Structure and conformation of the medium-sized chlorophosphazene rings

Article abstract of DOI:10.1021/ic500272b

Medium-sized cyclic oligomeric phosphazenes [PCl₂N]_m (where m = 5-9) that were prepared from the reaction of PCl₅ and NH₄Cl in refluxing chlorobenzene have been isolated by a combination of sublimation/extraction and column chromatography from the predominant products [PCl₂N]₃ and [PCl₂N]₄. The medium-sized rings [PCl₂N]_m have been characterized by electrospray ionization-mass spectroscopy (ESI-MS), their ³¹P chemical shifts have been reassigned, and their T₁ relaxation times have been obtained. Crystallographic data has been recollected for [PCl ₂N]₅, and the crystal structures of [PCl ₂N]₆, and [PCl₂N]₈ are reported. Halogen-bonding interactions were observed in all the crystal structures of cyclic [PCl₂N]_m (m = 3-5, 6, 8). The crystal structures of [P(OPh)₂N]₇ and [P(OPh)₂N]₈, which are derivatives of the respective [PCl₂N]_m, are also reported. Comparisons of the intermolecular forces and torsion angles of [PCl₂N]₈ and [P(OPh)₂N]₈ with those of three other octameric rings are described. The comparisons show that chlorophosphazenes should not be considered prototypical, in terms of solid-state structure, because of the strong influence of halogen bonding.

Full text of DOI:10.1021/ic500272b

Article
pubs.acs.org/IC
Structure and Conformation of the Medium-Sized
Chlorophosphazene Rings
David J. Bowers, Brian D. Wright, Vincenzo Scionti, Anthony Schultz, Matthew J. Panzner, Eric B. Twum,
Lin-Lin Li, Bryan C. Katzenmeyer, Benjamin S. Thome, Peter L. Rinaldi, Chrys Wesdemiotis,
Wiley J. Youngs, and Claire A. Tessier*
University of Akron, Department of Chemistry, Akron, Ohio 44325-3601, United States
S
* Supporting Information
ABSTRACT: Medium-sized cyclic oligomeric phosphazenes
[PCl2N]m (where m = 5−9) that were prepared from the
reaction of PCl5 and NH4Cl in refluxing chlorobenzene have
been isolated by a combination of sublimation/extraction and
column chromatography from the predominant products
[PCl2N]3 and [PCl2N]4. The medium-sized rings [PCl2N]m
have been characterized by electrospray ionization−mass
spectroscopy (ESI-MS), their 31P chemical shifts have been
reassigned, and their T1 relaxation times have been obtained.
Crystallographic data has been recollected for [PCl2N]5, and
the crystal structures of [PCl2N]6, and [PCl2N]8 are reported.
Halogen-bonding interactions were observed in all the crystal
structures of cyclic [PCl2N]m (m = 3−5, 6, 8). The crystal
structures of [P(OPh)2N]7 and [P(OPh)2N]8, which are
derivatives of the respective [PCl2N]m, are also reported. Comparisons of the intermolecular forces and torsion angles of
[PCl2N]8 and [P(OPh)2N]8 with those of three other octameric rings are described. The comparisons show that
chlorophosphazenes should not be considered prototypical, in terms of solid-state structure, because of the strong influence of
halogen bonding.
Scheme 1. Classic Synthesis of a Mixture of Cyclic and
Linear Chlorophosphazenes
INTRODUCTION
■
a
Polyphosphazenes encompass the largest class of inorganic
backbone oligomers and polymers, and applications for these
systems continue to accumulate.1,2 By way of substitution
reactions, chlorophosphazenes are utilized as precursors to
most of the hundreds of other phosphazene compounds. The
array of promising applications that exist for phosphazene
oligomers and polymers is attributed to the wide range of
substituents that can be placed on the P−N backbone.1,2
Therefore, chlorophosphazene chemistry is central to the
phosphazene field. Irreproducibility and other difficulties
associated with the synthesis and storage of chlorophospha-
zenes, especially high polymeric [PCl2N]n, have limited
commercial use of the entire class of polyphosphazenes.
Thus, fundamental research of chlorophosphazenes is essential
to overcome these problems.
Classically, cyclic chlorophosphazenes of the formula
[PCl2N]m (where m = oligomer = <20) are synthesized via
the reaction of PCl5 and NH4Cl in the presence of a high
boiling (∼132−145 °C) halogenated solvent (Scheme 1).3
Such reactions have been known since the 19th century4 and
give a mixture of cyclic and linear species.5 The cyclic species
[PCl2N]3 and [PCl2N]4 are the major products from Scheme 1.
These rings are important precursors to high polymeric
[PCl2N]n, because of their ability to undergo thermal ring-
a
Data taken from ref 3. The species that adds to the cationic
intermediates has been suggested to be Cl3PNH.
opening polymerization at elevated temperatures.6 The
medium-sized cyclic oligomeric phosphazenes [PCl2N]m, with
m ≥ 5, often have been foreshadowed by the lower-molecular-
weight homologues and are usually considered byproducts of
Scheme 1.
As part of a reinvestigation of Scheme 1, herein we report a
study of the medium-sized cyclic oligomeric chlorophospha-
zenes [PCl2N]m, with m = 5−9. There were several reasons why
we chose to investigate the medium-sized [PCl2N]m rings. An
Received: February 3, 2014
Published: August 11, 2014
© 2014 American Chemical Society
8874
dx.doi.org/10.1021/ic500272b | Inorg. Chem. 2014, 53, 8874−8886

Inorganic Chemistry
Article
early motivation concerned certain NMR parameters. There
was confusion in the literature with respect to the 31P NMR
chemical shift of several [PCl2N]m. In addition, quantification
of the products from Scheme 1 often had been done via
integration of 31P NMR signals. The underlying assumption
behind the use of integration is that the relaxation time T1 is
similar for all products. As part of our reinvestigation of Scheme
1, it was important to correctly identify the products from
Scheme 1 and determine their relative quantities.
suggest the presence of halogen bonding. Considering these
precedents, we wondered whether halogen bonding might play
a more general role in chlorophosphazene chemistry, including
its role in determining the conformations of rings and
polymers.
EXPERIMENTAL SECTION
■
General Procedures. All glassware was dried in an oven overnight
(∼115 °C) before use and reaction apparati were assembled hot. The
glassware used for the experiments was made with virtually greaseless
Fisher-Porter Solv-seal glass joints and high-vacuum valves on the
flasks were purchased from Kimble-Kontes.
This work on medium-sized [PCl2N]m rings also addresses
basic questions of structure and conformation and considers
how cyclic phosphazenes compare to other cyclic molecules
such as hydrocarbons, polysilanes, and polysiloxanes. Cyclic
phosphazenes are typically drawn with alternating single and
double P−N bonds. However, the bonds usually are all an
intermediate length, between those expected for single and
double bonds, and large P−N−P angles are observed. Two
bonding descriptions have been advanced for the multiple bond
character: one is based on P−N dπ−pπ bonding7 and, more
recently, the other is a combination of an ionic attraction
between charges on the ring coupled with hyperconjugation
effects.8 Either bonding description is rather different than
those applied to rings such as [CR2]2m and [SiR2]2m.9 More
similar to phosphazene rings are the various bonding
descriptions that have been applied to the isoelectronic siloxane
rings.10−13 In addition, steric considerations play an important
role in the conformations of rings. Unlike [CR2]2m and [SiR2]2m
that have a substituent on every atom of the ring, [SiMe2O]m
and [PCl2N]m have functional groups at every other atom of
the ring, thereby leading to different backbone alignments. For
these reasons, we were interested in determining the
conformations of the ring bonds of medium-sized [PCl2N]m.
Although the crystal structures and the configuration of the
P−N backbones of [PMe2N]m rings with m values in the range
3−12 have been described,14 the picture was far less complete
for the important [PCl2N]m rings, from which the [PMe2N]m
rings were derived. Only the crystal structures of [PCl2N]3,15
Materials. Chlorobenzene, hexanes, and THF (Fisher Scientific)
were dried and deoxygenated by alumina and copper columns in the
Pure Solv solvent system (Innovative Technologies, Inc.). Heptane
(ACS reagent grade), dodecanol (Fisher Scientific), pentane, aniline,
pyridine, PCl5, potassium hydride (Sigma−Aldrich), and CDCl3
(Cambridge Isotopes, 99.8%) were used as received. NH4Cl (Fisher
Scientific) was crushed with a mortar and pestle. Phenol was
recrystallized from petroleum ether and dried in vacuo. Thin layer
chromatography (TLC) was conducted on flexible sheets (Baker-flex)
precoated with SiO2 (IB2-F). Column chromatography was conducted
using or SiO2 (60−200 mesh) from Sorbent Technologies.
NMR Spectroscopy. Routine NMR spectra were obtained using a
Varian Gemini 300 MHz or VNMRS 500 MHz instruments at 25 °C.
31P NMR spectra were referenced to an external reference of 0.15 M
1
H3PO4 solution in a deuterated solvent (0 ppm). H NMR and 13C
NMR spectra were referenced to the residual protons (7.24 ppm) and
the 13C atoms (77.23 ppm) in the CDCl3 solvent. Integrations of 31P
NMR spectra were performed on a Varian Model VNMRS 500 MHz
instrument with a relaxation delay (d1) of 70 s. The 31P T1 experiment
was performed with the inversion−recovery (relaxation-180°−τ−90°-
acquire) technique. The experiment was performed with a spectral
window of 26 kHz, 20 s relaxation delay, 1.6 s acquisition time, and a
1
13.6 μs (90°) pulse width, with gated H (WALTZ-16, γBH/2π = 2.9
kHz) decoupling. A total of 17 τ delays ranging from 0.05 to 50.0 s
(which were randomized) were used. Data were fitted to an
exponential function using a three-parameter fit. The quantitative
31P NMR spectrum was obtained with an 13.7 kHz spectral window
with inverse-gated 1H decoupling (to suppress NOEs in order to retain
quantitative accuracy) using WALTZ-16 modulation (γBH/2π = 2.9
kHz). For the purpose of quantitation, a 70.0 s relaxation delay
(corresponding to 5 times the longest measured T1), a 1.6 s acquisition
time, and a 13.6 μs (90°) pulse width were used. The data were zero-
filled to 128 K, and line broadening was set to 0.5 Hz before Fourier
transformation.
17
[PCl2N]4,16 and [PCl2N]5 had been obtained. The unit cells
of the hexameric and octameric [PCl2N]m rings were known for
many years (as early as 1897!),18,19 but their crystal structures
or that of [PCl2N]7 had not been reported. As West and co-
workers have done for polysilanes,20 we expected that
comparisons among the structures of the medium-sized cyclic
[PCl2N]m rings and the [PCl2N]n high polymer would provide
important insights. The medium-sized rings should be better
models for the structure of the high polymer [PCl2N]n21,22 than
the smaller [PCl2N]m, because of the marked increase in
flexibility and the lower ring strain as the cyclics become larger.
By comparing the structures of the [PCl2N]m medium-sized
rings with rings in which the chlorides are substituted with
organic groups, we hoped to obtain insights into the structures
of the respective polymers. Another motivation to study the
structures of [PCl2N]m concerned halogen bonding. Halogen
bonds are intermolecular interactions that are due to
asymmetric electron distributions on halogen atoms and have
been compared to hydrogen bonds.23 The asymmetric electron
distribution creates a σ-hole so that the Cl atoms of the P−Cl
bonds are capable of acting as acceptors of lone pairs on
neighboring molecules. Recent computational studies indicate
the existence of halogen bonds between [PCl2N]3 and a variety
of Lewis basic substrates.24 Although we did not discuss this
point in our publication,25 there are short contacts in the crystal
structures of Group 13 Lewis acid adducts of [PCl2N]3, which
ESI-MS. ESI spectra were acquired in both positive and negative
mode, to detect the positive or negative ions, respectively, generated at
the source. The mass spectrometer utilized in this study was a
SYNAPT HDMS hybrid quadrupole/time-of-flight (Q/oa-ToF) mass
spectrometer (Waters, Beverly, MA) equipped with a Z-spray
electrospray source operated in both positive- and negative-ion
mode. The instrumental settings were tuned in order to optimize
both signal and resolution as follows: capillary voltage, 3.5 kV; cone
voltage, 35−40 V; and sampling cone voltage, 3−4 V. The source
temperature and the desolvation gas temperature were held to 90−60
°C and 240−110 °C, respectively. The sample solution was
electrosprayed at a flow rate of 15 μL/min. The data acquisition,
data processing, and theoretical isotope distribution generation were
performed using Waters’ MassLynx 4.1 software.
X-ray Crystallography. Crystal structure datasets were collected
on a Bruker SMART Apex or Bruker ApexII Duo diffractometers.
Unit-cell determination was achieved by using reflections from three
different orientations. An empirical absorption correction and other
corrections were done using multiscan SADABS. Structure solution,
refinement, and modeling were accomplished using the Bruker
SHELXTL package.26 The structures were obtained by full-matrix
least-squares refinement of F2 and the selection of appropriate atoms
from the generated difference map. X-ray wavelength, crystal data, and
8875
dx.doi.org/10.1021/ic500272b | Inorg. Chem. 2014, 53, 8874−8886

Inorganic Chemistry
Article
structure refinement for [PCl2N]6, [P(OPh)2N]7, [PCl2N]8, and
[P(OPh)2N]8 are given in Table 1, and data for [PCl2N]3, both forms
of [PCl2N]4, and [PCl2N]5 are given in the Supporting Information.
Synthesis and Isolation of a Crude Mixture of the Medium-
Sized Rings [PCl2N]m (m = 5−9). A mixture of cyclic and linear
chlorophosphazenes was synthesized in a chlorobenzene solution at
10%−20% scale of a literature procedure.27 In a representative
reaction, NH4Cl (0.672 g, 0.0126 mol), PCl5 (2.09 g, 0.0100 mol), and
dry chlorobenzene solvent (∼15 mL) was added to a 100-mL Schlenk
flask containing a 1-in. stir bar. A reflux condenser topped with a
drying tube containing Drierite was connected to the flask. The flask
was placed on a heating mantle and refluxed for ∼12 h. [Caution: This
literature procedure should be carefully monitored to ensure that
sublimation of PCl5 does not plug the narrower parts of the apparatus
and result in pressure buildup from the gaseous HCl product. Such pressure
buildup can lead to an explosion.] The pale yellow solution was filtered
from a solid, presumably unreacted NH4Cl. The volatile components
were removed under reduced pressure at 0.1 Torr resulting in a yellow
slurry. At this point and by 31P NMR, the yields of individual rings for
a representative reaction were as follows: [PCl2N]3, 64%; [PCl2N]4,
5%; [PCl2N]5, 7%; [PCl2N]6, 6%; [PCl2N]7, 3%; and [PCl2N]8,9, 2%.
The remainder of the mixture (∼13%) was a mixture of linear
oligomeric chlorophosphazenes. Individual reactions gave variable total
yields of cyclics [PCl2N]m (m = 3−9) in the range of 60%−92%.
[PCl2N]3 and most of the [PCl2N]4 were separated from the yellow
slurry by sublimation at 60 °C for 6 days with a vacuum line whose
ultimate vacuum was ∼10−4 Torr. The unsublimed material was
extracted with hexane in air to separate the soluble medium-sized rings
[PCl2N]m from the insoluble linear oligomers. The crude medium-
sized rings, which, at this point, still contain small amounts of
[PCl2N]4 and several higher-molecular-weight species, were collected
from ∼25 such reactions (total of ∼3 g), combined, and stored in air
(maximum exposure of ∼6 months).
Chromatographic Isolation of the Medium-Sized Rings
[PCl2N]m (m = 5−9) in Air. The combined crude medium-sized
rings (see above) were extracted with ∼50 mL heptane; the solution
was filtered, and the volatiles were removed under vacuum to give 2.5
g of a slightly brown oily solid. A concentrated heptane (∼5 mL)
solution of the oily solid was placed on a silica gel column (61 cm in
length, 4 cm in diameter). Chromatography of the mixture on silica gel
with heptane eluent resulted in the isolation of each cyclic species.
Thin-layer chromatography was used to monitor the separations and
determine Rf values (for [PCl2N]4, Rf = 0.44; for [PCl2N]5, Rf = 0.38;
for [PCl2N]6, Rf = 0.36; for [PCl2N]7, Rf = 0.32; for [PCl2N]8, Rf =
0.24; and for [PCl2N]9, Rf = 0.22) and visualization of the cyclic
oligomers was achieved by staining with a 2:1 (v/v) aniline−pyridine
mixture.28 A total of 1.308 g of medium-sized rings was isolated.
Yields, crystallization procedures (in air unless otherwise indicated),
and characterizational data for each medium-sized ring [PCl2N]m (m =
5−9) are given below. By collecting further heptane fractions, several
other higher-molecular-weight oligomeric species (m values up to ∼17,
as determined by mass spectrometry) were also collected (total of
0.067 g). A toluene flush was collected and an additional mixture of
chlorophosphazenes was isolated (0.02 g). Full characterization of the
individual products in the last heptane fractions and the toluene
fraction was not possible, because of the small quantity of material.
There was a yellow band of material that never moved more than
approximately one-third of the distance down the column, and
eventually the band turned green.
solution of [PCl2N]6 (0.100 g, 0.144 mmol), HCl (measured in a
known volume on the vacuum line: 55 mmHg, 295.75 K, 0.050 L, 0.15
mmol) and SbCl5 (0.019 mL, 0.148 mmol) in CHCl3 (10 mL) was
prepared and stirred for 12 h. After removal of the volatile components
on a vacuum line whose ultimate vacuum was ∼10−4 Torr, a colorless
solid was obtained. A 1H NMR spectrum in CDCl3 taken immediately
after the solid was isolated indicated protonation of the ring (δ NH =
6.13 ppm). After storage in the glovebox for 1 week, recrystallization
by slow removal of CHCl3 in vacuo at 0 °C yielded colorless crystals of
monoclinic [PCl2N]6 among a brown gelatinous solid. Colorless solid;
RF = 0.36, 31P NMR (500 MHz, CDCl3) −15.3 ppm (s); ESI-MS
−
(THF): for [PCl2N]6 , m/z 689.485; for [P6Cl11N6O]−, m/z 670.516;
and for [P6Cl11N6]−, m/z 654.535.
[PCl2N]7. Colorless liquid (0.322 g, 13%); RF = 0.32; 31P NMR (500
−
MHz, CDCl3) −17.0 ppm (s); ESI-MS (THF): for [PCl2N]7 , m/z
804.390; for [P7Cl13N7O]−, m/z 785.424; and for [P7Cl13N7]−, m/z
769.435.
[PCl2N]8. Crystals suitable for X-ray crystallography were obtained
by recrystallization from heptane. Colorless solid (0.140 g, 6%); RF =
0.24; 31P NMR (500 MHz, CDCl3) −17.7 ppm (s); ESI-MS (THF):
−
for [PCl2N]8 , m/z 919.307; for [P8Cl15N8O]−, m/z 900.314; and for
[P8Cl15N8]−, m/z 884.327.
[PCl2N]9. Colorless liquid (0.047 g, 2%); RF = 0.22; 31P NMR (500
−
MHz, CDCl3) −17.7 ppm (s); ESI-MS (THF): for [PCl2N]9 , m/z
1034.239; for [P9Cl17N9O]−, m/z 1015.247; and for [P9Cl17N9]−, m/z
999.250.
Synthesis of [P(OPh)2N]7. In the glovebox, KH (0.015 g, 0.37
mmol) and 10 mL of dry THF were added to a 100-mL Schlenk flask
containing a 1-in. stir bar. [PCl2N]7 (0.022 g, 0.19 mmol) and phenol
(0.036 g, 0.38 mmol) were placed in separate vials and each was
dissolved in 5 mL of dry THF. Using standard Schlenk techniques, the
phenol solution was added to the Schlenk flask under nitrogen. The
solution immediately evolved a gas, presumably H2, and became pink.
When gas evolution ceased, the [PCl2N]7 solution was added under
nitrogen. The flask was equipped with a condenser topped with a
bubbler and the contents were heated to reflux for ∼24 h. The pink
solution was filtered in air and the volatile components were removed
under reduced pressure at ∼0.1 Torr. The product was washed with
water (20 mL × 4) to remove unreacted KOPh and phenol. The
product was dried in vacuo, giving [P(OPh)2N]7 as a viscous colorless
oil. Yield: 86%. Crystals suitable for X-ray diffraction (XRD) were
obtained by adding a 50:50 heptane/n-dodecanol mixture to the oil.
The heptane was evaporated and the room temperature was cooled
below the melting point of dodecanol, leaving a small amount of a
colorless solid that was manually removed from the solid dodecanol.
31P NMR (CDCl3) δ −18.8 ppm (s). 1H NMR (CDCl3) δ 6.88 (d, 28
H) 6.96−7.02 (m, 42 H). 13C NMR (CDCl3) δ (ppm): 121.5(s),
123.9(s), 129.1(s), 152.0(s).
Synthesis of [P(OPh)2N]8. [P(OPh)2N]8 was synthesized utilizing
the same procedure as for [P(OPh)2N]7, but using 0.020 g of
[PCl2N]8 (0.022 mmol), 0.014 g of KH (0.35 mmol), and 0.034 g of
phenol (0.36 mmol). [P(OPh)2N]8 was obtained as colorless crystals.
Crystals suitable for XRD were obtained by recrystallization from
chloroform. Yield: 90%. Mp: 109−111 °C. 31P NMR (CDCl3) δ
(ppm): −20.1(s). 1H NMR (CDCl3) δ 6.83 (d, 32 H) 6.91−6.98 (m,
48 H). 13C NMR (CDCl3) δ (ppm): 121.2(s), 123.8(s), 129.0(s),
152.0(s).
Synthesis of [P(OPh)2N]9. [P(OPh)2N]9 was synthesized using
the same procedure as for [P(OPh)2N]7, but using 0.030 g of
[PCl2N]9 (0.029 mmol), 0.060 g of KH (1.50 mmol), and 0.034 g of
phenol (0.36 mmol). Yield: 87%. 31P NMR (CDCl3) δ (ppm):
[PCl2N]5. Crystals suitable for X-ray crystallography were obtained
by recrystallization from hexane. Colorless solid (0.407 g, 16%), RF =
0.38; 31P NMR (500 MHz, CDCl3) −15.1 ppm (s); ESI-MS (THF):
1
−21.2(s). H NMR (CDCl3) δ 6.79 (d, 36 H) 6.85−6.95 (m, 54 H).
13C NMR (CDCl3) δ (ppm): 121.2(s), 123.4(s), 128.7(s), 151.7(s).
−
for [PCl2N]5 , m/z 574.568; for [P5Cl9N5O]−, m/z 555.615; and for
[P5Cl9N5]−, m/z 539.624.
[PCl2N]6. Evaporation of the volatile components from the fraction
with RF = 0.36 gave a colorless solid (0.392 g, 16%). Triclinic crystals
suitable for X-ray crystallography were obtained by recrystallization
from hexane or CHCl3. Monoclinic crystals suitable for X-ray
crystallography were obtained during an attempt to isolate a derivative
by protonation of the ring. Using standard anaerobic techniques, a
RESULTS AND DISCUSSION
■
Synthesis and Spectral Characterization of Medium-
Sized Ring [PCl2N]m. Mixtures of oligomeric [PCl2N]m and
+
[Cl3PN(PCl2N)rPCl3 ]X− (X = Cl or PCl6) were prepared
from the reaction of PCl5 and NH4Cl in refluxing
8876
dx.doi.org/10.1021/ic500272b | Inorg. Chem. 2014, 53, 8874−8886

Inorganic Chemistry
Article
8877
dx.doi.org/10.1021/ic500272b | Inorg. Chem. 2014, 53, 8874−8886

Inorganic Chemistry
Article
chlorobenzene (Scheme 1). The following separation methods
were chosen such that the temperature did not exceed ∼100
°C, because it has been suggested that ring−ring conversions,
such as interconversion of [PCl2N]3 and [PCl2N]6, can occur
above 100 °C.5a,29 After [PCl2N]3 and [PCl2N]4 were largely
removed by sublimation at 60 °C under vacuum, the residue
was extracted with hexane in order to isolate the mixed
+
[PCl2N]m rings from the insoluble [Cl3PN(PCl2N)rPCl3 ]X−.
At this point, the mixture of [PCl2N]m rings can be stored in
dry air for long periods. Whereas sublimation and recrystalliza-
tion are popular techniques for purification of [PCl2N]3 and
[PCl2N]4, they are less useful in separating the higher [PCl2N]m
rings from each other. We decided to test reports that state that
the separation of the higher [PCl2N]m rings can be done via
thin-layer chromatography (TLC) on silica (hydrocarbon
elutants),30 despite a report that suggests that surface silanol
groups could displace Cl atoms in the rings, leading to covalent
binding to the surface of the silica.31 We found that both TLC
and column chromatography on silica with pure heptane eluent
of the mixture of higher [PCl2N]m rings can be used to isolate
[PCl2N]5, [PCl2N]6, [PCl2N]7, [PCl2N]8, and [PCl2N]9 in air.
Instead of heptane, hexanes also can be used as an elutant, but
the separation is not as good, with no separation of [PCl2N]4
and [PCl2N]5. According to NMR, MS and X-ray crystallo-
graphic data, which will be described below, the [PCl2N]m (m =
5−9) rings eluted in order of molecular weight, with the
smallest ring coming off first. Several reports indicate that
Scheme 1 produces cyclic oligomeric chlorophosphazenes that
are larger than dodecamers.32 Further elution of the column
with heptane and then toluene gave materials of higher
molecular weights but in quantities too small for complete
characterization. Some decomposition was evident because a
yellow band of material remained on the column.
The crude product mixture from Scheme 1 appears to
contain acidic impurities, and the polymer [PCl2N]n develops
acidic impurities upon storage.33 We have observed significantly
greater air sensitivity when the rings [PCl2N]m are impure; this
may be due to the presence of the more air-sensitive
[Cl3PN(PCl2N)rPCl3 ]X− or acidic impurities. Upon long-
+
term storage of the impure higher rings in air, degradation
begins to occur, as evidenced by the presence of POCl3,
H3PO4, and related impurities (by 31P NMR spectroscopy34).
Sublimation or filtration through a pad of silica are effective
methods to remove any acidic impurities or [Cl3PN-
(PCl2N)rPCl3 ]X− that may be present. Therefore, the success
+
of the column chromatography may be due, in part, to the
destruction of some impurities by the silica. Although the
purified medium-sized rings are fairly stable to air (∼1−2
months), we recommend that they be stored under a dry gas to
eliminate any degradation.
Previous reports of 31P chemical shifts for the medium-sized
cyclics [PCl2N]m are unclear and several reports are contra-
dictory,19,35 which includes incorrect assignments by our group
based on 31P DOSY NMR spectroscopy.36 Because the
chromatography allowed us to isolate each species and confirm
its identity with MS and X-ray crystallography (see below), the
31P NMR chemical shifts now can be assigned unambiguously.
Figure 1 shows the 31P NMR spectrum of each isolated cyclic
[PCl2N]m in order of elution from the chromatographic
column. The 31P chemical shifts were assigned as follows:
[PCl2N]5 at −15.1 ppm, [PCl2N]6 at −15.3 ppm, [PCl2N]7 at
−17.0 ppm, [PCl2N]8 at −17.7 ppm, and [PCl2N]9 at −17.7
ppm. The 31P chemical shift moves upfield as the ring gets
8878
dx.doi.org/10.1021/ic500272b | Inorg. Chem. 2014, 53, 8874−8886

Inorganic Chemistry
Article
[PCl2N]4 are exaggerated, unless a long relaxation delay was
used. To the best of our knowledge, this was not done in earlier
studies.
Except for an ESI-MS study of [PCl2N]3,40 several decades
have passed since the last report of the characterization of
chlorophosphazene rings by mass spectrometry.41 For this
reason, and to obtain unambiguous NMR assignments, the
samples used for NMR characterization were further diluted
and characterized by high-resolution electrospray ionization
mass spectrometry (ESI-MS). A few ions were observed from
the various [PCl2N]m in the positive ESI mode, but these data
were less useful than data obtained in the negative mode. As an
example, Figure 2 shows the high-resolution ESI spectrum of
Figure 1. 31P NMR spectra of (a) the cyclic pentamer [PCl2N]5, (b)
the cyclic hexamer [PCl2N]6, (c) the cyclic heptamer [PCl2N]7, (d)
the cyclic octamer [PCl2N]8, and (e) the cyclic nonamer [PCl2N]9, in
order of elution from a silica gel column.
larger, but it levels off for the higher rings. This general trend is
37
also observed for the rings [P(OPh)2N]m and largely for
[PMe2N]m and [PF2N]m.14g,38 The reassigned resonances of
[PCl2N]m offer evidence that [PCl2N]8 and [PCl2N]9 behave
most similarly to the high-molecular-weight polymer [PCl2N]n
in solution, because the 31P chemical shifts for the three
substances are virtually identical.
When several cyclic chlorophosphazene species are present
in a sample, integration of 31P NMR spectra has been used to
quantify the relative amount of each ring present.5,35,39
However, strained [PCl2N]3 and [PCl2N]4 would be expected
to have longer relaxation times than the unstrained larger rings.
A spin−lattice relaxation time (T1) 31P NMR spectroscopy
experiment for a sample of purified rings [PCl2N]m (m = 3−9)
in CDCl3 was completed and these data are shown in the
Supporting Information. The relaxation delay value is the
minimum relaxation delay time required to obtain accurate
integration values, and it is five times the largest T1 value.
[PCl2N]4 has the longest relaxation time at 14.0 s. The T1 value
for the remaining rings decreases as the rings increase in size:
[PCl2N]3, 13.4 s; [PCl2N]5, 11.2 s; [PCl2N]6, 8.2 s; and
[PCl2N]7, 7.8 s. The unresolved signal for [PCl2N]8 and
[PCl2N]9 relaxes with the shortest T1 value of 3.2 s. Therefore,
for accurate quantification of mixtures of [PCl2N]m (m = 3−9),
a relaxation delay time of 70 s (5T1 of [PCl2N]4) is required,
which drastically lengthens the time required to obtain a
spectrum. Based on several runs of the reaction in Scheme 1,
31P NMR spectra that are integrated without the proper
relaxation delay can result in great overestimation of the yields
of the higher [PCl2N]m (m > 5). All integrations reported in
this paper were run with a 70-s delay.
There are some important ramifications that result from the
above NMR studies. The reassignment of 31P NMR chemical
shifts of the medium-sized rings indicates that some earlier
mechanistic suggestions, with regard to the process in Scheme
1, may be in error. Because of the similarity of the T1 values,
reported relative quantities of [PCl2N]3 to [PCl2N]4 obtained
from the integrations of 31P NMR spectra are relatively
accurate, irrespective of the relaxation delay that was used to
obtain the spectrum. However, the relative quantities of the
medium-sized rings compared to those of [PCl2N]3 or
Figure 2. Negative-ion mode, ESI mass spectrum of the cyclic
heptamer [PCl2N]7. The cyclics [PCl2N]m (m = 5, 6, 8, 9) show
similar product ion patterns, and their spectra can be found in the
Supporting Information.
the [PCl2N]7 in the negative ion mode. The highest mass signal
arises from addition of one electron to neutral [PCl2N]7 to give
−
the radical anion [PCl2N]7 . The next highest mass signal at
18.97 Da lower mass has been assigned as the mono-
oxygenated heptamer [P7Cl13N7O]−, which arises by displace-
ment of one Cl atom of [PCl2N]7− by an O atom, whose source
presumably would be water. The third signal at 34.97 Da lower
mass from that of [PCl2N]7− is formed by loss of a Cl atom and
is assigned to [P7Cl13N7]−. All isolated [PCl2N]m rings (m = 5−
9) showed [PmCl2m Nm]− , [PmCl2 m− 1NmO]− , and
[PmCl2m−1Nm]− species in the negative mode ESI-MS (see
the Supporting Information) as the predominant species. This
observation suggests that no ring−ring equilibrium occurs in
the gas phase with any of the isolated cyclics, even under high-
energy conditions, and suggests that the medium-sized cyclic
oligomers have a relatively high level of stability toward
ionization. All the signals observed for [PCl2N]m rings (m = 5−
9) in the negative ESI mode showed the expected isotope
distributions (error between 3−15 ppm).
Single-Crystal X-ray Diffraction Studies of [PCl2N]m or
Their Derivatives. Crystallographic data were collected for
[PCl2N]3, as well as both forms of [PCl2N]4, [PCl2N]5,
[PCl2N]6, [PCl2N]8, [P(OPh)2N]7, and [P(OPh)2N]8. The
structures of the first three are known,15−17 but they were
redetermined in order to make comparisons of crystal
structures obtained under the same conditions and with
structures that have better R-values (see the Supporting
8879
dx.doi.org/10.1021/ic500272b | Inorg. Chem. 2014, 53, 8874−8886

Inorganic Chemistry
Article
Information). Crystal data and structure refinement parameters
for [PCl2N]6, [PCl2N]8, [P(OPh)2N]7, and [P(OPh)2N]8 are
shown in Table 1, and complete lists of bond lengths and
angles can be found in the Supporting Information. The crystal
structure of [PCl2N]6 was particularly challenging to obtain. An
orthorhombic form of [PCl2N]6, known since 1897, and two
triclinic forms of [PCl2N]6 had been reported, but none of the
crystal structures had been solved.18,19 Initially, triclinic crystals
of [PCl2N]6 were obtained that had a unit cell that was the
same as that reported by Paddock and co-workers.19 Our efforts
to solve the triclinic structure were unsuccessful, because of
high thermal motion of the Cl and N atoms even at low
temperatures. In order to obtain a derivative, [PCl2N]6 was
treated with HCl/SbCl5. NMR data showed that the
protonation reaction was successful. However, the protonated
crystals degraded in the glovebox over a week and monoclinic
crystals of [PCl2N]6 were obtained.42 The structure of the
monoclinic crystals of [PCl2N]6 was solved, over a century after
the first reported unit cell! Crystals of [PCl2N]8 were solved in
the same unit cell as that reported by Paddock in 1960.19 We
looked for the possibility of two crystalline forms for [PCl2N]8,
because the ring has an unusual shape (see below) but did not
find a second form. Because [PCl2N]7 and [PCl2N]9 are oils at
room temperature, their phenoxide derivatives were prepared
by reaction with excess KOPh. Only the crystals of
[P(OPh)2N]7 were suitable for X-ray diffraction. The derivative
[P(OPh)2N]8 was made from [PCl2N]8 to provide an
additional example for a study on how the crystal structure of
octameric rings is related to the substituent (see below).
The thermal ellipsoid plot of one of the two molecules of
monoclinic [PCl2N]6 is shown in Figure 3, and selected bond
isomers of the 12-membered ring of the thionyl phosphazene
[(PCl2N)2SOClN]4.44
The thermal ellipsoid plot of the crystal structure of
[PCl2N]8 is shown in Figure 4, and selected bond distances
and angles are given in Tables 2 and 3. The ring is elongated
and is nearly planar. This ring shape is rather different from
those observed for the crystal structures of other octameric
phosphazene rings,14f,45,46 including that of [P(OPh)2N]8 (see
Figure 7 below). Some of the larger P−N distances are
statistically different from the smaller ones, but there is no
alternating or other pattern to the different lengths, including
whether the bond is part of the more linear segment of the ring
or at the bent ends, which will be referred to as the kinks.
Average bond lengths and angles for the crystallographically
characterized [PCl2N]m (m = 3−6, 8) rings are shown in Tables
2 and 3, respectively. The P−N bond distances (Table 2) are
toward the longer end of the range of distances characteristic of
cyclophosphazenes and polyphosphazenes (1.47−1.62 Å).47
With the exception of [PCl2N]6, the P−N bond lengths
become shorter as the cyclics become larger. This correlates
with the observation that the larger rings have little or no ring
strain and thereby more idealized bonding between P and N
may occur. Except for [PCl2N]3, the average N−P−N angle
decreases as the ring size increases. No clear trend in the
average P−N−P angle and the size of the ring was observed,
and the average Cl−P−Cl and N−P−Cl angles vary little with
ring size. Interesting trends are observed from the ranges of
particular angles for the cyclic oligomers. Consistent with the
expectation that the larger rings are more flexible than the
smaller homologues, the range of N−P−N angles is ∼0° for
[PCl2N]3 and [PCl2N]4 (boat), ∼2° for [PCl2N]4 (chair), ∼4°
for [PCl2N]5, ∼6°[PCl2N]6, and ∼6° for [PCl2N]8. However,
this trend is not observed for the P−N−P angle. The P−N−P
angles are highly variable in [PCl2N]5 and [PCl2N]6 and are in
the range of 22° and ∼18°, respectively. In contrast, the range
of P−N−P angles is ∼10° for [PCl2N]8 and ∼7° for [PCl2N]4
(chair), and the range for either [PCl2N]3 or [PCl2N]4 (boat) is
<1°.
The crystal structure of highly stretched fibers of the polymer
[PCl2N]n has been reported by two different groups.21,22
Because one structure determination had problems with
disorder or crystal twinning, bond distances and angles from
the better-resolved structure are given in Tables 2 and 3. The
structural parameters for [PCl2N]n are most similar to those of
the larger rings, especially to those of [PCl2N]8. In contrast to
the structure of [PCl2N]n,22 no significant alternation in the
length of the P−N bonds is observed for either [PCl2N]6 or
[PCl2N]8.
In the crystal structures of the homologous series [PCl2N]m
(m = 3−6, 8), two different types of halogen bonds are
observed: P−Cl···Cl and P−Cl···N. The halogen bonding
interactions of each ring can be seen in Figure 5, and their
distances and the number per molecule are given in Table 4.
The distances in the halogen bonds fall within the sum of the
van der Waals radii, which are 3.6 and 3.3 Å for P−Cl···Cl and
P−Cl···N interactions, respectively. P−Cl···Cl halogen bonds
are consistently observed in all the rings, but the P−Cl···N
halogen bond is only observed for [PCl2N]3, [PCl2N]5, and
[PCl2N]6. The shortest P−Cl···Cl interactions are found in
[PCl2N]6, whereas the shortest P−Cl···N interactions are found
in [PCl2N]3. Halogen bonding is most prevalent in monoclinic
[PCl2N]6, where 12 or 14 separate interactions occur with the
two forms and least prevalent in the chair form of [PCl2N]4. It
Figure 3. Thermal ellipsoid plot of one of the two molecules of the
crystal structure of [PCl2N]6 drawn at the 50% probability level.
distances and angles are given in Tables 2 and 3. The overall
shapes of the two molecules are similar, but there are slight
differences in the structural parameters of the two rings.
Halogen bonding interactions may be responsible for the slight
differences (see below). Although the average P−N distances in
the two rings are similar (1.566 Å vs 1.567 Å), there are
significant differences in the ranges of distances [1.561(3)−
1.575(3) Å vs 1.556(3)−1.583(3) Å]. The nonplanar, bowl-
shaped molecules of [PCl2N]6 have rather differently shaped
43
rings than the double tub structures of [P(OMe)2N]6 and
14d
[PMe2N]6
but are similar to one of the two structural
8880
dx.doi.org/10.1021/ic500272b | Inorg. Chem. 2014, 53, 8874−8886

Inorganic Chemistry
Article
22
Table 2. Average Bond Lengths and Ranges in [PCl2N]m (m = 3−6, and 8) and [PCl2N]n
average P−N distance (range, Å)
average P−Cl distance (range, Å)
[PCl2N]3
1.58092 [1.5794(17)−1.5818(17)]
1.5710 [1.5710(12)−1.5711(12)]
1.565 [1.561(2)−1.570(2)]
1.5584 [1.547(3)−1.568(3)]
1.567 [1.556(3)−1.583(3)]
1.555 [1.544(5)−1.564(5)]
1.56 [1.44, 1.67]
1.9920 [1.9875(7)−1.9961(11)]
1.9952 [1.9951(7)−1.9952(6)]
1.997 [1.9902(9)−2.0028(9)]
1.9980 [1.9889(12)−2.0068(12)]
1.9987 [1.9870(12)−2.0097(12)]
2.002 [1.993(2)−2.0109(19)]
2.01 [1.97, 2.04]
[PCl2N]4 (boat)
[PCl2N]4 (chair)
[PCl2N]5
[PCl2N]6
[PCl2N]8
a
[PCl2N]n
a
Values taken from ref 22.
a
Table 3. Bond Angles of [PCl2N]m (m = 3−6, and 8) and [PCl2N]n
N−P−N avg. (range, deg)
P−N−P avg. (range, deg)
Cl−P−Cl avg. (range, deg)
N−P−Cl avg. (range, deg)
[PCl2N]3
118.38 [118.34(12)−118.41(10)]
120.20 [119.21(13)−121.18(13)]
121.20 [120.98(15)−121.41(11)]
134.70 [131.12(14)−138.27(15)]
101.75 [101.40(4)−102.09(3)]
103.38 [103.21(4)−103.55(4)]
108.74 [108.12(9)−109.27(7)]
107.97 [105.10(8)−111.15(9)]
[PCl2N]4
(chair)
[PCl2N]4
(boat)
120.46 [120.46(8)]
130.91 [130.91(8)]
103.19 [103.19(2)]
107.97 [105.02(5)−111.89(5)]
[PCl2N]5
[PCl2N]6
[PCl2N]8
117.43 [115.49(14)−119.93(15)]
116.84 [114.72(15)−120.77(14)]
114.5 [111.7(2)−117.5(2)]
140.3 [129.92(17)−152.0(2)]
133.79 [123.61(16)−141.30(19)]
138.3 [133.9(3)−143.1(3)]
131
102.49 [100.99(5)−103.56(5)]
102.53 [101.71(6)−103.59(5)]
101.46 [101.11(9)−102.26(8)]
99
108.95 [105.05(11)−112.28(12)]
109.12 [104.74(10)−112.62(11)]
110.0 [107.4(2)−111.8(2)]
a
b
b
[PCl2N]n
115
110 [108−113]
a
b
Values taken from ref 22. Values calculated from data reported in ref 22.
range of P−N bond distances of 1.545−1.614 Å,48 depending
on the method used in the calculation, and no studies could be
found for [PCl2N]6. In the crystal structure of stretched fibers of
polymeric [PCl2N]n, Cl···Cl contacts of 3.61, 3.72, 3.79, and
3.84 Å are observed that are equal to or longer than the sum of
the van der Waals distances. Therefore, no halogen bonds are
observed in stretched [PCl2N]n.22
The thermal ellipsoid plots of the structures of [P(OPh)2N]7
and [P(OPh)2N]8 are shown in Figures 6 and 7, respectively.
There are intermolecular C−H···O and C−H···π interactions in
both structures. Selected distances and angles are shown in
Table 31 of the Supporting Information, and parameters for the
crystal structures of [P(OPh)2N]3 and [P(OPh)2N]4 are
included therein for comparison.49 The shape of the P−N
ring of [P(OPh)2N]7 has an irregular tublike conformation, in
contrast to the spear-tip-shaped ring of [PMe2N]7.14e The P−N
ring of [P(OPh)2N]8 has two fairly planar regions separated by
46
Figure 4. Thermal ellipsoid plot of the cyclic octamer [PCl2N]8 drawn
at the 50% probability level.
a skewed step, somewhat like that of [P(OMe)2N]8 (see
Figure 8 below). The bond distances for [P(OPh)2N]8 are
similar to those of [P(OPh)2N]7, where the difference in
average P−N, P−O, and O−C bond distances are all less than
0.008 Å. The difference for average bond angles are <2°. As
with the chloro rings, the P−N distance of [P(OPh)2N]m
decreases as the ring becomes larger. The P−O bond distances
are shorter than a P−O single bond, which has been attributed
to partial double-bond character.50
Structural Comparisons among Phosphazene Oc-
tamers with Different Substituents. For historical and
synthetic reasons, many discussions of phosphazenes begin with
the chlorophosphazenes. This perspective gives the impression
that chlorophosphazenes are the prototypical phosphazene.
However, in several instances above, the P−N ring shape of
medium-sized [PCl2N]m is rather different than those of other
phosphazenes. This suggests chlorophosphazenes may not be
prototypical in terms of structure. Because there now are five
crystallographically characterized octamers, [PCl2N]8, [P-
(OPh)2N]8, [P(OMe)2N]8,46 [P(NMe2)2N]8,45 and
[PMe2N]8,14f a comparison of their structures, including the
is interesting to note that the more favored boat conformation
of [PCl2N]4 shows four halogen bonds per molecule, whereas
the less stable chair form only has two. The boat form of
[PCl2N]4 has longer average P−N bonds than the chair form,
whereas the [PCl2N]6 ring with the larger number of
interactions also has the greater range of P−N distances within
the ring. More intermolecular interactions would help to
stabilize a crystal, and, therefore, halogen bonding appears to be
responsible for the ranges in distances and angles and for the
preference of a particular conformation in [PCl2N]m rings.
There was no clear correlation between any of the bond lengths
and angles and the number of halogen bonds. The P−N
framework of the various [PCl2N]m rings may be flexible
enough that each molecule could have several different ways of
maximizing the number of halogen bonds. To probe these
possibilities, it would be useful to have the P−N bond distances
of [PCl2N]4 or [PCl2N]6 in the absence of halogen bonding.
However, the available theoretical studies of [PCl2N]4 give a
8881
dx.doi.org/10.1021/ic500272b | Inorg. Chem. 2014, 53, 8874−8886

Inorganic Chemistry
Article
Figure 5. Halogen bonding (red dashed lines) in the crystal structures of [PCl2N]3 A, [PCl2N]4 (chair B) and (boat C), [PCl2N]5 D, [PCl2N]6 E
and F (two molecules within the unit cell) and [PCl2N]8 G.
Table 4. Halogen Bond Distances from the Crystal Data for Cyclic [PCl2N]m
P−Cl···Cl
P−Cl···N
number of contacts
avg. (range, Å)
number of contacts
2
avg (range, Å)
3.141 (3.141)
[PCl2N]3, A
2
2
3.453 (3.453)
3.480 (3.480)
[PCl2N]4 (chair), B
[PCl2N]4 (boat), C
[PCl2N]5, D
4
3.497 (3.497)
2
3.409 (3.409)
2
4
2
3.213 (3.213)
3.267 (3.251−3.283)
3.250 (3.250)
[PCl2N]6, E
10
10
10
3.427 (3.381−3.486)
3.407 (3.291−3.486)
3.414 (3.399−3.421)
[PCl2N]6, F
[PCl2N]8, G
Figure 7. Thermal ellipsoid plot of [P(OPh)2N]8 drawn at the 50%
probability level. Hydrogen atoms omitted for clarity.
Figure 6. Thermal ellipsoid plot of [P(OPh)2N]7 drawn at the 50%
probability level. Hydrogen atoms omitted for clarity.
8882
dx.doi.org/10.1021/ic500272b | Inorg. Chem. 2014, 53, 8874−8886

Inorganic Chemistry
Article
a
Table 5. Conformations of the Ring P−N Bonds for Phosphazene Octamers
[PCl2N]8
[P(OPh)2N]8
[P(OMe)2N]8
[PMe2N]8
[P(NMe2)2N]8
number of ω range (deg) number of ω range (deg) number of ω range (deg) number of ω range (deg) number of ω range (deg)
S, SP
C
4
0−12
2
8
31
4
4
2
2
4
28−34
54−61
79
4
4
4
4
28
4
2
4
4
27−39
64
G
50−61
61
O
77
79
D
6
2
4
158−160
167
4
154−157
153
157
153−155
T
164−170
A
175−176
2
177
2
178
Dihedral angles for [P(OMe)2N]8,46 [P(NMe2)2N]8,45 and [PMe2N]814f were determined from the CIF files associated with the references in the
a
Cambridge Crystallographic Data Centre. The columns are arranged in order of increasing basicity.
configuration of the P−N chain in the ring, was undertaken.
Selected distances and angles and a summary of intermolecular
interactions for the five octamers are shown in Table 32 of the
Supporting Information (Table SI32). Based on the trend
observed for similarly substituted trimers,51 the columns in
Table SI32 (and in Table 5 and Figure 8 below) are arranged in
order of increasing basicity, starting with the weakly basic
[PCl2N]8 to highly basic [P(NMe2)2N]8. This was done to aid
in the discernment of electronic factors that may be involved in
determining the structures of the octamers, which is a factor
that has been considered by two theoretical studies of
chlorophosphazenes.52,53
The average metrical parameters in Table SI32 in the
Supporting Information for the five octameric rings show no
clear dependence on the basicity of the nitrogen atom in the
Figure 8. Two views of the P−N rings of [PCl2N]8, [P(OPh)2N]8,
ring, electronegativity of the substituent, or sterics. Except for
the structure of [P(OMe)2N]8, which has the largest estimated
standard deviations, the longer and shorter P−N bond
distances within each octameric ring are slightly different
from one other. The most variable angle among the octamers is
the P−N−P angle. However, the average P−N−P angles are all
within 2° of 138°, except the angle for [P(NMe2)2N]8 (156.5°).
Perhaps this larger angle is caused by steric hindrance among
substituents, but the next largest angle is for [PMe2N]8, which
contains small substituents. The average N−P−N angle for
[PCl2N]8 is 2°−4° less than the average angles of the other four
octamers. The average N−P−N angle (within the ring) for the
other octamers vary from each other by no more than a degree.
The average N−P−X (X = substituent and N is within the ring)
angles for the five octamers are similar and have a range of only
1.1°. The average X−P−X angles encompass a 3.6° range of
values but the differences are not explained by simple steric or
electronic arguments.
The shapes of the P−N rings and views perpendicular to the
planes of the P−N rings for the five octamers are shown in
Figure 8. The relative basicity of the ring nitrogen shows no
discernible effect on the structure of the ring. As described
above, [PCl2N]8 shows 10 Cl···Cl halogen bonds. The
structures of [P(OPh)2N]8 (eight C−H···O and 11 C−H···π
interactions) and [P(OMe)2N]8 (four CH···O interactions)
show contacts between molecules that are shorter than the sum
of the van der Waals radii. The most common shape of the P−
N rings among the five octamers is a chairlike or step structure,
in which there are two roughly planar regions of the ring. Both
[P(NMe2)2N]8 and [P(OMe)2N]8 have step structures and
[P(OPh)2N]8 shows a twisted-step structure. A combination of
steric interactions and intermolecular forces may be responsible
for the distortions observed in [P(OPh)2N]8 from a more
regular step structure. Of the five different substituents of the
[P(OMe)2N]8, [P(NMe2)2N]8, and [PMe2N]8 (P = yellow, N = blue).
The top view is of the best planes of the rings. The bottom view is
roughly perpendicular to the top view. The columns are arranged in
order of increasing basicity of the ring N atom. The substituents have
been removed for clarity. The drawings for [P(OMe)2N]8,46
14f
[P(NMe2)2N]8,45 and [PMe2N]8 were generated from the CIF
files associated with their references in the Cambridge Crystallographic
Data Centre.
octamers, chloride and methyl are the smallest and they are
roughly the same size.54 [PMe2N]8 has a crown-like ring that is
very different from the oblong, nearly planar ring in [PCl2N]8.
The latter, approximately planar ring stands apart from the
nonplanarity of the other four octameric rings. The only factor
that seems to account for the unusual structure of [PCl2N]8 is
halogen bonding.
The shapes of the five octamers can be compared to those of
nonphosphazene rings. As Paddock and co-workers have
pointed out, the distinctive crown shape of [PMe2N]8 is similar
that of [SiMe2O]8, but there are some subtle differences.55
Unfortunately, the crystal structures of other appropriately
substituted octameric siloxanes are not available to make
further comparisons. Paddock also noted that the shape of the
[PMe2N]m (m = 7−12) rings are related to the shapes of the
respective [CH2]m hydrocarbons rings, especially for the even
values of m. In these comparisons, each PMe2N unit is being
associated with a single CH2 unit. Except for the shape of
[PCl2N]8, the shapes of the other four phosphazene octamers
are similar to those of the higher-energy structures [CH2]8.56
There seems to be no agreement as to the most stable
structures of [CH2]16, but a rectangular structure similar to that
of [PCl2N]8 is a higher-energy form of [CH2]16 in some
calculations.57 The shape of [SiMe2]16 is different from any
shape exhibited by the five octamers.20
8883
dx.doi.org/10.1021/ic500272b | Inorg. Chem. 2014, 53, 8874−8886

Inorganic Chemistry
Article
Paddock and co-workers considered the conformations of
several medium-sized [PMe2N]m (m = 8, 9, 12) rings.14f,g
However, the nomenclature system they used does not describe
all the conformations of the five phosphazene octamers.
Therefore, the Michl−West system (A, anti, ω ≈ 180°; T,
transoid, ω ≈ 165°; D, deviant, ω ≈ 150°; O, ortho, ω ≈
90°; G, gauche, ω ≈ 60°; C, cisoid, ω ≈ 30°; S, syn, ω ≈
0°; E, eclipsed, ω ≈ 120°)58 and one additional conformation
(SP, syn-periplanar, 0° < ω > 30°59) will be used to describe
the five octameric phosphazene rings in Table 5 (pie-chart
representations are given in the graphical abstract and in Figure
14 of the Supporting Information). The structure of [PCl2N]8
has two very different types of torsion angles; large D, T, and A
angles for the linear portion of the molecule and S and SP for
the kinks. Once derivatized with organic functional groups, the
S and SP conformations are no longer observed and there are
fewer D, T, or A conformations. In their place, C and G
conformations are observed for the four organo-octamers and
O conformations are observed for all but [P(OPh)2N]8.
Does the above structural analysis of the five octamers
provide insights into the structures of the polymeric
phosphazenes? Unusual features in the structure of [PCl2N]8
include the nearly planar P−N ring, the conformations of the
P−N bonds, and the large number of intermolecular
interactions in the form of halogen bonding. The linear
fragment of the [PCl2N]8 molecule is close to ideal all-
trans(oid), a conformation that is so unlike those observed in
the other four octamers and of most polymers. Most
experiments and calculations on the structures of the polymers
[PCl2N]n (C: 31°, A: 175° for stretched fibers),22 [PMe2N]n (S:
2.39° and A: 174.1°),60 [P(OPh)2N]n,61 and other poly-
phosphazenes indicate that alternating cis−trans(oid) (S-A) or
distorted cis−trans(oid) (C-A) conformations are the most
stable.62,63 The cis−trans conformation is favored because it
minimizes steric interactions among the substituents on
adjacent P atoms that are separated by a N atom that has no
substituents. Interestingly, a molecular dynamics calculation of
unstretched [PCl2N]n indicates that, on average, the P−N chain
contains segments of roughly three (2.7) P−N bonds in
trans(oid) conformations that alternate with single cis−trans
units and result in a globular structure.53 This calculated
structure shares some similarity to the pattern of conformations
observed for the P−N bonds in the [PCl2N]8 ring. Each half of
the ring can be described as two cis−trans P−N bonds
separated by a segment of six trans(oid) P−N bonds.
Therefore, the structure of [PCl2N]8, which appears to be
stabilized by halogen bonds, may be a better model for the
structure of unstretched rather than stretched [PCl2N]n. As
mentioned earlier, no halogen bonds are observed in the crystal
structure of stretched fibers of polymeric [PCl2N]n.22 Because
there is no compelling reason to believe that [PCl2N]n would
not form halogen bonds like other chlorophosphazenes, the
lack of halogen bonds in stretched fibers of [PCl2N]n could
indicate that stretching ∼8 times from the original length
disrupted the halogen bonding in unstretched [PCl2N]n. If
unstretched [PCl2N]n is globular, then a constantly changing
network of Cl···Cl halogen bonds may be present, somewhat
reminiscent of the hydrogen-bonding network in liquid water.
This could occur because each chloride can act as both an acid
and a base in a halogen bond, and because there would be the
possibility of many such interactions due to the high chlorine
content. Lastly, the usefulness of the Michl−West notations for
conformations58 of the phosphazene octamers is notable and
surprising. In contrast to phosphazenes, the polysilanes that
inspired the Michl−West notations have two substituents at
every chain atom and all-trans(oid) conformations are the more
stable.
SUMMARY
■
Isolation of the individual medium-sized cyclic chlorophospha-
zenes [PCl2N]m (m = 5−9) has been completed through a
combination of separation techniques, where complete
separation was accomplished via column chromatography.
Each [PCl2N]m has been characterized by 31P NMR (chemical
shifts and relaxation times), ESI-MS, and if possible by
crystallographic data of it or a derivative. The crystal structures
of all the [PCl2N]m rings (m = 3, 4, 5, 6, 8) show that halogen
bonding interactions are present and, for most of the rings,
there are many interactions per ring. The conformations of the
P−N bonds for five octameric phosphazene rings largely can be
described by the Michl−West notations. The ring shape and
conformations of nearly planar [PCl2N]8 are quite different
from those of nonplanar [P(OPh)2N]8, [P(OMe)2N]8,
[PMe2N]8, and [P(NMe2)2N]8, and halogen bonding appears
to be the major contributing factor in the differences. The data
reported herein suggests that the medium-sized cyclic
chlorophosphazene [PCl2N]8 is a good model compound for
unstretched [PCl2N]n. The structures of chlorophosphazenes
should not be considered as prototypical of other phospha-
zenes.
ASSOCIATED CONTENT
■
S
* Supporting Information
NMR and mass spectra, X-ray diffraction data for all crystal
structures, a table of comparisons among phenoxy [P-
(OPh)2N]m (m = 3, 4, 7, 8), and a figure showing comparisons
among the five octameric phosphazene rings [P(X)2N]8 (X =
Cl, OPh, OMe, NMe2, Me). This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*Tel.: 330-972-5304. Fax: 330-972-7370. E-mail: tessier@
uakron.edu.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Christopher Allen (University of Vermont, retired),
John Rapko (St. Louis College of Pharmacy, St. Louis, MO),
and Zin-Min Tun (University of Akron, Akron, OH) for
reading the manuscript and providing useful suggestions. This
research was supported in part by ICL and the Ohio Board of
Regents. The latter also provided funds used to purchase
several instruments used in this work. We thank the National
Science Foundation (NSF) for support of early efforts (under
Grant No. CHE-0616601) and for providing funds for the
NMR instruments (Nos. CHE-0341701 and DMR-0414599),
the mass spectrometers (CHE-0821313 and CHE-1012636)
and the X-ray diffractometers (CHE-0116041 and CHE-
0840446). We thank Goodyear Corporation for donation of
an NMR instrument.
8884
dx.doi.org/10.1021/ic500272b | Inorg. Chem. 2014, 53, 8874−8886

Inorganic Chemistry
Article
(14) (a) Oakley, R. T.; Paddock, N. L.; Rettig, S. J.; Trotter, J. Can. J.
Chem. 1977, 55, 4206−4210. (b) Huynh, K.; Lough, A. J.; Forgeron,
M. A. M.; Bendle, M.; Soto, A. P.; Roderick E. Wasylishen, R. E.;
Manners, I. J. Am. Chem. Soc. 2009, 131, 7905−7916. (c) Dougill, M.
W.; Sheldrick, B. Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst.
Chem. 1977, B33, 295−297. (d) Oakley, R. T.; Paddock, N. L.; Rettig,
S. J.; Trotter, J. Can. J. Chem. 1977, 55, 3118−3123. (e) Gallicano, K.
D.; Oakley, R. T.; Paddock, N. L.; Rettig, S. J.; Trotter, J. Can. J. Chem.
1977, 55, 304−309. (f) Oakley, R. T.; Paddock, N. L.; Rettig, S. J.;
Trotter, J. Can. J. Chem. 1977, 55, 2530−2533. (g) Oakley, R. T.;
Rettig, S. J.; Paddock, N. L.; Trotter, J. J. Am. Chem. Soc. 1985, 107,
6923−6936.
REFERENCES
■
(1) Selected books and book sections: (a) Mark, J. E.; Allcock, H. R.;
West, R. Inorganic Polymers, 2nd Edition; Oxford University Press:
New York, 2005; Chapter 3. (b) Andrianov, A. K., ed.
Polyphosphazenes for Biomedical Applications; Wiley: New York, 2009.
(c) Allcock, H. R. The Chemistry and Applications of Polyphosphazenes;
Wiley−Interscience: New York, 2003; Chapters 12−19 (especially).
(d) Gleria, M., De Jeaeger, R., Eds. Phosphazenes A Worldwide Insight;
Nova Science: New York, 2004, Parts 2−4 (especially).
(2) Selected recent reviews: (a) Deng, M.; Laurencin, C. T.; Allcock,
H. R.; Kumbar, S. G. In Polymeric Biomaterials, 3rd Edition; Dumitriu,
S., Popa, V., Eds.; CRC Press: Boca Raton, FL, 2013; Vol. 1, pp 83−
134. (b) Andrianov, A. K.; Mutwiri, G. Vaccine 2012, 30, 4355−4360.
(c) Allcock, H. R.; Morozowich, N. L. Polym. Chem. 2012, 3, 578−590.
(d) Allen, C. W. Inorg. Chim. Acta 2011, 372, 32−36. (e) Eng, N. F.;
Garlapati, S.; Gerdts, V.; Potter, A.; Babiuk, L. A.; Mutwiri, G. K. Curr.
Drug Delivery 2010, 7, 13−20. (f) El Gouri, M.; El Harfi, A. J. Mater.
Environ. Sci. 2012, 3, 17−33.
(15) (a) Bartlett, S. W.; Coles, S. J.; Davies, D. B.; Hursthouse, M. B.;
́
̋
́
Ibiso̧ glu, H.; Kilic, A.; Shaw, R. A.; Un, I. Acta Crystallogr., Sect. B:
̆
Struct. Sci. 2006, B62, 321−329. (b) Brockway, L. O.; Bright, W. M. J.
Am. Chem. Soc. 1943, 65, 1551−1554. (c) Wilson, A.; Caroll, D. F. J.
Chem. Soc. 1960, 2548−2552. (d) Bullen, G. J. J. Chem. Soc. A 1971,
1450−1453.
(16) (a) Hazenkamp, R.; Michelsen, T.; Vos, A. Acta Crystallogr.
1962, 15, 539−543. (b) Wagner, A. J.; Vos, A. Acta Crystallogr., Sect. B:
Struct. Crystallogr. Cryst. Chem. 1968, B24, 707−713.
(17) (a) Schlueter, A. W.; Jacobson, R. A. J. Am. Chem. Soc. 1966, 65,
2051. (b) Schlueter, A. W.; Jacobson, R. A. J. Chem. Soc. A 1968,
2317−2325.
(18) (a) Stokes, H. N. Am. Chem. J. 1897, 19, 782−796. (b) Tassin,
W. Z. Kristallogr. 1899, 31, 304.
(19) Lund, L. G.; Paddock, N. L.; Proctor, J. E.; Searle, H. T. J. Chem.
(3) (a) Allcock, H. R. Phosphorus−Nitrogen Compounds; Academic
Press: New York, 1972; Chapter 4. (b) Allcock, H. R. Chem. Rev. 1972,
72, 315−356. (c) Jolly, W. L. The Synthesis and Characterization of
Inorganic Compounds; Prentice−Hall: Englewood Cliffs, NJ, 1970;
Chapter 30.
(4) (a) Liebig, J. Justus Liebigs Ann. Chem. 1834, 11, 139.
(b) Gerhardt, C. Ann. Chim. Phys. 1846, 18, 188. (c) Gladstone, J.
H.; Holmes, J. D. J. Chem. Soc. 1864, 17, 225. (d) Gladstone, J. H. J.
Chem. Soc. 1850, 2, 121. (e) Stokes, H. N. Am. Chem. J. 1896, 18, 629.
(5) (a) Emsley, J.; Udy, P. B. J. Chem. Soc. A 1970, 3025−3029.
(b) Emsley, J.; Udy, P. B. J. Chem. Soc. A 1971, 768−772.
Soc. 1960, 2542−2547.
(20) (a) Brough, L. F.; West, R. J. Am. Chem. Soc. 1981, 103, 3049−
3056. (b) Shafiee, F.; Haller, K. J.; West, R. J. Am. Chem. Soc. 1986,
108, 5478−5482. (c) Shafiee, F.; Damewood, J. R., Jr.; Haller, K. J.;
West, R. J. Am. Chem. Soc. 1985, 107, 6950−6956. (d) West, R. J.
Organomet. Chem. 2003, 685, 6−8.
(6) (a) Allcock, H. R. Chem. Rev. 1972, 72, 315−356. (b) De Jaeger,
R.; Gleria, M. Prog. Polym. Sci. 1998, 23, 179−276.
(7) (a) Huheey, J. E.; Keiter, E. A.; Keiter, R. L. Inorganic Chemistry, 4
Edition; Harper Collins: New York, 1993; pp 769−775. (b) Cha-
drasekhar, V. Inorganic and Organometallic Polymers; Springer: New
York, 2005; pp 108−112. (c) Allcock, H. R. The Chemistry and
Applications of Polyphosphazenes; Wiley−Interscience: New York,
2003; pp 399−406.
(21) (a) Allcock, H. R.; Arcus, R. A. Macromolecules 1979, 12, 1130−
1136. (b) Allcock, H. R.; Arcus, R. A.; Stroh, E. G. Macromolecules
1980, 13, 919−928.
(22) Chatani, Y.; Yatsuyanagi, K. Macromolecules 1987, 20, 1042−
1045.
(23) Selected recent reviews: (a) Cavallo, G.; Metrangolo, P.; Pilati,
T.; Resnati, G.; Sansotera, M.; Terraneo, G. Chem. Soc. Rev. 2010, 39,
́
3772−3783. (b) Erdelyi, M. Chem. Soc. Rev. 2012, 41, 3547−3557.
(c) Bertania, R.; Sgarbossa, P.; Venzo, A.; Lelj, F.; Amati, M.; Resnati,
G.; Pilati, T.; Metrangolo, P.; Terraneo, G. Coord. Chem. Rev. 2010,
254, 677−695.
(24) Xu, L.; Zou, J. W.; Yu, Q. S.; Sang, P. Int. J. Quantum Chem.
2009, 110 (6), 1245−1251.
(25) Tun, Z.-M.; Heston, A. J.; Panzner, M. J.; Medvetz, D. A.;
Wright, B. D.; Savant, D.; Dudipala, V. R.; Banerjee, D.; Rinaldi, P. L.;
Youngs, W. J.; Tessier, C. A. Inorg. Chem. 2011, 50, 8937−8945.
(26) Sheldrick, G. M. SHELX97: Programs for Crystal Structural
(8) Some examples: (a) Chaplin, A. B.; Harrison, J. A.; Dyson, P. J.
Inorg. Chem. 2005, 44, 8407−8417. (b) Sun, H. J. Am. Chem. Soc.
1997, 119, 3611−3618. (c) Vassileva, P.; Krastev, V.; Lakov, L.;
Peshev, O. J. Mater. Sci. 2004, 39, 3201−3202. (d) Waltman, R. J.;
Lengsfield, B.; Pacansky, J. Chem. Mater. 1997, 9, 2185−2196.
(e) Breza, M. Polyhedron 2000, 19, 389−397. (f) Kupka, T.; Pasterny,
K.; Pasterna, G.; Brandt, K. J. Mol. Struct.: THEOCHEM 2008, 866,
̌ ́ ̌
21−26. (g) Kapicka, L.; Kubacek, P.; Holub, P. J. Mol. Struct.:
THEOCHEM 2007, 820, 148−158. (h) Calichman, M.; Derecskei-
Kovacs, A.; Allen, C. W. Inorg. Chem. 2007, 46, 2011−2016.
(9) (a) Hengge, E.; Janoschek, R. Chem. Rev. 1995, 95, 1495−1526.
(b) Blinka, T. A.; West, R. Organometallics 1986, 5, 133−139.
(c) Mark, J. E.; Allcock, H. R.; West, R. Inorganic Polymers, 2nd
Edition; Oxford University Press: New York, 2005; pp 215−220.
(d) Michl, J.; West, R. In Silicon-Containing Polymers; Jones, R. G.,
Ando, W., Chojnowski, J., Eds.; Kluwer Academic Publishers: Boston,
2000; Chapter 18.
Analysis; University of Gottingen, Gottingen, Germany, 1997.
̈
̈
(27) Jolly, W. L. The Synthesis and Characterization of Inorganic
Compounds; Prentice-Hall: Englewood Cliffs, NJ, 1970; Chapter 30.
(28) Novobilsky, V.; Kolsky, V.; Wanek, W. Z. Anorg. Allg. Chem.
1976, 423, 273−279.
(10) Apeloig, Y In The Chemistry of Organic Silicon Compounds; Patai,
S., Rappoport, Z., Eds.; Wiley: New York, 1989: Chapter 2.
(11) (a) Voronkov, M. G.; Mileshkevich, V. P.; Yuzhelevskii, Yu. A.
The Siloxane Bond; Consultants Bureau: New York, 1978. (b) Cruick-
shank, D. W. J. J. Chem. Soc. 1961, 5486−5504. (c) Ebsworth, E. A. V.
In Organometallic Compounds of the Group IV Elements; MacDiarmid,
A. G., Ed.; Marcel Dekker: New York, 1968. (d) Burger, H. Angew,
Chem., Int. Ed. 1973, 12, 474−486.
(12) (a) Gillespie, R. J.; Johnson, S. A. Inorg. Chem. 1997, 36, 3031−
3039. (b) Grabowsky, S.; Hesse, M. F.; Paulmann, C.; Luger, P.;
Beckmann, J. Inorg. Chem. 2009, 48, 4384−4393.
(13) (a) Weinhold, F.; West, R. Organometallics 2011, 30, 5815−
5824. (b) Weinhold, F.; West, R. J. Am. Chem. Soc. 2013, 135, 5762−
5767.
(29) Mitropol’skaya, G. I.; Kireev, V. V. Russ. J. Inorg. Chem. 1983, 28
(8), 1914−1918.
(30) (a) Mitropol’skaya, G. I.; Kireev, V. V. Russ. J. Inorg. Chem.
́
́
̌
1983, 28 (8), 1914−1918. (b) Novobilsky, V. V.; Kolsky, V.; Wanek,
W. Z. Anorg. Allg. Chem. 1975, 416, 187−192. (c) Janos,
̌
P.; Broul, M.;
Novobilsky, V.; Kolsky, V. J. Chromatogr. A 1994, 676, 451−454.
́
́
(d) Novobilsky, V.; Kolsky, V.; Wanek, W. Z. Anorg. Allg. Chem. 1976,
423, 273−279. (e) Sinclair, D. P. U.S. Patent 4,242,316, 1980.
(31) Silvestrelli, P. L.; Gleria, M.; Milani, R.; Boscoletto, A. B. J. Inorg.
Organomet. Polym. Mater. 2006, 16, 327−341.
(32) (a) Allcock, H. R. The Chemistry and Applications of
Polyphosphazenes; Wiley−Interscience: New York, 2003; Chapter 1.
(b) Chapman, A. C.; Paddock, N. L.; Paine, D. H.; Searle, H. T.;
8885
dx.doi.org/10.1021/ic500272b | Inorg. Chem. 2014, 53, 8874−8886

Inorganic Chemistry
Article
Smith, D. R. J. Chem. Soc. 1960, 3608−3614. (c) Shaw, R. A.;
Fitzsimmons, B. W.; Smith, B. C. Chem. Rev. 1962, 62, 247−281.
(33) (a) Liu, H. Q.; Stannett, V. T. Macromolecules 1990, 23, 140−
144. (b) Andrianov, A. K.; Chen, J.; LeGolvan, M. P. Macromolecules
2004, 37, 414−420. (c) Tun, Z.; Panzner, M. J.; Scionti, V.; Medvetz,
D.; Wesdemiotis, C.; Youngs, W. J.; Tessier, C. A. J. Am. Chem. Soc.
2010, 132, 17059−17061.
(57) (a) Shannon, V. L.; Strauss, H. L.; Snyder, R. G.; Elliger, C. A.;
Mattice, W. L. J. Am. Chem. Soc. 1989, 111, 1947−1958. (b) Dale, J.
Acta Scand. 1973, 27, 1115−1129. (c) Allinger, N. L.; Gorden, B.;
Profeta, S., Jr. Tetrahedron 1980, 36, 859−864. (d) Anet, F. A. L.;
Cheng, A. K. J. Am. Chem. Soc. 1975, 97, 2420−2424.
(58) (a) Michl, J.; West, R. Acc. Chem. Res. 2000, 33 (12), 821−823.
(b) Michl, J.; West, R. In Silicon-Containing Polymers; Jones, R. G.,
Ando, W., Chojnowski, J., Eds.; Kluwer Academic Publishers:
Amsterdam, 2000; pp 499−529.
(34) Achmatowicz, M. M.; Thiel, O. R.; Colyer, J. T.; Hu, J.; Silva
Elipe, M. V.; Tomaskevitch, J.; Tedrow, J. S.; Larsen, R. D. Org. Process
Res. Dev. 2010, 14, 1490−1500.
(59) Jenkins, A. D. Pure Appl. Chem. 1981, 53, 733−752.
(60) Meille, S. V.; Poletti, A. R.; Gallazzi, M. C.; Gleria, M.; Bruckner,
S. Polymer 1992, 33, 2364−2369.
(35) (a) Allcock, H. R.; Crane, C. A.; Morrissey, C. T.; Olshavsky, M.
A. Inorg. Chem. 1999, 38, 280−283. (b) Allcock, H. R. The Chemistry
and Applications of Polyphosphazenes; Wiley−Interscience: New York,
2003; Chapter 4. (c) Borisov, A. S.; Hazendonk, P.; Hayes, P. G. J.
Inorg. Organomet. Polym. 2010, 20, 395−398. (d) Horn, H.-G. Z.
Anorg. Allg. Chem. 1974, 406, 199−208. (e) Dwek, R. A.; Richards, R.
E.; Taylor, D.; Shaw, R. A. J. Chem. Soc. A 1970, 1173−1180.
(f) Abouchaccra, T.; Helioui, M.; Puskaric, E.; de Jaeger, R.; Heubel, J.
J. Chem. Res. 1981, 5, 230−231.
(36) Heston, A. J.; Banerjee, D.; Rinaldi, P.; Tessier, C. A. J. Inorg.
Organomet. Polym. Mater. 2007, 17 (2), 477−481.
(37) Ikeda, N.; Utsumi, K.; Fushimi, T.; Tada, Y. Phosphorus, Sulfur,
Silicon 2010, 185, 1521−1525.
(61) Caminiti, R.; Gleria, M.; Lipkowitz, K. B.; Lombardo, G. M.;
Pappalardo, G. C. Chem. Mater. 1999, 11, 1492−1497.
(62) Allcock, H. R. Chemistry and Applications of Polyphosphazenes;
Wiley−Interscience: New York, 2003; Chapter 13.
(63) Pappalardo, G. C.; Lipkowitz, K. B. In Phosphazenes: A
Worldwide Insight; Gleria, M., De Jaeger, R., Eds.; Nova Science: New
York, 2004; Chapter 11, pp 237−255).
(38) Elias, A. J.; Shreeve, J. M. Adv. Inorg. Chem. 2001, 52, 335−358.
(39) Sirotin, I. S.; Bilichenko, Yu. V.; Suraeva, O. V.; Solodukhin, A.
N.; Kireev, V. V. Polym. Sci., Ser. B 2013, 55 (1−2), 63−68.
(40) Depree, C. V.; Ainscough, E. W.; Brodie, A. M.; Burrell, A. K.;
Lensink, C.; Nicholson, B. K. Polyhedron 2000, 19, 2101−2110.
(41) (a) Hirata, Y.; Matsumoto, K.; Takeuchi, T. Org. Mass Spectrom.
1978, 13, 111−112. (b) Horn, H. G. Z. Anorg. Allg. Chem. 1974, 406,
199−208. (c) Paddock, N. L.; Brion, C. E. J. Chem. Soc. A 1968, 388−
392. (d) Gleria, M.; Audisio, G.; Daolio, S.; Traldi, P.; Vecchi, E.
Macromolecules 1984, 17, 1230−1233.
(42) The reactions of the medium-sized rings with the superacids
HX/MX5 (M = Sb and P) will be the subject of an upcoming
publication.
(43) Doughill, M.; Paddock, N. L. J. Chem. Soc., Dalton Trans. 1974,
1022−1029.
(44) Ni, Y.; Lough, A. J.; Rheingold, A. L.; Manners, I. Angew. Chem.,
Int. Ed. 1995, 34, 998−1001.
(45) Calhoun, H. P.; Paddock, N. L.; Trotter, J. J. Chem. Soc., Dalton
Trans. 1976, 38−42.
(46) (a) Paddock, N. L.; Trotter, J.; Whitlow, S. H. J. Chem. Soc. A
1968, 2227−2236. (b) Allcock, H. R.; Dembek, A. A.; Mang, M. N.;
Riding, G. H.; Parvez, M.; Visscher, K. B. Inorg. Chem. 1992, 31,
2734−2739.
(47) Allcock, H. R. Phosphorus−Nitrogen Compounds; Academic
Press: New York, 1972; Chapter 3.
(48) (a) Chaplin, A. D.; Harrison, J. A.; Dyson, P. J. Inorg. Chem.
2005, 44, 8407−8417. (b) Bobrov, M. F.; Popova, G. V.; Tsirelson, V.
G. Russ. J. Phys. Chem. 2006, 80, 584−590. (c) Breza, M. Polyhedron
2003, 22, 3243−3248.
(49) (a) Marsh, W. C.; Trotter, J. J. Chem. Soc. A 1971, 169−173.
(b) Allcock, H. R.; Dembek, A. A.; Mang, M. N.; Riding, G. H.; Parvez,
M.; Visscher, K. B. Inorg. Chem. 1992, 31, 2734−2739.
(50) Wells, A. F. Structural Inorganic Chemistry, 5th Edition;
Clarendon Press: Oxford, U.K., 1984; Chapter 19.
(51) (a) Feakins, D.; Last, W. A.; Neemuchwala, N.; Shaw, R. A.
Chem. Ind. (London) 1963, 164−165. (b) Feakins, D.; Last, W. A.;
Shaw, R. A. J. Chem. Soc. 1964, 4464−4471. (c) Feakins, D.; Last, W.
A.; Neemuchwala, N.; Shaw, R. A. J. Chem. Soc. 1965, 2804−2811.
(52) Sun, H. J. Am. Chem. Soc. 1997, 119, 3611−3618.
(53) Tarazona, M. P.; Saiz, E. Polymer 2000, 41, 3337−3347.
(54) Harder, S. Coord. Chem. Rev. 1998, 176, 17−66.
(55) Paddock, N. L.; Rettig, S. J.; Trotter, J. Can. J. Chem. 1983, 61,
541−544.
(56) (a) Wiberg, K. B. J. Org. Chem. 2003, 68, 9322−932. (b) Martin,
S.; Thompson, A.; Coutsias, E. A.; Watson, J.-P. J. Chem. Phys. 2010,
132, 234115 (DOI: 10.1063/1.3445267).
8886
dx.doi.org/10.1021/ic500272b | Inorg. Chem. 2014, 53, 8874−8886