primary alkylamines at low temperatures formed immobilized
triazenes, which were able to convert carboxylic acids into
carboxylic esters. Subsequently, we have described that solid-
supported O-methylisourea, readily synthesized from com-
mercial carbodiimide resin and methanol, is an excellent
reagent for the formation of methyl esters from carboxylic
acids, with reaction times as short as 15 min when executed
under microwave irradiation.4
In this communication, we report our results regarding the
O-benzylation and O-allylation of carboxylic acids using the
corresponding solid-supported isoureas under both thermal
and microwave-assisted reaction conditions. In addition, we
disclose an improved, more practical synthesis of solid-
supported isoureas.
We have reported the preparation of solid-supported
O-methylisourea by heating polymer-supported carbodiimide
1 under microwave irradiation in methanol as a solvent. No
other reagents were used, obviating the need for any sub-
sequent resin washing operations. Following this successful
preparation, we attempted to synthesize O-benzyl and O-allyl
isoureas in a similar way.
Although the urea byproduct would not interfere in the
ester formation reaction, the result is a decrease in isourea
loading, and an alternative preparation was desirable. In
addition, the procedure using microwave irradiation is only
possible on a small scale (0.5 g), and a modified preparation
suitable for larger scale production of isourea resins was
required.
The formation of isoureas is known to be easily ac-
complished by reaction of a carbodiimide and an alcohol
under CuCl catalysis,7 a process that is possible at room
temperature. As we reasoned that the rearrangement was
accelerated at high temperature, a low-temperature process
could prove to be successful.
We selected copper(II) triflate as the catalyst, not only
because of its greater solubility in THF and easy handling
and stability but also because, in our experience, it is superior
to CuCl in its activity for isourea formation. Hence, carbo-
diimide resin 1 was treated with benzyl alcohol (9 equiv) in
THF and a catalytic amount (7 mol %) of Cu(OTf)2 (Scheme
2). Gratifyingly, we observed no detectable trace of urea
Hence, solid-supported carbodiimide resin 1 (Scheme 1)5
was suspended in benzyl alcohol and heated under micro-
Scheme 2. Copper-Catalyzed Synthesis of O-Alkylisoureas
Scheme 1. Microwave-Assisted Synthesis of O-Benzylisourea
formation by IR, while the carbodiimide band completely
disappeared.8 Identical results were obtained using allyl
alcohol as a reagent.
To remove the copper species after reaction, the resin was
subjected to solvent washings. Unfortunately, we were not
able to fully eliminate the copper species by repeated wash
cycles with a variety of solvents. The failure to remove the
catalyst represented a concern because ester formation using
isourea resins that still were contaminated with copper salts
did not proceed to completion. However, a mixture (10%
v/v) of N,N′-tetramethylethylenediamine (TMEDA) in di-
chloromethane (DCM) as a washing solvent resulted in
complete removal of the copper catalyst, through diamine
complexation to the copper species. As the copper-diamine
complex has an intense deep blue color, visual monitoring
of the washing operation is very straightforward. It is worth
noting that monoamine species were not able to effect
complete removal of the copper catalyst. While amine
hydrochloride salts and amino acids are known to react with
isoureas to form quaternary ammonium salts,9 neutral amines
do not react with isoureas. We have observed no degradation
wave irradiation until disappearance of the strong carbo-
diimide IR absorption band (2119 cm-1). However, from the
IR spectra, it was apparent that apart from the expected
isourea absorption bands (1654 and 1329 cm-1), another
strong absorption was present, indicating a urea group (1640
and 1555 cm-1). When O-methylisourea was prepared in this
way, no urea band was observed. The formation of the urea
bands could be explained by a rearrangement of 2a to the
corresponding immobilized N-benzyl urea 3. Resubjecting
this resin to the same conditions led to an increase in the
proportion of urea compared to the isourea bands. Although
we have no direct proof of the exact structure of the urea
byproduct, there is solution-phase precedent for similar
rearrangements of isoureas to the corresponding N-substituted
urea products.6
(4) (a) Crosignani, S.; White P. D.; Linclau, B. Org. Lett. 2002, 6, 1035-
1037. (b) Crosignani, S.; White P. D.; Linclau, B. Org. Lett. 2002, 6, 2961-
2963.
(5) We have used both commercial carbodiimide resin (loading 1.8 mmol/
g) and resin prepared “in-house” from aminomethyl polystyrene resin
(loading 3.2 mmol/g) in two steps: reaction with cyclohexylisocyanate in
THF (Weinshenker, N. M.; Shen, C. M.; Wong, J. Y. Org. Synth. 1977,
56, 95-99), followed by dehydration (PPh3, CBr4, TEA, DCM, 16 h; Lange,
U. E. W. Tetrahedron Lett. 2002, 43, 6857-6860).
(7) Mathias, L. J. Synthesis 1979, 561-576.
(8) Use of the copper catalyst (7 mol %) under microwave irradiation
accelerates the formation of both the desired isourea 2a and the urea
byproduct: after 5 min at 100 °C, the IR spectrum shows a complete
disappearance of the carbodiimide absorption band, while both isourea and
urea bands are present. After 30 min at 120 °C, only the urea bands are
visible.
(6) Tsuboi, S.; Stromguist, P.; Overman, L. E. Tetrahedron Lett. 1976,
17, 1144-1148.
(9) Musich J. A.; Rapoport H. J. Org. Chem. 1977, 42, 139-141.
854
Org. Lett., Vol. 5, No. 6, 2003