Selective cleavage of tert-butyldimethylsilylethers ortho to a carbonyl group by ultrasound

Article abstract of DOI:10.1016/S0040-4020(00)00053-3

A general method for the selective cleavage of tert- butyldimethylsilylethers ortho to a carbonyl group is established by sonication of a 0.1 M solution of the substrate in 1/1 (v/v) CH₃OH/CCl₄. Other phenolic tert-butyldimethylsilylethers are unaffected. This reaction performed on flavanoids is completed within 3 h and no special workup is required. Other substrates are also investigated and a mechanism is proposed. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4020(00)00053-3

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 1541–1549
Selective Cleavage of tert-Butyldimethylsilylethers ortho to a
Carbonyl Group by Ultrasound
*
`
Alex H. De Groot, Roger A. Dommisse and Guy L. Lemiere
Department of Chemistry, University of Antwerp (RUCA), Groenenborgerlaan 171, 2020 Antwerp, Belgium
Received 12 April 1999; revised 4 January 2000; accepted 13 January 2000
Abstract—A general method for the selective cleavage of tert-butyldimethylsilylethers ortho to a carbonyl group is established by
sonication of a 0.1 M solution of the substrate in 1/1 (v/v) CH3OH/CCl4. Other phenolic tert-butyldimethylsilylethers are unaffected.
This reaction performed on flavanoids is completed within 3 h and no special workup is required. Other substrates are also investigated
and a mechanism is proposed. ᭧ 2000 Elsevier Science Ltd. All rights reserved.
Introduction
several phenolic TBDMS ethers ortho to a carbonyl group
under ultrasonic conditions. To achieve this, the substrate
(0.1 M) was dissolved in a 1/1 (v/v) mixture of CH3OH/
CCl4 and sonicated in an ultrasonic cleaning bath at
50–55ЊC (Fig. 1). The reactions were monitored by TLC
and by mass spectrometry. After completion the solvent was
evaporated in vacuo. The fast mono-deprotections gave pure
end products. In the case of slower or consecutive desilyl-
ations, purification by preparative HPLC was sometimes
necessary.
Together with the increasing complexity of the molecules
synthesized, there is a growing demand for better protective
groups. These must be easy to introduce or to remove, and
they should also be resistant to a wide variety of reaction
conditions. The selectivity between several protecting
groups and even between similar groups is especially
valuable for the modern synthetic chemist. The tert-butyl-
dimethylsilyl (TBDMS) group has become very popular
during recent years because of all these characteristics.1
We present a new deprotection method involving sonication
with a classic ultrasonic cleaning bath.
This regioselective reaction was first observed on per-
silylated taxifolin (Table 1, entry 1a), whereas persilylated
catechin (entry 2a), the analogue of 1 lacking a carbonyl
group, showed no deprotection. These reactions suggest
that the carbonyl group plays a crucial role in the selective
desilylation reaction. According to Lee,5,7 sonication of a
CH3OH/CCl4 mixture generates HCl in situ in analogy to
the H2O/CCl4 mixture.8 Mason observed a similar reaction.9
By sonication of tert-butylchloride in aqueous ethanol he
obtained the corresponding ether and HCl. As Luche2 noted,
this reaction probably follows a radical pathway10 rather
than the ionic pathway proposed by Mason. This radical
pathway was proved by Eshuis,11 when CHBr3 was exposed
to ultrasound. In contrast to this last reaction, where HBr is
formed from two CHBr3 molecules, CCl4 needs a protic
cosolvent, which delivers the hydrogen atom necessary for
the formation of HCl. Consequently the created HCl
Ultrasound has found many interesting applications in
organic chemistry.2,3 This technique derives its success
not only from the simple rate enhancement of specific reac-
tions, in particular in organo-metallic chemistry, but it also
promotes radical pathways giving products which are
normally not accessible via ionic reactions. The investiga-
tion of the use of ultrasound for deprotection purposes has
only recently started. Schmittling applied sonication to
accelerate the deprotection of phenolic silylethers by
KF–Al2O3 in CH3CN4 and Lee cleaved primary TBDMS
ethers selectively by sonication in a CH3OH/CCl4 solution
without affecting phenolic groups.5 Since Cavelier obtained
some opposite results with this last method,6 we found it
worth investigating this technique within the scope of our
research for new selective protections of flavanoids.
Results and Discussion
We wish to report the highly regioselective deprotection of
Keywords: phenolics; protecting groups; sonochemistry; NMR.
Figure 1.
* Corresponding author. E-mail: dommisse@ruca.ua.ac.be
0040–4020/00/$ - see front matter ᭧ 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00053-3

1542
A. H. De Groot et al. / Tetrahedron 56 (2000) 1541–1549
Table 1. Desilylation of tert-butyldimethylsilylethers near to a conformationally fixed carbonyl group
temporarily protonates the carbonyl group and if the added
proton is near enough to a silylether, it can catalyze its
cleavage. Such a six-ring mechanism as shown in Fig. 2 is
well known for Lewis acids, which catalyze selective
deprotections ortho to a carbonyl group. Dean12 described
this mechanism for the selective deprotection of methyl
ethers ortho to a carbonyl with boron trichloride. Benzyl-
ethers near a carbonyl group were reported to be selectively
13
removed by AlCl3 and MgBr214. To the best of our
knowledge no selective deprotection of silylethers ortho to
a carbonyl has yet been reported.
The flavanoid skeleton guarantees an ideal distance between
the carbonyl group and the ortho-silylether and also a fixed
planar position of the carbonyl group. Products 1 and 3–6
indeed give a fast desilylation at the 5-position. Product 5
shows that the carbonyl-group-catalyzed deprotection of the
ortho-silylether is even faster than the desilylation of the
primary alcohol5 in the conditions mentioned above. As
indicated, we could remove the silylether ortho to the
carbonyl group, leaving the primary silylether unaffected.
The enolic silylether at the 3-position of persilylated
quercetin (3a) is cleaved after prolonged reaction time.
This result is also found for purpurogallin (7), which has a
Figure 2.

A. H. De Groot et al. / Tetrahedron 56 (2000) 1541–1549
1543
Table 2. Desilylation of tert-butyldimethylsilylethers near to a conformationally free carbonyl group
similar substitution pattern around the carbonyl group. The
oxygen atom of the enolic ethers is farther away from the
carbonyl than the ortho-silylether oxygen atom, and after
protonation only a five-membered ring can be formed.12 In
accordance with our proposed mechanism, the carbonyl
group catalyzes less effectively the deprotection of the
enolic silylether and so the ortho desilylated product can
be isolated (3b). The out of plane oriented aliphatic
silylether of the taxifolin derivative is not cleaved at all,
even after prolonged reaction time (Ͼ2 days). In this case
the silyl oxygen atom is clearly too far away from the
carbonyl group. The less reactive character of the aliphatic
ether must also be taken into account.
forms a stable hydrogen bond with the free hydroxyl
group making it unavailable for desilylation of the second
one. No further desilylation is seen even after three days of
sonication. In the other compounds (9–12) the bulkier
substituents prevent a planar conformation of the carbonyl
group,15,16 thereby destabilising an intramolecular hydrogen
bond. The carbonyl group is now more available for a
cleavage of the second ortho-silylether. This cleavage
occurs slowly, due to steric hindrance of the bulky sub-
stituent hindering the carbonyl group from adopting an
ideal planar conformation.
The aromatic silylester in 11 is also readily cleaved.
However, the ortho hydroxyl group does not play any role
here as can be seen from entry 13.
To test the generality of this desilylation, several
compounds (8–12, Table 2) with a carbonyl containing
functional group ortho to two silylethers and not fixed in a
ring structure were studied. The results suggest that the
selective desilylation reaction is general for all ortho-
hydroxy-carbonyl compounds, but they also show that the
reaction rate and selectivity decrease with increasing sub-
stituents on the carbonyl group. The proposed mechanism
suggests that the carbonyl group has to be in the plane of the
aromatic ring to play an ideal reaction promoting role. For
8–12 increasing sterical hindrance decreases the occurrence
of this planar orientation, thus increasing the required
reaction time. After deprotection of one of the ortho-
silylethers the carbonyl group in the benzaldehyde 8
In summary, aromatic and enolic tert-butyldimethylsilyl-
ethers ortho to a carbonyl group have been cleaved
selectively, using a very simple ultrasound treatment in a
CH3OH/CCl4 (1/1) mixture. Although the reaction seems to
be quite general, effects that hamper the planar position of
the carbonyl group relative to the ortho-silylether increase
reaction times.
Experimental
Melting points are determined in glass capillaries on a

1544
A. H. De Groot et al. / Tetrahedron 56 (2000) 1541–1549
Tottoli apparatus and are uncorrected. 1H, 13C, COSY
HETCOR and long-range HETCOR NMR spectra were
recorded on a Varian Unity 400 spectrometer in deuterated
chloroform using tetramethylsilane as an internal reference,
unless otherwise mentioned. 1H and 13C were assigned
with the aid of COSY, HETCOR and long-range HETCOR
spectra where necessary, the latter is listed when measured.
Assignments that could not be proved unambiguously, are
marked with an asterisk: ‘ء’. DCI-mass spectra were taken
on a Ribermag R10-10B quadrupole mass spectrometer,
using ammonia as reagent gas Preparative HPLC was
performed on a Merck Novaprep 5000, using a Merck
250×50 self-packing column packed with Merck
LiChroprep Si 60 (25–40 mm particle size).
under high vacuum and purified by straight-phase prepara-
tive HPLC using a heptane/ether gradient, yielding 769 mg
(98%) of pure 1b as a colourless oil. Anal. found: C, 61.41;
H, 9.03; C39H68O7Si4 requires C, 61.53; H 9.00; d 1H
(CDCl3) 11.78 (s, 1H, –OH5), 6.95 (d, 1H, Jˆ2.1 Hz,
H20), 6.91 (dd, 1H, Jˆ8.2,2.1 Hz, H60), 6.85 (d, 1H,
Jˆ8.1 Hz, H50), 6.02 (d, 1H, Jˆ2.1 Hz, H6), 5.96 (d, 1H,
Jˆ2.1 Hz, H8), 5.00 (d, 1H, Jˆ11.1 Hz, H2), 4.41 (d, 1H,
Jˆ11.1 Hz, H3) 1.002 (s, 9H, (CH3)3Si–), 0.997 (s, 9H,
(CH3)3Si–), 0.96 (s, 9H, (CH3)3Si–), 0.72 (s, 9H,
(CH3)3Si–), 0.24 (s, 6H, CH3Si–), 0.211 (s, 3H, CH3Si–),
0.205 (s, 3H, CH3Si–), 0.20 (s, 6H, CH3Si–), 0.07 (s, 3H,
CH3Si–), Ϫ0.27 (s, 3H CH3Si–); d 13C (CDCl3) 196.93
(C4), 164.85 (C7), 163.84 (C5), 162.48 (C9), 147.60
(C40), 146.90 (C30), 130.31 (C10), 120.98 (C60), 120.93
(C50), 120.52 (C20), 102.30 (C6), 102.30 (C10), 99.84
(C8), 83.87 (C2) 75.03 (C3), 26.01 ((CH3)3C–), 25.98
((CH3)3C–), 25.61 ((CH3)3C–), 25.51 ((CH3)3C–), 18.49
((CH3)3C–), 18.43 ((CH3)3C–), 18.32 ((CH3)3C–), 18.21
((CH3)3C–), Ϫ4.00 (CH3Si–), Ϫ4.06 (CH3Si–), Ϫ4.11
(CH3Si–), Ϫ4.15 (CH3Si–), Ϫ4.37 (CH3Si–), Ϫ4.45
(CH3Si–); Long-range HETCOR correlations: H5!C:5,
10,6; H20!C:40,60; H60!C:20; H50!C:30,10; H6!C:7,
5,10,8; H8!C:9,10; H2!C:4,60,3; H3!C:2. m/z (DCI):
761 [MϩH]ϩ 703 [MϪ57]ϩ; IR (KBr) nmax: 1650, 1513,
Commercial dimethylformamide (DMF) was dried over
calcium hydride and distilled. Commercial dichloromethane
˚
was dried over molecular sieves (4 A). All other products
were used as supplied by Acros, Aldrich or Sigma.
3,30,40,5,7-Penta-O-tert-butyldimethylsilyltaxifolin, 1a.
Racemic taxifolin (0.5 g, 1.64 mmol) was dissolved in dry
DMF (5 ml). Imidazole (1.1 g, 21 mmol) and tert-butyl-
dimethylsilyl chloride (1.4 g, 9 mmol) were then added
and the mixture was stirred at room temperature under a
nitrogen atmosphere for 16 h. The reaction mixture was
diluted with 40 ml ether and washed with water
(5×20 ml). After drying over MgSO4, the solvent was
evaporated giving a pale yellow oil. It was purified by
straight-phase preparative HPLC using a heptane/ether
gradient, yielding 1.34 g (93%) colourless oil of 1a. Anal.
found: C, 61.91; H, 9.20; C45H82O7Si5 requires C, 61.73; H
1256, 1172, 837, 784 cmϪ1
.
3,30,40,5,7-Penta-O-tert-butyldimethylsilylquercetin, 3a.
Quercetin (0.5 g, 1.48 mmol) was dissolved in dry dichloro-
methane (5 ml) and DBU (1.5 ml, 10 mmol) tert-butyl-
dimethylsilyl chloride (1.4 g, 9 mmol) was then added and
the mixture was stirred at room temperature under a nitro-
gen atmosphere for 6 h. The reaction mixture was diluted
with 20 ml dichloromethane and washed with water
(5×20 ml). After drying over MgSO4, the solvent was
evaporated giving a yellow-brown oil. It was purified by
straight-phase preparative HPLC using a heptane/dichloro-
methane gradient yielding 1.25 g (97%) of 1a as a yellow
oil. Anal. found: C, 61.64; H, 8.97; C45H80O7Si5 requires C,
1
9.44; d H (CDCl3) 6.97 (d, 1H, Jˆ2.4 Hz, H20), 6.92 (dd,
1H, Jˆ8.2,2.4 Hz, H60), 6.84 (d, 1H, Jˆ8.0 Hz, H50), 6.10
(d, 2H, Jˆ2.4 Hz, H8), 5.99 (d, 1H, Jˆ2.4 Hz, H6), 4.97 (d,
1H, Jˆ10.8 Hz, H2), 4.29 (d, 1H, Jˆ10.8 Hz, H3), 1.06 (s,
9H, (CH3)3Si–), 1.02 (s, 9H, (CH3)3Si–), 1.01 (s, 9H,
(CH3)3Si–), 0.98 (s, 9H, (CH3)3Si–), 0.70 (s, 9H,
(CH3)3Si–), 0.284 (s, 3H, CH3Si–), 0.242 (s, 9H, CH3Si–),
0.223 (s, 3H, CH3Si–), 0.217 (s, 3H, CH3Si–), 0.207 (s, 6H,
CH3Si–), 0.117 (s, 3H, CH3Si–), Ϫ0.17 (s, 3H, CH3Si–); d
13C (CDCl3) 189.66 (C4), 163.26 (C9), 161.85 (C7), 158.05
(C5), 147.27 (C40), 146.75 (C30), 131.03 (C10), 120.88 (C50),
120.81 (C60), 120.42 (C20), 107.81 (C10), 106.77 (C6),
101.82 (C8), 83.46 (C2), 76.20 (C3), 26.04 ((CH3)3C–),
26.01 ((CH3)3C–), 25.85 ((CH3)3C–), 25.69 ((CH3)3C–),
25.58 ((CH3)3C–), 18.53 ((CH3)3C–), 18.49 ((CH3)3C–),
18.44 ((CH3)3C–), 18.25 ((CH3)3C–), Ϫ4.00 (CH3Si–),
Ϫ4.06 (CH3Si–), Ϫ4.11 (CH3Si–), Ϫ4.15 (CH3Si–), Ϫ4.19
(CH3Si–), Ϫ4.25 (CH3Si–), Ϫ4.32 (CH3Si–), Ϫ4.33
(CH3Si–), Ϫ4.37 (CH3Si–), Ϫ5.76 (CH3Si–); Long-range
HETCOR correlations: H20!C:40,60,2; H60!C:40,20,2;
H50!C:30,10; H8!C:9,7,10,6; H6!C:7,5,10,8; H2!C:4,
10,60,20,3; H3!C:2; m/z (DCI): 875 [MϩH]ϩ, 817
[MϪ57]ϩ; IR (KBr) nmax: 1631, 1520, 1251, 1167, 839,
1
61.87; H 9.23; d H (CDCl3) 7.47 (dd, 1H, Jˆ8.4,2.1 Hz,
H60), 7.38 (d, 1H, Jˆ2.1 Hz, H20), 6.92 (d, 1H, Jˆ8.5 Hz
H50), 6.34 (d, 1H, Jˆ2.1 Hz, H8), 6.28 (d, 1H, Jˆ2.1 Hz,
H6), 1.02 (s, 9H, (CH3)3Si–), 1.01 (s, 9H, (CH3)3Si–), 1.00
(s, 9H, (CH3)3Si–), 0.87 (s, 10H, (CH3)3Si–), 0.86 (s, 9H,
(CH3)3Si–), 0.27 (s, 6H, CH3Si–), 0.25 (s, 6H, CH3Si–),
0.23 (s, 6H, CH3Si–), 0.15 (s, 6H, CH3Si–), 0.03 (s, 6H,
CH3Si–); d 13C (CDCl3) 178.20 (C4), 161.92 (C5), 161.77
ء
(C7), 156.49 (C9), 153.18 (C40 ), 147.51 (C2ء), 146.79
(C30), 135.69 (C3), 124.52 (C10), 123.21 (C60), 121.79
(C20), 120.81 (C50), 106.15 (C10), 102.95 (C6), 98.14
(C8), 25.96 ((CH3)3C–), 25.76 ((CH3)3C–), 25.73
((CH3)3C–), 25.67 ((CH3)3C–), 25.57 ((CH3)3C–), 18.62
((CH3)3C–), 18.59 ((CH3)3C–), 18.46 ((CH3)3C–), 18.23
((CH3)3C–), 18.11 ((CH3)3C–), Ϫ3.57 (CH3Si–), Ϫ4.03
(CH3Si–), Ϫ4.03 (CH3Si–), Ϫ4.15 (CH3Si–), Ϫ4.36
(CH3Si–); Long-range HETCOR correlations: H60!C:2,
40,20; H20!C:2,40,30,60; H50!C:30,10; H8!C:7,9,10,6;
H6!C:5,7,10,8; m/z (DCI): 873 [MϩH]ϩ, 815 [MϪ57]ϩ;
IR (KBr) nmax: 1644, 1611, 1499, 1350, 1313, 1253, 1181,
782 cmϪ1
.
3,30,40,7-Tetra-O-tert-butyldimethylsilyltaxifolin, 1b. A
solution of 1a (0.9 g, 1 mmol) in methanol (2 ml) and tetra-
chloromethane (2 ml) was sonicated in a Bransor 2200
cleaning bath (47 kHz) at 50–55ЊC. The reaction was moni-
tored by TLC and after 15 min the solvents were removed in
vacuo. The resulting light brown oil was held overnight
1014, 839, 779 cmϪ1
.
3,30,40,7-Tetra-O-tert-butyldimethylsilylquercetin, 3b.
Sonication of 3a following a procedure similar to the one

A. H. De Groot et al. / Tetrahedron 56 (2000) 1541–1549
1545
described for 1b gave 3b after 90 min. Purification by
straight-phase preparative HPLC using a heptane/dichloro-
methane gradient gave a yellow oil. Yield 1.01 g (93%).
Anal. found: C, 61.57; H, 8.83; C39H66O7Si4 requires C,
tions: H20ϩH60!C:2,40,60; H30ϩH50!C:50; H8!C:7,
9,10,6; H3!C:4,2,10,10; H6!C:7,5,10,8; m/z (DCI): 483
[MϩH]ϩ, 425 [MϪ57]ϩ; IR (KBr) nmax: 1642, 1619,
1450, 1352, 1170, 1100, 853, 840, 778 cmϪ1
1
61.69; H, 8.76; d H (CDCl3) 12.67 (s, 1H, –OH5), 7.45
(dd, 1H, Jˆ8.4,2.2 Hz, H60), 7.35 (d, 1H, Jˆ2.1 Hz, H20),
6.90 (d, 1H, Jˆ8.5 Hz, H50), 6.33 (d, 1H, Jˆ2.1 Hz, H8),
6.26 (d, 1H, Jˆ2.1 Hz, H6), 1.01 (s, 9H, (CH3)3Si–), 1.00
(s, 9H, (CH3)3Si–), 0.99 (s, 9H, (CH3)3Si–), 0.84 (s, 9H,
(CH3)3Si–), 0.26 (s, 6H, CH3Si–), 0.24 (s, 6H, CH3Si–),
0.21 (s, 6H, CH3Si–), 0.12 (s, 6H, CH3Si–); d 13C (CDCl3)
178.23 (C4), 161.91 (C5), 161.80 (C7), 156.50 (C9), 153.22
7-O-tert-Butyldimethylsilylchrysin, 4b. Sonication of 4a
following a procedure similar to the one described for 1b
gave 4b after 90 min. Recrystallization from ethanol
afforded a yellow crystalline material, mp 153.5ЊC. Yield
643 mg (95%). Anal. found C, 68.60; H, 6.57; C21H24O4Si
requires C, 68.45; H, 6.56; d 1H (CDCl3) 12.65 (s, 1H,
–OH5), 7.91 (m, 2H, H20ϩH60), 7.53 (m, 3H,
H30ϩH50ϩH40), 6.66 (s, 1H, H3), 6.45 (d, 1H, Jˆ2.0 Hz,
H8), 6.31 (d, 1H, Jˆ2.0 Hz H6), 1.00 (s, 9H, (CH3)3Si–),
0.28 (s, 6H, CH3Si–); d 13C (CDCl3) 182.50 (C4), 163.95
(C2), 162.36 (C7), 162.12 (C5), 157.62 (C9), 131.74 (C40),
131.36 (C10), 129.00 (C30ϩC50), 126.26 (C20ϩC60), 106.21
(C10), 105.82 (C3), 103.92 (C6), 98.74 (C8), 25.51
((CH3)3C–), 18.21 ((CH3)3C–), Ϫ4.38 (CH3Si–); Long-
range HETCOR correlations: H5!C:5,10,6; H20ϩH60!
C:2,40,60; H30ϩH50!C:50; H3!C:4,2,10,10; H8!C:7,9,
10,6; H6!C:7,10,8; m/z (DCI): 369 [MϩH]ϩ, 311
[MϪ57]ϩ; IR (KBr) nmax: 1652, 1618, 1451, 1349, 1169,
ء
(C40 ), 149.25 (C2ء), 146.81 (C30), 135.70 (C3), 124.52
(C10), 123.22 (C60), 121.75 (C20), 120.80 (C50), 106.17
(C10), 102.96 (C6), 98.17 (C8), 25.98 ((CH3)3C–), 25.76
((CH3)3C–), 25.73 ((CH3)3C–), 25.59 ((CH3)3C–), 18.63
((CH3)3C–), 18.61 ((CH3)3C–), 18.48 ((CH3)3C–), 18.26
((CH3)3C–), Ϫ4.02 (CH3Si–), Ϫ4.13 (CH3Si–), Ϫ4.34
(CH3Si–); Long-range HETCOR correlations: H5!C:5,
10,6; H60!C:2,40,20; H20!C:2,40(30,60; H50!C:3010;
H8!C:7,9,10,6; H6!C:5,7,10,8; m/z (DCI): 759
[MϩH]ϩ, 701 [MϪ57]ϩ; IR (KBr) nmax: 1652, 1601,
1509, 1352, 1309, 1253, 1188, 1164, 1004, 841, 781 cmϪ1
.
1102, 857, 839, 780 cmϪ1
.
30,40,7-Tri-O-tert-butyldimethylsilylquercetin, 3c. After
24 h 3c was obtained from 3b in 63% yield (0.54 g) as a
yellow oil by a procedure similar to the one described for
3b. Anal. found: C, 61.33; H, 8.01; C33H52O7Si3 requires C,
5-O-tert-Butyldimethylsilyl-7-O-(2-tert-butyldimethyl-
silyloxyethyl)chrysin, 5a. 5a was formed from 500 mg of
7-O-(2-hydroxyethyl)chrysin by a procedure similar to the
one described for 3a. Recrystallization from methanol/water
gave a deep yellow crystalline material, mp 92ЊC. Yield
840 mg (95%). Anal. found: C, 66.36; H, 8.15;
1
61.45; H, 8.13; d H (CDCl3) 11.70 (s, 1H, –OH5), 7.84 (d,
1H, Jˆ2.2 Hz, H20), 7.68 (dd, 1H, Jˆ8.5,2.3 Hz, H60), 6.96
(d, 1H, Jˆ8.6 Hz, H50), 6.64 (s, 1H, –OH3), 6.41 (d, 1H,
Jˆ2.0 Hz, H8), 6.30 (d, 1H, Jˆ2.0 Hz, H6), 1.03 (s, 9H,
(CH3)3Si–), 1.02 (s, 9H, (CH3)3Si–), 1.01 (s, 9H, (CH3)3Si–),
0.284 (s, 6H, CH3Si–), 0.275 (s, 6H, CH3Si–), 0.254 (s, 6H,
CH3Si–); d 13C (CDCl3) 175.30 (C4), 162.43 (C7ء), 160.86
(C5), 156.59 (C9ء), 149.17 (C40), 147.02 (C30), 145.58 (C2),
135.76 (C3), 123.94 (C10), 121.22 (C60), 121.05 (C50),
120.65 (C20), 104.40 (C10), 103.15 (C6), 98.61 (C8),
26.00 ((CH3)3C–), 25.89 ((CH3)3C–), 25.56 ((CH3)3C–),
18.56 ((CH3)3C–), 18.52 ((CH3)3C–), 18.27 ((CH3)3C–),
Ϫ4.06 (CH3Si–), Ϫ4.12 (CH3Si–), Ϫ4.34 (CH3Si–);
Long-range HETCOR correlations: H5!C:5, 10,6;
H60!C:40; H50!C:30,10; H3!C:2,3; H6!C:10; m/z
(DCI): 645 [MϩH]ϩ, 587 [MϪ57]ϩ; IR (KBr) nmax: 1659,
1602, 1512, 1352, 1309, 1254, 1186, 1167, 1009, 842,
1
C29H42O5Si2 requires C, 66.12; H 8.04; d H (CDCl3) 7.89
(m, 2H, H20ϩH60), 7.55 (m, 3H, H30ϩH50 and H40), 6.70
(d, 1H, Jˆ2.4 Hz, H8), 6.65 (s, 1H, H3), 6.46 (d, 1H,
Jˆ2.4 Hz, H6), 4.21 (t, 2H, Jˆ5.2 Hz, H10 0), 4.12 (t, 2H,
Jˆ5.2 Hz, H20 0), 1.08 (s, 9H, (CH3)3Si–), 0.92 (s, 9H,
(CH3)3Si–), 0.28 (s, 6H, CH3Si–), 0.12 (s, 6H, CH3Si–);
d
13C (CDCl3) 176.80 (C4), 162.53 (C7), 160.32 (C2),
159.05 (C9), 156.57 (C5), 130.77 (C10), 130.76 (C40),
128.59 (C30ϩC50), 125.57 (C20ϩC60), 111.43 (C10),
108.25 (C3), 105.45 (C6), 94.97 (C8), 69.73 (C10 0), 61.52
(C20 0), 26.59 ((CH3)3C–), 26.53 ((CH3)3C–), 19.44
((CH3)3C–), 19.15 ((CH3)3C–), Ϫ3.81 (CH3Si–), Ϫ4.71
(CH3Si–); Long-range HETCOR correlations: H20ϩH60!
C:2,40,60; H30ϩH50!C:40; H8!C:7,9,10; H3!C:2,10;
H6!C:7,5,10; H10 0!C:7; m/z (DCI): 527 [MϩH]ϩ, 469
[MϪ57]ϩ; IR (KBr) nmax: 1645, 1611, 1450, 1350, 1254,
777 cmϪ1
.
5,7-Di-O-tert-butyldimethylsilylchrysin, 4a. 4a was
formed from 500 mg of chrysin by a procedure similar to
the one described for 3a. Recrystallization from ethanol
afforded a yellow crystalline material, mp 152.5ЊC. Yield
885 mg (93%). Anal. found: C, 66.92; H, 8.17; C27H38O4Si2
requires C, 67.17; H 7.93; d 1H (CDCl3) 7.81 (m, 2H,
H20ϩH60), 7.44 (m, 3H, H30ϩH50ϩH40), 6.569 (d, 1H,
Jˆ2.2 Hz, H8), 6.565 (s, 1H, H3), 6.28 (d, 1H, Jˆ2.2 Hz,
H6), 1.07 (s, 9H, (CH3)3Si–), 0.99 (s, 9H, (CH3)3Si–), 0.28
(s, 6H, CH3Si–), 0.26 (s, 6H, CH3Si–); d 13C (CDCl3)
177.26 (C4), 160.73 (C2), 159.99 (C7), 159.05 (C9),
156.85 (C5), 131.71 (C10), 130.95 (C40), 128.75
(C30ϩC50), 125.85 (C20ϩC60), 112.14 (C10), 109.90
(C6), 108.38 (C3), 101.73 (C8), 25.87 ((CH3)3C–), 25.48
((CH3)3C–), 18.57 ((CH3)3C–), 18.16 ((CH3)3C–), Ϫ4.41
(CH3Si–), Ϫ4.42 (CH3Si–); Long-range HETCOR correla-
1174, 1136, 1102, 836, 777 cmϪ1
.
7-O-(2-tert-Butyldimethylsilyloxyethyl)chrysin, 5b. Soni-
cation of 5a following a procedure similar to the one
described for 1b gave 5b after 90 min. Recrystallization
from ethanol afforded a deep yellow crystalline material,
mp 107.5ЊC. Yield 585 mg (89%). Anal. found: C, 66.89;
H, 6.80; C23H28O5Si requires C, 66.96; H, 6.84; d 1H
(CDCl3) 13.95 (s, 1H, –OH5), 7.88 (m, 2H, H20ϩH60),
7.52 (m, 3H, H30ϩH50ϩH40), 6.66 (s, 1H, H3), 6.52 (d,
1H, Jˆ2.3 Hz, H8), 6.38 (d, 1H, Jˆ2.3 Hz, H6), 4.12 (t,
2H, Jˆ5.2 Hz, H10 0), 4.00 (t, 2H, Jˆ5.2 Hz, H20 0), 0.96
(s, 9H, (CH3)3Si–), 0.55 (s, 5H, CH3Si–); d 13C (CDCl3)
182.47 (C4), 165.07 (C7), 164.00 (C2), 162.25 (C5), 157.84
(C9), 131.78 (C40), 131.47 (C10), 129.08 (C30ϩC50), 126.32
(C20ϩC60), 105.95 (C3), 105.94 (C10), 98.78 (C6), 93.24

1546
A. H. De Groot et al. / Tetrahedron 56 (2000) 1541–1549
(C8), 70.03 (C10 0), 61.73 (C20 0), 25.88 ((CH3)3C–), 18.38
((CH3)3C–), Ϫ5.20 (CH3Si–); Long-range HETCOR corre-
lations: H5!C:5,10,6; H20ϩH60!C:2,40,60; H30ϩH50!
C:40; H3!C:2,10; H8!C:7,9,10; H6!C:7,5,10; H10 0!
C:7; m/z (DCI): 413 [MϩH]ϩ, 355 [MϪ57]ϩ; IR (KBr)
nmax: 1663, 1622, 1449, 1353, 1255, 1172, 1101, 838,
6.94 (d, 1H, Jˆ2.1 Hz, H20), 6.88 (d, 1H, Jˆ8.3 Hz, H50)
6.00 (d, 1H, Jˆ2.1 Hz, H6), 5.99 (d, 1H, Jˆ2.1 Hz, H8),
5.31 (dd, 1H, Jˆ12.7,2.9 Hz, H2), 3.82 (s, 3H, CH3), 3.06
(dd, 1H, Jˆ17.1,12.8 Hz, H3a), 2.78 (dd, 1H,
Jˆ17.0,2.9 Hz, H3b), 1.01 (s, 9H, (CH3)3Si–), 0.97 (s,
9H, (CH3)3Si–), 0.247 (s, 6H, CH3Si–), 0.174 (s, 3H,
CH3Si–), 0.169 (s, 3H, CH3Si–); d 13C (CDCl3) 196.09
(C4), 164.95 (C7), 163.99 (C5), 162.86 (C9), 151.45
(C40), 145.41 (C30), 131.07 (C10), 119.65 (C60), 119.10
(C20), 112.28 (C50), 103.72 (C10), 101.17 (C6), 99.90
(C8), 78.87 (C2), 55.58 (CH3), 43.31 (C3a), 25.73
((CH3)3C–), 25.51 ((CH3)3C–), 18.45 ((CH3)3C–), 18.19
((CH3)3C–), Ϫ4.37 (CH3Si–), Ϫ4.56 (CH3Si–); Long-
range HETCOR correlations: H5!C:5,10,6; H60!C:20;
H20!C:30,60; H50!C:30,10; H6!C:7; H8!C:7,9;
H3a!C:4,2; m/z (DCI): 531 [MϩH]ϩ, 473 [MϪ57]ϩ; IR
(KBr) nmax: 1698, 1609, 1509, 1256, 1172, 1094, 843,
771, 762 cmϪ1
.
7-O-(2-Hydroxyethyl)chrysin, 5c. Sonication of 5b
following a procedure similar to the one described for 1b
gave 5c after 20 h. Recrystallization from ethanol gave a
yellow crystalline material, mp 162ЊC. Yield 390 mg
(92%). Anal. found: C, 68.31; H, 4.65; C17H14O5 requires
1
C, 68.45; H, 4.73; d H (CDCl3) 12.65 (s, 1H, –OH5), 7.83
(m, 2H, H20ϩH60), 7.50 (m, 3H, H30ϩH50ϩH40), 6.47 (d,
1H, Jˆ2.1 Hz, H8), 6.34 (d, 1H, Jˆ2.1 Hz, H6), 4.14 (t, 2H,
Jˆ4.5 Hz, H10 0), 4.01 (br.q., 2H, Jˆ4.6 Hz, H20 0), 2.65
(br.s, 1H, –OH20 0); d 13C (CDCl3) 182.38 (C4), 164.66
(C7), 164.04 (C2), 162.19 (C5), 157.71 (C9), 131.80
(C40), 131.24 (C10), 129.03 (C30ϩC50), 126.25
(C20ϩC60), 105.86 (C10), 105.82 (C3), 98.69 (C6), 93.17
(C8), 69.94 (C10 0), 61.07 (C20 0); Long-range HETCOR
correlations: H5!C:5,10,6; H20ϩH60!C:2,40,60ϩ20;
H30ϩH50! C:10,50ϩ30; H8!C:7,9,10; H6!C:7,5,10;
H10 0!C:7; m/z (DCI): 299 [MϩH]ϩ; IR (KBr)
785 cmϪ1
.
2,7,8,9-Tetra-O-tert-butyldimethylsilylpurpurogallin, 7a.
7a was formed from 0.5 g of purpurogallin by a procedure
similar to the one described for 3a. Purification by straight-
phase preparative HPLC using a heptane/dichloromethane
gradient gave a brown oil. Yield 1.27 g (83%). Anal. found:
C, 61.88; H, 9.69. C35H64O5Si4 requires C, 62.07; H 9.53; d
1H (CDCl3) 6.77 (d, 1H, Jˆ11.3 Hz, H5), 6.66 (s, 1H, H6)
6.34 (d, 1H, Jˆ8.1 Hz, H3), 6.26 (dd, 1H, Jˆ11.4,8.8 Hz,
H4), 1.04 (s, 9H (CH3)3Si–), 1.03 (s, 9H, (CH3)3Si–), 1.02
(s, 9H, (CH3)3Si–), 0.98 (s, 9H, (CH3)3Si–), 0.30 (s, 6H,
CH3Si–), 0.29 (s, 6H, CH3Si–), 0.21 (s, 6H, CH3Si–), 0.08
(s, 6H, CH3Si–); d 13C (CDCl3) 184.72 (C1), 155.25 (C2),
150.73 (C7), 148.28 (C9), 140.40 (C8), 131.05 (C5), 130.48
(C11), 126.20 (C10), 121.77 (C4), 115.44 (C3), 115.05
(C6), 26.65 ((CH3)3C–), 26.44 ((CH3)3C–), 26.14
((CH3)3C–), 25.86 ((CH3)3C–), 18.77 ((CH3)3C–), 18.58
((CH3)3C–), 18.10 ((CH3)3C–), 17.98 ((CH3)3C–), Ϫ3.47
(CH3Si–), Ϫ3.61 (CH3Si–), Ϫ3.61 (CH3Si–), Ϫ3.83
(CH3Si–); Long-range HETCOR correlations: H5!C:10,
3,6; H6!C:7,8,10; H3!C:1,5; H4!C:2,11; m/z (DCI):
677 [MϩH]ϩ, 619 [MϪ57]ϩ; IR (KBr) nmax: 1619, 1432,
nmax
:
3500–3300(OH), 1665, 1615, 1451, 1380,
1174 cmϪ1
.
30,5,7-Tri-O-tert-butyldimethylsilylhesperetin, 6a. 6a was
formed from 500 mg of hesperetin by a procedure similar to
the one described for 3a. Recrystallization from methanol/
water gave a yellow crystalline material, mp 82ЊC. Yield
950 mg (89%). Anal. found: C, 63.13; H, 8.64. C34H56O6Si3
requires C, 63.31; H 8.75; d 1H (CDCl3) 6.97 (dd, 1H,
Jˆ8.3,2.2 Hz, H60), 6.91 (d, 1H, Jˆ2.3 Hz, H20), 6.85 (d,
1H, Jˆ8.4 Hz, H50), 6.11 (d, 1H, Jˆ2.3 Hz, H8), 5.94 (d,
1H, Jˆ2.3 Hz, H6), 5.24 (dd, 1H, Jˆ12.7,2.7 Hz, H2), 3.79
(s, 3H, CH3), 2.94 (dd, 1H, Jˆ12.9,16.4 Hz, H3a), 2.67 (dd,
1H, Jˆ16.4,2.8 Hz, H3b), 1.03 (s, 9H, (CH3)3Si–), 0.98 (s,
9H, (CH3)3Si–), 0.95 (s, 9H, (CH3)3Si–), 0.24 (s, 3H,
CH3Si–), 0.23 (s, 3H, CH3Si–), 0.21 (s, 6H, CH3Si–),
0.15 (s, 3H, CH3Si–), 0.14 (s, 3H, CH3Si–); d 13C
(CDCl3) 189.16 (C4), 164.05 (C9), 162.01 (C7), 158.25
(C5), 151.21 (C40), 145.28 (C30), 131.67 (C10), 119.55
(C60), 119.06 (C20), 112.23 (C50), 109.25 (C10), 106.63
(C6), 102.09 (C8), 78.59 (C2), 55.57 (CH3), 45.71 (C3a),
25.84 ((CH3)3C–), 25.73 ((CH3)3C–), 25.55 ((CH3)3C–),
18.51 ((CH3)3C–), 18.45 ((CH3)3C–), 18.22 ((CH3)3C–),
Ϫ4.26 (CH3Si–), Ϫ4.32 (CH3Si–), Ϫ4.33 (CH3Si–),
Ϫ4.56 (CH3Si–), Ϫ4.57 (CH3Si–); Long-range HETCOR
correlations: H60!C:40,20; H20!C:40,30,60; H50!C:30,
10,50; H8!C:9,7,10,6; H6!C:7,5,10,8; H2!C:4,60,20,2;
CH3!C:40; H3a!C:4,2; H3b!C:4; m/z (DCI): 645
[MϩH]ϩ, 587 [MϪ57]ϩ; IR (KBr) nmax: 1687, 1602,
1386, 1253, 1014, 851, 769 cmϪ1
.
2,7,8-Tri-O-tert-butyldimethylsilylpurpurogallin,
7b.
Sonication of 7a following a procedure similar to the one
described for 1b gave 7b after 3 h. Purification by straight-
phase preparative HPLC using a heptane/dichloromethane
gradient gave a brown oil. Yield 0.95 g (90%). Anal. found:
C, 61.64; H, 8.76. C29H50O5Si3 requires C, 61.87; H, 8.95; d
1H (CDCl3) 15.34 (s, 1H, –OH9), 7.09 (dd, 1H,
Jˆ12.3,0.8 Hz, H5), 6.90 (dd, 1H, Jˆ9.5,0.8 Hz, H3),
6.67 (s, 1H, H6), 6.48 (dd, 1H, Jˆ11.4,9.5 Hz, H4), 1.03
(s, 9H, (CH3)3Si–), 1.00 (s, 9H, (CH3)3Si–), 0.99 (s, 9H,
(CH3)3Si–), 0.29 (s, 5H, CH3Si–), 0.27 (s, 6H, CH3Si–),
0.24 (s, 6H, CH3Si–); d 13C (CDCl3) 186.49 (C1), 157.71
(C9), 154.92 (C2), 151.96 (C7), 136.71 (C8), 135.57 (C5),
133.12 (C11), 122.48 (C3), 122.43 (C4), 117.52 (C10),
114.87 (C6), 26.04 ((CH3)3C–), 26.02 ((CH3)3C–), 25.91
((CH3)3C–), 18.90 ((CH3)3C–), 18.81 ((CH3)3C–), 18.69
((CH3)3C–), Ϫ3.66 (CH3Si–), Ϫ3.75 (CH3Si–), Ϫ3.97
(CH3Si–); Long-range HETCOR correlations: H9!C:9,
8,10; H5!C:10,6; H3!C:1,2,4; H6!C:7,8,10; H4!C:2,
11; m/z (DCI): 563 [MϩH]ϩ, 505 [MϪ57]ϩ; IR (KBr)
1515, 1257, 1166, 1098, 839, 783 cmϪ1
.
30,7-Di-O-tert-butyldimethylsilylhesperetin, 6b. Sonica-
tion of 6a following a procedure similar to the one described
for 1b gave 6b after 3 h. Recrystallization from ethanol
afforded a light yellow crystalline material, mp 143ЊC.
Yield 711 mg (91%). Anal. found: C, 63.47; H, 8.04.
C28H42O6Si2 requires C, 63.36; H, 7.98. d 1H (CDCl3)
11.93 (s, 1H, –OH5), 6.99 (dd, 1H, Jˆ8.3,2.3 Hz, H60),
nmax: 1631, 1424, 1378, 1258, 1007, 845, 767 cmϪ1
.

A. H. De Groot et al. / Tetrahedron 56 (2000) 1541–1549
1547
1
7,8-Di-O-tert-butyldimethylsilylpurpurogallin, 7c. Soni-
cation of 7b following a procedure similar to the one
described for 1b gave 7c after 48 h. Purification by
straight-phase preparative HPLC using a heptane/dichloro-
methane gradient gave a brown oil. Yield 0.34 g (45%).
Anal. found: C, 61.73; H, 8.18. C23H36O5Si2 requires C,
C26H50O4Si3 requires C, 61.12; H 9.86; d H (CDCl3) 5.92
(s, 2H, H30ϩH50), 2.38 (s, 3H, H2), 0.92 (s, 9H, (CH3)3Si–),
0.90 (s, 18H, (CH3)3Si–), 0.15 (s, 18H, CH3Si–); d 13C
(CDCl3) 202.21 (C1), 157.08 (C40), 153.55 (C20ϩC60),
120.60 (C10), 104.81 (C30ϩC50), 32.46 (C2), 25.60
((CH3)3C–), 25.57 ((CH3)3C–), 18.19 ((CH3)3C–), 18.02
((CH3)3C–), Ϫ4.41 (CH3Si–); m/z (DCI): 511 [MϩH]ϩ,
453 [MϪ57]ϩ; IR (KBr) nmax: 1615, 1280, 1252, 1171,
1
61.57; H, 8.09; d H (CDCl3) 14.70 (s, 1H, –OH9), 8.49
(s, 1H, –OH2), 7.21 (dd, 1H, Jˆ12.3,0.7 Hz, H5), 7.10 (dd,
1H, Jˆ9.2,0.8 Hz, H3), 6.77 (s, 1H, H6), 6.70 (dd, 1H,
Jˆ11.2,9.2 Hz, H4), 1.04 (s, 9H, (CH3)3Si–), 1.01 (s, 9H,
(CH3)3Si–), 0.306 (s, 6H, CH3Si–), 0.247 (s, 6H, CH3Si–);
835, 781 cmϪ1
.
40-tert-Butyldimethylsilyloxy-20,60-dihydroxyaceto-
phenone, 9b. Sonication of 9a following a procedure simi-
lar to the one described for 1b gave 9b after 12 h.
Purification by straight-phase preparative HPLC using a
heptane/dichloromethane gradient gave a colourless oil.
Yield 615 mg (85%). Anal. found: C, 59.66; H, 7.92.
d
13C (CDCl3) 182.62 (C1), 156.94 (C9), 154.84 (C2),
153.05 (C7), 134.96 (C8), 134.60 (C5), 134.29 (C11),
124.06 (C4), 116.54 (C10), 115.67 (C3), 114.96 (C6),
26.01 ((CH3)3C–), 25.97 ((CH3)3C–), 18.74 ((CH3)3C–),
18.04 ((CH3)3C–), Ϫ3.65 (CH3Si–), Ϫ3.77 (CH3Si–);
Long-range HETCOR correlations: H9!C:9,8,10;
H5!C:10,6; H3!C:1,2,4; H6!C:7,8,10; H4!C:2,11;
1
C14H22O4Si requires C, 59.54; H, 7.85; d H (CDCl3) 5.94
(s, 2H, H30ϩH50), 2.75 (s, 3H, H2), 0.94 (s, 9H, (CH3)3Si–),
0.20 (s, 6H, CH3Si–); d 13C (CDCl3) 204.56 (C1), 163.61
(C20ϩC60), 163.42 (C40), 105.82 (C10), 100.08 (C30ϩC50),
32.52 (C2), 25.42 ((CH3)3C–), 18.107 ((CH3)3C–),
Ϫ4.48 (CH3Si–); m/z (DCI): 283 [MϩH]ϩ, 225
[MϪ57]ϩ; IR (KBr) nmax: 1632, 1282, 1252, 1167, 830,
m/z (DCI): 449 [MϩH]ϩ, 391 [MϪ57]ϩ; IR (KBr) nmax
:
1635, 1427, 1377, 1254, 1003, 846, 763 cmϪ1
.
2,4,6-Tri-tert-butyldimethylsilyloxybenzaldehyde, 8a. 8a
was formed from 0.5 g of 2,4,6-trihydroxybenzaldehyde by
a procedure similar to the one described for 1a. Purification
by straight-phase preparative HPLC using a heptane/
dichloromethane gradient gave a light brown oil. Yield
1.5 g (93%). Anal. found: C, 60.10; H, 9.91. C25H48O4Si3
777 cmϪ1
.
203,40,60-Tetra-tert-butyldimethylsilyloxy-4-methoxy-
chalcone, 10a. 10a was formed from 80 mg of 3,20,40,60-
tetrahydroxy-4-methoxychalcone by a procedure similar to
the one described for 3a. Purification by straight-phase
heptane/dichloromethane
gradient gave a yellow oil. Yield 175 mg (87%). Anal.
found: C, 63.11; H, 9.44; C40H70O6Si4 requires C, 63.27;
1
requires C, 60.43; H 9.74; d H (CDCl3) 10.27 (s, 1H, H1),
5.93 (s, 2H, H30ϩH50), 1.00 (s, 18H, (CH3)3Si–), 0.97 (s,
9H, (CH3)3Si–), 0.24 (s, 12H, CH3Si–), 0.23 (, 6H, CH3Si–
preparative HPLC using
a
13
); d C (CDCl3) 188.07 (C1), 166.33 (C40), 159.89
(C20ϩC60), 114.89 (C10), 106.48 (C30ϩC50), 25.61
((CH3)3C–), 25.49 ((CH3)3C–), 18.41 ((CH3)3C–), 18.19
((CH3)3C–), Ϫ4.38 ((CH3Si–), Ϫ4.42 ((CH3Si–). m/z
(DCI): 497 [MϩH]ϩ, 439 [MϪ57]ϩ; IR (KBr) nmax: 1627,
1
H 9.29 d H (CDCl3) 7.20 (d, 1H, Jˆ16.1 Hz, Hb), 7.02
(dd, 1H, Jˆ8.2,2.1 Hz, H6), 7.00 (d, 1H, Jˆ2.0 Hz, H2)
6.80 (d, 1H, Jˆ8.2 Hz, H5), 6.74 (d, 1H, Jˆ16.0 Hz, Ha),
6.00 (s, 2H, H30ϩH50), 3.80 (s, 3H, CH3), 0.99 (s, 9H,
(CH3)3Si–), 0.98 (s, 9H, (CH3)3Si–), 0.86 (s, 18H,
(CH3)3Si–), 0.22 (s, 6H, CH3Si–), 0.15 (s, 12H, CH3Si–)
1604, 1477, 1319, 1251, 1160, 1078, 846, 779 cmϪ1
.
4,6-Di-tert-butyldimethylsilyloxy-2-hydroxybenzaldehyde,
8b. Sonication of 8a following a procedure similar to the
one described for 1b gave 8b after 10 min. Purification by
straight-phase preparative HPLC using a heptane/ether
gradient gave a light brown oil. Yield 1.12 g (97%). Anal.
found: C, 59.53; H, 8.98. C19H34O4Si2 requires C, 59.64; H,
13
0.14 (s, 6H, CH3Si–); d C (CDCl3) 194.38 (Cb 0), 157.10
(C40), 154.41 (C20ϩC60), 153.00 (C4), 145.24 (C3) 144.38
(Cb), 127.47 (Ca), 123.21 (C6), 119.82 (C2), 118.88 (C10),
111.82 (C5), 104.89 (C30ϩC50), 55.35 (CH3) 25.66
((CH3)3C–), 25.65 ((CH3)3C–), 25.51 ((CH3)3C–), 18.38
((CH3)3C–), 18.24 ((CH3)3C–), 17.96 ((CH3)3C–), Ϫ4.36
(CH3Si–), Ϫ4.37 (CH3Si–), Ϫ4.69 (CH3Si–); Long-range
HETCOR correlations: Hb!C:b 0,6,2 H6!C:4,3,b,2;
H2!C:4,3,b,6; H5!C:3,1; Ha!C:1; H30ϩH50!
C:40,20ϩ60; CH3!C:4; m/z (DCI): 759 [MϩH]ϩ, 701
[MϪ57]ϩ; IR (KBr) nmax: 1615, 1509, 1253, 1157, 1104,
1
8.96; d H (CDCl3) 12.11 (s, 1H, OH20), 10.10 (d, 1H,
Jˆ0.5 Hz, H1), 5.99 (dd, 1H, Jˆ2.3,0.7 Hz, H30), 5.80 (d,
1H, Jˆ2.1 Hz, H50) 1.01 (s, 9H, (CH3)3Si–), 0.97 (s, 9H,
(CH3)3Si–), 0.29 (s, 6H, CH3Si–), 0.25 (s, 6H, CH3Si–), d
13C (CDCl3) 192.44 (C1), 165.50 (C20), 164.82 (C40),
160.78 (C60), 109.13 (C10), 102.41 (C50), 101.30 (C30),
25.64 ((CH3)3C–), 25.52 ((CH3)3C–), 18.35 ((CH3)3C–),
18.26 ((CH3)3C–), Ϫ4.40 (CH3Si–), Ϫ4.43 (CH3Si–);
837, 764 cmϪ1
.
Long-range
HETCOR
correlations:
H1!C:20,10;
3,40-Di-tert-butyldimethylsilyloxy-20,60-dihydroxy-4-
methoxychalcone, 10b. Sonication of 10a following a
procedure similar to the one described for 1b gave 10b
after 20 h. Purification by straight-phase preparative
HPLC using a heptane/dichloromethane gradient gave a
brown oil. Yield 103 mg (84%). Anal. found: C, 63.41; H,
H20ϩH60!C:20,40,10,30; H30ϩH50!C:20,40,10,50; H50!C:
40,60,10,30; m/z (DCI): 383 [MϩH]ϩ, 325 [MϪ57]ϩ; IR
(KBr) nmax: 1642, 1605, 1475, 1321, 1255, 1156, 1078,
845, 776 cmϪ1
.
1
20,40,60-Tri-tert-butyldimethylsilyloxyacetophenone, 9a.
9a was formed from 0.5 g of 20,40,60-trihydroxyaceto-
phenone by a procedure similar to the one described for
3a. Purification by straight-phase preparative HPLC using
a heptane/dichloromethane gradient gave a colourless oil.
Yield 1.31 g (95%). Anal. found: C, 60.96; H, 9.97.
8.04; C28H42O6Si2 requires C, 63.36; H, 7.98; d H (CDCl3)
10.5 (br.s, 2H, –OH20ϩ60), 7.95 (d, 1H, Jˆ15.6 Hz, Hb),
7.58 (d, 1H, Jˆ15.4 Hz, Ha), 7.11 (dd, 1H, Jˆ8.6,2.3 Hz,
H6), 7.03 (d, 1H, Jˆ2.1 Hz, H2), 6.81 (d, 1H, Jˆ8.5 Hz,
H5), 5.84 (s, 2H, H30ϩH50), 3.77 (s, 3H, CH3), 0.92 (s, 9H,
(CH3)3Si–), 0.89 (s, 9H, (CH3)3Si–), 0.17 (s, 6H, CH3Si–),

1548
A. H. De Groot et al. / Tetrahedron 56 (2000) 1541–1549
13
0.09 (s, 6H, CH3Si–); d C (CDCl3) 192.78 (Cb 0), 163.26
(C20ϩC60), 162.65 (C40), 153.28 (C4), 145.32 (C3), 143.47
(Cb), 124.91 (Ca), 123.64 (C6), 120.35 (C2), 111.98 (C5),
106.20 (C10), 100.54 (C30ϩC50), 55.51 (CH3), 25.75
((CH3)3C–), 25.57 ((CH3)3C–), 18.49 ((CH3)3C–), 18.23
((CH3)3C–), Ϫ4.31 (CH3Si–), Ϫ4.56 (CH3Si–); Long-
range HETCOR correlations: H20ϩH60!C:10,20ϩ
60,30ϩ05; Hb!C:b 0,6,2; Ha!C:1; H6!C:4,3,b,2;
20,4,40-Tri-tert-butyldimethylsilyloxy-2-hydroxybenzo-
phenone, 12b. Sonication of 12a following a procedure
similar to the one described for 1b gave 12b after 24 h.
Purification by straight-phase preparative HPLC using a
heptane/dichloromethane gradient gave a yellow oil. Yield
962 mg (84%). Anal. found: C, 63.09; H, 8.83. C31H52O5Si3
1
requires C, 63.21; H, 8.90; d H (CDCl3) 12.555 (s, 1H,
–OH2), 7.30 (d, 1H, Jˆ9.0 Hz, H6), 7.20 (d, 1H,
Jˆ8.3 Hz, H60), 6.55 (dd, 1H, Jˆ8.3,2.4 Hz, H50), 6.40
(d, 1H, Jˆ2.3 Hz, H3), 6.36 (d, 1H, Jˆ2.1 Hz, H30), 6.28
(dd, 1H, Jˆ8.8,2.4 Hz, H5), 1.00 (s, 9H, (CH3)3Si–),
0.98 (s, 9H, (CH3)3Si–), 0.73 (s, 9H, (CH3)3Si–), 0.24
(s, 6H, CH3Si–), 0.23 (s, 6H, CH3Si–), 0.07 (s, 6H,
CH3Si–), d 13C (CDCl3) 200.26 (CˆO), 165.16 (C2),
H2!C:4,3,b,6;
H5!C:3,1;
H30ϩH50!C:4020ϩ60;
CH3!C:4; m/z (DCI): 531 [MϩH]ϩ, 473 [MϪ57]ϩ;
IR (KBr) nmax: 1607, 1510, 1256, 1154, 1100, 839,
761 cmϪ1
.
tert-Butyldimethylsilyl
20,40,60-tri-tert-butyldimethyl-
ء
ء
silyloxybenzoate, 11a. 11a was formed from 0.5 g of
20,40,60-trihydroxybenzoic acid by a procedure similar to
the one described for 1a. Recrystallization from methanol/
water gave a white crystalline material, mp 184ЊC. Yield
1.466 g (88%). Anal. found: C, 59.20; H, 10.18;
162.93 (C4), 158.69 (C20 ), 153.77 (C40 ), 135.95
(C6), 130.17 (C60), 124.72 (C10), 115.51 (C1), 115.51
(C50), 111.94 (C5), 111.65 (C30), 107.78 (C3), 25.66
((CH3)3C–), 25.59 ((CH3)3C–), 25.32 ((CH3)3C–),
18.27 (2X, (CH3)3ϩC–), 17.86 ((CH3)3C–), Ϫ4.34
(CH3Si–), Ϫ4.36 (CH3Si–), Ϫ4.54 (CH3Si–); Long-
1
C31H62O5Si4 requires C, 59.37; H 9.96; d H (CDCl3) 5.94
(s, 2H, H30ϩH50), 0.964 (s, 9H, (CH3)3Si–), 0.960 (s, 18H
(CH3)3Si–), 0.93 (s, 9H (CH3)3Si–), 0.34 (s, 6H, CH3Si–),
0.22 (s, 12H, CH3Si–), 0.19 (s, 6H CH3Si–); d 13C (CDCl3)
166.72 (C1), 156.46 (C20ϩC60), 153.81 (C40), 115.46 (C10),
104.36 (C30ϩC50), 25.87 ((CH3)3C–), 25.70 ((CH3)3C–),
25.65 ((CH3)3C–), 18.38 ((CH3)3C–), 18.23 ((CH3)3C–),
17.66 ((CH3)3C–), Ϫ4.03 (CH3Si), Ϫ4.36 (CH3Si); m/z
(DCI): 627 [MϩH]ϩ, 569 [MϪ57]ϩ; IR (KBr) nmax: 1629,
range
HETCOR
correlations:
H2!
C:2,1,3;
H6!C:CO,2,4;
H60!C:CO,20,40;
H50!C:10,30;
H3!C:1,5; H30!C:10,50; H5!C:1; m/z (DCI): 589
[MϩH]ϩ, 531 [MϪ57]ϩ; IR (KBr) nmax: 1627, 1496,
1259, 1102, 839, 783 cmϪ1
.
4,40Di-tert-butyldimethylsilyloxy-2,20-dihydroxybenzo-
phenone, 12c. Sonication of 12b following a procedure
similar to the one described for 1b gave 12c after six
days. Purification by straight-phase preparative HPLC
using a heptane/dichloromethane gradient gave a yellow
oil. Yield 289 mg (37%; the remaining fraction consists of
the mono-silylether and the unprotected benzophenone deri-
vative). Anal. found: C, 63.17; H, 8.03. C25H38O5Si2
requires C, 63.25; H, 8.07; d 1H (CDCl3) 11.12 (s,
–OH2ϩ20), 7.51 (d, H, Jˆ8.8 Hz, H6 (ϩ60)), 6.48 (d, H,
Jˆ2.4 Hz, H3 (ϩ30)), 6.40 (dd, H, Jˆ8.8,2.4 Hz, H5 (ϩ50)),
1.00 (s, H, Jˆ18.3 Hz, (CH3)3Si–), 0.27 (s, H, Jˆ10.2 Hz,
1284, 1164, 1065, 834 cmϪ1
.
40-tert-Butyldimethylsilyloxy-20,60-dihydroxybenzoic
acid, 11b. Sonication of 11a following a procedure similar
to the one described for 1b gave 11b after 30 h. Recrystal-
lization from methanol/water afforded a white crystalline
material, mp 199ЊC. Yield 493 mg (74%). Anal. found: C,
1
54.88; H, 7.08. C13H20O5Si requires C, 54.91; H, 7.09; d H
(acetone-d6) 5.87 (s, 2H, H30ϩH50), 0.94 (s, 9H, (CH3)3Si–),
13
0.20 (s, 6H, CH3Si–); d C (acetone-d6) 172,03 (C10),
13
163,56 (C40), 160,03 (C20ϩC60), 100,02 (C10), 95,48
(C30ϩC50), 25,89 ((CH3)3C–), 18,76 ((CH3)3C–), Ϫ4,28
(CH3Si–); m/z (CDI): 285 [MϩH]ϩ, 241 [MϪ44]ϩ, 227
[MϪ57]ϩ; IR (KBr) nmax: 1637, 1472, 1279, 1254, 1166,
CH3Si–); d C (CDCl3) 199.53 (CˆO), 164.30 (C2ϩC20),
162.48 (C4ϩC40), 134.52 (C6ϩC60), 114.39 (C1ϩC10),
111.81 (C5ϩC50), 108.63 (C3ϩC30), 25.57 ((CH3)3C–),
18.26 ((CH3)3C–), Ϫ4.31 (CH3Si–); Long-range HETCOR
correlations: H2ϩH20!C:2; H6ϩH60!C:2,4; H3ϩH30!
C:1,5; H5ϩH50!C:3; m/z (DCI): 475 [MϩH]ϩ, 417
[MϪ57]ϩ; IR (KBr) nmax: 1632, 1509, 1259, 1172, 1092,
1066, 837, 740 cmϪ1
.
2,20,4,40-Tetra-tert-butyldimethylsilyloxybenzophenone,
12a. 12a was formed from 0.5 g of 2,20,4,40-tetrahydroxy-
benzophenone by a procedure similar to the one described
for 3a. Purification by straight-phase preparative HPLC
using a heptane/dichloromethane gradient gave a yellow
oil. Yield 1.37 g (96%). Anal. found: C, 63.03; H, 9.33.
838, 786 cmϪ1
.
tert-Butyldimethylsilyl 30,40,50-tri-O-tert-butyldimethyl-
silylgallate, 13a. 13a was formed from 0.5 g of gallic acid
by a procedure similar to the one described for 1a. 1H NMR
did not show any impurities and product was further used
without purification. Yield 1.63 g (98%). Anal. found: C,
59.32; H, 9.85. C31H62O5Si4 requires C, 59.37; H 9.96; d
1H (CDCl3) 7.23 (s, 2H, H20ϩH60), 1.02 (s, 9H, (CH3)3Si–),
1.01 (s, 9H, (CH3)3Si–), 0.96 (s, 18H, (CH3)3Si–), 0.36 (s,
5H, CH3Si–), 0.25 (s, 12H, CH3Si–), 0.15 (s, 6H, CH3Si–);
1
C37H66O5Si4 requires C, 63.19; H 9.46; d H (CDCl3) 7.35
(d, 2H, Jˆ8.4 Hz, H6 (ϩ60)), 6.42 (dd, 2H, Jˆ8.5,2.3 Hz,
H5 (ϩ50)), 6.24 (d, 2H, Jˆ2.2 Hz, H3 (ϩ30)), 0.96 (s, 18H,
(CH3)3Si–), 0.79 (s, 18H, (CH3)3Si–), 0.19 (s, 12H, CH3Si–),
0.05 (s, 12H, CH3Si–); d 13C (CDCl3) 193.89 (CˆO),
ء
ء
159.27 (C2ϩC20 ), 155.83 (C4ϩC40 ), 132.80 (C6ϩC60),
126.41 (C1ϩC10), 112.90 (C5ϩC50), 111.35 (C3ϩC30),
25.66 ((CH3)3C–), 25.59 ((CH3)3C–), 18.18 ((CH3)3C–),
18.15 ((CH3)3C–), Ϫ4.37 (CH3Si), Ϫ4.41 (CH3Si–);
Long-range HETCOR correlations: H6ϩH60!C:CO,2,4;
H5ϩH50!C:1,3; H3ϩH30!C:2,4,1,5; m/z (DCI): 703
[MϩH]ϩ, 645 [MϪ57]ϩ; IR (KBr) nmax: 1619, 1497,
d
13C (CDCl3) 166.12 (C1), 148.33 (C30ϩC50), 143.19
(C40), 123.40 (C10), 116.05 (C20ϩC60), 26.24 ((CH3)3C–
), 26.18 ((CH3)3C–), 25.66 ((CH3)3C–), 18.81 ((CH3)3C–),
18.49 ((CH3)3C–), 17.72 ((CH3)3C–), Ϫ3.60 (CH3Si–),
Ϫ3.83 (CH3Si–), Ϫ4.75 (CH3Si–); m/z (DCI): 627
[MϩH]ϩ, 569 [MϪ57]ϩ; IR (KBr) nmax: 1627, 1496,
1257, 1168, 1106, 852, 775 cmϪ1
.
1259, 1180, 1104, 839, 783 cmϪ1
.

A. H. De Groot et al. / Tetrahedron 56 (2000) 1541–1549
1549
30,40,50-Tri-O-tert-butyldimethylsilylgallic acid, 13b.
Sonication of 13a following a procedure similar to the one
described for 1b gave 13b after 10 h. Recrystallization from
methanol afforded white needles, mp 237ЊC. Yield 1.31 g
(98%). Anal. found: C, 58.60; H, 9.43. C25H48O5Si3 requires
3. Ley, S. V.; Low, C. M. R. Ultrasound in Synthesis, Springer:
Berlin, 1989.
4. Schmittling, E. A.; Sawyer, J. S. Tetrahedron Lett. 1991, 32,
7207–7210.
5. Lee, A. S. Y.; Yeh, H. C.; Tsai, M. H. Tetrahedron Lett. 1995,
36, 6891–6894.
1
C, 58.54; H, 9.43; d H (CDCl3) 7.29 (s, 2H, H20ϩH60),
1.00 (s, 9H, (CH3)3Si–), 0.96 (s, 18H, (CH3)3Si–), 0.25 (s,
12H, CH3Si–), 0.16 (s, 6H, CH3Si–); d 13C (CDCl3) 171.75
(C1), 148.53 (C30ϩC50), 144.10 (C40), 121.13 (C10), 116.17
(C20ϩC60), 26.22 ((CH3)3C–), 26.15 ((CH3)3C–), 18.83
((CH3)3C–), 18.54 ((CH3)3C–), Ϫ3.61 (CH3Si–), Ϫ3.84
(CH3Si–); m/z (DCI): 513 [MϩH]ϩ, 469 [MϪ44]ϩ, 455
[MϪ57]ϩ.
6. Cavelier, F.; Enjalbal, C. Tetrahedron Lett. 1996, 37, 5131–
5134.
7. Lee, A. S. Y.; Yeh, H. C.; Yeh, M. K.; Tsai, M. H. J. Chin.
Chem. Soc. 1995, 42, 919–922.
8. Fitzgerald, M. E.; Griffing, V.; Sullivan, J. J. Chem. Phys. 1956,
25, 926.
9. Lorimer, J. P.; Mason, T. M.; Mistry, B. P. Ultrasonics 1987,
25, 23.
10. Luche, J.-L.; Einhorn, C.; Einhorn, J. Tetrahedron Lett. 1990,
31, 4125–4130.
Acknowledgements
11. Eshuis, J. J. W. Tetrahedron Lett. 1994, 35, 7833–7836.
12. Dean, F. M.; Goodchild, J.; Houghton, L. E. Tetrahedron Lett
1966, 7, 4153–4159.
13. Horie, T.; Kobayashi, T.; Kawamura, Y. Bull. Chem. Soc. Jpn
1995, 68, 2033–2041.
We wish to thank F. Boers and I. Daniel for their assistance
and E. Esmans and J. Verreydt for the mass spectra. This
work was supported by the Flemish Government ‘concerted
action 92/94-90 and by the FWO.
14. Haraldsson, G. G.; Baldwin, J. E. Tetrahedron 1997, 53, 215–
224.
References
15. Montaudo, G.; Finocchiaro, P.; Maravigna, P. J. Am. Chem.
Soc. 1971, 93, 4214–4217.
16. Buchanan, G. W.; Montaudo, G.; Finocchiaro, P. Can. J.
Chem. 1973, 51, 1053–1059.
1. For a review: Nelson, T. D.; Crouch, R. D. Synthesis 1996
1031–1069.
2. Einhorn, C.; Einhorn, J.; Luche, J.-L. Synthesis 1989, 787–813.