2736
Q. Wang et al. / Tetrahedron Letters 42 (2001) 2735–2737
appear attractive as catalysts. Among various possible
spiro aza crowns, the one consisting of two 1,5,9-triaza-
cyclododecane rings is of particular interest since it has
previously been shown to exhibit catalytic activity supe-
rior to that of other monocyclic aza crowns.10 We now
report the synthesis of this compound, 2,6,10,14,18,22-
hexaazaspiro[11.11]tricosane, as the first example of the
preparation of spiro aza crowns.
butoxyformic anhydride ((Boc)2O) in a mixture of THF
and aqueous sodium hydroxide. The protected com-
pound, which according to LC/ESI-MS contained four
Boc groups, was purified on a silica gel column using a
1:1 mixture (v/v) of dichloromethane and ethyl acetate
(Rf=0.58) as eluent. The Boc protecting groups were
finally removed with aqueous hydrogen chloride (1 mol
L−1, 2 h, 40°C), giving pure 1 as the hexahydrochloride
in 32% yield (from 6).21 The free base form was
obtained by passing the hydrochloride through an ion
exchange column (Dowex lx2, OH− form). The oily
The synthetic route is outlined in Scheme 1. 2-Phenyl-
5,5-bis(hydroxymethyl)-1,3-dioxane 2, prepared from
pentaerythritol as described previously,11 was converted
into 2-phenyl-5,5-bis(toluene-4-sulfonyloxymethyl)-1,3-
dioxane 3 with toluene-4-sulfonyl chloride in pyridine.
The product was purified by column chromatography,
and the benzylidene protection was removed by cata-
lytic hydrogenation on Pd/C in ethanol, giving 2,2-bis-
(toluene-4-sulfonyloxymethyl)propane-1,3-diol12 4 as a
solid foam in virtually quantitative yield. The reason
for the removal of the benzylidene protection was that
1
product was characterized by H and 13C NMR spec-
troscopy, LC/ESI-MS and high resolution mass
spectrometry.22
The spiro carbon of 1 resonates at l 41.45 in NaOD/
D2O (at l 38.41 for the hydrochloride in D2O). It is
worth noting that the four methylene carbons bonded
to the spiro carbon (C1, C11, C13, C23) exhibit only
one resonance signal (l 52.38 in NaOD/D2O and l
52.74 for the hydrochloride in D2O), while the methyl-
ene groups of the fragments ꢀHN(CH2)3NHꢀ clearly
show two distinct set of signals in the 13C NMR
spectrum. In NaOD/D2O, one set of the carbon reso-
nances appears at l 32.52, 39.61 and 47.16, each of the
signals being coupled with a single methylene group,
the corresponding proton resonances occur at l 1.52,
2.55 and 2.50, respectively. The other set is found at l
25.58, 48.50 and 49.50, the respective methylene proton
resonance being at l 1.61, 2.77 and 2.69. The protons
are coupled to each other within each set, the coupling
constants being 5.8 and 4.2 Hz, respectively. Similar
double signals for these three methylene carbons are
observed on recording the spectrum of the hydrochlo-
ride of 1 in pure D2O. However, in this case, the proton
resonance signals of the methylene groups overlap
extensively. The reason for the observed multiplicity of
the signals of the ꢀHN(CH2)3NHꢀ fragments remains
obscure.
4
proved to be more soluble than 3 in 1,2-
dimethoxyethane (DME), which was used as a solvent
in the subsequent nucleophilic displacement of the tosyl-
oxy groups.
The well established13–16 displacement of tosyloxy
groups with amine nucleophiles was applied in a step-
wise manner to form the two triaza macrocycle moie-
ties. Accordingly, the two tosyloxy groups of 4 were
first displaced with a 1,5,7-triazabicyclo[4.4.0]dec-5-ene
(TBD) in DME.17–19 The reaction took place almost
quantitatively in 24 h at 40°C when TBD was used in a
twofold excess. The progress of the reaction was fol-
lowed by appearance of a LC/ESI-MS signal at m/z 240
[M]+, referring to formation of the 2,2-bis(hydroxy-
methyl)-1H,4H,7H-pentahydro-3a,6a,9a-triazaphena-
lenium cationic salt 5. Crude 5 was then tosylated at
low temperature with toluene–4-sulfonyl chloride in a
mixture of pyridine and dichloromethane using 4-di-
methylaminopyridine (DMAP) as a catalyst. Reduc-
tion with sodium borohydride in THF and purification
by column chromatography (silica gel, 5–10% MeOH in
DCM) then gave 2,2-bis(toluene-4-sulfonyloxymethyl)-
1H,4H,7H-hexahydro-3a,6a,9a-triazaphenalene20 6 in
33% overall yield (from 2).
Acknowledgements
The authors wish to thank Dr. Kari Neuvonen for his
skilful technical help and fruitful discussions concerning
the NMR spectroscopy of the compounds prepared.
Financial aid from the Academy of Finland is grate-
fully acknowledged.
The
second 1H,4H,7H-hexahydro-3a,6a,9a-triaza-
phenalene ring was introduced similarly by displace-
ment of the tosyloxy groups with TBD and subsequent
hydride reduction. The formation of spiro
bis(1H,4H,7H-hexahydro-3a,6a,9a-triazaphenalene)
7
was followed by the appearance of a LC/ESI-MS signal
at m/z 347 [M+H]+. Crude 7 was then hydrolyzed in
aqueous hydrogen chloride (6 mol L−1) to the desired
2,6,10,14,18,22-hexaazaspiro[11.11]tricosane 1, which
took one week at refluxing temperature. However, the
progress of the hydrolysis was followed by the appear-
ance of a LC/ESI-MS signal at m/z 327 [M+H]+. The
purification of 1 proved difficult. The pure product was
obtained by protecting the amino functions with tert-
butoxycarbonyl (Boc) groups and purifying the Boc
protected compound on a silica gel column. The protec-
tion was carried out by treating crude 1 with tert-
References
1. Weber, E. J. Org. Chem. 1982, 47, 3478–3486.
2. Ouchi, M.; Inoue, Y.; Sakamoto, H.; Yamahira, A.;
Yoshinaga, M.; Hakushi, T. J. Org. Chem. 1983, 48,
3168–3173.
3. Bashkin, J. Chem. Rev. 1998, 98, 939–960.
4. Koike, T.; Kimura, E. J. Am. Chem. Soc. 1991, 113,
8935–8941.
5. Kuusela, S.; Lo¨nnberg, H. J. Phys. Org. Chem. 1993, 6,
347–356.
Wang, Qi
