Buffer-induced, selective mono-C-alkylation of phloroglucinol: Application to the synthesis of an advanced intermediate of catechin

Article abstract of DOI:10.1016/j.tet.2004.06.026

A straightforward mono-selective and C-specific alkylation of phloroglucinol with activated alkyl halides is presented. The use of water as solvent limits the amount of over-alkylated by-products. Provided some minor changes in the experimental conditions, hydrophobic cinnamyl halides can also be reacted, thus giving a direct access to advanced intermediates of natural flavonoids.

Full text of DOI:10.1016/j.tet.2004.06.026

Tetrahedron 60 (2004) 6807–6812
Buffer-induced, selective mono-C-alkylation of phloroglucinol:
application to the synthesis of an advanced intermediate
of catechin
a
Arnaud Gissot,a,b Alain Wagnera and Charles Mioskowski
,
,
*
*
a
` ´ ´
Laboratoire de Synthese bioorganique, Universite Louis Pasteur de Strasbourg, UMR 7514 du CNRS, Faculte de Pharmacie,
74 route du Rhin-BP 24-F-67401, Illkirch-Graffenstaden, France
bThe Scripps Research Institute, MB 210, 10550 North Torrey Pines Rd, 92037 La Jolla, CA, USA
Received 16 January 2004; revised 17 May 2004; accepted 7 June 2004
Abstract—A straightforward mono-selective and C-specific alkylation of phloroglucinol with activated alkyl halides is presented. The use of
water as solvent limits the amount of over-alkylated by-products. Provided some minor changes in the experimental conditions, hydrophobic
cinnamyl halides can also be reacted, thus giving a direct access to advanced intermediates of natural flavonoids.
q 2004 Elsevier Ltd. All rights reserved.
1. Introduction
glucinol.4 These methods are multi-step and require careful
control of the reaction conditions. Most interestingly, the
selective C-alkylation of phloroglucinol has never been
reported, the only examples found in the literature
concerning reactions with protected versions of phloro-
glucinol.5 Though apparently simple, this transformation
offers several synthetic challenges. As phloroglucinol
contains six nucleophilic sites, an ideal reaction will be
mono- and C-specific. Undoubtedly, such a reaction would
constitute an easy and straightforward approach toward the
synthesis of natural unprotected flavonoids like catechin
whose retrosynthesis is depicted in the following scheme.
The flavonoids are a family of polyphenolic compounds
found in the plant kingdom. They feature interesting anti-
oxidative properties responsible for the health-benefits
associated with a diet rich in vegetables and are found in
large quantities in red wines and green teas.1 However, due
to the lack of reliable chemical methods for their synthesis,
only natural extracts with ill-defined chemical compositions
are usually evaluated for their biological activity. Undoubt-
edly, the poor overall stability of flavonoids is responsible
for the lack of general methods for their chemical synthesis.
Consequently, protecting-group free syntheses of poly-
phenols are scarce. The target molecules cannot often
withstand the conditions required for removal of most of the
conventional protecting groups. To the best of our knowl-
edge, only benzyl derivatives have been used successfully
as protecting groups in the total synthesis of natural
flavonoids.2 The phloroglucinol motif (1,3,5-benzenetriol)
is ubiquitous in all natural flavonoids structures. As such, it
constitutes the ultimate starting material en route to the
synthesis of elaborated (un)natural polyphenols. Unfortu-
nately, the benzylation of phloroglucinol is not selective and
gives a mixture of both O- and C-benzylated products.3
Hence, many efforts have been aimed at methods to achieve
the specific mono-, di-, or tri-O-benzylation of phloro-
Our approach to catechin is very concise and involves the
intramolecular ring closing between a phenol of the
phloroglucinol core and the epoxide as the last step. This
epoxide intermediate is ideally synthesized in two steps
starting from the selective alkylation of phloroglucinol
followed by the epoxidation of the double bond. In this
communication, we show that phloroglucinol can be
C-specifically and mono-selectively alkylated with various
activated alkyl halides in buffered aqueous solutions. The
use of buffered aqueous solutions as the reaction media
proved crucial to control the regio-selectivity of the
reaction.
Keywords: Flavonoids; Polyphenol compounds; Phloroglucinol;
Alkylation; Synthetic methodology.
2. Results and discussion
*
Corresponding authors. Tel.: þ33-3-88-67-68-63; fax: þ33-3-88-67-88-
The reaction of phloroglucinol with allyl bromide was
investigated first (Table 1, entries 1–6). Interestingly,
91; e-mail address: alwag@bioorga.u-strasbg.fr;
mioskowsi@bioorga.u-strasbg.fr
0040–4020/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2004.06.026

6808
A. Gissot et al. / Tetrahedron 60 (2004) 6807–6812
Scheme 1. Retrosynthetic scheme of catechin.
Table 1. Conditions tested for the alkylation of phloroglucinol in water
Entry
Base (equiv.)
x
R
Solvent
T (8C)
Time (h)
Yields (%)a
1
2
1
2
3
4
5
6
7
8
9
None
None
5
5
1.1
1.1
4
4
2
2
2
Allyl
Allyl
Allyl
Allyl
Allyl
Allyl
Benzyl
Benzyl
Benzyl
EtOH40%
H2O
H2O
H2O
H2O
H2O
H2O
EtOH
THF
20
20
20
20
20
10
20
20
20
72
72
12
5
4
10
4
36
68
45
30
23
17
NaOH (1)
NaOH (2)
Bufferc
Bufferc
Bufferc
NaOH (2)
TEA (3)
n.d.b
62
24
12
21
26
23
35
54
40
42
4
5
a
Isolated yields.
Not determined, see text.
Phosphate buffer 0.2 M, pH 7.8.
b
c
phloroglucinol is nucleophilic enough to undergo alkylation
in aqueous ethanol in the absence of base, with allyl
phloroglucinol 1a being obtained in 36% yield (entry 1). No
products resulting from the O-alkylation of phloroglucinol
are detected in the crude reaction mixture. However, the
reaction is not mono-selective and gives substantial
amounts of the di-alkylated product 2a. The extent of
mono-alkylation as well as the yield of the desired
compound 1a is greatly improved in pure water (entry 2).
The lower solubility of 2a in water in comparison to the
starting phloroglucinol may account for this better
selectivity.
conditions. Alternatively, use of a buffered, ca. neutral,
aqueous solution proved very successful. Since the pH is
maintained throughout the course of the reaction, the
reaction time is short and the yield of 1a remains quite
good (entry 5). Reducing the reaction temperature to 10 8C
not only slows product formation but also results in lower
selectivity and overall yield (entry 6).
Under the same conditions (0.2 M phosphate buffer, pH 7.8,
25 8C), the reaction of phloroglucinol with benzyl bromide
gave 1b and 2b in 54 and 21% yield, respectively. As for
allyl bromide (results not shown), no O-benzylated
phloroglucinol was obtained in ethanol or THF (entry 8
and 9),6 and the reaction must be carried out in DMF to
observe the formation of benzyl ethers (results not shown).7
Hence, water does not influence the C-selectivity of the
reaction. Indeed, both the phloroglucinol and activated alkyl
halides are known to favor C-alkylation.8,9 This is in sharp
contrast to the reaction of non-activated alkyl bromides with
phloroglucinol where water has a marked influence both on
the C- and on the mono-selectivity.10
Although 1a is obtained in good yield (68%, entry 2) in
water, the pH of the reaction medium becomes increasingly
acidic as the allyl phloroglucinol is formed. Since most of
the polyphenols are pH-sensitive, milder conditions are
required for the synthesis of more elaborate adducts. Using
stoichiometric amounts of sodium hydroxide, the phenolate
of phloroglucinol is formed and the reaction is faster (entry
3). Yet, the mono-alkylation selectivity is low. This can be
ascribed to the better solubility of the phenolate of 1a in
water in comparison to its neutral form. Consequently, 1a is
more likely to over-react in basic solutions and lesser mono-
selectivity is observed (entry 3). A complex mixture of
products is obtained with excess quantities of base (entry 4).
Partial migration of the double bond of 1a and 2a in
(conjugate) benzylic position takes place under these
The reaction of phloroglucinol with cinnamyl halides
(Table 2) readily gives access to the skeleton of natural
flavonoids (Scheme 1). Yet, these adducts are quite acid-
sensitive. This may in turn explain why the synthesis of
simple intermediates like 1c have only been reported in a
low 30% yield,9,11 or not achieved at all.12 The reaction of

A. Gissot et al. / Tetrahedron 60 (2004) 6807–6812
Table 2. Cinnamylation of phloroglucinol in aqueous solutions
6809
Entry
Conditions
x
Yields (%)
1c
1d
1
2
Phosphate buffer 0.2 M, pH 7.8
NaOH (1.25 equiv.), H2O/EtOH 15/85 (v/v)
2
1.25
35
69
35
Traces
cinnamyl halides is actually even more challenging. In
addition to the classical mono- and C-/O-alkylation issues
encountered so far with allyl and benzyl halides, cinnamyl
halides can give two additional regioisomers upon reaction
with a nucleophile. Indeed, both regioisomers 1c and 1d are
formed in equal amounts under our buffered reaction
conditions (Table 2, entry 1).
halide leading to 1c. These two opposing effects, hydro-
phobicity and solvation of the phenols, may account for the
absence of selectivity in water.
With this result in hand, we then synthesized the advanced
intermediate of catechin 1e starting from the highly
hydrophobic functionalized cinnamyl chloride 3 (Table 3).
This compound is efficiently synthesized in four steps
starting from the commercially available 3,4-dihydroxy-
benzaldehyde (Scheme 3).
We found the desired isomer 1c is cleanly obtained in a
mixture of ethanol and aqueous sodium hydroxide. Several
factors may account for this selectivity. First, the dielectric
constant of the solution decreases with increasing amounts
of ethanol. Hence, the dissociation of ion pairs is less likely
in solutions containing higher concentration of ethanol. The
sodium counter-ion of the phenolate and the chlorine of
cinnamyl chloride may be in close vicinity throughout the
reaction pathway in ethanol-rich aqueous solutions as in
transition state A, which leads to 1c (Scheme 2). However,
the strong hydrophobic effect and high cohesive energy
density expected in water alone as a solvent should favor
transition state B and thus produce 1d.13 Yet, the phenols
are also thoroughly solvated in water. This should favor the
attack on the less crowded terminal position of the alkyl
The cinnamyl chloride 3 was then reacted with the
phloroglucinol under a variety of aqueous ethanolic
conditions (Table 3).
In contrast to allyl bromide and cinnamyl chloride, the
substituted cinnamyl chloride 3 is not soluble in ethanol or
water. Consequently, no adduct is formed under the
conditions optimized with cinnamyl chloride (entry 1).
Under ultrasonic irradiation, the desired product 1e is
formed, yet not selectively (entry 2), the other regioisomer
1f being obtained in substantial amounts.14 When a solution
of 3 in a minimum of THF is added slowly to the
Scheme 2. Solvent-dependant transition states in the cinnamylation of phloroglucinol.

6810
A. Gissot et al. / Tetrahedron 60 (2004) 6807–6812
Table 3. Synthesis of an advanced intermediate of catechin
Conditions
x
Yields (%)
1e
1f
1
2
3
NaOH (1.3 equiv.), EtOH90% (vol.), 2 h
NaOH (1.3 equiv.), EtOH90% (vol.), 0.1 h
NaOH (2 equiv.), EtOH/H2O/THF 72/14/14, 20–50 8C, 9 h
1.25
1.25
0.33a
0
45
53
0
12
Traces
a
3 in a minimum of THF was added slowly to the reaction over a 4 h period.
3. Experimental
3.1. General
3.1.1. 2-Allylphloroglucinol (1a). To a solution of
phloroglucinol·2H2O (0.1 g, 0.61 mmol) in 10.98 mL
0.2 M Na2HPO4 and 1.02 mL of 0.2 M NaH2PO4 is added
the allyl bromide (0.21 mL, 2.44 mmol). The reaction is
stirred for 3 h and the aqueous phase is extracted twice with
ether. The combined organic layers are then washed with
brine, dried over Na2SO4, filtered and concentrated under
vaccum. The residue was purified by flash chromatography
on silica gel (25–35% AcOEt/hexane). C9H10O3 (166.06);
Scheme 3. Synthesis of the cinnamyl chloride 3. Reagents: (a) NaH, BnBr,
DMF; (b) NaH, triethylphosphonoacetate, THF; (c) LAH, Et2O, 215 8C;
(d) SOCl2, Net3, CH2Cl2, 0 8C.
1
Rf 0.45 (40% AcOEt/hexane); H (200 MHz, 1 CD3OD/
CDCl3): d¼3.34 (d, 3J¼5.2 Hz, 2H, H5), 5.00–5.14 (m, 2H,
H7), 5.85–6.07 (m, 1H, H6), 5.91 (s, 2H, H2). 13C (75 MHz,
1 CD3OD/CDCl3): d¼27.6, 96.2, 104.6, 116.0, 136.5,
153.5, 154.4. IR (neat): nmax¼3454 (br, OH), 2924 (Ar.),
1621 (Ar-O), 1452, 1212 (br), 1112 cm21. MS (NH4þ): m/z:
167 [MþH]þ.
phloroglucinol in aqueous ethanol under gentle heating,
only trace amounts of the undesired 1f are observed and 1e
is obtained pure in a satisfactory 53% yield after flash
chromatography (entry 3). This product has already been
synthesized in low yield (17%) from phloroglucinol and the
corresponding palladium p-acetate.8 This advanced inter-
mediate of natural flavonoids is obtained in our case in 53%
yield under very simple and straightforward conditions.
3.1.2. 2,4-Diallylphloroglucinol (2a). C12H14O3 (206.24);
Rf 0.7 (60% AcOEt/hexane); 1H (200 MHz, CDCl3):
d¼3.42 (m, 4H), 5.13–5.25 (m, 4H), 5.94–6.07 (m, 2H),
6.01 (s, 1H); 13C (75 MHz, CDCl3): d¼27.1, 95.8, 104.2,
115.5, 136.6, 155.5, 155.9; IR: nmax¼3492 (br, OH), 2924
(Ar.), 1623 (Ar-O), 1464, 1150 cm21; MS (NH4þ): m/z: 207
[MþH]þ.
In conclusion, a mild, high-yielding C-specific and mono-
selective alkylation of phloroglucinol with activated alkyl
halides has been developed. The C-specificity of this
reaction comes from the fact that both phloroglucinol and
activated alkyl halides favor C-alkylation. On the other
hand, the solubility of mono-C-adducts are limited in water,
thereby, preventing overalkylation reactions and giving
good mono-alkylation selectivity. With minor changes in
the experimental conditions, these aqueous conditions have
been utilized with highly hydrophobic substrates to afford
an advanced intermediate of catechin. The subsequent
epoxidation of the double bond of 1e has not been successful
so far due to the high reactivity of the phloroglucinol toward
oxidants. Yet, preliminary results indicate that iodination of
the double bond is feasible. We are now concentrating our
efforts on trying to make this reaction more selective and
then achieve a straightforward, protection-free, total
synthesis of catechin.
3.1.3. 2-Benzylphloroglucinol (1b). Same procedure as for
1a (benzyl bromide as the electrophile). C13H12O3 (216.23);
Rf 0.4 (60% AcOEt/hexane); 1H (200 MHz, 1 CD3OD/
CDCl3): d¼3.96 (s, 2H), 5.95 (s, 2H), 7.12–7.26 (m, 5H);
13C (50 MHz, 1 CD3OD/CDCl3): d¼28.4, 96.0, 106.5,
126.2, 128.2, 128.6, 140.3, 155.4, 155.8; IR (neat):
nmax¼3359 (br, OH), 2927 (Ar.), 1615 (br, Ar-O), 1456,
1144 cm21; MS (NHþ4 ): m/z: 217 (100) [MþH]þ, 234
(49.8) [MþNH4]þ, 251 (10.8) [MþNH3þNH4]þ.
3.1.4. 2,4-Dibenzylphloroglucinol (2b). C20H18O3
(306.36); Rf 0.7 (60% AcOEt/hexane); 1H (200 MHz,
CDCl3): d¼3.99 (s, 4H), 4.81 (s, 3H, OH), 5.99 (s, 1H),
7.12–7.30 (m, 10H); 13C (50 MHz, CDCl3): d¼28.9, 96.2,

A. Gissot et al. / Tetrahedron 60 (2004) 6807–6812
6811
106.7, 126.4, 128.2, 128.7, 139.8, 153.4, 154.1; IR (neat):
nmax¼3526 (br, OH), 3028, 2924 (Ar.), 1619 (br, Ar-O),
1447, 1186, 1039 cm21; MS (NHþ4 ): m/z: 307 (100)
[MþH]þ, 323 (54.6) [MþNH4]þ.
6.30–6.41 (m, 1H), 6.77–6.90 (m, 3H), 7.26–7.43 (m,
10H); MS (NHþ4 ): m/z: 472 [MþNH4]þ.
3.1.9. 3,4-Dibenzyloxybenzaldehyde (4). A first portion of
sodium hydride (60%, 7 g) is added to the 3,4-dihydroxy-
benzaldehyde (18.65 g, 135 mmol) in 300 mL of dry DMF.
After the evolution of gas had ceased, the benzyl bromide
(32.92 mL) is added dropwise on the solution. After 1 h, the
rest of the NaH is added (5.88 g, 297 mmol total). The
reaction is stirred for four additional hours and 600 g of ice
is added to the reaction mixture. The aqueous solution is
extracted three times with ether. The combined organic
layers are then washed with brine, dried over Na2SO4,
filtered and concentrated under vaccum. The brown solid
thus obtained is triturated with methanol to afford 36.4 g
(85%) of 4. C21H18O3 (318.37); Rf 0.55 (30% AcOEt/
3.1.5. 2-Cinnamylphloroglucinol (1c). The cinnamyl
chloride (7.62 mL, 54.00 mmol) is added to the
phloroglucinol·2H2O (7 g, 43.17 mmol) in solution in
200 mL of ethanol and 25 mL of 2 M NaOH. After 3 h,
the ethanol is partially evaporated under reduced pressure.
The resulting aqueous solution is first washed with 200 mL
of hexane and extracted three times with methylene
chloride. The combined organic layers are then washed
with brine, dried over Na2SO4, filtered and concentrated
under vaccum. The residue was purified by flash chroma-
tography on silica gel (25–40% AcOEt/hexane). C15H14O3
1
1
(242.27); Rf 0.45 (60% AcOEt/hexane); H (200 MHz, 1
3
hexane); H (200 MHz, CDCl3): d¼5.22 (s, 2H), 5.26 (s,
CD3OD/CDCl3): d¼3.57 (d, J¼5.4 Hz, 2H), 6.01 (s, 2H),
2H), 7.03 (d, 3J¼8.0 Hz, 1H), 7.26–7.51 (m, 11H), 9.82 (s,
1H); 13C (50 MHz, CDCl3): d¼70.8, 71.0, 112.5, 113.1,
126.6, 127.0, 127.3, 128.0, 128.1, 128.5, 128.6, 130.3,
136.2, 136.5, 149.2, 154.3, 190.8; IR (neat): nmax¼3034,
2825, 2729, 1683, 1262, 1130 cm21; MS (NH4þ): m/z: 319
[MþH]þ.
6.36 (dt, 3J1¼15.9 Hz, 3J2¼5.6 Hz, 1H), 6.53 (d,
3J¼15.9 Hz, 1H), 7.22–7.39 (m, 5H); 13C (75 MHz, 1
CD3OD/CDCl3): d 26.2, 96.0, 104.7, 126.2, 127.2, 128.0,
128.5, 130.7, 137.4, 155.4, 155.8; IR (neat): nmax¼3381 (br,
OH), 2986, 2870 (Ar.), 1612 (Ar-O), 1142 cm21; MS
(NHþ4 ): m/z: 243 (56.9) [MþH]þ, 260 (100) [MþNH4]þ,
277 (20.3) [MþNH3þNH4]þ.
3.1.10. 3-(3,4-Bis-benzyloxy-phenyl)-acrylic acid ethyl
ester (5). NaH (5.26 g, 130 mmol) is added to a solution of
triethylphosphonoacetate (24.05 mL, 120 mmol) in 300 mL
of THF. After the evolution of gas had ceased, the 3,4-
dibenzyloxybenzaldehyde (36.4 g, 114 mmol) is added and
the reaction is stirred for 10 min. 400 g of ice is added to the
reaction mixture and the aqueous solution is extracted three
times with ethyl acetate. The combined organic layers are
then washed with brine, dried over Na2SO4, filtered and
concentrated under vaccum. The residue was rapidly filtered
(AcOEt) on a short pad of silica to afford 44.2 g (100%) of
the acrylate. C25H24O4 (388.46); Rf 0.45 (20% AcOEt/
hexane); 1H (200 MHz, CDCl3): d¼1.33 (t, 3J¼7.2 Hz, 3H),
4.25 (q, 3J¼7.2 Hz, 2H), 5.19 (s, 2H), 5.20 (s, 2H), 6.25 (d,
3J¼15.9 Hz, 1H), 6.92 (d, 3J¼8.4 Hz, 1H), 7.07 (dd,
3J1¼8.4 Hz, 4J2¼2.2 Hz, 1H), 7.13 (d, 4J¼2.2 Hz, 1H),
7.32–7.48 (m, 10H), 7.80 (d, 3J¼15.9 Hz, 1H); 13C
(50 MHz, CDCl3): d¼14.4, 60.4, 71.0, 71.3, 113.7, 114.3,
116.2, 122.8, 127.2, 127.3, 128.0, 128.6, 136.8, 144.4, 148.9,
151.0, 167.2. IR (neat): nmax¼2972, 2926, 2866, 1739, 1716,
1511, 1230 cm21. MS (NHþ4 ): m/z: 389 [MþH]þ.
3.1.6. 2-(1-Phenyl-allyl)-phloroglucinol (1d). C15H14O3
(242.27); Rf 0.5 (60% AcOEt/hexane); 1H (300 MHz, 1
CD3OD/CDCl3): d¼5.07–5.39 (m, 2H), 5.21 (br s, 2H,
3
OH), 5.31 (d, J¼5.9 Hz, 1H), 5.96 (s, 2H), 7.24–7.34 (m,
5H); IR (neat): nmax¼3446 (br, OH), 2983 (Ar.), 1615
(Ar-O), 1217 cm21; MS (NHþ4 ): m/z: 243 [MþH]þ, 260
[MþNH4]þ.
3.1.7. 2-[3-(3,4-Bisbenzyloxy-phenyl)-allyl]-phloro-
glucinol (1e). The cinnamyl chloride 3 (0.3 g, 0.82 mmol)
in 1.5 mL of THF is slowly added (0.35 mL/h) to the
phloroglucinol·2H2O (0.4 g, 2.47 mmol) in 8 mL of ethanol
and 1.5 mL of water and NaOH (0.2 g, 5 mmol). The
reaction is stirred at 50 8C for 4 h and the ethanol is partially
evaporated under reduced pressure. The resulting aqueous
solution is first washed with 200 mL of hexane and
extracted three times with methylene chloride. The
combined organic layers are then washed with brine, dried
over Na2SO4, filtered and concentrated under vaccum. The
residue was purified by flash chromatography on silica gel
(30–40% AcOEt/hexane). C29H26O5 (454.51); Rf 0.45
(60% AcOEt/hexane); 1H (300 MHz, 1 CD3OD/CDCl3):
d¼3.39 (d, 3J¼5.0 Hz, 2H), 5.02, 5.03 (s, 4H), 5.89 (s, 2H),
6.11 (dt, 3J1¼15.9 Hz, 3J2¼5.9 Hz, 1H), 6.28 (d,
3.1.11. 30,40-Dibenzyloxycinnamyl alcohol (6). LiAlH4
(0.195 g, 5.15 mmol) is added over 5 min to a solution of the
acrylate (2.00 g, 5.15 mmol) in 60 mL anhydrous ether at
215 8C (RM: to avoid the precipitation of the acrylate, the
addition must start right after the flask is cooled). The
reaction is stirred for 2 h and a few drops of conc. Na2S2O5
and 5 g of Na2SO4 are added. The residue is rapidly filtered
(AcOEt) on a short pad of silica to afford 1.34 g (75%) of
pure 6. C23H22O3 (346.42); Rf 0.25 (40% AcOEt/hexane);
1H (200 MHz, CDCl3): d¼4.27 (dd, 3J1¼5.9 Hz, 4J2¼
1.5 Hz, 2H), 5.18, 5.17 (s, 4H), 6.18 (dt, 3J1¼15.9 Hz,
3
4
3J¼15.9 Hz, 1H), 6.78 (dd, J1¼8.4 Hz, J2¼1.9 Hz, 1H),
3
4
6.84 (d, J¼8.4 Hz, 1H), 6.97 (d, J¼1.9 Hz, 1H), 7.23–
7.41 (m, 10H); 13C (75 MHz, 1 CD3OD/CDCl3): d¼26.8,
72.3, 72.4, 95.4, 106.1, 113.5, 116.3, 120.5, 128.4, 128.5,
128.6, 129.1, 129.2, 133.7, 138.3, 148.7, 149.9, 156.9,
157.4; IR (neat): nmax¼3388 (br, OH), 3032, 2925 (Ar.),
1606 (Ar-O), 1509, 1262, 1133 cm21; MS (NH4þ): m/z: 472
[MþNH4]þ.
3
4
3J2¼5.9 Hz, 1H), 6.50 (dt, J1¼15.9 Hz, J2¼1.5 Hz, 1H),
6.90–7.04 (m, 3H), 7.28–7.51 (m, 10H); 13C (50 MHz,
CDCl3): d¼61.9, 69.4, 69.5, 111.1, 113.1, 118.5, 125.0,
125.4, 125.5, 126.0, 126.1, 126.7, 126.8, 128.6, 129.0,
135.4, 147.0, 147.2; IR (neat): nmax¼3065 (br, OH), 3035,
2920, 1512, 1259 cm21; MS (NH4þ): m/z: 364 [MþNH4]þ.
3.1.8. 2-[3-(3,4-Bisbenzyloxy-phenyl)-allyl]-phloro-
glucinol (1f). C29H26O5 (454.51); Rf 0.55 (60% AcOEt/
hexane); 1H (300 MHz, CDCl3): d¼5.01–5.28 (m, 2H),
3
5.07, 5.10 (s, 4H), 5.19 (d, J¼6.2 Hz, 1H), 5.92 (s, 2H),

6812
A. Gissot et al. / Tetrahedron 60 (2004) 6807–6812
Chan, T. H. Org. Lett. 2001, 3, 739. (d) Kozikowski, A. P.;
Tu¨ckmantel, W.; Hu, Y. J. Org. Chem. 2001, 66, 1287.
3. Curtis, W. D.; Stoddart, J. F.; Jones, G. H. J. Chem. Soc.
Perkin Trans. 1 1977, 785.
3.1.12. 3-(3,4-Bis-benzyloxy-phenyl)-1-chloro-prop-2-
ene (3). Thionyl chloride (0.93 mL, 12.71 mmol) is added
dropwise to the cinnamyl alcohol (4 g, 11.56 mmol) and
triethylamine (1.85 mL, 13.29 mmol) in 100 mL of
dichloromethane at 0 8C. After 1 h, 100 mL of water is
added and the resulting aqueous solution is extracted three
times with methylene chloride. The combined organic
layers are then washed with brine, dried over Na2SO4,
filtered and concentrated under vaccum. The residue was
purified by a rapid flash chromatography on silica gel (50%
AcOEt/hexane) to give 3.72 g (88%) of the cinnamyl
4. (a) Deme, E. J. Org. Chem. 1976, 41, 3769. (b) Bu¨chi, G.;
Francisco, M. A.; Liesch, J. M.; Schuda, P. J. Am. Chem. Soc.
1981, 103, 3497. (c) Kawamoto, H.; Nakatsubo, F.; Murakami,
K. Synth. Commun. 1996, 26, 531. (d) Nagvekar, D. S.;
Gibson, H. W. Org. Prep. Proced. Int. 1997, 29, 240.
5. (a) Nakahira, H.; Sunagawa, M. Tetrahedron Lett. 1997, 38,
4443. (b) Forget-Champagne, D.; Mondon, M.; Fonteneau, N.;
Gesson, J. P. Tetrahedron Lett. 2001, 42, 7229.
1
chloride 3. C23H21Cl1O2 (364.86). H (200 MHz, CDCl3):
3
3
d¼4.22 (dd, J1¼7.3 Hz, J2¼1.3 Hz, 2H), 5.18, (s, 4H),
6.14 (dt, 3J1¼15.7 Hz, 3J2¼7.3 Hz, 1H), 6.55 (d,
3J¼15.7 Hz, 1H), 6.88–7.03 (m, 3H), 7.31–7.50 (m,
10H); 13C (50 MHz, CDCl3): d¼46.1, 71.6, 71.9, 113.6,
115.3, 121.1, 123.6, 127.7, 127.8, 128.3, 128.9, 130.0,
134.3, 137.5, 149.5, 149.8; IR (neat): nmax¼3032, 2936,
1511, 1263, 1135 cm21; MS (NH4þ): m/z: 382 [MþNH4]þ.
6. These results ruled out the possibility that the previous
allylphloroglucinol was obtained following a Claisen
rearrangement of the O-allyl phloroglucinol.
7. O/C-benzylation¼8/2 according to the NMR analysis of the
crude mixture.
8. Le Noble, W. J. Synthesis 1970, 1.
9. Nay, B.; Peyrat, J. F.; Vercauteren, J. Eur. J. Org. Chem. 1999,
2231.
10. Gissot A.; Wagner A.; Mioskowski C. Unpublished results.
11. Jurd and coll. have reported the condensation of phloro-
glucinol with cinnamyl alcohol in aqueous acetic acid. Yet, the
adduct is only described as its triacetate ester. It is not clear
whether the protection of the phenols was necessary or was
just used to obtain crystalline products. Yet, from our own
experience it is likely that the unprotected adduct might be
difficult to isolate in that case as we have always experienced
degradation of 1c during the work-up procedure when the
reaction was carried out under acidic conditions. See Jurd, L.
Tetrahedron 1969, 25, 1407.
Acknowledgements
We thank Shannon Biros for reviewing this manuscript and
`
the Ministere de la recherche et de l’enseignement
superieur for financial support through a MRE grant to
´
Arnaud Gissot.
References and notes
1. (a) Jovanovic, S. V.; Steenken, S.; Tosic, M.; Marjanovic, B.;
Simic, M. G. J. Am. Chem. Soc. 1994, 116, 4846. (b) Wright,
J. S.; Johnson, E. R.; DiLabio, G. A. J. Am. Chem. Soc. 2001,
123, 1173.
12. Nay, B.; Collet, M.; Lebon, M.; Cheze, C.; Vercauteren, J.
Tetrahedron Lett. 2002, 43, 2675.
´
13. Lubineau, A.; Auge, J.; Queneau, Y. Synthesis 1994, 741.
14. As seen before, the different cinnamyl phloroglucinols such as
1c or 1e are acid-sensitive. They can nevertheless be purified
on silica gel provided a fast elution of the products. Hence,
both regioisomers 1e and 1f are barely separable on silica gel.
2. (a) Kawamoto, H.; Nakatsubo, F.; Murakami, K. J. Wood
Chem. Technol. 1989, 9, 35. (b) Nay, B.; Arnaudinaud, V.;
´
Peyrat, J. F.; Nuhrich, A.; Deffieux, G.; Merillon, J. M.;
Vercauteren, J. Eur. J. Org. Chem. 2000, 1279. (c) Li, L.;