Preparation and reaction of uracil substituted cyclen and cyclam: formation of tricyclic guanidinium and dihydroimidazolium salts

Article abstract of DOI:10.1016/j.tet.2009.12.038

Di-uracil substituted cyclen derivatives were prepared by the reaction of cyclen with 6-chloro-1-methyluracil or 6-chloro-1,3-dimethyluracil. The reaction of cyclam with 6-chloro-1,3-dimethyluracil gave a similar di-uracil substituted cyclam. The 1,7-di-uracil substituted cyclen was converted to the tricyclic guanidinium salt and acylurea upon heating in DMSO in the presence of weak acid. The 1,8-di-uracil substituted cyclam gave a tricyclic dihydroimidazolium salt under the same conditions. These reactions can be explained by an intramolecular uracil ring-breaking reaction mechanism.

Full text of DOI:10.1016/j.tet.2009.12.038

Tetrahedron 66 (2010) 1196–1201
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Preparation and reaction of uracil substituted cyclen and cyclam: formation
of tricyclic guanidinium and dihydroimidazolium salts
,
b
,
a
a
a
Soichiro Watanabe a , Ayako Komori , Masayuki Saito , Akira Uchida
*
a Department of Biomolecular Science, Faculty of Science, Toho University, Miyama 2-2-1, Funabashi, Chiba 274-8510, Japan
b Research Center for Materials with Integrated Properties, Toho University, Miyama 2-2-1, Funabashi, Chiba 274-8510, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Di-uracil substituted cyclen derivatives were prepared by the reaction of cyclen with 6-chloro-1-meth-
yluracil or 6-chloro-1,3-dimethyluracil. The reaction of cyclam with 6-chloro-1,3-dimethyluracil gave
a similar di-uracil substituted cyclam. The 1,7-di-uracil substituted cyclen was converted to the tricyclic
guanidinium salt and acylurea upon heating in DMSO in the presence of weak acid. The 1,8-di-uracil
substituted cyclam gave a tricyclic dihydroimidazolium salt under the same conditions. These reactions
can be explained by an intramolecular uracil ring-breaking reaction mechanism.
Received 9 October 2009
Received in revised form
15 December 2009
Accepted 16 December 2009
Available online 22 December 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Cyclen
Cyclam
Uracil
Tricyclic onium salt
1. Introduction
nucleobase conjugates in which the two components are directly
attached to each other.
Azamacrocycles have been studied for a number of reasons
ranging from host–guest chemistry to biological and medical use.1
Cyclen (1,4,7,10-tetraazacyclododecane) is one of the most impor-
tant macrocycles for use as metal chelators.2,3 Their derivatives
have been widely used for the synthesis of useful materials, such as
MRI constant reagents,4,5 protein active site mimics,6 supramolec-
ular building units,7,8 and so on. Azamacrocycles provide a fruitful
chemistry because of their selective complexation with metal cat-
ions as well as the unique interaction of the metal–azamacrocycle
complex with other molecules. Intermolecular interactions be-
tween a zinc–cyclen complex and a nucleobase,9 as well as those
within the cyclen–nucleobase conjugates,10 have already been
reported. In these reported systems, the two units, a cyclen and
a nucleobase, are present in a different molecule or linked by
a phenylene dimethylene group, respectively. In the latter case,
cyclen–uracil conjugates form a unique and stable complex with
a zinc cation to afford supramolecular structure, which can cleave
plasmid DNA. The relationship between the structure and the
function of these molecules is not very clear, and a variety of
cyclen–uracil conjugates need to be synthesized and tested. Con-
trary to these flexible systems, there have been no reported cyclen–
In our study to develop new supramolecular building units, we
focused on the sterically-restricted cyclen–nucleobase conjugates
because this steric restriction can cause a change in the binding
mode, leading to new properties. In addition to our interest in su-
pramolecular behavior of these molecules, they are expected to
give us a clue to explain the structure–function relationship of
aforementioned cyclen–uracil conjugates that can cleave plasmid
DNA.10 We also focused on the ring size effect because a subtle
change in the ring size sometimes causes considerable effects in
their chemical properties, especially in the host–guest chemistry.
Cyclam (1,4,8,11-tetraazacyclotetradecane) is a good alternative to
cyclen because it has two more carbon atoms.
Unfortunately, it has not been possible to obtain any evidence
suggesting a supramolecular interaction as yet. However, newly
synthesized uracil substituted cyclen and cyclam gave exotic tri-
cyclic guanidinium and dihydroimidazolium salts by heating in the
presence of weak acid. Here we report the preparation and unique
reactions of uracil substituted cyclen and cyclam.
2. Results and discussion
The introduction of uracil derivatives on to the nitrogen atoms
in cyclen 1A (n¼2) and cyclam 1B (n¼3) were carried out by the
reaction of 1 with 6-chloro-1-methyluracil (2a) or 6-chloro-1,3-
dimethyluracil (2b) (Scheme 1).
* Corresponding author. Tel./fax: þ81 47 472 1298.
E-mail address: soichiro@biomol.sci.toho-u.ac.jp (S. Watanabe).
0040-4020/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2009.12.038

S. Watanabe et al. / Tetrahedron 66 (2010) 1196–1201
1197
O
H
R
(CH2)n
(CH2)n
N
O
R
(CH2)n
(CH2)n
O
N
H
H
H
H
O
CH3
(CH2)n
(CH2)n
N
H
O
N
R
O
reflux
N
N
N
N
N
N
N
N
N
H
H
O
+
+
N
N
N
N
N
H3C
CH3CN
or EtOH
CI
CH3
N
O
CH3
N
H
R
1A: (n=2)
1B: (n=3)
2a: R=H
3Aa: n=2, R=H
4Aa: n=2, R=H
2b: R=CH3
3Ab: n=2, R=CH3
3Bb: n=3, R=CH3
4Ab: n=2, R=CH3
4Bb: n=3, R=CH3
Scheme 1. Synthesis of mono- and di-uracil substituted azamacrocycles.
The reaction of cyclen 1A with 2 equiv of 2a afforded a mono-
substituted product (3Aa) and a di-substituted product (4Aa) in 11%
and 13% yield, respectively (Table 1). This reaction gave a complex
mixture containing both starting material and byproducts together
with their target compounds. The isolation of 3Aa and 4Aa required
a tedious recrystallization-washing process resulting in low yields of
the two products. The reaction of cyclen 1A with 1 equiv of 2b gave
the mono-substituted product (3Ab) in 42% yield, whereas the same
reaction, this time with 2 equiv of 2b, afforded a di-substituted
product (4Ab)11 in 61% yield. A methyl group on uracil in the 3-po-
sition made the isolation of products practicable. However, it was
not possible to obtain the tri- or tetra-substituted derivatives
probably due to steric congestion on the cyclen ring. The reaction of
cyclen 1A with 6-chlorouracil yielded only a complex mixture. A
cyclam–uracil conjugate was also prepared in the same manner, but
this time, only the dimethyluracil derivative 2b was used because
the purification of a dimethyluracil substituted cyclamwas expected
to be easier than the purification of a monomethyl derivative. An-
other reason why only 2b was used for the synthesis of the cyclam
derivative was that both 1-methyl and 1,3-dimethyluracil derivat-
ized cyclams were expected to have similar reactivity according to
the results obtained using cyclen derivatives (vide infra). In this re-
action, only the di-uracil substituted cyclam 4Bb was obtained, the
yield of which increased with increasing amount of uracil derivative
used. No mono-substituted cyclam 3Bb was obtained, probably
because the cyclam ring has two more carbon atoms compared with
cyclen. The larger ring structure is considered to make the second
attack on the cyclam ring easier. A UV/Vis spectroscopy study was
conducted to estimate the intramolecular uracil–uracil stacking in-
In the 1H NMR spectra of 3Aa and 3Ab, several new peaks
appeared, which grew with increasing amount of zinc perchlorate.
This change is probably due to the complex formation of zinc with
cyclen. An intramolecular interaction between zinc and N1 in py-
rimidine is thought to be difficult to achieve because of the steric
congestion caused by the methyl group on the N1 position. Another
possible explanation for the observed change in the 1H NMR spectra
can be derived from the intermolecular interaction that causes di-
merization or oligomerization. This possibility is inspired by Kimura
and Aoki’s work, which shows that the zinc–cyclen complex has
a high affinity for uracil by forming a zinc–N interaction at the N3
position in the uracil.9 In our system, the effect of this intermolecular
interaction is not important, because both 3Aa with a free NH group
and 3Ab with an NCH3 group at the N3 position on the uracil ring
showed similar changes in the 1H NMR spectra.
When zinc perchlorate was added into the NMR samples of 4Aa
and4Ab inthe rangeof 0.1–3 equiv, the 1HNMRspectra showed only
a small change in chemical shift. This result indicates that com-
pounds 4A are not a good host molecule for zinc cations, because of
the steric congestion caused by the two uracil substituents on the
cyclen ring. Weak complexation of 4A with zinc can also be
explained by the fact that only two nitrogen atoms in the cyclen ring
are available for coordination to zinc because two nitrogen atoms
attached with uracil substituents hardly coordinate to zinc cation.13
An X-ray crystallographic analysis of a single crystal prepared from
a mixed solution of 4Ab and zinc perchlorate gave an ammonium
salt of 4Ab without a zinc ion accommodated in the azamacrocycle
(data not shown), indicating a poor host–guest chemistry.
VT 1H NMR measurements (up to 80 ꢀC) of 4Aa in the presence
of a zinc salt showed only a slight change in chemical shift, and the
profile of the spectra was almost the same as that measured at
room temperature. During the VT NMR experiments, we noticed
that new signals appeared upon prolonged heating at 80 ꢀC. The
reaction of both 4Aa and 4Ab in DMSO-d6 in the presence of 1 equiv
of zinc perchlorate at 120 ꢀC was monitored by 1H NMR spectros-
copy. The 1H NMR spectral change showed that compounds 4Aa
and 4Ab were almost quantitatively converted to a tricyclic gua-
nidinium salt 5 and acylurea 6 (isolated yields of 5a and 5b are 97%
and 64%, respectively) as shown in Scheme 2. The structure of 5a
was confirmed by X-ray crystallographic analysis (Fig. 1).
teraction of compound 4A. The absorption maxima of 3Aa (lmax
:
281.5 nm) and 4Aa (lmax: 282.5 nm) in water showed only a small
difference indicating that the two uracil groups in 4Aa are not effi-
ciently stacked in aqueous solution.
Table 1
1a
2b
Molar ratio(2/1)
Yield (%)
3
4
A
A
A
B
B
B
a
b
b
b
b
b
2
1
2
1
2
5
11
42
d
d
d
d
13
d
61
9
31
65
O
R
N
O
a
A: n¼2, B:n¼3.
N
Zn(ClO4)2
b
N
N
N
N
O
O
a:R¼H, b:R¼CH3.
CH3
4A
+
+
DMSO-d
N
R
N
H
6
–
Complex formation with zinc ions is one of the most attractive
ClO4
events among the metal chelating chemistry of azamacrocycles.12
To investigate the cyclen–zinc interaction in our system, a 1H NMR
analysis in DMSO-d6 was conducted in the presence and absence of
zinc perchlorate.
5a: R=H
5b: R=CH3
6a: R=H
6b: R=CH3
Scheme 2. Reaction of 4Aa and 4Ab upon heating in the presence of zinc perchlorate.

1198
S. Watanabe et al. / Tetrahedron 66 (2010) 1196–1201
While the use of more than 0.5 equiv of zinc perchlorate caused no
harmful effect in this reaction, the use of less than 0.5 M equiva-
lents of zinc perchlorate could not complete the reaction. The
whole reaction can be described as follows:
2$4AaDZnðClO4Þ2D2H2O/2$5aD2$6aDZnðOHÞ2
The same result was obtained when compound 4Ab was heated
under the same reaction conditions.
The present reaction of 4Aa in DMSO-d6 with a few drops of D2O
showed the involvement of water in this reaction (Fig. 2(c)). Deu-
terium atoms were incorporated into the methyl group in an acetyl
position in acylurea 6a (1.98 ppm). A control experiment showed
that no H–D exchange occurred at the acetyl position of 6a under
the same conditions. We also noticed that the integration of the
proton signal associated with the C5 position of uracil in product 5a
(5.24 ppm) also decreased. A H–D exchange at the C5 position, as
well as the incorporation of D atoms into acylurea 6a, must provide
a clue to understand the reaction process.
Figure 1. ORTEP drawing of guanidinium salt 5a. A perchlorate ion was omitted for
clarity.
We also confirmed that no reaction occurred after heating of
4Aa with zinc hydroxide. This result means that zinc hydroxide, an
expected byproduct of this reaction, is not involved in the forma-
tion of 5a.
In the present reaction, a guanidinium moiety was formed with
a loss of one of the uracil groups on the cyclen ring in 4A, accom-
panied by the formation of acylurea. To obtain insights into the
requirement for this reaction, a number of control experiments
were carried out. Simple heating of 4Aa in DMSO-d6 at 120 ꢀC
without zinc perchlorate did not show any change in the 1H NMR
spectrum, indicating that a zinc ion is necessary for this reaction.
Although the addition of zinc perchlorate into a mixed solution of
cyclen and 1-methyluracil in DMSO-d6 caused a change in the 1H
NMR spectrum, due to the complex formation of zinc–cyclen with
uracil, no reaction was observed after heating of this sample at
120 ꢀC. This result suggests that a uracil unit must be attached di-
rectly to the nitrogen atom on cyclen for the present reaction. A
0.5 equiv of zinc perchlorate compared with 4Aa was enough to
complete this reaction, indicating that one zinc ion could convert
two 4Aa compounds to the guanidinium salt (Fig. 2 (a) and (b)).
Tricyclic guanidinium salts or tricyclic orthoamides were pre-
pared by the reaction of cyclen with dimethylformamide dialkyl
acetal14–17 or ethylorthocarbonate.18,19
A similar reactive in-
termediate must exist in the present reaction. We believe that the
reaction starts with intramolecular nucleophilic addition of the
nitrogen atom in the cyclen ring at the C6 position in uracil. No
reaction occurred from the simple heating of 4A in DMSO-d6,
suggesting that this reaction requires a proton donor to proceed. A
zinc ion can provide a proton via Zn2þ-bound H2O. As a result,
compound 4A forms an ammonium salt from which we could ob-
tain a single crystal (vide supra).
Our hypothesis is that compound 4A itself does not have a good
proton source to complete the reaction after the initial nucleophilic
Figure 2. 1H NMR spectra of (a) 4Aa, (b) 4Aa after reaction with 0.5equiv of zinc perchlorate, (c) 4Aa after reaction with 0.5 equiv of zinc perchlorate in the presence of D2O. Asterisk
(*) indicates signals of DMSO and H2O.

S. Watanabe et al. / Tetrahedron 66 (2010) 1196–1201
1199
O
H
O
O
R
R
R
O
N
R
N
N
N
H
N
O
CH3
O
HX
H
N
H
O
CH3
O
CH3
H
H
U
N
N
N
N
N
H
U
N
N
N
N
N
N
+
HN
CH3
4A
5
6
N
N
N
N
N
N
U
U
H
H
X–
X–
X–
X–
Scheme 3. Proposed reaction mechanism for the reaction of 4A in the presence of weak acid. U denotes uracil substituents.
ions. This prediction was confirmed by a series of reactions in the
presence of several acids. The reaction with ammonium chloride,
acetic acid, and barbituric acid in DMSO proceeded well, whereas
the use of hydrochloric acid gave only a complex mixture. Our
proposed reaction mechanism is depicted in Scheme 3.
O
N
CH3
O
N
Zn(ClO4)2
DMSO
N
N
N
N
4Bb
O
CH3
+
CH3
H3C
N
H
N
H
O
–
ClO4
The reaction remains in equilibrium until one of the uracil rings
in 4A opens. In this mechanism, H–D exchange at the C5 position of
the uracil ring occurs during the protonation of an enolate in-
termediate. This H–D exchange leads to the incorporation of deu-
terium in both the C5 position of uracil in 5 and the acyl methyl
position in 6, being consistent with the results shown in Figure 2(c).
With regards to the pyrimidine nucleotide metabolism in eu-
karyotes, the ring-opening reaction is achieved by hydrolysis of the
amide bond between the N3 and C4 positions of the dihydropyr-
imidine ring. This same bond breaks by the reaction of a pyrimidine
nucleobase with hydrazine in the Maxam–Gilbert sequencing
method. In the present reaction, the C6 position in the uracil ring is
attacked twice by the nitrogen of the cyclen resulting in the
cleavage of two of the bonds (N1–C6 and C5–C6) in the uracil. The
difference in the reaction mode is explained when it is considered
that the nucleophilic nitrogen atoms in cyclen are restricted to the
proximity of the C6 position of uracil and hardly approach the
carbonyl carbon. The C5–C6 bond cleavage in the pyrimidine ring is
unique in this compound.
8
7 (50%)
Scheme 4. Reaction of 4Bb upon heating in the presence of zinc perchlorate.
The importance of the steric restriction effect was also con-
firmed when an azamacrocycle having a slightly bigger ring size
was used. Although a similar reaction was observed to take place
when the cyclam–uracil conjugate 4Bb was used instead of cyclen–
uracil derivatives 4Aa and 4Ab (Scheme 4), the nitrogen on the
azamacrocycle attacked the uracil at a different position. The
product this time was the dihydroimidazolium salt 7, the structure
of which was determined by X-ray crystallographic analysis (Fig. 3).
We also confirmed by 1H NMR that dimethylurea 8 was formed
after the reaction.
Figure 3. ORTEP drawing of dihydroimidazolium salt 7. A perchlorate ion was omitted
for clarity.
addition, whereas the di-ammonium ion of 4A lacks a good nu-
cleophile. Only a mono-ammonium ion can possess both a nucleo-
philic nitrogen atom and a good proton donor (ammonium ion) in
close vicinity around the C6 position in uracil. Based on this hy-
pothesis, the present reaction does not always require a zinc ion,
but requires a weak proton donor to produce mono-ammonium
O
O
N
O
CH3
CH3
O
CH3
O
N
N
N
H
N
H
N
H
U
H
H
HX
N
N
N
N
N
CH3
O
N
N
N
N
N
CH3
CH3
N
N
H
U
H
U
H
X–
X–
O
N
O
CH3
CH3
N
HN
N
N
N
N
O
H
O
N
N
O
N
N
N
N
CH3
O
+
CH3
H3C
CH3
N
H
N
H
U
N
N
N
U
H
U
H
X–
X–
X–
Scheme 5. Proposed reaction mechanism for the reaction of 4Bb in the presence of weak acid. U denotes uracil substituents.

1200
S. Watanabe et al. / Tetrahedron 66 (2010) 1196–1201
The plausible reaction mechanism is shown in Scheme 5. To
(s, 6H), 5.50 (s, 2H); 13C NMR (75 MHz, D2O):
(t), 92.5 (d), 154.3 (s), 163.8 (s), 166.9 (s); UV/VIS (H2O), lmax 282.5 (
27,200) nm. Anal. Calcd for C18H28N8O4: C, 51.41; H, 6.71; N, 26.65.
Found: C, 51.16; H, 6.66; N, 27.11%.
d 33.0 (q), 47.5 (t), 50.4
confirm the reaction mechanism, we tried the same experiment for
the cyclen–uracil derivative 4A. The same reaction of 4Bb shown in
Scheme 4 occurred with barbituric acid as the weak acid instead of
zinc perchlorate. When D2O was added to the reaction mixture in
the NMR sample tube, the integral of the proton signal due to H5 in
the uracil ring as well as the methylene next to the dihy-
droimidazolium ring were diminished. These results are consistent
with the reaction mechanism shown in Scheme 5. The difference in
products between the cyclen and cyclam derivatives was consid-
ered as follows. Although the first nucleophilic attack occurred at
the C6 position in the uracil ring in both cases, the second attack
took place at a different carbon atom. The C6 carbon was attacked
by the nitrogen atoms in the cyclen, whereas the C4 carbonyl car-
bon was attacked in the cyclam derivatives. This is because the
additional carbon in the cyclam ring compared with the cyclen ring
provides enough flexibility for the azamacrocycle to attack the C4
carbon in the uracil ring.
3
4.3. 1,3-Dimethyluracil-6-yl-cyclen (3Ab)
A
solution of cyclen (0.8631 g, 5.010 mmol), 6-chloro-1,3-
dimethyluracil (0.8772 g, 5.025 mmol), and sodium carbonate
(0.5231 g, 4.936 mmol) in acetonitrile (25 ml) was stirred under
reflux for one day. After the filtration of precipitates, the filtrate was
concentrated under reduced pressure. The residue was purified by
preparative GPC to afford 3Ab (0.6440 g, 2.077 mmol, 41.5%).
Compound 3Ab: white crystals. Mp: 104.5–106.0 ꢀC; 1H NMR
(300 MHz, D2O):
d
2.66 (t, J¼5.8 Hz, 4H), 2.77 (m, 8H), 3.13 (t,
J¼5.8 Hz, 4H), 3.21 (s, 3H) 3.43 (s, 3H), 5.61 (s, 1H); 13C NMR
(75 MHz, D2O): d 27.9 (q), 33.0 (q), 44.1 (t), 44.8 (t), 45.7 (t), 49.0 (t),
90.6 (d), 154.1 (s), 160.8 (s), 165.7 (s). Anal. Calcd for
C14H26N6O2$1.5H2O: C, 49.83; H, 8.66; N, 24.91. Found: C, 50.30; H,
8.23; N, 24.71%.
3. Conclusion
In summary, we have synthesized uracil derivatized azamacro-
cycles. The reaction of di-substituted cyclen and cyclam in the
presence of weak acid gave tricyclic guanidinium and dihy-
droimidazolium salt together with urea, respectively. The differ-
ence in the products was due to the ring size of azamacrocycles.
Although host–guest chemistry between azamacrocycles and metal
cations often depends on the ring size of the macrocycles, this ar-
ticle shows that the reaction of azamacrocycle derivatives also
depends on the subtle change in the ring size.
4.4. 1,7-Bis(1,3-dimethyluracil-6-yl)-cyclen (4Ab)
A solution of cyclen (0.6088 g, 3.534 mmol), 6-chloro-1,3-
dimethyluracil (1.3094 g, 7.500 mmol), and sodium carbonate
(0.7451 g, 7.030 mmol) in acetonitrile (20 ml) was stirred under
reflux for one day. After a thermal filtration, white precipitate 4Ab
(0.9648 g, 2.154 mmol, 61.0%) was formed from the filtrate. Com-
pound 4Ab: white crystals. Mp: 223.0–224.5 ꢀC; 1H NMR (400 MHz,
DMSO-d6):
d
2.68 (t, J¼6.0 Hz, 8H), 3.07 (t, J¼6.0 Hz, 8H), 3.11 (s,
6H), 3.60 (s, 6H), 5.43 (s, 2H); 13C NMR (100 MHz, DMSO-d6):
d
27.2
4. Experimental
(q), 32.7 (q), 46.2 (t), 49.5 (t), 88.9 (d), 152.7 (s), 159.5 (s), 162.2 (s).
Anal. Calcd for C20H32N8O4: C, 53.55; H, 7.19; N, 24.99. Found: C,
53.43; H, 7.10; N, 25.12%.
4.1. General methods
Melting points were determined on a Yanaco micro melting
point apparatus. All melting points were uncorrected. Preparative
gel permeation chromatography was performed by LC-908 (Japan
Analytical Industry) with a JAIGEL GS-310 column with MeOH or
JAIGEL 1Hþ2H columns with CHCl3 as solvent. UV–vis spectra were
recorded on a HITACHI U-3010 spectrophotometer. 1H and 13C NMR
spectra were measured in CDCl3, DMSO-d6, CD3OD, or D2O with
a JEOL ECP 400 or Bruker Avance 300 spectrometer using tetra-
methylsilane as an external standard. Elemental analyses were
performed by the Instrument Analysis Center of School of Phar-
maceutical Sciences, Toho University.
4.5. 1,8-Bis(1,3-dimethyluracil-6-yl)-cyclam (4Bb)
A solution of cyclam (0.6007 g, 2.998 mmol), 6-chloro-1,3-
dimethyluracil (2.5859 g, 14.81 mmol), and sodium carbonate
(1.5925 g, 15.02 mmol) in acetonitrile (40 ml) were stirred under
reflux for one day. The reaction mixture was cooled with ice-water
bath and the precipitate was filtered. This solid was dissolved in
CHCl3 (50 ml) and the precipitate was removed by filtration. The
filtrate was concentrated under reduced pressure. The resulting
solid was dissolved in MeOH (40 ml), white precipitate was formed
to afford 4Bb (0.9212 g, 1.933 mmol, 64.5%). Compound 4Bb: white
crystals. Mp: 245.5–247.5 ꢀC; 1H NMR (300 MHz, CDCl3):
4H), 2.59 (t, J¼5.4 Hz, 4H), 2.95 (s, 8H), 3.24 (t, J¼5.8 Hz, 4H), 3.34 (s,
6H), 3.50 (s, 6H), 5.38 (s, 2H); 13C NMR (75 MHz, CDCl3):
24.9 (t),
d 1.75 (m,
4.2. 1-Methyluracil-6-yl-cyclen (3Aa) and 1,7-bis(1-
methyluracil-6-yl)-cyclen (4Aa)
d
27.8 (q), 32.0 (q), 43.8 (t), 46.1 (t), 48.0 (t), 54.8 (t), 90.8 (d), 153.1 (s),
158.3 (s), 163.1 (s). Anal. Calcd for C22H36N8O4: C, 55.44; H, 7.61; N,
23.51, Found: C, 55.03; H, 7.49; N, 23.11%.
A solution of cyclen (1.7248 g, 10.01 mmol) and 6-chloro-1-
methyluracil20 (3.2110 g, 20.00 mmol) in ethanol (40 ml) were
stirred under reflux for 2 h. The suspension was filtered and the
precipitate was washed with hot solvent (water/ethanol¼1:1). The
residual solid was dried under reduced pressure to afford 4Aa
(0.5594 g, 1.331 mmol, 13.3%). From the filtrate derived from re-
action mixture, white precipitates were formed. After filtration, the
precipitates were washed with hot ethanol, hot acetonitrile and
then dried under reduced pressure to afford 3Aa (0.3323 g,
1.121 mmol, 11.2%). Compound 3Aa: white crystals. Mp: 198.0–
4.6. 10-(1-Methyluracil-6-yl)-1,4,7,10-
tetraazatricyclo[5.5.1.04,13]tridecanium perchlorate (5a)
A mixed solution of 4Aa (0.0529 g, 0.126 mmol) in DMSO (1 ml)
and 0.50 M zinc perchlorate in DMSO (0.25 ml, 0.13 mmol) was
heated at 120 ꢀC for 1 h. After cooled to room temperature, chlo-
roform was added to the reaction mixture. Filtration of the pre-
cipitates afforded 5a (0.0497 g, 0.125 mmol, 97%). Compound 5a:
white crystals. Mp: 292.0–293.0 ꢀC; 1H NMR (300 MHz, DMSO-d6):
199.0 ꢀC; 1H NMR (300 MHz, D2O):
d
2.99 (m, 8H), 3.05 (t, J¼6.0 Hz,
4H), 3.25 (t, J¼6.0 Hz, 4H), 3.40 (s, 3H), 5.55 (s, 1H); 13C NMR
(75 MHz, D2O):
d
32.3 (q), 43.5 (t), 44.2 (t), 45.1 (t), 48.7 (t), 91.2 (d),
d
3.19 (s, 3H), 3.24 (t, J¼5.7 Hz, 4H), 3.44 (t, J¼7.5 Hz, 4H), 3.80 (t,
153.6 (s), 161.9 (s), 166.1 (s); UV/VIS (H2O), lmax 281.5 (
Compound 4Aa: white crystals. Mp: 261.0–262.5 ꢀC; 1H NMR
(300 MHz, D2O):
2.89 (t, J¼6.0 Hz, 8H), 3.25 (t, J¼6.0 Hz, 8H), 3.35
3
11,000) nm.
J¼5.7 Hz, 4H), 4.05 (t, J¼7.5 Hz, 4H), 5.23 (s, 1H), 11.13 (s, 1H); 13C
NMR (75 MHz, DMSO-d6):
d 31.0 (q), 42.1 (t), 45.2 (t), 51.0 (t), 55.5
d
(t), 91.1 (d), 151.9 (s), 162.4 (s), 162.9 (s), 163.9 (s).

S. Watanabe et al. / Tetrahedron 66 (2010) 1196–1201
1201
4.7. 10-(1,3-Dimethyluracil-6-yl)-1,4,7,10-
CCDC, 12 Union Road, Cambridge CB2 1EZ, UK, (fax: þ44 1223
336033 or e-mail: deposit@ccdc.cam.ac.uk).
tetraazatricyclo[5.5.1.04,13]tridecanium perchlorate (5b)
Crystal data for 5a: C14H21N6O2$ClO4, Mr¼404.82, T¼100(1) K,
A mixed solution of 4Ab (0.0569 g, 0.127 mmol) in DMSO (1 ml)
and 0.50 M zinc perchlorate in DMSO (0.25 ml, 0.13 mmol) was
heated at 120 ꢀC for 1 h. After the solvent was evaporated under
reduced pressure, water and ether was added to the residue. The
aqueous layer was evaporated under reduced pressure and the
residue was washed with cold methanol to afford 5b (0.0339 g,
0.0807 mmol, 63.6 %). Compound 5b: white crystals. Mp: 232.0–
monoclinic, space group C2/c, Z¼8, a¼23.995(3) Å, b¼8.5986(12) Å,
c¼19.341(3) Å,
b
¼120.971(3)ꢀ, V¼3421.5(8) Å 3, Dx¼1.572 Mgmꢁ3
,
R¼0.0606, wR(F2)¼0.1629.
Crystal data for 7: C19H29N6O3$ClO4, Mr¼488.93, T¼100(2) K,
triclinic, space group P-1, Z¼2, a¼7.2524(5) Å, b¼10.1157(7) Å,
c¼14.6156(9) Å,
a
¼89.1360(10)ꢀ,
b
¼84.8900(10)ꢀ,
g
¼82.0580(10)ꢀ,
V¼1057.74(12) Å 3, Dx¼1.535 Mgmꢁ3, R¼0.0467, wR(F2)¼0.1303.
233.0 ꢀC; 1H NMR (300 MHz, D2O):
d 3.22 (s, 3H), 3.35 (s, 3H), 3.40
(t, J¼6.0 Hz, 4H), 3.47 (t, J¼7.8 Hz, 4H), 3.81 (t, J¼6.0 Hz, 4H), 4.08 (t,
Acknowledgements
J¼7.8 Hz, 4H), 5.48 (s, 1H); 13C NMR (75 MHz, D2O):
d 27.9 (q), 33.2
(q), 42.6 (t), 45.2 (t), 51.0 (t), 55.7 (t), 90.5 (d), 154.0 (s), 162.1 (s),
164.1 (s), 165.7 (s). Anal. Calcd for C15H23N6O6Cl$H2O: C, 41.24; H,
5.77; N, 19.24. Found: C, 40.86; H, 5.28; N, 18.68%.
This work was supported by ‘High-Tech Research Center’ Project
for Private Universities: matching fund subsidy from MEXT (Min-
istry of Education, Culture, Sports, Science and Technology), 2005–
2009.
4.8. 11-(1,3-Dimethyluracil-6-yl)-1,4,8,11-tetraaza-17-oxo-
tricyclo[6.6.3.04,15]heptadecanium perchlorate (7)
References and notes
1. Delgado, R.; Felix, V.; Lima, L. M. P.; Price, D. W. Dalton Trans. 2007, 2734.
2. Curtis, N. F. In Comprehensive Coordination Chemistry II; McCleverty, A., Thomas,
T. J., Eds.; Elsevier: Oxford, UK, 2004; Vol. 1, pp 447–474.
3. Alexander, V. Chem. Rev. 1995, 95, 273.
4. Caravan, P.; Ellison, J. J.; McMurry, T. J.; Lauffer, R. B. Chem. Rev. 1999, 99, 2293.
5. Bianchi, A.; Calabi, L.; Corana, F.; Fontana, S.; Losi, P.; Maiocchi, A.; Paleari, L.;
Valtancoli, B. Coord. Chem. Rev. 2000, 204, 309.
6. Aoki, S.; Kimura, E. In Comprehensive Coordination Chemistry II; McCleverty, A.,
Thomas, T. J., Eds.; Elsevier: Oxford, UK, 2004; Vol. 8, pp 601–640.
7. Aoki, S.; Shiro, M.; Kimura, E. Chem.dEur. J. 2002, 8, 929.
8. Aoki, S.; Zulkefeli, M.; Shiro, M.; Kimura, E. Proc. Natl. Acad. Sci. U.S.A. 2002,
99, 4894.
9. Aoki, S.; Kimura, E. Chem. Rev. 2004, 104, 769.
10. Xia, C.-Q.; Jiang, N.; Zhang, J.; Chen, S.-Y.; Lin, H.-H.; Tan, X.-Y.; Yue, Y.; Yu, X.-Q.
Bioorg. Med. Chem. 2006, 14, 5756.
11. The X-ray crystallographic analysis data of 4b has already been reported, see:
Uchida, A.; Komori, A.; Watanabe, S. Acta Crystallogr. 2007, E63, o648.
12. Archibald, S. J. In Comprehensive Coordination Chemistry II; McCleverty, A.,
Thomas, T. J., Eds.; Elsevier: Oxford, UK, 2004; Vol. 6, pp 1147–1251.
13. Aoki, S.; Kagata, D.; Shiro, M.; Takeda, K.; Kimura, E. J. Am. Chem. Soc. 2004,
126, 13377.
A solution of 4Bb (0.2388 g, 0.503 mmol) and zinc perchlorate
hexahydrate (0.3894 g, 1.046 mmol) in DMSO (2 ml) was heated at
120 ꢀC for 1 h. After the reaction mixture was cooled to room
temperature, water and ether was added. The aqueous layer was
evaporated under reduced pressure and the residue was recrys-
tallized from methanol to afford 7 (0.1086 g, 0.222 mmol, 44.2 %).
Compound 7: white crystals. Mp: decomp. below 300 ꢀC; 1H
NMR (300 MHz, DMSO-d6):
d 1.40–1.61 (m, 1H), 1.93–2.11 (m,
1H), 2.12–2.41 (m, 3H), 2.77 (d, J¼12.7 Hz,1H), 2.96 (d, J¼12.7 Hz,1H),
3.00–3.59 (m, 5H), 3.13 (s, 3H), 3.54 (s, 3H), 3.74 (d, J¼14.9 Hz, 1H),
3.65–3.92 (m, 4H), 3.97–4.20(m, 3H), 4.51(d, J¼14.9 Hz, 1H), 4.86 (t,
J¼12.7 Hz, 1H), 5.25 (s, 1H); 13C NMR (75 MHz, DMSO-d6):
d 22.9 (t),
27.3 (t), 27.8 (q), 32.9 (q), 34.2 (t), 43.3 (t), 45.6 (t), 47.0 (t), 49.8 (t),
50.2 (t), 54.5 (t), 88.0 (d),152.5 (s),159.3 (s),162.1 (s),164.2 (s),166.7
(s). Anal. Calcd for C19H29N6O7Cl: C, 46.67; H, 5.98; N, 17.19. Found:
C, 46.24; H, 5.87; N, 16.80%.
14. Atkins, T. J. J. Am. Chem. Soc. 1980, 102, 6364.
15. Anelli, P. L.; Murru, M.; Uggeri, F.; Virtuani, M. J. J. Chem. Soc., Chem. Commun.
1991, 1317.
16. Platzek, J.; Blaszkiewicz, P.; Gries, H.; Luger, P.; Michl, G.; Mu¨ller-Fahrow, A.;
Radu¨ chel, B.; Su¨ lzle, D. Inorg. Chem. 1997, 36, 6086.
4.9. X-ray crystallographic analysis
17. Anelli, P. L.; Calabi, L.; Dapporto, P.; Murru, M.; Paleari, L.; Paoli, P.; Uggeri, F.;
Verona, S.; Virtuani, M. J. Chem. Soc., Perkin Trans. 1 1995, 2995.
18. Richman, J. E.; Simmons, H. E. Tetrahedron 1974, 30, 1769.
19. Li, Z.; Undheim, K. Acta Chem. Scand. 1998, 52, 1247.
Crystallographic data for 5a and 7 have been deposited with the
Cambridge Crystallographic Data Center as supplementary publi-
cation number CCDC 654509 and CCDC 743682, respectively.
Copies of the data can be obtained, free of charge, on application to
20. Youssef, S.; Pfleiderer, W. J. J Heterocycl. Chem. 1998, 35, 949.