High energy density materials from azido cyclophosphazenes

Article abstract of DOI:10.1039/b510924h

Azido substituted cyclophosphazenes were prepared and their standard heats of formation were calculated based on experimentally determined heats of combustion. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b510924h

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
High energy density materials from azido cyclophosphazenes{
K. Muralidharan, Bamidele A. Omotowa, Brendan Twamley, Crystal Piekarski and Jean9ne M. Shreeve*
Received (in Berkeley, CA, USA) 1st August 2005, Accepted 2nd September 2005
First published as an Advance Article on the web 20th September 2005
DOI: 10.1039/b510924h
obtained from the reaction of sodium azide with bis[2,4-
(dibutylamino)]hexachlorocyclophosphazene tetramer, is given.
Increasing the number of nitrogen atoms in heterocycles results
in considerable gain in the standard enthalpy of formation in the
resulting compounds.11 Each phosphorus atom in a cyclophos-
phazene ring can typically accommodate two azido or amino
groups. With the aim of preparing high energy materials based
on cyclophosphazene azides, we have reacted N3P3Cl6 (1),
N3P3Cl5(NMe2) (2)12 and N3P3Cl2[1,1-(NH2)2] (3)13 with sodium
azide. Reactions at 25 uC in acetone led to complete replacement
of chlorine by azide leading to the formation of N3P3(N3)6 (4),14
N3P3(NMe2)(N3)5 (5)15 and N3P3(N3)4[1,1-(NH2)2](6)16 (Scheme 1).
The reaction mixture was filtered through filter paper and the
solvent evaporated to leave the spectroscopically pure product.
Compounds 4 and 5 are liquids, while at 25 uC 6 is a solid. Each
was characterized by NMR, HRMS, and infrared spectroscopy.
Compound 4 was reported to be highly friction sensitive and
was known to detonate without warning.14 This has been our
experience. However, the bis(dimethylamino) (5) and diamino (6)
compounds are substantially more stable. Introduction of smaller
amounts of hydrogen and carbon enhances the good oxygen
balance and reduces sensitivity to detonate due to friction. As a
result, 5 and 6 were significantly easier to handle than 4 and, in our
hands, there have been no signs of instability. The density of 5 and
6 are 1.67 and 1.79 g cm23, respectively.
Azido substituted cyclophosphazenes were prepared and their
standard heats of formation were calculated based on
experimentally determined heats of combustion.
Compounds that contain azide are most often highly energetic and
some tend to be unstable. However, over the last years, relatively
stable azido substituted carbon-nitrogen heterocycles have been
reported with densities .1.7 g cm23 and with very high heats
of formation,1 e.g., 4,49,6,69-tetra(azido)azo-1,3,5-triazine with
DfHu 5 +6164 kJ kg21. The high heats of formation arise from
the large number of inherently energetic N–N and C–N bonds.
This energy source is in contrast to that for the classic energetic
materials such as HMX (1, 3, 5, 7-tetranitro-1, 3, 5, 7-tetra-
azacyclooctane) and TNT (trinitrotoluene),2 where energy is
obtained from oxidation of the carbon backbone. In the former
cases, because a higher percentage of the decomposition products
will be dinitrogen, these nitrogen-rich compounds tend to be
promising high energetic materials and may be significantly more
environmentally benign.3
Despite their high enthalpies of formation and accompanying
tendency to be sensitive to friction, impact and heat, the introduc-
tion of amino groups into the nitrogen rich organic heterocycles
appeared to enhance their thermal stability.4 However, until now,
most research is based on compounds with both carbon-azido and
carbon-amino or nitrogen-amino substitutents.5 Yet the chemistry
and properties of energetic materials with P-azido and P-amino
moieties have been discussed only rarely in the scientific literature.6
The spirocyclic compounds of cyclophosphazene trimers and
tetramers were suggested to have higher energy compared to
their ansa isomers.7 A few reports have suggested use of inorganic
heterocycles such as spirocyclic fluorophosphazenes8 and azido
substituted cyclophosphazenes9 as high energy materials. There is
no evidence to support the high energy of these cyclophosphazene
compounds in the form of energy calculations based on experi-
mentally obtained thermodynamic data.10 The focus of the present
work is the syntheses of energetic materials based on cyclophos-
phazenes which may offer alternative energetic compositions.
For the first time, we report calculated enthalpies of formation
for cyclophosphazene azides based on experimentally determined
heats of combustion. While there are a few reports of azido sub-
stituted cyclophosphazenes, none of these azido substituted cyclo-
phosphazene tetramers has been structured crystallographically.
In this work, the crystal structure of an azidocyclophosphazene,
Recent results from our group17 and others18 have shown the
application of low-melting salts prepared from triazole and
tetrazole as high-energy materials. But, attempted quaternization
reactions of 5, 6 and 8 (vide infra) with methyl iodide and other
electrophiles were unsuccessful. The electron withdrawing nature
of cyclophosphazenes may account for this failure.19
The standard enthalpy of formation of an energetic compound
is a very important characteristic which can be calculated from
heats of combustion using Hess thermochemical equations. The
standard enthalpies of formation of the chlorophosphazene trimer
and tetramer were reported as 2811.33 and 21083.03 kJ mol21
respectively.20 In
calorimetric experiment in an oxygen
,
a
Department of Chemistry, University of Idaho, Moscow, Idaho, 83844-
2343, USA. E-mail: jshreeve@uidaho.edu
{ Electronic supplementary information (ESI) available: Experimental
details of 4 and 7, crystal structure, chair conformation and CIF for 8. See
http://dx.doi.org/10.1039/b510924h
Scheme 1
This journal is ß The Royal Society of Chemistry 2005
Chem. Commun., 2005, 5193–5195 | 5193

View Article Online
atmosphere, the cyclophosphazenes were reported to burn
according to eqn (1).20 Based on this, 4–6 were assumed to burn
in the presence of oxygen according to eqns (2)–(4), respectively. In
order to confirm the formation of P4O10, the solid obtained after
burning was dissolved in D2O and the 31P NMR spectrum was
recorded. The spectrum showed only a strong peak at 0 ppm
indicating the formation of phosphoric acid.
(PNCl2)n(s) + 5n/4 O2 (g) A n/4 P4O10 (s) + n Cl2 (g)
(1)
Scheme 2
[P3N3(N3)6] (s) + 14/4 O2 (g) A 3/4 P4O10 (s) + 42/4 N2 (g) (2)
for P3N3Cl6.20 The energy released on burning of one kg of a
compound is an important parameter to be considered for
energetic applications. Hence, it is significant to compare the
heats of formation in kJ kg21 as is shown in Table 1. These
values indicate that the compounds 4–6 would expect to release a
considerable amount of energy on burning, thus may potentially
have interesting uses.
[P3N3(N3)5(NMe2)] (s) + 29/4 O2 (g) A
3/4 P4O10(s) + 38/4 N2 (g) + 2 CO2 + 12 H2O (l)
(3)
[P3N3(N3)4(NH2)2]2] (s) + 17/4 O2 (g) A
3/4 P4O10 (s) + 34/4 N2 (g) + 4 H2O (l)
(4)
The standard molar enthalpy of combustion (DcHmu) of the
samples can be derived from the constant volume combustion
energy (DcUm) by means of the following equation:
Based on the opportunity to introduce additional azide
functionalities into a single stable moiety, we have synthesized
N4P4Cl6[2,6-(t-BuNH)]2 (7) according to the literature procedure
from the reaction of N4P4Cl8 with n-dibutylamine.22 When the
latter was further reacted with sodium azide in acetonitrile at 25 uC,
N4P4Cl2(N3)4[2,6-(t-BuNH)]2(8), which contains two unsubstituted
chlorine atoms, was obtained23 (Scheme 2).
DcHmu 5 DcUm + DnRTDn 5 gini(products,g) 2
gini(reactants,g)
Here Sni is the total molar amount of the gases in products or
reactants. Typical calorimetric experiment21 results (averaged over
three measurements each) of the constant volume combustion
energy (DcUm) of 4, 5 and 6 are given in Table 1. From these values
and eqns (2), (3) and (4), the standard molar enthalpies of
combustion (DcHmu) for 4, 5 and 6 are 24142.3 25409.0 and
24393.3 kJ mol21, respectively The literature values for standard
The presence of chlorine atoms diminished our interest in
determining thermodynamic properties and the release of the
chlorine gas is harmful from an environmental perspective.
However, since 8 is a stable solid which provided an opportunity
for structural study. The solid-state structure of 8 was obtained24
which is the first reported crystal structure of a tetrameric
azidocyclophosphazene. The structure (Figure 1) shows that the
gem N3 substitution and the Cl ligands are trans-oriented. The
orientation of both gem-substituted N3 groups is cis to each other
and point towards the t-Bu groups. This orientation is preferred to
prevent steric crowding as the NH groups form a hydrogen bond
molar enthalpies of formation of CO2 (g) 5 2395.51 kJ mol21
H2O (l) 5 2285.83 kJ mol21 and P4O10 (s) 5 22984 kJ mol21
,
.
Substituting these values in the designed Hess thermochemical
equations for 4, 5 and 6 (eqns (5)–(7)) the molar enthalpies of
formation are calculated and are shown in Table 1.
DfHmu [P3N3(N3)6] (s) 5 DfHmu [P4O10] (s) 2
(5)
DcHmu [P3N3(N3)6] (s)
DfHmu [P3N3(N3)5(NMe2)] (s) 5 {3/4 DfHmu [P4O10] (s) +
2 DfHmu [CO2] (g) + 3 DfHmu[H2O] (1)} 2
(6)
DcHmu [P3N3 (N3)5(NMe2)] (s)
DfHmu [P3N3(N3)4(NH2)2] (s) 5 (3/4 DfHmu [P4O10] (s) +
(7)
DfHmu[H2O] (1)) 2 DcHmu [P3N3 (N3)4(NH2)2] (s)
The positive heats of formation are indicative of rather high
energies associated with azido derivatives. Not surprisingly, the
heats of formation of the azido-substituted cyclophosphazene
trimers are positive compared with the negative value reported
Table 1 Standard molar enthalpies of formation of the azido
derivatives of cyclophosphazene trimers
Density/
g cm23
DcUm/
kJ mol21
DfHmu/
DfHmu/
Compounds
kJ mol21
kJ kg21
Fig. 1 Thermal ellipsoid plot (30%) showing the hydrogen bonding
…
4
5
6
—
1.67
1.79
24158.9
25419.4
24403.7
1904.3
1528.6
1869.5
4920.7
3929.6
5580.7
association via N9–H9 N2. Only two molecules and only the N9
hydrogen atoms are shown for clarity. Hydrogen bonding interaction
indicated by dashed lines.
5194 | Chem. Commun., 2005, 5193–5195
This journal is ß The Royal Society of Chemistry 2005

View Article Online
i
…
15 Compound 2 (1.16 g, 3.25 mmol) was dissolved in dry acetone and
sodium azide was added (2.11 g, 32.45 mmol). The mixture was stirred
at 25 uC for two days. It was filtered to remove NaCl, and the solvent
was removed to give mono(dimethyl-amino)pentaazido cyclotriphos-
˚
with a neighboring ring nitrogen (N9–H9 N2 5 3.143(3) A; i 5
2x,y,2z + 1/2; see electronic supplementary information (ESI){)
generating a twisted intermolecular eight membered ring. The
phosphazene core in 8 is in the typical chair conformation and has
1
phazene (5). Yield: 1.1 g, 87.3%. IR: 2152 cm21 (vs) N3; NMR: H, d
2.75 [d, 6H, NMe2 (3JPH 5 13.24 Hz)]; 13C, d 36.89 [d, NMe2, (2JPC
5
˚
˚
similar but slightly elongated P–N bonds (0.01 A–0.02 A) and
more acute P–N–P (5u–6u) angles than the parent halogenated
cyclotetraphosphazene, N4P4Cl8.25 The azido P–N bond and angle
are similar to the only other azido phosphazene,1,3,5-N3-1,3,5-R3-
cyclotriphosphazene (R 5 2,6(bis(4-tert-butylphenyl)phenyl).26
In conclusion, we have synthesized azido substituted cyclophos-
phazenes and for the first time calculated the standard heats of
formation based on experimentally determined heats of combus-
tion. The magnitude of these positive heats of formation suggests
that these materials may have energetic applications. While
the azido phosphazene trimers are liquids, we are able to report
the first crystal structure of an azido substituted cyclophosphazene
tetramer.
3 Hz)]; 31P{1H}, d 18.65 [t, P(NMe2), (J 5 49 Hz)], 12.09 {d, [P(N3)2],
(J 5 49 Hz)}; HRMS: Calcd for C2H7P3N19: 390.0345, Found:
390.0293.
16 Compound 6 was synthesized according to ref. 6. MP 81 uC; Yield
24%.; Spectral data for 6: IR: 2183 cm21 (s, N3); NMR 1H, d 4.66
(d, 4H, P–NH2); 31P {1H} d 14.66 (t, P–NH2, J 5 49.3 Hz), 12.42
(d, P–N3, J 5 49.9 Hz); HRMS: Calcd for P3N17H5: 336.0126,
Found: 336.0136.
17 (a) H. Xue, Ye Gao, B. Twamley and J. M. Shreeve, Chem. Mater.,
2005, 17, 191–198; (b) H. Xue, W. Arritt Sean, B. Twamley and
J. M. Shreeve, Inorg. Chem., 2004, 43, 7972–7977.
18 (a) G. Drake, T. Hawkins, A. Brand, L. Hall, M. Mckay, A. Vij and
I. Ismail, Propellants, Explos., Pyrotech., 2003, 28, 4, 174–180; (b) Ionic
Liquids III B: Fundamentals, Progress, Challenges and Opportunities,
ed. R. D. Rogers and K. R. Seddon, American Chemical Society,
Washington, D.C., 2005.
The authors gratefully acknowledge the support of the NSF
(Grant CHE0315275). We are particularly indebted to Dr. Michio
Sasaoka, Otsuka Chemical Co., Ltd. for gifts of phosphazene
trimer and tetramer.
19 R. A. Saraceno, G. H. Riding, H. R. Allcock and A. G. Ewing, J. Am.
Chem. Soc., 1988, 110, 80–982.
20 (a) B. V. Lebedev, T. G. Kulagina and D. R. Tur, J. Chem. Thermodyn.,
1999, 31, 697–710; (b) S. B. Hartley, N. L. Paddock and H. T. Searle,
J. Chem. Soc., 1961, 430–432.
21 The heat of combustion was determined using a Parr (series 1425)
semimicro oxygen bomb calorimeter. The substances were burned in an
oxygen atmosphere at a pressure of 3.04 MPa. The energy equivalent of
the calorimeter was determined with a standard reference sample of
benzoic acid (SRM 39i, NIST). Since a Parr 45C10 alloy fuse wire was
used, a correction of 2.3 (IT) cal cm21 of wire burned has been applied
in all standardization and calorific value determinations. Acid correction
has been omitted for all semimicro samples. The bomb was examined
for evidence of unburned compound after each run, and, if more than a
slight trace was present, the run was discarded.
Notes and references
1 M. V. Huynh, M. A. Hiskey, E. L. Hartline, D. P. Montoyo and
R. Gilardi, Angew. Chem., Int. Ed., 2004, 43, 4924–4928.
2 M. X. Zhang, P. E. Eaton and R. Gilardi, Angew. Chem., Int. Ed., 2000,
39, 401–404.
3 (a) D. E. Chavez, M. A. Hiskey and R. D. Gilardi, Angew. Chem., Int.
Ed., 2000, 39, 10, 1791–1793; (b) A. N. Ali, S. F. Son, M. A. Hiskey and
D. L. Naud, J. Propul. Power, 2004, 20, 120–126.
22 S. S. Krishnamurthy, A. C. Sau, A. R. Vasudeva Murthy, R. Keat,
A. R. Shaw and M. J. Woods, J. Chem. Soc., Dalton Trans., 1977,
1980–1985.
23 The reaction of N4P4Cl6[2,6-(t-BuNH)]2 (7) (4.30 g, 8.0 mmol) and
excess sodium azide (6.50 g, 100.0 mmol) in 50 mL acetonitrile was
carried out at 25 uC for 24 h. Separation of pure N4P4(N3)4Cl2[2,6-
(t-BuNH)]2 (8) was achieved when crystals of this product grew slowly
as the chloroform solution was evaporated. The isolated yield was 90%.
4 J. P. Agrawal, Prog. Energy Combust. Sci., 1998, 24, 1–30.
5 (a) A. Hammerl and T. M. Klapo¨tke, Inorg. Chem., 2002, 41, 4,
906–912; (b) H. H. Krause, in Energetic Materials, ed. U. Teipel, Wiley-
VCH, Weinheim, 2005, p. 1–25.
6 M. S. Chang and A. J. Matusko, J. Am. Chem. Soc., 1960, 82,
5756–5757.
7 V. Chandrasekhar, M. G. Muralidhara and I. Selvaraj, Heterocycles,
1990, 31, 2231.
White solid, MP 129 uC IR: 2155 (vs) nN , 2976 (m) na (Me), and 3265
8 (a) F. Forohar, P. R. Dave, T. Axenrod, C. D. Bedford, R. Gilardi and
C. George, Phosphorus, Sulfur, Silicon, 1995, 101, 161–166; (b)
P. R. Dave, F. Forohar, T. Axenrod, C. D. Bedford, M. Chaykosky,
M.-K. Rho, R. Gilardi and C. George, Phosphorus, Sulfur, Silicon, 1994,
90, 175–184.
9 (a) P. R. Dave, F. Forohar, M. Chaykovsky, C. D. Bedford, US Pat.,
5,352 829 1994; (b) F. Forohar, P. R. Dave and S. Iyer, US Pat.,
6,218,554 and 6,232, 479, 2001.
3
(m, br) cm21 na (N–H). NMR: 1H, d 1.41 [s, 9H (t-Bu)], 3.16 [d
(J 5 3 Hz) N–H]; 13C, d 31.1 (d, J 5 6 Hz); 31P{1H}, d 22.3 (t,
J 5 43 Hz), 26.8 (t, J 5 43).
˚
24 Crystal data for 8 at 84 K with Mo Ka (l 5 0.71073 A) radiation:
C8H20Cl2N18P4, M 5 563.20, Monoclinic, space group C2/c (no. 15),
˚
a 5 18.452(4), b 5 9.5500(19), c 5 14.060(3) A, b 5 103.67(3)u, V 5
2407.4(9) A , Z 5 4, r 5 1.554 mg m23, m 5 0.574 mm21, v scans,
3
˚
h range 5 2.27 to 27.50u, 15727 reflections measured, GoF on F2 5 1.061
for 2774 unique observed data, (Rint 5 0.0233), 6 restraints and 159
parameters, R1 5 0.0444, wR2 5 0.1134 for [I . 2s(I)]. CCDC 278747.
See http://dx.doi.org/10.1039/b510924h for crystallographic data in CIF
or other electronic format. The t-Bu group was disordered and modeled
in two positions with occupancies of 0.74 and 0.16. Loose restraints were
applied to keep C–C distances reasonable.
10 (a) A. J. Elias and J. M. Shreeve, Adv. Inorg. Chem., 2001, 52, 335–358;
(b) V. Chandrasekhar and V. Krishnan, Adv. Inorg. Chem., 2002, 53, 159.
11 R. Haiges, S. Schneider, T. Schroer and K. O. Christe, Angew. Chem.
Int. Ed., 2004, 43, 4919–4924.
12 J. M. E. Goldschmidt and J. Weiss, J. Inorg. Nucl. Chem., 1964, 26,
2023–2024.
13 G. R. Feistel, M. K. Feldt, R. L. Dieck and T. Moeller, Inorg. Synth.,
1973, 14, 23–26.
14 (a) C. Grundmann and R. Ra¨tz, Z. Naturforsch., B, 1955, 10, 116–117;
(b) IR: 2156 cm21 (vs) N3; NMR 31P{1H}: d 11.38 ppm (s); HRMS:
Calcd. for P3N21 + 1: 387.9937, Found: 387.9951.
25 A. J. Wagner and A. Vos, Acta Crystallogr., Sect. B, 1968, B24,
707–713.
26 R. J. Wehmschulte, M. A. Khan and S. I. Hossain, Inorg. Chem., 2001,
40, 2756–2762.
This journal is ß The Royal Society of Chemistry 2005
Chem. Commun., 2005, 5193–5195 | 5195