B. Kelly, I. Rozas / Tetrahedron Letters 54 (2013) 3982–3984
3983
which is subsequently S-methylated by methyl iodide and cleaved
S
in two steps by ammonia to yield the guanidine.10 The 2-aminoim-
idazoline moiety of clonidine is introduced from the corresponding
aryl amine by first converting it into the thiourea using ammonium
thiocyanate, then S-methylating with iodomethane and reacting
with the powerfully nucleophilic ethylenediamine.11 This is a clean
synthesis but only works for the preparation of 2-aminoimidazo-
lines and, like the synthesis of famotidine, still requires the use
of light-sensitive and highly toxic iodomethane. Clearly, there is
a need for safe and efficient syntheses of guanidines considering
their potential as therapeutics and synthetically useful reagents.
While many methods exist for the introduction of the guanidine
functionality, the prevailing problem is that they only work with
aliphatic amines, and not with deactivated aryl amines. Mukaiy-
ama’s reagent12 can also promote the coupling of amines with
thiourea derivatives, while several reagents, such as protected pyr-
azole-1-carboximidamides, have a leaving group on a protected
guanidine, which can react with amines yielding protected guani-
dines (Fig. 2).9
Thus, we sought to find a method which would allow the gua-
nidylation of unreactive aryl amines without the need for mercury
salts. The use of copper for guanidylation reactions has been ex-
plored by different authors. For example, Lai and co-workers used
a CuI/N-methylglycine-catalysed reaction to couple guanidine ni-
trate with aryl iodides or bromides at 70ꢀ100 °C to afford symmet-
rical N,N0-diaryl guanidines with good to excellent yields.13
Additionally, Terada and co-workers have reported, under mild
conditions, the amination of S-methyl-N,N0-bis-Boc-isothiourea
with either primary or sterically hindered secondary amines pro-
moted by copper(I) chloride and K2CO3 resulting in N,N0-bis-Boc
protected guanidines in good to excellent yields.14
During the synthesis of a related family of 1,4-dihydroquinazo-
line compounds,15 copper(II) oxide was used in catalytic amounts
as a desulfurizing agent to promote the intramolecular coupling
of an aryl amine and N,N0-di-(tert-butoxycarbonyl)thiourea. With
this in mind an investigation into the possibility of extending this
methodology to intermolecular guanidine formation was under-
taken. Under the same catalytic conditions, only trace amounts of
product were isolated (<5%). Similarly, using a stoichiometric
amount of copper(II) oxide gave very low yields (<5%). Use of a
stoichiometric oxidant [N-methylmorpholine N-oxide (NMO)]
was also attempted to examine if the catalytic species was being
consumed in the reaction, but no yield improvements were ob-
served. It seemed that a leaving group was required on the cop-
per(II) species for the reaction to occur. Revisiting copper(II)
chloride, which had been mentioned in Kim and Qian’s original
Letter,7 but was thought to be less effective than mercury(II) chlo-
ride, the guanidylation of test compound aniline using stoichiom-
etric copper(II) chloride and 2 equiv of Et3N (Scheme 1) was
effected in high yield (73%).
Boc
Boc
Boc
N
H
N
H
N
1
R
R
Boc
NH2
N
H
N
H
Et3N, CuCl2
2-11
Scheme 1. General procedure for the guanidylation reaction of amines promoted
by CuCl2.
butoxycarbonyl) thiourea (1) can be synthesised on large scale
and in high yield (90%) from thiourea and di-tert-butyl dicarbon-
ate.16 To investigate the general applicability of this method
(Scheme 1) different amines of varied reactivity were subjected
to the conditions (Table 1).
Where available, yields were compared to published values for
the corresponding HgCl2-promoted guanidylation. Reactions were
carried out at room temperature in dichloromethane, but for more
deactivated amines slightly more forceful conditions using dimeth-
ylformamide as the solvent and heating at 60 °C were employed,
also yielding products in acceptable yields. It is worth noting that
the guanidylation of para-nitroaniline to give 10, which was not
achieved in other studies, was accomplished using our method
(41% yield). Furthermore, 4-aminopyridine was successfully gua-
nidylated (26%) using CuCl2 to give 11, where this had not been
possible in our hands with HgCl2.
Based on these results we report that this method is at least as
efficient and as high yielding as the HgCl2-promoted guanidylation,
and vastly more desirable as it obviates the need to use mercury
salts. It is also applicable to both aliphatic and aryl amines, which
are essential to its practicality.
Having shown that the method was successful in generating
N,N0-di-Boc-protected guanidines we wanted to extend it to
the synthesis of N,N0-di-Boc-protected 2-aminoimidazolines and
Table 1
CuCl2 promoted coupling of amines with N,N0-di-(tert-butoxycarbonyl)thiourea (1).
Published yields using HgCl2 are given in parentheses
R
Conditions
Product
Yield (%)
89
CH2Cl2, rt, 3 h
2
CH2Cl2, rt, 3 h
CH2Cl2, rt, 3 h
CH2Cl2, rt, 5 h
3
4
5
89 (40)17
83 (59)18
76 (76)19
N
EtO
CH2Cl2, rt, 16 h
CH2Cl2, rt, 16 h
6
7
73
52
Notably, the yield matched exactly, published values for the
same reaction promoted by HgCl2 (73%).5b It should be noted that
the commercially available N,N0-bis-(tert-butoxycarbonyl)-S-
methyl isothiourea, which can often be used instead of N,N0-di-
(tert-butoxycarbonyl)thiourea in HgCl2 promoted guanidylations,
was not successful when CuCl2 was used. Besides, N,N0-di-(tert-
F3C
Cl
CH2Cl2, rt, 16 h
CH2Cl2, rt, 16 h
8
9
67 (71)19
N
H3C
F
62
Boc
N
Boc
N
N
N
H
Cl
N
DMF, 60 °C, 16 h
DMF, 60 °C, 16 h
10
11
41
26
I
O2N
N
Figure 2. Reagents used in the synthesis of guanidines from aliphatic amines:
Mukaiyama’s reagent (left), and Boc-protected pyrazol-1-ylcarboximidamides
(right).