Development of a one-pot in situ synthesis of poly(dichlorophosphazene) from PCl3

Article abstract of DOI:10.1021/ma0489772

A method for the one-pot synthesis of poly(dichlorophosphazene) from PCl₃ was developed. The method was a combination of synthesis of Cl₃P=NSiMe₃ monomer and living cationic polymerization method. The one-pot synthesis method provided an alternative way for the preparation of [N=PCl₂]_n. The results show that the yields of moisture stable derivative polymers [N=P(OCH₂CF ₃)₂]_n are in 40-50% range starting from PCl₃.

Full text of DOI:10.1021/ma0489772

Macromolecules 2005, 38, 643-645
643
The living cationic condensation of phosphoranimine
Cl3PdNSiMe3 at room temperature can produce
[NdPCl2]n with controlled molecular weight and narrow
polydispersities and offers possibilities to synthesize
block copolymers.8 This method (Scheme 1) represents
a significant advance over the previous methods to
prepare [NdPCl2]n. The scientific and technological
importance of the living polymerization from Cl3Pd
NSiMe3 presents great incentives to enhance the yield
of the highly reactive monomer, which was initially
Development of a One-Pot in Situ Synthesis
of Poly(dichlorophosphazene) from PCl3
Bin Wang*
Departments of Chemistry & Physics and Chemical
Engineering, Lamar University, PO Box 10022,
Beaumont, Texas 77710
Received May 24, 2004
Revised Manuscript Received September 1, 2004
Introduction
produced from the reaction of PCl5 with either LiN-
(SiMe3)2
or N(SiMe3)3.12c These methods of making
12a,b
Polyphosphazenes are inorganic polymers based on
the repeating unit [NdPR2]n, where R can be organic
or organometallic groups.1 These polymers incorporate
a large class of macromolecules with a variety of useful
properties, which are determined primarily by the side
groups attached to the polymer backbone. Several
methods have been developed for the synthesis of
polyphosphazenes. Ring-opening polymerization of [Nd
PCl2]3 at 250 °C can afford poly(dichlorophosphazene),
[NdPCl2]n.2 Macromolecular substitution of [NdPCl2]n
with organic groups provides hydrolytically stable poly-
(organophosphazenes) with alkoxy, aryloxy, or amino
side groups.3 Another general method of polyphos-
phazene synthesis is through condensation polymeri-
zation. Thermal condensation polymerization of pre-
formed alkyl- or aryl-substituted phosphoranimines can
produce poly(alkyl arylphosphazenes) which cannot be
obtained via the macromolecular substitution route.4
Other monomers, viz. alkoxy- or aryloxy-substituted
phosphoranimines,5 OdPCl2-NdPCl3,6 phosphinoazides,7
and Cl3PdNSiMe3,8 have been employed to produce
polyphosphazenes via condensation polymerization.
Among the methods for preparing poly(organophos-
phazenes), the most well-developed pathway is the
macromolecular substitution of [NdPCl2]n. This route
provides access to several hundred kinds of polymers
and is commercially used.
Practical, economical methods for the synthesis of
[NdPCl2]n are highly valuable. Some recent progress
in synthesis of [NdPCl2]n particularly focuses on the so-
called one-pot method for preferable industrial applica-
tions. At molten state, the mixture of PCl5 and NH4SO4
forms OdPCl2-NdPCl3 at ∼165 °C; continuing heating
of this species at ∼225 °C to distill off OPCl3 can lead
to the formation of [NdPCl2]n.9 Precaution has to be
taken when heating the mixture at high temperatures
so as to avoid cross-linking of [NdPCl2]n,10 which leads
to uncharacterized and useless products. In another
development, a 1,2,4-trichlorobenzene solution of PCl5
and NH4Cl is heated to reflux (∼214 °C), and high
molecular weight [NdPCl2]n is produced in solution.11
However, sublimation of PCl5 at temperatures above
180 °C causes blockage of reaction vessels, thus pre-
venting scale-up attempts of this route. Carriedo et al.
have experienced explosion hazards due to this subli-
mation-caused blockage problem. Also, use of chlori-
nated solvents can be problematic due to environmental
concerns. Therefore, neither of the aforementioned one-
pot synthesis methods has been shown to be a poten-
tially satisfactory route for large-scale, industrial ap-
plications.
Cl3PdNSiMe3 from PCl5 give relatively low product
yields, and the reason seems to be related to the
polymerization mechanism (Scheme 1). PCl5 is a known
initiator for the polymerization of Cl3PdNSiMe3; thus,
any product formed in solution would be in immediate
contact with PCl5 and instant, concurrent reaction of
the monomer into oligomeric or cyclic phosphazenes
would be inevitable. Also, a polymerization inhibitor
ClN(SiMe3)2 is often formed; this species needs to be
separated via additional steps from Cl3PdNSiMe3 to
afford suitable monomer for polymerization.8b
To circumvent this concurrent polymerization di-
lemma, a new method for synthesizing Cl3PdNSiMe3
avoids the use of PCl5.13 Starting from PCl3, a stable
phosphine intermediate Cl2P-N(SiMe3)2 is prepared in
solution and subsequently oxidized by SO2Cl2 to afford
Cl3PdNSiMe3 (Scheme 2). 31P NMR monitoring is used
to ensure that the sequential formations of both Cl2P-
N(SiMe3)2 and Cl3PdNSiMe3 are clean processes. There-
fore, by excluding PCl5 from the reactants, the degree
of concurrent polymerization is greatly reduced. Also,
8b
this route does not generate ClN(SiMe3)2 so the
additional purification operation is not required. Phos-
phoranimine has been isolated at high yields (>80%)
on large laboratory scales (>0.3 mol).
Because of intriguing recent progress in one-pot
synthesis of [NdPCl2]n,9,11 a more robust preparation
method of [NdPCl2]n from commercially available ma-
terials becomes interesting to study. Of particular
interest is to study the possibility of combining the two
well-established techniques: living cationic polymeri-
zation for [NdPCl2]n and the new method to synthesize
Cl3PdNSiMe3. The objective of such an approach is to
construct a coherent, one-pot fashion synthesis route,
which involves room temperature solution synthesis of
[NdPCl2]n without vacuum distillation isolation of the
monomer. Conceptually, this kind of new synthesis
method will present great potential for large-scale,
industrial polyphosphazene preparations. The experi-
mental details of the method development of such an
approach are discussed below.
Results and Discussion
Literature Examination of the Cl3PdNSiMe3
and [NdPCl2]n Preparation Methods. Recently, a
few reports14 have used Cl3PdNSiMe3 synthesized via
the improved, high-yield method13 as a precursor.
However, this highly versatile phosphoranimine is
isolated through stepwise, variable temperature vacuum
distillation. In a typical experiment, after the LiCl is
filtered, the solvent (e.g., Et2O) and ClSiMe3 are distilled
off at 0 °C under ∼20 mmHg. The phosphoranimine is
* E-mail: wangbx@hal.lamar.edu.
10.1021/ma0489772 CCC: $30.25 © 2005 American Chemical Society
Published on Web 12/24/2004

644 Notes
Scheme 1. Living, Cationic Polymerization for
Macromolecules, Vol. 38, No. 2, 2005
Scheme 3. One-Pot in Situ Synthesis of [NdPCl2]n
[NdPCl2]n and Preparation of Cl3PdNSiMe3 from PCl5
from PCl3
Scheme 2. Preparation of Cl3PdNSiMe3 from PCl3
Table 1. Results of the One-Pot Synthesis of [NdPCl2]n
from PCl3
then collected by distillation at room temperature under
∼0.1 mmHg. The isolation step is often the most likely
factor for low product yields in laboratory preparations,
and it also appears less favored from industrial view-
points compared to procedures requiring no vacuum
distillation. After distillation, a yellow gellike residue
is often observed left in the vessel. 31P NMR of this
residue has revealed a single peak around -18 ppm,
indicating the formation of [NdPCl2]n.
time (before PCl3, LiN(SiMe3)2, PCl5,
b
run adding PCl5) mmol
mmol
mmol
Mw
PDIb
1
N/A
218
220
0
103 300 2.35
59 600 1.26
49 600 1.24
20 900 1.61
2
3
4
5
1 ha
1 h
10.3
10.3
8.2
0.51
0.17
8.1
8.3
6 h
8.3
0.41 124 000 2.14
3.98 245 100 3.95
2 days
80.2
81.3
a Stirred at 0 °C. b Mw and PDI are for [NdP(OCH2CF3)2]n.
During the preparation of Cl3PdNSiMe3, major byprod-
ucts include ClSiMe3, LiCl, and gaseous SO2 that
escapes from the reaction medium.13 After purification,
Cl3PdNSiMe3 can begin condensation polymerization
when initiated by a Lewis acid such as PCl5. Therefore,
it is interesting to examine whether Cl3PdNSiMe3
prepared in situ can start polymerization in the pres-
ence of all of these byproducts in the same reaction
medium. When the monomer is prepared, low boiling
point solvents such as ether and pentane have been
chosen to facilitate the distillation process for better
monomer yield. In polymerization reactions, CH2Cl2,
toluene, benzene, and dioxane have been employed.15
Toluene appears to be the most economical and feasible
to handle among the solvents used.
Polymerization of Cl3PdNSiMe3 Prepared in
Situ from PCl3. In toluene, Cl2P-N(SiMe3)2 was
prepared by reacting PCl3 and LiN(SiMe3)2 (which can
be acquired from commercial sources or prepared from
HN(SiMe3)2 and n-BuLi) and confirmed by 31P NMR.
This compound was stable in solution and had been
stored in the reaction medium for a few days without
decomposition. Addition of SO2Cl2 to the solution con-
taining Cl2P-N(SiMe3)2 led to the formation of Cl3Pd
NSiMe3. The reactivity of this phosphoranimine in the
same reaction medium was tested consequently. The
mixture, mainly containing Cl3PdNSiMe3, ClSiMe3, and
LiCl, was added with a trace amount of PCl5 or no PCl5
at all. Stirring overnight at room temperature produced
a product that demonstrated one sharp 31P NMR
resonance (-17.6 ppm) characteristic of polymer 1
(Scheme 3). From the observations, the multistep con-
version of PCl3 to [NdPCl2]n was essentially quantita-
tive in one reaction medium. No inhibition of polymer-
ization had been observed. The question of the role of
PCl5 in the in situ polymerization was revealed to some
degree through molecular weight measurements (vide
infra). In our laboratory, [NdPCl2]n has been routinely
prepared through this one-pot method on scales from a
few to a few hundred millimoles. After filtration and
solvent removal, [NdPCl2]n can be subsequently treated
to produce other valuable derivatives, stored in bulk in
an inert atmosphere, or preserved in a diglyme solu-
tion16 for future modification.
Molecular Weight Measurements of the Deriva-
tive Polymers [NdP(OCH2CF3)2]n (2). To examine
the properties of the [NdPCl2]n formed through the one-
pot method, nucleophilic substitution was performed to
afford hydrolytically stable derivatives. Polymers 1 were
treated with excess NaOCH2CF3, and the resultant
species yielded a 31P NMR singlet characteristic of the
known polymer [NdP(OCH2CF3)2]n (2) (Scheme 3).
Starting from PCl3, the yields of [NdP(OCH2CF3)2]n are
in the 40-50% range (see Experimental Section). Gel
permeation chromatography (GPC) analysis of a few
examples of [NdP(OCH2CF3)2]n indicated that all poly-
mers possessed high molecular weight fractions only
(Table 1). The initiation of polymerization in the mixture
was analyzed. In run 1, the Cl3PdNSiMe3 mixture was
left stirring at room temperature overnight to form [Nd
PCl2]n. The resultant [NdP(OCH2CF3)2]n demonstrated
a bimodal distribution. This suggests that there are at
least two species in the reaction medium to have
initiated polymerization independently. To prevent the
spontaneously initiated polymerization by unspecified
species, the formed Cl3PdNSiMe3 was kept at 0 °C
before addition of PCl5 (run 2). The derivative [Nd
P(OCH2CF3)2]n demonstrated a relatively narrow PDI.
When the Cl3PdNSiMe3 mixture was left stirring at
room temperature for ∼1 h (run 3) before addition of
PCl5, the resultant [NdP(OCH2CF3)2]n showed a little
wider molecular weight distribution (1.61 vs 1.24 from
run 2). When the monomer was left stirring at room
temperature for ∼6 h (run 4) or left standing for 2 d
(run 5) before addition of PCl5, the resultant [Nd
P(OCH2CF3)2]n showed a broader molecular weight
distribution. In total, we have 22 runs under the
reaction conditions outlined above, and all gave similar
results. Only a few examples were sent for GPC analysis
with the results shown in Table 1. PCl5 seems to have
played a very competitive role among the initiators.
Promptly added PCl5 provides better control over poly-
mer molecular weight distribution. In general, because

Macromolecules, Vol. 38, No. 2, 2005
Notes 645
dropwise over 10 min to this suspension at 0 °C. The reaction
was allowed to proceed at 0 °C for ∼1 h. Cl3PdNSiMe3 (-54.5
ppm, toluene/CDCl3) was the only product present by 31P NMR.
PCl5 (106 mg, 0.51 mmol) was then added, and the resultant
mixture was stirred overnight at room temperature. [NdPCl2]n
was observed from 31P NMR (-17.6 ppm, toluene/CDCl3). The
mixture was then filtered through Celite (dried at ∼110 °C
for >48 h prior to use), which was then washed with toluene
(2 × 5 mL). The volatiles from the resulting pale yellow filtrate
were removed under reduced pressure to give a yellow gellike
solid. THF (∼40 mL) was used to dissolve the solid. To this
resulting solution, 10 mL of 2.5 M NaOCH2CF3 (prepared from
NaH and HOCH2CF3 in dioxane) was added, and the mixture
was stirred overnight at room temperature and [NdP(OCH2-
CF3)2]n was the only product according to 31P NMR (-7.2 ppm,
THF/CDCl3). The reaction mixture was concentrated by rotary
evaporation, and the polymer was purified by multiple pre-
cipitations into acidified water (pH ∼ 5) and hexanes followed
by drying on a vacuum line. Yield ) 1.0 g (41%).
of the competition among various initiation species, this
route does not provide the same level of control over
molecular weight (PDI < 1.1) as evidenced in the living
polymerization route starting with purified Cl3Pd
NSiMe3.8 However, polyphosphazenes prepared through
the traditional ring-opening mechanism, which possess
wide molecular weight distribution (∼19),2 have gained
significant industrial applications. This new one-pot
method has prepared [NdPCl2]n with narrower molec-
ular weight distribution than that prepared via the ring-
opening route. Therefore, the new method could have
similar industrial application potentials as the tradi-
tional ring-opening technique regarding molecular weight
distributions.
Conclusion
We report a new route to [NdPCl2]n based on a
convenient combination of the recently reported im-
proved synthesis of the Cl3PdNSiMe3 monomer and the
living cationic polymerization method for this monomer.
The new route proves to be a beneficial alternate
method to prepare [NdPCl2]n. Starting from PCl3, this
route proceeds through a multistep, quantitative con-
version, one-pot fashion under ambient temperature and
pressure to synthesize [NdPCl2]n. The yields of the
moisture stable derivative polymers [NdP(OCH2CF3)2]n
are in 40-50% range starting from PCl3. The whole
process does not require demanding vacuum distillation
of the highly reactive monomer and thus can be viewed
as a more cost-effective alternate to the living cationic
polymerization route using purified monomer. The
conditions used in this study are likely to be modified
for large-scale, industrial applications; therefore, it
possesses distinct advantages over other one-pot syn-
thesis routes, which face acute scale-up problems. This
method warrants favorable applications for both indus-
trial and laboratory preparation of polyphosphazenes.
Acknowledgment. We sincerely thank A. Bartole-
Scott (University of Toronto) for performing GPC, Prof.
S. P. Fearnley for using the laboratory, Prof. J. L.
Gossage for helpful discussions, and Lamar University
for financial support (REG #214302).
References and Notes
(1) Mark, J. E.; Allcock, H. R.; West, R. Inorganic Polymers;
Prentice Hall: Englewood Cliffs, NJ, 1992; Chapter 3.
(2) Allcock, H. R.; Kugel, R. L. J. Am. Chem. Soc. 1965, 87,
4216-4217.
(3) Allcock, H. R.; Kugel, R. L.; Valan, K. J. Inorg. Chem. 1966,
6, 1709-1715. Allcock, H. R.; Kugel, R. L. Inorg. Chem.
1966, 6, 1716-1718.
(4) Wisian-Neilson, P.; Neilson, R. H. J. Am. Chem. Soc. 1980,
102, 2848-2849. Neilson, R. H.; Wisian-Neilson, P. Chem.
Rev. 1988, 88, 541-562.
(5) Montague, R. A.; Matyjaszewski, K. J. Am. Chem. Soc. 1990,
112, 6721-6723. White, M. L.; Matyjaszewski, K. J. Polym.
Sci., Part A: Polym. Chem. 1996, 34, 277-289.
(6) Abouchaccra, T.; Helioui, M.; Puskaric, E.; De Jaeger, R.;
Heubel, J. J. Chem. Res., Synop. 1981, 230-231. D’Halluin,
G.; De Jaeger, R.; Chambretter, J. P.; Pottin, Ph. Macro-
molecules 1992, 25, 1254-1258.
Experimental Section
Materials and Equipment. LiN(SiMe3)2 (Aldrich, 97%),
SO2Cl2 (Aldrich, 97%), PCl3 (Eastman, 98%), PCl5 (Aldrich,
95%), NaH (Aldrich, 60% dispersion), HOCH2CF3 (Aldrich,
99%), n-BuLi (Acros, 2.5 M in hexanes), and HN(SiMe3)2
(Avocado Research Chemicals Ltd., 98%) were used as re-
ceived. THF was distilled over sodium and benzophenone prior
to use, and dioxane and toluene were distilled over CaH2. The
reactions were performed using standard Schlenk techniques
under an atmosphere of nitrogen (Aeriform). 31P{1H} spectra
were recorded on a Bruker AC250 NMR spectrometer operated
at 101.1 MHz, or on a Bruker EFT90 NMR spectrometer
operated at 36.4 MHz, with the chemical shifts externally
referenced to 85% phosphoric acid.
Molecular weight estimates were obtained using gel per-
meation chromatography (GPC) using a Waters Associates
2690 separations module equipped with a column heater,
Ultrastyragel columns with pore sizes of 103-105 Å, in-line
degasser, and a differential refractometer. The GPC possesses
a triple detection system (refractive index, light scattering,
viscosity) such that absolute molecular weights are reported
for homopolymers. A flow rate of 1.0 mL/min was used, and
the eluent was THF with 0.1% n-Bu4NBr (w/w).
A Typical Procedure for the Synthesis of [NdP-
(OCH2CF3)2]n from PCl3. LiN(SiMe3)2 (1.73 g, 10.3 mmol) was
dissolved in 40 mL of toluene, and the solution was cooled to
0 °C. PCl3 (0.9 mL, 10.3 mmol) was then added dropwise over
10 min. The resulting mixture was stirred for 30 min at the
same temperature followed by stirring at room temperature
for ∼1 h, giving a white suspension. 31P NMR revealed
complete conversion of PCl3 to Cl2P-N(SiMe3)2 (188.5 ppm,
toluene/CDCl3). SO2Cl2 (0.85 mL, 10.5 mmol) was then added
(7) Paciorek, K. L.; Kratzer, R. Inorg. Chem. 1964, 3, 594-595.
Franz, U.; Nuyken, O.; Matyjaszewski, K. Macromolecules
1993, 26, 3723-3725. Kojima, M.; Magill, J. H.; Cypryk, M.;
White, M. L.; Franz, U.; Matyjaszewski, K. Macromol.
Chem. Phys. 1994, 195, 1823-1842.
(8) (a) Honeyman, C. H.; Manners, I.; Morrissey, C. T.; Allcock,
H. R. J. Am. Chem. Soc. 1995, 117, 7035-7036. (b) Allcock,
H. R.; Crane, C. A.; Morrissey, C. T.; Nelson, J. M.; Reeves,
S. D.; Honeyman, C. H.; Manners, I. Macromolecules 1996,
29, 7740-7747.
(9) Allen, C. W.; Hneihen, A. S.; Peterson, E. S. US Patent 6,-
309,619, 2001.
(10) Allen, C. W., personal communications.
(11) Carriedo, G. A.; Alonso, F. J. G.; Go´mez-Elipe, P.; Fidalgo,
J. I.; AÄ lvarez, J. L. G.; Presa-Soto, A. Chem.sEur. J. 2003,
9, 3833-3836.
(12) (a) Niecke, E.; Bitter, W. Inorg. Nucl. Chem. Lett. 1973, 9,
127-129. (b) Honeyman, C. H.; Lough, A. J.; Manners, I.
Inorg. Chem. 1994, 33, 2988-2993. (c) Allcock, H. R.; Crane,
C. A.; Morrissey, C. T.; Olshavsky, M. A. Inorg. Chem. 1999,
38, 280-283.
(13) Wang, B.; Rivard, E.; Manners, I. Inorg. Chem. 2002, 41,
1690-1691.
(14) Rivard, E.; Huynh, K.; Lough, A. J.; Manners, I. J. Am.
Chem. Soc. 2004, 126, 2286-2287. Rivard, E.; Lough, A. J.;
Chivers, T.; Manners, I. Inorg. Chem. 2004, 43, 802-811.
Rivard, E.; Lough, A. J.; Manners, I. Inorg. Chem. 2004,
43, 2765-2767.
(15) Allcock, H. R.; Reeves, S. D.; de Denus, C. R.; Crane, C. A.
Macromolecules 2001, 34, 748-754.
(16) Andrianov, A. K.; Chen, J.; LeGolvan, M. P. Macromolecules
2004, 37, 414-420.
MA0489772