21
F. Sournies et al./Journal of Molecular Structure 412 (1997) 19–26
chemical content of the loop, at 81.0 MHz, in
CDCl3 with H3PO4 85% as the standard),
for the cis-ANSA structure, a false singlet
centred on 23–24 ppm (depending on the che-
mical content of the arch, under the same
experimental conditions as above),
for the trans-ANSA structure, a doublet
centred on 21–22 ppm (PClNH units) and a
triplet centred on 23–24 ppm (PCl2 unit)
(depending on the chemical content of the
arch, same experimental conditions as above),
and for the mono-BINO structure, a doublet
centred on 21–22 ppm (PCl2 units) and a tri-
plet centred on 19–21 ppm (PClNH unit)
(depending on the chemical content of the
bridge, same experimental conditions as
above).
soluble in n-hexane whereas the expected phosphorus
chemical is not.) The final product is the mono-
dangling entity, N3P3Cl5 [HN–(CH2)5–OH] (micro-
crystalline white powder, 6.71 g, 81% yield), as
demonstrated by a combined use of 31P NMR (doublet
centred on 21.3 ppm and triplet centred on 18.6 ppm,
2JPP = 46.1 Hz) and of DCI/NH3 mass spectrometry
(molecular peak m/z 362 with a satellite distribution
confirming the presence of 5 chlorine atoms in the
molecule).
•
•
In other words, the length of the methylenic chain
in the difunctional reagent investigated here is such
that the dangling intermediate is stabilized and does
not generate further macrocycle-bearing entities.
•
4. The case of the 2-amino 1-ethanol, H2N–
(CH2)2–OH
Therefore, there is no real difficulty in knowing if
the dangling moiety (doublet centred on 21–22 ppm
(PCl2 units) and a triplet centred on 19–21 ppm
(PClNH unit) depending on the chemical content of
the tentacle, same experimental conditions as above)
has generated a spiro, a cis-ansa and/or a trans-ansa
configuration. Conversely, the assignment of a ‘‘21–
22 ppm doublet plus a 19–21 ppm triplet’’ to either a
dangling or to a bino architecture will be critical and
will obviously need the help of a technique such as
mass spectrometry where both the molecular peak m/z
values and their isotopic distributions (5 or 10 chlor-
ine atoms) will allow the unambiguous choice of the
actual structure.
1.22 g (0.020 mole) of 2-amino 1-ethanol in 20 ml
of THF were added dropwise over 1 h to a mixture of
6.96 g (0.020 mole) of N3P3Cl6 and 4.04 g (0.040
mole) of Et3N in 150 ml of THF. The reaction takes
6 h. Hydrochloride was then filtered off and the sol-
vent removed in vacuo. The residue is recrystallized
from dry CH2Cl2, yielding crystalline samples (5.1 g,
76% yield). A combined use of 31P NMR (singlet at
23.9 ppm) and of DCI/NH3 mass spectrometry (mole-
cular peak m/z 337 with a satellite distribution indicat-
ing the presence of 4 chlorine atoms in the molecule)
would have been in favour of an cis-ansa chemical,
according to the NMR patterns we detailed above.
Actually, X-ray crystallography proved that the so-
called ansa derivative was its spiro isomer, N3P3Cl4
[HN–(CH2)2–O].
Now we must consider what happens for each of the
amino alcohols we investigated.
3. The case of the 5-amino 1-pentanol, H2N–
(CH2)5–OH
4.1. Crystal data—structure determination and
refinement
2.06 g (0.020 mole) of 5-amino 1-pentanol in 100
ml of Et2O were added dropwise to a mixture of 6.96 g
(0.020 mole) of N3P3Cl6 and 2.02 g (0.020 mole) of
Et3N in 300 ml of Et2O. The medium was stirred for
15 min at room temperature.
Hydrochloride was then filtered off and the solvent
removed in vacuo at 30ЊC. The residue was washed
twice with 20 ml of n-hexane in order to eliminate
traces of unreacted N3P3Cl6. (N3P3Cl6 is highly
A perspective view of this loop-bearing cyclo-
phosphazene is shown in Fig. 2. The compound
crystallized in the orthorhombic Pna21 (group 33)
space group. Unit cell parameters at 298 K were
˚
˚
as follows: a = 11.683 (1) A, b = 8.107 (2) A and c
3
−3
˚
˚
= 12.437 (1) A; V = 1178 (1) A , rx = 1.76 g cm , rexp
= 1.74 g cm−3, Z = 4. Direct methods were used to
determine the structure. Isotropic refinement gave a
reliability index of 0.045. Introduction of anisotropic