Elastomeric Polyphosphazenes with Phenoxy-Cyclotriphosphazene Side Groups

Article abstract of DOI:10.1021/acs.macromol.5b01892

New polymers with a phosphazene backbone and both 2,2,2-trifluoroethoxy- and phenoxy-functionalized cyclotriphosphazene substituents exist in three phases depending on the side group ratios. At low concentrations of the bulky substituents (up to ～7 mol %), the polymers are semicrystalline thermoplastics, with properties that are minor variations of poly[bis(2,2,2-trifluoroethoxy)phosphazene]. However, after the incorporation of between ～7 mol % and ～20 mol % of the bulky cyclic trimeric side groups, the polymers lose their semicrystalline properties and become amorphous elastomers. At still higher trimer loadings (>20 mol %) the materials develop gum-like behavior. The elastomeric phase appears to be generated by interdigitation or agglomeration of the bulky aryloxy-cyclotriphosphazene side groups, which act as quasi-physical cross-links between the polymer chains. The presence of these interactions allows the materials to experience high strain values before rupture (up to 1000%), and elastic recovery of more than 85% of the original dimensions when stressed up to 60% of the break elongation over four cycles. In addition, the chemical and physical nature of the substituents on the cyclic trimeric side groups alters the physical characteristics of the polymer in a way that provides a facile method to tune the properties.

Full text of DOI:10.1021/acs.macromol.5b01892

Article
pubs.acs.org/Macromolecules
Elastomeric Polyphosphazenes with Phenoxy−Cyclotriphosphazene
Side Groups
Tomasz Modzelewski,† Emily Wilts,† and Harry R. Allcock*
Department of Chemistry, The Pennsylvania State University University Park, Pennsylvania 16802, United States
S
* Supporting Information
ABSTRACT: New polymers with a phosphazene backbone and both 2,2,2-
trifluoroethoxy- and phenoxy-functionalized cyclotriphosphazene substituents exist in
three phases depending on the side group ratios. At low concentrations of the bulky
substituents (up to ∼7 mol %), the polymers are semicrystalline thermoplastics, with
properties that are minor variations of poly[bis(2,2,2-trifluoroethoxy)phosphazene].
However, after the incorporation of between ∼7 mol % and ∼20 mol % of the bulky
cyclic trimeric side groups, the polymers lose their semicrystalline properties and
become amorphous elastomers. At still higher trimer loadings (>20 mol %) the
materials develop gum-like behavior. The elastomeric phase appears to be generated by
interdigitation or agglomeration of the bulky aryloxy-cyclotriphosphazene side groups,
which act as quasi-physical cross-links between the polymer chains. The presence of
these interactions allows the materials to experience high strain values before rupture
(up to 1000%), and elastic recovery of more than 85% of the original dimensions when
stressed up to 60% of the break elongation over four cycles. In addition, the chemical
and physical nature of the substituents on the cyclic trimeric side groups alters the physical characteristics of the polymer in a way
that provides a facile method to tune the properties.
INTRODUCTION
■
The utility of a large part of polyphosphazene chemistry rests
on the ease of linkage of a wide variety of organic side groups to
the inorganic backbone.1−11 This has allowed the development
of a substantial number of polymers with properties that vary
from semicrystalline10,11 or amorphous thermoplastics, to low
Tg elastomers, hydrophobic films and fibers, polymer dyes,
water-soluble polymers, and biomedically useful materials.
Thus, the ease of chemical manipulation is one of the most
important features of this field. Particularly interesting is the
influence of bulky side groups on the physical properties of the
polymers
In a recent publication,12 we described the first members of a
novel class of polymers based on a polyphosphazene backbone
(0.7−22 mol %) of cyclotriphosphazene units each decorated
with phenoxy substituents. In principle, the aryloxy-function-
alized cyclophosphazene cosubstituents could increase the
interchain interactions in three ways. First, the increased steric
bulk and rigidity of the aryloxy units, compared to the
trifluoroethoxy counterparts, could generate stronger inter- and
intramolecular steric interactions. Second, the phenoxy groups
might undergo π−π interactions, which could further increase
the strength of the interchain forces. Third, the presence of
both aryloxy and trifluoroethoxy substituents may cause them
to phase separate, with the aryloxy units occupying clusters or
with 2,2,2-trifluoroethoxy side groups and a small percentage
(0.5−20 mol %) of bulky cyclotriphosphazene units which
themselves bear trifluoroethoxide units (structure 1). The bulky
cyclic trimeric side units in that system act as anchoring points
or interdigitation sites, and function as physical “cross-links” to
impart elastomeric properties to polymers that would normally
be either microcrystalline thermoplastics or noncrystalline,
nonelastomeric gums. Several questions arose from that study,
including the possible effects of linking even bulkier
cosubstituent groups to the polyphosphazene backbone and
the influence that larger side groups might have on the
elastomeric properties.
Thus, in this paper we report the synthesis of a new series of
phosphazene polymers (structure 2) that bear a majority of
2,2,2-trifluoroethoxy side groups plus small to medium ratios
Received: August 30, 2015
Revised: October 5, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.macromol.5b01892
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
a
Scheme 1. Synthesis of the Aryloxy-Functionalized Cyclotriphosphazene Side Group 5/5a
a
5a is the deprotonated unit after coupling to the polymer skeleton.
Scheme 2. Synthesis of Polymers 6−10
agglomerates that resemble the more traditional physical cross-
links in block copolymers.
removed the methyl ether unit and generated the required
monohydroxyl-functionalized 5. The structure of 5 was
1
confirmed by both H and 31P NMR spectroscopy.
Analysis of the newly synthesized macromolecules revealed
that the aryloxy-functionalized cyclotriphosphazene groups do
indeed have a significant influence on the mechanical properties
of the polymers. Specifically, the new polymers have a higher
Young’s modulus and tensile strength than series 1 polymers at
similar loadings. This change is evidence of increased
interactions between the bulky side units as well as a decrease
in overall polymer chain flexibility. Thus, in the following
sections the emphasis is on the similarities and differences
between polymers of types 1 and 2.
Synthesis of Polymers 6− 10. These polymers are
variations of structure 2 with five different side group ratios.
The synthesis of these species followed three-step procedures,
with the preparation of polymer 6 following a slightly different
procedure than the one used for polymers 7 − 10.
First, for polymers 7−10, which were designed to have 5−20
mol % of 5a, respectively, 95−80 mol % of the chlorine atoms
along the polymer backbone were replaced by the sodium salt
of trifluoroethanol to generate a partially substituted
intermediate. To favor a random distribution of the remaining
chlorine atoms, the sodium salt solution was added dropwise to
a vigorously stirred polymer solution at room temperature. This
procedure allowed the required number of chlorine atoms to be
left unreacted. Thus, their subsequent replacement by the
sodium salt of 5 would control the loading of the cyclic trimeric
groups along their backbone. The progress of these and future
steps were monitored using 31P NMR spectroscopy. The
reaction mixture was then treated with a 4-fold excess of the
sodium salt of 5 to replace the remaining chlorine atoms. This
large excess allowed a near-complete replacement of the
remaining chlorine atoms by 5. Finally, any remaining P−Cl
bonds, below the detection limit of 31P NMR signals, were
replaced by a second treatment of the polymer with the sodium
trifluoroethoxide. This protocol ensured the formation of
chlorine-free, un-cross-linked polymer.
RESULTS AND DISCUSSION
■
Synthesis of Aryloxy Functionalized Cyclotriphospha-
zene Side Group, 5. Compound 5 was synthesized using a
three step procedure (Scheme 1), which resembles the protocol
employed for the trifluoroethoxy-substituted counterpart.12
First, hexachlorocyclotriphosphazene was treated with the
sodium salt of 4-methoxyphenol to replace one of the six
chlorine atoms to generate intermediate 3. The progress of this
reaction and subsequent steps were monitored by 31P NMR
spectroscopy. Once the substitution was complete, the product
was purified extensively to eliminate any impurity with more
than one arylmethoxy group attached to the cyclotriphospha-
zene ring. Any multifunctional impurities would ultimately be a
source of covalent cross-linking during subsequent synthesis
steps.
The synthesis of polymer 6 followed a slight variation of the
above procedure due to the extremely low loading of 5. In the
first step only 90 mol % of the chlorine atoms were replaced
using sodium trifluoroethoxide. This was necessary due to the
After the monosubstituted intermediate 3 had been isolated
and purified, it was treated with an excess of sodium phenoxide
to replace the remaining five chlorine atoms on the
cyclotriphosphazene ring to give 4. Treatment with BBr3
B
DOI: 10.1021/acs.macromol.5b01892
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
a
Table 1. Characterization Data for Polymers 6−10
composition (mol %)
CTP
Mw
(kDa)
yield
(%)
polymer
1H NMR (ppm)
31P NMR (ppm)
Tg/T1 (°C)
−62.0/79.7
PDI
BTFP 4.43 (s)
−7.05 (s)
0
6
4.42 (s, 283.7 H), 6.54−7.33 (m,
8.73 (3.00 P), −7.52 (70.1 P)
0.7
−61.0 1.1
4400
2100
2000
1600
1100
1.95
2.35
2.13
1.79
1.87
51
18
67
41
19
29.00 H)
/60.9 1.6
7
4.57 (s, 28.64 H), 6.57−7.41 (m,
9.05 (3.00 P), −7.65 (8.40 P)
6.3
10.9
17.0
22.3
−21.4 0.9
29.00 H)
/56.5 4.3
8
4.39 (s, 16.32 H), 6.69−7.33 (m,
8.68 (3.00 P), −8.06 (3.11 P), −12.40
−19.2 0.9
−3.2 0.4
4.8 0.1
29.00 H)
(0.99 P)
9
4.44 (s, 13.22 H), 6.50−7.32 (29.00 8.68 (3.00 P), −8.66 (1.82 P), −12.42
H) (1.00 P)
10
4.48 (s, 6.67 H), 6.59−7.39 (29.00 8.77 (3.00 P), −8.97 (1.37 P), −12.53
H) (1.04 P)
a
CTP: Cyclotriphosphazene (5a). BTFP: poly(bis-2,2,2-trifluoroethoxy)phosphazene.
difficulty of preparing an intermediate with only 0.5 mol % of
the chlorine atoms remaining. By leaving 10 mol % unreacted
chlorine atoms, it was then possible to incorporate the precise
amount of 5a in the second step, and to then replace the
remaining chlorine atoms by trifluoroethoxy units to generate
compound 6.
interactions, which reduces torsional freedom and yields a
stiffer polymer.8 Apart from purely steric interactions the
possibility also exists that the aromatic nature of the
substituents on 2 may further limit polymer flexibility though
π−π interactions.
This is a similar trend to that found for polymers of type 1.12
Figure 1 shows a comparison of the Tg values as a function of
These synthetic procedures yielded fully substituted
polymers in moderate to good yields (15−67%) and with
high molecular weights (≥1000 kDa). The total concentration
of the cyclotriphosphazene side group was determined for each
polymer using both 31P and 1H NMR spectroscopy.
1
Representative examples of the 31P and H NMR spectra for
polymer 7 are provided in the Supporting Information.
Characterization data are provided in Table 1
Solubility and Surface Character. All the polymers
described here are soluble in tetrahydrofuran, acetone, methyl
ethyl ketone, and a number of other traditional organic
solvents, and showed little deviation from the solubility profile
of the parent [NP(OCH2CF3)2]n. The water contact angles for
all polymers are in the range of 82−97 deg, with no discernible
influence based on the concentration of 5a. This too reflects the
properties of the parent [NP(OCH2CF3)2]n. Thus, the bulky
phenoxy-functionalized cyclotriphosphazene groups have no
appreciable influence on the surface properties of these
polymers.
Thermal Transitions. Table 1 shows characterization data
for these polymers, with the DSC traces provided in the
Supporting Information. The Tg/T1 data are from DSC
experiments. The important trends are (1) the glass transition
temperatures rise as increasing numbers of the bulky side
groups are present along the chains, and (2) the T1 transition,
which is a well-known feature of the parent polymer
[NP(OCH2CF3)2]n,10,11 is detected in samples with a loading
of the bulky groups of less than ∼6−7 mol %. Above this
loading the T1 transition is no longer observed. As discussed in
our previous publications12,13 the observed decrease in the
intensity of the T1 transition, and its eventual disappearance,
with an increase in the concentration of the bulky cyclo-
triphosphazene side groups is due to disruption of chain
packing due to the cyclotrimeric side group’s steric bulk. Thus,
the presence or absence of the T1 transition can be used to
determine if the sample is microcrystalline or completely
amorphous, respectively. A further indication of the loss of the
T1 transition, and its correlation with the transition of the
polymers from semi crystalline to amorphous, is evident from
the X-ray scattering data (see later). The increase of the Tg is
almost certainly a consequence of steric restrictions to
backbone torsional motions caused by intra- and intermolecular
Figure 1. Glass transition temperatures of polyphosphazene
elastomers bearing phenoxy- and trifluoroethoxy12-functionalized
cyclotriphosphazene cosubstituents.
side group concentration for the polymers of types 1 and 2.
The Tg values measured for the polymers of type 2 are on
average 20−30 °C higher than those of their counterparts of
type 1 at comparable concentrations. Figure 2 shows a 3D
comparison of the size of the two bulky side groups, together
with the 2,2,2-trifluoroethoxy unit for comparison.
Although the Tg values of the polymers are influenced by the
bulk of the cyclotriphosphazene groups, the T1 transition
appears to be relatively insensitive to the size of the larger
substituents. Earlier reports have suggested that this transition
represents a conversion of the microcrystallites found in the
parent [NP(OCH2CF3)2]n to a secondary mesophase.10,13,14
Thus, because the larger side groups are associated with the
noncrystalline sections of the chains it is understandable that
changes in the size of the bulky side groups have a minimal
effect on the behavior of the crystallites.12 Thus, the presence
or absence of this transition is a simple indicator of whether the
polymer retains crystallinity or is amorphous. Only polymers 6
C
DOI: 10.1021/acs.macromol.5b01892
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Figure 2. 3D rendering of the 2,2,2-trifluoroethoxy- (left), 2,2,2-trifluoroethoxycyclotriphosphazene (middle), and phenoxycyclotriphosphazene
(right) side groups. These groups extend from the polyphosphazene backbone by ∼4.8, 14.8, and 16.5 Å, respectively.
and 7 show evidence of this transition, while polymers 8−10 do
not. This suggests that the loading of bulky groups required to
eliminate crystallinity (between 6 and 10 mol %) is fairly
constant regardless of the size or character of the bulky groups,
at least within this polymer system.
WAXS Data and Crystallinity. Additional evidence of the
change from semicrystalline to amorphous was provided by
wide-angle X-ray scattering (WAXS) analysis. Figure 3 shows
those of polymers 8−10. This provides additional confirmation
that polymers 6 and 7 are semicrystalline, while 8−10 are most
likely amorphous. These observations are further supported by
the DSC data which show the presence or absence of the
semicrystalline nature of these polymers though the presence or
absence of the T1 transition, respectively. All of these
observations reinforce the conclusion that the loading of the
bulky cosubstituents required to transform a semicrystalline
polymer, like [NP(OCH2CF3)2]n to an amorphous material
(5−10 mol %), is more a function of its concentration than its
size.
Second, although the steric bulk of the cosubstituent group
appears not to alter the loading at which the transition from
semicrystalline to amorphous occurs, it does influence the
degree to which the polymer chain packing is disrupted at
higher cosubstituent loadings. Specifically, the WAXS trace of
polymer 10, which contains 22.3 mol % of 5a, shows a nearly
complete elimination of the two main signals at 2θ = 9 and 21
deg. This is in contrast to its counterparts 112 and the oligo-p-
phenylene containing polyphosphazenes13 which retains
stronger crystalline signals at a similar bulky cosubstituent
loading. Thus, the larger size of 5 appears to have a greater
disruptive effect on chain packing than does its counterpart of
series 1 and the oligo-p-phenylene containing polyphospha-
zenes.
The properties may be analogous to those observed for more
traditional poly(dimethylsiloxane) (PDMS), poly-
(diethylsiloxane) (PDES) and poly(dipropylsiloxane)
(PDPS).19−21 Of these polymers, PDES and PDPS, show the
presence of a liquid-crystalline phase when heated above their
Tg, but below their Tm, while PDMS does not show this
transition. In these polymers, the ability of the longer side
groups in PDES and PDPS to become oriented may allow the
materials to show the presence of the liquid-crystalline phase,
while in the case of PDMS the side chains are too short to allow
these interactions to occur. A similar phenomenon may be
occurring in this phosphazene system, in which the bulky
cyclotriphosphazene side groups may interact with each other,
and allow a similar liquid-crystalline like phase to exist. This
could account for the retention of some of the order in the
samples as indicated by the decrease, but not complete
elimination, of the two main signals in the WAXS traces of
the polymers that contained higher cyclotriphosphazene
loadings.
Figure 3. WAXS patterns for poly(bis-trifluoroethoxy)phosphazene
(TFE) and polymers 6−10 (Note: the patterns for polymers 8 and 9
show very minor, sharp signals from traces of sodium chloride
impurity that are present even after extensive purification.)
the patterns collected for polymers 6−10 together with the
profile for the parent poly(bis-2,2,2-trifluoroethoxy)-
phosphazene. All the patterns reveal the presence of two
signals centered at 2θ ≈ 9 and 21 deg. The same signals are also
found for the previously studied polymers of series 112 as well
as the more recently studied oligo-p-phenylene containing
polyphosphazenes.13 On the basis of the earlier work, these
diffraction peaks represent the interchain spacing and the near-
cis−trans planar repeat unit lengths, respectively.14−18
The data suggest two main conclusions. First, the WAXS
patterns obtained for polymers 6, 7, and poly(bis-2,2,2-
trifluoroethoxy)phosphazene contain sharper signals than
D
DOI: 10.1021/acs.macromol.5b01892
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Mechanical Properties. The mechanical properties of
polymers 6−10 were examined using an Instron tensile testing
instrument. The data are summarized in Table 2, with
representative stress−strain curves shown in Figure 4.
domains, π−π stacking of the aryl groups, or interdigitation of
the cyclotriphosphazene units. All three influences could
generate properties that would normally be ascribed to covalent
or physical cross-links.
Once the microcrystallinity is eliminated by the incorpo-
ration of more than ∼10 mol % of side group 5a (polymers 8
and 9), the mechanical properties undergo a major change
which becomes manifest in the appearance of elastomeric
behavior. Thus, amorphous polymers 8 and 9 show a good
retention of tensile strength (2.2 and 2.1 MPa, respectively),
while their Young’s modulus decreases (0.24 and 0.48 MPa,
respectively). In addition, the elongation at break (1400 and
2200%, respectively) is two to three times larger than that
found for the semicrystalline 6 and 7. These changes are
attributable to the elimination of crystallinity, which is
confirmed by both WAXS and DSC analyses. In addition, the
tensile strength and Young’s modulus are two- and 10-fold
larger (respectively) than in the polymer 1 series for an
equivalent loading of the cyclo-trimeric side group. This
increased tensile strength and Young’s modulus probably
originate from the ability of the bulky side groups to coassociate
and form stronger interchain interactions than in series 1.
However, once the content of 5a reaches 22.6 mol %
(polymer 10), the polymers take on the morphology of self-
adhesive, low Tg gums, with a drastic loss of tensile strength
(0.47 MPa) and a further increase in the break elongation
(3,100%). These results are similar to those reported for the
polymers based on structure 1.12 The change in physical
behavior is accompanied by an almost complete elimination of
chain packing order, as detected in the WAXS data. This could
be a result of steric crowding of 5a units along the polymer
backbone, and the elimination of free volume needed to permit
interdigitation. Thus, to summarize, it appears that the
incorporation of ∼20 mol %, or more of bulky cosubstituent
groups causes polyphosphazenes to lose their elastomeric
properties and to behave like uncross-linked gums. Elasticity
can then be retained only by the presence of covalent cross-
links introduced, for example, by cross-coupling through
unsaturated cosubstituents.
Table 2. Mechanical Properties of Polymers 6−10
tensile
strength
(MPa)
break
elongation
(%)
Young’s
modulus
(MPa)
yield strength
(MPa)
polymer
6
5.6 0.5
3.9 0.3
2.2 0.3
2.1 0.1
0.47 0.04
1.8 0.1
1.0 0.1
440 40
720 80
19.5 2.3
12.2 2.7
0.24 0.05
0.48 0.15
0.42 0.15
7
8
0.05 0.01
0.04 0.01
0.03 0.01
1,400 120
2,200 70
3,100 300
9
10
Figure 4. Representative stress−strain curves for polymers 6−10 and
poly(bis-2,2,2-trifluoroethoxy)phosphazene (TFE).
The mechanical properties of the polymers follow the trends
seen in the DSC and X-ray results and fall into three main
groups at room temperatures: (a) semicrystalline (6 and 7), (b)
amorphous elastomers (8 and 9) and (c) low Tg gum (10).
These are similar results to those obtained for our previously
published systems, and may represent a general trend for other
polymers of this type.12,13
However, the main point is that the increase in both the
tensile strength and Young’s modulus of the elastomeric
polymers 8 and 9, compared to their counterparts of type 1,
illustrates that the mechanical properties can be controlled by
the size and structure of the bulky substituents, and that this
offers opportunities for the fine-tuning of useful properties.
Elastomeric Recovery. The elastomeric recovery of the
polymers was examined by elongation of dog-bone shaped
samples up to 60% of their break elongation, before retraction
back to their original length. This cycling was performed four
times, and the elastic recovery was calculated using the degree
of nonrecoverable permanent deformation remaining after the
fourth cycle. The data are summarized in Table 3. The degree
The semicrystalline polymers 6 and 7, which contain 0.7 and
6.3 mol % of 5a, respectively, are relatively tough materials.
This is a consequence of their microcrystallinity, which is
responsible for the highest tensile strength (3.9 to 5.6 MPa)
and Young’s modulus (12.2 to 19.5 MPa) and the shortest
elongation at break (440 to 720%). For polymer 6, these
properties are a minor variation of those found for its
counterpart of type 1, and indeed of the single-substituent
[NP(OCH2CF3)2]n. Thus, within this range, the major
contributing factor to the mechanical behavior is the semi-
crystalline structure, with a minimal influence by the small
amounts of the bulky cosubstituents. However, for polymer 7
the mechanical properties are changed significantly compared
to those for series 1. Specifically, the tensile strength and
Young’s modulus are increased by a factor of 2 and 100,
respectively, and the elongation at break is only half of the value
measured for its counterpart in series 1. This change almost
certainly originates from the onset of interactions between the
aryloxy-cyclotriphosphazene side groups. These interactions
could involve phase separation between aryloxy and fluorinated
Table 3. Elastic Recovery of Polymers 5−9 after Four
Elongation Cycles up to 60% of their Break Elongation
polymer
elastic recovery (%)
6
27
43
7
8
88
9
87
10
NA
E
DOI: 10.1021/acs.macromol.5b01892
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
vigorously stirred solution of hexachlorocyclotriphosphazene (300 g,
0.863 mol) in tetrahydrofuran (1.5 L), and the resulting reaction
mixture was stirred for 48 h at room temperature. Once the
substitution was complete, the solvent was removed under reduced
pressure and the resulting oil was purified by vacuum distillation at 170
°C and 10−2 mbar. The collected material contained 3 in 37% yield
with no multifunctional groups detected by 31P NMR. 31P NMR (145
MHz, CDCl3), δ (ppm): 22.9 (d, 2P), 13.4 (t, 1P).
to which the samples were capable of retracting to their original
shape correlates well with the concentration of the cyclo-
triphosphazene side group 5a. The elastomeric recovery
information confirms the conclusions described above.
Thus, the semicrystalline polymers 6 and 7 had the lowest
ability to retract after being elongated (27% and 43%,
respectively). This low elastic recovery is similar to that
measured for [NP(OCH2CF3)2]n (∼25%), and is almost
certainly a result of the semicrystalline character of the samples.
However, once the concentration of 5a is increased to above
∼7 mol %, the polymers become amorphous, and they show a
drastic increase in their degree of elastic recovery (>85%),
which resembles the behavior of polymers of series 1 at similar
cyclic trimeric side group loading. The ability of these materials
to retract back to almost their original shape exemplifies the
cross-link-like interactions caused by the cyclotriphosphazene
side groups 5a. However, once the concentration of 5a is higher
than ∼20 mol %, the polymers become self-adhesive gum-like
materials with a minimal ability to undergo elastic recovery.
This appears to be a result of the nearly complete loss of chain
orientation as indicated in the WAXS study described earlier,
and the probable decreased ability of the cyclotriphosphazene
side groups to interdigitate, thus leading to a decrease in
“crosslink” efficiency.
Synthesis of Cyclotriphosphazene Side Group 5. To a
suspension of NaH (60% w/w in mineral oil) (16.6 g, 0.416 mol)
in tetrahydrofuran (0.5 L) at 0 °C, was slowly added phenol (39.1 g,
0.416 mol). After the NaH was consumed, intermediate 3 (30.0 g,
0.070 mol) was added to the main reaction solution which was stirred
at reflux for 24 h The progress of the reaction was monitored using 31P
NMR spectroscopy. After this reaction was complete, the solvent was
removed under reduced pressure and the resulting oil was redissolved
in dichloromethane (0.5 L) and was washed with water (2×) and brine
(2×), before drying the organic phase over anhydrous magnesium
sulfate and removal of the solvent under reduced pressure. The
resulting product was centrifuged, and the mineral oil was removed
from the surface of 4, which was a viscous oil. The product was dried
further under high vacuum to remove any remaining moisture. Once
dry, 4 was redissolved in DCM (0.15 L), to which was added BBr3
(10.0 mL, 0.11 mol), and the solution was stirred at room temperature
overnight. The solution was then poured into water (0.2 L, 0 °C) and
the mixture was stirred vigorously for 30 min. The organic phase was
isolated, washed with 0.1 M sodium carbonate (2X), and brine (2X),
before removal of the solvent to afford 5 as an oil in 89.5% yield. 31P
1
EXPERIMENTAL SECTION
NMR (145 MHz, CDCl3) δ (ppm): 9.39 (m). H NMR (360 MHz,
■
CDCl3), δ (ppm): 7.26−7.10 (m, 15 H), 6.95 (d, 10 H), 6.76 (d, 2H),
6.64 (d, 2H), 5.66 (s, 1H).
Synthesis of Polymer 6. To a suspension of sodium hydride
(60% w/w in mineral oil) (2.97 g, 77.58 mmol) in tetrahydrofuran
(200 mL) was slowly added 2,2,2-trifluoroethanol (5.65 mL, 77.58
mmol). The resultant solution of sodium trifluoroethoxide was added
Reagents and Equipment for Polymer Synthesis. The
syntheses were carried out using standard Schlenk-line techniques
under a dry argon atmosphere, unless specified otherwise. All
glassware was dried in an oven at 140 °C overnight before use.
Tetrahydrofuran (THF) was dried using solvent purification columns.
2,2,2-Trifluoroethanol (Halocarbon) was distilled over sodium and
stored over molecular sieves (type 4A, EMD). Phenol (TCI) was
sublimed under reduced pressure (10−2 mbar) and stored under dry
argon before use. Dichloromethane (EMD), hexanes (EMD), sodium
carbonate (Alfa Aesar), sodium hydride (60% w/w in mineral oil,
Aldrich), acetone (EMD), methanol (EMD), and terta-n-butylammo-
nium nitrate (Alfa Aesar) were used as delivered. Dialysis was
accomplished using Spectra/Por molecular porous cellulose dialysis
membranes with molecular weight cutoffs of 12 000−14 000 Da.
Hexachlorocyclotriphosphazene (Fushimi Chemical Company, Japan)
was purified by recrystallization from hexanes followed by vacuum
sublimation (10−2 mbar) at 50 °C. Poly(dichlorophosphazene) was
synthesized by the uncatalyzed thermal ring-opening polymerization of
the purified hexachlorocyclotriphosphazene at 250 °C in an evacuated
and sealed Pyrex tube. The unreacted trimer was removed by vacuum
sublimation (10−2 mbar) at 50 °C, to leave the polymer as a colorless
elastomeric material.
dropwise to
a vigorously stirred solution of poly-
(dichlorophosphazene) (5.00 g, 43.1 mmol) in tetrahydrofuran (750
mL). The mixture was then stirred at room temperature for 24 h. The
progress of the substitution was monitored using 31P NMR
spectroscopy. In a separate vessel, which contained a suspension of
sodium hydride (60% w/w in mineral oil) (0.02 g, 0.43 mmol) in
tetrahydrofuran (100 mL), was added 5 (0.30 g, 0.43 mmol). After the
NaH was consumed, this solution was added to the polymer solution
and the reaction mixture was stirred at room temperature for 48 h. In a
separate vessel, 2,2,2-trifluoroethanol (0.62 mL, 8.63 mmol) was
added to a suspension of sodium hydride (60% w/w in mineral oil)
(0.33 g, 8.63 mmol) in tetrahydrofuran (100 mL). Again, after the
NaH was consumed, the mixture was added to the polymer solution.
This suspension was stirred at room temperature for 2 h, after which
time the product was concentrated under reduced pressure and was
precipitated into water. The solid polymer was redissolved in acetone
and dialyzed against 20% v/v methanol in acetone for 4 days. The
solution was then concentrated under reduced pressure and was
precipitated into water before being dried under high vacuum to afford
the final material in 51% yield. Characterization data for all polymers
are provided in Table 1.
Synthesis of Polymers 7−10. Polymers 7−10 were synthesized
by followed a slightly different procedure, with a representative
example provided for species 8. To a suspension of sodium hydride
(60% w/w in mineral oil) (1.18 g, 31.0 mmol) in tetrahydrofuran (100
mL) was added 2,2,2-trifluoroethanol (2.26 mL, 31.0 mmol). After the
NaH was consumed, the solution was added dropwise to a vigorously
stirred solution of poly(dichlorophosphazene) (2.00 g, 17.2 mmol) in
tetrahydrofuran (300 mL). The resultant mixture was stirred at room
temperature for 24 h while the progress of the reaction was monitored
using 31P NMR spectroscopy. In a separate vessel, which contained a
suspension of sodium hydride (60% w/w in mineral oil) (0.53 g, 13.8
mmol) in tetrahydrofuran (200 mL) was added compound 5 (9.77 g,
13.8 mmol). After hydrogen evolution had ceased, this mixture was
added to the polymer solution and the reaction mixture was stirred for
1H and 31P NMR spectroscopy made use of a Bruker AV-360
instrument operated at 360 and 145 MHz, respectively. 31P NMR
shifts are reported in ppm relative to 85% H3PO4 at 0 ppm. Thermal
transitions were measured using a TA Instruments Q10 DSC unit
operated at a heating rate of 10 °C/min, under a nitrogen stream with
a sample size of 10−15 mg. Data analysis was carried out with use of a
TA Instruments Universal Analysis 2000 Software. Polymer molecular
weights were obtained using gel permeation chromatography using a
Hewlett-Packard 1047A refractive index detector and two Phenom-
enex Phenogel linear 10 columns, eluted at a rate of 1 mL/min using a
10 mM solution of tetra-n-butylammonium nitrate in tetrahydrofuran.
The elution times were calibrated using polystyrene standards.
Synthesis of Cyclotriphosphazene Intermediate 3. To a
suspension of sodium hydride (60% w/w in mineral oil) (34.5 g, 0.863
mol) in tetrahydrofuran (0.25 L) at 0 °C was slowly added a solution
of 4-methoxyphenol (107 g, 0.863 mol) in tetrahydrofuran (0.25 L).
Once the addition was complete, the reaction mixture was allowed to
warm slowly to room temperature and was stirred overnight. After the
sodium hydride was consumed, the mixture was added dropwise to a
F
DOI: 10.1021/acs.macromol.5b01892
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
48 h at room temperature. In a separate vessel, 2,2,2-trifluoroethanol
(0.24 mL, 3.45 mmol) was added to a suspension of sodium hydride
(60% w/w in mineral oil) (0.13 g, 3.45 mmol) in tetrahydrofuran (100
mL). Once the NaH had been consumed, the sodium trifluoroethoxide
solution was added to the polymer solution and the reaction mixture
was stirred for an additional 2 h at room temperature. This ensured the
total replacement of the remaining chlorine. The product was then
concentrated under reduced pressure and precipitated into water. The
solid was redissolved in acetone and dialyzed against 20% v/v
methanol in acetone for 4 days. The resultant solution was
concentrated under reduced pressure, precipitated into water, and
the precipitate was dried under high vacuum to afford the final material
in 67% yield. Characterization data for all the polymers are provided in
Table 1
Evaluation of the Mechanical Properties of Polymers 6−10.
The polymers were solvent cast from acetone (1.5 g in 50 mL) in
Bytac coated dishes (6 × 6 cm). The solvent was allowed to evaporate
slowly in an acetone-saturated atmosphere. Once solid, the films were
dried under high vacuum for an additional 48 h. The resultant films
were cut into dog-bone shaped samples according to ASTM D-1708
using a PioneerDietics die. The samples were kept under vacuum
before testing.
Tensile Tests. Evaluation of mechanical properties was carried out
with the dog-bone shaped films described above. The samples were
pulled at a fixed cross-head speed of 100 mm/min until failure, using
an Instron 5866 Tensile Testing System at ambient temperature using
a 100 N load cell. The results are summarized in Table 2.
Comparison of Elasticity. The samples were cast and cut in a
similar manner as described above. They were pulled at a fixed cross-
head speed of 100 mm/min up to 60% of their previously determined
break elongation, and retracted back to their original length. This
cycling was performed 4 times per sample, and the percent recovery
after the fourth cycle is reported in Table 3.
Wide Angle X-ray Scattering. Wide angle X-ray scattering
(WAXS) patterns were obtained using a PANalytical X-Pert PRO
MPD Theta−Theta goniometer with Cu Kα radiation, and fixed slit
incidence (0.25 deg divergence, 0.50 deg antiscatter, specimen length
10 mm) and diffracted (0.25 deg antiscatter, 0.02 mm nickel filter)
optics. Sample films were cast using a method described in a previous
section. The film samples were placed in a zero background holder for
testing, and the data were collected at 45 kV and 40 mA from 5−50
deg in 2θ using a PIXcel detector in scanning mode with a PSD length
of 3.35 degrees 2θ and 255 channels for a duration of ∼20 min. The
resulting patterns were analyzed using Jade+10 software. The WAXS
patterns for polymers 5−9 are provided in Figure 3.
3D Structure Rendering. 3-Dimensional structures of 2,2,2-
trifluoroethoxy-, phenoxy-, and 2,2,2-trifluoroethoxy-functionalized
cyclotriphosphazene side groups (Figure 2) were generated using
ChemBio3D Ultra 11.0, utilizing the MM2 minimization function to
generate the lowest energy structure.
them to interact though phase separation or interdigitation,
possibly augmented by π−π interactions. However, larger
amounts of the bulky groups eliminate the elastic properties
and generate gums. Thus, the shape and size of the
cyclotriphosphazene cosubstituent groups are the main factors
that control the mechanical properties of the polymers. A very
wide range of organic groups can be linked to phosphazene
rings.2,16−19 As a result, the opportunities for property tuning
are almost unprecedented and provide a broad portfolio of
options for the further development of this field.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.macro-
mol.5b01892.
■
S
1H and 31P NMR spectra for polymer 7 and Differential
scanning calorimetry data for polymers 6−10 (PDF)
AUTHOR INFORMATION
Corresponding Author
*E-mail: hra1@psu.edu (H.R.A.).
■
Author Contributions
†These authors contributed equally to this manuscript.
Notes
The authors declare no competing financial interest.
REFERENCES
■
(1) Allcock, H. E.; Kwon, S.; Riding, G. H.; Fitzpatrick, R. J.; Bennett,
J. L. Biomaterials 1988, 9, 509.
(2) Mack, L. L.; Fitzpatrick, R. J.; Allcock, H. R. Langmuir 1997, 13,
2123.
(3) Maher, A. E.; Ambler, C. M.; Powell, E. S.; Allcock, H. R. J. Appl.
Polym. Sci. 2004, 92, 2569.
(4) Allcock, H. R.; Steely, L. B.; Singh, A. Polym. Int. 2006, 55, 621.
(5) Fei, S.-T.; Allcock, H. R. J. Power Sources 2010, 195, 2082.
(6) Allcock, H. R.; Morozowich, N. L. Polym. Chem. 2012, 3, 578.
(7) Allcock, H. R.; Kugel, R. L. J. Am. Chem. Soc. 1965, 87, 4216.
(8) Allcock, H. R. Chemistry and Applications of Polyphosphazenes;
John Wiley & Sons, Inc.: Hoboken, NJ, 2003.
(9) Phosphazenes A Worldwide Insight; Gleria, M., Jaeger, R. D., Eds.;
Nove Science Publishers, Inc.: New York, 2004.
(10) Magill, J. H.; Kojima, M. In Phosphazenes A Worldwide Insight;
Gleria, M., Jaeger, R. D., Eds.; Nova Science Publishers, Inc.: New
York, 2004; p 305.
(11) King, A.; Presnall, D.; Steely, L. B.; Allcock, H. R.; Wynne, K. J.
Polymer 2013, 54, 1123.
CONCLUSIONS
■
(12) Modzelewski, T.; Allcock, H. R. Macromolecules 2014, 47, 6776.
(13) Modzelewski, T.; Wonderling, N. M.; Allcock, H. R.
Macromolecules 2015, 48, 4882.
(14) Schneider, N. S.; Desper, C. R.; Singler, R. E. J. Appl. Polym. Sci.
1976, 20, 3087.
(15) Plate, N.; Shklyaruk, B.; Kuptsov, S.; Zadorin, A.; Antipov, E.
Macromol. Symp. 1996, 104, 33.
(16) Giglio, E.; Pompa, F.; Ripamonti, A. J. Polym. Sci. 1962, 59, 293.
(17) Allcock, H. R.; Arcus, R. A.; Stroh, E. G. Macromolecules 1980,
13, 919.
(18) Allcock, H. R. Phosphorus-Nitrogen Compounds: Cyclic, Linear,
and High Polymeric Systems; Academic Press: New York, 1972.
(19) Allen, G.; Lewis, C. J.; Todd, S. M. Polymer 1970, 11, 44.
(20) Godovsky, Y. K.; Makarova, N. N.; Papkov, V. S.; Kuzmin, N. N.
Makromol. Chem., Rapid Commun. 1985, 6, 443.
(21) Godovsky, Y. K.; Papkov, V. S.; Dusek, K. Adv. Polym. Sci. 1989,
88, 129.
New phosphazene polymers were synthesized with phenoxy-
functionalized cyclotriphosphazene side groups at various
concentrations (0.7−22 mol %), with the remaining sites
along the polymer backbone occupied by 2,2,2-trifluoroethoxy
groups. The incorporation of ∼10−22 mol % of the cyclo-
trimeric species disrupts the crystallinity characteristic of the
parent poly[bis(2,2,2-trifluoroethoxy)phosphazene] polymer,
and leads to the appearance of elastomeric properties. Because
of the aromatic nature of the substituent groups present on the
trimer units, and their incompatibility with the trifluoroethoxy
units along the polymer backbone, the new polymers are
tougher materials in terms of their tensile strength and Young’s
modulus compared to their counterparts with trifluoroethoxy-
functionalized cyclic trimeric units at comparable concen-
trations. This strengthening is attributed to the rigid aromatic
nature of the cyclotriphosphazene side groups, which allows
G
DOI: 10.1021/acs.macromol.5b01892
Macromolecules XXXX, XXX, XXX−XXX