by starting from bis(pro-
cyclization of the retrochalcone C1. Finally, the resulting pynoyl)benzene derivatives, a double condensation with two
cyclic hemiketal C2 could aromatize after protonation and equivalents of phloroglucinol derivatives could also provide
elimination of water, creating the aromatic benzopyrylium flavylium-benzopyrylium salts.
nucleus of the flavylium salt. From our results, it is evident
By analogy, the application of our condensation process
that the condensation process is effective whatever the elec- required the initial preparation of an aryl diethynyl dike-
trophilic partner 12, as long as the phenolic derivative 9 tone pattern. Thus, by starting from the commercially avail-
is nucleophilic enough, the C-nucleophilicity of the phenol able isophthalic carboxaldehyde (24) and terephthaldehyde
clearly emerging as the key parameter of the process. Never- (25), the synthesis of two aryl diethynyl ketone patterns 26
theless, this fact is not decisive for explaining the mechan- and 27 was achieved through the same two-step sequence as
istic pathway, because the C-nucleophilic character of the described above for the preparation of aryl ethynyl ketone
phenol could suggest either Pathway A involving the forma- synthons 12 (Scheme 4). Each of the two aryl diethynyl di-
tion of A1 or Pathway B with the formation of B2. More- ketone patterns 26 and 27 were then engaged in a condensa-
over, the regioselectivity observed with resorcinol (9d) or tion step in the presence of HPF6 with phloroglucinol (9e)
1,2,4-trihydroxybenzene (9g) could also be explained by or phloroglucinol dimethyl ether (9f). In this manner, three
both proposed pathways. In these two cases, the regioselec- new flavylium-derived pigments 28, 30 and 31 were ob-
tivity would result from the selective formation of the C–C tained in excellent yields. Unfortunately, the reaction be-
bond, which could, for electronic and steric reasons, pro- tween one equivalent of 26 and two equivalents of 9e gave
gress either through the intermolecular alternative, to afford a hard-to-purify mixture of organic salts, among which the
A1 (Pathway A), but also through the intramolecular alter- expected dicationic derivative 29 was detected but only in a
native from the corresponding intermediate B1 (Pathway very limited amount.
B). In the case of 9g, it is also possible to imagine that the
corresponding intermediate B1 could be produced through
the nucleophilic addition of 9g through the phenolic posi-
tion 4, which could be seen as its most reactive phenolic
Conclusions
site. Consequently, the interpretation of the mechanism of
this condensation process is not facile, Pathway A and Path- ple and very efficient method for the synthesis of 3-deoxy-
way B both being plausible and difficult to differentiate. anthocyanidins through reactions between activated phe-
In conclusion, we have reactivated and optimized a sim-
To expand the scope of this condensation further, we en- nols and aryl ethynyl ketones in acetic acid in the presence
visaged the application of our procedure to the synthesis of HPF6. At the same time, by the same method, we have
of bispyrylium salts and, more precisely, those possessing a succeeded in the synthesis of novel dicationic flavylium-
flavylium-benzopyrylium skeleton 20 (see 3a in Figure 3). benzopyrylium skeletons.
To the best of our knowledge, no such sophisticated skel-
Further works, not only to study the behaviour of the
eton has ever been encountered in natural pigments but it newly synthesized dicationic pigments in aqueous media
Figure 3. The flavylium-benzopyrylium skeleton: nature and the only synthetic approach developed until now.
Eur. J. Org. Chem. 2007, 2438–2448
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2443

S. Chassaing, M. Kueny-Stotz, G. Isorez, R. Brouillard
FULL PAPER
Scheme 4. Synthesis of dicationic flavylium-benzopyrylium derivatives.
13 mmol, 1.3 equiv.) was added at 0 °C to a solution of the corre-
sponding benzaldehyde 10a–g (10 mmol) in THF (20 mL). After
the mixture had been stirred for 2 h at room temperature, a satu-
rated solution of NH4Cl (20 mL) was added to the solution and
the THF was evaporated under vacuum. The aqueous phase was
extracted three times with ethyl acetate and the organic layers were
washed with water and brine and then dried with Na2SO4. After
evaporation of the solvent, the resulting crude product was purified
by column chromatography (heptane/ethyl acetate) to give 11a–g
in pure form.
but also to continue to broaden the scope of this reaction,
are now underway.
Experimental Section
General Remarks: Most reactions were performed under argon with
magnetic stirring and in anhydrous solvents. All starting materials
were commercially available and were used without further purifi-
cation. The reactions were monitored by thin-layer chromatog-
raphy carried out on silica plates (silica gel 60 F254, Merck) with
use of UV light for detection. Column chromatography was per-
formed on silica gel 60 (0.040–0.063 mm, Merck) with the eluents
indicated. Melting points (m.p.) were determined with a Bibby–
Stenlin Stuart SMP3 melting point apparatus and are uncorrected.
IR spectra were recorded with a Perkin–Elmer FTIR 1600 spec-
trometer (KBr disc) and values are reported in cm–1. UV/Vis spec-
tra were measured with a 8452A Hewlett–Packard spectrophotom-
eter in the solvents indicated. 1H and 13C NMR spectra were re-
corded with a Bruker AC 300 spectrometer at 300 and 75 MHz,
respectively. Chemical shifts (δ) and coupling constants (J) are
given in ppm and Hz, respectively. Low- and high-resolution mass
spectra were obtained with a Bruker micro-TOF instrument (ESI).
1-Phenylprop-2-yn-1-ol (11a): Yellow oil. Yield 79%. 1H NMR
(300 MHz, CDCl3, 25 °C): δ = 2.49 (d, 3JOH–1 = 6.2 Hz, 1 H, OH),
4
3
4
2.66 (d, J3–1 = 2.2 Hz, 1 H, 3-H), 5.45 (dd, J1–OH = 6.2 Hz, J1–3
= 2.2 Hz, 1 H, 1-H), 7.37 (m, 3 H, 2Ј-H/ 4Ј-H/ 6Ј-H), 7.55 (m, 2
H, 3Ј-H and 5Ј-H) ppm. 13C NMR (75 MHz, CDCl3, 25 °C): δ =
64.4 (C-1), 74.9 (C-2), 83.5 (C-3), 126.6 (C-2Ј and C-6Ј), 128.6 (C-
4Ј), 128.7 (C-3Ј and C-5Ј), 140.0 (C-1Ј) ppm. IR (KBr): ν = 3335
˜
(O–H), 3290 (ϵC–H), 2115 (CϵC) cm–1. MS (ESI, positive mode):
m/z (%) = 139 (100) [M + Li]+, 111 (40). HRMS: calcd. for
C9H8LiO: 139.0730; found: 139.0730.
1-(3Ј,4Ј-Dimethoxyphenyl)prop-2-yn-1-ol (11c): White solid. Yield
90%. M.p. 98 °C. 1H NMR (300 MHz, CDCl3, 25 °C): δ = 2.32 (d,
3JOH–1 = 5.9 Hz, 1 H, OH), 2.67 (d, 4J3–1 = 2.2 Hz, 1 H, 3-H), 3.87
3
(s, 3 H, OCH3), 3.89 (s, 3 H, OCH3), 5.41 (dd, J1-OH = 5.9 Hz,
4J1–3 = 2.2 Hz, 1 H, 1-H), 6.85 (d, 3J5Ј–6Ј = 8.8 Hz, 1 H, 5Ј-H), 7.08
4
3
4
(d, J2Ј–6Ј = 2.2 Hz, 1 H, 2Ј-H), 7.08 (dd, J6Ј–5Ј = 8.8 Hz, J6Ј–2Ј
=
2.2 Hz, 1 H, 6Ј-H) ppm. 13C NMR (75 MHz, CDCl3, 25 °C): δ =
55.8 (OCH3), 55.9 (OCH3), 64.0 (C-1), 74.6 (C-2), 83.8 (C-3), 109.8
(C-2Ј), 110.9 (C-5Ј), 119.0 (C-6Ј), 132.8 (C-1Ј), 148.9/149.0 (C-3Ј
General Procedure for the Preparation of 1-Arylprop-2-yn-1-ols 11a–
g: A solution of ethynylmagnesium bromide (0.5  in THF, 26 mL,
and C-4Ј) ppm. IR (KBr): ν = 3330 (O–H), 3300 (ϵC–H), 2110
˜
(CϵC) cm–1. MS (ESI, positive mode): m/z (%) = 215 (40) [M +
2444
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2438–2448

3-Deoxyanthocyanidins and Novel Dicationic Derivatives
FULL PAPER
Na]+, 199 (100) [M + Li]+, 175 (15). HRMS: calcd. for C11H12LiO3
199.0941; found: 199.0952.
(100) [M + Li]+. HRMS: calcd. for C15H20LiO2Si 267.1387; found:
267.1386.
1-(4Ј-tert-Butyldimethylsilyloxyphenyl)prop-2-yn-1-ol (11e): Yellow
1
oil. Yield 94%. H NMR (300 MHz, CDCl3, 25 °C): δ = 0.20 [s, 6
3
H, Si(CH3)2], 0.98 [s, 9 H, SiC(CH3)3], 2.24 (d, JOH–1 = 6.2 Hz, 1
4
3
H, OH), 2.66 (d, J3–1 = 2.2 Hz, 1 H, 3-H), 5.41 (dd, J1–OH
=
4
6.2 Hz, J1–3 = 2.2 Hz, 1 H, 1-H), 6.84 (m, AAЈ part of an
AAЈMMЈ system, 2 H, 3Ј-H and 5Ј-H), 7.42 (m, MMЈ part of an
AAЈMMЈ system, 2 H, 2Ј-H and 6Ј-H) ppm. 13C NMR (75 MHz,
CDCl3, 25 °C): δ = 4.4, 18.2, 25.7, 64.1 (C-1), 74.7 (C-2), 83.6 (C-
3), 120.2 (C-3Ј and C-5Ј), 128.0 (C-2Ј et C-6Ј), 132.9 (C-1Ј), 156.0
7-Hydroxy-2-phenylbenzopyrylium Hexafluorophosphate (13a): An
excess of aqueous hexafluorophosphoric acid (50% in water) was
added to a solution of 9d (55 mg, 0.5 mmol) and 12a (65 mg,
0.5 mmol) in a minimum of acetic acid. The mixture immediately
became dark yellow and was stirred for 48 h at room temperature.
The mixture was then plunged into diethyl ether (10 mL) and the
flavylium salt precipitated. The resulting yellow powder was reco-
vered by filtration and washed with diethyl ether to give 13a in
(C-4Ј) ppm. IR (KBr): ν = 3330 (O–H), 3310 (ϵC–H), 2120
˜
(CϵC) cm–1. MS (ESI, positive mode): m/z (%) = 269 (100) [M +
Li]+, 245 (35). HRMS: calcd. for C15H22LiO2Si 269.1544; found:
269.1534.
pure form (103 mg, 0.28 mmol). Yield 56%. M.p. 117 °C. 1H NMR
3
(300 MHz, CD3CN/1% [D1]TFA, 25 °C): δ = 7.50 (dd, J6–5
=
=
4
4
5
9.1 Hz, J6–8 = 2.2 Hz, 1 H, 6-H), 7.58 (dd, J8–6 = 2.2 Hz, J8–4
0.7 Hz, 1 H, 8-H), 7.71 (m, AAЈ part of an AAЈBMMЈ system, 2
H, 3Ј-H and 5Ј-H), 7.83 (m, B part of an AAЈBMMЈ system, 1 H,
3
3
4Ј-H), 8.19 (d, J5–6 = 9.1 Hz, 1 H, 5-H), 8.33 (d, J3–4 = 8.4 Hz, 1
H, 3-H), 8.39 (m, MMЈ part of an AAЈBMMЈ system, 2 H, 2Ј-H
and 6Ј-H), 9.17 (dd, 3J3–4 = 8.4 Hz, 5J4–8 = 0.7 Hz, 1 H, 4-H) ppm.
13C NMR (75 MHz, CD3CN/1% [D1]TFA, 25 °C): δ = 102.6 (C-
8), 113.4 (C-3), 120.4 (C-10), 122.3 (C-6), 128.9 (C-1Ј), 129.1 (C-3Ј
and C-5Ј), 129.9 (C-2Ј and C-6Ј), 133.1 (C-5), 136.2 (C-4Ј), 155.5
General Procedure for the Preparation of Aryl Ethynyl Ketones 12a–
g: IBX (2.8 g, 10 mmol, 2 equiv.) was added in one portion to a
solution of the corresponding 1-arylprop-2-yn-1-ol 11a–g (5 mmol,
1 equiv.) in ethyl acetate (25 mL). The mixture was heated at 80 °C
and stirred overnight. After recooling to room temperature, the
mixture was filtered and ethyl acetate was evaporated under vac-
uum. The resulting crude product was purified by column
chromatography (heptane/ethyl acetate) to give 12a–g in pure form.
(C-4), 159.9/169.2/172.4 (C-2/C-7/C-9) ppm. IR (KBr): ν = 3360
˜
(O–H), 1635 (C=O), 860 (P–F) cm–1. UV/Vis (MeOH/5% 1 
HCl): λmax (ε)
= 264 (15600), 356 (4700), 436 (21100) nm
(–1·cm–1). MS (ESI, positive mode): m/z (%) = 223 (100) [M]+.
Phenyl Ethynyl Ketone (12a): Yellow solid. Yield 94%. M.p. 55 °C.
1H NMR (300 MHz, CDCl3, 25 °C): δ = 3.47 (s, 1 H, 3-H), 7.47
(m, AAЈ part of a AAЈBMMЈ system, 2 H, 3Ј-H and 5Ј-H), 7.59
(m, MMЈ part of a AAЈBMMЈ system, 1 H, 4Ј-H), 8.13 (m, AAЈ
part of a AAЈBMMЈ system, 2 H, 2Ј-H and 6Ј-H) ppm. 13C NMR
(75 MHz, CDCl3, 25 °C): δ = 80.2 (C-2), 81.0 (C-3), 128.7/129.7
(C-2Ј, C-3Ј, C-5Ј and C-6Ј), 134.5 (C-4Ј), 136.1 (C-1Ј), 177,4 (C-
HRMS: calcd. for C15H11O2 223.0754; found: 223.0760.
5,7-Dihydroxy-2-phenylbenzopyrylium Hexafluorophosphate (14):
An excess of aqueous hexafluorophosphoric acid (50% in water)
was added to a solution of 9e (dihydrated, 811 mg, 5 mmol) and
12a (651 mg, 5 mmol) in a minimum of acetic acid. The mixture
immediately became dark red and was stirred for 48 h at room
temperature. The mixture was then plunged into diethyl ether
(100 mL) and the flavylium salt precipitated. The resulting orange
powder was recovered by filtration and washed with diethyl ether
1) ppm. IR (KBr):
ν = 3230 (ϵC–H), 2090 (CϵC), 1645
˜
(C=O) cm–1. UV/Vis (MeOH): λmax (ε)
= 264 (18900) nm
(–1·cm–1). MS (ESI, positive mode): m/z (%) = 137 (100) [M +
Li]+. HRMS: calcd. for C9H6LiO 137.0573; found: 137.0568.
to give 14 in pure form (1.748 g, 0.455 mmol). Yield 91%. 1H NMR
4
3,4-Dimethoxyphenyl Ethynyl Ketone (12c): White solid. Yield 98%.
(300 MHz, CD3CN/1% [D1]TFA, 25 °C): δ = 6.81 (d, J6–8
=
M.p. 118 °C. 1H NMR (300 MHz, CDCl3, 25 °C): δ = 3.38 (s, 1 H,
4
5
2.2 Hz, 1 H, 6-H), 7.11 (dd, J8–6 = 2.2 Hz, J8–4 = 0.7 Hz, 1 H, 8-
3
3-H), 3.93 (s, 3 H, OCH3), 3.96 (s, 3 H, OCH3), 6.92 (d, J5Ј–6Ј
=
H), 7.72 (AAЈ part of an AAЈBMMЈ system, 2 H, 3Ј-H and 5Ј-H),
4
3
8.4 Hz, 1 H, 5Ј-H), 7.59 (d, J2Ј–6Ј = 2.2 Hz, 1 H, 2Ј-H), 7.86 (dd,
7.84 (B part of an AAЈBMMЈ system, 1 H, 4Ј-H), 8.13 (d, J3–4
=
3J6Ј–5Ј = 8.4 Hz, J6Ј–2Ј = 2.2 Hz, 1 H, 6Ј-H) ppm. 13C NMR
4
8.4 Hz, 1 H, 3-H), 8.35 (MMЈ part of an AAЈBMMЈ system, 2 H,
3
5
(75 MHz, CDCl3, 25 °C): δ = 56.0 (OCH3), 56.2 (OCH3), 80.0 (C-
2), 80.3 (C-3), 110.1 (C-2Ј and C-5Ј), 126.1 (C-6Ј), 129.7 (C-1Ј),
2Ј-H and 6Ј-H), 9.25 (dd, J4–3 = 8.4 Hz, J4–8 = 0.7 Hz, 1 H, 4-
H) ppm. 13C NMR (75 MHz, CD3CN/1% [D1]TFA, 25 °C): δ =
95.7 (C-6), 102.9 (C-8), 111.2 (C-3), 114.4 (C-10), 128.9 (C-3Ј and
C-5Ј), 129.1 (C-1Ј), 130.0 (C-2Ј and C-6Ј), 135.8 (C-4Ј), 150.5 (C-
4), 158.4/159.4/171.1/171.4 (C-2, C-5, C-7 and C-9) ppm. IR (KBr):
149.1/154.7 (C-3Ј and C-4Ј), 176,0 (C-1) ppm. IR (KBr): ν = 3235
˜
(ϵC–H), 2090 (CϵC), 1635 (C=O) cm–1. UV/Vis (MeOH): λmax
(ε) = 294 (7600), 330 (9200) nm (–1·cm–1). MS (ESI, positive
mode): m/z (%) = 197 (100) [M + Li]+, 179 (10). HRMS: calcd. for
C11H10LiO3 197.0785; found: 197.0782.
ν = 3410 (O–H), 1640 (C=O), 835 (P–F) cm–1. UV/Vis (MeOH/5%
˜
1  HCl): λmax (ε) = 274 (35400), 474 (36200) nm (–1·cm–1). MS
(ESI, positive mode): m/z (%) = 239 (100) [M+]. HRMS: calcd. for
C15H11O3 239.0703; found: 239.0696.
4-(tert-Butyldimethylsilyloxy)phenyl Ethynyl Ketone (12e): White
1
solid. Yield 94%. H NMR (300 MHz, CDCl3, 25 °C): δ = 0.24 [s,
6 H, Si(CH3)2], 0.93 [s, 9 H, SiC(CH3)3], 3.37 (s, 1 H, 3-H), 6.90
(m, AAЈ part of an AAЈMMЈ system, 2 H, 3Ј-H and 5Ј-H), 8.08
5,7,4Ј-Trimethoxy-2-phenylbenzopyrylium
(15): An excess of aqueous hexafluorophosphoric acid (50% in
Hexafluorophosphate
(m, MMЈ part of an AAЈMMЈ system, 2 H, 2Ј-H and 6Ј-H) ppm. water) was added to a solution of 9f (771 mg, 5 mmol) and 12b
13C NMR (75 MHz, CDCl3, 25 °C): δ = –4.3, 18.2, 25.5, 80.0 (C-
(801 mg, 5 mmol) in a minimum of acetic acid. The mixture imme-
2), 80.4 (C-3), 120.1 (C-3Ј and C-5Ј), 130.0 (C-4Ј), 132.1 (C-2Ј and diately became dark red and was stirred for 48 h at room tempera-
C-6Ј), 161.8 (C-1Ј), 176.0 (C-1) ppm. IR (KBr): ν = 3210 (ϵC–H), ture. The mixture was then plunged into diethyl ether (100 mL) and
˜
2090 (CϵC), 1630 (C=O) cm–1. UV/Vis (MeOH): λmax (ε) = 300
(17800) nm (–1·cm–1). MS (ESI, positive mode): m/z (%) = 267
the flavylium salt precipitated. The resulting purple powder was
recovered by filtration and washed with diethyl ether to give 15 in
Eur. J. Org. Chem. 2007, 2438–2448
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2445

S. Chassaing, M. Kueny-Stotz, G. Isorez, R. Brouillard
FULL PAPER
pure form (2.057 g, 0.465 mmol). Yield 93%. M.p. 201 °C. 1H
IBX (7.0 g, 25 mmol, 2.5 equiv.) was added in one portion to a
solution of the purified diol (935 mg, 5 mmol) in ethyl acetate
(25 mL). The mixture was heated at 80 °C and stirred overnight.
NMR (300 MHz, CD3CN/1% [D1]TFA, 25 °C): δ = 3.96 (s, 3 H,
4
OCH3), 4.07 (s, 3 H, OCH3), 4.09 (s, 3 H, OCH3), 6.80 (d, J6–8
=
2.2 Hz, 1 H, 6-H), 7.19 (m, AAЈ part of an AAЈMMЈ system, 2 H, After recooling to room temperature, the mixture was filtered and
4
5
3Ј-H and 5Ј-H), 7.23 (dd, J8–6 = 2.2 Hz, J8–4 = 0.7 Hz, 1 H, 8-
ethyl acetate was evaporated under vacuum. The resulting crude
product was purified by column chromatography (heptane/ethyl
3
H), 8.05 (d, J3–4 = 8.8 Hz, 1 H, 3-H), 8.32 (m, MMЈ part of an
3
AAЈMMЈ system, 2 H, 2Ј-H and 6Ј-H), 9.07 (dd, J4–3 = 8.8 Hz,
acetate, 4:1) to give 26 in pure form as a yellow solid (756 mg,
5J4–8 = 0.7 Hz, 1 H, 4-H) ppm. 13C NMR (75 MHz, CD3CN/1% 4.2 mmol). Yield 84%. M.p. 213 °C. H NMR (300 MHz, CDCl3,
1
[D1]TFA, 25 °C): δ = 55.9/57.0/57.2 (OCH3), 93.4 (C-6), 99.8 (C- 25 °C): δ = 3.55 (s, 2 H, 5-H), 8.27 (s, 4 H, 2-H) ppm. 13C NMR
8), 111.3 (C-3), 113.4 (C-10), 115.8 (C-3Ј and C-5Ј), 120.8 (C-1Ј),
131.9 (C-2Ј and C-6Ј); 148.7 (C-4), 158.8/159.0/167.0/171.4/171.7
(75 MHz, CDCl3, 25 °C): δ = 80.0 (C-4), 82.2 (C-5), 129.9 (C-1),
140.0 (C-2), 176.5 (C-3) ppm. IR (KBr): ν = 3220 (ϵC–H), 2090
˜
(C-2, C-5, C-7, C-9 and C-4Ј) ppm. IR (KBr): ν = ca. 1640 (C=O),
(CϵC), 1655 (C=O) cm–1. UV/Vis (MeOH): λmax (ε) = 280
(25100) nm (–1·cm–1). MS (ESI, positive mode): m/z (%) = 183
˜
835 (P–F) cm–1. UV/Vis (MeOH/5% 1  HCl): λmax (ε) = 266
(24900), 394 (18000), 460 (13600) nm (–1·cm–1). MS (ESI, positive (100) [M + H]+, 149 (15). HRMS: calcd. for C12H7O2: 183.0441;
mode): m/z (%) = 297 (100) [M]+. HRMS: calcd. for C18H17O4 found: 183.0447.
297.1121; found: 297.1109.
6,7-Dihydroxy-2-phenylbenzopyrylium Hexafluorophosphate (19):
An excess of aqueous hexafluorophosphoric acid (50% in water)
was added to a solution of 9g (48 mg, 0.38 mmol) and 12a (50 mg,
0.38 mmol) in a minimum of acetic acid. The mixture immediately
became dark red and was stirred for 48 h at room temperature.
The mixture was then plunged into diethyl ether (10 mL) and the
flavylium salt precipitated. The resulting red powder was recovered
by filtration and washed with diethyl ether to give 19 in pure form
(125 mg, 0.327 mmol). Yield 86%. 1H NMR (300 MHz, CD3CN/
1% [D1]TFA, 25 °C): δ = 6.81 (s, 1 H, 5-H), 7.67 (d, 5J8–4 = 0.7 Hz,
1 H, 8-H), 7.73 (AAЈ part of an AAЈBMM’ system, 2 H, 3Ј-H and
5Ј-H), 7.81 (B part of an AAЈBMM’ system, 1 H, 4Ј-H), 8.31 (d,
3J3–4 = 8,1 Hz, 1 H, 3-H), 8.34 (MMЈ part of an AAЈBMMЈ sys-
m-Dipropynoylbenzene (27): A solution of ethynylmagnesium bro-
mide (0.5  in THF, 40 mL, 20 mmol, 2.7 equiv.) was added at 0 °C
to
a solution of isophthalic dicarboxaldehyde 25 (1.006 g,
7.5 mmol) in THF (75 mL). The solution was stirred at room tem-
perature for 4 h, a saturated solution of NH4Cl (40 mL) was added,
and THF was evaporated under vacuum. The aqueous phase was
extracted three times with ethyl acetate and the organic layers were
washed with water and brine and then dried with Na2SO4. After
evaporation of the solvent, the resulting crude product was purified
by column chromatography (heptane/ethyl acetate 2:1) to give the
expected diol in pure form as a white solid (1.103 g, 5.9 mmol).
3
5
tem, 2 H, 2Ј-H and 6Ј-H), 9.04 (dd, J4–3 = 8.1 Hz, J4–8 = 0.7 Hz,
1 H, 4-H) ppm. 13C NMR (75 MHz, CD3CN/1% [D1]TFA, 25 °C):
δ = 102.9 (8-C), 110.4 (5-C), 113.7 (3-C), 121.8 (10-C), 128.5 (3Ј-C
and 5Ј-C), 129.2 (1Ј-C), 129.9 (2Ј-C and 6Ј-C), 135.4 (4Ј-C), 148.7
(6-C), 152.7 (4-C), 155.7/159.9/169.7 (2-C/7-C/9-C) ppm. IR (KBr):
1
4
Yield 79%. H NMR (300 MHz, CDCl3, 25 °C): δ = 2.64 (d, J7–5
4
= 2.3 Hz, 2 H, 7-H), 3.57 (s, 2 H, OH), 5.38 (d, J5–7 = 2.3 Hz, 2
H, 5-H), 7.31–7.36 (m, 1 H, 4-H), 7.44–7.49 (m, 2 H, 3-H), 7.61–
7.63 (m, 1 H, 1-H) ppm. 13C NMR (75 MHz, CDCl3, 25 °C): δ =
64.0 (C-5), 75.1 (C-7), 83.4 (C-6), 125.0 (C-1), 126.8 (C-3), 129.0
ν = 3410 (O–H), 1640 (C=O), 835 (P–F) cm–1. UV/Vis (MeOH/5%
˜
1  HCl): λmax (ε) = 282 (14600), 440 (23500) nm (–1·cm–1). MS
(ESI, positive mode): m/z (%) = 239 (100) [M]+. HRMS: calcd. for
C15H11O3: 239.0703; found: 239.0717.
(C-4), 140.5 (C-2) ppm. IR (KBr): ν = 3350 (O–H), 3265 (ϵC–H),
˜
2120 (CϵC) cm–1. MS (ESI, positive mode): m/z (%) = 193 (100)
[M + Li]+. HRMS: calcd. for C12H10LiO2: 193.0836; found:
193.0802.
IBX (7.0 g, 25 mmol, 2.5 equiv.) was added in one portion to a
solution of the purified diol (935 mg, 5 mmol) in ethyl acetate
(25 mL). The mixture was heated at 80 °C and stirred overnight.
After recooling to room temperature, the mixture was filtered and
ethyl acetate was evaporated under vacuum. The resulting crude
product was purified by column chromatography (heptane/ethyl
acetate, 4:1) to give 27 in pure form as a yellow solid (656 mg,
p-Dipropynoylbenzene (26): A solution of ethynylmagnesium bro-
mide (0.5  in THF, 40 mL, 20 mmol, 2.7 equiv.) was added at 0 °C
to a solution of terephthalaldehyde (24, 1.006 g, 7.5 mmol) in THF
(75 mL). After the mixture had been stirred for 4 h at room tem-
perature, a saturated solution of NH4Cl (40 mL) was added and
the THF was evaporated under vacuum. The aqueous phase was
extracted three times with ethyl acetate and the organic layers were
washed with water and brine and then dried with Na2SO4. After
evaporation of the solvent, the resulting crude product was purified
by column chromatography (heptane/ethyl acetate 2:1) to give the
expected diol in pure form as a yellow oil (1.229 g, 6.6 mmol). Yield
1
3.4 mmol). Yield 67%. M.p. 114 °C. H NMR (300 MHz, CDCl3,
3
25 °C): δ = 3.56 (s, 2 H, 7-H), 7.66 (t, J4–3 = 8.0 Hz, 1 H, 4-H),
4
4
8.39 (dd, 3J3–4 = 8.0 Hz, J3–1 = 1.7 Hz, 2 H, 3-H), 8.89 (d, J1–3
=
1.7 Hz, 1 H, 1-H) ppm. 13C NMR (75 MHz, CDCl3, 25 °C): δ =
79.8 (6-C-6), 82.0 (7-C-7), 129.3/130.9 (C-1 and C-4), 134.8 (C-3),
136.6 (C-2), 176.2 (C-5) ppm. IR (KBr): ν = 3240 (ϵC–H), 2090
˜
(CϵC), 1650 (C=O) cm–1. MS (ESI, positive mode): m/z (%) = 183
(100) [M + H]+. HRMS: calcd. for C12H7O2: 183.0441; found:
183.0445.
1
88%. M.p. 105 °C. H NMR (300 MHz, CD3OD, 25 °C): δ = 2.99
4
4
(s, J5–3 = 2.3 Hz, 2 H, 5-H), 5.40 (d, J3–5 = 2.3 Hz, 2 H, 3-H),
2,2Ј-(p-Phenylene)bis(5,7-dimethoxybenzopyrylium)
Hexafluoro-
7.53 (s, 4 H, 2-H) ppm. 13C NMR (75 MHz, CD3OD, 25 °C): δ =
phosphate (28): Compound 9f (390 mg, 2.50 mmol, 2.1 equiv.) and
63.0 (C-3), 74.0 (C-5), 83.7 (C-4), 126.4 (C-1), 141.0 (C-2) ppm. IR an excess of aqueous hexafluorophosphoric acid (50% in water)
(KBr): ν = 3350 (O–H), 3280 (ϵC–H), 2120 (CϵC) cm–1. MS (ESI, were added at room temperature to a solution of 26 (220 mg,
˜
positive mode): m/z (%) = 193 (100) [M + Li]+, 175 (15). HRMS: 1.20 mmol, 1.0 equiv.) in a minimum of acetic acid. The mixture
calcd. for C12H10LiO2 193.0836; found: 193.0835.
2446
immediately became dark red and was stirred for 48 h at room
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2438–2448

3-Deoxyanthocyanidins and Novel Dicationic Derivatives
FULL PAPER
give 31 in pure form (2.139 g, 3.10 mmol). Yield 91%. 1H NMR
4
(300 MHz, CD3CN/1% [D1]TFA, 25 °C): δ = 6.87 (d, J6–8
=
4
5
2.1 Hz, 2 H, 6-H), 7.23 (dd, J8–6 = 1.8 Hz, J8–4 = 0.6 Hz, 2 H, 8-
H), 8.05 (t, J4Ј–3Ј = 8.1 Hz, 1 H, 4Ј-H), 8.32 (d, J3–4 = 8.4 Hz, 2
H, 3-H), 8.69 (dd, J3Ј–4Ј = 8.1 Hz, J3Ј–2Ј = 1.8 Hz, 2 H, 3Ј-H),
9.13 (t, J2Ј–3Ј = 1.8 Hz, 1 H, 2Ј-H), 9.38 (dd, J4–3 = 8.5 Hz, J4–8
3
3
3
4
4
3
5
= 0.6 Hz, 2 H, 4-H) ppm. IR (KBr): ν = 3390 (O–H), 1645 (C=O),
˜
855 (P–F) cm–1. UV/Vis (MeOH/5% 1  HCl): λmax (ε) = 264
(9500), 298 (7300), 396 (10000), 476 (8800) nm (–1·cm–1). MS
(ESI, positive mode): m/z (%) = 399 (90) [M – H]+, 200 (100)
[M]2+. HRMS: calcd. for C24H16O6 [M2+]: 200.0468; found:
200.0494.
temperature. The mixture was then plunged into diethyl ether
(50 mL) and the flavylium salt precipitated. The resulting purple
powder was recovered by filtration and washed with diethyl ether
to give 28 in pure form (797 mg, 1.07 mmol). Yield 89%. 1H NMR
(300 MHz, CD3CN/1% [D1]TFA, 25 °C): δ = 4.14 (s, 6 H, OCH3),
4
4.18 (s, 6 H, OCH3), 6.93 (d, J6–8 = 2.0 Hz, 2 H, 6-H), 7.41 (d,
3
4J8–6 = 2.0 Hz, 2 H, 8-H), 8.47 (d, J3–4 = 8.5 Hz, 2 H, 3-H), 8.63
Acknowledgments
3
(s, 4 H, 2Ј-H), 9.39 (d, J4–3 = 8.5 Hz, 2 H, 4-H) ppm. 13C NMR
(75 MHz, CD3CN/1% [D1]TFA, 25 °C): δ = 58.8 (OCH3), 59.1
(OCH3), 95.0 (C-6), 102.2 (C-8), 114.5 (C-3), 121.5 (C-10), 131.0
(C-2Ј), 135.5 (C-1Ј), 152.1 (C-4), 160.8/161.8/169.9 (C-5, C-7 and
We thank the Centre National de la Recherche Scientifique
(CNRS) and the Ministère de l’Education Nationale de la Recher-
che et de la Technologie (MNERT) for funding. Three of us (S. C.,
M. K. and G. I.) thank the MNERT for doctoral fellowships.
C-9), 175.4 (2-C) ppm. IR (KBr): ν = 1643 (C=O), 855 (P–F) cm–1.
˜
UV/Vis (MeOH/5% 1  HCl): λmax (ε) = 370 (7000), 420 (11200),
502 (16000) nm (–1·cm–1). MS (ESI, positive mode): m/z (%) = 228
(100) [M]2+. HRMS: calcd. for C28H24O6 [M]2+: 228.0781; found:
228.0777.
[1] a) T. Goto, T. Kondo, Angew. Chem. Int. Ed. Engl. 1991, 30,
17–33; b) R. Brouillard, O. Dangles, in The Flavonoids: Ad-
vances in Research since 1986, Chapman & Hall, London, 1994,
pp. 565–588.
[2] a) C. F. Timberlake, P. Bridle, in The Flavonoids, Academic
Press, New York, 1975, pp. 214–266; b) G. Hrazdina, in The
Flavonoids: Advances in Research, Chapman & Hall, London,
1982, pp. 135–188; c) J. B. Harborne, R. J. Grayer, in The Fla-
vonoids: Advances in Research since 1980, Chapman & Hall,
London, 1988, pp. 1–20; d) D. Strack, V. Wray, in The Flavono-
ids: Advances in Research since 1986, Chapman & Hall, Lon-
don, 1994, pp. 1–22; e) J. B. Harborne, C. A. Williams, Nat.
Prod. Rep. 1998, 15, 631–652; f) J. B. Harborne, C. A. Williams,
Nat. Prod. Rep. 2001, 18, 310–333; g) C. A. Williams, R. J.
Grayer, Nat. Prod. Rep. 2004, 21, 539–573; h) O. M. Andersen,
M. Jordheim, in Flavonoids: Chemistry, Biochemistry and Appli-
cations, CRC Press, Boca Raton, FL, USA 2006, pp. 471–551.
[3] E. Gonzalez, A. Fougerousse, R. Brouillard, Phytochemistry
2001, 58, 1257–1262.
[4] a) P. Coggon, G. A. Moss, H. N. Graham, G. W. Sanderson, J.
Agric. Food Chem. 1973, 21, 727–733; b) G. Mazza, E. Miniati,
in Anthocyanins in Fruits, Vegetables and Grains, CRC Press,
Boca Raton, FL, USA 1993, pp. 225–248.
[5] a) G. A. Iacobucci, J. G. Sweeny, Tetrahedron 1983, 39, 3005–
3038; b) O. Dangles, H. Elhajji, Helv. Chim. Acta 1994, 77,
1595–1610.
[6] a) P. Darmenton, M. Philippe, L’Oreal S. A. FR 2757383, 1996;
Chem. Abstr. 1998, 129, 140458n; b) G. Sauter, H.-J. Braun, R.
Brouillard, A. Fougerousse, C. Roehri-Stoeckel, Wella S. A.,
WO 03/000214 A1, 2003.
[7] P. Czerney, G. Graness, E. Birckner, F. Vollmer, W. Rettig, J.
Photochem. Photobiol. A 1995, 89, 31–36.
[8] a) N. J. Cherepy, G. P. Smestad, M. Grätzel, J. Z. Zhang, J.
Phys. Chem. 1997, 101, 9342–9351; b) M. Grätzel, J. Chem.
Educ. 1998, 75, 752–756.
[9] A. Roque, C. Lodeiro, F. Pina, M. Maestri, R. Ballardini, V.
Balzani, Eur. J. Org. Chem. 2002, 2699–2709.
2,2Ј-(m-Phenylene)-bis(5,7-dimethoxybenzopyrylium) Hexafluoro-
phosphate (30): Compound 9f (300 mg, 1.95 mmol, 2.1 equiv.) and
an excess of aqueous hexafluorophosphoric acid (50% in water)
were added at room temperature to a solution of 27 (170 mg,
0.93 mmol, 1.0 equiv.) in a minimum of acetic acid. The mixture
immediately became dark red and was stirred for 48 h at room
temperature. The mixture was then plunged into diethyl ether
(50 mL) and the flavylium salt precipitated. The resulting red pow-
der was recovered by filtration and washed with diethyl ether to
give 30 in pure form (594 mg, 0.79 mmol). Yield 85%. 1H NMR
(300 MHz, CD3CN/1% [D1]TFA, 25 °C): δ = 4.14 (s, 6 H, OCH3),
4
4.19 (s, 6 H, OCH3), 6.92 (d, J6–8 = 1.8 Hz, 2 H, 6-H), 7.48 (d,
4J8–6 = 1.8 Hz, 2 H, 8-H), 8.06 (t, 3J4Ј–3Ј = 8.1 Hz, 1 H, 4Ј-H), 8.46
3
3
4
(d, J3–4 = 8.4 Hz, 2 H, 3-H), 8.71 (dd, J3Ј–4Ј = 8.1 Hz, J3Ј–2Ј
=
4
1.8 Hz, 2 H, 3Ј-H), 9.21 (t, J2Ј–3Ј = 1.8 Hz, 1 H, 2Ј-H), 9.39 (d,
3J4–3 = 8.5 Hz, 2 H, 4-H) ppm. 13C NMR (75 MHz, CD3CN/1%
[D1]TFA, 25 °C): δ = 58.7 (OCH3), 59.1 (OCH3), 95.1 (6-C-6),
102.1 (8-C-8), 114.1 (3-C-3), 121.5 (10-C-10), 129.7 (2Ј-C-2Ј or 4Ј-
C-4Ј), 132.2 (1Ј-C-1Ј), 132.9 (2Ј-C-2Ј or 4Ј-C-4Ј), 135.9 (3Ј-C-3Ј),
152.3 (4-C-4), 160.8/161.7/170.4 (C-5, C-7 and C-9), 175.2 (C-
2) ppm. IR (KBr): ν = 1645 (C=O), 855 (P–F) cm–1. UV/Vis
˜
(MeOH/5% 1  HCl): λmax (ε) = 394 (18800), 468 (18000) nm
(–1·cm–1). MS (ESI, positive mode): m/z (%) = 228 (100) [M]2+
.
HRMS: calcd. for C28H24O6 [M]2+: 228.0781; found: 228.0776.
2,2Ј-(m-Phenylene)-bis(5,7-dihydroxybenzopyrylium)
Hexafluoro-
[10] A. C. Fernandes, C. C. Romao, C. P. Rosa, V. P. Vieira, A.
Lopes, P. F. Silva, A. L. Maçanita, Eur. J. Org. Chem. 2004,
4877–4883.
[11] a) M. N. Clifford, J. Sci. Food Agric. 2000, 80, 1063–1072; b)
J. M. Kong, L. S. Chia, N. K. Goh, T. F. Chia, R. Brouillard,
Phytochemistry 2003, 64, 923–933; c) D. X. Hou, Curr. Mol.
Med. 2003, 3, 149–159; d) M. A. Lila, in Plant Pigments and
their Manipulation, CRC Press, Boca Raton, FL, USA 2004,
pp. 248–274.
phosphate (31): Compound 9e (1.162 g, 7.04 mmol, 2.0 equiv.) and
an excess of aqueous hexafluorophosphoric acid (50% in water)
were added at room temperature to a solution of 27 (622 mg,
3.40 mmol, 1.0 equiv.) in a minimum of acetic acid. The mixture
immediately became dark red and was stirred for 48 h at room
temperature. The mixture was then plunged into diethyl ether
(50 mL) and the flavylium salt precipitated. The resulting red pow-
der was recovered by filtration and washed with diethyl ether to
Eur. J. Org. Chem. 2007, 2438–2448
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2447

S. Chassaing, M. Kueny-Stotz, G. Isorez, R. Brouillard
FULL PAPER
[12] a) R. Brouillard, J. E. Dubois, J. Am. Chem. Soc. 1977, 99,
1359–1364; b) R. Brouillard, B. Delaporte, J. Am. Chem. Soc.
1977, 99, 8461–8468; c) R. Brouillard, in Anthocyanins as Food
Colors, Academic Press, New York, 1982, pp. 1–40; d) R. Brou-
illard, Phytochemistry 1983, 22, 1311–1323.
[13] O. Dangles, G. Fargeix, C. Dufour, J. Chem. Soc., Perkin Trans.
2 2000, 1653–1663.
N. Koga, H. Iwamura, J. Am. Chem. Soc. 1995, 117, 5550–
5560; c) L. De Luca, G. Giacomelli, A. Porcheddu, J. Org.
Chem. 2001, 66, 7907–7907; d) Y. Maeda, N. Kakiushi, S. Mat-
sumura, T. Nishimura, S. Uemura, Tetrahedron Lett. 2001, 42,
8877–8879; e) Y. Maeda, N. Kakiushi, S. Matsumura, T. Nishi-
mura, T. Kawamura, S. Uemura, J. Org. Chem. 2002, 67, 6718–
6724; f) L. Yu, M. L. Trudell, Tetrahedron Lett. 2003, 44, 2553–
2555; g) Y. Maeda, Y. Washitake, T. Nishimura, K. Iwai, T.
Yamauchi, S. Uemura, Tetrahedron 2004, 60, 9031–9036; h) M.
Hunsen, Tetrahedron Lett. 2005, 46, 1651–1653.
[14] a) D. Pratt, R. Robinson, J. Chem. Soc. 1922, 1577–1585; b)
D. D. Pratt, R. Robinson, J. Chem. Soc. 1923, 745–758; c) D.
Pratt, R. Robinson, J. Chem. Soc. 1924, 188–199; d) D. D.
Pratt, R. Robinson, J. Chem. Soc. 1924, 199–207; e) D. D.
Pratt, R. Robinson, J. Chem. Soc. 1925, 1128–1138.
[15] N. Kuhnert, M. N. Clifford, A.-G. Radenac, Tetrahedron Lett.
2001, 42, 9261–9263.
[20] J. D. More, N. S. Finney, Org. Lett. 2002, 4, 3001–3003.
[21] a) A. R. Katritzky, P. Czerney, J. R. Level, W. Du, Eur. J. Org.
Chem. 2001, 4451–4456; b) C. Fichtner, G. Remennikov, H.
Mayr, Eur. J. Org. Chem. 2001, 4451–4456.
[22] K. M. Doxsee, M. Feigel, D. S. Kent, J. W. Canary, C. B.
Knobler, D. Cram, J. Am. Chem. Soc. 1987, 109, 3098–3107.
[23] G. A. Reynolds, J. A. Van Allan, J. Heterocycl. Chem. 1969, 6,
623–626.
[16] T. Mas, Synthesis 2003, 12, 1878–1880.
[17] a) J. G. Sweeny, G. A. Iacobucci, Tetrahedron 1977, 33, 2923–
2926; b) J. G. Sweeny, G. A. Iacobucci, Tetrahedron 1977, 33,
2927–2932.
[18] a) A. W. Johnson, R. R. Melhuish, J. Chem. Soc. 1947, 346–
349; b) J. W. Gramshaw, A. W. Johnson, T. J. King, J. Chem.
Soc. 1958, 4040–4049.
Received: December 9, 2006
Published Online: April 4, 2007
[19] a) E. V. Vedejs, J. R. Pribish, J. Org. Chem. 1988, 53, 1593–
1599; b) K. Matsuda, N. Nakamura, K. Takahashi, K. Inoue,
2448
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2438–2448