A study on the synthesis of calix[4 and 6]arenes using alternately arranged phloroglucinols and p-tert-butylphenols

Article abstract of DOI:10.1248/cpb.56.1173

A method for the synthesis of calix[4 and 6]arenes with two or three alternately arranged phloroglucinols and p-tert-butylphenols was studied using 3+1 and 5+1 approaches, compared to a simple one-pot synthesis based on a 1+1 approach. By using Yb(OTf)₃ as a catalyst, each calix[4 and 6]arenes was afforded, in an overall yield of 20.7% and 11.8% in the 3+1 and 5+1 approaches, respectively, and 24.6% and 19.9% using a 1+1 approach with chlorobenzene and toluene as refluxing solvents, respectively.

Full text of DOI:10.1248/cpb.56.1173

August 2008
Notes
Chem. Pharm. Bull. 56(8) 1173—1176 (2008)
1173
A Study on the Synthesis of Calix[4 and 6]arenes Using Alternately
Arranged Phloroglucinols and p-tert-Butylphenols
Shingo SATO,* Akihito YASHIRO, Naoki TSUSHIMA, and Kenta HOSAKA
Graduate School of Science and Engineering, Yamagata University; 4–3–16 Jonan, Yonezawa, Yamagata 992–8510,
Japan. Received March 8, 2008; accepted May 19, 2008; published online May 23, 2008
A method for the synthesis of calix[4 and 6]arenes with two or three alternately arranged phloroglucinols
and p-tert-butylphenols was studied using “3ꢀ1” and “5ꢀ1” approaches, compared to a simple one-pot synthesis
based on a “1ꢀ1” approach. By using Yb(OTf)3 as a catalyst, each calix[4 and 6]arenes was afforded, in an over-
all yield of 20.7% and 11.8% in the “3ꢀ1” and “5ꢀ1” approaches, respectively, and 24.6% and 19.9% using a
“1ꢀ1” approach with chlorobenzene and toluene as refluxing solvents, respectively.
Key words calix[4]arene; calix[6]arene; phloroglucinol; p-tert-butylphenol; 1ꢀ1 approach
Various calixarenes composed of phenol derivatives have forded a calix[4]arene 6 in the best yield of 37%,6) as shown
been synthesized and their functionality such as inclusion by Entry 2 in Table 2. When the concentration was either too
property has been developed in molecular recognition and dilute or too concentrated, the yield was not good. Scan-
supramolecular chemistry.1—5) However, to our knowledge, dium(III) trifluoromethanesulfonate [Sc(OTf)3] or ytterbium
calixarenes including phloroglucinols (1,3,5-tri-hydroxyben- (III) trifluoromethanesulfonate [Yb(OTf)3] was used as a cat-
zenes) have not been synthesized. Recently we have first syn- alyst in the place of p-TsOH·H2O, however, the yield was no
thesized a calix[4]arene (1) with two alternately arranged improvement.
phloroglucinols and p-tert-butylphenols and explored its
We next examined the synthesis of calix[6]arene (8) by
property such as determination of the pKa-values by a poten- “5ꢀ1” approach applying the above “3ꢀ1” approach. Linear
tiometric titration and a photometric pH titration methods, pentamer 7 was synthesized by condensation of trimer 5 and
observation of the structure’s change with the change of pH 3 with an equimolar amount of 4 in the yield of 31% along
by NMR spectroscopy, solvent extraction of some alkali with 6 in a yield of 13% under the reaction conditions of stir-
metal ions, and measurement of the cyclic voltammogram.6)
That is due to the property of phlorogrucinol which is easy to
change into keto-form in alkaline solution. In the previous
paper,6) it was found that 1 showed pKa values to be 3—4,
7.5, and the other chelated hydroxyls 11, and caused the
structure-change over pH 11, and extracted only Liꢀ among
alkaline metals at pH 11.
We have been interested in the calix[6]arene (2) which is
expected to be able to extract larger ions than Liꢀ, because 2
has a larger cavity than 1. We have also been interested in the
simple and effective synthetic method for the larger amount
of supply of these calix[4 and 6]arenes (1 and 2) for the
above study.2)
Results and Discussion
The synthesis of calix[4]arene 6 was performed by “3ꢀ1”
approach.1—6) Trimer 5 could be prepared by the condensa-
tion of 2 eq of trimethoxybenzenes (3) and 1 eq of 2,6-di-hy-
droxymethyl-4-tert-butylphenol (4) in the presence of cat-
alytic amounts (0.1 eq) of p-TsOH·H2O, as shown in Table
1. The reaction was carried out by dropwise addition of a so-
lution of 4 in toluene to a solution of 3 and p-TsOH·H2O in
toluene at room temperature for 10 min—1 h and an addi-
tional heating. Even when the temperature and time was
changed, respectively, the yield was scarcely changed. The
best yield of 5 was 56% with a linear pentamer 7 of 21%
yield by stirring at 110 °C for 1 h after 20 min-stirring at
room temperature (Table 1, Entry 3).
The next “3ꢀ1” approach, in which trimer 5 and equimo-
lar amount of 4 were condensed by azeotropic refluxing in
toluene under the conditions of high-dilute concentrations
(0.59 mM) in the presence of 0.45 eq of p-TsOH·H2O,3—5) af-
Fig. 1.
∗ To whom correspondence should be addressed. e-mail: Shingo-s@yz.yamagata-u.ac.jp
© 2008 Pharmaceutical Society of Japan

1174
Vol. 56, No. 8
Table 1. Synthesis of Trimer 5
Chart 1. Synthesis of Linear Pentamer 7
Yield (%)
p-TsOH·H2O
Entry
Temp./Time
(eq)
5
7
Chart 2. Synthetic Attempt of Calix[6]arene 8 Using “5ꢀ1” Approach
under High-Dilution Conditions
1
2
3
4
5
0.1
0.1
0.1
0.2
0.05
rt, 1 h
→60 °C, 5 h
47
52
56
47
44
18
22
21
23
21
rt, 30 min→100 °C, 1 h
rt, 20 min→110 °C, 1 h
rt, 10 min→100 °C, 40 min
rt, 10 min→100 °C, 40 min
Table 3. Synthesis of Calix[6]arene 8 Using “5ꢀ1” Approach under High-
Dilution Conditions
Entry
Conc. (mM)
Yield of 8 (%)
Table 2. Synthesis of Calix[4]arene 6
1
2
3
4
0.408
0.815
1.63
28.8
30.7
28.2
16.9
Entry
Catalyst (eq)
Conc. (mM) Yield of 6 (%)
3.26
1
2
3
4
5
6
p-TsOH·H2O (0.45)
p-TsOH·H2O (0.45)
p-TsOH·H2O (0.45)
Sc(OTf)3 (0.10)
Yb(OTf)3 (0.10)
Yb(OTf)3 (0.20)
0.44
0.59
0.88
3
3
0.44
32
37
26
11
18
22
Table 4. Synthesis of Calix[4 and 6]arenes (6 and 8) Using “1ꢀ1” Ap-
proach under High-Dilution Conditions
Yield (%)
ring in toluene in the presence of a 0.1 eq of p-TsOH·H2O
within a range of room temperature to 60 °C for 1 h (Chart
1). The “5ꢀ1” approach was performed in the same manner
as the above “3ꢀ1” approach, however, these conditions
gave no 8 and only a calix[4]arene 6 in the yield of 21%
(Chart 2). Since these results suggested that scission of the 7
proceeded under these conditions, the reaction conditions
were made to be milder, such as the use of benzene in the
Entry
Conc. (mM)
5
6
7
8
1
2
3
4
1.63
2.45
3.26
6.52
8.7
4.7
2.7
3.5
22.3
20.2
17.9
13.5
2.9
6.1
5.2
1.4
8.5
9.1
9.7
1.7
place of toluene as a refluxing solvent, and the use of Yb and products (5, 6, 7, 8) was measured at regular intervals
(OTf)3 in the place of p-TsOH·H2O as an acid catalyst (Table from the start until 12 h, using sampling and HPLC analysis
3). The best condition was 7-h refluxing in benzene (concen- from some different refluxing solvents under a 3.26 mM con-
tration: 0.815 mM) in the presence of 0.15 eq of Yb(OTf)3 centration (Figs. 2a, b, c, d, e). The reaction was completed as
and gave 8 in the yield of 30.7% (Table 3, Entry 2). When follows: the dropwise-addition of a solution of equimolar 3
the concentration was either too dilute or too concentrated, and 4 to a refluxing solution of Yb(OTf)3 was continued dur-
the yield was not good.
ing 6 h, with 6 h of additional refluxing. When benzene was
Further, we tried a simpler one-step synthetic method, a used as a solvent, the formation of 6 was begun immediately,
“1ꢀ1” approach using the condensation reaction of equimo- while the formation of 8 occurred at 4.5 h (Fig. 2a). The pro-
lar amounts of 3 and 4 under high-dilution concentrations duction ratios of 6 and 8 were 38 and 18%, respectively, and
(Table 4).2,8) Under the same conditions as the above that of this ratio did not change from 6 to 12 h. The same reaction
8, the reaction was carried out using the four different con- under the higher refluxing temperature using toluene
centrations to result in 5, 6, 7, and 8 as a main product. The (bpꢁ118 °C) as a solvent was carried out (Fig. 2b). Both pro-
best yield of 8 was 9.7% with 6 of 17.9% in the concentra- duction ratios for 6 and 8 were the same, and gradually in-
tion of 3.26 mM (Table 3, Entry 3), and the best yield of 6 creased reaching the maximum of 32% after 5 h—beyond
was 22.3% with 8 of 8.5% in the concentration of 1.63 mM which the ratio did not change. The reaction was finished
(Table 4, Entry 1). “1ꢀ1” Approach gave many products, after 12 h and the each reaction product was isolated to af-
however, purification of calix[4 and 6]arenes (6 and 8) was ford 6 in 22.0% and 8 in 19.9% (Table 5). When xylenes
relatively easy because of ease of crystallization of the cyclic (bpꢁ138—145 °C) were used as a solvent, trimers 5ꢂ, which
oligomers. After calix[4]arene 6 was firstly afforded by re- were condensed between 4 and xylenes in the place of 3, was
crystallization from EtOAc, the filtrate was crude-separated produced as a main product (35—50%). Trimers 5ꢂ may not
by silica-gel column chromatography (n-hexane–EtOAc) and be able to react further due to its own steric hindrance. Next,
the resulted fractions including 8 afforded pure 8 as crystals the same reaction using chlorobenzene (bpꢁ132 °C) as a sol-
by recrystallization from EtOAc. The production ratio of 3 vent which is the more polar and does not react itself, was

August 2008
1175
Table 5. Synthetic Study of Calix[4 and 6]arenes (6 and 8) by “1ꢀ1” Ap-
proach under High-Dilution Reaction Conditions Using Some Refluxing
Solvents
Yield (%)
Solvent
5
6
7
8
Benzene
Toluene
Chlorobenzene
Ethylbenzene
2.8
Trace
Trace
27.3
17.9
22.0
24.6
23.4
5.2
9.3
14.0
8.3
9.7
19.9
12.8
6.1
Chart 3
was low, at 10%. When the refluxing temperature was high,
the starting material 3 almost reacted immediately. The reac-
tion products after 12 h were isolated in the same manner.
Cyclic tetramer 6 and hexamer 8 was afforded in 24.6% and
12.8% yield, respectively (Table 5). Finally ethylbenzene,
having the high boiling point (bpꢁ136 °C, Fig. 2e) was used
as a refluxing solvent in the place of xylenes and chloroben-
zene. However, unexpectedly the unreactive 5 and 7 were re-
mained at the ratio of 14% and 9%, respectively, and the pro-
duction ratio of 6 and 8 was low, at 25% and 8%, respec-
tively. The isolated yield of 6 and 8 was 23.4% and 6.1%, re-
spectively (Table 5). For the one-step synthesis using a
“1ꢀ1” approach, when the more polar chlorobenzene was
used as a refluxing solvent, the isolated yield of cyclic
tetramer 6 reached the maximum of 24.6%. The maximum
yield of the more unstable cyclic hexamer 8 was 19.9% when
toluene was used as a refluxing solvent.
Deprotection of 8 was achieved, in the same manner as
that of 6,6) by BBr3 at ꢃ78 °C to room temperature to give 2
in the yield of 90%. Since 1H- and 13C-NMR spectra of 8 and
2 both gave a symmetry singlet peak for each proton and car-
bon of the tert-butylphenol, phloroglucinol, and methylene
moieties, it was assumed that the structures of both 8 and 2
are symmetrical cyclic in analogy with 6 and 1.
Conclusion
Fig. 2. Product Ratio vs. Reaction Time in the Condensation Reaction of
Trimethoxybenzene (3) with Bis-hydroxymethyl-tert-butylphenol (4) in the
Presence of Yb(OTf)3 in Refluxing Solvent (3: ꢀ, 5: ꢁ, 5ꢂ: ꢂ, 6: ꢃ, 7: ꢄ, 8:
ꢅ)
We studied on the simple and effective synthesis of the
calix[4 and 6]arenes including two or three alternately
arranged phloroglucinols and p-tert-butylphenols using ei-
ther a “3ꢀ1” or a “5ꢀ1” approach, and a simple “1ꢀ1” ap-
proach for both using a refluxing solvent. The “3ꢀ1” ap-
Solvents: (a) benzene, (b) toluene, (c) xylenes, (d) chlorobenzene, (e) ethylbenzene.
carried out (Fig. 2d). Under the condition of refluxing in proach requiring a two-step reaction gave the cyclic tetramer
chlorobenzene, the production ratio of 6 was immediately 6 in the overall yield of 20.7%, and the “5ꢀ1” approach
reached at a yield of 50—60% and the production ratio of 8 requiring a three-step reaction gave the cyclic hexamer 8 in

1176
Vol. 56, No. 8
der. The filtrate was separated by silica-gel column chromatography (n-
hexane–EtOAcꢁ3 : 1) and recrystallized from CHCl3 to give 8 (69 mg,
11.2%) as a colorless prism.
“1ꢀ1” Approach To a refluxed solution of Yb(OTf)3 (446.4 mg,
0.72 mmol) in dry benzene (900 ml), a solution of trimethoxybenzene (3,
the overall yield of 11.8%. On the other hand, the simple
“1ꢀ1” approach, by changing the reaction solvent, gave 6
and 8 in the maximum yield of 24.6% and 19.9%, respec-
tively. Chlorobenzene and toluene were the most effective
solvents in the one-pot synthesis of the calix[4 and 6 ]arenes 806 mg, 4.8 mmol) and bis-hydroxymethyl-4-tert-butylphenol (4, 1008 mg,
4.8 mmol) in dry benzene (600 ml) was added dropwise for 6 h, and addi-
tional 1-h refluxing. Na2HPO4 (102 mg, 0.72 mmol) was added to the cooled
using the “1ꢀ1” approach, respectively.
reaction mixture and benzene was removed under vacuum until 100 ml. The
residual benzene solution was washed with water and brine, and then dried
with anhydrous Na2SO4. Benzene was removed under vacuum and the resid-
ual solid was crystallized from EtOAc to afford 6 (588 mg, 17.9%). The fil-
trate was separated by silica gel column chromatography (n-hexane–
EtOAcꢁ3 : 1) to give fractions including 8, along with 5 (66 mg, 2.7%) and
7 (213 mg, 5.2%), which were crystallized from EtOAc to give 8 (478 mg,
9.7%) as a white powder.
Experimental
Sc(OTf)3 and Yb(OTf)3, purchased from Taiheiyo Kinzoku Co., Ltd. and
TOKYO KASEI Co., Ltd. respectively, and were directly used without any
further manipuration. The solvents used in this reaction were purified by dis-
tillation. Reactions were monitored by TLC, on 0.25-mm Silica Gel F254
plates (E. Merck) using UV light, and a 7% ethanolic solution of phospho-
molybdic acid with heat as a coloration agent. For separation and purifica-
tion, flash column chromatography was performed on silica gel (230—400
mesh, Fuji-Silysia Co., Ltd., BW-300). Melting points were determined on a
AS ONE ATM-01 melting point apparatus and are uncorrected. IR spectra
were recorded on a Horiba FT-720 IR spectrometer in the form of KBr disks.
NMR spectra were recorded on a Varian Inova 500 spectrometer using
Me4Si as the internal standard. Mass spectral data were obtained by fast-
atom bombardment (FAB) using m-nitrobenzylalcohol as a matrix on a
JEOL JMS-AX505HA instrument. Elemental analyses were performed on a
Perkin-Elmer PE 2400 II instrument.
HPLC Analysis HPLC was performed on a Hitachi L-7100 and L-
4200H system using the following conditions; column, GL Sciences Inc., In-
ertsil ODS-3 column (4.6ꢄ250 mm), solvent, methanol–H2Oꢁ92 : 8: flow
rate, 1 ml/min; and wavelength, UV 254 nm. Sampling (0.8 ml) for HPLC
analysis was carried out at each 0.5 h from the reaction start time until the
finish (12 h). After evaporation of the sample, the residue was dissolved in
AcOEt (ca. 110 ml), and 10 ml was subjected to the HPLC analysis. Reten-
tion time (min) was as follows: 3; 4.34, 5; 7.76, 6; 26.6, 7; 16.6, 8; 18.3.
Synthesis The syntheses and data of trimethoxybenzene (1), bis-hy-
droxymethyl-4-tert-butylphenol (2), and trimer (5) were showed in the previ-
ous paper.6)
Data for 8: Colorless prism (CHCl3). mp ꢅ300 °C. IR (KBr) cmꢃ1: 3400,
1
2952, 2835, 1601, 1485, 1464, 1408, 1325, 1198, 1101. H-NMR (CDCl3)
d: 1.10 (278H, s, tert-Buꢄ3), 3.39 (9H, br s, OMeꢄ3), 3.80 (12H, s, CH2ꢄ
3), 3.87 (18H, s, OMeꢄ6), 6.36 (3H, s, OHꢄ3), 6.84 (6H, s, ArHꢄ6), 7.49
(6H, s, ArHꢄ3). 13C-NMR (CDCl3, at 50 °C) d: 23.8 (C-2, 8, 14, 20, 26,
32), 31.6 (tert-Butyl methyl C), 33.8 (tert-butyl quaternary C), 92.4 (C-11,
23, 35), 115.8 (C-1, 9, 13, 21, 25, 33), 124.4 (C-4, 6, 16, 18, 28, 30), 126.5
(C-3, 7, 15, 19, 27, 31), 141.31 (C-5, 17, 29), 149.56 (C-38, 40, 42), 157.26
(C-10, 12, 22, 24, 34, 36), 157.42 (C-37, 39, 41). FAB-MS m/z: 1027 (Mꢀ
H)ꢀ. Anal. Calcd for C63H78O12·0.6CHCl3: C, 69.51; H, 7.21. Found: C,
69.56; H, 7.15.
11,23,35-Tri-tert-butyl-4,6,16,18,28,30,37,38,39,40,42-dodecahydroxy-
calix[6]arene (2) To a solution of 8 (220 mg, 0.214 mmol) in dry CH2Cl2
(3 ml), BBr3 (912 ml, 9.65 mmol) was added at ꢃ78 °C under argon. The re-
action mixture was stirred at room temperature for 2 d, and then poured into
ice-cold saturated NaHCO3 aqueous solution, then extracted three times with
AcOEt. The organic layer was washed with water and brine, and dried over
anhydrous Na2SO4 and then evaporated under vacuum to give 2 (185 mg,
90%) as a pale-yellow powder.
Data for 2: Pale-yellow powder (EtOAc–diethyl ether). mp ꢅ300 °C. IR
(KBr) cmꢃ1: 3245, 2960, 2870, 1622, 1485, 1446, 1188, 1111. 1H-NMR
(DMSO-d6, at 50 °C) d: 1.07 (27H, s, tert-Buꢄ3), 3.64 (12H, br s, CH2ꢄ6),
6.06 (3H, s, ArHꢄ3), 6.95 (6H, br s, ArHꢄ6), 8.86 (6H, br s, OHꢄ6), 9.08
(6H, br s, OHꢄ6). 13C-NMR (DMSO-d6ꢀD2O, at 50 °C) d: 23.6 (C-2, 8, 14,
20, 26, 32), 31.3 (tert-butyl methyl C), 33.3 (tert-butyl quaternary C), 95.2
(C-11, 23, 35), 115.4 (C-1, 9, 13, 21, 25, 33), 124.5 (C-4, 6, 16, 18, 28, 30),
126.6 (C-3, 7, 15, 19, 27, 31), 141.1 (C-5, 17, 29), 148.5 (C-37, 39, 41),
153.1 (C-38, 40, 42). FAB-MS m/z: 901 (MꢀH)ꢀ. Anal. Calcd for
C54H60O121.2B(OH)3: C, 66.50; H, 6.57. Found: C, 66.62; H, 6.83.
2,6-Bis[4-tert-butyl-2-hydroxy-3-(2,4,6-trimethoxybenzyl)]benzyl-
1,3,5-trimethoxybenzene (7) A solution of 4 (420 mg, 2 mmol) in dry
toluene (20 ml) was added dropwise for 1 h to a stirred solution of 5 (2.04 g,
4 mmol), 3 (672 mg, 4 mmol), and p-TsOH·H2O (38 mg, 0.2 mmol) in dry
toluene (10 ml) at room temperature under argon atmosphere, and then
stirred at 60 °C for 4 h. After cooling, to the reaction mixture 5% aqueous
NaHCO3 solution was added until the reaction’s pH to be 5. The resulting
mixture was extracted with EtOAc twice. The organic layer was washed with
water and brine, and dried with anhydrous Na2SO4, and then removed under
vacuum. The residue was separated by silica-gel column chromatography (n-
hexane–EtOAcꢁ3 : 1) to give 7 (1063 mg, 31.2%) as a white powder.
Data for 7: Colorless powder. mp 182 °C. IR (KBr) cmꢃ1: 3435, 2956,
2835, 1601, 1483, 1464, 1205, 1187, 1122. 1H-NMR (CDCl3) d: 1.14 (18H,
s, tert-Buꢄ2), 3.51 (3H, s, OMeꢄ1), 3.77 (6H, s, OMeꢄ2), 3.80 (6H, s,
OMeꢄ2), 3.84 (12H, s, OMeꢄ4), 3.86 (4H, s, CH2ꢄ2), 3.91 (4H, s, CH2ꢄ
2), 6.16 (4H, s, ArHꢄ4), 6.35 (1H, s, ArHꢄ1), 6.81 (2H, d, Jꢁ2.4 Hz,
ArHꢄ2), 7.02 (2H, d, Jꢁ2.4 Hz, ArHꢄ2), 7.56 (2H, s, OHꢄ2). FAB-MS
m/z: 855 (MꢀH)ꢀ. Anal. Calcd for C51H64O11: C, 71.80; H, 7.56. Found: C,
71.52; H, 7.46.
11,23,35-Tri-tert-butyl-37,39,41-trihydroxy-4,6,16,18,28,30,38,40,42-
nonamethoxycalix[6]arene (8). “5ꢀ1” Approach To a refluxing solu-
tion of Yb(OTf)3 (55.8 mg, 0.09 mmol) in dry benzene (450 ml), a solution
of 7 (512 mg, 0.6 mmol) and bis-hydroxymethyl-4-tert-butylphenol (4, 21
mgꢄ6, 0.6 mmol) in dry benzene (50 mlꢄ6) was added dropwise for 6 h.
The resulting water was removed with a Dean–Stark trap. After additional
refluxing for 1 h, the reaction mixture was cooled and then benzene was re-
moved under vacuum until 100 ml. The residual benzene solution was
washed with a 5% NaHCO3 aqueous solution and water, and then dried over
anhydrous Na2SO4. After removing the benzene under vacuum, the residue
was recrystallized from EtOAc to give 8 (120 mg, 19.5%) as colorless pow-
References
1) Thondorf I., Shivanyuk A., Böhmer V., “Calixarenens 2001,” ed. by
Asfari Z., Böhmer V., Harrowfield J., Vicens J., Kluwer Academic
Publishers, Dordrecht, 2001, pp. 26—53.
2) Böhmer V., Angew. Chem. Int. Ed. Engl., 34, 713—745 (1995).
3) Gutsche C. D., “Calixarenens 2001,” ed. by Asfari Z., Böhmer V., Har-
rowfield J., Vicens J., Kluwer Academic Publishers, Dordrecht, 2001,
pp. 1—25.
4) Happel G., Mathiasch B., Kämmerer H., Makromol. Chem., 176,
3317—3334 (1975).
5) Böhmer V., Chhim P., Kämmerer H., Makromol. Chem., 180, 2503—
2506 (1979).
6) Sato S., Iijima H., Haga N., Osono K., Mizuguchi H., Kijima T., Ono-
dera J.-I., Supramolecular Chemistry, 18, 39—46 (2006).
7) Fukazawa Y., Deyama K., Usui S., Tetrahedron Lett., 33, 5803—5806
(1992).
8) Böhmer V., Jung K., Schön M., Wolff A., J. Org. Chem., 57, 790—792
(1992).