Debenzylation of tertiary amines using phosgene or triphosgene: An efficient and rapid procedure for the preparation of carbamoyl chlorides and unsymmetrical ureas. Application in carbon-11 chemistry

Article abstract of DOI:10.1021/jo0346297

Efficient and rapid preparations of carbamoyl chlorides and unsymmetrical ureas from tertiary amines and phosgene or its safe equivalent triphosgene [bis(trichloromethyl)carbonate, BTC] are described. First, the reaction of stoichiometric amounts of phosgene with secondary amines was revisited, and it was shown that the formation of carbamoyl chlorides in high yields required careful adjustments of experimental conditions and the use of pyridine as an HCl scavenger. A phosgene-mediated dealkylation of triethylamine was observed when this base was used instead of pyridine. Taking advantage of this observation, a strategy of synthesis of carbamoyl chlorides from tertiary amines and phosgene has been developed. N-Alkyl-N-benzyl(substituted)tetrahydroisoquinolines, -piperazines, -piperidines, or -anilines were treated with stoichiometric amounts of phosgene (or BTC) in CH₂Cl₂. Tertiary amines bearing electron-enriched benzyl group(s) afforded carbamoyl chlorides in excellent yields and without any contamination by symmetrical ureas. Subsequent additions of primary or secondary amines to these carbamoyl chlorides produced unsymmetrical ureas in single-pot and high-yielding operations. This methodology was applied in ¹¹C-chemistry. From [ ¹¹C]phosgene, a common precursor used in the preparation of radiotracers for positron emission tomography, a rapid and efficient synthesis of ¹¹C-carbamoyl chlorides and ¹¹C-unsymmetrical ureas derived from tetrahydroisoquinoline and piperazine is described. The first example of ¹¹C-amide formation from the reaction of a ¹¹C-carbamoyl chloride and an organometallic (cyanocuprate or a Grignard reagent in the presence of a nickel catalyst) is also presented.

Full text of DOI:10.1021/jo0346297

Deben zyla tion of Ter tia r y Am in es Usin g P h osgen e or
Tr ip h osgen e: An Efficien t a n d Ra p id P r oced u r e for th e
P r ep a r a tion of Ca r ba m oyl Ch lor id es a n d Un sym m etr ica l Ur ea s.
Ap p lica tion in Ca r bon -11 Ch em istr y
Laurent Lemoucheux,† J acques Rouden,*,† Me´ziane Ibazizene,‡ Franck Sobrio,‡ and
Marie-Claire Lasne*,†
Laboratoire de Chimie Mole´culaire et Thioorganique, CNRS UMR 6507, ENSICAEN,
Universite´ de Caen-Basse Normandie, 6 Boulevard du Mare´chal J uin, 14050 Caen Cedex, France, and
Groupe de De´veloppements Me´thodologiques en Tomographie par EÄ mission de Positons, EA 2609,
UMR CEA, Universite´ de Caen-Basse Normandie, Centre Cyceron, 15 Boulevard Henri Becquerel,
14070 Caen Cedex, France
lasne@ismra.fr
Received May 12, 2003
Efficient and rapid preparations of carbamoyl chlorides and unsymmetrical ureas from tertiary
amines and phosgene or its safe equivalent triphosgene [bis(trichloromethyl)carbonate, BTC] are
described. First, the reaction of stoichiometric amounts of phosgene with secondary amines was
revisited, and it was shown that the formation of carbamoyl chlorides in high yields required careful
adjustments of experimental conditions and the use of pyridine as an HCl scavenger. A phosgene-
mediated dealkylation of triethylamine was observed when this base was used instead of pyridine.
Taking advantage of this observation, a strategy of synthesis of carbamoyl chlorides from tertiary
amines and phosgene has been developed. N-Alkyl-N-benzyl(substituted)tetrahydroisoquinolines,
-piperazines, -piperidines, or -anilines were treated with stoichiometric amounts of phosgene (or
BTC) in CH2Cl2. Tertiary amines bearing electron-enriched benzyl group(s) afforded carbamoyl
chlorides in excellent yields and without any contamination by symmetrical ureas. Subsequent
additions of primary or secondary amines to these carbamoyl chlorides produced unsymmetrical
ureas in single-pot and high-yielding operations. This methodology was applied in 11C-chemistry.
From [11C]phosgene, a common precursor used in the preparation of radiotracers for positron
emission tomography, a rapid and efficient synthesis of 11C-carbamoyl chlorides and 11C-
unsymmetrical ureas derived from tetrahydroisoquinoline and piperazine is described. The first
example of 11C-amide formation from the reaction of a 11C-carbamoyl chloride and an organometallic
(cyanocuprate or a Grignard reagent in the presence of a nickel catalyst) is also presented.
In tr od u ction
11C in positron emission tomography (PET), are valuable
radionuclides since they permit the synthesis of radio-
labeled versions of the compound of interest. The syn-
theses with these radioisotopes share several common
features: limited number of labeled precursors, sub-
micromolar amounts of these starting materials, and a
need for the introduction of the radioisotope as late as
possible in the synthesis. All these reasons have re-
stricted complex radiosyntheses. Whereas carbon-14 ra-
diotracers or reagents2 can be stored indefinitely with no
loss of radioactivity (14C-half-life: 5700 years), the short
half-life of carbon-11 (20 min) requires the rapid prepara-
tion and purification of carbon-11 labeled molecules.
Those have to be carried out immediately before use and
from cyclotron produced precursors ([11C]CO2, [11C]CO,
or [11C]CH4) or rapidly prepared from them ([11C]CH3I,
[11C]COCl2, [11C]HCN).3 Because of the increased molec-
ular complexity and diversity of biologically active com-
pounds, there is still a need for new methodologies which
Radiolabeled tracers play a critical role throughout the
drug discovery process and medical imaging. They are
used in pharmaceutical research in screening against
new targets for labeling novel enzymes or receptors and
for plasma protein binding experiments. During the lead
optimization stage, radiolabeled drug candidates are
useful for the identification of circulating metabolites,
determination of drug absorption, distribution, metabo-
lism, and excretion. In the field of nuclear medicine,
positron emission tomography (PET) is a noninvasive
technique and a method of choice to provide anatomical
distribution and localization of radiolabeled drugs in
living human organs, especially the brain.1 The radio-
isotopes of carbon, 14C in pharmaceutical research and
* To whom correspondence should be addressed. Tel: 33 2 31 45 28
92. Fax: 33 2 31 45 28 77.
† Laboratoire de Chimie Mole´culaire et Thioorganique.
‡ Groupe de De´veloppements Me´thodologiques en Tomographie par
EÄ mission de Positons.
(1) Fowler J . S., Wolf A. P. Acc. Chem. Res. 1997, 30, 181-188.
(2) McCarthy, K. E. Curr. Pharm. Design 2000, 6, 1057-1083.
10.1021/jo0346297 CCC: $25.00 © 2003 American Chemical Society
Published on Web 08/27/2003
J . Org. Chem. 2003, 68, 7289-7297
7289

Lemoucheux et al.
SCHEME 1a
give access in short time and high yield to radioactive
11C- or 14C-probes.
Pharmaceutically active molecules often contain ureas,
carbamates, or related functions. In stable isotope chem-
istry, carbamoyl chlorides are common intermediates for
the introduction of these functional groups.4 Moreover,
they can also be transformed into amides by reaction with
aryllithiums.5 Recently, we have developed their nickel-
catalyzed cross-coupling reactions with Grignard re-
agents as an alternative route to amides.6 Carbamoyl
chlorides are often synthesized by reaction of a secondary
amine with an excess of phosgene.7 These conditions are
detrimental to the environment, and they cannot be used
under radioactive conditions. Moreover, they often lead
to byproducts. The pallado-catalyzed carbonylation of
chloramines8 or selenium-mediated carbonylation of
amines or alcohols9 under carbon monoxide pressure
could be envisaged. However, this possibility was not
considered since it required either a pressure of carbon
monoxide or, in carbon-11 chemistry, a special technol-
ogy.10 Formamides are also direct precursors of carbamoyl
chlorides by reaction with a chlorinating reagent,11 but
their 11C-labeled counterparts have not yet been de-
scribed. Methods based on the use of carbon dioxide and
again a chlorinating agent have been employed both in
14C-carbon12 and 11C-carbon13 chemistries. However, yields
and selectivities remained moderate. [11C]Phosgene, eas-
ily prepared from cyclotron-produced [11C]methane,14
remains the precursor of choice for rapid and efficient
radiosyntheses of symmetrical15 or cyclic 11C-ureas16 and
a
Reagents and conditions: (a) excess COCl2, rt, toluene; (b)
[
11C]COCl2, -78 °C, THF; (c) BTC or COCl2 and NEt3, dichlo-
romethane.
11C-carbamates via the formation of the corresponding
11C-isocyanates under high dilution of the starting amine.17
[11C]Phosgene was used to prepare 11C-labeled ethylchlo-
roformate in low yields,18 [11C]methanol being preferred
for the 11C-labeling of methylchloroformate.19 To our
knowledge, only one 11C-carbamoyl chloride has been
isolated and characterized due to the low reactivity of
the used amine.20 Indeed, if carbamoyl chlorides are the
products formed by reaction of secondary amines and
phosgene under normal conditions,21 11C-ureas are the
only products isolated in 11C-chemistry. Our preliminary
labeling experiments confirmed the difficulty in prepar-
ing 11C-labeled carbamoyl chlorides. The reaction of
tetrahydroisoquinoline 1 or N-benzylpiperazine 4 (1.5
mmol or 15 µmol) in THF or toluene (400 µL) with [11C]-
phosgene was performed at -10 or -78 °C for 5 min.
Radio-TLC or HPLC analyses of the crude reaction
mixture showed the exclusive formation of the corre-
sponding 11C-ureas 3 or 6 (Scheme 1).
(3) For general references on practical PET chemistry, see: (a)
Fowler, J . S., Wolf, A. P. In Positron Emission Tomography and
Autoradiography: Principles and Applications for the Brain and Heart;
Phelps, M., Mazziota, J ., Schelbert, H., Eds.; Raven Press: New York,
1986; Chapter 9. (b) Radiopharmaceuticals for Positron Emission
Tomography: Methodological Aspects; Sto¨cklin, G., Pike, V. W., Eds.;
Kluwer Academic Publishers: Dordrecht, The Netherlands, 1993. (c)
Langstro¨m, B.; Kihlberg, T.; Bergstro¨m, M.; Antoni, G.; Bjo¨rkman, M.;
Forngren, B. H.; Forngren, T.; Hartvig, P.; Markides, K.; Yngve, U.;
O¨ gren, M. Acta Chem. Scand. 1999, 53, 651-669.
(4) Hegarty, A. F. In Comprehensive Organic Chemistry; Sutherland,
I. O., Ed.; Pergamon Press: Oxford, 1979; Vol. 2, Chapter 9.10.
(5) Mills, R. J .; Taylor, N. J .; Snieckus, V. J . Org. Chem. 1989, 54,
4372-4385.
(6) Lemoucheux, L.; Rouden, J .; Lasne, M.-C. Tetrahedron Lett.
2000, 41, 9997-1001.
(7) (a) Holmes, D. L.; Smith, E. M.; Nowick, J . S. J . Am. Chem. Soc.
1997, 119, 7665-7669. (b). Wang, G. T.; Chen, Y.; Wang, S.; Sciotti,
R.; Sowin, T. Tetrahedron Lett. 1997, 38, 1895-1898. (c) Nowick, J .
S.; Holmes, D. L.; Noronha, G.; Smith, E. M.; Nguyen, T. M.; Huang,
S.-L. J . Org. Chem. 1996, 61, 3929-3934.
(8) Saegusa, T.; Tsuda, T.; Isegawa, Y. J . Org. Chem. 1971, 36, 858-
860.
(9) Kihlberg, T.; Karimi, F.; Langstro¨m, B. J . Org. Chem. 2002, 67,
3687-3692.
(10) Karimi, F.; Langstro¨m, B. Org. Biomol. Chem. 2003, 1, 541-
(16) (a) Crouzel, C.; Mestelan, G.; Kraus, E.; Lecomte, J . M.; Comar,
D. Appl. Radiat. Isot. 1980, 31, 545-548. (b) Roeda, D.; Tavitian, B.;
Coulon, C.; David, F.; Dolle´, F.; Fuseau, C.; J obert, A.; Crouzel, C.
Bioorg. Med. Chem. 1997, 5, 397-403. (c) Hammadi, A.; Crouzel, C.
J . Labelled Compd. Radiopharm. 1991, 29, 681-689.
(17) (a) Dolle´, F.; Valette, H.; Hinnen, F.; Vaufrey, F.; Demphel, S.;
Coulon, C.; Ottaviani, M.; Bottlaender, M.; Crouzel, C. J . Labelled
Compd. Radiopharm. 2001, 44, 785-795. (b) Crouzel, C.; Hinnen, F.;
Maitre, E. Appl. Radiat. Isot. 1995, 46, 167-170. (c) Conway, T.; Diksic,
M. J . Nucl. Med. 1988, 29, 1957-1960.
(18) Rousseau, A. H.; Diksic, M.; Van Lier, J . E. Int. J . Appl.
Instrum. B 1992, 19, 175.
(19) Ravert, H. T.; Mathews, W. B.; Musachoio, J . L.; Dannals, R.
F. J . Labelled Compd. Radiopharm. 1995, 36, 365-371.
(20) Lasne, M.-C.; Barre´, L.; Huard, C.; Ducandas, C.; McKenzie,
E. T. J . Labelled Compd. Radiopharm. 1993, 38, 425-426.
(21) Gymer, G. E.; Narayanaswani, S. N. In Comprehensive Organic
Functional Group Transformations; Katritzky, A. R., Meth-Cohn, O.,
Rees, C. W., Eds.; Pergamon Press: Oxford, 1995; Vol. 6, p 443.
546 and references therein.
(11) (a) Schindler, N.; Plo¨ger, W. Chem Ber. 1971, 104, 969-971.
(b) Hasserodt, U. Chem Ber 1968, 101, 113-120.
(12) (a) Volford, J .; Knausz, D.; Meszticzky, A.; Horvath, L.; Csak-
vari, B. J . Labelled Compd. Radiopharm. 1981, 18, 555. (b) Hoshino,
O.; Saito, K.; Ishizaki, M.; Umezawa, B. Synth. Commun. 1987, 17,
1887-1892. (c) Birkofer, L.; Krebs, K. Tetrahedron Lett. 1968, 885-
888. (d) Dean, D. C.; Wallace, M. A.; Marks, T. N.; Melillo, D. G.
Tetrahedron Lett. 1997, 38, 919-922.
(13) Schirbel, A.; Holschbach, M. H.; Coenen, H. H. J . Labelled
Compd. Radiopharm. 1999, 42, 537-551.
(14) (a) Landais, P.; Crouzel, C. Appl. Radiat. Isot. 1987, 38, 297-
300. (b) Link, J . M.; Krohn, K. A J . Labelled Compd. Radiopharm.
1997, 40, 306-308.
(15) (a) Lidstro¨m, P.; Bonasera, T. A.; Marquez-M, M.; Nilsson, S.;
Bergstro¨m, M.; Langstro¨m, B. Steroids 1998, 63, 228-234. (b) Lid-
stro¨m, P.; Bonasera, T. A.; Nilsson, S.; Langstro¨m, B. J . Labelled
Compd. Radiopharm. 1997, 40, 788-789.
7290 J . Org. Chem., Vol. 68, No. 19, 2003

Debenzylation of Tertiary Amines Using Phosgene or Triphosgene
TABLE 1. Rea ction of BTC w ith Secon d a r y Am in es: In flu en ce of th e Ad d ed Ba se
entry
aminea
base
equiv
solvent
T (°C)
timeb (h)
products
relative ratio (%)c
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
4
NEt3
NEt3
NEt3
NEt3
Na2CO3
Na2CO3
Na2CO3
C5H5N
C5H5N
C5H5N
C5H5N
C5H5N
2.2
2.2
2.2
2.2
1
THF
THF
-78 to 20
0.15 then 16
0.10 then 1.25
0.25 then 1
1 then 3
30 then 0.75
0.15 then 2
0.3 then 2
0.5 then 3
0.16 then 3
2 then 2
2/3/8/7
2/3/8/7
2/3/8/7
2/3/8/7
2/3
2/3
2/3
2/3
2/3
46/0/54/0
93/7/traces/0
9/46/38/7
4/44/38/14
84/16
78/22
84 (68)/16
100/0
100 (88)/0
82 (60)/18
100 (48)/0
(81)e
20
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
THF
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
20
20
0
-78
20
20
-78 to 20
20
0
1
1
2.2
2.2
1.2
1.2
2.5
10d
11
12
2/3
5
0.15 then 1
0.15 then 1
20
13
C5H5N
2.5
CH2Cl2
-50
0.15 then 1
(60)
a
All of the reactions were carried out with BTC (0.37 equiv) except entry 1 (reaction with phosgene): To BTC in CH2Cl2 at a given
temperature were added the base then the secondary amine over a given period of time. After reaction, the mixture was hydrolyzed with
b
1 N HCl and extracted with CH2Cl2. Time of addition followed by the reaction time. c Percentage of the reaction products determined
d
by 1H NMR. The isolated yields are given in parentheses. BTC was added to the amine. e Crude yield, carbamoyl chloride 10 decomposed
on silica gel.
As a part of our program on the synthesis of cholinergic
radioligands20,22 and the search for new 11C-labeled pre-
cursors or new methodologies usable in radiolabeling23
we have developed a rapid, efficient, and selective method
to prepare carbamoyl chlorides from tertiary amines
using triphosgene in stable isotope chemistry and [11C]-
phosgene under radioactive conditions. Applications to
the synthesis of unsymmetrical ureas, carbamates, and
amides in both types of chemistry are described as well.
romethyl)carbonate, BTC], a solid and safe surrogate of
phosgene, appeared as the reagent of choice for preparing
carbamoyl chlorides.24
Amine 1 was slowly added to a solution of 0.35-0.40
equiv of BTC (equal to a slight excess of phosgene) in
THF or CH2Cl2 in the presence of a base (triethylamine,
sodium carbonate, or pyridine). Table 1 presents the
1
results as the H NMR ratio of the four products eventu-
ally detected in the crude mixture: carbamoyl chloride
2 and three side compounds, ureas 3 and 7 and dialkyl-
carbamoyl chloride 8. Urea 3 was the major expected
byproduct in both experiments conducted with triethyl-
amine in CH2Cl2 (Table 1, entries 3 and 4). Using this
organic base, carbamoyl chlorides 8 and unsymmetrical
ureas 7 coming from the dealkylation of the tertiary base
were formed (Table 1, entries 1, 3, and 4). The specific
reaction between phosgene and triethylamine, already
described,25 has been rarely mentioned as a side reaction
by authors using the same conditions as ours. Hetero-
geneous conditions carried out with sodium carbonate
generated carbamoyl chloride 2 in 68% isolated yield,
without avoiding the formation of urea 3 (Table 1, entries
5-7). Pyridine, which was the base used originally with
BTC,26 gave the best result. Comparison of entries 1, 2
and 8, 9 (Table 1) shows that both triethylamine and the
secondary amine 1 gave the corresponding carbamoyl
chlorides (8 and 2, respectively) when the reagents were
mixed at -78 °C. In the presence of triethylamine, the
distribution of the products was strongly dependent on
the temperature. At 20 °C, no carbamoyl chloride 8 was
detected whereas tetrahydroisoquinoline urea 7 was
formed. Such an influence of the temperature was not
Resu lts a n d Discu ssion
Since conventional spectroscopic analysis of 11C-labeled
molecules is not possible, a new radiotracer is usually
characterized by comparison of its radio-TLC or HPLC
properties with those of the stable isotope reference
compounds.3 Therefore, an efficient synthesis of unla-
beled carbamoyl chlorides 2 and 5 as model compounds
was developed. Tetrahydroisoquinoline 1 was first chosen
as a substrate because it offers several advantages: its
structure is frequently encountered in biologically active
compounds, it has a high boiling point, and the presence
of an aromatic ring allows for reaction monitoring and
analysis by TLC and HPLC. Finally, the corresponding
carbamoyl chloride 2 is reasonably stable in aqueous
conditions and can be purified by chromatography on
silica gel without any loss. It has been prepared by
reaction of amine 1 with an excess of phosgene in toluene
solution in the presence of an excess of triethylamine.
Using these conditions, the carbamoyl chloride 2 was
contaminated with another product which was later
identified as diethylcarbamoyl chloride 8 (Scheme 1,
Table 1, entry 1). Because some carbamoyl chlorides are
unstable and are best used without any purification, we
searched for a cleaner and more efficient reaction. After
a quick survey of the literature, triphosgene [bis(trichlo-
(24) (a) Cotarca, L.; Delogu, P.; Nardelli, A.; Sunjic, V. Synthesis
1996, 553-576. (b) Bigi, F., Maggi, R., Sartori, G. Green Chem. 2000,
2, 140-148.
(25) (a) Itaya, T.; Kanai, T. Heterocycles 1997, 46, 101-104. (b)
Strepikheev, Y. A.; Perlova, T. G.; Zhivechkova, L. A. J . Org. Chem.
USSR 1968, 41, 1826-1829.
(22) Marrie`re, E.; Rouden, J .; Tadino, V.; Lasne, M.-C. Org. Lett.
2000, 2, 1121-1124.
(23) Aubert, C.; Huard-Perrio, C.; Lasne, M.-C. J . Chem. Soc., Perkin
Trans. 1 1997, 2837-2842 and references therein.
(26) Eckert, H.; Forster, B. Angew. Chem., Int. Ed. Engl. 1987, 26,
894-895.
J . Org. Chem, Vol. 68, No. 19, 2003 7291

Lemoucheux et al.
SCHEME 2
observed when the reaction was performed in the pres-
ence of pyridine in dichloromethane.
The reaction can be performed cleanly by slowly adding
tetrahydroisoquinoline 1 to a mixture of BTC and pyri-
dine in CH2Cl2, for example, to isolate carbamoyl chloride
2 in 88% yield on a 17 g scale experiment (entry 9). The
addition order of the reagents is critical for the sole
formation of the carbamoyl chloride as shown by entry
10. Under the best conditions (entry 11), N-benzylpiper-
azine 4 led quantitatively to carbamoyl chloride 5 which
was isolated in 48% yield after column chromatography.
This low yield is due to the partial instability of the
carbamoyl chloride 5 on silica gel. Two more carbamoyl
chlorides 10 and 1227 were synthesized according to the
same conditions from amines 9, 11, and BTC.
The whole results presented in Table 1 show that the
synthesis of carbamoyl chlorides from secondary amines
and BTC in the presence of a base, often considered as a
straightforward reaction, requires careful adjustments
in order to avoid contamination by the corresponding
ureas. The formation of diethylcarbamoyl chloride 8,
when the reaction was carried out in the presence of
triethylamine, demonstrates that a tertiary amine can
be dealkylated by phosgene to generate a carbamoyl
chloride. This latter was unreactive toward the tertiary
amine as proved by entries 3 and 4 where diethylcar-
bamoyl chloride 8 did not react with the excess of
triethylamine to give tetraethylurea. This observation is
crucial when considering the experimental conditions in
radioactive chemistry where the 11C-labeled carbamoyl
chloride will be generated in the presence of a large
excess of the starting amine. We then focused our efforts
on the possible generation of carbamoyl chlorides from
phosgene (or BTC) and tertiary amines.
rapidly with the nitrogen lone pair of the tertiary amine
to afford an acylammonium chloride. Then decomposition
of this ionic intermediate to the carbamoyl chloride and
alkyl chloride occurs either by an SN1 process when allyl,
benzyl, or tertiary groups are present or, more generally,
by an SN2 process.35
We undertook our study using the same amine, tet-
rahydroisoquinoline 1, which was converted into several
tertiary amines, the methyl, benzyl, 4-methoxybenzyl,
2,4-dimethoxybenzyl, 4-dimethylaminobenzyl derivatives
13a -e, to test the dealkylation using phosgene or BTC
(Scheme 2), its selectivity, and to choose the substrate
the most favorable in terms of time and efficiency. The
synthesis of these starting materials was carried out
according to standard procedures.36 Table 2 summarizes
the different conditions of stoichiometry, solvent, tem-
perature, and time. The crude mixtures were analyzed
1
by H NMR, and the proportions of the chemical species
which were detected and characterized are reported. It
should be noted that none of the starting amines were
present in the final mixtures, even when the stoichiom-
etry phosgene/amine was 1/1.
N-Methyl- and N-benzylamines 13a and 13b reacted
immediately with phosgene (or its equivalent BTC) to
form the corresponding quaternary ammonium salts 14a
or 14b, insoluble (or slightly soluble) in toluene, ether,
and THF and soluble in dichloromethane. From the
reactions with N-methylamine 13a , dichloromethane was
the best solvent but the proportion of the carbamoyl
chloride 2 reached only 54% of the mixture. Increasing
reaction time and/or temperature and adding a chloride
salt did not improve significantly the conversion. A
competitive attack of the chloride anion on the benzylic
position of the ring system was not observed as it was in
the case of phthalideisoquinoline alkaloids treated with
thiophosgene.37 However, our results compare well with
the phosgene-mediated rearrangement of a benzoquino-
lizidine where the dealkylation was governed by stereo-
electronic factors.33
P h osgen e-Med ia t ed Dea lk yla t ion of Ter t ia r y
Am in es. Dealkylations of tertiary amines are usually
performed under the von Braun conditions28 or using acid
chlorides or alkyl chloroformates.29 Vinyl chloroformate
and chloroethyl chloroformate have proved to be more
selective and produce cleaner reaction products.30 Beside
dealkylation of triethylamine,25 we found only a few
reports on phosgene-induced dealkylation,31 often de-
methylation,32 of particular tertiary amines with the aim
of generating a secondary amine, or to cleave a cyclic
tertiary amine.33 The possibility of using triphosgene for
debenzylation of tertiary amines has been reported
recently.34
From a mechanistic point of view, the reaction can be
viewed as a two-step reaction. First, phosgene reacts
(27) Freer, R.; McKillop, A. Synth. Commun. 1996, 2; 331-349
(28) (a) Von Braun, J . Chem. Ber. 1900, 33, 1438-1452. (b) Hage-
man, H. A. Org. React. 1953, 7, 198-262.
When N-benzylamine 13b in CH2Cl2 was treated with
phosgene (Table 2, entry 10), the ammonium salt 14b
(29) For a review on amine dealkylations with acyl chlorides, see:
Cooley, J . H.; Evain, E. J . Synthesis 1989, 1-7.
(33) (a) Maryanoff, B. E.; Molinari, A. J .; Wooden, G. P.; Olofson,
R. A. Tetrahedron Lett. 1982, 23, 2829-2832. (b) Maryanoff, B. E.;
Molinari, A. J .; McComsey, D. F.; Maryanoff, C. A.; Wooden, G. P.;
Olofson, R. A. J . Org. Chem. 1983, 48, 5074-5080.
(34) (a) J orand-Lebrun, C.; Valognes, D.; Halazy, S. Synth. Commun.
1998, 28, 1189-1195. (b) During the preparation of our manuscript,
another group published a similar study: Banwell, M. G.; Coster, M.
J .; Harvey, M. J .; Moraes, J . J . Org. Chem. 2003, 68, 613-616.
(35) Howarth, N. M.; Malpass, J . R.; Smith, C. R. Tetrahedron 1998,
54, 10899-10914.
(36) See the Supporting Information for complete experimental
procedures and characterization data for all new compounds.
(37) Baradarani, M. M.; Prager, R. H. Tetrahedron Lett., 1999, 40,
7403-7406.
(30) (a) Olofson, R. A.; Schnur, R. C.; Bunes, L.; Pepe, J . P.
Tetrahedron Lett. 1977, 1567-1570. (b) Olofson, R. A.; Martz, J . T.;
Senet, J .-P.; Piteau, M.; Malfroot, T. J . Org. Chem. 1984, 49, 2081-
2082.
(31) (a) Holtschmidt, H. Angew. Chem. 1962, 74, 848-855. (b)
Bo¨hme, H.; Koch, L.; Ko¨hler, E. Chem. Ber. 1962, 95, 1849-1858.
(32) (a) Sauter, F.; Stanetty, P.; Hetzl, E.; Furhmann, F. J . Hetero-
cycl. Chem. 1983, 20, 1477-1480. (b) Wolf-Pflugmann, M.; Lambrecht,
G.; Wess, J .; Mutschler, E. Arzneim.-Forsch./ Drug Res. 1989, 39, 539-
544; Chem. Abstr. 1989, 111, 49920. (c) Banholzer, R.; Heusner, A.;
Schulz, W. Liebigs Ann. Chem. 1975, 2227-2231. (d) Von Banholzer,
R.; Pook, K.-H. Arzneim.-Forsch./ Drug Res. 1985, 35, 217-228; Chem.
Abstr. 1985, 103, 196293.
7292 J . Org. Chem., Vol. 68, No. 19, 2003

Debenzylation of Tertiary Amines Using Phosgene or Triphosgene
TABLE 2. Dea lk yla tion of Ter tia r y Am in es 13 Usin g P h osgen e or BTC
entrya
amine
reagent
ratio amine/reagentb
solvent
T (°C)
time (h)
14c (%)
2c (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
13a
BTC
BTC
BTC
BTC
BTC
BTC
BTC
BTC
COCl2
COCl2
COCl2
COCl2
COCl2
COCl2
COCl2
BTC
1/0.4
1/0.4
1/0.4
1/0.4
1/0.4
1/0.4
1/0.4
1/0.4
1/1
1/1
1/1
5/1
1/1
Et2O
Et2O
THF
THF
20
20
20
68
68
20
20
20
20
20
20
20
20
20
20
20
20
20
1
18
1
24
24
1
21
360
18
1
1
1
0.16
1
1
97
96
93
80
81d
57
46
48
99
47
75
0
1
0
0
0
0
0
3
4
7
20
19
43
54
52
1
THF
CH2Cl2
CH2Cl2
CH2Cl2
Toluene
CH2Cl2
(ClCH2)2
THF
13b
13c
53
25
100
THF
99
1/1
1/1
1/0.4
1/0.4
1/0.4
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
100 (65)
100 (65)
100 (60)
100 (85)
100
0.5
0.5
0.5
13d
13e
BTC
BTC
a
To a mixture of tertiary amine in CH2Cl2 was added BTC in dichloromethane. The mixture was stirred for a given period of time and
b
at a chosen temperature. The crude product was extracted with CH2Cl2. For practical reasons, phosgene was added to the amine. Molar
ratio amine/BTC. c Percentages of compounds 14 and 2 measured in the crude 1H NMR mixture. The isolated yields are given in parentheses.
d
Addition of LiCl (30 equiv).
and the expected product 2 were formed in proportions
similar to those observed for N-methylamine 13a . Using
a 5-fold excess of starting amine over phosgene (entry
12), the conversion into the carbamoyl chloride 2 was
quantitative in less than 1 h. No trace of the salt 14b
was detected by 1H NMR. As anticipated, the excess
starting tertiary amine did not react with the carbamoyl
chloride to produce the symmetrical urea. Appending
electron-donating groups on the benzyl group facilitated
the cleavage.38 The 4-methoxy-, 2,4-dimethoxy-, and
4-dimethylaminobenzyl derivatives 13c, 13d , and 13e
were all converted directly to the carbamoyl chloride 2
using a stoichiometric amount of phosgene, and it was
not possible to isolate the phosgene adduct 14. In one
case, after a short reaction time (10 min, entry 13) traces
of the salt 14c could be visualized on the 1H NMR
spectrum. This in good agreement with an SN1-type of
mechanism where electron-rich aromatics stabilize a
benzylic carbocation and therefore increase the rate of
the overall process. The reaction was fast and reproduc-
ible leading to the carbamoyl chloride 2 in isolated yields
of ranging from 60 to 85%, similar to those obtained using
tetrahydroisoquinoline 1, BTC, and pyridine.
SCHEME 3
SCHEME 4
The selectivity of the reaction was tested with the
piperazine derivative 15 treated with BTC in CH2Cl2.
After 1 h at rt, only the cleavage of the 4-methoxybenzyl
group was observed by 1H NMR along with the formation
of 4-methoxybenzyl chloride 16 (Scheme 3). The resulting
carbamoyl chloride 5 was isolated in 53% yield after
purification on silica gel, with a yield similar to that
obtained previously from piperazine 4 (see Table 1, entry
11). The limitation of the tertiary amine dealkylation was
evaluated by using a less nucleophilic amine. The aniline
17 reacted completely with BTC to afford after 30 min a
64/36 mixture of the carbamoyl chloride 12 and the salt
18. The former was isolated with a 44% yield. Optimiza-
tion of the conditions would require in this particular case
the presence of a dimethoxy-, trimethoxy-, or dimethyl-
aminobenzyl group for an efficient transformation into
carbamoyl chloride 12 by reaction with phosgene. In view
of these results it became clear that electron-enriched
benzyl groups are well suited as cleavable groups from
tertiary amines by triphosgene. By comparison, the
benzyl moiety is not always easily removed from a
tertiary amine (Table 2, entry 10) unless an excess of
amine versus phosgene was used (Table 2, entry 12).34b
Triphosgene has been used previously for the prepara-
tion of unsymmetrical ureas either from primary39 or
(38) Miller, M. W.; Vice, S. F.; McCombie, S. W. Tetrahedron Lett.
1998, 39, 3429-3432.
J . Org. Chem, Vol. 68, No. 19, 2003 7293

Lemoucheux et al.
TABLE 3. Com p a r a ison of Meth od s A a n d B To Syn th esize Ur ea s
a
b
Isolated yield. Method A: To the carbamoyl chloride in CH2Cl2 was added the tertiary amine and NEt3 at rt.The mixture was
stirred for 12 h and then treated with a saturated aqueous NaHCO3 solution and extracted with dichloromethane. The crude product was
purified by flash chromatography. c Method B: To the tertiary amine in CH2Cl2 was added, at -10 °C, BTC in one portion. The reaction
mixture was stirred for 1 h at rt. The secondary amine then NEt3 were added and the mixture was stirred for 15 h before workup as in
method A.
secondary amines.40 Our method was applied to the one-
pot synthesis of unsymmetrical ureas from tertiary
amines and compared with the reaction of carbamoyl
chlorides with secondary amines (Scheme 4). Table 3
described the results obtained from both procedures:
directly by reaction between a carbamoyl chloride and
(39) Majer, P.; Randad, R. S. J . Org. Chem. 1994, 59, 1937-1938.
7294 J . Org. Chem., Vol. 68, No. 19, 2003

Debenzylation of Tertiary Amines Using Phosgene or Triphosgene
TABLE 4. In flu en ce of Tem p er a tu r e, Am ou n t of Su bstr a te, a n d Solven t on th e F or m a tion of [11C]2
amine
(µmol)
11C-carbamoyl
chloride
crude rcya
(%)
11C-carbamoyl
chlorideb (%)
entry
amine
solvent(s)
c (g‚mL-1
)
T (°C)
1
2
3
4
5
6
7
8
9
13c
13c
13c
13c
13c
13c
13c
13c
13c
13d
15
154
79
59.3
79
toluene
0.13
0.1
0.1
0.1
0.1
0.05
0.01
0.01
0.01
0.01
0.1
20
-78
-78
-30
20
20
20
20
20
2
2
2
2
2
2
2
2
2
2
5
5
59
42
45
c
54
41
56
c
(74)
(76)
31 (28)
(58)
81
40
18
52
50
60
94
85
THF/CH2Cl2
THF/CH2Cl2
THF/CH2Cl2
THF/CH2Cl2
THF/CH2Cl2
THF/CH2Cl2
THF/CH2Cl2
CH2Cl2
79
39.5
7.9
3.95
7.9
7
100
100
90
10
11
12
CH2Cl2
THF/CH2Cl2
CH2Cl2
20
-30
-30
6.7
50.5
15
0.1
100
a
(5-10%, decay corrected, calculated from the amount of radioactivity in the organic layer and the radioactivity of [11C]phosgene
b
used in the reaction. Each reaction was run at least twice. The isolated yields (HPLC) are given in parentheses. Percentage (( 10%) of
[
11C]carbamoyl chloride in the organic layer. c Not determined.
lecular oxygen on N-acyltetrahydroisoquinolines42 has
never been described on urea derivatives.
SCHEME 5
Overall, the debenzylation, especially the cleavage of
electron-enriched benzyl groups, should be preferred over
the demethylation in order to generate a carbamoyl
chloride without urea production when utilizing a tertiary
amine in the presence of phosgene (or BTC). This efficient
and rapid method appeared tailor-made for the synthesis
of 11C-carbamoyl chlorides and was thus validated by the
preparation and characterization of a few 11C carbamoyl
chlorides, 11C ureas, one 11C carbamate, and one 11C
amide.
Syn th esis a n d Ch a r a cter iza tion of Tw o 11C-Ca r -
ba m oyl Ch lor id es. As model compounds, the [11C]-
carbamoyl chlorides 2 and 5 were synthesized via [11C]-
phosgene-promoted debenzylation of tertiary amines 13c
and 15. [11C]Phosgene was prepared from [11C]methane
according to a described procedure.14 The reaction of [11C]-
phosgene with N-benzyltetrahydroisoquinoline 13b was
first studied under different conditions of temperature
(-10, 20 °C), solvent (toluene, dichloromethane/THF),
and amount of substrate (up to 10 mg). Radio-TLC
analysis only showed the formation of traces of the
expected [11C]carbamoyl chloride 2 in a mixture of
unidentified products.
[11C]Phosgene-mediated debenzylation of 4-methoxy-
benzylamine 13c, 2,4-dimethoxybenzylamine 13d , and
piperazine 15 were then studied. The results are pre-
sented in Table 4. When [11C]phosgene was allowed to
react with amine 13c in a 1/1 mixture of toluene/
dichloromethane (Table 4, entry 1), more than 80% of
[11C]carbamoyl chloride 2 were present in the crude
product. No trace of the corresponding [11C]urea 3 or the
salt [11C]14c was detected. Due to the efficiency of THF
to trap [11C]phosgene, and of dichloromethane to carry
out the debenzylation, a 1/1 mixture of these solvents was
then used. In this reaction medium, 50% of the total
radioactivity was again recovered in the organic layer.
The ratio of the expected [11C]carbamoyl chloride 2
increased with the temperature until an optimum value
at -30 °C (30% at -78 °C, about 50% at -30 and 20 °C)
was reached. Comparison of entries 5-8 (Table 4) showed
that the lower the amine concentration was (entry 7), the
an amine (method A) or by a one-pot procedure involving
the reaction of phosgene with a tertiary 4-methoxyben-
zylamine and subsequent addition of a primary or
secondary amine (method B). The synthesis of tertiary
amines bearing a 4-methoxybenzyl group was carried out
by simple alkylation of the corresponding secondary
amine with 4-methoxybenzyl chloride.36
Both methods gave the same yields of ureas starting
with the tetrahydroisoquinoline derivatives (R ) 4-meth-
oxybenzyl or carbamoyl chloride, entries 1-8). As pointed
out previously, tetrahydroisoquinoline carbamoyl chloride
2 was generated quantitatively from 13c and triphosgene
using method B. Piperazine-based ureas were obtained
in good to excellent yields from method B. This method
is particularly advantageous in this case, the carbamoyl
chloride being unstable (entries 9-16). The lower yields
obtained for these ureas derived from piperazine using
method A were due to partial hydrolysis of carbamoyl
chloride 5 prior its reaction with amines. The synthetic
utility of method B was further exemplified in two more
cases with nipecotate 25 and pipecolate 27 (entries 17
and 18). Method B appears a good alternative to other
methods41 for the preparation of unsymmetrical tri- or
tetrasubstituted ureas. Some of these ureas were later
used as reference compounds in radiolabeled compounds.
A point worth noticing is the instability of urea 19
(entries 1 and 2): after purification, this compound was
oxidized rapidly (few days) upon standing in an open
flask to the isoquinolinone 29 (Scheme 5). Such a reactiv-
ity toward molecular oxygen seems dependent on the
degree of substitution of the urea. No benzylic oxidation
was observed with urea 7, and traces of R-oxo urea were
observed with tetrahydroisoquinoline 20 after two months.
To our knowledge, such an oxidation promoted by mo-
(40) Huang, K.-T.; Sun, C.-M. Biorg. Med. Chem. Lett. 2001, 11,
271-273.
(41) See, i.e.: (a) Katritzky, A. R., Pleynet, D. P. M.; Yang, B. J .
Org. Chem. 1997, 62, 4155-4158. (b) Lamothe, M.; Perez, M.; Colovray-
Gotteland, V.; Halazy, S. Synlett. 1996, 507-508.
(42) (a) Ochiai, M.; Kajishima, D.; Sueda, T. Tetrahedron Lett. 1999,
40, 5541-5544. (b) Minisci, F.; Punta, C.; Recupero, F.; Fontana, F.;
Pedulli, G. F. J . Org. Chem. 2002, 67, 2671-2676.
J . Org. Chem, Vol. 68, No. 19, 2003 7295

Lemoucheux et al.
11C-productd (%)
TABLE 5. Rea ction s of [11C]Ca r ba m oyl Ch lor id es 2 or 5 w ith Am in es or w ith a n Alcoh ola te
entrya
11C-substrateb
reagent
mmol
11C-product
crude rcyc (%)
1
2
3
4
5
6
7
8
9
2
2
2
2
2
5
5
5
5
Et2NH
n-PrNH2
1
aniline
EtONa
Et2NH
nPrNH2
4
1
7
19
3
21
30
23
22
6
70
57
nd
75
94
45
38
74
79
73
77
64
0
74
83
96
79
75
0.85
0.56
2.1
0.15
1
1.27
0.43
0.56
1
24
a
b
d
Each reaction was carried out at least twice. In CH2Cl2. c Decay corrected, ( 10%, from [11C]phosgene. Percentage of the expected
11C-product in the crude mixture (from radio-TLC).
TABLE 6. Rea ction of [11C]-La beled Ca r ba m oyl Ch lor id e 2 w ith Or ga n om eta llics
amine
entry (µmol)
crude rcyd
(%)
[
11C]31e
(%)
[
11C]2e
(%)
solventa (v/v/v)
reagentb
µmol
catalystc
T (°C)
1
2
3
4
5
6
7
8
9
39.5
39.5
592
79
7
7.9
39.5
79
3.9
7
PhMe/THF/CH2Cl2 (5/1/1) PhMgBr
PhMe/THF/CH2Cl2 (5/2/1) PhMgBr
PhMe/THF/CH2Cl2 (5/3/1) PhMgBr
70
210
105
70
140
140
140
70
20
20
20
-30
20
-30
-30
-30
-30
-30
0
0
0
42
51
44
50
55
63
54
57
67
66
21
34
18
10
13
36
43
NiCl2(PPh3)2
THF/CH2Cl2 (3/2)
THF
PhMgBr
PhMgBr
51
66
THF/CH2Cl2 (1/1)
THF/CH2Cl2 (1/1)
THF/CH2Cl2 (3/2)
THF/CH2Cl2 (1/1)
THF
PhMgBr
PhMgBr
PhMgBr
PhMgBr
NiCl2(PPh3)2
NiCl2(PPh3)2
NiCl2(PPh3)2
NiCl2(PPh3)2
24
56
52
140
56
10
Ph2CuMgBr‚BrMgCN
a
b
Solvent or mixture of solvents used for the preparation of [11C]2. After addition of the reagent, and eventually the catalyst, the
mixture was allowed to react for 5 min before addition of a saturated ammonium chloride solution and extraction with dichloromethane.
d
c 2 mg (3.06 µmol). From [11C]2, decay corrected to EOB. e Ratio of [11C]31 or [11C]2 in the organic layer (from radioTLC).
higher the ratio of [11C]carbamoyl chloride 2 was. Finally,
SCHEME 6
when [11C]phosgene was trapped in dichloromethane and
the debenzylation carried out in the same solvent, the
[11C]carbamoyl chloride 2 was isolated in 74% radio-
chemical yield [decay corrected to the end of bombard-
ment (EOB), 16 min total synthesis time]. Under similar
conditions, the debenzylation of the 2,4-dimethoxybenzyl
tetrahydroisoquinoline 13d led to [11C]carbamoyl chloride
2 in 76% isolated radiochemical yield (decay corrected
to EOB, 16 min total synthesis time). Whereas the
presence of two donor groups on the aromatic ring
improved significantly the yields in stable isotope chem-
istry, no difference was noticed here. This could be
explained by the large excess of tertiary amine (>7 µmol,
entries 9 or 10) compared to the sub-micromolar amount
of [11C]phosgene. Only the 11C-labeled carbamoyl chloride
5 was formed in the reaction of [11C]phosgene with
piperazine 15. The purities (> 96%) of the 11C-labeled
carbamoyl chlorides 2 and 5 allowed their use, in further
reactions, without any purification. Due to the low
amounts of radioactivity used, no measurement of specific
radioactivity was attempted.
2 occurred with sodium ethanolate (Table 5, entry 5), and
more than 94% of the carbamate 30 was formed in the
crude reaction mixture. When [11C]carbamoyl chloride 5
was allowed to react with amines for 5 min, more than
75% of 11C-ureas 6 and 22-24 were present in the crude
product, the remaining 25% being again the unreacted
starting material.
Preliminary identifications of the [11C]carbamoyl chlo-
rides 2 and 5 were performed using LC and radio-TLC
with co-injection of non-radioactive compounds. The
unambiguous assigment of the 11C-carbamoyl chlorides
was deduced from the subsequent synthesis of [11C]-
derivatives (Scheme 6). In particular, [11C]ureas 7 and
19 were formed by addition of the [11C]carbamoyl chloride
2 to the corresponding amines in dichloromethane. The
urea was shown only contaminated by the unreacted
labeled starting material. Table 5 summarizes the main
results. Under the conditions used, no reaction was
observed with the non-nucleophilic aniline (Table 5, entry
4). However, a rapid reaction of [11C]carbamoyl chloride
Finally, [11C]carbamoyl chloride 2 was treated with an
organometallic under the conditions described in Table
6. A survey of the data shows the role of the solvent in
the reaction of [11C]-labeled carbamoyl chloride 2 with
phenylmagnesium bromide at 20 °C. No [11C]-labeled
amide 31 was formed when toluene was used as the
cosolvent whatever the reaction conditions: presence of
the absence of catalyst (entries 1-3), high loading of
amine (entry 3) or reagent (entry 2). Heating the mixture
under reflux for 5 min did not transform the [11C]-
carbamoyl chloride 2 into the expected amide [11C]31.
When the reaction was performed in THF or THF/CH2-
Cl2, at -30 °C or at 20 °C for 5 min (Table 6, entries 4
7296 J . Org. Chem., Vol. 68, No. 19, 2003

Debenzylation of Tertiary Amines Using Phosgene or Triphosgene
and 5), formation of the [11C]amide 31 was observed with
some unreacted starting material and unidentified la-
beled compounds. A relative improvment in the yield of
the [11C]amide 31 was obtained using a small amount of
amine and addition of NiCl2(PPh3)2 (3 µmol) to the
Grignard reagent at -30 °C. However, in no case was
the reaction complete (Table 6, entries 6-10). Attempts
to catalyze the reaction using methyldiphenylphosphine
failed. Finally, the cyanocuprate Ph2CuMgBr‚BrMgCN
was found to give the best results, the reaction mixture
being easy to purify by HPLC (Table 6, entry 10), more
than 97% of the radioactivity being recovered either as
the target compound or the starting material.
They allow the preparation of carbamoyl chlorides de-
rived from less reactive tertiary benzyl anilines. In
addition, we have demonstrated the utility of this phos-
gene-mediated debenzylation of tertiary amines by pre-
paring two 11C-labeled carbamoyl chlorides in 58-75%
yields (decay corrected, 20 min total synthesis time from
[11C]CH4). They were characterized by their transforma-
tions into 11C-labeled unsymmetrical acyclic ureas or
carbamate and one 11C-amide. The simplicity and ef-
ficiency of the method has a great potential both in
organic synthesis and in any type of carbonyl-labeling
chemistry when phosgene must be used in a limited
amount. The paramethoxybenzyl group, already known
as a protecting group for secondary amines, can be
considered as a direct group relay for the synthesis of
carbamoyl chlorides and subsequently of unsymmetrical
ureas, as well as carbamates and amides.6
Con clu sion
In this paper, we have presented a versatile and
efficient one-pot synthesis of carbamoyl chlorides. The
key step, which was known as a side reaction until
recently, was the phosgene (or triphosgene, its safe
substitute) induced dealkylation of benzyl-substituted
tertiary amines. An attractive feature of this methodology
is that no symmetrical ureas are generated in the
reaction. Moreover, it provides a useful strategy for the
one-pot synthesis of unsymmetrical ureas and is particu-
larly valuable when carbamoyl chlorides are unstable.
In addition, electron-enriched benzyl substituents on the
amine proved to be superior to benzyl or alkyl group.
Ack n ow led gm en t. We thank the French Ministery
of Research and Education (MENRT) for a fellowhip to
L.L. and CNRS for financial support.
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures and full charaterization for new compounds are
reported. This material is available free of charge via the
Internet at http://pubs.acs.org.
J O0346297
J . Org. Chem, Vol. 68, No. 19, 2003 7297