Phosphazene vs. diazaphospholene PN-bond cleavage in spirocyclic cyclodiphosphazenes

Article abstract of DOI:10.1039/b717219b

Thermolysis of 2-azido-1,3,2-diazaphospholenes offers access to novel and rare spirocyclic cyclodiphosphazenes. The spectroscopic data and X-ray structure of one representative of the 2-azido-1,3,2-diazaphospholenes reveals an ionic bonding situation explaining sufficiently its rather high thermal stability. The cyclodiphosphazenes were characterised by NMR, mass spectrometry, and X-ray diffractometry. The results of ESI-FT-ICR studies demonstrate the potential of these compounds to undergo reductive elimination at a phosphazene unit via [1,4]-cycloreversion of a λ⁵-diazaphospholene ring, as well as symmetrical cleavage of the P₂N₂-unit. The unexpected inclusion of benzene in the crystal of one of the cyclodiphosphazenes was interpreted in terms of molecular recognition. Chemical reaction studies comprise the proof of double N-protonation at a phosphazene ring, and hydrolytic degradation via selective cleavage of a phosphazene P-N bond.

Full text of DOI:10.1039/b717219b

View Article Online / Journal Homepage / Table of Contents for this issue
This paper is published as part of a Dalton Transactions theme issue on:
Main group chemistry: from molecules to materials
Guest Editor Simon Aldridge
University of Oxford, UK
Published in issue 26, 2008 of Dalton Transactions
Images reproduced with permission of Daniel M. German (left) and Vadim Kessler (right)
Papers published in this issue include:
Heteroleptic metal alkoxide “oxoclusters” as molecular models for the sol–gel synthesis of
perovskite nanoparticles for bio-imaging applications
Gulaim A. Seisenbaeva, Vadim G. Kessler, Robert Pazik and Wieslaw Strek, Dalton Trans., 2008,
DOI: 10.1039/b801351a
Hydride encapsulation by molecular alkali-metal clusters
Joanna Haywood and Andrew E. H. Wheatley, Dalton Trans., 2008, DOI: 10.1039/b717563a
Manipulation of molecular aggregation and supramolecular structure using self-assembled lithium
mixed-anion complexes
J. Jacob Morris, Dugald J. MacDougall, Bruce C. Noll and Kenneth W. Henderson, Dalton Trans.,
2008, DOI: 10.1039/b719565f
Toward selective functionalisation of oligosilanes: borane-catalysed dehydrogenative coupling of
silanes with thiols
Daniel J. Harrison, David R. Edwards, Robert McDonald and Lisa Rosenberg, Dalton Trans., 2008,
DOI: 10.1039/b806270f
Visit the Dalton Transactions website for more cutting-edge inorganic and organometallic research
www.rsc.org/dalton

PAPER
www.rsc.org/dalton | Dalton Transactions
Phosphazene vs.diazaphospholene PN-bond cleavage in spirocyclic
cyclodiphosphazenes†
Sebastian Burck,b Dietrich Gudat,*a Martin Nieger,c Christoph A. Schalleyd and Torsten Weilandte
Received 7th November 2007, Accepted 25th January 2008
First published as an Advance Article on the web 28th February 2008
DOI: 10.1039/b717219b
Thermolysis of 2-azido-1,3,2-diazaphospholenes offers access to novel and rare spirocyclic
cyclodiphosphazenes. The spectroscopic data and X-ray structure of one representative of the
2-azido-1,3,2-diazaphospholenes reveals an ionic bonding situation explaining sufficiently its rather
high thermal stability. The cyclodiphosphazenes were characterised by NMR, mass spectrometry, and
X-ray diffractometry. The results of ESI-FT-ICR studies demonstrate the potential of these compounds
to undergo reductive elimination at a phosphazene unit via [1,4]-cycloreversion of a
k5-diazaphospholene ring, as well as symmetrical cleavage of the P2N2-unit. The unexpected inclusion
of benzene in the crystal of one of the cyclodiphosphazenes was interpreted in terms of molecular
recognition. Chemical reaction studies comprise the proof of double N-protonation at a phosphazene
ring, and hydrolytic degradation via selective cleavage of a phosphazene P–N bond.
phinonitrene monomer, R2PN.7 However, species of this type
Introduction
are unstable toward oligomerisation, and it was predicted that
Stable phosphazenes are composed of alternating R2P- and N-
cyclotrimerisation of transient phosphinonitrenes is thermody-
namically favoured over cyclodimerisation.8 Nonetheless, cy-
clodiphosphazenes B canbe obtainedwhenstericallybulky ligands
on the phosphorus atoms are present, and the first derivative
was prepared more than two decades ago via thermolysis or
photolysis of a diamino–azido phosphine and dimerisation of the
transient phosphinonitrene formed.7 Although some further stable
cyclodiphosphazenes became known later on,9 compounds of this
type remain scarce until today, and their chemical properties have
not been studied in depth.
moieties that form either rings with sizes between two or up to
forty, or polymers with as much as 15 000 repeating units.1 The
first reported procedure for the synthesis of such polymers started
from the cyclotriphosphazene A (Scheme 1) which was prepared by
reaction of NH4Cl and PCl5, and was the first phosphazene to be
ever made.2 This cyclic trimer whose structure was first proposed
in 1895 by Stokes3 is since then used to access polyphosphazenes
via ring-opening polymerisation in the molten state4 or in the
presence of cationic catalysts.5 Subsequent displacement of the
chlorine substituents on the polymer backbone offers access to
a broad range of inorganic polymers with adjustable properties
which make useful functional materials for various applications.6
Some time ago we reported on the synthesis of 2-amino-1,3,2-
diazaphospholenes D and provided quantum chemical evidence
for a perceptible weakening of the endocyclic PN-bonds in these
compounds10 that is in stark contrast to the observed polarisation
of exocyclic bonds in P-hydrogen-,11 P-cyclopentadienyl-,12 and
P-phosphinyl-substituted 1,3,2-diazaphospholenes.13 This effect
was explained by hyperconjugation between the lone-pair at the
exocyclic nitrogen atom and the r*-orbitals of the endocyclic PN-
bonds and resulted at the same time in relative strengthening
of the exocyclic PN-bond.10 In principle, this interaction should
facilitate [1,4]-cycloreversion of the heterocycle in D to afford a
1,4-diazabutadiene and a (transient) phosphinidene14 which has a
reputation as a highly reactive intermediate.15 By analogy to this
process, fragmentation of a five-membered ring in a cyclodiphos-
phazene C might afford new r2,k3-phosphines, or, when both
Scheme 1
The most simple building block for polyphosphazenes is
obviously not the cyclotriphosphazene A but rather the phos-
aInstitut fu¨r Anorganische Chemie, Universita¨t Stuttgart, Stuttgart, Ger-
many. E-mail: gudat@iac.uni-stuttgart.de; Fax: +49 711 685 64241;
Tel: +49 711 685 64186
16
diazaphospholene rings are cleaved, the elusive molecule P2N2
or polymeric (PN)x-materials.17
These considerations prompted us to investigate this chemistry
in more detail. We report here on the synthesis and characterisa-
tion of spirocyclic representatives of rare cyclodiphosphazenes,
and on mass spectrometric studies which were carried out to
establish the fragmentation pathway of the spirocyclic assembly.
The results are expected to provide insight if compounds of type
C may undergo selective PN-bond activation to yield new types of
phosphorus–nitrogen compounds.
bSection Organic and Inorganic Chemistry, Vrije Universiteit Amsterdam,
De Boelelaan 1083, 1081 HV, Amsterdam, the Netherlands
cLaboratory of Inorganic Chemistry, University of Helsinki, A. I. Virtasen
aukio 1, Helsinki, Finland
dInstitut fu¨r Organische Chemie, Freie Universita¨t Berlin, Berlin, Germany
eKekule` Institut fu¨r Organische Chemie und Biochemie, Universita¨t Bonn,
Bonn, Germany
† CCDC reference numbers 664556–664560. For crystallographic data in
CIF or other electronic format see DOI: 10.1039/b717219b
3478 | Dalton Trans., 2008, 3478–3485
This journal is
The Royal Society of Chemistry 2008
©

The 31P NMR signal of 2a (d 118.3) is more shielded than in
other diamino azidophosphines,18 and the 1H and 13C NMR
data are similar to those of the corresponding 2-chloro-1,3,2-
diazaphospholene, 1a.19 The IR spectrum displays an intense
band attributable to mNN at 2080 cm−1 in the typical region for
azidophosphines.20
The crystals of 2a contain isolated molecules without significant
intermolecular interactions (Fig. 1). The five-membered ring
displays a flat envelope conformation with an angle between the
PN2 and C2N2 planes of 12◦. The P1–N2/N5 bond distances in
Results and discussion
Syntheses
Thermal decomposition of azidophosphines is a useful method for
the generation of cyclodiphosphazenes (Scheme 2).7,9 The appro-
priate 2-azido-1,3,2-diazaphospholenes 2a–c were easily accessed
by salt metathesis between the 2-chloro-1,3,2-diazaphospholenes
1a–c and sodium azide in the presence of a catalytic amount
of LiCl to increase the solubility of the azide ion in THF. 31P
NMR spectra indicated that the conversion was complete after
the reaction mixtures had been stirred for 16 h at ambient
temperature. The solvents were then evaporated, and the residues
dissolved in toluene, followed by filtration of the insoluble parts.
Further purification of the products 2a–c is not necessary for
the consecutive reaction, but 2a was isolated in pure form in
91% yield after removal of the solvent, and recrystallisation from
diethylether/THF at −20 ◦C.
˚
the ring (1.682(1), 1.683(1) A) match those in 2-chloro-1,3,2-
19
˚
diazaphospholenes (1.66–1.68 A ). The P1–N11 bond to the
˚
azide substituent (1.862(1) A) is significantly longer than both
the endocyclic PN bonds in 2a and the corresponding P–N(azide)
21
˚
bonds in known acyclic derivatives (R2N)2P–N3 (1.78–1.79 A )
but compares to the bond length in a tricyclic azidophosphine
reported by Schranz et al.22 The azide substituent shows a quasi
linear arrangement with an NNN-angle of 177.8(2)◦, and the N11–
˚
˚
N12 (1.228(2) A) and N12–N13 distances (1.135(2) A) are normal.
The relation between the endocyclic P1–N2/N5 and the exocyclic
P1–N11 bonds in 2a suggests the presence of a similar n(P–N)–
r*(P–X) (X = N3) interaction as in 2-chlorodiazaphospholenes19,23
and may explain the relatively high thermal stability of 2a.
Scheme
2
Synthesis of 2-azido-1,3,2-diazaphospholenes 2a–c, cy-
clodiphosphazenes 3a–c, and 4, 5 (a: Rꢀ = 2,6-Me2C6H3, R = H; b, 4:
Rꢀ = 2,4,6-Me3C6H2, R = H; c, 5: Rꢀ = 2,6-iPr2C6H3, R = Me). Reagents
and conditions: i. NaN3, LiCl, 16 h; ii. 6 h at 110 ◦C, iii. excess HOTf,
CH2Cl2, iv. neat DMF, 2 h at 145 ◦C.
Fig. 1 Molecular structure of 2a in the crystal. Thermal ellipsoids
are drawn at 50% probability level and H atoms◦have been omitted
˚
for clarity. Selected bond lengths (A) and angles ( ): P1–N5 1.682(1),
P1–N2 1.683(1), P1–N11 1.862(1), N11–N12 1.228(2), N12–N13 1.135(2),
N2–C3 1.415(2), C3–C4 1.329(2), C4–N5 1.409(2); N5–P1–N2 89.4
(1), N5–P1–N11 103.0(1), N2–P1–N11 98.7(1), N12–N11–P1 113.0(1),
N13–N12–N11 177.8(2).
For the transformation of 2a–c into cyclophosphazenes, the
toluene solutions were heated to 110 ◦C for several hours until
the 31P NMR spectra showed that the reaction was complete.
Careful inspection of the 31P NMR spectra of the reaction mixtures
showed no signals indicating the formation of side-products
and, furthermore, mass spectra of the residues obtained after
evaporation of the solvent gave no evidence for the presence of
cyclotriphosphazenes or polymeric products. Crystallisation from
concentrated reaction mixtures finally afforded the cyclodiphosp-
hazenes 3a–c in 75–95% yield.
Characterisation of the cyclodiphosphazenes 3a–c
The 31P NMR-spectra of the cyclodiphosphazenes 3a–c display
singlets with a chemical shift of 33 ppm which is very similar
to that reported for [(iPr2N)2PN]2.7 The 1H NMR spectra of 3a,b
display AAꢀXXꢀ-type multiplets for the olefinic protons in the five-
membered heterocycles; all other data are normal.
Characterisation of the P-azido-1,3,2-diazaphospholene 2a
Suitable crystals for X-ray diffraction studies of 3a,b were grown
from cooled toluene solutions. The structure of 3a is shown in
Fig. 2. Surprisingly, 3b was found to crystallise as a 1 : 1 solvate
with one molecule of benzene (see Fig. 3) which was present as
a trace impurity in the solvent used. The molecules occupy sites
with crystallographic Ci (3a) and C2h symmetry (3b). Selected bond
lengths and angles are summarised in Table 1.
The pure 2-azido-1,3,2-diazaphospholene 2a was isolated as
colourless, air- and moisture-sensitive crystalline material and was
characterised by spectroscopic techniques and an X-ray diffraction
study. As already known for bulkily substituted diamino azi-
dophosphines, 2a is a stable and non-explosive solid,18 but should
nevertheless always be handled with suitable safety precautions.
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 3478–3485 | 3479
©

◦
˚
The molecular structures of both compounds show very sim-
ilar features. The 1,3,2-diazaphospholene rings and the central
phosphazene ring are all planar and are arranged mutually
orthogonal. The flanking aryl rings lie perpendicular to the
diazaphospholene rings and are nearly parallel to each other.
The special ring stacking motif allows the methyl groups in 2,6-
position of opposite aromatic rings to evade both each other and
the adjacent nitrogen atoms in the phosphazene ring and yields
thus minimal steric repulsion. Looking at the arrangement of the
different building blocks, it is easily envisaged that attachment of
six peripheral aryl rings to a cyclotriphosphazene unit must result
in greater crowding, suggesting that the observed preference for the
cyclodiphosphazene structure owes mainly to steric constraints.
The phosphorus nitrogen bonds in the central phosphazene
Table 1 Selected bond lengths (A) and angles ( ) of 3a,b
3a
3b
P1–N1
P1–N1#a
P1–N2
P1–N5
N2–C3
N5–C4
C3–C4b
1.642(2)
1.639(2)
1.666(2)
1.666(2)
1.405(3)
1.408(3)
1.325(3)
1.641(2)
1.642(2)
1.661(2)
1.414(2)
1.326(4)
P1–N1–P1#a
85.0(1)
95.0(1)
91.4(1)
84.9(1)
95.1(1)
91.8(1)
N1–P1–N1#a
N2–P1–N5(N2#)
a N1, N1# and P1, P1# indicate pairs of atoms related by crystallographic
Ci symmetry. b C3–C3# in the case of 3b.
˚
rings display values around 1.64 A and are by some 2 pm shorter
than the PN-bonds in the five-membered rings. All PN distances
are shorter than typical single bond lengths24 and resemble the
values reported for known tetraamino-cyclodiphosphazenes.7,9
The NC and CC bond lengths are similar as in 2a and in other
1,3,2-diazaphospholenes.11,19
The inclusion of one molecule of benzene per formula unit
of 3b results in an interesting packing structure as shown in
Fig. 3. The phosphazene nitrogen atoms N1, N1# display short
˚
intermolecular distances of 3.580(4) A to the C1B and C1B#
atoms in the benzene molecule that are interpreted as a sign of
weak C–H · · · donor hydrogen bonds25 which link phosphazene
and benzene molecules in a one-dimensional strand parallel to the
crystalline c-axis.
The inclusion of benzene molecules in the solvate 3b·C6H6 owes
presumably to both the presence of appropriately sized cavities
between molecules of 3b and the possibility to form hydrogen
bridges, which is facilitated by the increased charge density at
the nitrogen atoms of a cyclodiphosphazene.7,9 If one considers
that neither incorporation of the bulk solvent (toluene) nor the
occurrence of a similar inclusion phenomenon in crystalline 3a
was observed, it is clear that the inclusion is a highly selective
process which can be described as molecular recognition.26
Fig. 2 Molecular structure of 3a. Thermal ellipsoids are drawn at 50%
probability level and H atoms have been omitted for clarity.
Mass spectrometric studies of the cyclodiphosphazenes 3a–c
The EI mass spectra of3a–c show beside the peaks of the molecular
radical cations those of characteristic fragment ions arising from
elimination of 1,4-diazabutadienyl radicals. The fragmentation
process can be explained by cleavage of a diazabutadiene unit
under simultaneous shift of a hydrogen atom from the ortho-
methyl group of a N-aryl substituent and concomitant formation
of a fragment ion with an even electron number; the loss of a
neutral diazabutadiene is not observed as this would end up in the
formation of another radical cation and is obviously energetically
less favourable.
To obtain more detailed information on the decay of cy-
clodiphosphazenes in the gas phase, compound 3b was studied
also by ESI-FT-ICR mass spectrometry using MS2/MS3-CID and
double resonance experiments. The use of the ESI technique was
considered as a particular advantage as it allows to suppress the
formation of radical ions. The ESI mass spectrum of 3b shows a
peak at m/z 675.3 corresponding to a protonated molecular ion I
(cf . Scheme 3) and an additional peak of a fragment ion at m/z
383.2 (Fig. 4(a)).
Fig. 3 Representation of the intermolecular interactions between two
molecules of 3b and one benzene molecule in the solvate 3b·C6H6. The
carbon atoms in the aryl rings of 3b that are not directly attached to
the ring nitrogen atoms and all hydrogen atoms of 3b have been omitted
for clarity. Intermolecular distances (A) and angle ( ): N1–C1B 3.580(4),
N1–H1B 2.63; N1–H1B–C1B 176.0.
◦
˚
3480 | Dalton Trans., 2008, 3478–3485
This journal is
The Royal Society of Chemistry 2008
©

Fig. 4 ESI-MS spectra of 3b recorded in the positive ion mode sprayed
from a 50 lM methanol solution: (a) only two peaks attributable to the
protonated cation I and a fragment ion II are visible; (b) CID-spectrum
of the mass selected monocation I; (c) double resonance experiment of the
mass selected cation I with all ions with m/z = 383 being removed from the
collision cell during the fragmentation period; (d) MS3 experiment with
mass selected II which results from previous fragmentation of mass selected
I; (e) MS3 experiment with mass selected IV resulting from previous
fragmentation of mass selected I.
Scheme 3 Fragmentation pathway of N-protonated monocation I.
Additional information on the fragmentation mechanism was
secured from the CID and double resonance experiments. Upon
mass selection of the parent ion followed by collisional activation,
cation I shows two different decay pathways (Fig. 4(b) and (c)). The
first one proceeds by the loss of a 1,4-diazabutadiene III in a [1,4]-
cycloreversion and affords cation II. The second fragmentation
pathway involves cleavage of the phosphazene unit by loss of
a neutral phosphinonitrene and gives the N-monoprotonated
phosphinonitrene IV. The same ion is also obtained via further
fragmentation of II under extrusion of PN. An alternative
decay pathway of II involves the loss of a neutral fragment of
composition C2H3N2P to give the phosphorus analogue of a
carbodiimide VI (Fig. 4(d)). In contrast, decay of the protonated
monophosphazene cation IV by loss of HN/C2H2 which would
likewise give VI was not observed (Fig. 4(e)). Likewise, a second
[1,4]-cycloreversion of II resulting in the formation of a diazabu-
tadiene III and a cation HN2P2 (V) could not be detected.
are conclusively interpreted by assuming that both types of P–N
bonds are accessible for cleavage, thus giving way to two possible
decay mechanisms. The breaking of bonds in the phosphazene ring
recovers a phosphinonitrene and reverses thereby the formation of
3b. To the best of our knowledge there is no evidence for an equi-
librium between cyclo- or poly-phosphazenes and a monomeric
phosphinonitrene building block in solution. As an alternative,
the observed loss of 1,4-diazabutadiene provides evidence that
the postulated [1,4]-cycloreversion of a diazaphospholene unit
can actually take place. Fragmentation via reductive elimination
at phosphorus has not yet been reported for cyclophosphazenes
whose dominant decay process in the gas phase involves the loss
of a negatively charged exocyclic substituent.27 The previously
reported finding of a biphenyl radical anion being produced in
the reduction of P3N3Ph6 with alkali metals may be explained
by a reductive elimination pathway, but this interpretation is
questionable as the fate of the phosphorus compound in this
reaction remains unknown.28 The proposed molecular structure
of VI is supported by the fact that isolable Lewis-base adducts
These gas-phase studies reveal that breaking phosphorus ni-
trogen bonds is the favourite decay channel for protonated 3b.
Support for the initial protonation occurring at a phosphazene ni-
trogen atom comes from the fact that a stable cation of this type has
been isolated (see below). The observed fragmentation processes
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 3478–3485 | 3481
©

of such cations are known.29 Likewise, iminophosphenium ions of
type IV are known, although they are only marginally stable.18a,30
In contrast, there is no precedence for the spirocyclic skeleton of
II; if one considers that the overall mass loss during the formation
of a daughter ion VI can be rationalised by elimination of a stable
1,3,2-diazaphosphole, it may be speculated that the creation of the
fragment ions II, VI is accompanied by extensive rearrangement
processes.
N-substituted 1,3,2-diazaphospholene moieties, the 1H NMR
spectrum displays an additional doublet with a splitting by 2JPH
=
2.9 Hz that is assigned to an NH2-group. The proposed structure
of 5 is confirmed by the results of an X-ray diffraction study
that was carried out on a crystal of a toluene solvate obtained
from toluene solution at low◦temperature (Fig. 6). The O1–P1–
P1ꢀ–N1ꢀ torsional angle is 45 . One of the NH2-hydrogen atoms
(H1ꢀB) is connected by an intramolecular hydrogen bridge with the
ꢀ
˚
oxygen atom O1 (H1 B · · · O1 2.44(4) A), thus forming a twisted
Reactivity of cyclodiphosphazene 3c
six-membered ring. The P–N bonds in the 1,3,2-diazaphospholene
˚
moieties are in the range 1.68–1.70 A and are only slightly longer
That the phosphazene nitrogen atoms constitute the most basic
site in the cyclodiphosphazenes under study is demonstrated
by reaction of 3b with excess triflic acid to give the salt 4
(Scheme 2). The product was isolated after work-up in 89%
yield and characterised by NMR data and a single-crystal X-
ray diffraction study of a solvate 4·2HOTf that was obtained by
crystallisation from CH2Cl2 at −20 ◦C. The 31P NMR signal of 4
(d 5.0) exhibits an upfield shift as compared to both 3b and known
tetraaminophosphonium ions which display signals between 20
and 45 ppm.31 The crystalline solvate contains cations (Fig. 5)
that feature crystallographic C2 symmetry and are connected with
the two triflate anions via NH · · · O hydrogen bridges. The overall
disposition of the spirocyclic skeleton of the cation is similar as in
3b but the two structures differ in a lengthening of the P1–N1/N1#
bonds in the phosphazene ring by approx. 2 pm and a concomitant
shortening of the P1–N2/N5 distances in the diazaphospholene
moiety by approx. 1 pm. Since at the same time the N–C bonds are
by 2–3 pm shorter and the C–C bonds in the diazaphospholene
ring by 2 pm shorter than in 3b, the structural data point to a more
pronounced bond localisation.
than in 3a,b. The bridging N1 atom features distances of 1.565(4)
and 1.617(4) A to the adjacent phosphorus atoms of which the
one to the P1ꢀ atom bearing the amino group is clearly shorter.
˚
ꢀ
ꢀ
˚
The P1 –N1 bond in the NH2 substituent of 1.633(4) A is likewise
shorter than the corresponding bond in an NH2-substituted 1,3,2-
diazaphospholene.10 The P1–O1 bond of 1.489(3) A lies in the
˚
typical range for phosphine oxides.32
Fig. 6 Representation of the molecular structure of 5. Thermal ellipsoids
are drawn at 50% probability level and H atoms and solvent molecules
◦
˚
have been omitted for clarity. Selected bond lengths (A) and angles ( ).
O1–P1 1.489(3), N1–P1ꢀ 1.565(4), N1–P1 1.617(4), P1–N5 1.675(4), P1–N2
1.697(3), N2–C3 1.435(5), C3–C4 1.329(6), C4–N5 1.438(5), N1ꢀ–P1ꢀ
1.633(4), N1ꢀ–H1ꢀA 0.91(2), N1ꢀ–H1ꢀB 0.91(2), P1ꢀ–N5ꢀ 1.670(3), P1ꢀ–N2ꢀ
1.682(4), N2ꢀ–C3ꢀ 1.443(5), C3ꢀ–C4ꢀ 1.324(5), C4ꢀ–N5ꢀ 1.446(5); P1ꢀ–N1–P1
132.0(2), N5–P1–N2 91.5(2), N5ꢀ–P1ꢀ–N2ꢀ 91.6(2).
The reaction of 3c with water is remarkable since the preferred
reaction mode of cyclophosphazenes involves cleavage of exocyclic
bonds with conservation of the heterocyclic ring.33 A similar hy-
drolytic ring opening was, however, reported for a diazadiphosphe-
tidinium salt.34 A plausible reason for the particular reactivity of
3c lies in the combination of high ring strain in the P2N2-ring with
efficient protection of the PN-bonds in the 1,3,2-diazaphospholene
rings by the flanking 2,6-diisopropylphenyl groups.
Fig. 5 Representation of the cation in the solvate 4·2HOTf. Thermal
ellipsoids are drawn at 50% probability level and H atoms except those at
N1 and N1# to N have been omitted for clarity. N1, N1# and P1, P1#
indicate pairs of atoms in symmetry equivalent halves of the molecules
related by a symmetry operation 0.5 − x, 1.5 − y, 1 − z. Selected bond
˚
lengths (A): N1–P1 1.666(5), P1–N2 1.601(6), P1–N5 1.607(5), P1–N1#
1.657(5)
By heating of 3c in wet DMF a product 5 arising from hydrolytic
cleavage of one P–N bond in the phosphazene ring was obtained.
The 31P NMR spectrum shows two doublets of an AB-type spin
system with chemical shifts at 5.2 and 3.0 ppm and a coupling of
31.7 Hz. Besides the signals attributable to two distinguishable
Conclusions
The selective synthesis of novel and rare spirocyclic cy-
clodiphosphazenes is achieved by thermolysis of 2-azido-1,3,2-
diazaphospholenes that were obtained by salt metathesis from
3482 | Dalton Trans., 2008, 3478–3485
This journal is
The Royal Society of Chemistry 2008
©

readily available 2-chloro-1,3,2-diazaphospholenes and sodium
azide. Although the NMR and crystal structural data show no
significant differences to known monocyclic cyclodiphosphazenes,
the highly specific inclusion of benzene present as trace impurity in
the solvent in crystalline 3b is a surprising result which can be inter-
preted as an example for molecular recognition in the solid state. A
comprehensive mass spectrometric study reveals the accessibility
of both types of PN-bonds for fragmentation reactions of N-
protonated cations in the gas phase. Two observable decay patterns
include the splitting of the diphosphazene unit into monomeric
phosphinonitrenes and the rupture of the diazaphospholene rings
by [1 + 4]-cycloreversion under formal reductive elimination at
phosphorus and are both highly unusual for cyclophosphazenes.
Realising these fragmentations with neutral molecules (in the gas
phase or even in solution) rather than protonated cations would
make the cyclodiphosphazenes under study suitable precursors for
the generation for phosphinonitrenes or novel (PN)x-compounds.
Finally, the hydrolytic ring opening reaction reveals a highly un-
usual reaction mode for a cyclophosphazene which is presumably
owed to the sterical protection of the peripheral P–N bonds by the
bulky N-aryl substituents.
which were filtered of and dried in vacuum. Yield: 1.22 g (91%),
mp 126 C; H NMR (C6D6): d 6.94 (s, 6 H, m/p-CH), 5.60 (d,
◦
1
3
1
2 H, JPH = 1.9 Hz, N–CH), 2.33 (s, br, 12 H, o-CH3); 13C{ H}
2
NMR (C6D6): d 138.3 (d, JPC = 12.6 Hz, ipso-C), 129.1 (s, br,
o-C), 128.3 (s, p-C), 127.8 (d, 4JPC = 2.3 Hz, p-C), 117.9 (d, 2JPC
=
7.6 Hz, N–CH), 18.8 (s, o-CH3), 18.7 (s, o-CH3); 31P{ H} NMR
(C6D6): d 118.3 (s); MS (EI, 70 eV; 405 K) m/z (%) = 327.1 ([M]+,
5.4), 295.1 ([M − N3]+, 100.0), 249.1 ([M − C3H10N3]+, 14.1). Calc.
for C18H20N5P: C 64.08, H 5.98, N 20.76. Found: C 64.24, H 6.04,
N 20.19%; IR (KBr, Nujol): m˜/cm−1 = 2081 (s).
1
General procedure for the synthesis of 3a–c. 10 mmol of
the appropriate 2-chloro-1,3,2-diazaphospholene 1a–c, 10 mmol
(0.65 g) sodium azide, and 1 mmol (0.04 g) lithium chloride
were dissolved in THF (100 mL) and stirred for 12 h at room
temperature. The reaction mixture was then evaporated to dryness,
the residue dissolved in toluene (100 mL) and the insoluble parts
filtered off. The solution was then heated to 110 ◦C for 2 h. After
cooling to room temperature the solution was concentrated to a
volume of 15 mL and stored at −20 ◦C for crystallisation. The
colourless crystals formed were filtered off and dried in vacuum.
1,4,8,11-Tetrakis(2,6-dimethylphenyl)-1,4,6,8,11,12-hexaaza-
5k5,7k5-diphosphadispiro[4.1.4.1]dodeca-2,9-diene (3a). Yield:
2.31 g (75%); mp 313 ◦C; 1H NMR (CDCl3): d 6.95–6.70 (m, 12 H,
Experimental
1
m/p-CH), 5.62 (m, 4 H, N–CH), 2.20 (s, 24 H, o-CH3); 13C{ H}
General considerations
2/4
NMR (CDCl3): d 139.0 (dd,
J
= 1.2 Hz, ipso-C), 135.6 (dd,
PC
3/5
J
J
= 1.7 Hz, o-C), 128.6 (s, m-CH), 128.2 (s, p-CH), 115.4 (dd,
All manipulations were carried out under protective gas at-
mosphere (argon) and in flame dried glassware. Solvents were
dried prior to use by common procedures. 1a–c were synthesised
according to reported procedures.19 All other chemicals were
purchased from commercial suppliers. NMR Spectra: Bruker
Avance 400 (1H: 400.13 MHz, 31P: 161.9 MHz, 13C: 100.4 MHz) at
PC
= 7.6 Hz, N–CH), 19.8 (s, o-CH3); 31P{ H} NMR (CDCl3):
3/4
1
PC
d 33.3 (s); MS: (EI, 70 eV, 430 K): m/z (%) = 618.3 ([M]+, 100.0),
571.2 ([M − C3H11]+, 33.5), 474.2 ([M − C10H10N]+, 15.0), 355.1
([M − C18H19N2]+, 82.9), 300.0 ([M − C18H19N3P]+, 11.6), 295.1
([M − C18H20N4P]+, 15.8), 262.2 ([M − C18H20N4P2]+, 32.9). Calc.
for C36H40N6P2: C 69.89, H 6.52, N 13.58. Found: C 69.93, H
6.44, N 13.12%.
◦
30 C; chemical shifts referred to external TMS (1H, 13C), 85%
H3PO4 (N = 40.480747 MHz, 31P); positive signs of chemical
shifts denote shifts to lower frequencies; coupling constants
are given as absolute values; prefixes ipso-, o-, m-, p- denote
atoms of aryl-substituents. ESI mass spectra were recorded on a
Bruker APEX IV Fourier-transform ion cyclotron resonance (FT-
ICR) mass spectrometer with an Apollo electrospray ion source
equipped with an off-axis 70◦ spray needle. Analyte solutions were
introduced into the ion source with a syringe pump (Cole–Parmers
Instruments, Series 74900) at flow rates of ca. 3 − 4 lL/min. For
CID experiments, the monoisotopic parent ions of interest were
mass-selected in the FT-ICR cell and then collided with argon
in order to increase the internal energy of the ions and provoke
their decomposition. EI-MS: Varian MAT 711, 70 eV. Elemental
analysis: Perkin Elmer 2400CHSN/O Analyser. Melting points
were determined in sealed capillaries.
1,4,8,11-Tetramesityl-1,4,6,8,11,12-hexaaza-5k5,7k5-diphospha-
dispiro[4.1.4.1]dodeca-2,9-diene (3b). Yield: 3.22 g (95%); mp
315 ◦C; 1H NMR (CDCl3): d 6.56 (s, 8 H, M–CH), 5.92 (m, 4 H,
1
N–CH), 2.30 (s, 12 H, p-CH3), 2.17 (s, 24 H, o-CH3); 13C{ H}
NMR (CDCl3): d 133.3 (pseudo-t, 4/6JPC = 1.2 Hz, m-CH), 131.4
(pseudo-t, 3/5JPC = 0.8 Hz, o-C), 127.8 (pseudo-t, 2/4JPC = 1.7 Hz,
ipso-C), 123.5 (s, p-C), 110.0 (pseudo-t, 3/5JPC = 7.8 Hz, N–CH),
1
16.1 (s, p-CH3), 13.3 (s, o-CH3); 31P{ H} NMR (CDCl3): d 33.4
(s); MS: (EI, 70 eV, 450 K) m/z (%) = 674.3 ([M]+, 100.0), 383.1
([M − C20H23N2]+, 95.2). Calc. for C40H48N6P2: C 71.20, H 7.17,
N 12.45. Found: C 71.01, H 7.04, N 12.28%.
1,4,8,11-Tetra(2,6-diisopropylphenyl)-2,3,9,10-tetramethyl-
1,4,6,8,11,12-hexaaza-5k5,7k5-diphospha-dispiro[4.1.4.1]-dodeca-
2,9-diene (3c). Yield: 4.18 g (93%); mp 298 ◦C; 1H NMR (C6D6):
d 7.01 (s, 8 H, M–CH), 6.98 (s, 4 H, p-CH), 3.34 (sep, 4 H, 3JHH
=
Syntheses
3
6.8 Hz, CH), 1.18 (d, 12 H, JHH = 6.8 Hz, CH3), 1.17 (d, 12 H,
1
2-Azido-1,3-bis(2,6-dimethylphenyl)-1,3,2-diazaphospholene (2a).
4 mmol (1.32 g) 1a, 4 mmol (0.26 g) sodium azide and 0.1 mmol
(4 mg) lithium chloride were dissolved in THF (50 mL) and stirred
for 24 h at room temperature. The reaction mixture was then
evaporated to dryness, the residue taken up in toluene (50 mL) and
the inorganic solids filtered off. The solution was again evaporated
to dryness and the residue dissolved in a mixture of Et2O (15 mL)
and THF (10 mL). Storage at −20 ◦C afforded colourless crystals
3JHH = 6.8 Hz, CH3), 1.12 (s, 12 H, CH3); 13C{ H} NMR (C6D6): d
149.2 (pseudo-t, 2/4JPC = 1.3 Hz, ipso-C), 135.2 (pseudo-t, 3/5JPC
=
1.6 Hz, o-C), 129.3 (s, o-C), 128.5 (s, m-CH), 125.6 (s, m-CH),
2/4
124.5 (s, p-CH), 118.9 (pseudo-t,
J
= 7.0 Hz, N–C), 28.6 (s,
PC
CH), 24.8 (s, CH3), 24.0 (s, CH3), 12.2 (pseudo-t, 3/5JPC = 2.1 Hz,
N–CCH3); 31P{ H} NMR (C6D6): d 33.7 (s); MS: (EI, 70 eV, 450
1
K) m/z (%) = 898.6 ([M]+, 4.8), 495.3 ([M − C28H39N2]+, 10.3),
452.3 ([M − C28H40N4P + OH]+, 100.0), 435.4 ([M − C28H40N4P]+,
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 3478–3485 | 3483
©

12.3), 361.3 ([M − C25H33N4P2]+, 43.2), 202.2 ([M − C42H60N5P2]+,
81.5). Calc. for C56H60N6P2: C 74.80, H 8.97, N 9.35. Found: C
74.86, H 9.05, N 8.98%.
2a. Colourless crystals, C18H20N5P, M = 337.4, crystal size
0.45 × 0.40 × 0.35 mm, monoclinic, space group P21/n (no. 14):
◦
˚
a = 12.7383(8), b = 10.8003(8), c = 12.8025(6) A, b = 90.415(4) ,
3
−3
˚
V = 1761.3(2) A , Z = 4, Dc = 1.272 Mg m , F(000) = 712, l =
1,4,8,11-Tetramesityl-1,4,6,8,11,12-hexaaza-5,7-diphosphonia-
dispiro[4.1.4.1]dodeca-2,6,9-triene bis(trifluormethanesulfonate)
(4). A solution of◦1 mmol (674 mg) of 3b in toluene (30 mL)
was cooled to −78 C. Triflic acid (2 mmol, 300 mg) was added
dropwise. The solution was stirred for additional 15 min and
then allowed to warm to ambient temperature. The solvent was
evaporated in vacuum and the residue dissolved in 30 mL of
CH2Cl2. Storage of the solution at −20 ◦C produced colourless
crystals which were filtered off and dried in vacuum. Yield
0.165 mm−1, 16649 reflections (2hmax = 55◦), 4028 unique (Rint
=
0.048), 221 parameters, R1 (I > 2r(I)) = 0.040, wR2 (all data) =
−3
˚
0.108, largest diff. peak/hole 0.212/−0.318 e A .
3a. Colourless crystals, C36H40N6P2, M = 618.7, crystal size
0.25 × 0.15 × 0.06 mm, monoclinic, space group P21/n (no. 14):
◦
˚
a = 8.3779(3), b = 17.0975(7), c = 11.4017(6) A, b = 96.728(2) ,
3
−3
˚
V = 1622.0(1) A , Z = 2, Dc = 1.267 Mg m , F(000) = 656, l =
0.170 mm−1, 9486 reflections (2hmax = 50◦), 2855 unique (Rint
=
1
0.054), 203 parameters, R1 (I > 2r(I)) = 0.040, wR2 (all data) =
869 mg (89%); H NMR (CDCl3): d 7.13 (s, br, 2H, NH), 6.68
−3
˚
0.090, largest diff. peak and hole 0.275/−0.387 e A .
(s, br, 8H, M–CH), 6.08 (s, br, 4 H, N–CH), 2.29 (s, br, 12 H,
1
p-CH3), 2.04 (s, br, 24 H, o-CH3); 13C{ H} NMR (CDCl3): d
3b·C6H6. Yellowish crystals, C46H54N6P2 (C40H48N6P2·C6H6),
M = 752.9, crystal size 0.30 × 0.15 × 0.05 mm, orthorhombic,
space group Cmca (no. 64): a = 22.6662(8), b = 14.4954(5), c =
141.1 (pseudo-t, 2/4JPC = 1.0 Hz, ipso-C), 137.1 (pseudo-t, 3/5JPC
=
=
1
1.7 Hz, o-C), 130.4 (s, m-CH), 126.7 (s, p-C), 119.5 (q, JCF
2/4
317.8 Hz, CF3), 118.4 (pseudo-t,
J
= 11.8 Hz, N–CH), 21.6
PC
3
−3
˚
˚
12.2951(4) A, V = 4039.6(2) A , Z = 4, Dc = 1.238 Mg m ,
F(000) = 1608, l = 0.149 mm−1, 13927 reflections (2hmax = 50◦),
1829 unique (Rint = 0.067), 130 parameters, R1 (I > 2r(I)) = 0.039,
1
(s, p-CH3), 17.9 (s, o-CH3); 31P{ H} NMR (CDCl3): d 5.0 (s). Due
to the highly hygroscopic nature no elemental analysis could be
obtained.
−3
˚
wR2 (all data) = 0.098, largest diff. peak/hole 0.287/−0.354 e A .
1,3-Bis(2,6-diisopropylphenyl)-4,5-dimethyl-2-[1,3-bis(2,6-
4. Colourless crystals, C44H52F12N6O12P2S4 ([C40H50N6P2]2+-
2[CF3SO3]−·2CF3SO3H), M = 1275.1, crystal size 0.30 × 0.15 ×
0.10 mm, monoclinic, space group C2/c (no. 15), a = 31.662(2),
diisopropylphenyl)-4,5-dimethyl-2-oxo-2k5-1,3,2-diazaphosphol-
enylimino]-2k5-1,3,2-diazaphospholenylamine
(5).
5
mmol
(450 mg) of 3c were dissolved in DMF (15 mL) and heated to
150 ◦C for 8 h. After cooling to room temperature all volatiles
were removed in vacuum and the remaining solid was taken up
in toluene (5 mL). Storage at −20 ◦C afforded colourless crystals
◦
˚
b = 12.3344(13), c = 16.2491(17) A, b = 108.536(3) , V =
3
−3
˚
6016.6(10) A , Z = 4, Dc = 1.408 Mg m , F(000) = 2624, l =
0.307 mm−1, 13155 reflections (2hmax = 50◦), 5235 unique (Rint
=
0.061), 368 parameters, R1 (I > 2r(I)) = 0.096, wR2 (all data) =
that were filtered off and dried in vacuum. Yield: 437 mg (95%);
−3
˚
0.309, largest diff. peak/hole 1.458/−0.721 e A .
◦
1
3
mp 324 C; H NMR (C6D6): d 4.18 (sep, 1 H; JHH = 6.7 Hz,
CH), 4.06 (sep, 1 H; 3JHH = 6.7 Hz, CH), 3.96 (sep, 1 H; 3JHH
=
5·2C6H5CH3. Colourless crystals, C70H98N6OP2 (C56H82N6-
6.8 Hz, CH), 3.82 (sep, 1 H; 3JHH = 6.9 Hz, CH), 3.38 (sep, 1 H;
3JHH = 6.7 Hz, CH), 3.35 (sep, 1 H; 3JHH = 6.7 Hz, CH), 3.08 (sep,
1 H; 3JHH = 6.9 Hz, CH), 2.70 (sep, 1 H; 3JHH = 6.7 Hz, CH), 2.29
(d, 2 H, 2JPH = 2.9 Hz, NH2), 1.62 (d, 6 H, 3JHH = 6.6 Hz, CH3)
OP2·2C6H5CH3), M = 1101.5, crystal size 0.20 × 0.15 × 0.10 mm,
¯
triclinic, space group P1 (no. 2), a = 13.279(3), b = 14.542(3), c =
◦
˚
17.751(4) A, a = 102.26(3), b = 103.09(3), c = 99.46(3) , V =
3
−3
˚
3180.2(12) A , Z = 2, Dc = 1.150 Mg m , F(000) = 1196, l =
0.115 mm−1, 37376 reflections (2hmax = 50◦), 11069 unique (Rint
=
3
3
1.59 (d, 6 H, JHH = 6.7 Hz, CH3), 1.58 (d, 6 H, JHH = 6.7 Hz,
3
3
0.103), 692 parameters, 92 restraints, R1 (I > 2r(I)) = 0.080, wR2
CH3), 1.50 (d, 6 H, JHH = 6.7 Hz, CH3), 1.37 (d, 6 H, JHH
=
−3
3
˚
(all data) = 0.219, largest diff. peak/hole 0.475/−0.474 e A .
6.9 Hz, CH3), 1.30 (d, 6 H, JHH = 6.8 Hz, CH3), 1.23 (d, 6 H,
3JHH = 6.9 Hz, CH3), 1.13 (d, 6 H, 3JHH = 6.7 Hz, CH3), 1.00 (dd,
4
6
3
CCDC reference numbers 664556 (2a), 664557 (3a), 664558 (3b),
664559 (4) and 664560 (5).
For crystallographic data in CIF or other electronic format see
DOI: 10.1039/b717219b
3 H, JPP = 6.8 Hz, JPP = 1.6 Hz, CH3), 0.95 (dd, 3 H, JHH
=
=
6.4 Hz, 6JPP = 5.1 Hz, CH3), 0.84 (dd, 3 H, 3JHH = 6.6 Hz, 6JPH
1.4 Hz, CH3), 0.50 (d, 3 H, 3JHH = 6.7 Hz, CH3); 31P{ H} NMR
1
2
2
(C6D6): d 5.3 (d, JPP = 31.7 Hz, PO), 2.0 (d, JPP = 31.7 Hz,
PNH2); MS (EI, 70 eV, 590 K) m/z (%) = 916.6 ([M]+, 100.0),
512.2 ([M − C28H40N2]+, 17.2), 202.1 ([M − C28H42N4P2O]+, 7.0).
Calc. for C56H82N6P2O·C4H8O: C 72.84, H 9.17, N 8.49. Found:
C 72.74, H 9.04, N 8.74%.
Notes and references
1 (a) R. T. Oakley, S. J. Rettig, N. L. Paddock and J. Trotter, J. Am. Chem.
Soc., 1985, 107, 6923; (b) H. R. Allcock and R. L. Kugel, J. Am. Chem.
Soc., 1965, 87, 4216.
2 (a) J. H. Gladstone and J. D. Holmes, J. Chem. Soc., 1864, 17, 225;
(b) J. H. Gladstone and J. D. Holmes, Ann. Chim. Phys., 1864, 3, 465.
3 H. N. Stokes, Chem. Ber., 1895, 28, 437.
4 H. R. Allcock and R. L. Kugel, J. Am. Chem. Soc., 1965, 87, 4216.
5 R. De Jaeger and M. Gleria, Prog. Polym. Sci., 1998, 23, 179.
6 (a) J. E. Mark, H. R. Allcock and R. West, Inorganic Polymers,
Prentice-Hall, New Jersey, 1992; (b) H. R. Allcock, Chemistry and
Applications of Polyphosphazenes, Wiley, Hoboken, 2003; (c) M. Gleria,
R. De Jaeger, Phosphazenes - A World Wide Insight, Nova Science,
New York, 2004; (d) M. Gleria, R. De Jaeger, Applicative Aspects
of Poly(organophosphazenes), Nova Science, New York, 2004; (e) M.
Gleria, R. De Jaeger, Top. Curr. Chem., 2005, 250, 165.
Crystal structure determinations
All single-crystal X-ray diffraction studies were carried out on
a Nonius Kappa-CCD diffractometer at 123(2) K using Mo-
35
˚
Ka radiation (k = 0.71073 A). Direct methods (SHELXS-97 )
were used for structure solution and refinement (SHELXL-97,36
full-matrix least squares on F2). No absorption corrections were
applied, and H atoms were localized by difference electron density
determination and refined using a riding model (except the NH2
group in 5, in which H(N) were refined free).
3484 | Dalton Trans., 2008, 3478–3485
This journal is
The Royal Society of Chemistry 2008
©

7 A. Baceiredo, G. Bertrand, J.-P. Majoral, G. Sicard, J. Jaud and J. Galy,
J. Am. Chem. Soc., 1984, 106, 6088.
20 (a) O. J. Scherer and N. Kuhn, Chem. Ber., 1974, 107, 2123; (b) I. Yu.
Budilova, N. I. Gusar’ and Yu. G. Gololobov, J. Gen. Chem. USSR
(Engl. Transl.), 1984, 54, 1771; (c) R. Kalgutkar, A. S. Ionkin, L. D.
Quin and P. M. Lahti, Tetrahedron Lett., 1994, 35, 3889.
21 (a) U. Englert, P. Paetzold and E. Eversheim, Z. Kristallogr., 1993, 208,
307; (b) A. H. Cowley, F. B. Gabbai, G. Bertrand, C. J. Carrano and
M. R. Bond, J. Organomet. Chem., 1995, 493, 95.
8 G. Trinquier, J. Am. Chem. Soc., 1986, 108, 568.
9 (a) R. Reau, A. Baceiredo and G. Bertrand, Phosphorus Sulfur Silicon
Relat. Elem., 1994, 93–94, 1; (b) J. Tirree´, D. Gudat, M. Nieger and E.
Niecke, Angew. Chem., Int. Ed., 2001, 40, 3025; (c) R. J. Wehmschulte,
M. A. Khan and S. I. Hossain, Inorg. Chem., 2001, 40, 2756.
10 S. Burck, D. Gudat, F. Lissner Na¨ttinen, M. Nieger and T. Schleid,
Z. Anorg. Allg. Chem., 2005, 631, 2738.
22 I. Schranz, L. P. Grocholl, L. Stahl, R. J. Staples and A. Johnson, Inorg.
Chem., 2000, 39, 3037.
11 (a) D. Gudat, A. Haghverdi and M. Nieger, Angew. Chem., Int. Ed.,
2000, 39, 3084; (b) S. Burck, D. Gudat, M. Nieger and W.-W. du Mont,
J. Am. Chem. Soc., 2006, 128, 3946.
12 S. Burck, D. Gudat, M. Nieger and J. Tirree´, Dalton Trans., 2007, 1891.
13 (a) S. Burck, D. Gudat and M. Nieger, Angew. Chem., Int. Ed., 2004,
43, 4801; (b) S. Burck, D. Fo¨rster and D. Gudat, Chem. Commun., 2006,
2810; (c) S. Burck, D. Gudat and M. Nieger, Angew. Chem., Int. Ed.,
2007, 46, 2919.
14 Z. Benko˝, D. Gudat and L. Nyulaszi, Chem.–Eur. J., 2008, 14, 902.
15 (a) U. Schmidt, Angew. Chem., Int. Ed. Engl., 1975, 14, 523; (b) F.
Mathey, in Multiple Bonds and Low Coordination Chemistry in Phos-
phorus Chemistry, ed. M. Regitz and O. J. Scherer, Thieme, Stuttgart,
1990, p. 33.
16 (a) O. Kwon, P. M. Almond and M. L. McKee, J. Phys. Chem. A, 2002,
106, 6864; (b) H. Schno¨ckel, T. Mehner, H. S. Plitt and S. Schunck,
J. Am. Chem. Soc., 1989, 111, 4578.
17 (a) Y. Furukawa, O. Mikami, M. Okamura and Y. Hirota, Proc.
Electrochem. Soc., 1986, 86, 34; (b) Y. Hirota and T. Kobayashi, J. Appl.
Phys., 1982, 53, 5037; (c) O. Mikami, M. Okamura, E. Yamaguchi, Y.
Hirota and Y. Furukawa, Jpn. J. Appl. Phys., 1984, 23, 1408; (d) Y. H.
Jeong, S. Takagi, F. Arai and T. Sugano, J. Appl. Phys., 1987, 62,
2370; (e) Y. Matsumoto, T. Hanajiri, T. Sugano, Le. T. T. Tuyen and T.
Katoda, Thin Solid Films, 1995, 269, 4146.
18 (a) G. Alcaraz, A. Baceiredo, F. Dahan and G. Bertrand, J. Am. Chem.
Soc., 1994, 116, 1225; (b) G. Sicard, A. Baceiredo, G. Bertrand and J.-P.
Majoral, Angew. Chem., Int. Ed. Engl., 1984, 23, 459; (c) A. Baceiredo,
G. Bertrand, J.-P. Majoral, F. El Anba and G. Manuel, J. Am. Chem.
Soc., 1985, 107, 3945; (d) J. Tejeda, R. Reau, F. Dahan and G. Bertrand,
J. Am. Chem. Soc., 1993, 115, 7880.
23 D. Gudat, A. Haghverdi, H. Hupfer and M. Nieger, Chem.–Eur. J.,
2000, 3414.
24 B. F. Allen, O. Kennard, D. G. Watson, L. Brammer, A. G. Orpen and R.
Taylor, in International Tables for Crystallography, ed. A. J. C. Wilson,
Vol. C, Kluwer Academic Publishers, Dordrecht, 1992, 685.
25 G. R. Desiraju, Acc. Chem. Res., 2002, 35, 565.
26 (a) D. J. Cram, Science, 1988, 240, 760; (b) J. D. Dunitz and A.
Gavezzotti, Angew. Chem., Int. Ed., 2005, 44, 1766; (c) F. Toda, Adv.
Supramol. Chem., 1992, 2, 141.
27 (a) S. Daolio, P. Traldi, E. Vecchi and M. Gleria, Org. Mass Spectrom.,
1985, 20, 492; (b) M. Gleria, S. Daolio, A. M. Maccioni, E. Vecchi
and P. Traldi, Org. Mass Spectrom., 1985, 20, 686; (c) M. Gleria, G.
Audisio, S. Daolio, P. Traldi and E. Vecchi, Macromolecules, 1984, 17,
1230; (d) M. Gleria, G. Audisio, P. Traldi, S. Daolio and E. Vecchi,
J. Chem. Soc., Chem. Commun., 1983, 1380.
28 D. Chapman, S. H. Glarum and A. G. Massey, J. Chem. Soc., 1963,
3140.
29 M. Bla¨ttner, M. Nieger, A. Ruban, W. W. Schoeller and E. Niecke,
Angew. Chem., 2000, 112, 2876.
30 M. Mulliez, J.-P. Majoral and G. Bertrand, J. Chem. Soc., Chem.
Commun., 1984, 284.
31 G. N. Koidan, A. P. Marchenko, A. A. Kudryatsev and A. M. Pinchuk,
J. Gen. Chem. USSR, 1982, 52, 1779.
32 Result of a query in the CSD on P–O bond length in phosphine oxides
˚
of type (R2N)3PO: 1.474 0.016 A.
33 (a) H. R. Allcock and T. J. Fuller, J. Am. Chem. Soc., 1981, 103, 2250;
(b) D. G. Gabler and J. F. Haw, Inorg. Chem., 1990, 29, 4018.
34 H. Tho¨nnessen, P. G. Jones and R. Schmutzler, Z. Anorg. Allg. Chem.,
2003, 629, 1265.
19 S. Burck, D. Gudat, K. Na¨ttinen, M. Nieger, M. Niemeyer and D.
Schmid, Eur. J. Inorg. Chem., 2007, 5112.
35 G. M. Sheldrick, Acta Crystallogr., Sect. A, 1990, 46, 467.
36 G. M. Sheldrick, University of Go¨ttingen, 1997.
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 3478–3485 | 3485
©