Solid-phase synthesis of cyclic polyamines

Article abstract of DOI:10.1016/j.tet.2008.05.119

A method for the synthesis of cyclic polyamines based on solid-phase chemistry is shown. Linear polyamines are stepwise synthesized on solid support from the center by repetitive alkylations at benzylic N-atoms. Cyclizations at the resins were effected conventionally by direct intramolecular S_N2 reactions between sulfonyl-protected terminal amino groups and primary alkyl bromides or by intramolecular Mitsunobu reactions between sulphonamides and primary alcohols. Particularly the latter transformation proved to be powerful for the construction of medium- as well as large-sized rings.

Full text of DOI:10.1016/j.tet.2008.05.119

Tetrahedron 64 (2008) 7531–7536
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Solid-phase synthesis of cyclic polyamines
*
Pascal Bisegger, Nikolay Manov, Stefan Bienz
Institute of Organic Chemistry, University of Zurich, Winterthurerstrasse 190, 8057 Zurich, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 11 April 2008
Received in revised form 26 May 2008
Accepted 29 May 2008
Available online 3 June 2008
A method for the synthesis of cyclic polyamines based on solid-phase chemistry is shown. Linear
polyamines are stepwise synthesized on solid support from the center by repetitive alkylations at
benzylic N-atoms. Cyclizations at the resins were effected conventionally by direct intramolecular SN2
reactions between sulfonyl-protected terminal amino groups and primary alkyl bromides or by intra-
molecular Mitsunobu reactions between sulphonamides and primary alcohols. Particularly the latter
transformation proved to be powerful for the construction of medium- as well as large-sized rings.
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
the center.8–10 Based on this strategy we planned to form cyclic
polyazacompounds of different sizes on solid support by ring
The construction of cyclic polyamines is an important task in
organic synthesis because such polyamines are widely used as
radiopharmaceuticals1 and MRI contrast agents2 and also have
diagnostic applications.3 Furthermore, macrocyclic polyamines
continue to be important components in coordination chemistry.4
Azacrowns have been utilized as additives to regulate enzyme
reactivity: some cyclic polyamines significantly enhance both
enantioselectivity and efficiency of enzymatic reactions, even for
substrates that are otherwise not favored by the enzyme.5 In-
teresting is also the fact that phenylenebis(methylene)-linked di-
mers of cyclic polyamines were described as a new class of antiviral
agents that exhibit potent inhibitory effects on several strains of
human immunodeficiency virus type 1 (HIV-1) and type 2 (HIV-2)
replication with high selectivity.6 A high-yielding synthesis to link
the two polyamine rings together was described by Le Baccon et al.7
The classical synthesis of macrocyclic polyamines has two major
drawbacks: (i) polyamines are compounds of rather high polarity,
which often enough impedes their purification and (ii) macro-
cyclizations to close their rings usually require high-dilution tech-
niques to avoid dimerization and polymerization processes and are
thus uneconomical. These disadvantages can be overcome by using
solid-phase strategies, as shown in this paper.
closing reactions with appropriately a,u-functionalized precursors.
Due to the limited loading capacity of resins and the statistical
distribution of the active sites, a natural dilution effect is realized
for solid-phase reactions. Limitations for the cyclizations were only
expected for the formation of medium-sized or very large rings. For
the closure of such rings, reaction rates might be too low, allowing
competing decompositions, and, for the formation of ‘very large’
rings, ‘cross-alkylation’ of neighboring polyamine moieties on the
solid support could additionally be problematic. In a first instance,
we concentrated on the synthesis of 8-, 12-, and 16-membered
cyclic polyamines.
The cascades of repetitive alkylations for the preparations of the
targeted cyclization precursors 3, 6, 7, 8, 9, and 10 started with the
coupling of N-(3-aminopropyl)-2-nitrobenzenesulfonamide (1)11
to the Merrifield polymer (200–400 mesh, loading capacity
0.73 mmol gꢀ1 or 0.32 mmol gꢀ1), performed in the presence of
diisopropylethylamine (DIEA), leading to resin 2 (Scheme 1). Al-
kylation of this resin with 1,3-dibromopropane in presence of DIEA
delivered the first cyclization precursor, resin 3. The Ns-protected
amine does not react under these conditions.12 Substitution of the
Br-atom of resin 3 by reaction with either benzylamine or N,N0-
dibenzylpropane-1,3-diamine in presence of DIEA, followed by
alkylation of the resultant resins 4 and 5 with 1,3-dibromopropane
as above, afforded the second and third cyclization precursor, resins
6 and 7.
2. Results and discussion
The hydroxy analogs 8, 9, and 10 were prepared from 2, 4, and 5,
respectively, by their alkylation with 3-bromopropanol.
The basis of our study is the previously described solid-phase
synthesis of linear polyamines and polyamine derivatives, con-
structing the polyamine framework in a convergent manner from
Cyclization of the solid-bound a-NHNs-u-bromoazaalkanes was
performed by treatment of the resins 3, 6, and 7 with K2CO3 or
Cs2CO3 in DMF at 50 ꢁC to form 11, 13, and 14 (Scheme 2). The
progress of the reactions was checked at several times by concur-
rent liberation of the mono-Ns-protected cyclic polyamines 12, 15,
* Corresponding author. Tel.: þ41 44 635 4245; fax: þ41 44 635 6812.
E-mail address: sbienz@oci.uzh.ch (S. Bienz).
0040-4020/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2008.05.119

7532
P. Bisegger et al. / Tetrahedron 64 (2008) 7531–7536
the resin-bound N-atom represented the problem in the cleavage
process.
The products 12, 15, and 16 were also expected to be obtained by
Mitsunobu reaction of the precursors 8, 9, and 10, followed by
cleavage of the cyclized products 11, 13, and 14 (Scheme 3), and we
were confident that the Mitsunobu reaction would be more efficient
a)
b) or
c)
NHNs
N
NHNs
NHNs
HN
N
H
X
1
2
3
X=Br
8
X= OH
d)
b) or
c)
than the direct substitution reaction of the
a-Ns-u-bromo com-
N
N
NHNs
NHNs
pounds. Intramolecular Mitsunobu reactions on solid phase were
already shown by Kung and Swayze.14 Treatment of 8 with DEAD
and PPh3 in THF, followed by the usual cleavage procedure, afforded
in fact the desired product 12 in 35.5% yield. The yield was just
slightly higher than that obtained with the previous procedure, but
the optimal reaction time of approximately 20 h for the cyclization
was found to be much more convenient. Similarly, the ring closures
of 9 and 10 under Mitsunobu conditions proceeded efficiently, de-
livering after cleavage compounds 15 and 16 in yields of 34.0 and
25.4%, respectively.
Y
Y
Z
4
Y = NHBn
Y = BnN(CH2)3NHBn
6
7
9
Y = NBn, Z = Br
5
Y = BnN(CH2)3NBn, Z = Br
Y = NBn, Z = OH
10 Y = BnN(CH2)3NBn, Z = OH
Scheme 1. (a) Merrifield resin, DIEA, 50 ꢁC. (b) 1,3-Dibromopropane, DIEA, 50 ꢁC. (c) 3-
Bromopropanol, DIEA, 50 ꢁC. (d) Benzylamine or N,N0-dibenzylpropane-1,3-diamine,
DIEA, 50 ꢁC.
and 16 from their solid support and from their benzyl protective
groups using a known procedure.13 Since the alkylation and sub-
stitution reactions with the resins usually proceed with almost
quantitative yields, except for the first alkylation at the benzylic N-
atom attached to the resin, the overall yields of the products 12, 15,
and 16 can be taken as a measure for the efficiency of the cycliza-
tion reaction.
a)
b)
N
NHNs
OH
N
HN
HN
NNs
NNs
8
11
12
a)
b)
NNs
N
NNs
N
NHNs
b)
a)
X
N
OH
N
NHNs
NHNs
HN
NNs
NNs
X
X
Br
X
9
X = NBn
13 X = NBn
15 X = NH
16 X = HN(CH2)3NH
3
11
12
10 X = BnN(CH2)3NBn
14 X = BnN(CH2)3NBn
Scheme 3. (a) PPh3, DEAD, THF, 23 ꢁC. (b) (i) ACE-Cl, DIEA, DCE; (ii) MeOH, reflux.
a)
b)
NNs
HN
N
N
NNs
Br
During all these cyclizationsdperformed with resins up to
a loading level of 0.73 mmol gꢀ1d‘cross-alkylation’ of neighboring
polyamine moieties on the solid support was not observed.
X
X
6
X = NBn
X = BnN(CH2)3NBn
13 X = NBn
15 X = NH
7
14 X = BnN(CH2)3NBn 16 X = HN(CH2)3NH
To further improve the reaction sequence, we attempted to
perform the cyclizations with the aid of an even more activating
sulfonyl group. It is known that the 2,4-dinitrobenzenesulfonyl
group can be used as an alternative for the 2-nitrobenzenesulfonyl
group in the Mitsunobu reaction, leading to higher reaction rates.15
Unfortunately, the use of this sulfonyl group proved cumbersome
because it was not possible to follow an analogous reaction se-
quence as before due to facile Smiles rearrangements. Already the
synthesis of N-(3-aminopropyl)-2,4-dinitrobenzenesulfonamide
was not possible to realize analogous to that of 1. When diamine 17
was treated with sulfonylchloride 18 in presence of NEt3 not the
desired product 19, but rather the aniline derivative 20 was
obtained. It is assumed that compound 19 was in fact formed ini-
tially, but that this product, in the form of the free amine, sub-
sequently underwent a rapid Smiles rearrangement, similarly as
that described by Kan et al.16 (Scheme 4). The hydro trifluoroacetic
salt of compound 19 was prepared17,18 and shown to undergo im-
mediate rearrangement to 20 upon treatment with base.
Scheme 2. (a) Cs2CO3, DMF, 50 ꢁC. (b) (i) ACE-Cl, DIEA, DCE; (ii) MeOH, reflux.
The optimal reaction time for the cyclizations was found to be
approximately 135 h. Performing the cyclization reaction for this
period of time, resin 3 delivered finally the eight-membered cyclic
diaza compound 12 in 33% overall yield. After five reaction steps 12-
and 16-membered cyclic polyazacompounds 15 and 16 were
obtained in overall yields of 30%. When the resins were reacted for
a markedly shorter period of time, the conversions were in-
complete, and prolonged treatment of the resins with the bases
resulted in lower yields of the subsequently liberated products. It is
interesting to mention at this point that the nature of the cation
present during cyclization (Kþ or Csþ) had no effect on the outcome
of the reactions. For both bases used, K2CO3 or Cs2CO3, almost the
same yields of the final products were obtained when the reactions
were performed under otherwise the same conditions. Thus, it can
be concluded that no template effect through chelation of the
cations is necessary or even supportive for the cyclization.
In the initial phase of the investigation we have experienced
that the yields of the liberated polyamines were constantly low at
approximately 6–8%, when the classical cleavage procedure was
applied (treatment with ACE-Cl followed by methanolysis).13 The IR
spectra of the resins recovered after the cleavage procedure
revealed that still significant amounts of material bearing a sul-
fonamide group remained on the polymer, thus, that the product
was not quantitatively liberated from the solid support. The prob-
lem of incomplete cleavage was assumed to be caused by quarter-
nization of the resin-bound N-atom, either due to alkylation or
protonation. By the addition of some DIEA to the reagent mixture,
the cleavage of the polyamines from the solid support resulted in
markedly improved yields. Thus, it was shown that protonation of
NO2
NO2
O
O
a)
S
S
+
NH2
H2N
Cl
H2N
N
O
O
NO2
NO2
H
17
19
18
–SO2
NO2
N
H2N
H
NO2
20
Scheme 4. (a) Et3N, CH2Cl2, 0 ꢁC, evaporation of CH2Cl2 at 23 ꢁC.

P. Bisegger et al. / Tetrahedron 64 (2008) 7531–7536
7533
Thus, we had to find another pathway to explore the potential of
2,4-dinitrobenenesulfonamide to be used as protecting and acti-
vating group for the Mitsunobu cyclization (Scheme 5). For this
reason we alkylated the Merrifield polymer under the usual con-
ditions with 3-aminopropanol to obtain resin 21. The possibility of
insertion of aminoalcohols on the solid support as a building block
for polyamines was described by Renault et al.19 The benzylic amine
of this substrate was alkylated with N-(3-bromopropyl)ph-
thalimide and the sulfonyl group was finally introduced by removal
of the phthalimide protecting group of 22 by treatment with
H2NNH2 to form resin 23, which then was reacted with 18 to form
resin 24. Cleavage of the polyamine derivative from the resin at this
step showed that up to this moment no Smiles rearrangement had
occurred.
cyclization to the eight-membered ring might be explained with
a ‘Thorpe–Ingold effect’ or the ‘gem dialkyl effect’21–26 provided by
the proximate resin. Due to repulsive interactions of the func-
tionalized alkyl chains with the polymeric resin framework, the
mobility of the chains and hence the entropy of the system might
be reduced.
To test whether such a ‘Thorpe–Ingold effect’ is operative, we
introduced a spacer in-between the resin and the attached poly-
amine portion. As this spacer unit, we have chosen 4-hydroxy-
benzyl alcohol to form Wang resin 29, which was then converted
to resin 30 (Scheme 6).27
OH
Cl
O
a)
a)
b)
Cl
N
H
OH
N
OH
Merrifield resin
29 (Wang resin)
NPht
21
b)
22
c)
Br
O
d)
e)
N
OH
NH2
N
OH
N
NO2
O
S
N
NH
30
O S O
O
Scheme 6. (a) 4-Hydroxybenzyl alcohol, NaOCH3. (b) PPh3Br2.
NO2
NO2
25
24
23
By Volhard titration28 of the original Merrifield resin, Wang resin
29, and the bromo resin 30, it was found that the involved sub-
stitutions proceeded quantitatively. To secure that the following
reactions are fully comparable, they were all performed in parallel
procedures. Resins 8 and 9 were prepared anew from the Merrifield
resin, and in parallel, resin 30 was converted to resins 31 and 32.
These four materials were exposed to the Mitsunobu condition.y
Treatment of the resultant cyclized products 11 and 33 or 13 and 34,
respectively, with the cleavage reagents under the usual conditions
gave finally the free cyclized products 12 and 15 as shown in
Scheme 7.
f)
NO2
Cl
NO2
HN
HN
HN
NO2
H
N
NO2
O
N
N
S
+
+
NO2
O
NO2
NO2
26
28
27
Scheme 5. (a) 3-Aminopropanol, DIEA, 50 ꢁC. (b) N-(3-Bromopropyl)phthalimide,
DIEA, 50 ꢁC. (c) N2H4$H2O, 80 ꢁC. (d) 2,4-Dinitrobenzenesulfonylchloride, 2,6-lutidine,
DCM, 23 ꢁC. (e) PPh3, DEAD, THF, 23 ꢁC. (f) (i) ACE-Cl, DCE; (ii) MeOH, reflux.
The yields of the final products 12 and 15 obtained this way (21.0
and 14.4%, respectively) were markedly lower than those obtained
with the Merrifield resin (35.5 and 24.0%, respectively). The in-
troduction of the linker resulted for both products in a decrease of
the yields by a factor of 40%. Thus, it seems that a ‘Thorpe-Ingold
effect’ is in fact operative for cyclizations performed with substrates
directly connected to the Merrifield resin. Possibly due to smaller
cyclization rates, substrates connected to the Wang resin had the
opportunity to undergo side reactions to a greater extent, which
finally resulted in the lower yields.
With resin 24 at hand, cyclization under Mitsunobu conditions
was attempted under the usual conditions, and the resultant resin
25 was cleaved by treatment with ACE-Cl in DCEdwithout the
addition of base to minimize the formation of the Smiles product
27dfollowed by methanolysis. The reaction finally delivered
compounds 26, 27, and 28, however, with the desired compound 26
as a minor component only. The low overall yield of 5.8% for
compound 26 can be explained on the one hand by the fact that we
did not use a base during the cleavage process, but also by the
production of the undesired compound 27 (6.2%) and the non-
cyclized product 28 (6.0%).
3. Conclusion
To avoid Smiles rearrangement that still might occur during the
cleavage process of the final product from the resin, standard
methods to remove the 2,4-dinitrobenenesulfonyl group from the
N-atomdtreatment of resin 25 with mercaptoacetic acid and DIEA
or with thiophenol and K2CO3dwere applied. Unfortunately both
procedures resulted in the formation of 27 as the major product
showing that the Smiles rearrangement most probably already
occurred during the Mitsunobu reaction.
We have demonstrated that cyclic polyamines can be con-
structed by solid-phase synthesis on the Merrifield resin as an
efficient alternative to ‘in-solution’ methodologies. The cyclizations
under Mitsunobu conditions showed to be considerably faster and
more convenient than the corresponding cyclizations by intra-
molecular substitution reactions of bromides. The efficiency of the
cyclizations appears to be enhanced through a Thorpe–Ingold effect
provided by the sterically demanding resin. Since the Thorpe–Ingold
effect usually needs irreversibly bound alkyl substituents, the
reversibly resin-bound alternative provides a possibility for the
efficient construction of non-substituted small- and medium-sized
rings. With the flexible route to the precursors for cyclic
2.1. ‘Thorpe–Ingold effect’
We were rather astonished that the yields of the products 12, 15,
and 16 were virtually the same, independent of the ring sizes
formed upon cyclization. Particularly the overall yield of the eight-
membered diamine derivative 12 (33–36%) is surprisingly high,
considering the strain, that is, introduced into the system upon
closure of the medium-sized ring.20 The high efficiency of the
y
The time of 20 h showed to be optimal. Shorter reaction times resulted in in-
complete cyclizations, longer reaction times of up to 80 h did not improve the
overall yield for the cyclization with and without the linker.

7534
P. Bisegger et al. / Tetrahedron 64 (2008) 7531–7536
b)
a)
Cl
N
N
HN
NHNs
OH
NNs
NNs
X
X
X
Merrifield resin
11 X = CH2
13 X = CH2NBn(CH2)2
8 X = CH2
9 X = CH2NBn(CH2)2
12 X = CH2
15 X = CH2NH(CH2)2
35.5%
24.0%
decrease in yield
by a factor of 40%
a)
b)
N
Br
N
HN
NNs
X
NHNs
NNs
X
X
OH
30
33 X = CH2
34 X = CH2NBn(CH2)2
12 X = CH2
15 X = CH2NH(CH2)2
21.0%
14.4 %
31 X = CH2
32 X = CH2NBn(CH2)2
= spacer
Scheme 7. (a) PPh3, DEAD, THF, 23 ꢁC. (b) (i) ACE-Cl, DIEA, DCE; (ii) MeOH, reflux.
compounds with varying ring sizes, numbers of N-atoms, and
numbers of methylene units in-between the amino functions,
the described method represents a versatile tool for the preparation
of cyclic polyamine derivatives. Due to non-desired Smiles
rearrangement, Mitsunobu reactions with 2,4-dinitrobenzene-
sulfonamide proved not to be applicable for the construction of
cyclic polyamines.
a stirred soln of 1,3-diaminopropane (9.4 ml, 0.113 mol) and Et3N
(5.2 ml, 37.5 mmol) in CH2Cl2 (100 ml) at 0 ꢁC. After stirring at 23 ꢁC
for 2 h, the mixture was washed with satd aq NaCl soln and the
organic phase was dried with Na2SO4. The solvent was evaporated
in vacuo, and the residue purified by chromatography (CH2Cl2/
MeOH/NH3 (25% in H2O) 9:0.5:0.05) to give 20 as a dark yellow oil
(6.10 g, 20.0 mmol, 53%). IR: 3620w, 3363s, 3109m, 2940m, 2871m,
1623s, 1587s, 1524s, 1498s, 1470m, 1424s, 1366s, 1335s, 1310s,
1279s, 1181w, 1143s, 1135s, 1102m, 1087m, 1059m, 922m, 832m,
822m, 764w, 744s, 712m, 650w. 1H NMR (CDCl3): 9.12 (d, 4J¼2.7, 1
arom. H), 9.02 (br s, NH), 8.26 (dd, 3,4J¼9.5, 2.7, 1 arom. H), 6.96 (d,
3J¼9.5, 1 arom. H), 3.54 (dt, 3J¼6.7, 5.2, 2H), 2.95 (t, 3J¼6.4, 2H), 1.91
(quint., 3J¼6.6, 2H), 1.28 (s, NH2). 13C NMR (CDCl3): 148.4 (s, 2C),
135.9 (s, 1C), 130.3, 124.4, 113.9 (3d, 3 arom. C), 42.1, 39.8, 31.5 (3t,
3C). CI-MS: 241 (100, [MþH]þ).
4. Experimental part
4.1. General
Unless otherwise stated, starting materials were obtained from
commercial suppliers and were used without further purification.
For the solid-phase reactions, an Advanced ChemTech PLS 6 Organic
Synthesizer was used. IR spectra were recorded on a Perkin–Elmer
1600 Series FT-IR spectrometer or on a Perkin–Elmer IR ‘Spectrum
4.2.2. Derivatization of Merrifield resin with N-(3-aminopropyl)-2-
One’;
n
in cmꢀ1. Solid supports: Merrifield peptide resin 200–400
nitrobenzenesulfonamide (resin 2)
mesh, 1% divinylbenzene, loading 0.73 mmol gꢀ1 or 0.32 mmol gꢀ1
Merrifield resin (24.00 g, 7.68 mmol, 0.32 mmol gꢀ1
) was
(Advanced ChemTech). Chromatography: Merck silica gel 60 (40–
63 mm) or by preparative HPLC; chromatograms were recorded
with a Dynamax solvent delivering system model SD-300 coupled
with a Dynamax absorbance detector model UV-1. Column used:
Kromasil KR100-10C18. 1H NMR spectra: in CDCl3 or MeOH-d4;
swelled in N-methyl-2-pyrrolidone (NMP, 180 ml) for 15 min. After
addition of N-(3-aminopropyl)-2-nitrobenzensulfonamide11 (18.67 g,
72.00 mmol) and ethyldiisopropylamine (DIEA, 20.5 ml, 0.120 mol),
the mixture was stirred for 24 h at 50 ꢁC. The resin was filtered
off, washed successively with NMP, CH2Cl2, and MeOH, and dried in
vacuo to give 26.48 g of resin 2. Its loading, 0.29 mmol gꢀ1 (100%),
was measured by Volhard titration.8,28 IR: 1690, 1542 (NO2), 1347
(SO2), 1167 (SO2).
Bruker AC-300 (300 MHz);
or MeOH (
4.87), J in hertz. 13C NMR spectra in CDCl3 or MeOH-d4;
Bruker ARX-300 (75.5 MHz); in parts per million rel to CDCl3
77.23) or MeOH-d4 49.15); multiplicities from DEPT-135 and
d in parts per million rel to CHCl3 (d 7.26)
d
d
(d
(d
DEPT-90 experiments. CI-MS: Finnigan MAT95 (San Jose, CA, USA);
ionization energies of 70 eV for EI and 150 eV for CI (with NH3 as
reactant gas); quasi-molecular ions and characteristic fragments in
m/z (rel %). Proof of structure and purity of the final polyamine
derivatives is provided by their 1H and 13C NMR spectra. Com-
pounds loaded on the resin are characterized by IR (only charac-
teristic signals or signals, which undergo a change in intensity are
given). Elemental analyses are not appropriate for polyamine de-
rivatives since the compounds arose, as waxy or glassy solids only,
from which the last solvent molecules can hardly be removed. The
hydrochloric salts are rather hygroscopic, and the uptake of water
falsifies the elemental analyses. Although the resins were dried
after every step in a desiccator (silica gel with indicator) overnight,
the mass of crude products from solid-phase syntheses is generally
unsuitable for estimating the real yields, because these products are
often contaminated with significant amounts of salts and solvents.
4.2.3. Alkylation of resin 2 with 1,3-dibromopropane (resin 3)
Resin 2 (1.16 mmol) was swelled in NMP (30 ml) for 15 min.
1,3-Dibromopropane (1.2 ml, 11.60 mmol) and DIEA (2.0 ml,
11.60 mmol) were added, and the mixture was agitated for 20 h at
50 ꢁC. Resin 3 was filtered off, washed with NMP and CH2Cl2, and
dried in vacuo. IR: 1542 (NO2), 1347 (SO2), 1167 (SO2).
4.2.4. Elongation of resin 3 (resin 4)
Resin 3 (1.16 mmol) was swelled in NMP (30 ml), and benzyl-
amine (1.3 ml, 11.60 mmol) and DIEA (2.0 ml, 11.60 mmol) were
added. After agitation for 20 h at 50 ꢁC, resin 4 was filtered off,
washed with NMP and CH2Cl2, and dried in vacuo. IR: 1686, 1542
(NO2), 1363 (SO2), 1297, 1155 (SO2).
4.2.5. Elongation of resin 3 (resin 5)
Analogous to Section 4.2.4, resin 3 (0.51 mmol) was treated with
N,N0-dibenzylpropane-1,3-diamine (0.78 g, 3.08 mmol) and DIEA
(0.88 ml, 5.14 mmol) for 20 h at 50 ꢁC delivering resin 5. IR: 1686,
1542 (NO2), 1362 (SO2), 1166 (SO2).
4.2. Construction of the polyamine backbones on the solid
support and building blocks for the solid-phase synthesis
4.2.1. N-(2,4-Dinitrophenyl)-1,3-diaminopropane (20)
4.2.6. Elongation of resin 4 (resin 6)
A
solution of 2,4-dinitrobenzenesulfonylchloride (10.00 g,
Analogous to Section 4.2.3, resin 4 (1.16 mmol) was treated with
1,3-dibromopropane (1.2 ml, 11.60 mmol) to give resin 6.
37.5 mmol) in CH2Cl2 (150 ml) was added dropwise over 2 h to

P. Bisegger et al. / Tetrahedron 64 (2008) 7531–7536
7535
4.2.7. Elongation of resin 5 (resin 7)
4.2.16. Modification of resin 29 with Ph3PBr2 (resin 30)
Analogous to Section 4.2.3, resin 5 (1.16 mmol) was alkylated
further with 1,3-dibromopropane (1.2 ml, 11.60 mmol) to give
resin 7.
Resin 29 (1.40 mmol) was swelled in CH2Cl2 (30 ml). After the
slow addition of Ph3PBr2 (1.76 g, 4.20 mmol) suspended in CH2Cl2
(15 ml), the mixture was agitated for 3 h at 23 ꢁC under Ar. The
resin was filtered off, washed successively with CH2Cl2 and MeOH,
and dried in vacuo. Volhard titration8,28 showed complete
conversion.
4.2.8. Alkylation of resin 2 with 3-bromopropanol (resin 8)
Resin 2 (0.51 mmol) was swelled in NMP (15 ml). 3-Bromopro-
panol (0.28 ml, 3.08 mmol) and DIEA (0.88 ml, 5.14 mmol) were
added, and the mixture was agitated for 20 h at 50 ꢁC. The resin was
filtered off, washed with NMP and CH2Cl2, and dried in vacuo. IR:
1542 (NO2), 1362 (SO2), 1167 (SO2).
4.3. Cyclization of the polyamines on the resins
4.3.1. Cyclization of resin 3 (resin 11)
Resin 3 (0.58 mmol) was swelled in dry DMF (15 ml) and Cs2CO3
(2.84 g, 8.70 mmol) was added to the reaction mixture. The sus-
pension was heated to 50 ꢁC and agitated for 135 h, the resin was
filtered off, washed successively with NMP, NMP/H2O (1:1), NMP,
and CH2Cl2, and dried in vacuo at 50 ꢁC. IR: 1685, 1542 (NO2), 1374
(SO2), 1166 (SO2).
4.2.9. Elongation of resin 4 (resin 9)
Analogous to Section 4.2.3, resin 4 (0.51 mmol) was treated
with 3-bromopropanol (0.28 ml, 3.08 mmol) and DIEA (0.88 ml,
5.14 mmol) for 20 h at 50 ꢁC delivering resin 9. IR: 3418 (OH), 1683,
1542 (NO2), 1401, 1371 (SO2), 1168 (SO2), 1112, 984.
4.2.10. Elongation of resin 5 (resin 10)
4.3.2. Cyclization of resin 6 (resin 13)
Resin 6 (0.58 mmol) was cyclized according to Section 4.3.1 to
give 13. IR: 1686, 1542 (NO2), 1371 (SO2), 1166 (SO2).
Analogous to Section 4.2.3, resin 5 (0.51 mmol) was treated with
3-bromopropanol (0.28 ml, 3.08 mmol) to give resin 10. IR: 1686,
1542 (NO2), 1401, 1363 (SO2), 1167 (SO2).
4.3.3. Cyclization of resin 7 (resin 14)
Resin 7 (0.87 mmol) was cyclized analogous to Section 4.3.1
delivering 14. IR: 1686, 1542 (NO2), 1371 (SO2), 1165 (SO2).
4.2.11. Derivatization of Merrifield resin with 3-aminopropanol
(resin 21)
Merrifield resin (20.00 g, 15.60 mmol, 0.78 mmol gꢀ1
) was
swelled in NMP (140 ml). After addition of 3-aminopropanol
(18.67 g, 72.00 mmol) and DIEA (20.5 ml, 0.120 mol), the mixture
was stirred for 24 h at 50 ꢁC. The resin was filtered off, washed
successively with NMP, CH2Cl2, and MeOH, and dried in vacuo. Vol-
hard titration8,28 revealed 100% loading. IR: 3520–3100 (OH), 1069.
4.3.4. Cyclization of resin 8 (resin 11)
The well-dried resin 8 (0.51 mmol) was swelled in 18 ml NMP
for 15 min. After addition of PPh3 (674 mg, 2.57 mmol) and DEAD
(448 mg, 2.57 mmol) the mixture was agitated for 20 h at 23 ꢁC. The
resin was filtered off, washed with CH2Cl2, NMP, and CH2Cl2, and
dried in vacuo to give resin 11.
4.2.12. Alkylation of resin 21 with N-(3-bromopropyl)phthalimide
(resin 22)
4.3.5. Cyclization of resin 9 (resin 13)
Resin 9 (0.51 mmol) was cyclized analogous to Section 4.3.4
delivering 13.
Resin 21 (1.17 mmol) was suspended in NMP (10 ml). N-(3-
Bromopropyl)phthalimide (1.88 g, 7.02 mmol) and DIEA (2.0 ml,
12.00 mmol) were added, and the mixture was agitated for 20 h at
50 ꢁC. Resin 22 was filtered off, washed with NMP, CH2Cl2, and
MeOH, and dried in vacuo. IR: 1712 (CO), 1394, 1364.
4.3.6. Cyclization of resin 10 (resin 14)
Resin 10 (0.51 mmol) was cyclized analogous to Section 4.3.4
delivering 14.
4.2.13. Removing of the phthalimide protective group from resin 22
(resin 23)
4.3.7. Cyclization of resin 24 (resin 25)
Resin 24 (0.59 mmol) was cyclized analogous to Section 4.3.4
delivering 25.
Resin 22 (0.59 mmol) was swelled in NMP (5 ml), and N2H4$H2O
(2.8 ml, 58.50 mmol) was added. The mixture was agitated for 3 h
at 80 ꢁC, the resin was filtered off, washed with NMP, dioxane, di-
oxane/H2O (1:1), dioxane, CH2Cl2, MeOH, CH2Cl2, and MeOH, and
dried in vacuo. IR: 3520–3100 (OH, NH).
4.4. Cleavage of the cyclic polyamine derivatives from the
resins
4.2.14. Alkylation of resin 23 with 2,4-dinitrobenzene-
sulfonylchloride (resin 24)
Resin 23 (0.59 mmol) was swelled in CH2Cl2 (10 ml). 2,4-Dini-
trobenzenesulfonylchloride (156 mg, 0.59 mmol) was added. After
4.4.1. 1-[(2-Nitrobenzene)sulfonyl]-1,5-diazacyclo-octane (12)
Resin 11 (0.58 mmol) was swelled in 1,2-dichloroethane (15 ml)
and 1-chloroethyl chloroformate (ACE-Cl; 1.3 ml, 11.60 mmol) was
added, followed by DIEA (0.5 ml, 2.90 mmol). After agitation for 3 h
at 23 ꢁC, the resin was filtered off and washed with CH2Cl2. The
organic solutions were combined and evaporated to dryness. The
residue was dissolved in MeOH, and the resulting solution was
refluxed for 3 h. The solvent was removed, and the residue was
purified by chromatography (CH2Cl2/MeOH/NH3 (25% in H2O)
9:2:0.05) to give 12 as the free base (58 mg, 0.19 mmol; overall yield
33% with respect to 2). IR: 3380w, 3100w, 2940s, 1595m, 1550s,
1470m, 1380s, 1345s, 1170s, 1065m, 1000m, 860m, 785m, 750m. 1H
NMR (CDCl3): 7.99–7.90 (m, 1 arom. H), 7.74–7.56 (m, 3 arom. H),
3.47–3.39 (m, 4H), 3.00–2.94 (m, 4H), 2.65 (s, NH), 1.91–1.81 (m,
4H). 13C NMR (CDCl3): 148.5 (s, CNO2), 133.6 (d, 1 arom. C), 132.5 (s,
CSO2), 131.6, 130.7, 124.2 (3d, 3 arom. C), 47.8, 46.4, 30.2 (3t, 6C). CI-
MS: 300 (100, [MþH]þ).
5 min, 2,6-lutidine (67 ml, 0.59 mmol) was added, and the mixture
was agitated for 5 h at 23 ꢁC. The resin was filtered off, washed with
CH2Cl2, DMF, CH2Cl2, MeOH, CH2Cl2, and MeOH, and dried in vacuo.
IR: 1550 (NO2), 1537 (NO2), 1348 (SO2), 1252, 1210, 1168 (SO2).
4.2.15. Derivatization of Merrifield resin with 4-hydroxybenzyl
alcohol (resin 29)
Merrifield resin (10.00 g, 2.80 mmol, 0.28 mmol gꢀ1
) was
swelled in N,N-dimethylacetamide (60 ml). After addition of 4-
hydroxybenzyl alcohol (0.45 g, 3.64 mmol) and NaOMe (200 mg,
3.73 mmol), the mixture was stirred for 24 h at 80 ꢁC. The resin was
filtered off, washed successively with DMF, dioxane, CH2Cl2, and
MeOH, and dried in vacuo. Volhard titration8,28 showed complete
loading.

7536
P. Bisegger et al. / Tetrahedron 64 (2008) 7531–7536
The analogous cleavage of resin 11 (0.51 mmol) obtained by the
Mitsunobu cyclization delivered compound 12 (54 mg, 0.18 mmol)
with an overall yield of 35.5%.
26.3 (t, 2C). CI-MS: 345 (95, [MþH]þ), 315 (78), 281 (100,
[MþHꢀSO2]þ).
4.4.4.2. Data of 27. IR: 2974m, 2359m, 1683s, 1604m, 1507m,
1333s, 1201m, 1138m, 1053w, 1033w, 910w, 832w, 720w. 1H NMR
(MeOH-d4): 8.57 (d, 4J¼2.8, 1 arom. H), 8.29 (dd, 3,4J¼9.5, 2.8, 1
arom. H), 7.42 (d, 3J¼9.5, 1 arom. H), 3.58 (t, 3J¼6.1, 4H), 3.36–3.30
(m, 4H), 2.24–2.16 (m, 4H). 13C NMR (MeOH-d4): 149.6 (s, 1C), 139.7,
139.4 (2s, 2CNO2), 128.9, 124.3, 119.9 (3d, 3 arom. C), 51.8 (t, 2C),
44.8 (t, 2C), 24.6 (t, 2C). CI-MS: 281 (100, [MþH]þ), 251 (12).
4.4.2. 1-[(2-Nitrobenzene)sulfonyl]-1,5,9-triazacyclodo-decane (15)
Analogous to Section 4.4.1, resin 13 (0.58 mmol) delivered after
purification by chromatography (CH2Cl2/MeOH/NH3 (25% in H2O)
9:1.5:0.2) 15 as the free base (62 mg, 0.17 mmol; overall yield 30%
with respect to 2). IR: 3350w, 2980m, 2880m, 1565s, 1500w, 1400s,
1370s, 1190s, 1160m, 945w, 885w. 1H NMR (CDCl3): 8.04–7.97 (m, 1
arom. H), 7.74–7.56 (m, 3 arom. H), 3.49 (t, 3J¼7.0, 4H), 2.83–2.75
(m, 4H), 2.73–2.64 (m, 4H), 2.16 (br s, 2NH), 1.78–1.65 (m, 4H), 1.55
(quint., 3J¼5.4, 2H). 13C NMR (CDCl3): 148.4 (s, CNO2), 133.6 (s,
CSO2), 133.5, 131.7, 130.8, 124.2 (4d, 4 arom. C), 45.8 (t, 2C), 45.4 (t,
2C), 44.2 (t, 2C), 28.8 (t), 26.1 (t, 2C). CI-MS: 357 (100, [MþH]þ).
The analogous cleavage of resin 13 (0.51 mmol) obtained by the
Mitsunobu cyclization delivered compound 15 (62 mg, 0.17 mmol)
with an overall yield of 34.0%.
4.4.4.3. Data of 28. Mp: 131.1–133.6 ꢁC (yellow solid after lyophy-
lization). IR (KBr): 3329m, 3097m, 3045m, 2842m, 1662s, 1624s,
1586s, 1526s, 1474m, 1422s, 1352s, 1318m, 1281s, 1270s, 1196s,
1171s, 1040s, 1104m, 1055w, 1022w, 921w, 833m, 826m, 803m,
780w, 768w, 745m, 721s, 689w, 660w. 1H NMR (MeOH-d4): 9.05 (d,
4J¼2.7,1 arom. H), 8.32 (dd, 3,4J¼9.6, 2.7,1 arom. H), 7.21 (d, 3J¼9.6,1
arom. H), 3.71–3.61 (m, 4H), 3.32–3.14 (m, 4H), 2.19–2.08 (m, 4H).
13C NMR (MeOH-d4): 149.5 (s, CSO2), 137.4 (s, CNO2), 132.2 (s,
CNO2), 131.2, 124.8, 115.6 (3d, 3 arom. C), 46.6 (t, 2C), 42.1 (t, 2C),
41.0 (t, 2C), 30.2 (t, 2C), 26.6 (t, 2C). CI-MS: 317 (25, [MþH]þ), 281
(25, [MþHꢀHCl]þ), 251 (18).
4.4.3. 1-[(2-Nitrobenzene)sulfonyl]-1,5,9,13-tetraazacyclo-
hexadecane (16)
Analogous to Section 4.4.1, resin 14 (0.87 mmol) delivered
compound 16 as the free base (109 mg, 0.26 mmol; overall yield
30% with respect to 2) after rinsing the crude mixture with CH2Cl2
and purifying the rest by chromatography (CH2Cl2/MeOH/NH3 (25%
in H2O), with the following gradient 10:2:0.1/4:1:0.1/7:3:1). IR:
3290m, 2940s, 1550s, 1470m, 1375s, 1345s, 1165s, 1130m, 1020w,
855w. 1H NMR (MeOH-d4): 8.14–8.06 (m, 1 arom. H), 7.94–7.80 (m,
3 arom. H), 3.48 (t, 3J¼7.4, 4H), 2.84–2.69 (m, 12H), 1.94–1.70 (m,
8H). 13C NMR (MeOH-d4): 149.9 (s, CNO2), 135.3 (d, 1 arom. C), 133.9
(s, CSO2), 133.2, 131.4, 125.5 (3d, 3 arom. C), 49.4 (t, 2C), 48.9 (t, 2C),
47.9 (t, 2C), 47.0 (t, 2C), 29.4 (t, 2C), 29.3 (t, 2C). CI-MS: 414 (100,
[MþH]þ), 396 (29), 384 (92), 350 (22).
Acknowledgements
We like to thank the Swiss National Science Foundation for their
generous financial support.
References and notes
1. Caravan, P.; Ellison, J. J.; McMurry, J.; Lauffer, R. B. Chem. Rev. 1999, 99, 2293–
2352.
2. Anderson, C. J.; Welch, M. J. Chem. Rev. 1999, 99, 2219–2234.
3. Guo, Z.; Sadler, P. J. Angew. Chem., Int. Ed. 1999, 38, 1512–1531.
4. Reichenbach-Klinke, R.; Koenig, B. J. Chem. Soc., Dalton Trans. 2002, 121–130.
5. Itoh, T.; Takagi, Y.; Maurakami, T.; Hiyama, Y.; Tsukube, H. J. Org. Chem. 1996, 61,
2158–2163.
The analogous cleavage of resin 14 (0.51 mmol) obtained by the
Mitsunobu cyclization delivered compound 16 (54 mg, 0.13 mmol)
with an overall yield of 25.4%.
6. Bridger, G. J.; Skerlj, R. T.; Thornton, D.; Padmanabhan, S.; Martellucci, S. A.;
Henson, G. W.; Abrams, M. J.; Yamamoto, N.; De Vreese, K.; Pauwels, R.;
De Clercq, E. J. Med. Chem. 1995, 38, 366–378.
7. Le Baccon, M.; Chuburu, F.; Toupet, L.; Handel, H.; Soibinet, M.; De´champs-
Olivier, I.; Barbier, J.-P.; Aplincourt, M. New J. Chem. 2001, 25, 1168–1174.
8. Manov, N.; Bienz, S. Tetrahedron 2001, 57, 7893–7898.
9. Manov, N.; Tzouros, M.; Chesnov, S.; Bigler, L.; Bienz, S. Helv. Chim. Acta 2002,
85, 2827–2846.
10. Manov, N.; Tzouros, M.; Bigler, L.; Bienz, S. Tetrahedron 2004, 60, 2387–2391.
11. Hidai, Y.; Kan, T.; Fukuyama, T. Chem. Pharm. Bull. 2000, 48, 1570–1576.
12. Anderson, T. F.; Strømgaard, K. Tetrahedron Lett. 2004, 45, 7929–7933.
13. Conti, P.; Demont, D.; Cals, J.; Ottenheijm, H. C. J.; Leysen, D. Tetrahedron Lett.
1997, 38, 2915–2918.
14. Kung, P.; Swayze, E. Tetrahedron Lett. 1999, 40, 5651–5654.
15. Hone, N. D.; Payne, L. J. Tetrahedron Lett. 2000, 41, 6149–6152.
16. Fukuyama, T.; Cheung, M.; Jow, C.-K.; Hidai, Y.; Kan, T. Tetrahedron Lett. 1997, 38,
5831–5834.
4.4.4. 1-[(2,4-Dinitrobenzene)sulfonyl]-1,5-diazacyclo-octane (26),
1-(2,4-dinitrobenzene)-1,5-diazacyclo-octane (27), and N-(3-
chloropropyl)-N0-(2,4-dinitrophenyl)-propane-1,3-diamine (28)
Resin 25 (0.59 mmol) was swelled in 1,2-dichloroethane (10 ml)
and 1-chloroethyl chloroformate (0.65 ml, 5.85 mmol) was added.
After agitation for 3 h at 23 ꢁC under Ar, the product resin was fil-
tered off and washed with CH2Cl2. The organic solutions were
combined and evaporated to dryness. The residue was dissolved in
MeOH, and the resulting solution was refluxed for 3 h. The solvent
was removed, and the residue was purified by HPLC (RP C-18;
solvent (MeCN/H2O/TFA 25:74.95:0.05)) to give pure 26 as the TFA
salt (15.6 mg, 0.034 mmol; overall yield 5.8% with respect to 21),
the Smiles product 27 (14.6 mg, 0.037 mmol; overall yield 6.2% with
respect to 21), and the non-cyclic compound 28 (15.1 mg,
0.035 mmol; overall yield 6.0% with respect to 21).
17. Bisegger, P. University of Zurich 2007, unpublished results.
18. Lorand, L.; Parameswaran, K. N.; Stenberg, P.; Tong, Y. S.; Velasco, P. T.; Jo¨nsson,
N. A.; Mikiver, L.; Moses, P. Biochemistry 1979, 18, 1756–1765.
19. Renault, J.; Lebranchu, M.; Lecat, A.; Uriac, P. Tetrahedron Lett. 2001, 42,
6655–6658.
4.4.4.1. Data of 26. IR (KBr): 3434m, 3102m, 3034m, 2990m,
2972m, 2926m, 2875m, 2820m, 2762m, 2565m, 2474w, 1676s,
1621s, 1606m, 1563s, 1540s, 1472m, 1464m, 1430m, 1371s, 1356s,
1197s, 1176s, 1162s, 1144m, 1020s, 1051w, 1034m, 1007m, 970w,
904m, 851w, 835m, 824w, 798m, 751s, 721m, 696m, 658w. 1H NMR
(MeOH-d4): 8.75 (s, 1 arom. H), 8.61 (d, 3J¼8.8, 1 arom. H), 8.29 (d,
3J¼8.6, 1 arom. H), 3.58–3.54 (m, 4H), 3.39–3.36 (m, 4H), 2.18–2.16
(m, 4H). 13C NMR (MeOH-d4): 151.8, 150.0 (2s, 2CNO2), 137.2 (s,
CSO2), 133.4, 127.7, 121.3 (3d, 3 arom. C), 48.6 (t, 2C), 44.9 (t, 2C),
20. Illuminati, G.; Mandolini, L. Acc. Chem. Res. 1981, 14, 95–102.
21. Allinger, N. M.; Zalkow, V. J. Org. Chem. 1960, 25, 701–704.
22. Kirby, A. J. Adv. Phys. Org. Chem. 1980, 17, 208–222.
23. DeTar, D. F.; Luthra, N. P. J. Am. Chem. Soc. 1980, 102, 4505–4512.
24. Jager, J.; Graafland, T.; Schenck, H.; Kirby, A. J.; Engberts, J. B. F. N. J. Am. Chem.
Soc. 1984, 106, 139–143.
25. Sternbach, D. D.; Rossana, D. M.; Onan, K. D. Tetrahedron Lett. 1985, 26, 591–594.
26. Beesley, R. M.; Ingold, C. K.; Thorpe, J. F. J. Chem. Soc. 1915, 107, 1080–1106.
27. Ngu, K.; Patel, D. V. Tetrahedron Lett. 1997, 38, 973–976.
28. Lu, G.-s.; Mojsov, S.; Tam, J. P.; Merrifield, R. B. J. Org. Chem. 1981, 46,
3433–3436.