Asymmetric cyclization of 2′-hydroxychalcones to flavanones: Catalysis by chiral Bronsted acids and bases

Article abstract of DOI:10.1002/ejoc.200700682

The asymmetric cyclization of 2′-hydroxychalcones to flavanones is a basic, enzyme-catalyzed step in the biosynthesis of flavonoid natural products, but poses a long-standing problem for asymmetric catalysis with small molecule catalysts. Earlier claims concerning the realization of an asymmetric flavanone synthesis by means of camphorsulfonic acid as chiral Bronsted acid catalysts were reinvestigated using accurate HPLC methods for quantification of enantiomer ratios. The previous claims of asymmetric induction were thus shown to be untenable. On the other hand, cinchona alkaloids serve as chiral Bronsted base mediators for the asymmetric cyclization of either 6′-substituted 2′-hydroxychalcones or 2′,6′- dihydroxychalcones. 2′,6′-Dihydroxy-4,4′-dimethoxychalcone, for instance, cyclized to give the naturally occurring naringenin-4′,7- dimethyl ether in up to 64 % ee at 81 % conversion. The catalysis shows a marked dependency of the enantiomeric excess of the product on the catalyst, solvent and reactant concentration. Based on these successful examples of asymmetric cyclizations of 2′-hydroxychalcones to flavanones, requirements for more active asymmetric catalysts can be defined. Wiley-VCH Verlag GmbH & Co. KGaA, 2007.

Full text of DOI:10.1002/ejoc.200700682

FULL PAPER
DOI: 10.1002/ejoc.200700682
Asymmetric Cyclization of 2Ј-Hydroxychalcones to Flavanones: Catalysis by
Chiral Brønsted Acids and Bases
Claudia Dittmer,[a] Gerhard Raabe,[a] and Lukas Hintermann*[a]
Keywords: Asymmetric catalysis / Chiral Brønsted acid / Cyclization / Flavanones / Oxygen heterocycles
The asymmetric cyclization of 2Ј-hydroxychalcones to flav-
anones is a basic, enzyme-catalyzed step in the biosynthesis
of flavonoid natural products, but poses a long-standing
problem for asymmetric catalysis with small molecule cata-
lysts. Earlier claims concerning the realization of an asym-
metric flavanone synthesis by means of camphorsulfonic acid
as chiral Brønsted acid catalysts were reinvestigated using
accurate HPLC methods for quantification of enantiomer ra-
tios. The previous claims of asymmetric induction were thus
shown to be untenable. On the other hand, cinchona alka-
loids serve as chiral Brønsted base mediators for the asym-
cones or 2Ј,6Ј-dihydroxychalcones. 2Ј,6Ј-Dihydroxy-4,4Ј-di-
methoxychalcone, for instance, cyclized to give the naturally
occurring naringenin-4Ј,7-dimethyl ether in up to 64% ee at
81% conversion. The catalysis shows a marked dependency
of the enantiomeric excess of the product on the catalyst, sol-
vent and reactant concentration. Based on these successful
examples of asymmetric cyclizations of 2Ј-hydroxychalcones
to flavanones, requirements for more active asymmetric cata-
lysts can be defined.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
metric cyclization of either 6Ј-substituted 2Ј-hydroxychal- Germany, 2007)
1. Introduction
The cyclization of 2Ј-hydroxychalcones to flavanones
(Figure 1) is both a simple model for the addition of an
oxygen nucleophile to an alkene and a fundamental step in
the biosynthesis of flavonoids.[1] The reaction produces the
chromane core that is characteristic of many flavonoids,
which are, inter alia, responsible for the colorful appearance
of many flowers (as floral pigments), the bitter taste of
grapefruit, the color of red wine, and serve for UV and
antimicrobial protection of plants.[2] In plants and also in
fungi, molds and bacteria,[3] the reaction is catalyzed by
the enzyme chalcone isomerase (CHI; EC 5.5.1.6),[4,5] which
generates 2S-flavanones[4] with an astonishing enantio-
selectivity estimated of S/R = 100Ј000 (ee = 99.998%) from
Figure 1. Cyclization of 2Ј-hydroxychalcones to flavanones. a)
Most simple model reaction. b) Flavonoid biosynthesis.
kinetic measurements (Figure 1, b).[4d]
This oxa-Michael type cyclization takes place spontane-
ously in solution to give an equilibrium mixture of chalcone
and flavanone, but the process is typically slow under ambi-
ent and neutral conditions, especially for the nonsubstituted
2Ј-hydroxychalcone 1.[6] Kostanecki realized the transfor-
mation 1Ǟ2 (Figure 1, a) for the first time in 1904 with a
mineral acid catalyst.[7] In the meantime, a wide range of
catalytic reagents and conditions have been found to induce
the cyclization, including: aqueous buffers at variable
pH,[6,8] mineral acids,[7,9] acidic ion-exchange resin,[10] acetic
acid,[11] alkali metal hydroxide[6,8b,12] (sometimes combined
with phase-transfer catalysts),[13,14] sodium acetate,[8m,15]
potassium fluoride,[16] amine bases,[14,17,18] amino acids,[14]
photoirradiation,[19] thermal reaction at 60 °C in the solid
state[20] or in the melt at 230 °C,[21] catalysis by CoIII-salen
complexes,[22] Lewis acids,[11a,23] SiO2,[24] or under electro-
chemical conditions.[25] All of these methods have in com-
mon that they give racemic flavanone (2). There is no exam-
ple of an asymmetric cyclization. Indeed, asymmetric syn-
theses of flavanones[26,27] require multistep-sequences in-
volving chiral auxiliaries and protecting groups, which is in
stark contrast to the waste-free, simple and elegant one-step
enzymatic asymmetric synthesis mediated by CHI! Scheidt
[a] Institute of Organic Chemistry, RWTH Aachen University,
Landoltweg 1, 52074 Aachen, Germany
E-mail: lukas.hintermann@oc.rwth-aachen.de
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
5886
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 5886–5898

Catalysis by Chiral Brønsted Acids and Bases
Figure 2. Experiments of Fujise et al. concerning the asymmetric cyclization of chalcones to flavanones. Chalcone precursors of 7 and 8
were the respective chalcone triacetates.
and co-workers have recently shown that a catalytic asym- reinvestigation was deemed necessary. A second study with
metric cyclization of hydroxychalcones can be achieved in- apparent success concerns the cinchona-catalyzed cycliza-
directly by introducing an activating ester group at C-2 of tion of ortho-tigloylphenols to 2,3-dimethylchroman-4-ones
the substrate; asymmetric cyclization is then mediated by a that has been shown to proceed with up to 98% ee for the
cinchona-alkaloid-derived catalyst containing urea/tertiary cis isomer, at a diastereomeric excess of 52% (vide infra,
amine functional groups; the flavanone-3-carboxylic esters Scheme 4).[30,31] This method has not yet been applied to
were decarboxylated in acidic medium to flavanones, which substrates bearing an aryl group at C-3 of the alkenone unit
were obtained in up to 94% ee.[28] Still, this interesting pro- (i.e., chalcones); therefore, it is not clear if an asymmetric
tocol requires additional synthetic steps for substrate acti- synthesis of flavanones is possible by this approach. The
vation and product deprotection. In that sense, the funda- object of the present study was therefore to investigate the
mental problem of performing a direct asymmetric cycliza- fundamental viability of an asymmetric cyclizations of 2Ј-
tion of 2Ј-hydroxychalcones to the desired flavanones is cir- hydroxychalcones to flavanones by i. reinvestigating the re-
cumvented, and this reaction remains one of those enzyme- ported CSA-catalyzed reactions (as examples of chiral
catalyzed processes that proceedes with high enantio- Brønsted acid catalysis) and ii. investigating the catalytic
selectivity but yet cannot be imitated with a small-molecule action of chiral Brønsted bases, exemplified by the cinchona
asymmetric catalyst. Attempts towards finding such a cata- alkaloids. Along these lines, we will now show that the
lyst have been undertaken repeatedly over the years: i. - CSA-catalyzed cyclization of 2Ј-hydroxychalcones does not
Proline is a catalyst for the synthesis of flavanones from lead to enantiomerically enriched flavanones within the lim-
hydroxyacetophenones and aryl-substituted aldehydes, its of detection of HPLC method (Ͻ1% ee), and that while
where 2Ј-hydroxychalcones are intermediates.[29] No optical the alkaloid-catalyzed asymmetric cyclization of ortho-tig-
induction was observed. ii. Enantiopure amino acids were loylphenols cannot be generally transferred to 2Ј-hydroxy-
tested as catalysts for the cyclization of 2Ј-hydroxychalcones chalcones, a number of substrates have been found to un-
to flavanones in a solid-liquid interphase reaction in water, dergo asymmetric cyclizations with notable levels of induc-
with no reported induction.[14] iii. The microwave-assisted tion. Thus we present here the first cases of catalytic asym-
cyclization of 2Ј-hydroxychalcones was performed in the metric cyclizations of 2Ј-hydroxychalcones to flavanones by
presence of chiral additives such as tartaric acid or N-ben- a nonenzymatic method with small-molecule catalysts,
zyl-1-phenylethylamine, with no observed enantio- opening the door to the solution of a problem in asymmet-
selectivity.[18] iv. Chiral cobalt-salen complexes were found ric catalysis that had already been explicitly defined in
to catalyze the cyclization of a range of substituted 2Ј-hy- 1938.[32]
droxychalcones, giving equilibirum mixtures of chalcones
and flavanones, with no reported induction.[22] However, in
2. Results
two cases, asymmetric catalytic variants of the direct chal-
cone/flavanone cyclization have apparently been realized: A
series of papers by Fujise et al. report that camphorsulfonic
acid (CSA, 3) catalyzes the ring closure of a handful of 2Ј-
hydroxychalcones to the corresponding flavanones, which
2.1. Reinvestigation of CSA-Catalyzed Asymmetric
Cyclizations of 2Ј-Hydroxychalcones
In a series of papers between 1938 and 1951, Fujise and
were found to be optically active (Figure 2). For several coworkers had reported several examples of the cyclization
reasons, these results appeared to contradict current expec- of substituted 2Ј-hydroxychalcones (or acetylated deriva-
tations regarding catalysis with chiral Brønsted acids, and a tives) in the presence of (+)-10-camphorsulfonic acid (CSA,
Eur. J. Org. Chem. 2007, 5886–5898
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5887

C. Dittmer, G. Raabe, L. Hintermann
FULL PAPER
3) in substoichiometric amounts in alcoholic solution, in
which the resulting flavanones were found to be optically
active (Figure 2).[32–35] In the first reported example,
matteucinol-chalcone triacetate 4 was heated in EtOH un-
der pressure to give diacetyl-matteucinol 5 in moderated
yield with an optical activity close to that of “optically
pure” matteucinol diacetate ([α]D = +32.8), prepared from
plant-derived (–)-matteucinol (2S-6).[36]
Later, Tatsuta applied the protocol to the synthesis of
flavanones 7–9 in optically active form. Strangely, in these
experiments, fully de-acetylated flavanones were obtained
from the fully acetylated chalcones.[33] The optical induc-
tions were variable. Nevertheless, 7-hydroxyflavanone (9)
was formed with [α]D = +29.3, which implies an optical
purity of around 35%.[37] In later experiments, Fujise simi-
larly obtained flavanones 10 and 11 with rather small op-
tical rotations.[34,35] In spite of its simplicity, this method
has not found any applications, and has in fact hardly been
recognized.[38,39] The topic of asymmetric catalysis with chi-
ral Brønsted acids has remained little explored for a long
time,[40] though the situation has dramatically changed in
recent years[41] with the successful application of binaph-
thol-derived phosphoric acids.[42] Camphorsulfonic acid 3,
a readily available chiral acid, has not been reported to give
high inductions in catalysis,[40] which is probably related to
its conformational flexibility and a limited abililty for par-
ticipating in secondary interactions with substrates. The
success of CSA (3) as catalyst of the chalcone-flavanone
cyclization is therefore surprising, especially because the re-
actions were performed in ethanol, whereas modern exam-
ples of asymmetric Brønsted acid catalysis almost invariably
require apolar solvents which will not rupture weak hydro-
gen bonding interactions between catalyst and sub-
strate.[41,42] We therefore deemed a reinvestigation of the re-
action necessary, applying current knowledge and modern
methods for enantiomer analysis. Chalcone 1 was heated to
120 °C in the presence of 20 mol-% of 3 in n-butanol. These
conditions are intended to simulate those from the early
experiments (EtOH, 115–120 °C, pressure tube), but make
sampling for analysis easier. The course of the reaction was
followed by chiral HPLC and progress curves were fitted to
a pseudo first order rate law for the forward and backward
reactions (d[1]/dt = –k[1] + kЈ[2]; Keq = k/kЈ). As is evident
from Figure 3, the reaction progress was consistent with
this rate law.[43] From the kinetic analysis, an equilibrium
constant of Keq = 3.82Ϯ0.07 is deduced, corresponding to
an equilibrium composition of 21% of hydroxychalcone (1)
and 79% of flavanone (2).[44] Over the whole range of the
reaction, no significant (Ն 0.7%) enantiomeric excess was
detected (see part a of Figure 4 for a typical HPLC trace of
the reaction mixture). A blank reaction in the absence of
Figure 3. Reaction progress curves for cyclization of 1 to 2 in
nBuOH at 120 °C. Solid lines: amount of 1 (%), dashed lines: con-
version to 2 (%). a) Filled symbols: reaction in the presence of
20 mol-% of CSA. b) Open symbols: reaction without added cata-
lyst (background reaction).
Figure 4. HPLC traces of reaction mixtures from CSA-catalyzed
cyclizations of 2Ј-hydroxychalcones to flavanones. a) Cyclization of
1 to 2. Chiralcel OD, n-heptane/iPrOH, 97:3, 1 mL/min, λ =
254 nm. Peaks for 1 (integral 36.29%), 2 (integral 31.86%), ent-2
(integral 31.85%). b) Cyclization of 12 to 9. Chiralcel OJ, n-hep-
tane/EtOH, 95:5, 1.2 mL/min, λ = 277 nm; peaks for 9 (integral
40.24%), ent-9 (integral 40.62%), 12 (integral 19.13%). c) Cycliza-
tion of 13 to 10. Chiralcel OD, n-heptane/iPrOH = 95:5, 1 mL/min,
λ = 254 nm, peaks for 13 (integral 36.03%), 10 (integral 31.93%),
ent-10 (integral 32.04%).
Next, the reaction solution was left standing for about
CSA was also performed. Surprisingly, it turns out that the two hours at 4 °C. Flavanone 2 crystallized directly from
CSA-catalyzed reaction was only faster by a factor of 1.6 the reaction solution, but neither crystals nor the remaining
(based on pseudo first-order rate constants); consequently, material in the mother liquor displayed an enantiomeric ex-
the amount of flavanone formed by CSA catalysis is small cess. In another experiment, pure racemic flavanone was
(0.6/1.6 ≈ 38%) relative to the product formed by the “non- heated to 120 °C in nBuOH in the presence of CSA, and,
catalyzed” pathway. Any potential induction due to asym- after formation of a significant amount of 2Ј-hydroxychal-
metric catalysis will be small from the start.
cone 1, the reaction mixture was analyzed by HPLC and
5888
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 5886–5898

Catalysis by Chiral Brønsted Acids and Bases
the remaining flavanone 2 found to be racemic. These ex-
periments imply that there is no induction in the CSA-cata-
lyzed cyclization of 1 to 2, and that neither preferential
crystallization nor kinetic resolution in the presence of CSA
lead to an enantiomeric excess. We now turned our atten-
tion to 2Ј,4Ј-dihydroxychalcone (12), whose cyclization to
flavanone 9 had been described as being accompanied by a
significant optical induction (see above).[33] The experiment
was performed according to the literature directions by
heating chalcone 12 with (+)-CSA in EtOH for 60 h at 115–
120 °C in a pressure tube. A sample from the resulting solu-
tion was analyzed by HPLC and gave the trace shown in
Figure 4 (b). Even though the peaks are not completely
baseline separated, the method is sufficiently accurate to
exclude an ee of Ͼ2%.
Likewise, chalcone 13 was refluxed for 50 h in EtOH in
the presence of 28 mol-% of CSA according to the reported
conditions.[34] The HPLC trace of the reaction mixture is
shown in part c of Figure 4: it implies again that flavanone
10 was formed with an enantiomer excess below the limits
of detection of the analytical method (Ͻ1% ee). The failure
to observe induction in these simple experiments shed
doubts on the conclusions reached in the older literature.
Nevertheless, we felt that a definite decision as to whether
asymmetric cyclizations with CSA are fact or artefact
would rely on a reinvestigation of the original example,
namely Fujise’s 1938 cyclization of the chalcone triacetate
4 to matteucinol-diacetate 5. It was this original report, on
which later experiments were modeled, and the substrate
was a quite specific one: matteucinol (6; Figure 2), isolated
from the fern Matteuccia orientalis in 1924,[45] is the first
flavanone from plant material which was shown to be op-
tically active.[46] In the absence of plant material, matteuci-
nol is not readily available, but a total synthesis is in reach.
Syntheses of 6 have already been reported, but experimental
details are not readily available.[47] Favorable retrosynthetic
disconnections from 6 lead to dimethyl-trihydroxy-aceto-
phenone (14) and anisaldehyde (15), or alternatively to di-
methyl phloroglucinol (16) and p-methoxycinnamoyl chlo-
ride (17) (Scheme 1, a).
Scheme 1.
may be obtained either via nitration of meta-xylene, re-
duction and hydrolysis,[52] or via selective diformylation of
phloroglucinol[53,54] and Clemmensen reduction (Zn/Hg,
HCl aq.).[54,53a] We followed the second route (Scheme 2):
Reaction of phloroglucinol (18) with two equivalents of
Vilsmeyer reagent and hydrolysis produced dialdehyde 21 in
high yield. The Clemmensen reduction was modified by
using a zinc–copper pair instead of toxic zinc amalgam; this
approach was partially successful, in that the procedure fur-
nished sufficient material for our synthesis, but the yields
of the reaction were variable (Scheme 2).
It has been reported that phloroglucinol (18) can be
methylated under basic conditions to give 19 in 20% yield
after acetylative workup (Scheme 1, b), and then converted
to 14 by Fries rearrangement.[48] On repeating the methyl-
ation/acetylation of 18 we did indeed isolate a nicely crystal-
line compound in 17% yield, but this turned out to be the
cyclohexatrione enolacetate 20, as evident from spectral
analyses and an X-ray crystal structure determination (see
the electronic supporting information). We cannot readily
explain the discrepancy between our and the literature re-
sult for this reaction. Earlier studies on the methylation of
resorcinol,[49] phloroglucinol (18)[49,50] or trihydroxyaceto-
phenone[51] in basic media with methyl iodide have shown
that the nature and yields of products are critically variable
Scheme 2. Synthesis of matteucinol.
Friedel–Crafts acylation of 16 with acid chloride 17 gave
with small changes of temperature, reaction time, concen- matteucinol 6 in low, but reproducible yield (Scheme 2).
trations and other factors.[50d]
Modified conditions for the Friedel–Crafts reaction on
An alternative retrosynthetic disconnection of
6
acidic montmorillonite clay[55] returned the isomeric dihy-
(Scheme 1, a) leads to dimethyl phloroglucinol (16) which drocoumarine 22 as the major product instead. Racemic
Eur. J. Org. Chem. 2007, 5886–5898
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5889

C. Dittmer, G. Raabe, L. Hintermann
FULL PAPER
matteucinol forms yellow crystals from acetone. Its identity
was secured by literature comparison of melting point, mass
spectral and NMR spectroscopic data. Matteucinol (6) was
derivatized according to Fujise’s conditions (Scheme 3):
mild acetylation gave diacetate 23 as colorless needles,
harsh acetylation by refluxing with Ac2O/NaOAc, gave
chalcone triacetate 24.[32] The acetate 24 was carefully ana-
lyzed in order to exclude that enol ester 25 had been ob-
tained instead. This reaction mode has been observed at
least once with a similar flavanone.[56] If Fujise had mis-
taken enol ester 25 for its isomer 24, then one could easily
understand why refluxing of 25 with CSA in ethanol will
give 23 with an intact stereogenic center and optical ac-
tivity, since all of the starting materials were prepared from
optically active 6 (extracted from plants), and the sequence
6Ǟ25Ǟ23 (instead of 6Ǟ24Ǟ23) would of course lead to
an optically active product. However, our spectroscopic
data show that the acetylation product is indeed the achiral
chalcone acetate 24, and had the same melting point as re-
ported by Fujise.
Figure 5. HPLC traces of 23. a) Racemic reference. HPLC: Chi-
ralcel-OJ, EtOH/n-heptane, 1:1 (v/v), 0.7 mL/min, λ = 280 nm, α =
1.707, RS = 2.265; integrals 50.1% and 49.9%. b) Product from the
CSA-catalyzed cyclization of 24; impurity at tR = 19 min; integrals
49.9% and 48.7%.
2.2. Chiral Base-Catalyzed Approaches to Flavanone
Synthesis
Ishikawa and co-workers have recently described that
cinchona-catalyzed cyclizations of some ortho-tigloylphen-
ols (25) to 2,3-dimethylchroman-4-ones (26) proceed with
very high enantiomeric excess (Scheme 4 a).[30] The reaction
is characterized by mild conditions and a marked solvent
dependency of enantioselectivity: best results are achieved
in nonhydrogen bonding polar aromatic solvents such as
chlorobenzene. The products 26 are formed as mixtures of
cis and trans diastereomers, and the diastereoselectivity of
the reaction depends critically on the alkaloid catalyst and
the solvent, in a way that is difficult to predict. Judging
from the published examples (Scheme 4, a) it was not clear
whether the reaction had any generality beyond the specific
cases of coumarin-derived tigloyl-phenols 25. Thus we ap-
Scheme 3. Derivatization of matteucinol.
Chiral HPLC analysis of 23 was successfully carried out plied the reaction conditions to hydroxychalcone 1, which
on Chiralcel-OJ (Figure 5 a). We were now in a position to was stirred with quinine or cinchonidine at room temp. for
repeat the original asymmetric cyclization experiment: 24 two weeks in either EtOAc or CH2Cl2, without giving a
was heated in EtOH with 5 mol-% of (+)-CSA in a pressure trace of product 2. Upon stirring of 1 with 20 mol-% of
tube.[32] According to TLC analysis, matteucinol diacetate quinine in chlorobenzene (10 mg/mL) for 1 day at 80 °C, a
23 was formed as the major reaction product, which could conversion 1Ǟ2 of 0.43% (TOF = 0.00083 h–1) was mea-
be isolated in 40% yield. The results of the HPLC measure- sured by HPLC, with an ee of 2 of Յ 3%. On heating for
ments for either the reaction mixture or the purified prod- another 18 h at 110 °C, the conversion rose to 1.9% (TOF
uct show that an enantiomeric excess was not present = 0.0042 h–1) and the product was essentially racemic
within the limits of detection of the analytical method (eeϽ1.3%).[57]
(Ͻ1% ee). In contrast, Fujise et al. had isolated 23 in 36%
It is now evident that the cinchona-catalyzed reaction is
yield with an optical purity of close to 100% ([α]D = +32.1 limited in its substrate spectrum and not general with re-
vs. +32.7° for 23, which had been prepared from natural spect to the asymmetric synthesis of flavanones. Neverthe-
6).[32] According to our results, the claim of an asymmetric less, it appeared to us that the kinetic barrier in the way of
synthesis of flavanones by CSA-catalyzed cyclization of hy- catalysis might be overcome by either using a more active
droxychalcones (or chalcone acetates) cannot be upheld, catalysts, or by choosing a more reactive substrate. Both
and the optical activity earlier observed must have been due approaches eventually turned out to be successful, and here
to reasons other than an enantiomeric excess of the reaction we present the results of the second approach, where we
product.
stuck to cinchona alkaloids as catalysts, but searched for
5890
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 5886–5898

Catalysis by Chiral Brønsted Acids and Bases
are available in one step from commercial precursors.[8k]
The high tendency towards cyclization of 27–29 means that
it is best to first synthesize the corresponding flavanones
and perform base-induced ring-opening reactions followed
by an acid or TMSCl quench,[58] giving the chal-
cones.[8k,58,60] Indeed, chalcones 27–29 underwent catalytic
cyclizations in the presence of cinchona alkaloids (see
Tables 1, 2, and 3). The substrates were combined with
0.5 equiv. of alkaloid in a solvent at room temperature, and
the reaction progress was followed quantitatively by chiral
HPLC analysis. In view of the small scale of the experi-
ments and the slow kinetics, no particular attempt has been
undertaken to reduce catalyst loading. Table 1 presents re-
sults for the catalytic cyclization of chalcone 27 to naring-
enin dimethyl ether (30). The enantioselectivity of the cata-
lytic asymmetric cyclization was dependent on the nature
of the alkaloid catalyst and on the solvent. Dichlorometh-
ane as a polar, nonprotic solvent gave relatively fast reac-
tions with low enantioselectivity (entry 1). Reactions in less
polar solvents were not successful because of the insolubil-
ity of substrate and catalyst, but both ortho-dichloroben-
zene (o-C6H4Cl2) and chlorobenzene provide a suitable bal-
ance of polarity and solubility. A screening with o-C6H4Cl2
as solvent with all four cinchona alkaloids (quinine–QN;
quinidine–QD; cinchonine–CN; cinchonidine–CD) showed
that the pseudo-enantiomeric[61] pair QN/QD bearing
methoxy groups gave higher selectivity than CN/CD (en-
tries 2–5). Interestingly, CN produced the same enantiomer
in excess as did its “pseudoenantiomer” CD, though with
lower absolute selectivity and more slowly, probably due to
the low solubility of that alkaloid (entry 4). At extended
reaction times, substrate conversions reached Ͼ99.8% in
either CH2Cl2 or o-C6H4Cl2, which corresponds to
Scheme 4. Cinchona catalysis of alkenoyl-phenol cyclization. a) Re-
sults of Ishikawa and co-workers.[30] b) Unsuccessful extension to
2Ј-hydroxychalcone.
more active (i.e. destabilized) substrates in order to reduce
the energy gap between the ground and transition state.
From preparative and kinetic investigations, it is known
that 2Ј-hydroxychalcones bearing an additional C-6Ј hy-
droxy group cyclize more readily than those with a hydro-
gen at C-6Ј.[8i,8f,58,59] This reactivity difference has usually
been explained by assuming a hydrogen-bonding stabli-
zation of the flavanone product[59] or the transition-state
(Figure 6, a).[8f,8i] In addition, a range of 2Ј-hydroxychal-
cones bearing alkyl or annelated arene substituents at C-6Ј
also undergo facile cyclization,[20] sometimes during at-
tempted recrystallization.[8k] We have not found explana-
tions for the effect of the 6Ј-substituent in the literature,
but presumably there is a destabilization of the substrate by
repulsion between the 6Ј-substituent and the α-CH bond of
the alkenone portion, which disrupts the strong intramolec-
ular hydrogen bond (Figure 6, b).
K
eq Ͼ500, considerably higher than for the cyclization of 1
to 2 (Keq ≈ 4). As a consequence, the reversibility of the
reaction is kinetically unimportant and the enantiomer ex-
cess does not deteriorate at high conversions (entries 1, 8).
Additional experiments revealed an unusual dependency
of enantiomeric excess of the product on the initial concen-
tration of the chalcone substrate, which is strong in o-
C6H4Cl2 as solvent (entries 6–9), but somewhat less pro-
nounced in chlorobenzene (entries 10–13). Because the sub-
strate concentration decreases as the reaction progresses,
there is also a conversion dependency of the product enan-
tiomeric excess (entry 8). Since chlorobenzene had turned
out to induce higher enantioselectivities in the reaction (en-
tries 10–13), catalyses with all four alkaloids were repeated
in that solvent at an optimal substrate concentration of
10 gL–1 (entries 12, 14–16). Once again, the pseudo-
enantiomers CN and CD show the same absolute sense of
induction; after 11 hours of reaction time, the inductions
had in fact reached identical levels (entries 15a, 16a), but
as the reaction progressed, the value decreased faster for
CN (entry 16b). The influence of catalyst concentration was
Figure 6. a) Hydrogen-bonding stabilization of the chalcone cycli-
zation transition-state. b) Destabilizing effects of a 6Ј-substitution.
c) Examples of ground-state destabilized chalcones used in this
study.
As substrates for our studies, we chose 4,4Ј-dimethoxy- also investigated (entries 17–21) and found to be important
2Ј,6Ј-dihydroxychalcone 27, which is synthesized from nar- at high substrate to catalyst ratios (entry 17), where it led
ingenin (5,7,4Ј-trihydroxyflavanone) in two steps (see exper- to a lower enantioselectivity. The reason for the substrate-
imental),[60] and the chalcones 28 and 29 (Figure 6, c) which concentration dependency of enantiomeric excess is unclear
Eur. J. Org. Chem. 2007, 5886–5898
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5891

C. Dittmer, G. Raabe, L. Hintermann
FULL PAPER
Table 1. Asymmetric cyclization of 27 to 30 with cinchona catalysts.
Entry
Catalyst[a]
CD
Solvent
CH2Cl2
Conc. / gL–1
10
Time / h
% Conv.[b]
% ee[b][c]
1a
1b
2
3
4
5
6
7
8a
8b
8c
8d
8e
9
10
11
12
13
14a
14b
15a
15b
16a
16b
17
18
19
20
21
22
23
24
96
70
70
70
96
Ͼ99
79
23
74
69
92
92
66
90.2
97.4
99.0
99.7
81
68
82
10 (+)
9 (+)
CD
CN
QN
QD
QN
QN
QN (30%)
o-C6H4Cl2
o-C6H4Cl2
o-C6H4Cl2
o-C6H4Cl2
o-C6H4Cl2
o-C6H4Cl2
o-C6H4Cl2
10
10
10
10
1.6
3.2
5
31 (+)
11 (+)
52 (+)
47 (–)
23 (+)
34 (+)
38 (+)
37 (+)
35 (+)
34 (+)
34 (+)
53 (+)
52 (+)
59 (+)
64 (+)
61 (+)
57 (–)
58 (–)
33 (+)
13 (+)
33 (+)
30 (+)
46 (+)
58 (+)
61 (+)
61 (+)
59 (+)
19 (+)
5 (+)
70
295
170
96
192
339
465
672
70
90
90
70
90
11
70
11
70
11
70
44
44
QN
QN
QN
QN
QN
QD
o-C6H4Cl2
PhCl
PhCl
PhCl
PhCl
10
2.5
5
10
17
10
81
91
[d]
PhCl
–
78
Ͻ3
10
CN
CD
PhCl
PhCl
10
10
[d]
–
70
11
39
72
81
80
85
65
QN (13%)
QN (30%)
QN (50%)
QN (70%)
QN (100%)
QN + ROH[e]
QN + ROH[f]
PhCl
PhCl
PhCl
PhCl
PhCl
PhCl
PhCl
10
10
10
10
10
10
10
44
44
44
90
90
[a] 50 mol-%. [b] Conversion determined by chiral HPLC, see Exp. Sect. [c] Sign of optical rotation of the excess enantiomer (in acetone,
λ = 589 nm) given in parentheses; (+)-flavanones usually show 2R configuration.[62] [d] Accurate conversions could not be determined
due to inhomogeneous reaction mixtures. [e] In the presence of 2 equiv. of 2,6-dihydroxyacetophenone. [f] In the presence of 3.3 equiv. of
2,2,2-trifluoroethanol.
in the absence of mechanistic knowledge, but it is evident The solvent influence was smaller, with reactions performed
that hydrogen-bonding interactions are important: the sub- in chlorobenzene, o-C6H4Cl2 or CDCl3 giving similar re-
strate 27, the product 30 and the alkaloid catalysts all con- sults. Reactions in CH2Cl2 were faster, but less selective (en-
tain hydrogen-bond donor hydroxy groups. The presence of tries 6,7,11). An influence of substrate concentration (and
hydrogen-bond donors lowers the enantioselectivity of the therefore: conversion) on enantioselectivity was essentially
reaction, as is visible from the experiments at high catalyst absent (entries 8,9). This implies that while noncomplexed
loadings (enty 21 vs. 20) or high substrate-to-catalyst ratios OH groups disturb the reaction selectivity, the same is not
(entry 17). This was further shown by performing catalytic true for OH groups which are involved in a strong hydrogen
runs in the presence of the hydrogen-bond donors 2,6-di- bond with a carbonyl group, as is the substrates 28/29.
yhdroxyacetophenone (a substance with a similar hydroxy
substitution pattern as the substrate) or trifluoroethanol. In
both cases, the enantioselectivity of the reaction dropped
(entries 22, 23). The benzo-annelated hydroxychalcones 28
3. Discussion
and 29 also underwent asymmetric cyclizations, but slower
The 2Ј-hydroxychalcone/flavanone cyclization can be
and with lower selectivities than 27 (see Tables 2 and 3). considered a model reaction for an olefin hydroalkoxylation
Both substrates gave very similar results, therefore the fol- reaction. The enzyme chalcone isomerase is capable of cata-
lowing discussion (including Table entry numbers) is valid lyzing this particular transformation with astonishing levels
for either of them. Highest selectivities towards flavanones of enantioselectivity. Attempts towards achieving this reac-
31 and 32 were also obtained with quinine (QN) as catalyst. tion enantioselectively with a small-molecule chiral cata-
5892
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 5886–5898

Catalysis by Chiral Brønsted Acids and Bases
Table 2. Asymmetric cyclization of 28 to 31 with cinchona alkaloid lysts have been undertaken, but have not been successful,
catalysts.
unless additional π-acceptor groups were attached to the
olefin moiety.[28] In the present work, we have investigated
and analyzed two approaches towards realizing the title re-
action, one based on chiral Brønsted acid catalysis, the
other on chiral Brønsted base catalysis. Early studies be-
tween 1938 and 1951 on the cyclization of several hydroxy-
chalcones (or their acetates) catalyzed by camphorsulfonic
Entry Catalyst[a] Solvent[b]
Time / h % Conv.[c] % ee[c][d]
acid (CSA) have claimed optical induction and thus a very
simple and potentially very useful asymmetric synthesis of
flavanones.[32–35] From a current perspective the mechan-
istic basis for these results is unclear. We have now rein-
vestigated several CSA-catalyzed cyclizations of hydroxy-
chalcones to flavanones and analyzed the reaction products
by means of accurate chiral HPLC methods. Our experi-
ments on the model reaction 1Ǟ2 reveal that CSA displays
only limited catalytic activity in the cyclization reaction of
hydroxychalcones, and the background reaction in the ab-
sence of an added catalyst leading to racemic product is
quite fast. It was also found that the reaction times used in
the historical work were too long, such that the reversibility
of the chalcone/flavanone equilibrium would have led to a
marked erosion of any initial enantiomeric excess. Having
established these unfavorable preconditions, we were little
surprised to find that the flavanones obtained from CSA-
catalyzed reactions were racemic within the limits of accu-
racy of the HPLC analytical method. We went as far as
repeating the original report from 1938, relating to the
CSA-catalyzed synthesis of matteucinol chalcone triacetate
(23) to ascertain whether that specific substrate was critical
for the success of the asymmetric reaction. Along these
lines, we synthesized matteucinol (6) and repeated step by
step the original procedures,[32] without observing induction
in the CSA-catalyzed reaction. Why the discrepancy with
the historical work? We have excluded one potential expla-
nation for the generation of optically active reaction prod-
ucts, namely the intermediacy of enol ester 25 instead of
chalcone acetate 23 (see Scheme 3). Now, it can be stated
that most of the early work on asymmetric synthesis relied
on the determination of optical rotations, sometimes at
levels close to the limits of detection. Systematic errors have
plagued such measurements, and the reproducibility of
some of the early results have later been questioned.[63] In
the absence of any plausible explanation, we must therefore
assume that the original observation of optical activity in
refs.[32–35] was probably due to a systematic error, but cer-
tainly not an induction caused by the asymmetric Brønsted
acid. On the other hand, the investigation of chiral
Brønsted base-catalyzed cyclization of hydroxychalcones
has turned out to provide for the first time a solution to
the long-standing problem of imitating the asymmetric
chalcone isomerase reaction: Taking as a starting point the
work of Ishikawa and coworkers on the asymmetric cycliza-
tion of ortho-tigloylphenols to 2,3-dimethylchroman-4-
ones,[30] we find that this reaction is not general, since the
cyclization 1Ǟ2 is not catalyzed to any relevant degree.
However, by the use of ground-state destabilized 2Ј-hy-
droxychalcone starting materials, the catalytic cyclization of
1
2
3
4
5
6
7
8a
8b
9a
9b
9c
10a
10b
11a
11b
12a
12b
CD
CN
QN
QD
CD
CD
CN
QN
o-C6H4Cl2
o-C6H4Cl2
o-C6H4Cl2
o-C6H4Cl2
PhCl
CH2Cl2
CH2Cl2
PhCl
250
250
72
72
250
72
59
32
36.5
32
58.5
57
33
58
81
31
48
60
22
61
24
82
60
84
13 (+)
3 (+)
29 (+)
6.2 (–)
14 (+)
12 (+)
1.8 (–)
32 (+)
33 (+)
31 (+)
33 (+)
33 (+)
34 (+)
35 (+)
21 (+)
17 (+)[e]
11 (–)
12 (–)
72
97
162
97
162
212
27
77
27
77
97
QN
PhCl (5 gL–1
)
QN
QN
QD
CDCl3
CH2Cl2
PhCl
162
[a] 50 mol-%. [b] Substrate concentration c = 10 gL–1. [c] Deter-
mined by HPLC. [d] Sign of optical rotation of the excess enanti-
omer in acetone solution (λ = 589 nm). [e] Increasing concentration
in the course of reaction due to solvent evaporation.
Table 3. Asymmetric cyclization of 29 to 32 with cinchona alkaloid
catalysts.
Entry
Catalyst[a] Solvent[b]
Time / h % Conv.[c] % ee[c][d]
1
2
3
4
5
6
7
8a
8b
9a
9b
10a
10b
11a
11b
12a
12b
CD
CN
QN
QD
CD
CD
CN
QN
o-C6H4Cl2
o-C6H4Cl2
o-C6H4Cl2
o-C6H4Cl2
PhCl
CH2Cl2
CH2Cl2
PhCl
250
70
70
70
240
70
65
0
15 (+)
–
26
36
67
58
33
59
78
17.5
34
73
90
39
99[e]
23
62
25 (+)
8.5 (–)
19.4 (+)
16 (+)
3.4 (–)
32 (+)
32 (+)
31 (+)
33 (+)
13 (–)
13 (–)
20 (+)
16 (+)[e]
33 (+)
33 (+)
70
108
162
49
QN
QD
QN
QN
PhCl (c = 5)
PhCl
99
108
162
27
77
27
CH2Cl2
CDCl3
77
[a] 50 mol-%. [b] Concentration c = 10 gL–1. [c] Determined by
HPLC. [d] Sign of optical rotation of the excess enantiomer in ace-
tone solution (λ = 589 nm). [e] Increasing concentration in the
course of reaction due to solvent evaporation.
Eur. J. Org. Chem. 2007, 5886–5898
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5893

C. Dittmer, G. Raabe, L. Hintermann
FULL PAPER
2Ј-hydroxychalcones 27–29 to the corresponding flavanones currently in sight. However, the enzyme chalcone isomerase
30–32 has been achieved in variable selectivities up to 64% (CHI) is capable of catalyzing cyclizations of similarly non-
enantiomeric excess. Incidentally, (2S)-naringenin 4Ј,7-di- activated substrates (part b of Figure 1, R = H). In this
methyl ether (30) is a naturally occurring substance which context, it is interesting to analyze the decisive mechanistic
has been isolated many times from plants,[64] and the results factors of the enzymatic reaction, which have recently been
in Table 3 represent the first asymmetric syntheses of that deduced from X-ray structure data and kinetic analyses.[5] It
material. The ground-state properties which affect the ease has been found that the hydroxychalcone enters the enzyme
of cyclization of 2Ј-hydroxyalkenones A to chromanones B active site as the chalconate monoanion, where it binds to
(Figure 7) can now be summarized as follows: i. Aryl sub- the active site through interactions with several amino acid
stituents at C-3 (R3) stabilize the ground state by conjuga- residues. This forces upon the achiral anion a chiral confor-
tion, decreasing Keq = [B]/[A] and slowing down cyclization. mation, from which cyclization occurs to an enzyme-bound
For a series of chalcones, π- and σ-accepting substituents flavanone enolate, which is intercepted by protonation to
RArЈ increase the acceptor strength, translating into a release (2S)-flavanone. A critical factor for acceleration of
higher Keq according to a Hammett correlation.[65] ii. The the cyclization is the deprotonation of the substrate prior to
presence of a substituent R6Ј-C-6Ј destabilizes the chalcone binding to the enzyme: the deprotonation removes the hy-
starting material by repulsion with R2 (including R2 = H), drogen bridge, which stabilizes the ground state, thereby in-
increasing Keq. The repulsion between R6Ј and the carbonyl creasing both conformational freedom and nucleophilicity
group in B is less severe, because the carbonyl group tilts of the phenolate anion.[5] If these observations are trans-
out of the aromatic plane, away from R6Ј. iii. If both C-2 lated into requirements for a small-molecule catalyst for the
and C-3 in A are substituted, steric repulsion will destabilize asymmetric cyclization of nonactivated hydroxychyclones, it
A and increase Keq. If any of these substituents is an ac- follows that we need a catalyst that is sufficiently basic to
ceptor, the chalcone is electronically destabilized, increasing fully deprotonate the substrate, and which folds the achiral
Keq. iv. The presence of a 6Ј-hydroxy group facilitates cycli- chalconate anion into a chiral conformation by providing
zation (see discussion above). In addition to the traditional multi-point noncovalent interactions, such as hydrogen
explanations (flavanone stabilization/intramolecular acid bonding, arene π-stacking, and ion pairing. This line of rea-
catalysis by the hydrogen bond), we propose that the acce- soning has indeed led to the discovery of an asymmetric
lerating effect of the 6Ј-OH may also be due to ground-state catalyst for the conversion 1Ǟ2, and progress along these
destabilization, as discussed more generally for 6Ј-substi- lines will be reported in due course.
tuted chalcones above.
4. Conclusion
In conclusion, the asymmetric catalytic cyclization of 2Ј-
hydroxychalcones to flavanones using chiral Brønsted acids
and -bases has been investigated. Earlier reports on a cam-
phor sulfonic acid catalyzed asymmetric cyclizations of hy-
droxychalcones in alcoholic solvent have been disproved.
Ground-state-destabilized hydroxychalcones with a low ki-
netic barrier towards cyclization were found to undergo
asymmetric catalytic cyclizations (ee up to 64%) to flav-
anones in the presence of cinchona alkaloids, providing for
Figure 7. Structural factors in cyclization equilibriums leading to
chromanones/flavanones.
the first time an example of the direct catalytic asymmetric
2Ј-hydroxychalcone to flavanone transformation with a
small-molecule catalyst.
Based on these generalizations, an analysis of published
experimental data is possible: The ortho-tigloylphenol sub-
strates used by Ishikawa and co-workers profit from the
Experimental Section
presence of repelling substituents R2 and R3 in the sub-
strate, the presence of a substituent R6Ј, and the absence of Reagent grade solvents were used for reactions and distilled techni-
an aryl group R3, which explains the high reactivity of the
cal grade solvents for column chromatography on SiO2. Melting
points were measured in open capillaries on a metal heating block
substrates towards catalytic cyclization. Likewise, the suc-
with a digital thermometer. The following substances were ob-
tained from commercial sources: 1 (Fluka), 2 (Acros), (+)-CSA (3)
cess in the recent report on asymmetric catalytic cycliza-
tions of 2-substituted-chalcones by Scheidt and co-workers
(Fluka), rac-naringenin (Acros). p-Methoxycinnamoyl chloride
relies not only on the electronic activation provided by an
(17) was prepared by a literature procedure.[66] The chalcones and
ester group in R2 (CO2tBu), but also the repulsion (in A)
flavanones in this work are known substances which were prepared
according to literature references and/or identified by comparison
of NMR spectra and melting points to literature data: 2Ј,4Ј-dihy-
droxychalcone,[67] 7-hydroxyflavanone (9),[8j] 3,4-methylenedioxy-
between groups R2 and R3. None of the above mentioned
facilitating factors apply to the most simple transformation
1Ǟ2, for which catalysis by the weakly basic cinchona alka-
loids is not feasible, and no other asymmetric catalyst is 2Ј-hydroxychalcone,[68] 3Ј,4Ј-methylenedioxyflavanone (10),[34,9b,69]
5894
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 5886–5898

Catalysis by Chiral Brønsted Acids and Bases
5Ј,6Ј-benzo-2Ј-hydroxychalcone (28) and 5,6-benzoflavanone 31,[8k]
4-methoxy-5Ј,6Ј-benzochalcone (29) and 4Ј-methoxy-5,6-benzofla-
vanone 32,[70] naringenin 4Ј,7-dimethyl ether chalcone (27)[8i] and
naringenin 4Ј,7-dimethyl ether (30).[60,71]
orange crystalline powder (729 mg, 69%). Spectral data corre-
sponded to literature values.[8i]
5-Acetoxy-2,2,4,4,6-pentamethylcyclohex-5-en-1,3-dione (20); At-
tempted Synthesis of Dimethylphloroglucine (16): Dry phloroglu-
cinol (22.74 g, 0.16 mol) was dissolved in degassed MeOH
(150 mL). A solution of KOH (19.13 g, 85%, 0.29 mol) in a mini-
mum of water (15 mL, CAUTION, dissolves with heating!) was
added. After cooling to room temp., MeI (25 mL, 0.40 mol) was
added in small portions over 2 h. The mixture was stirred for 1 d
at room temp. Another addition of solid KOH (20.7 g, 0.31 mol,
CAUTION, see above) in small portions and of MeI (30 mL,
0.48 mol) in small portions was followed by dilution with MeOH
to a volume of 500 mL and stirring for 2 d at room temp. The
reaction was quenched by addition of NEt3 (10 mL; destroys excess
toxic MeI before workup) and the solvents evaporated to dryness
(60 °C/15 mbar). Azeotropic drying of the residue with toluene
(3ϫ50 mL) and tBuOMe (3ϫ50 mL) with evaporation after each
addition was followed by addition of tBuOMe (100 mL), pyridine
(60 mL) and acetic anhydride (60 mL). The resulting suspension
was stirred overnight at room temp. and another day at 60 °C. The
mixture was diluted with EtOAc (400 mL) and Water (400 mL), the
aqueous phase extracted with EtOAc (3ϫ100 mL) and the com-
bined organic phases washed with 2  HCl aq. (4ϫ100 mL), water
(2ϫ100 mL) and NaHCO3 aq. (4ϫ200 mL). The organic phase
was dried with Na2SO4, evaporated, and the residue submitted to
column chromatography (tBuOMe/hexanes, 1:5–1:3) to give the
least polar and main product 20 as colorless crystals (6.30 g, 17%).
X-ray quality crystals were obtained from hexane. Rf = 0.4 (SiO2;
tBuOMe/hexanes, 1:5); m.p. 91–92 °C. 1H NMR (400 MHz,
CDCl3): δ = 1.35 (s, 6 H), 1.38 (s, 6 H), 1.73 (s, 3 H), 2.31 (s, 3 H)
ppm. 13C NMR (100 MHz, CDCl3): δ = 10.7 (CH3), 20.7 (CH3),
24.1 (CH3), 24.9 (CH3), 48.5 (C), 56.8 (C), 83.3 (C), 122.8 (C),
General Procedure for the Cinchona-Catalyzed Cyclization of 2Ј-Hy-
droxychalcones to Flavanones: The 2Ј-hydroxychalcone (5 mg, ca
0.017 mmol) and the alkaloid (2.5 mg for CN and CD; 2.7 mg for
QN and QD; 0.0085 mmol, 50 mol-%) are stirred in the solvent
(0.5 or 1 mL; filtered through neutral Al2O3) at room temperature.
Conversions and ee were determined by HPLC (see the Supporting
Information for conditions) and through comparison with test-
mixtures containing known quantities of chalcone and flavanone.
The reaction was quenched by addition of a drop of HOAc, evapo-
rated in vacuo and the residue separated by column chromatog-
raphy (SiO2, EtOAc/hexanes or toluene). Products were identified
by comparison of their NMR spectroscopic data to literature val-
ues.
NMR Spectroscopic Data not Reported in the Literature
Compound 29: 1H NMR (400 MHz, CDCl3): δ = 3.85 (s, 3 H,
OMe), 6.93 (d, J = 8.8 Hz, 2 H Arl), 7.18 (d, J = 9.8 Hz, 1 H),
7.36–7.43 (m, 1 H), 7.37 (d, J = 16.2 Hz, 1 H), 7.49–7.61 (m, 3 H),
7.80 (d, J = 8.1 Hz, 1 H), 7.69–7.93 (m, 2 H), 8.07 (d, J = 9.4 Hz,
1 H), 12.57 (s, 1 H, OH) ppm. 13C NMR (100 MHz, CDCl3): δ =
55.5 (CH3), 114.5 (CH), 116.0 (C), 119.4 (CH), 123.8 (CH), 124.7
(CH), 125.1 (CH), 127.5 (C), 127.7 (CH), 128.6 (C), 129.2 (CH),
130.5 (CH), 131.5 (C), 136.4 (CH), 143.0 (CH), 161.8 (C), 162.4
(C), 194.4 (C) ppm.
Compound 31: 1H NMR (400 MHz, CDCl3): δ = 2.97 (dd, J = 16.5,
3.1 Hz, 1 H), 3.22 (dd, J = 16.5, 13.8 Hz, 1 H), 5.59 (dd, J = 13.8,
3.0 Hz, 1 H), 7.17 (d, J = 9.0 Hz, 1 H), 7.36–7.53 (m, 6 H), 7.65
(ddd, J = 8.7, 6.9, 1.4 Hz, 1 H), 7.76 (br. d, J = 8.1 Hz, 1 H), 7.94
(d, J = 8.9 Hz, 1 H), 9.49 (d, J = 8.7 Hz, 1 H) ppm. 13C NMR
(100 MHz, CDCl3): δ = 45.8 (CH2), 79.7 (CH), 112.6 (C), 118.8
(CH), 124.9 (CH), 125.9 (CH), 126.2 (CH), 128.4 (CH), 128.8
(CH), 128.8 (CH), 129.2 (C), 129.7 (CH), 131.4 (C), 137.6 (CH),
138.5 (C), 163.6 (C), 192.9 (C) ppm.
163.4 (C), 166.5 (C), 199.3 (C), 211.3 (C) ppm. IR (KBr): ν = 2995
˜
(m), 2942 (m), 2878 (w), 1765 (s), 1724 (m), 1674 (m), 1468 (m),
1376 (m), 1331 (m), 1184 (s), 1099 (s), 1007 (m), 876 (m), 581 (m)
cm–1. MS (EI): m/z (%) = 238 (30) [M]+, 196 (31), 178 (25), 168
(22), 126 (100). C13H18O4 (238.28): calcd. C 65.53, H 7.61; found
C 65.53, H 7.82.
Compound 32: 1H NMR (400 MHz, CDCl3): δ = 2.93 (dd, J = 16.5,
3.0 Hz, 1 H), 3.23 (dd, J = 16.5, 13.8 Hz, 1 H), 3.83 (s, 3 H, OMe),
5.52 (dd, J = 13.8, 3.0 Hz, 1 H), 6.94–7.00 (m, 2 H Arl), 7.14 (d, J
= 9.0 Hz, 1 H Arl), 7.39–7.49 (m, 3 H Arl), 7.64 (ddd, J = 8.7, 7.0,
1.5 Hz, 1 H Arl), 7.73–7.78 (m, 1 H Arl), 7.92 (d, J = 8.9 Hz, 1 H
Arl), 9.48 (d, J = 8.6 Hz, 1 H Arl) ppm. 13C NMR (100 MHz,
CDCl3): δ = 45.6 (CH2), 55.5 (CH3), 79.4 (CH), 112.5 (C), 114.2
(CH), 118.9 (CH), 124.8 (CH), 125.8 (CH), 127.8 (CH), 128.3
(CH), 129.2 (C), 129.6 (CH), 130.5 (C), 131.4 (C), 137.5 (C), 159.9
(C), 163.7 (C), 193.1 (C) ppm.
X-ray Crystal Structure of 20: C13H18O4, Mr = 238.28, monoclinic,
space group P21/n (14), Dcalcd. = 1.238 g/cm3, Z = 4, a =
8.3494(18) Å, b = 10.392(4) Å, c = 14.746(4) Å, β = 92.441(11)°,
Vcell = 1278.3(7) Å3. Enraf–Nonius CAD4, Cu-Kα radiation, λ =
1.54179 Å,
T = 298 K. Approximate crystal dimensions =
0.3ϫ0.3ϫ0.3 mm3. Numbers of measured, independent and ob-
served reflections: 5121, 2310, 1797. Rint = 0.056. Final R = 0.072,
Rw = 0.080 for 154 parameters and 1775 reflections with IϾ2σ(I)
and Θmax = 67.81°.
CCDC-652790 contains the supplementary crystallographic data
for this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.au.uk/
data_request/cif.
New
Synthesis
of
1-(2,6-Dihydroxy-4-methoxyphenyl)-3-(4-
methoxyphenyl}prop-2-en-1-one (27): The procedure cited in ref.[8i]
gave mixtures of flavanone and chalcone, with low yields of chal-
cone. The TMSCl quenching method[58] was advantageous: To 30
(1.06 g, 3.53 mmol) in CH2Cl2 (10 mL) and chlorotrimethylsilane
(5.0 mL, 39.6 mmol) was added dropwise with stirring and cooling
in a room temp. water bath DBN (3.5 mL, 28.3 mmol) over 30 min.
The homogeneous solution was stirred for 3 days (not optimized),
poured onto EtOAc (50 mL) and 2.4  aq. HCl (50 mL) and the
mixture shaken for 10 minutes. The organic phase was washed with
2.4  aq. HCl (1ϫ) and water (4ϫ), then evaporated to dryness.
The orange residue was suspended in toluene/CH2Cl2 (1:1, 15 mL).
The suspension was reduced to a volume of ca 5 mL by rotatory
evaporation and overlayered with hexanes (10 mL). After standing
(5 h), the product was filtered and washed with hexanes to give an
Phloroglucinoldicarbaldehyde (21):[54] a) Vilsmeyer reagent: POCl3
(85 mL, 0.93 mol) was added dropwise with strong stirring to N,N-
dimethylformamide (72 mL, 0.93 mol) at 20–35 °C with occasional
cooling (ice/water bath). After complete addition and 30 min of
stirring at room temp., the resulting yellow viscous liquid was trans-
ferred into a dropping funnel. b) Phloroglucinol dihydrate was
dried to constant weight in an open beaker at 150 °C (20 h). Anhy-
drous phloroglucinol (18; 56.5 g, 0.448 mol) was dissolved in diox-
ane (185 mL; distilled from KOH/FeSO4) with heating, then cooled
to room temp. The Vilsmeyer reagent was slowly added to the
phloroglucinol/dioxane solution with stirring and occasional cool-
Eur. J. Org. Chem. 2007, 5886–5898
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5895

C. Dittmer, G. Raabe, L. Hintermann
FULL PAPER
ing at a reaction temperature below 30 °C. After complete addition,
the mixture was set aside overnight, after which it had crystallized
to a yellow solid. This material was transferred into a 2 L Erlen-
meyer with plenty of ice and little water, adjusting the total volume
of the mixture to 1.8 L by addition of more ice. The mixture was
slowly warmed to room temp. with stirring. From the initially
formed yellow solution, a yellow solid precipitated. After stirring
for 3 h at room temp., the mixture was filtered and the solid washed
with water. The solid was suspended in water (300 mL) and heated
to boiling, followed by cooling in an ice bath with stirring. Fil-
tration and washing with little water gave an orange-yellow solid
which was dried in an oven at 90 °C to constant weight (21; 72.7 g,
89%). The material may be purified by recrystallization from hot
EtOAc in the presence of some added water for increasing solubil-
ity (!), even though on cooling, crystals of the anhydrous substance
separate from the solution. Rf = 0.75 (SiO2; CH2Cl2/MeOH, 6:1);
m.p. above 220 °C (dec.). 1H NMR (300 MHz, [D6]DMSO): δ =
5.90 (s, 1 H Arl), 10.00 (s, 2 H, CHO), 12.50 (br. s, 2 H, OH), 13.58
(br. s, 1 H, OH) ppm. 13C NMR (76 MHz, [D6]DMSO): δ = 94.6
(CH), 104.2 (C), 169.5 (C), 169.9 (C), 191.8 (CH) ppm. IR (KBr):
180 mg, 33%). Data for Matteucinol (6): Rf = 0.6 (SiO2; EtOAc/
hexanes, 1:2); m.p. 171.0–172.8 °C (ref.[72] 172.5–173 °C; 172–
172.5 °C).[32] 1H NMR (400 MHz, [D6]acetone): δ = 2.03 (s, 3 H),
2.04 (s, 3 H), 2.79 (dd, J = 17.0, 3.1 Hz, 1 H), 3.12 (dd, J = 17.0,
12.6 Hz, 1 H), 3.82 (s, 3 H, OMe), 5.46 (dd, J = 12.5, 2.7 Hz, 1 H),
6.99 (d, J = 8.7 Hz, 2 H Arl), 7.49 (d, J = 8.7 Hz, 2 H Arl), 8.45
(br. s, 1 H, OH), 12.42 (s, 1 H, OH) ppm. 13C NMR (100 MHz,
[D6]acetone): δ = 7.5 (CH3), 8.2 (CH3), 43.5 (CH2), 55.5 (CH3),
79.2 (CH), 103.0 (C), 103.2 (C), 104.1 (C), 114.6 (CH), 128.5 (CH),
132.1 (C), 158.5 (C), 159.9 (C), 160.6 (C), 162.8 (C), 197.4 (C) ppm.
Analytical data were in accord with literature values.[73]
Matteucinolchalcone Triacetate (24): Matteucinol (6; 79.5 mg,
0.253 mmol), NaOAc (1.10 g, 13.4 mmol) and Ac2O (5 mL) were
heated to 160–170 °C (oilbath) and refluxed for 5 h. After cooling,
the mixture was transferred to a separatory funnel with EtOAc
(50 mL) and water. The organic phase was washed with water (2ϫ)
and NaHCO3 aq. (2ϫ). Evaporation and purification by column
chromatography (EtOAc/hexanes, 1:3–1:2) gave product-fractions,
which were evaporated and crystallized from CH2Cl2/hexanes by
overlayering and standing at 4 °C to give faint yellow crystals that
were washed with hexanes (85.7 mg, 77%). Rf = 0.11 (SiO2; EtOAc/
hexanes, 1:3); m.p. 152–153 °C (ref.[32] 152–153 °C). 1H NMR
(500 MHz, CDCl3): δ = 1.97 (s, 6 H, 2 Me), 2.14 (s, 6 H, 2 Me),
2.38 (s, 3 H, Me), 3.84 (s, 3 H, OMe), 6.79 (d, J = 16.1 Hz, 1 H),
6.88–6.92 (m, 2 H Arl), 7.41 (d, J = 16.1 Hz, 1 H), 7.48–7.52 (m,
2 H Arl) ppm. 13C NMR (125 MHz, CDCl3): δ = 10.5, 20.4, 20.5,
55.4, 114.5, 123.0, 124.1, 125.4, 126.8, 130.5, 144.9, 147.1, 149.6,
ν = 1618 s (br), 1442 (m), 1396 (m), 1273 (m), 1181 (m), 808 (m),
˜
611 (m) cm–1. MS (EI): m/z (%) = 182 (100) [M]+, 154 (52), 153
(63), 108 (5). C8H6O5 (182.13): calcd. C 52.76, H 3.32; found C
52.76, H 3.16.
Reduction of 21 to Dimethylphloroglucin (16): The literature pro-
cedure is recommended.[53a,54]
Acylation of 16 with p-Methoxycinnamoyl Chloride (17) and Mont-
morillonite to 6,8-Dimethyl-5,7-dihydroxy-4-(p-methoxyphenyl)chro-
man-2-one (22): Dimethylphloroglucine (16; 100 mg, 0.65 mmol),
17 (140 mg, 0.7 mmol) and montmorillonite KSF (300 mg) were
stirred in PhNO2 (2 mL) at 110 °C for 20 h. Dilution with EtOAc,
followed by filtration and washing of the organic phase with water
and evaporation gave a residue that was purified by column
chromatography (EtOAc/hexanes, 1:3–1:1) to give matteucinol (6;
17.0 mg, 8%) and a beige solid (22; 90 mg, 44%). An analytical
sample was recrystallized from CH2Cl2/hexanes. Data for 22: Rf =
0.24 (SiO2; EtOAc/hexanes, 1:2); m.p. 187–188 °C. 1H NMR
(400 MHz, [D6]acetone): δ = 2.16 (s, 3 H, Me), 2.17 (s, 3 H, Me),
2.92 (dd, J = 15.7, 1.9 Hz, 1 3-H), 3.08 (dd, J = 15.7, 6.8 Hz, 1 3-
H), 3.74 (s, 3 H, OMe), 4.64 (d, J = 6.5, 1.6 Hz, 1 4-H), 6.80–6.86
(A2 of A2B2, 2 H Arl), 7.02–7.08 (B2 of A2B2, 2 H Arl), 7.49 (s, 1
162.0, 167.9, 168.2, 191.0 ppm. IR (KBr): ν = 2942 (w), 1769 (s),
˜
1647 (s), 1604 (s), 1374 (m), 1253 (m), 1190 (s), 1086 (s), 821 (m)
cm–1. MS (EI): m/z (%) = 440 (36) [M]+, 398 (26), 356 (100), 338
(21), 314 (44), 222 (18), 180 (35). C24H24O8 (440.44): calcd. C 65.45,
H 5.49; found C 65.36, H 5.37.
Matteucinol Diacetate (23): Matteucinol (6; 58.0 mg, 0.185 mmol),
Ac2O (1 mL) and a small drop of concd. H2SO4 (ca. 15 mg) were
stirred for 45 min at room temp. The mixture was diluted with
water and EtOAc, the organic phase washed (2ϫNaHCO3 aq.,
1ϫwater), dried (MgSO4) and the solvents evaporated. The residue
was dissolved in CH2Cl2 (3 mL) and tBuOMe (15 mL), followed
by slow rotatory evaporation to a volume of 2 mL and overlayering
with hexanes (6 mL). The product slowly crystallized at –20 °C.
Filtration and washing with hexanes gave colorless needles of 23
H, OH), 7.51 (s, 1 H, OH) ppm. 13C NMR (76 MHz, [D6]acetone): (60.3 mg, 82%). Rf = 0.33 (SiO2; EtOAc/hexanes, 1:3); m.p. 177–
1
δ = 8.7 (CH3), 9.3 (CH3), 35.1 (CH), 38.1 (CH2), 55.4 (CH3), 105.0
178 °C. H NMR (500 MHz, CDCl3): δ = 1.95 (s, 3 H, Me), 2.03
(C), 106.2 (C), 108.5 (C), 114.7 (CH), 128.9 (CH), 134.9 (C), 149.6 (s, 3 H, Me), 2.37 (s, 3 H, Me), 2.42 (s, 3 H, Me), 2.75 (dd, J =
(C), 150.8 (C), 154.1 (C), 159.5 (C), 168.4 (C) ppm. IR (KBr): ν = 16.7, 2.3 Hz, 1 H), 3.03 (dd, J = 16.6, 13.7 Hz, 1 H), 3.83 (s, 3 H,
˜
3467 s (br), 2924 (w), 1758 (s), 1620 (m), 1470 (m), 1115 (s), 833 OMe), 5.42 (d, J = 13.6 Hz, 1 H), 6.93–6.99 (m, 2 H Arl), 7.35–
(m) cm–1. MS (EI): m/z (%) = 314 (80) M+, 271 (55), 241 (23), 206 7.41 (m, 2 H Arl) ppm. 13C NMR (125 MHz, CDCl3): δ = 9.2, 9.6,
(100), 178 (17). C18H18O5 (314.33): calcd. for C18H18O5 + 0.3H2O:
C 67.62, H 5.86; found C 67.53, H 5.95.
20.4, 21.0, 45.2, 55.4, 79.0, 111.6 (br), 114.2, 117.8, 118.0 (br),
127.5, 130.6, 146.5 (br), 153.6, 159.1, 159.9, 167.9, 169.4 (br), 190.4
(br) ppm. MS (EI): m/z (%) = 398 (30) [M]+, 356 (100), 314 (35),
Acylation of 16 with p-Methoxycinnamoyl Chloride (17) and AlCl3
to Matteucinol (6): Dimethylphloroglucine (16; 540 mg, 3.5 mmol)
was suspended in PhNO2 (10 mL) and AlCl3 (650 mg, 4.9 mmol)
was added in portions. The mixture was heated to 60 °C and a
solution/suspension of 17 (690 mg, 3.5 mmol) in PhNO2 (6 mL)
was slowly added. The reaction mixture was stirred for 2 h at 60 °C,
cooled to room temp. and quenched by addition of water, 2  aq.
HCl and EtOAc. The aqueous phase was extracted with EtOAc
(2ϫ) and the collected organic phases washed with 2  aq. HCl,
NaHCO3 aq. (2ϫ) and water (2ϫ). Evaporation and separation of
the residue by column chromatography (tBuOMe/hexanes, 1:10,
then EtOAc/hexanes, 1:3–1:2–1:1) gave matteucinol (6; 246 mg,
22%), lactone 22 (180 mg, 16%) and dimethylphloroglucine (16;
222 (20), 180 (51), 134 (20). IR (KBr): ν = 2932 (w), 1761 (s), 1684
˜
(m), 1609 (m), 1516 (m), 1450 (m), 1355 (m), 1219 (s) cm–1.
C22H22O7 (398.41): calcd. C 66.32, H 5.57; found C 66.19, H 5.36.
CSA-Catalyzed Cyclization of 24 to 23. Repetition of the Fujise Ex-
periment: Chalcone triacetate 24 (38 mg, 0.086 mmol) and (+)-cam-
phorsulfonic acid (1.0 mg, 0.0043 mmol, 5 mol-%) were dissolved
in EtOH abs. (3 mL) and heated to 110 °C in an open thick-walled
glas pressure tube. When the volume of the reaction solution had
reached 2 mL by evaporation of EtOH, the tube was closed and
kept for 24 h at 110 °C. After cooling, toluene (3 mL) was added,
the mixture partially evaporated to 2 mL and separated by column
chromatography (EtOAc/hexanes, 1:4–1:1) to give matteucinol-di-
5896
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 5886–5898

Catalysis by Chiral Brønsted Acids and Bases
M. Iinuma, S. Matsuura, T. Tanaka, Chem. Pharm. Bull. 1984,
32, 1472–1476.
[10] T. Patonay, G. Litkei, M. Zsuga, A. Kiss, Org. Prep. Proced.
Int. 1984, 16, 315–319.
acetate (25; 11.7 mg, 40%). According to the HPLC-analysis (for
conditions see the electronic supporting information), the product
was racemic within the limits of accuracy of the analysis (eeϽ2%).
Supporting Information (see also the footnote on the first page of
this article): Conditions for chiral HPLC analysis of flavanones 2,
9, 10, 23, 30–32, and figures for the X-ray structural analysis of
compound 20.
[11] a) G. Solladié, N. Gehrold, J. Maignan, Eur. J. Org. Chem.
1999, 2309–2314; b) R. G. Harvey, J.-T. Hahn, M. Bukowska,
H. Jackson, J. Org. Chem. 1990, 55, 6161–6166.
[12]
a) N. S. Poonia, K. Chhabra, C. Kumar, V. W. Bhagwat, J. Org.
Chem. 1977, 42, 3311–3313; b) I. Thomsen, K. B. G. Torssell,
Acta Chem. Scand., Ser. B 1988, 42, 303–308; c) L. Reichel,
J. Steudel, Justus Liebigs Ann. Chem. 1942, 553, 83–97; d) T.
Oyamada, Bull. Chem. Soc. Jpn. 1935, 10, 182–186.
a) J. K. Makrandi, S. Bala, Synth. Commun. 2000, 30, 3555–
3558; b) H. García, S. Iborra, M. A. Miranda, J. Primo, Het-
erocycles 1985, 23, 1983–1989.
Acknowledgments
[13]
L. H. thanks the Deutsche Forschungsgemeinschaft (DFG) for
funding within the Emmy Noether Programm, and Prof. Carsten
Bolm, RWTH Aachen, for continuing support.
[14]
[15]
K. Tanaka, T. Sugino, Green Chem. 2001, 3, 133–134.
G. Zemplén, R. Bognár, L. Mester, Ber. Dtsch. Chem. Ges.
1942, 75, 1432–1438.
[1] a) G. Forkmann, W. Heller, in Comprehensive Natural Products
Chemistry, vol. 1, Elsevier, 1999, pp. 713–748; b) K. Springob,
J. Nakajima, M. Yamazaki, K. Saito, Nat. Prod. Rep. 2003, 20,
288–303.
[2] a) Ø. M. Andersen, K. R. Markham, Flavonoids: Chemistry,
Biochemistry and Applications, CRC Press, Boca Raton, 2006;
b) T. A. Geissman, The Chemistry of Flavonoid Compounds,
Pergamon Press, Oxford, 1962.
[3] a) M. Gensheimer, A. Mushegian, Protein Sci. 2004, 13, 540–
544; b) C. Herles, A. Braune, M. Blaut, Arch. Microbiol. 2004,
181, 428–434.
[4] Chalcone isomerase: a) E. Wong, E. Moustafa, Tetrahedron
Lett. 1966, 26, 3021–3022; b) E. Moustafa, E. Wong, Phyto-
chemistry 1967, 6, 625–635; c) K. Hahlbrock, H. Zilg, H.
Grisebach, Eur. J. Biochem. 1970, 15, 13–18; d) R. A. Bednar,
J. R. Hadcock, J. Biol. Chem. 1988, 263, 9582–9588.
[16]
[17]
a) D. J. Macquarrie, R. Nazih, S. Sebti, Green Chem. 2002, 4,
56–59; b) L. M. Harwood, G. C. Loftus, A. Oxford, C. Thom-
son, Synth. Commun. 1990, 20, 649–657.
a) A. Aitmambetov, C. Dalimov, A. Kubzheterova, Chem. Nat.
Compd. 2002, 37, 421–423; b) M. Majewski, G. Bantle, Tetrahe-
dron Lett. 1989, 30, 6653–6656; c) S. W. Sathe, K. N. Wadod-
kar, Indian J. Chem. Sect. B 1982, 21, 153–154; d) M. B. Ba-
gade, A. W. Thool, P. D. Lokhande, B. J. Ghiya, Indian J.
Chem. Sect. B 1991, 30, 973–975.
[18]
[19]
T. Patonay, R. S. Varma, A. Vass, A. Lévai, J. Dudás, Tetrahe-
dron Lett. 2001, 42, 1403–1406.
a) K. Kaneda, T. Arai, Org. Biomol. Chem. 2003, 1, 2041–2043;
b) Y. Maki, K. Shimada, M. Sako, K. Hirota, Tetrahedron
1988, 44, 3187–3194; c) G. Pandey, A. Krishna, G. Kumaras-
wamy, Tetrahedron Lett. 1987, 28, 4615–4616; d) R. Matsush-
ima, H. Kageyama, J. Chem. Soc. Perkin Trans. 2 1985, 743–
748; e) R. Matsushima, I. Hirao, Bull. Chem. Soc. Jpn. 1980,
53, 518–522; f) F. R. Stermitz, J. A. Adamovics, J. Geigert, Tet-
rahedron 1975, 31, 1593–1595.
[5] a) J. M. Jez, J. P. Noel, J. Biol. Chem. 2002, 277, 1361–1369; b)
J. M. Jez, M. E. Bowman, J. P. Noel, Biochemistry 2002, 41,
5168–5176; c) J. M. Jez, M. E. Bowman, R. A. Dixon, J. P.
Noel, Nat. Struct. Biol. 2000, 7, 786–791.
[6] L. Reichel, W. Burkart, Ber. Dtsch. Chem. Ges. 1941, 74, 1802–
1805.
[20]
[21]
[22]
[23]
B. S. Goud, K. Panneerselvam, D. E. Zacharias, G. R. Desi-
raju, J. Chem. Soc. Perkin Trans. 2 1995, 325–330.
Y. Hoshino, N. Takeno, Bull. Chem. Soc. Jpn. 1986, 59, 2903–
2904.
K. Maruyama, K. Tamanaka, A. Nishinaga, A. Inada, T. Nak-
anishi, Tetrahedron Lett. 1989, 30, 4145–4148.
a) N. Ahmed, W. H. Ansari, J. Chem. Res. Synop. 2003, 572–
573; b) N. Ahmed, H. Ali, J. E. van Lier, Tetrahedron Lett.
2006, 46, 253–256.
[7] S. Kostanecki, W. Szabranski, Ber. Dtsch. Chem. Ges. 1904, 37,
2634–2635.
[8] The kinetics, thermodynamics and the mechanism of the 2Ј-
hydroxychalcone/flavanone interconversion, mostly in aque-
ous/alcoholic solution, have been studied intensively: a) E. A.
González, M. A. Nazareno, C. D. Borsarelli, J. Chem. Soc. Per-
kin Trans. 2 2002, 2052–2056; b) A. Cisak, C. Mielczarek, J.
Chem. Soc. Perkin Trans. 2 1992, 1603–1607; c) V. L. Arcus,
C. D. Simpson, L. Main, J. Chem. Res. (S) 1992, 80–81; d)
R. G. Button, P. J. Taylor, J. Chem. Soc. Perkin Trans. 2 1992,
1571–1580; e) C. M. Brennan, I. Hunt, T. C. Jarvis, C. D. John-
son, P. D. McDonnell, Can. J. Chem. 1990, 68, 1780–1785; f)
C. O. Miles, L. Main, J. Chem. Soc. Perkin Trans. 2 1989, 1623–
1632; g) C. O. Miles, L. Main, J. Chem. Soc. Perkin Trans. 2
1988, 195–198; h) J. J. P. Furlong, N. S. Nudelman, J. Chem.
Soc. Perkin Trans. 2 1988, 1213–1217; i) C. O. Miles, L. Main,
J. Chem. Soc. Perkin Trans. 2 1985, 1639–1642; j) J. J. P. Fur-
long, N. S. Nudelman, J. Chem. Soc. Perkin Trans. 2 1985, 633–
639; k) K. B. Old, L. Main, J. Chem. Soc. Perkin Trans. 2 1982,
1309–1312; l) A. I. Panasenko, O. I. Kachurin, S. P. Starkov,
Izv. Vyssh. Uchebn. Zaved., Khim. Khim. Tekhnol. 1975, 18,
1203–1207 (C. A. 1975, 83, 205454); m) G. Litkei, R. Bognár,
Z. Dinya, E. R. Dávid, in Topics in Flavonoid Chemistry and
Biochemistry (Eds.: L. Farkas, M. Gábor, F. Kállay), Elsevier,
Amsterdam, 1975, pp. 110–118; n) A. Grouiller, P. Thomassery,
H. Pacheco, Bull. Soc. Chim. Fr. 1973, 3448–3451; o) A.
Corvaisier, Bull. Soc. Chim. Fr. 1962, 528–535; p) M. Shimoko-
riyama, J. Am. Chem. Soc. 1957, 79, 4199–4202.
[24]
[25]
[26]
N. K. Sangwan, B. S. Varma, K. S. Dhindsa, Chem. Ind. (Lon-
don) 1984, 271–272.
Z. Sanicanin, I. Tabakovic, Tetrahedron Lett. 1986, 27, 407–
408.
a) Y. Noda, M. Watanabe, Helv. Chim. Acta 2002, 85, 3473–
3477; b) K. J. Hodgetts, Tetrahedron Lett. 2001, 42, 3763–3766;
c) K. J. Hodgetts, ARKIVOC 2001, 74–79; d) G. Solladié, N.
Gehrold, J. Maignan, Tetrahedron: Asymmetry 1999, 10, 2739–
2747; e) A. V. Rama Rao, A. S. Gaitonde, K. R. C. Prakash, S.
Prahlada Rao, Tetrahedron Lett. 1994, 35, 6347–6350.
a) M. Kawasaki, H. Kakuda, M. Goto, S. Kawabata, T. Kome-
tani, Tetrahedron: Asymmetry 2003, 14, 1529–1534; b) T. Todo-
roki, A. Saito, A. Tanaka, Biosci. Biotechnol. Biochem. 2002,
66, 1772–1774.
[27]
[28]
[29]
[30]
M. M. Biddle, M. Lin, K. A. Scheidt, J. Am. Chem. Soc. 2007,
129, 3830–3831.
S. Chandrasekhar, K. Vijeender, K. V. Reddy, Tetrahedron Lett.
2005, 46, 6991–6993.
a) E. Sekino, T. Kumamoto, T. Tanaka, T. Ikeda, T. Ishikawa,
J. Org. Chem. 2004, 69, 2760–2767; b) T. Ishikawa, Y. Oku, T.
Tanaka, T. Kumamoto, Tetrahedron Lett. 1999, 40, 3777–3780;
c) T. Tanaka, T. Kumamoto, T. Ishikawa, Tetrahedron: Asym-
metry 2000, 11, 4633–4637; d) T. Tanaka, T. Kumamoto, T.
Ishikawa, Tetrahedron Lett. 2000, 41, 10229–10232.
[9] a) H. Kagawa, T. Takahashi, M. Uno, S. Ohta, Y. Harigaya,
Chem. Pharm. Bull. 2004, 52, 953–956; b) M. De, D. P. Majum-
dar, N. G. Kundu, J. Indian Chem. Soc. 1999, 76, 665–674; c)
Eur. J. Org. Chem. 2007, 5886–5898
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5897

C. Dittmer, G. Raabe, L. Hintermann
FULL PAPER
[31] R. M. de Figueiredo, M. Christmann, Eur. J. Org. Chem. 2007,
2575–2600.
[32] S. Fujise, H. Sasaki, Ber. Dtsch. Chem. Ges. 1938, 71, 341–344.
[33] H. Tatsuta, Bull. Chem. Soc. Jpn. 1941, 16, 327–329.
[34] S. Fujise, S. Sasaki, Nippon Kagaku Zasshi 1951, 72, 462–463.
[35] S. Fujise, M. Suzuki, Nippon Kagaku Zasshi 1951, 72, 1073–
1075.
Monatsh. Chem. 1890, 11, 104–119; d) J. Herzig, B. Erthal,
Monatsh. Chem. 1910, 31, 827–831.
[51] a) W. Riedl, J. Nickl, K. H. Risse, R. Mitteldorf, Chem. Ber.
1956, 89, 1849–1863; b) W. Riedl, K. H. Risse, Justus Liebigs
Ann. Chem. 1954, 585, 209–219; c) F. H. Curd, A. Robertson,
J. Chem. Soc. 1933, 437–444.
[52] H. Weidel, F. Wenzel, Monatsh. Chem. 1898, 19, 236–248.
[53] a) W. Gruber, Ber. Dtsch. Chem. Ges. 1942, 75, 29–33; b) S. B.
Bharate, I. P. Singh, Tetrahedron Lett. 2006, 47, 7021–7024.
[54] a) E. Bombardelli, B. Gabetta, M. J. Magistretti, EP 122053,
1984; b) E. Bombardelli, B. Gabetta, M. J. Magistretti, U. S.
Patent 4636569, 1987.
[55] J. Lee, T. Tseng, Y. Lee, Synthesis 2001, 2247–2254.
[56] Y. Takizawa, T. Mitsuhashi, J. Heterocycl. Chem. 1978, 15,
701–701.
[36] Even if enrichment during crystallization were taking place, an
initial enantiomeric excess of more than 35% must be assumed.
[37] The highest optical rotation for samples of 9 known at the time
was [α]D = +33.3; in subsequent years, higher [α]D values of
–85.3 (c = 1.2, MeOH) or 74 were reported: a) B. Su, E. J.
Park, J. S. Vigo, J. G. Graham, F. Cabieses, H. H. S. Fong, J. M.
Pezzuto, A. D. Kinghorn, Phytochemistry 2003, 63, 335–341;
b) G. Cardillo, L. Merlini, G. Nasini, P. Salvadori, J. Chem.
Soc. C 1971, 3967–3970.
[57] Determinations of ee values are more accurate at higher con-
versions because of the better signal to noise ratio.
[38] a) L. Velluz, J. Valls, J. Mathieu, Angew. Chem. 1967, 79, 774–
785; b) H. Pracejus, Fortschr. Chem. Forsch. 1967, 8, 493–553; [58] C. O. Miles, L. Main, B. K. Nicholson, Aust. J. Chem. 1989,
c) J. Mathieu, J. Weill-Raynal, Bull. Soc. Chim. Fr. 1968, 1211–
1244.
42, 1103–1113.
[59] N. Narasimhachari, T. R. Seshadri, Proc. Indian Acad. Sci.
1948, 27A, 223–239 (Chem. Abstr. 1950, 44, 7577).
[60] a) Naringenin 4Ј,7-dimethyl ether: T. A. Geissman, R. O. Clin-
ton, J. Am. Chem. Soc. 1946, 68, 697–700; b) For conversion
[39] We found one unsuccessful contemporary attempt to apply the
CSA method: H. Brockmann, K. Maier, Justus Liebigs Ann.
Chem. 1938, 535, 149–175.
[40] Early example of asymmetric catalysis with camphorsulfonic
acid: P. Maitland, W. H. Mills, J. Chem. Soc. 1936, 987–998.
[41] Asymmetric Brønsted acid catalysis: a) B. List, J. W. Yang, Sci-
ence 2006, 313, 1584–1586; b) C. Bolm, T. Rantanen, I. Schif-
fers, L. Zani, Angew. Chem. 2005, 117, 1788–1793; Angew.
Chem. Int. Ed. 2005, 44, 1758–1763.
[42] Chiral phosphoric acid diesters: a) T. Akiyama, J. Itoh, K.
Yokota, K. Fuchibe, Angew. Chem. 2004, 116, 1592–1594; An-
gew. Chem. Int. Ed. 2004, 43, 1566–1568 and subsequent
works; b) D. Uraguchi, M. Terada, J. Am. Chem. Soc. 2004,
126, 5356–5357 and subsequent works; c) M. Rueping, E. Su-
giono, C. Azap, T. Theissmann, M. Bolte, Org. Lett. 2005, 7,
3781–3783 and subsequent works; d) S. Hoffmann, A. Seayad,
B. List, Angew. Chem. 2005, 117, 7590–7593; Angew. Chem.
Int. Ed. 2005, 44, 7424–7427 and subsequent works; e) R. I.
Storer, D. E. Carrera, Y. Ni, D. W. C. MacMillan, J. Am. Chem.
Soc. 2006, 128, 84–86 and subsequent works.
[43] Note, that the absence of an enantiomeric excess is a necessary
precondition for the correctness of the simple first-order ki-
netic fit; if kR and kS are not equal, the reaction progress curve
assumes a different shape, which will be discussed elsewhere.
[44] The value Keq = 3.5 was found in an aqueous alcoholic buffer
at pH7 and 80 °C: A. Grouiller, P. Thomassery, H. Pacheco,
Bull. Soc. Chim. Fr. 1973, 3448–3451.
into chalcone, see ref.[8i]
.
[61] K. Kacprzak, J. Gawronski, Synthesis 2001, 961–998.
[62] H. Arakawa, M. Nakazaki, Justus Liebigs Ann. Chem. 1960,
636, 111–117.
[63] For examples, see: a) C. A. Mead, A. Moscowitz, J. Am. Chem.
Soc. 1980, 102, 7301–7302; b) J. M. O’Gorman, J. Am. Chem.
Soc. 1944, 66, 1041–1041; c) J. W. Cornforth, R. H. Cornforth,
M. J. S. Dewar, Nature 1944, 153, 317–317.
[64] A search in the Beilstein database gave around 20 literature
references for isolation of 30 from natural sources. The natural
product displays an optical rotation of αD = –182° (solvent not
available) and has therefore the
S configuration: G. P.
Kononenko, S. A. Popravko, N. S. Vul’fson, Bioorg. Khim.
1975, 1, 506–511 (Chem. Abstr. 1975, 83, 128765).
[65] Y. Hoshino, N. Takeno, Bull. Chem. Soc. Jpn. 1986, 59, 2903–
2904.
[66] E. B. Lee, K. H. Shin, W. S. Woo, J. Med. Chem. 1968, 11,
1262–1263.
[67] a) S. Sogawa, Y. Nihro, H. Ueda, A. Izumi, T. Miki, H. Matsu-
moto, T. Satoh, J. Med. Chem. 1993, 36, 3904–3909; b) H. Tak-
agaki, S. Nakanishi, M. Abe, H. Ohki, Y. Sano, U. S. Patent
5068364, 1991.
[68] a) W. Feuerstein, St. v. Kostanecki, Ber. Dtsch. Chem. Ges.
1899, 32, 315–317; b) L. Reichel, W. Burkart, K. Müller, Justus
Liebigs Ann. Chem. 1942, 550, 146–151.
[45] T. Munesada, Yakugaku Zasshi 1924, 505, 185–189 (Chem.
Abstr. 1924, 18, 13671).
[46] S. Fujise, T. Kubota, Ber. Dtsch. Chem. Ges. 1934, 67, 1905–
1908.
[47] a) S. Fujise, Scient. Pap. Inst. Phys. Chem. Res. 1929, 11, 111–
118 (Chem. Abstr. 1930, 24, 30312); b) J. Xie, L. Wang, C. Liu,
D. Ge, C. A 1988, 108, 204365; c) L. Wang, J. Kong, Chem.
Abstr. 1989, 110, 7898.
[48] M. Hauteville, R. Stromberg, P. Gaillard, M.-C. Duclos, Lie-
bigs Ann. 1995, 1707–1709.
[69] A. Aitmambetov, A. Kubzheterova, Russ. J. Bioorg. Chem.
2002, 28, 165–166.
[70] M. G. Marathey, J. Org. Chem. 1955, 20, 563–571.
[71] a) K. Oyama, T. Kondo, J. Org. Chem. 2004, 69, 5240–5246;
b) H. Chu, H. Wu, Y. Lee, Tetrahedron 2004, 60, 2647–2656.
[72] H. R. Arthur, W. H. Hui, J. Chem. Soc. 1954, 2782–2784.
[73] a) M. do. C. M. Miraglia, A. P. de Padua, A. A. L. Mesquita,
O. R. Gottlieb, Phytochemistry 1985, 24, 1120–1120; b) N.
Tanaka, T. Murakami, H. Wada, A. B. Gutierrez, Y. Saiki, C.
Chen, Chem. Pharm. Bull. 1985, 33, 5231–5238.
Received: July 24, 2007
[49] A. R. Stein, Can. J. Chem. 1965, 43, 1493–1507.
[50] a) O. Marguelis, Monatsh. Chem. 1888, 9, 1045–1054; b) O.
Marguelis, Monatsh. Chem. 1889, 10, 459–463; c) A. Spitzer,
Published Online: October 8, 2007
5898
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 5886–5898