Study of the O-glycidylation of natural phenolic compounds. the relationship between the phenolic structure and the reaction mechanism

Article abstract of DOI:10.1016/j.tet.2012.11.079

The O-alkylation reaction by epichlorohydrin of some natural phenolic compounds such as 4-methylcatechol, gallic acid, protocatechuic acid, pyrogallol and resorcinol was investigated. Phenolic compounds reacted first with epichlorohydrin in the presence of benzyltriethylammonium chloride as phase transfer catalyst. Then, an aqueous solution of sodium hydroxide was added. It was demonstrated that the two competitive mechanisms involved in the O-alkylation reaction were highly dependent of the starting material. The O-alkylated products obtained in this reaction could be further used as bisphenol A substitutes in the synthesis of epoxy resins pre-polymers.

Full text of DOI:10.1016/j.tet.2012.11.079

Tetrahedron 69 (2013) 1345e1353
Contents lists available at SciVerse ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Study of the O-glycidylation of natural phenolic compounds.
The relationship between the phenolic structure and the reaction
mechanism
,
b
,
c
,
,
b
,
c
d
a,b,c
ꢀ a
ꢀ ꢁ
, Sylvain Caillol , Helene Fulcrand
Chahinez Aouf a
, Christine Le Guerneve
*
a INRA, UMR 1083 Sciences pour l’œnologie, F-34060 Montpellier, France
b Montpellier SupAgro, UMR 1083 Sciences pour l’œnologie, F-34060 Montpellier, France
c
ꢀ
Universite Montpellier I, UMR 1083 Sciences pour l’œnologie, F-34060 Montpellier, France
Institut Charles Gerhardt, UMR CNRS 5253, Equipe Ingenierie et Architecture Macromoleculaire, ENSCM, 34296 Montpellier, France
d
ꢀ
ꢀ
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 27 August 2012
Received in revised form 23 November 2012
Accepted 24 November 2012
Available online 30 November 2012
The O-alkylation reaction by epichlorohydrin of some natural phenolic compounds such as
4-methylcatechol, gallic acid, protocatechuic acid, pyrogallol and resorcinol was investigated. Phenolic
compounds reacted first with epichlorohydrin in the presence of benzyltriethylammonium chloride as
phase transfer catalyst. Then, an aqueous solution of sodium hydroxide was added.
It was demonstrated that the two competitive mechanisms involved in the O-alkylation reaction were
highly dependent of the starting material. The O-alkylated products obtained in this reaction could be
further used as bisphenol A substitutes in the synthesis of epoxy resins pre-polymers.
Ó 2012 Elsevier Ltd. All rights reserved.
Keywords:
Natural phenolic compounds
O-glycidylation
Reactivity
Mechanism
Bio-based epoxy resin pre-polymers
1. Introduction
between bisphenol A (BPA) and epichlorohydrin, yielding diglycidyl
ether of bisphenol A (DGEBA). Suspected of being hazardous to
Bio-based polymer products derived from annually renewable
agricultural and biomass feedstock have become increasingly im-
portant as sustainable and eco-efficient products, which will po-
tentially replace the products based exclusively on petroleum
feedstock.1,2 Of all polymers produced, the thermosets represent an
important class in the industry due to their high flexibility for tai-
loring desired ultimate properties, leading to their high modulus,
strength, durability, and thermal and chemical resistances as pro-
vided by high cross-linking density.3e7 Paradoxically, thermosets
prepared from renewable resources have been the subject of lim-
ited investigations.8
Epoxy resins are some of the most important thermosetting
resins, and are extensively used as coatings, adhesives, electronic
materials, and for structural applications because of their out-
standing mechanical properties ranging from extreme flexibility to
high strength and hardness, high adhesion strength, good heat
resistance and high electrical resistance.9,10 Nowadays, almost 90%
of the world production of epoxy resins is based on the reaction
humans, concerns about the use of BPA in consumer products have
been regularly reported in the news media since 2008 after several
scientists and governments questioned its safety.11e14 Therefore,
there is a huge interest in developing a bio-based epoxy monomer
or oligomer able to replace the traditional DGEBA by offering high
performance materials.
In the manufacturing of sustainable epoxy resins, epoxidized
plant oils (soybean oil, linseed oil, castor oil.) and fatty acids have
been largely utilized as reported in the literature.15e17
However, these networks present disadvantages in heat
resistance and mechanical properties due to the long aliphatic
chains. Thus, this type of epoxy resin is applied only in limited fields
of applications. Recently, the use of natural polysaccharides in the
synthesis of bio-based epoxy resins retained industrial and scien-
tific interest. Indeed, the conversion of maltitol, and sorbitol into
multifunctional epoxy monomers has been reported.18 Glycerol
polyglycidyl ether, usually employed in textile and paper as pro-
cessing agent and as reactive diluent, was reacted with curing
agent to produce bio-based epoxy resin.19 Isosorbide produced
from the double dehydration of sorbitol was also used as a sub-
stitute of bisphenol A to produce epoxy material with good net-
work properties.20,21 Unfortunately, glycidyl ether derivatives of
* Corresponding author. Tel.: þ33 4 99 61 24 54; fax: þ33 4 99 61 28 57; e-mail
address: aouf@supagro.inra.fr (C. Aouf).
0040-4020/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tet.2012.11.079

1346
C. Aouf et al. / Tetrahedron 69 (2013) 1345e1353
polysaccharide monomers are hygroscopic and the presence of
water may deteriorate the properties of the networks.
Although the experimental conditions of the reaction are not
totally the same between the work that we reported earlier on
catechin and those of the Haruo Tomita’s patent on gallic acid,
the products obtained after reaction suggest a difference in re-
activity between the two phenolic compounds towards epi-
chlorohydrin. Therefore, we decided to investigate further the
role of the phase transfer catalyst (PTC) on one hand, and, on the
other hand, the glycidylation of model phenolic compounds by
epichlorohydrin in order to ultimately optimize the functionali-
zation reactions of natural tannins. For this purpose, some
characteristic mononuclear aromatic phenolic compounds,
bearing two or three phenolic hydroxyls in ortho- or meta-posi-
tions, with or without additional carboxylic group, were reacted
with epichlorohydrin. The structures of the glycidylated products
were determined, mainly from NMR experiments. The glycidy-
lation mechanism was then discussed on the basis of the struc-
tural characterizations.
The heat resistance and the mechanical and electrical perfor-
mance properties of organic compounds are generally attributed to
the aromatic ring structure. Most of the naturally occurring aro-
matic compounds are often believed to be originated from lignin.
For this reason, many researchers have made their efforts to apply
these macromolecular compounds to the production of epoxy
resins. Until now, methods to synthesize lignin-based epoxy resin
can be summarized into two categories as follows: (i) blending
derivatives of lignin with epoxy resin directly;22 (ii) modifying
lignin derivatives (organosolv lignin) to improve their reactional
ability, followed by epoxidation.23,24 However, all lignins created
problems with organic solventdsolubility, and thus required spe-
cial solvent mixtures for formulation. Moreover their chemical
structures, which are not yet entirely elucidated, could make the
direct functionalization of this polymer difficult.
The other terrestrial source of phenolic compounds is tannins.
Tannins are defined as water-soluble plant phenolic compounds
having molecular weight ranging from 500 Da to 3000e4000 Da.
Furthermore, the compounds should undergo the usual phenolic
reactions and have the ability to precipitate some alkaloids, gelatin
and other proteins from solutions.25e27 They constitute an impor-
tant group within the phenolic compounds and may be subdivided
into hydrolyzable and condensed tannins. The former are esters of
gallic acid (gallo- and ellagitannins), while the later (also known as
proanthocyanidins) are polymers of polyhydroxyflavan-3-ol
monomers. Another subdivision, the phlorotannins derived from
the oligomerizing dehydrogenative coupling of phloroglucinol,
have been isolated from several genera of red-brown algae.26,28e30
Thus, tannins exhibit a wide structural diversity and their molec-
ular structure is a key determinant of their chemical reactivity.
The objective of our ongoing research is to study the feasibility
of replacing bisphenol A by such natural phenolic compounds in
the synthesis of novel bio-based reactive pre-polymer systems.
In our previous paper,31 one of the building blocks of polymeric
condensed tannins, namely catechin, was reacted with epichloro-
hydrin in alkaline medium to lead to the expected tetraglycidy-
lether of catechin along with a benzodioxane derivative (Scheme 1).
The formation of this cyclic by-product seemed to be directly re-
lated to the ortho-positions of the two phenolic OH groups carried
by the B-ring of catechin.31
In 1985, the glycidylation of gallic acid (building block of gal-
lotannins) had been claimed in a patent from Haruo Tomita et al.32
The use of a phase transfer catalyst (PTC) and a two step-procedure
in Tomita’s protocol make the noticeable differences compared to
the protocol that we used for the catechin glycidylation. The O-al-
kylation product obtained by the experimental conditions given in
the patent is reported as being the product of substitution reaction
of epichlorohydrin onto the carboxyl group and at least onto one
phenolic hydroxyl group. The formation of side-products such as
the benzodioxane derivative that we obtained with catechin was
not mentioned in this work in spite of the presence of three hy-
droxyls groups in ortho-position on the aromatic ring of gallic acid.
The model phenolic compounds studied in the present work
are: 4-methylcatechol, gallic acid, protocatechuic acid, pyrogallol
and resorcinol.
2. Results and discussion
2.1. Confirmation of the benzodioxane structure in the gly-
cidylation of 4-methylcatechol
In our previous work related to the functionalization of cate-
chin,31 the reaction of 4-methylcatechol 1 (as representative of the
B-ring of catechin) with epichlorohydrin in alkaline medium
revealed the formation of the benzodioxane derivative 3 besides
the expected glycidylation product (Scheme 2). This cyclic by-
product was tentatively identified by NMR analyses among a mix-
ture of other impurities.
In order to unequivocally establish the chemical structure of the
hypothetical benzodioxane derivative, the glycidylation reaction of
4-methylcatechol 1 was repeated in the same reaction conditions
mentioned in our anterior work.31 The crude product was purified
by silica gel chromatography allowing the isolation of products 2
and 3 in equal amount. These products definitely correspond, re-
spectively, to the diglycidylether and the benzodioxane derivative
of 4-methylcatechol, (Scheme 2).
Detailed structure characterization of product 3 was performed
by NMR spectroscopy. The 1H NMR spectrum (Fig. 1) displays res-
onance signals easily attributed using both their chemical shifts and
coupling pattern. The 6.6e6.8 ppm region contains an ABX system
corresponding to the 3 aromatic ring protons whereas the aliphatic
signals in the 3.5e5.1 ppm range arise from the dioxane group. The
dioxane hydroxyl group gives a characteristic signal at 5.03 ppm,
the methylene proton Ha and Ha0 resonance signals are located at
3.96 and 4.27 ppm while the signal at 4.09 ppm is assigned to the
methyne proton Hb.
The HMBC spectrum of product 3 (Fig. 1) allows to attribute the
quaternary carbons of the aromatic ring C1, C2 and C3 from their
correlations with both the aromatic, the methyne and the methyl
Scheme 1. Glycidylation of catechin.

C. Aouf et al. / Tetrahedron 69 (2013) 1345e1353
1347
Scheme 2. Glycidylation of 4-methylcatechol in the presence of ethanolic solution of NaOH.
resins curing.35e37 Only Tomita et al.32 achieved the O-glycidylation
of gallic acid in the presence of PTC. However, the reaction products
were poorly characterized and the reactivity of gallic acid was not
discussed. Therefore, we carried out the reaction with gallic acid
following carefully the experimental procedure disclosed in Tomi-
ta’s invention. Thus, gallic acid was mixed with epichlorohydrin
(molar ratio hydroxyl group/epichlorohydrin, 1:4) in the presence
of benzyltriethylammonium chloride (BnEt3NCl) (molar ratio sub-
strate/catalyst, 1:0.05) and the mixture was heated 1 h at 100 ꢀC.
Then, the solution was cooled to 30 ꢀC. The previous amount of
phase transfer catalyst (PTC) was added again along with an
aqueous solution of sodium hydroxide (molar ratio hydroxyl group/
NaOH, 1:2). The reaction mixture was further stirred at 30 ꢀC for
90 min. The purification of the crude product by silica gel chro-
matography allowed the isolation in fair yield (68%) of compound 5
(Scheme 3).
The formation of the tetraglycidylated derivative of gallic acid 5
was established from 1H NMR spectrum (see Experimental section).
Whereas the characteristic phenolic and carboxyl hydroxyl proton
resonance signals (at w9.5 ppm and w12.5 ppm, respectively) cannot
be observed, numerous aliphatic signals arising from oxirane groups
are identified. The methylene and methyne ring proton signals ap-
pear in the 2.73e4.65 ppm range and the CH2eO protons give res-
onance signals in the 3.92e4.65 ppm spectral range. The average
number of oxirane groups per gallic acid ring, calculated from the
ratio of 1H surface signal integrations, was equal to w4. In-
terestingly, in contrast to what was observed with methylcatechol,
no benzodioxane derivative was detected, despite the ortho-posi-
tions of the three phenolic hydroxyls carried by gallic acid.
It was verified that the formation of benzodioxane derivatives
from glycidylation reaction of methylcatechol was not due to dif-
ferences in reaction conditions, especially the lack of PTC
(BnEt3NCl). For this, 4-methylcatechol 1 was reacted with epi-
chlorohydrin in the same experimental conditions than the ones
applied to gallic acid (i.e., with PTC). Products 2 and 3 were ob-
tained in equivalent amounts, with a total reaction yield of 67%.
Therefore, the use of PTC in glycidylation reaction does not prevent
the formation of benzodioxane derivatives. It only increases the
reaction yield.
Fig. 1. HMBC spectrum of products 3A and 3B in DMSO-d6 showing the correlations
between the methylene protons Ha and Ha0 and the quaternary carbons C1 and C2.
protons. From the intensity of the long-range correlations, it is also
possible to demonstrate the presence of two regio-isomers A and B
as shown in Scheme 2. The methylene protons Ha and Ha0 give
indeed strong correlations with both quaternary carbons C1 and C2.
Since the intensity of 1He13C long-range correlations are directly
related to the n-bond coupling constants for a given proton and
a given delay of the HMBC spectrum,33,34 it means that these cross-
peaks correspond both to 3JCH correlations.
The absence of benzodioxane derivative during the gallic acid
glycidylation prompted us to investigate in greater details the
source of the difference in reactivity of this phenolic compound
towards epichlorohydrin. For this purpose, two phenolic com-
pounds, protocatechuic acid 6 and pyrogallol 7, whose chemical
structures are close to that of gallic acid, were reacted with epi-
chlorohydrin using the procedure described in Scheme 3.
2.2. Glycidylation of gallic acid
Gallic acid 4 is a phenolic compound encountered in different
plants in various forms, as gallic acid derivatives (esters, glycosides)
or as the acyl groups of some polyols (glucose, quinic acid). Some
studies have reported the use of this phenolic compound in ther-
mosetting applications, especially as an adjuvant in the epoxy
2.3. Glycidylation of protocatechuic acid 6 and pyrogallol 7
The treatment of protocatechuic acid 6 by epichlorohydrin in the
presence of BnEt3NCl as PTC led to the triglycidylated derivative 8
as the main product (60% yield) along with a small amount (9%

1348
C. Aouf et al. / Tetrahedron 69 (2013) 1345e1353
Scheme 3. Glycidylation of 4, 6 and 7 in the presence of PTC followed by basic treatment.
yield) of benzodioxane derivative 9. The structures of products 8
and 9 were determined by NMR spectroscopy. As explain below for
products 3A and 3B, the position of the dioxane ring of product 9
was determined from the long range 1He13C correlations, which
showed that only one regio-isomer was formed as displayed in
Scheme 3.
the glycidylation reaction was carried out with or without BnEt3NCl
and stopped prior to the addition of the base solution.
Thus, gallic acid 4 and epichlorohydrin were mixed in the
presence of BnEt3NCl. Initially, the solid gallic acid was observed to
be only slightly soluble in epichlorohydrin but after 1 h at 100 ꢀC,
a
homogenous solution was obtained. After epichlorohydrin
Similarly, pyrogallol 7 was reacted with epichlorohydrin to give
a mixture of two products, which could not be separated. NMR
analyses of this mixture revealed the presence of the triglycidy-
lether of pyrogallol 10 and the benzodioxane derivative 11 in
equivalent amounts (Scheme 3). NMR data of product 11 showed
that only one regio-isomer was formed.
These results suggest that the presence or absence of the car-
boxyl group on the phenolic compound influences strongly the
distribution of the products obtained as demonstrated by the re-
action of pyrogallol that lacks this substituent and led to the for-
mation of the benzodioxane derivative, in amounts comparable to
those obtained with 4-methylcatechol 1. In order to better un-
derstand the reactivity of these compounds, the glycidylation re-
action was investigated step by step.
evaporation and silica gel filtration, the residue was analysed by
NMR spectroscopy, which revealed that products 12a, 12b and 12c
were formed (Scheme 4). These products are evidence of the re-
action process: a nucleophilic attack of the phenolate ions onto the
electrophilic methylene carbon of the epichlorohydrin oxirane,
thereby causing the opening of the oxirane ring to occur.
In the absence of BnEt3NCl, the substrate was entirely recovered
without any detectable traces of any other products, suggesting
complete unreactivity of phenolic compounds with epichlorohy-
drin in the absence of PTC. These results point out the crucial role of
the catalyst in the formation of phenolate ions. This reaction system
is called the solideliquid phase transfer catalysis. Yang and Wu42
proposed a conceptual scheme describing the reaction mecha-
nism in solideliquid phases. In their system, O-phthalic acid (ArH2)
was reacted with potassium hydroxide (KOH) to produce potassium
phthalate ArK2.
Without the addition of any water, phase transfer catalyst QX
can solvate and react with solid ArK2 to produce ArQ2 dissolved in
organic phase. Then, the active intermediate ArQ2 (org) reacts with
the organic substrate RX to form the desired product ArR2. This
mechanistic model could be applied to our reactional system with
the only difference being that gallic acid hydroxyls are able to form
ion-pairs with PTC even in absence of base.
2.4. Investigation of the glycidylation reaction mechanism
The O-alkylation of alcohols or phenolic compounds in the
presence of PTC such as ammonium salt was widely described in
the literature.38e41 Nevertheless, in most studies, the phenolic
compound is dissolved in an organic solvent and then an aqueous
solution of base such as NaOH is added in the presence of PTC. In
this aqueous-organic two-phase system, the reacting nucleophile is
provided by the aqueous phase and is usually insoluble or slightly
soluble in the organic phase under operating conditions.41
In contrast, our glycidylation reaction proceeds in two steps: the
first step consists in heating a suspension of solid phenolic com-
pound (nucleophile) and liquid epichlorohydrin (electrophile) at
100 ꢀC in the presence of PTC for 1 h. In the second step, an aqueous
solution of NaOH containing PTC is added to the previous reaction
medium. In order to clarify the role of the phase transfer catalyst,
The overall reaction of solid gallic acid ArOH and organic epi-
chlorohydrin RX under the action of phase transfer catalyst QX,
could be formally expressed as follows:
ArOHðsolÞ þ RXðorgÞ þ QXðorgÞ/ArORðorgÞ þ HXðorgÞ þ QXðorgÞ
The role of PTC being better clarified, the reaction of pyrogallol 7
with epichlorohydrin and BnEt3NCl was also performed.

C. Aouf et al. / Tetrahedron 69 (2013) 1345e1353
1349
Scheme 4. Reaction of 4, 7 and 14 with epichlorohydrin in the presence of PTC. Products analyses were done before the addition of the base solution.
A detailed structural characterization of the mixture recovered
after 1 h at 100 ꢀC was performed using several NMR spectroscopy
experiments (see Experimental section).
The abundantion at m/z 275, which occurred after30min revealsthat
the production of the benzodioxane derivative requires more time.
The substitution of phenolic hydroxyl proton by a methyloxirane
group (m/z 389) was observed at the end of the reaction (Fig. 3).
According to these results, we noticed that the number and
the nature of substituents on the phenyl ring influence the re-
activity of phenolic compounds towards epichlorohydrin and
the regio-selectivity of this reaction. For this reason, the O-
glycidylation reaction of resorcinol 14, which possesses two
phenolic hydroxyl groups at meta position from each other was
studied. Applying the same experimental procedure used for
gallic acid and pyrogallol, the reaction of resorcinol with epi-
chlorohydrin in the presence of PTC led to product 15 in 81%
yield (Scheme 4).
Although the phenolic hydroxyls of resorcinol are equivalent,
they react in different way towards epichlorohydrin (two different
substituents). This result proves unequivocally that, in reaction
mixture, a permanent competition exists between two different
mechanisms (Scheme 5). In the mechanism A, the direct nucleo-
philic substitution (SN2) of phenolate ion (ArOꢁ) on epichlorohy-
drin occurs, concomitantly with the cleavage of CeCl bond. In the
mechanism B, the attack of ArOꢁ causes the opening of epichlo-
rohydrin ring followed by the protonation of the resulting alco-
holate 16 formed in situ. Thus, both mechanisms A and B are
involved in the glycidylation reaction of resorcinol and pyrogallol.
However, only methylchlorohydrin derivatives (12ae12c) have
been generated by the reaction of gallic acid with epichlorohydrin
revealing the prevalence of mechanism B in this case.
The 1D and 2D NMR spectra indicate the presence, in equivalent
amount, of two types of molecules, which are different in their ar-
omatic 1H spin systems. For one type of molecules, the aromatic ring
displays an AB spin systems (i.e., one triplet and one doublet rep-
resenting one and two protons, respectively), revealing symmetric
aromatic rings whereas the molecules of the other type exhibit ABC
spin systems (i.e., three multiplets, each representing one proton),
the signature of asymmetric aromatic ring. Besides aromatic pro-
tons, NMR spectra show a number of aliphatic 1H and 13C resonance
signals. Most of them were assigned from NMR spectra allowing the
distinction of three different substituents: oxirane, dioxane and
chlorohydrin groups. The linkage position of these groups on the
aromatic rings was then determined using 1He13C long range cor-
relations between the eCH2eOe protons of the substituent groups
and quaternary carbons of the aromatic rings as shown in Fig. 2. We
thus found that, in symmetric aromatic rings, both hydroxyls 1 and 3
are substituted, by either oxirane or chlorohydrin groups. The hy-
droxyl 2 position is also partly substituted but only byoxirane group.
The asymmetric aromatic rings are characterized by the presence of
a dioxane group, linked to two adjacent hydroxyl positions of the
phenyl ring. The protons CH2eO of the dioxane groups give corre-
lations only with C2B (and not C3B) showing that only one regio-
isomer was formed as displayed in Fig. 2. The third hydroxyl posi-
tion is substituted by either an oxirane or a chlorohydrin group.
These products 13ae13d are shown in Scheme 4 (Fig. 2).
Besides NMR analyses, the kinetics of the pyrogallol O-glycidy-
lation reaction was monitored and products were analysed by
a tandem UPLC/ESI-MS in positive mode (Fig. 3).
From gallic acid, the mixture obtained (12ae12c) indicates that
the glycidylation occurred first in carboxylic function, which is
transformed into ester.
The three substituents (oxirane, dioxane and chlorohydrin) iden-
tified byNMR analyses, successivelyappearedinthereactionmedium
beginning by the methylchlorohydrin derivative (m/z 219 and 311).
Thus, we can deduce that the presence of this electron-with-
drawing group on phenolic ring may promote mechanism B
(Scheme 5).

1350
C. Aouf et al. / Tetrahedron 69 (2013) 1345e1353
3. Conclusions
In our ongoing research program, the feasibility to substitute
bisphenol A by natural phenolic compounds in the synthesis of
epoxy resins pre-polymers is studied. The results of the present
work show the versatility of two phenolic compounds, 4-
methylcatechol and gallic acid, to react with epichlorohydrin.
The mechanism of O-glycidylation of several model phenolic
compounds was investigated enabling us to establish a relation-
ship between the chemical structures of phenolic compounds
and their behaviour towards epichlorohydrin. An understanding
at molecular-level of the reactivity of phenolic monomers to-
wards glycidylation constitutes an essential step in the de-
velopment of bio-based epoxy resins using tannin polymers as
phenolic sources.
4. Experimental section
4.1. General
Gallic acid (97.5%), 4-methylcatechol (ꢂ95.0%), resorcinol
(ꢂ99.0%), protocatechuic acid (ꢂ97.0%), pyrogallol (ꢂ99.0%), epi-
chlorohydrin (99.0%), BnEt3NCl (ꢂ98.0%) and sodium hydroxide
(ꢂ98.0%) were purchased from SigmaeAldrich France.
All the reactions were monitored by TLC performed on silica
gel 60 F254. NMR spectra were acquired on VARIAN Unity-Inova
500 MHz spectrometer (Varian NMR instruments, Palo Alto,
CA), operating at 500.05 MHz for 1H and 125.75 MHz for 13C,
using a 5 mm indirect detection Z-gradient probe. All samples
were dissolved in DMSO-d6. The chemical shifts were reported to
that of internal DMSO-d6, at 2.5 ppm for 1H and 39.5 ppm for 13C.
Assignments of both proton and carbon resonances, identifica-
tion and structure characterization of products were performed
using several of NMR experiments: Homonuclear 1H 1D and 2D
(DOSY, COESY, ROESY and 1D selective TOCSY) and heteronuclear
1He13C 2D (HSQC and HMBC) experiments. Spectra were pro-
cessed and analysed using either VNMRJ or ACD/Labs software.
ESI-MS analyses in positive mode were performed using
a Brucker Daltonics Ion trap mass spectrophotometer (Bremen,
Germany) coupled to Reversed-Phase Ultra Performance Liquid
Chromatography (UPLC). Samples (5 mL) were directly injected to
the UPLC system coupled to ESI-MS. The liquid chromatography
system was an Acquity UPLC (Waters, Milford, MA) equipped
with a Photodiode Array Detector. The column was a BEH C8
1.0ꢃ50 mm, 1.7
m
m. The flow rate was 0.30 mL minꢁ1 and the
gradient conditions were solvent A (H2O); solvent B (CH3CN);
initial 0.1% B; 0e3 min, 30% B linear; 3e6 min, 60% B linear;
6e8 min, 99.9% B linear.
Fig. 2. HMBC spectrum of compounds 13a and 13d in DMSO-d6 showing correlations
between methylene protons of the substituent groups and quaternary carbons of the
aromatic rings.
The predominance of mechanism B was also observed by Pchelka
et al.43 during the synthesis of 3-aryloxy-1,2-epoxypropanes. In-
deed, 1-chloro-3-aryloxypropan-2-ols 17 were mainly obtained
when the phenol substituents were electron withdrawing groups.
On the other hand, the study of the O-glycidylation mechanism
of these different phenolic compounds allowed us to determine the
origin of the formation of benzodioxane derivative. Indeed, the
oxirane ring introduced via mechanism A in the first substitution
step (in the case of 4-methylcatechol 1 and pyrogallol 7) undergoes
an intramolecular nucleophilic attack from the adjacent phenolate
anion to yield this cyclic by-product. In contrast, the glycidylation of
gallic acid exclusively implies the mechanism B, thus preventing
the production of benzodioxane derivatives despite the presence of
ortho phenolic hydroxyls.
4.2. General procedure for the glycidylation of phenolic
compounds
A 100 mL two-necked flask equipped with a condenser, a sep-
tum cap and magnetic stirring bar was charged with 5 mmol of
phenolic compound in epichlorohydrin (4 M equiv/OH), the sus-
pension was heated at 100 ꢀC and benzyltriethylammonium chlo-
ride (0.05 M equiv/substrate) was added. After 1 h, the solution was
cooled to 30 ꢀC and the aqueous solution of NaOH 20 wt %
(2 M equiv/OH) with 0.05 M equiv of phase transfer catalyst
(BnEt3NCl) were added. The mixture was stirred vigorously for
90 min. Water (30 mL) was added to the reaction mixture and the
aqueous phase was extracted with 3ꢃ30 mL of ethyl acetate. The
organic phase was washed with 40 mL of brine then dried over
MgSO4 and vacuum concentrated. Crude products were purified by
silica gel chromatography using petroleum ether/ethyl acetate (PE/
EA) system.
Finally, alcoholate 16 formed during the first step of O-glycidy-
lation underwent the NaOH assisted intramolecular cyclization to
give glycidylated derivatives.

C. Aouf et al. / Tetrahedron 69 (2013) 1345e1353
1351
Fig. 3. Positive mode ESI-MS mass spectra of pyrogallol O-glycidylation.
4.2.1. 2,20-(4-Methyl-1,2-phenylene)bis(oxy)bis(methylene)dioxirane
130.34 (C4), 145.57 (C1), 147.61 (C2) ppm. HRMS calcd for C13H17O4
2. PE/EA, 70:30, colourless oil, 1.7 mmol, 34% yield. 1H NMR
[MþH]þ: 237.1127, found 237.1119.
(500 MHz, DMSO-d6)
d
¼2.22 (s, 3H, Me), 2.69 (m, 2H, a0,d0), 2.83 (m,
2H, a,d), 3.32 (m, 2H, b,c), 3.78 (m, 1H, f0), 3.81 (m, 1 H, c0), 4.25 (dd,
J¼11.5, 2.7 Hz, 1H, f), 4.29 (dd, J¼11.4, 2.5 Hz, 1H, c), 6.69 (dd, J¼8.1,
1.2 Hz, 1H, H5), 6.82 (d, J¼1.2 Hz, 1H, H3), 6.87 (d, J¼8.1 Hz, 1H,
4.2.2. (6-Methyl-2,3-dihydrobenzo[b][1,4]dioxin-2-yl)methanols 3A
and 3B. PE/EA, 70:30, colourless oil, 1.66 mmol, 33% yield. 1H NMR
(500 MHz, DMSO-d6)
d
¼2.17 (s, 3H, Me), 3.59 (m, 2H, c), 3.94 (ddd,
H6) ppm. 13C NMR (125 MHz, DMSO-d6)
d
¼20.31 (Me), 43.5 (2C,
J¼11.3, 7.4, 1.9 Hz, 1H, a0), 4.08 (ddd, J¼7.4, 6.4, 1.9 Hz, 1H, b), 4.26
a,d), 49.62 (2C, b,e), 69.7 (2C, c,f), 114.1 (C6), 114.82 (C3), 121.22 (C5),
(m, 1H, a), 5.02 (t, J¼5.7 Hz, 1H, OH), 6.60 (m, 1H, H5), 6.65 (d,
Scheme 5. Mechanisms of phenolic compounds O-glycidylation.

1352
C. Aouf et al. / Tetrahedron 69 (2013) 1345e1353
J¼2.4 Hz, 1H, H3), 6.72 (dd, J¼8.1, 2.4 Hz, 1H, H6) ppm. 13C NMR
4.3. Reactions of phenolic compounds with epichlorohydrin
in the presence of benzyltriethylammonium chloride
(125 MHz, DMSO-d6)
d
¼20.38, 20.41 (Me), 60.01, 60.06 (c), 65.18,
65.32 (b), 73.66, 73.75 (a), 116.64, 116.80 (C6), 117.28, 117.44 (C3),
121.59, 121.91 (C5), 130.17, 130.51 (C4), 140.9, 140.97 (C1), 142.79,
142.87 (C2) ppm. HRMS calcd for C10H13O3 [MþH]þ: 181.0865,
found 181.0857.
A mixture of phenolic compounds (10 mmol), epichlorohydrin
(4 M equiv/OH) and benzyltriethylammonium chloride (0.05 M
equiv/substrate) was stirred at 100 ꢀC for 1 h in a 100 mL flask
equipped with a reflux condenser. The reaction mixture was cooled
to room temperature and thereto was added 40 mL of hexane to
remove the excess of epichlorohydrin. This operation was repeated
twice. The crude product was then filtrated over silica using ethyl
acetate to yield the following products:
4.2.3. 3,4,5-Triglycidylether glycidyl benzoate 5. PE/EA, 40:60e30:70,
colourless oil, 3.4 mmol, 68% yield. 1H NMR (500 MHz, DMSO-d6)
d
¼2.62e2.86 (m, 8H, aa0,dd0,gg0), 3.29e3.38 (m, 4H, b,e,h), 3.90e3.94
(m, 3H, f0,i0), 4.08 (dd, J¼12.4, 6.3 Hz, 1H, c0), 4.26 (qd, J¼11.7, 2.9 Hz,
1H, i), 4.44 (dd, J¼11.3, 2.1 Hz, 2H, f), 4.63 (dd, J¼12.4, 2.5 Hz, 1H, c),
7.30 (s, 2H, H2,H6) ppm. 13C NMR (125 MHz, DMSO-d6)
d¼43.4 (a),
4.3.1. Mixture of 3-chloro-2-hydroxypropyl 4-(3-chloro-2-hydroxypr-
opoxy)-3,5-dihydroxybenzoate 12a (32% yield), 3-chloro-2-
hydroxypropyl 3,4-bis(3-chloro-2-hydroxypropoxy)-5-hydroxybenzoate
12b (23% yield) and 3-chloro-2-hydroxypropyl 3,4,5-tris(3-chloro-2-
hydroxypropoxy)benzoate 12c (15% yield). Brown oil. 1H NMR
43.6 (2C, d), 43.8 (g), 49.0 (b), 49.7 (2C, e), 50.1 (h), 65.5 (c), 70.1 (2C,
f), 74.0 (i), 108.4 (2C, C2, C6), 124.6 (C1), 141.6 (C4), 151.8 (2C, C3,C5),
164.0 (COO) ppm. HRMS calcd for C19H23O9 [MþH]þ: 395.1342,
found 395.1341.
(500 MHz, DMSO-d6)
d
¼3.75 (m, 18H, cc0,kk0,ff0), 4.01 (m, 23H,
4.2.4. Oxiran-2-ylmethyl 3,4-bis(oxiran-2-ylmethoxy)benzoate 8. PE/
aa0,gg0,dd0,e,h and b), 4.23 (m, 4H, aa0), 5.59 (m, 7H, OH), 7.01 (s, 1H,
EA, 40:60, colourless oil, 3 mmol, 60% yield. 1H NMR (500 MHz,
H
Ar(A)), 7.10 (d, J¼1.8 Hz, 1H, H6(B)), 7.20 (d, J¼1.9 Hz, 1H, H2(B)), 7.31
DMSO-d6)
d
¼2.72e2.87 (m, 6H, aa0,dd0,gg0), 3.32e3.40 (m, 3H, b,e,h),
(s, 1H, HAr(C)), 9.51 (s, 2H, OH(A)), 9.64 (s, 1H, OH(B)) ppm.
3.88 (dd, J¼11.3, 6.6 Hz, 1H, f0), 3.94 (dd, J¼11.4, 6.5 Hz, 1H, c0), 4.05
(dd, J¼12.4, 6.4 Hz, 1H, i0), 4.42 (m, 2H, c,f), 4.60 (dd, J¼12.3, 2.4 Hz,
1H, i), 7.12 (d, J¼8.5 Hz, 1H, H5), 7.51 (d, J¼1.8 Hz, 1H, H2), 7.61 (dd,
J¼8.5, 1.8 Hz, 1H, H6) ppm. 13C NMR (125 MHz, DMSO-d6)
4.3.2. Mixture of 13ae13d. Brown oil, 77% total yield. 1H NMR
(500 MHz, DMSO-d6)
d
¼2.59 (m, f0), 2.68 (m, o0), 2.72 (m, c0),
2.74e2.76 (m, f), 2.82e2.83 (m, o), 2.83e2.84 (m, c), 3.28 (m, e),
3.31 (m, n), 3.33 (m, b), 3.59e3.64 (m, ll0), 3.65e3.77 (m, ii0 and
rr0), 3.78e3.80 (m, m0), 3.79e3.83 (m, d0), 3.85e3.87 (m, a0),
3.93e3.94 (m, pp0), 3.95e3.97 (m, j0), 3.97e3.98 (m, gg0),
4.01e4.05 (m, h and q), 4.10e4.13 (m, d), 4.12 (m, k), 4.24e4.27 (m,
m), 4.30e4.33 (m, a), 4.32e4.34 (m, j), 5.04 (m, OH dioxane), 5.51
(m, OH chlorohydrin), 6.49e6.51 (m, H4B), 6.54e6.57 (m, H6B),
6.68e6.71 (m, (H4-H6)A), 6.72e6.74 (m, H5B), 6.95e6.99 (m, H5A).
d
¼43.6e43.8 (3C, a,d,g), 49.0 (h), 49.5e49.7 (2C, b,e), 65.2 (i),
69.7e70.0 (2C, c,f), 112.7 (C5), 114.0 (C2), 121.9 (C1), 123.7 (C6), 147.5
(C3),152.3 (C4),165.0 (COO) ppm. HRMS calcd for C16H19O7 [MþH]þ:
323.1131, found 323.1141.
4.2.5. Oxiran-2-ylmethyl 3-(hydroxymethyl)-2,3-dihydrobenzo[b][1,4]
dioxine-6-carboxylate 9. PE/EA, 40:60, colourless oil, 0.45 mmol, 9%
yield. 1H NMR (500 MHz, DMSO-d6)
d
¼2.71 (dd, J¼5.0, 2.6 Hz,1H, a0),
13C NMR (125 MHz, DMSO-d6)
d
¼43.67e43.78 (c,f,o), 46.81 (i,r),
2.83 (t, J¼4.6 Hz, 1H, a), 3.31 (m, 1H, b), 3.64 (td, J¼11.0, 5.4 Hz, 2H,
dd0), 4.05 (dd, J¼12.3, 6.3 Hz, 1H, c0), 4.10 (dd, J¼11.0, 4.7 Hz, 1H, f0),
4.20 (m,1H, e), 4.41 (dd, J¼11.5, 2.2 Hz,1H, f), 4.58 (dd, J¼12.3, 2.6 Hz,
1H, c), 5.10 (t, J¼5.5 Hz, 1H, OH), 6.99 (d, J¼8.6 Hz, 1H, H5), 7.43 (d,
J¼1.9 Hz 1H, H2), 7.48 (dd, J¼8.6, 1.9 HZ, 1H, H6) ppm. 13C NMR
49.74 (b,e,n), 59.83 (l), 64.86 (j), 68.63 (h,q), 69.15e69.89
(a,d,g,m,p), 73.46 (k), 105.72e105.92 (C6 B), 107.12e107.43
((C4eC6) A), 109.96 (C4B), 120.06 (C5B), 123.74 (C5A),
133.98e132.12 (C2 B), 137.22 (C2 A), 147.77e143.8 ((C1eC3) B),
151.9e152.27 ((C1eC3) A) ppm.
(125 MHz, DMSO-d6)
d
¼43.82 (a), 49.05 (b), 59.69 (d), 65.13 (c) 65.51
(f), 73.59 (e), 117.12 (C5), 118.06 (C2), 122.40 (C1), 122.82 (C6), 142.88
(C3), 147.57 (C4), 164.82 (C00) ppm. HRMS calcd for C13H15O6
[MþH]þ: 267.0869, found 267.0861.
4.3.3. 1-Chloro-3-(3-(oxiran-2-ylmethoxy)phenoxy)propan-2-ol
15. Brown oil, 8.1 mmol, 81% yield. 1H NMR (500 MHz, DMSO-d6)
d
¼2.69 (m,1H, c0), 2.83 (m,1H, c), 3.31 (dd, J¼6.6, 3.3 Hz,1H, b), 3.66
(dd, J¼11.1, 5.4 Hz, 1H, f0), 3.74 (dd, J¼11.1, 4.6 Hz, 1H, f), 3.80 (dd,
J¼11.4, 6.6 Hz, 1H, a0), 3.96 (m, 2H, d,d0), 4.02 (dd, J¼10.3, 5.1 Hz, 1H,
e), 4.30 (d, J¼11.4 Hz, 1H, a), 5.54 (d, J¼5.2 Hz, 1H, OH), 6.54e6.56
(m, 3H, H2, H4 and H6), 7.18 (t, J¼8.4 Hz, 1H, H5) ppm. 13C NMR
4.2.6. 2,20,200-(Benzene-1,2,3-triyltris(oxy))tris(methylene)trioxirane
10. PE/EA, 40:60, colourless oil, 33% yield. 1H NMR (500 MHz,
DMSO-d6)
d
¼2.59e2.83 (m, 6H, aa0, dd0), 3.29e350 (m, 3H, b,e),
3.82 (m, 1H, f0), 3.87 (m, 2 H, c0), 4.12 (m, 1H, f), 4.32 (m, 2H, c),
(125 MHz, DMSO-d6)
d¼43.4 (c), 46.4 (f), 49.3 (b), 66.59 (e), 68.96
6.69 (d, J¼8.4 Hz, 2H, H4 and H6), 6.74 (t, J¼8.4 Hz, 1H, H5) ppm.
(2C, c and d), 101.27 (C2), 106.8 (2C, C6 and C4), 130.02 (C5), 159.41
(C1), 159.57 (C3) ppm. HRMS calcd for C12H16ClO4 [MþH]þ:
259.0737, found 259.0735.
13C NMR (125 MHz, DMSO-d6)
d
¼43.65 (2C, a), 43.50 (1C, d),
49.78 (2C, b), 50.26 (1C, e), 69.75 (2C, c), 73.95 (1C, f), 107.43
(2C, C4 and C6), 123.75 (C5), 137.31 (C2), 152.39 (2C, C1 and
C3) ppm. HRMS calcd for C15H19O6 [MþH]þ: 295.1123, found
295.1136.
Supplementary data
Supplementary data associated with this article can be found in
the online version, at http://dx.doi.org/10.1016/j.tet.2012.11.079.
4.2.7. (5-(Oxiran-2-ylmethoxy)-2,3-dihydrobenzo[b][1,4]dioxin-2-yl)
methanol 11. PE/EA, 40:60, colourless oil, 33% yield. 1H NMR
(500 MHz, DMSO-d6)
d
¼2.68e2.83 (m, 2H, gg0), 3.31 (m, 1H, h),
References and notes
3.62 (m, 2H, jj0), 3.78 (dd, J¼11.3, 6.6 Hz, 1H, i0), 3.96 (dd, J¼11.2,
2.1 Hz, 1H, l0), 4.10 (dd, J¼5.7, 2.8 Hz, 1H, k), 4.26 (dd, J¼11.3, 2.6 Hz,
1H, i), 4.33 (m, 1H, l), 5.05 (t, J¼5.6 Hz, 1H, OH), 6.50 (dd, J¼8.3,
1 Hz, 1H, H40), 6.54 (dd, J¼8.3, 1 Hz, 1H, H60), 6.94 (t, J¼8.3 Hz, 1H,
1. Biopolymers from Renewable Resources; Kaplan, D. L., Ed.; Springer: Berlin, 1998;
pp 1e26.
2. Mohanty, A. K.; Misra, M.; Hinrichsen, G. Macromol. Mater. Eng. 2000, 276, 1e24.
3. Feldman, D. In Composites, Thermosetting Polymers; Salamone, J. C., Ed.; CRS: FL,
1996; pp 277e278.
4. Gourichon, B.; Deleglise, M.; Binetruy, C.; Krawczak, P. Comp. Part A: Appl. Sci.
Manuf. 2008, 39, 46e55.
5. Guilleminot, J.; Comas-Cardona, S.; Kondo, D.; Binetruy, C.; Krawczak, P. Com-
H50) ppm. 13C NMR (125 MHz, DMSO-d6)
d
¼43.78 (1C, g), 49.70
(1C, h), 59.83 (1C, j), 64.86 (1C, l), 69.94 (1C, i), 73.52 (1C, k), 105.73
(C60), 109.94 (C40), 120.06 (C50), 132.98 (C20), 143.97 (C30), 147.82
(C10) ppm. HRMS calcd for C12H15O5 [MþH]þ: 239.0919, found
239.0912.
pos. Sci. Technol. 2008, 68, 1777e1786.
6. Krawczak, P.; Pabiot, J. J. Compos. Mater. 1995, 29, 2230e2253.

C. Aouf et al. / Tetrahedron 69 (2013) 1345e1353
1353
7. Yousefi, A.; Lafleur, P. G.; Gauvin, R. Polym. Compos. 1997, 18, 157e168.
8. Gandini, A. In Epoxy Polymers Based on Renewable Resources in Epoxy Polymers:
New Materials and Innovations; Pascault, J. P., Williams, R. J. J., Eds.; Wiley-VCH:
Weinheim, 2010.
9. Epoxy Resins, Chemistry and Technology; May, C. A., Ed.; Marcel Dekker: New
York, NY/Basel, 1988.
10. Bauer, R. S. Epoxy Resin Chemistry, Advances in Chemistry Series; American
Chemical Society: Washington, DC, 1979.
11. Gore, A. C. In Endocrine-Disrupting Chemicals: From Basic Research to Clinical
Practice (Contemporary Endocrinology); Totowa, NJ: Humana Press, 2007;
pp 100e125.
24. Zhao, B. Y.; Chen, G.; Liu, Y.; Hu, K.; Wu, R. J. J. Mater. Sci. Lett. 2001, 20,
859e862.
25. Haslam, E.; Cai, Y. Nat. Prod. Rep. 1994, 11, 41e66.
26. Quideau, S.; Deffieux, D.; Douat-Casassus, C.; Pouysegu, L. Angew. Chem., Int. Ed.
2011, 50, 586e621.
27. Swain, T.; Bate-Smith, E. C. In Comparative Biochemistry; Mason, H. S., Florkin,
H. M., Eds.; Academic: New York, NY, 1962; Vol. 3, pp 755e809.
28. Balasundram, N.; Sundram, K.; Samman, S. Food Chem. 2006, 99, 191e203.
29. Bravo, L. Nutr. Rev. 1998, 56, 317e333.
30. Porter, L. J. In Tanins; Harborn, J. B., Ed.; Academic: London, 1989; Vol. 1,
pp 389e419.
12. O’Connor, J. C.; Chapin, R. E. Pure Appl. Chem. 2003, 75, 2099e2123.
13. Okada, H.; Tokunaga, T.; Liu, X.; Takayanagi, S.; Matsushima, A.; Shimohigashi,
Y. Environ. Health Perspect. 2008, 116, 32e38.
14. vom Saal, F. S.; Myers, J. P. J. Am. Med. Assoc. 2008, 300, 1353e1355.
15. Guner, F. S.; Yagci, Y.; Erciyes, A. T. Prog. Polym. Sci. 2006, 31, 633e670.
16. Meier, M. A. R.; Metzger, J. O.; Schubert, U. S. Chem. Soc. Rev. 2007, 36,
1788e1802.
17. Puglisi, J. S.; Chaudhari, M. A. In Epoxies (EP); Epel, J. N., Margolis, J. M., Newman,
S., Seymour, R. B., Eds.; Engineered materials handbook; Metals Park, OH: ASM
International, 1988; Vol. 2, pp 240e241.
18. Aciero, D.; Russo, P.; Savarese, R. Conference on Polymer Progress Society,
Salerno, Italy, 2008; PPS-24.
31. Nouailhas, H.; Aouf, C.; Le Guerneve, C.; Caillol, S.; Boutevin, B.; Fulcrand, H. J.
Polym. Sci., Part A: Polym. Chem. 2011, 49, 2261e2270.
32. Tomita, H.; Yonezawa, K. Epoxy resin and process for preparing the same. U.S.
Patent 4,540,802, 1985.
33. Ohuchi, M. JOEL News 1988, 24A, 2e7.
34. Liu, X.; Cui, Y. Molecules 2007, 12, 1117e1124.
35. Hayashi, M.; Tsuneta, K.; Sasaki, H.; Takada, H. Chelate-forming aqueous resin
composition, U. S. Patent 4067838, 1978.
36. Nagase, R.; Yasuda, A.; Kawashima, M. Epoxy resin/imidazole/polyphenol ad-
duct as epoxy curing agent. U.S. Patent 5,077,355, 1991.
37. Qureshi, S. P.; Gardner, H. C. Epoxy/aromatic amine resin systems containing
aromatic trihydroxy compounds as cure accelerators. U.S. Patent 4,593,056,
1986.
19. Takada, Y.; Shinbo, K.; Someya, Y.; Shibata, M. J. Appl. Polym. Sci. 2009, 113,
479e484.
38. Kishi, H.; Akamatsu, Y.; Noguchi, M.; Fujita, A.; Matsuda, S.; Nishida, H. J. Appl.
20. Chrysanthos, M.; Galy, J.; Pascault, J.-P. Polymer 2011, 52, 3611e3620.
21. Feng, X.; East, A. J.; Hammond, W. B.; Zhang, Y.; Jaffe, M. Polym. Adv. Technol.
2011, 22, 139e150.
22. Lora, J. H.; Glasser, W. G. J. Polym. Environ. 2002, 10, 39e48.
23. Sun, G.; Sun, H.; Liu, Y.; Zhao, B.; Zhu, N.; Hu, K. Polymer 2007, 48,
330e337.
Polym. Sci. 2011, 120, 745e751.
39. Senthamizh Selvi, R.; Nanthini, R.; Sukanyaa, G. IJSTR 2012, 1, 61e63.
40. Yadav, G. D.; Lande, S. V. Appl. Catal. A: Gen. 2005, 287, 267e275.
41. Yadav, G. D.; Reddy, C. A. Ind. Eng. Chem. Res. 1999, 38, 2245e2253.
42. Yang, H. M.; Wu, H. E. Ind. Eng. Chem. Res. 1998, 37, 4536e4541.
43. Pchelka, B. K.; Loupy, A.; Petit, A. Tetrahedron 2006, 62, 10968e10979.