Propellants, Explosives, Pyrotechnics 27, 1 ± 6 (2002) 1

Octanitrocubane: A New Nitrocarbon

Philip E. Eaton* and Mao-Xi Zhang
Department of Chemistry, The University of Chicago, Chicago, Illinois (USA)

Richard Gilardi
Laboratory for the Structure of Matter, Naval Research Laboratory, Washington, D.C. (USA)

Nat Gelber, Sury Iyer, Rao Surapaneni

US Army ARDEC, Picatinny Arsenal, New Jersey (USA)

Dedicated to the memory of Everett E. Gilbert 1914 ± 2000

Summary

Octanitrocubane, the first new nitrocarbon in 18 years, is

introduced as a potential explosive of great power. Its synthesis
and characterization are described.

1. Introduction

With few exceptions, e.g. the perchlorates, azides and

peroxides, powerfully explosive, carbon-based molecules
contain one or more nitro (NO2) groups. In the detonation of
such an explosive the nitro group provides (at least in part)
oxygen to convert the molecule×s hydrogen to water and its
Figure 1. Nitro groups can be attached to a molecular skeleton
carbon to carbon dioxide or carbon monoxide. In a ™perfect∫ in four quite different ways.
explosion the nitrogen atom of the nitro group would be
converted to dinitrogen. These products are gaseous and are
released quickly, hot, and in voluminous amounts. A nitro detonation (shock, impact, friction sensitivity, etc.) is related
group can be attached to a molecular skeleton in four quite to the weakness of its weakest bond. Certainly, C-nitro
different ways: 1) by a bond from its nitrogen to a carbon explosives like TNT are ™rocks∫ by comparison to O-nitro
atom, as in TNT; 2) by a bond from its nitrogen to an oxygen explosives like nitroglycerin. The first (though certainly
part of the rest of the molecule, as in nitrate esters like deserving great respect) was once commonly used as an
nitroglycerin; 3) by a bond from its nitrogen to another unknown in undergraduate qualitative organic analysis
nitrogen in the molecule, as in the nitramines HMX and CL- laboratories; the second was the stuff responsible for
20 or finally 4) by a bond from its oxygen to a carbon atom, as numerous disasters until Alfred Nobel calmed it with
in nitrites like methyl nitrite (Figure 1). sawdust (as in dynamite).
The bond strength of the ™weakest bond∫ in these four TNT was first made in 1863 and by the First World War
quite different situations decreases in the order CNO2 > was the explosive of choice (238,000 tons were produced in
CNNO2 > CONO > CONO2 with bond dissociation ener- Britain alone)(3). However, through less sensitive to shock,
gies of roughly 293,(1) 197, 184, and 167 kJ ¥ mol 1,(2) respec- TNT is not as powerful an explosive as the nitramines RDX
tively. It is not unlikely that the ™trigger∫ for an explosive and HMX. By the end of the Second World War these
detonation is the rupture of one of these weak bonds; the (sometimes in combination with TNT) became the explosifs
rest of the molecule being made of CC, CH and CO bonds ordinaires of military establishments world-wide, finding
which are substantially stronger. It is much less sure, but use as high power propellant ingredients for gun shells and
probably not unrealistic, to say that an explosive×s ease of rockets as well. In the late 1980×s Nielson successfully
synthesized the nitramine CL-20,(4) presently the best next-
* Corresponding author; e-mail: eaton@uchicago.edu generation explosive(5).

¹ WILEY-VCH Verlag GmbH, 69469 Weinheim, Germany, 2002 0721-3113/02/2701/0001 $ 17.50+.50/0

2 Eaton, Zhang, Gilardi, Gelber, Iyer, Surapaneni Propellants, Explosives, Pyrotechnics 27, 1 ± 6 (2002)

What about new C-nitro explosives? These should be less

shock sensitive than CL-20, but are there candidates that
might be more powerful? ™Quantitative∫ evaluation of the
potential of a candidate explosive before synthesis would be
a very valuable ability, but calculating such is very difficult
and opens to much argument. It relies on the empirically-
derived Kamlet-Jacobs equations(6a) or, more recently, on
sophisticated ™first principle∫ techniques(6b) for an estima-
tion of detonation velocity and detonation pressure, the key
measures of the effectiveness of an explosive.
In these equations, the explosive×s density, the heat
released, the number of moles of gas produced, and the Figure 2. Structure of octanitrocubane.
molecular weight of these gases are all critical factors.
Estimations of the heat of formation(7) and of the density(8) corresponding to 74 kJ per (CNO2)-mole(9b). The highest
of even very strange organic compounds are fairly straight- calculated heat of formation of solid hexanitrobenzene
forward. However, it is not possible to predict quantitatively [(CNO2)6], one of the most energetic explosives known,(11) is
the composition of the products unless one makes many about 200 kJ ¥ mol 1,(9b,14) corresponding to 33 kJ per (CNO2)-
assumptions about detonation efficiency: How much of the mole, less than half that of ONC. The energy content of
carbon goes to CO2 and/or CO rather than to soot, C60 and hexanitrobenzene is lowered by the stabilizing resonance
diamond? Is there complete equilibration amongst prod- energy of the aromatic system. The energy content of
ucts? etc. None of this can be done with real confidence, and octanitrocubane is increased by the strain energy of its
this has led to much discussion in the literature,(9) par- unique carbon skeleton. The most recent theoretical
ticularly in regard to explosives that have no hydrogen in estimate of the detonation velocity for ONC is 9.90 km ¥
their molecular makeup. sec 1 (10). The experimental value for hexanitrobenzene is
Nonetheless, certain qualitative generalizations are rea- 9.5 km ¥ s 1;(15) for HMX, 9.1 km ¥ s 1;(16a) for CL-20, 9.2 km ¥
sonable. Of two molecules with the same formula, the one s 1 (16b).
with the higher heat of formation will probably release more
energy. The one with the higher density will probably be a
superior explosive. Density affects an explosive×s detona- 2. Cubane
tion velocity, a specialized ™linear∫ rate of reaction that
relates to how fast the energy is released, and its maximum Cubane, the hydrocarbon (CH)8, is named appropriately;
detonation pressure, a direct measure of the power of an its skeleton is in the shape of a cube.(17) At each corner of this
explosive. The overall dependence of detonation pressure cube there is a carbon atom (carrying a hydrogen) bound to
on density is greater than quadratic, so even a small increase three identical neighboring carbons. The internuclear CCC
in density can have a major impact on performance(10). angles are 908, far off from the standard 109.58 typical of
At the start of our work only five fully nitrated organic tetravalent, sp3-hybridized carbons like those in common
compounds had ever been made and characterized. These, hydrocarbons. This departure from normal is energetically
the nitrocarbons, are tetranitromethane, hexanitroethane, expensive. Cubane is ™strained∫ by about 695 kJ ¥ mol 1
tetranitroethylene, hexanitrobenzene and decanitrobiphen- corresponding to a substantial weakening of its bonds. The
yl. Their chemistry (and explosive properties) have been large bond angle deformations in cubane make it a power-
reviewed recently by Arnold Nielson,(11) one of the most house of stored energy. Each strained bond is like the spring
notable contributors to the synthesis of high energy com- in a set mousetrap. It was once thought that cubane would be
pounds. In the early 1980×s Everett Gilbert of the U.S. Army impossible to keep or even make. Neither proved to be the
Armament Research and Development Command (now case.
ARDEC) pointed out that the nitrocarbon octanitrocubane The cubane system and cubane itself were first synthe-
(ONC), then unknown, has a perfect oxygen balance (see sized in 1964 at The University of Chicago(17). It crystallizes
Figure 2), and ± in light of the properties of the parent as small rhombs and is one of the most dense hydrocarbons
hydrocarbon cubane (see below) ± should have a very high known, 1.29 g ¥ cm 3. The measured heat of formation of
heat of formation per CNO2 unit and an exceptionally high solid cubane is ‡ 620 kJ ¥ mol 1. This is a notably high value
density as well. His colleagues Jack Alster, Oscar Sandus and quite unlike that of ™everyday∫ hydrocarbons; these
and Norman Slagg at ARDEC provided theoretical support have negative heats of formation. The thermodynamic
for Gilbert×s brilliant insight and estimated that ONC would instability of cubane is not, however, accompanied by
have a detonation pressure significantly greater than kinetic instability. Cubane and most of its derivatives are
HMX(12). Later, both statistical and computational ap- amazingly stable. The energy of activation for thermal
proaches predicted a density of 2.1 ± 2.2 g ¥ cm 3 for octani- decomposition of cubane in the gas phase is very large, about
trocubane, greater than any other C, N, O compound(13). 180 kJ ¥ mol 1 at 230 ± 260 8C, so the molecule decomposes
The most recently published calculated value for the heat only slowly at such temperatures. The reason for this is that
of formation of solid ONC [(CNO2)8] is 594 kJ ¥ mol 1, there are no kinetically viable paths along which cubane can
Propellants, Explosives, Pyrotechnics 27, 1 ± 6 (2002) Octanitrocubane: A New Nitrocarbon 3

rearrange thermally. On one hand, orbital symmetry con- above(20a, d). TNC is an extraordinary material. It is a
siderations raise the energy of concerted two-bond ring crystalline solid with a very high density (1.814 g ¥ cm 3),
opening reactions. On the other, there is little to be gained quite up to prediction. Although a thermodynamic power-
by breaking just one bond, as there is only a small house, it is kinetically quite stable, not melting until 270 8C,
concomitant change in geometry and thus the resulting and then with decomposition, not detonation. Tiny samples
biradical is still very strained.(17) on the lab bench stone top survive hammer blows. Deto-
nation has been achieved purposefully under controlled
conditions at ARDEC; the explosive energy release ex-
3. Synthesis of Nitrocubanes ceeded expectation, but the quantitative results are not yet
available openly. Calculations indicate that TNC may well
Like cubane once, octanitrocubane was first thought to be prove to be a better monopropellant than ONC(21).
impossible to keep and maybe even impossible to make. How about the synthesis of more highly nitrated cubanes?
How might one approach the synthesis of octanitrocubane? It is easy enough to use the methodology developed at
Nitrations (i.e., replacing a hydrogen by a nitro group) of Chicago to introduce more than four carboxylic acid groups
aromatics and other unsaturated systems are well-known onto cubane (indeed the full complement of eight have been
reactions. However, this approach is inapplicable to the put on). But, as soon as there are more than four carboxylic
cubane system, as it has no unsaturated carbons. In most acid groups on cubane at least two must be adjacent.
cases, nitro groups on saturated systems are introduced by Conversion of adjacent acid groups to adjacent nitro groups
functional group transformations. The most standard is the necessarily involves adjacent amine groups (cf. Scheme 1).
sequence of introducing a carboxylic acid substituent There is nothing wrong with this, even on the cubane
(COOH) in place of hydrogen, changing it to an amine skeleton; 1,2-diaminocubanes have been made. They are
(NH2) by one version or another of the classic Curtius relatively stable compounds, but ring cleavage reactions
rearrangement, and then oxidizing the amine to the desired intervene during attempted oxidation to the corresponding
nitro group. Simple enough. But at the time this project was 1,2-dinitrocubane. This sort of skeletal cleavage always
started there was no way of introducing carboxylic acid happens when attempts are made to place an electron-
groups directly onto the cubane skeleton. Indeed, method- donating group (e.g. NH2) on a cubane carbon adjacent to
ology for systematic substitution on strained systems was one bearing an electron-withdrawing substituent (e.g. NO2).
almost unknown. It was finally worked out at Chicago, in the It is unavoidable and is one of the very few limitations
quest for ONC; the details are described fully elsewhere(18). imposed on manipulation of substitution/substitution pat-
It is rewarding to report that this new methodology, tern in the cubanes(17).
developed specially for the synthesis of one of the most Putting more than four nitro groups onto cubane required
unusual organic compounds, now finds wide use for learning how to directly nitrate tetranitrocubane. Chemi-
introducing substituents in place of hydrogen on common cally the most exciting thing about TNC is its Br˘nsted
systems like cyclopropanes and aromatics (18b, 19). acidity. The measured pKa is  21, at least 20 powers of ten
Once made,(20) cubane-1,3,5,7-tetracarboxylic acid was more acidic than cubane itself(20d). Much of this enhanced
converted to the corresponding tetranitrocubane (TNC) by acidity must be the result of the tremendous inductive
a much modernized version (Scheme 1) of the classic electron withdrawing power of the nitro groups stabilizing
COOH ! NO2 functional group manipulation mentioned the anion. Salts of TNC are readily formed. Reaction of
these with different electrophiles has been used with great
success to add one to four additional substituents (e.g.
methyl, carboxyl, trimethyltin) to the carbon skeleton of
tetranitrocubane(20d, 22). Unfortunately, it is not nearly so
simple to introduce additional nitro groups.
Before the present work nitration of an undelocalized
anion was impossible to achieve in useful yield. For example,
reaction under various conditions of phenyllithium with any
of numerous nitrating agents (N2O4, NO2BF4, N2O5, etc.)
gives, if anything, complex mixtures containing little nitro-
benzene. The same is true for aliphatic organometallics.
However, interfacial nitration, a new nitration process
discovered serendipitously at Chicago goes a long way
toward solving this old problem(23). In this approach, a
solution of an anion salt is frozen to a glass and solid N2O4 is
deposited on its surface. The reactants come into productive
combination when the glass is allowed to melt. So, for
example, reaction of phenyllithium in frozen tetrahydrofur-
an with N2O4 on its surface proceeds successfully as melting
Scheme 1. occurs (ca. 115 8C). This gives nitrobenzene in good yield,
4 Eaton, Zhang, Gilardi, Gelber, Iyer, Surapaneni Propellants, Explosives, Pyrotechnics 27, 1 ± 6 (2002)

whereas little nitrobenzene is formed when N2O4 is reacted 78 8C in 1 : 1 THF/a-methyl THF. After the mono-sodium
with phenyllithium in a very cold fluid solution of THF just salt formed, the solution was cooled to between 125 and
above its melting point. It is not understood why interfacial 130 8C giving a clear, but very viscous fluid. This was
nitration works so well. Whatever the explanation, the stirred vigorously as excess N2O4 in cold isopentane was
method is widely successful and has been used to convert the added. After one minute the base was quenched with nitric
sodium salt of TNC to pentanitrocubane (PNC). PNC is acid in cold ether, and then the whole mixture was added to
about a thousand times more acidic than TNC, and its water. This resulted reproducibly in almost complete
sodium salt was converted in turn by interfacial nitration to conversion of TNC (1 gram scale) to HpNC (95% by
hexanitrocubane (HNC), a still more acidic material. Both NMR), isolated crystalline in 74% yield. When lesser
PNC and HNC are stable, crystalline materials of high amounts of base were used the conversion could be
density (Figure 3)(20d, 22). controlled to give predominantly penta- or hexanitrocu-
PNC is the first example of a nitrocubane with nitro bane. For the conversions of TNC to HpNC to have
groups on adjacent carbons. X-ray single crystal analysis at happened, many sequential reactions (Scheme 2) must
the Naval Research Laboratory showed, contrary to the have occurred very rapidly in the viscous solvent at about
worries of some, that no significant distortion within the 125 8C. That all these come about so quickly is truly
nitro groups, or in the N to C bonds, or in the cubane core are remarkable and testimony to the extraordinary power of
necessary to accommodate a nitro group surrounded this new methodology; the reaction of an anion with N2O4 at
entirely by other nitro groups. Bond angles and lengths in very low temperature is a very effective nitration procedure.
PNC and HNC are little different from those in TNC(24). The sequence of Scheme 2 ends (prior to quenching) with
As good as the interfacial nitration process is, it proved formation of the salt of HpNC. This salt is quite a powerful
deficient for further nitration of hexanitrocubane. Different nucleophile as shown by its reaction at 125 8C with methyl
experimental methodology had to be developed to make triflate to give methylheptanitrocubane. Nonetheless, even
heptanitrocubane (HpNC)(25). In this procedure TNC was when an excess of a nitrating agent (N2O4 or many others)
treated with at least 4 equiv. of the base NaN(TMS)2 at was substituted for methyl triflate, no indication of any
formation of octanitrocubane was ever seen.
That ONC is not formed was very strange and very
frustrating. Maybe some of the more pessimistic theorists
had been right, and there was something ™wrong∫ with
ONC. This notion was soon discarded in favor of the idea
that the anion of HpNC is too stabilized for nitration with
N2O4 (HpNC is much more acidic than HNC). This concept
led to the use of the more powerful oxidant nitrosyl chloride.
Figure 3. Structure of pentanitrocubane (left) and hexanitrocu- Addition of excess NOCl to a solution of the lithium salt of
bane (right). HpNC in dichloromethane at 78 8C followed by ozonation

Scheme 2.
Propellants, Explosives, Pyrotechnics 27, 1 ± 6 (2002) Octanitrocubane: A New Nitrocarbon 5

at 78 8C gave the long-sought ONC in 45 ± 55% isolated showed, as with HpNC, that the intramolecular bond lengths
yield(25). Octanitrocubane is the sixth known nitrocarbon and angles are not unusual for cubanes (Figure 5)(24).
and the first new one to be made in the last 18 years. The D4 symmetric equilibrium conformation of isolated
ONC, computed by Hrovat, Kahr and Borden,(29) is
essentially the same as the conformation found in the
4. Properties of Hepta- and Octanitrocubane crystalline state. This effectively interdigitates the eight
nitro substituents on the cubane skeleton, thereby avoiding
Both hepta- and octanitrocubane have decomposition close intramolecular O..O contacts. These remain very small
points well above 200 8C. Octanitrocubane sublimes un- throughout 1808 rotations about the CNO2 bonds, proceed-
changed at atmospheric pressure at 200 8C. Samples were ing by nearly barrierless, correlated disrotation of all eight
not detonated by hammer taps. Little is yet known about nitro groups through a D2d transition state; two tetrahedral
long term storage behavior, but at least one sample of ONC subsets of four nitro groups turn in opposite directions. This
sealed in a glass tube has survived unchanged for 14 months. threshold mechanism provides for enantiomerization of the
Heptanitrocubane forms beautiful, colorless crystals chiral equilibrium conformation.
when its solution in fuming nitric acid is diluted with sulfuric The density of the one ONC polymorph now in hand is
acid. Single crystal X-ray analysis at the Naval Research very high, 1.979 g ¥ cm 3, but still lower than that calculated.
Laboratory confirmed the assigned structure and provided Even a simple extrapolation from the observed X-ray
an accurate density at 21 8C of 2.028 g ¥ cm 3, impressively densities of the other nitrocubanes leads to an expected
high for a C,H,N,O compound. There are only two or three value of 2.06 g ¥ cm 3. The latest and most sophisticated
other known compounds composed of these elements that calculation, using all the available data from polynitrocu-
(perhaps) have higher densities(26). Although ONC catches bane X-ray crystal structures, predicts a density above 2.1 g ¥
the imagination with its symmetry, HpNC currently is cm 3 for the most stable polymorph of ONC.(13, 30) This
significantly easier to make than ONC. It may prove to be a encourages belief in the existence of a crystal form of ONC
more powerful explosive and/or monopropellant as it is much more dense than the one now in hand. Many nitro
more dense, has a calculated(9b) heat of formation only about compounds, e.g. HMX and CL-20, have multiple poly-
6% less than ONC, and will produce some lighter gaseous morphs of remarkably different density. The four known
products as it is not perfectly oxygen balanced. polymorphs of CL-20 have densities ranging from 1.91 to
It is quite reasonable to expect that oxidative coupling of 2.044 g ¥ cm 3 (4). There is no magic recipe for obtaining the
HpNC would produce tetradecanitrobicubyl (14 nitro most dense polymorph. Typically, the search is done by
groups, Figure 4). Early estimates(27) put its heat of forma- trying different crystallization solvents, temperatures, etc.
tion at 1013 kJ ¥ mol 1. Its parent (in a formal sense), the This work is underway.
hydrocarbon bicubyl, is already known; it has a density of Although the first signs are good ± ONC exists and is
1.336 g ¥ cm 3, significantly higher than cubane (1.29 g ¥ cm 3). stable ± significant quantities must still be made for the
Other polycubyls are known and together make an inter- exploration of its properties as an explosive and/or propel-
esting set of stable rigid rod rulers to which each cubyl unit lant. Presently, ONC is an expensive compound to make.
adds 0.42 nm to the length of the rod(28). Nitro derivatives of Research is now focused on finding an economical synthesis.
these are not yet known. Fortunately, there are many possibilities. One particularly
Octanitrocubane is a white solid, somewhat soluble in enticing approach recognizes that four molecules of dini-
hexane and readily soluble in polar organics. Solvent-free troacetylene are a stoichiometric (atom-by-atom) equiva-
crystals for X-ray analysis were obtained from trifluoro- lent of one molecule of ONC. Perhaps ONC can be made
acetic acid, from fuming nitric acid/sulfuric acid, or from directly by tetramerization of dinitroacetylene. Politzer has
scrupulously dried cyclohexane. X-ray structural analysis applied the state-of-the-art theoretical calculations to this

Figure 5. The molecular structure of octanitrocubane as it occurs

in the crystal. The ellipsoids represent the magnitudes of atomic
Figure 4. Structure of tetradecanitrobicubyl and polycubyl. displacements that are primarily due to harmonic vibrations.
6 Eaton, Zhang, Gilardi, Gelber, Iyer, Surapaneni Propellants, Explosives, Pyrotechnics 27, 1 ± 6 (2002)

possibility(31). His conclusion is spectacular: the free energy (16) a) As formulated in LX-14 (95.5% HMX), see Ref. 5.
change for dinitroacetylene to octanitrocubane is thermo- b) As formulated in RX-39-AB (95.5% e-CL-20), see Ref. 5.
(17) For a review and references to the synthesis, properties and
dynamically downhill by nearly 417 kJ ¥ mol 1. Of course, chemistry of the cubane system see P. E. Eaton, Angew.
these calculations do not address the kinetics of the Chem. Int. Ed. Engl. 104, 1447 ± 1462 (1992).
tetramerization or provide a clue as to how to make it (18) a) P. E. Eaton, G. Castaldi, J. Am. Chem. Soc. 107, 724 (1985).
happen. But it is time to open the investigation. Unfortu- b) P. E. Eaton, C.-H. Lee, and Y. Xiong, J. Am. Chem. Soc.
nately, dinitroacetylene is not yet a known compound. The 111, 8016 ± 8018 (1989).
c) P. E. Eaton, Y. Xiong, and C.-H. Lee, J. Chinese Chem. Soc.
acidity of acetylene, its hydrocarbon parent, is not unlike (Taiwan) 38, 303 ± 306 (1991).
that of tetranitrocubane, whose anion has been nitrated with (19) a) P. E. Eaton, R. G. Daniels, D. Casucci, G. T. Cunkle, and P.
great success (see above). Perhaps a similar approach will Engel, J. Org. Chem. 52, 2100 ± 2102 (1987).
provide for the simple conversion of acetylene, a cheap b) P. E. Eaton, K. Lukin, J. Am. Chem. Soc. 115, 11370 ±
11375 (1993).
starting material, to dinitroacetylene and thence to octani- (20) a) P. E. Eaton, Y. Xiong, and R. Gilardi, J. Am. Chem. Soc.
trocubane. 115, 10195 ± 10202 (1993).
b) A. Bashir-Hashemi, Angew. Chemie Int. Ed. Engl. 32,
612 ± 613 (1993).
5. References c) A. Bashir-Hashemi, J. Li, N. Gelber, and H. Ammon, J.
Org. Chem. 60, 698 ± 702 (1995).
d) K. A. Lukin, J. Li, P. E. Eaton, N. Kanomata, J. Hain, E.
(1) This value pertains to nitro attached to sp2 unsaturated Punzalan, and R. Gilardi, J. Am. Chem. Soc. 119, 9591 ± 9602
carbon as in nitroaromatics. The C-NO2 bond in nitrocubanes (1997).
is very much like that in nitroaromatics. The bond strength is (21) P. Politzer, J. S. Murray, M. E. Grice, P. Sjoberg in: G. A.
less,  250 kJ ¥ mol 1, for a nitro group on an unstrained Olah, D. A. Squire, (eds), ™Chemistry of Energetic Materials∫,
saturated carbon.(3) Academic Press, San Diego 1991, p 89.
(2) C. F. Melius in: S. N. Bulusu (ed), ™Chemistry and Physics of (22) K. A. Lukin, J. Li, P. E. Eaton, and R. Gilardi, J. Org. Chem.
Energetic Materials∫, Kluwer Academic Publishers 1990, 62, 8490 ± 8496 (1997).
pp 30 ± 32. (23) K. Tani, K. Lukin, and P. E. Eaton, J. Am. Chem. Soc. 119,
(3) A. Bailey, S. G. Murray, ™Explosives, Propellants & Pyro- 1476 ± 1477 (1997).
technics∫, Brassey (UK), London 1989, pp 6 ± 7. (24) Details of the X-ray crystallographic structures for all the
(4) A. T. Nielsen, A. P. Chafin, S. L. Christian, D. W. Moore, nitrocubanes reported in this paper can be obtained on
M. P. Nadler, R. A. Nissan, D. J. Vanderah, R. D. Gilardi, request from the Director, Cambridge X-Ray Crystallo-
C. F. George, and J. L. Flippen-Anderson, Tetrahedron 54, graphic Data Centre.
11793 ± 11812 (1998). (25) M. X. Zhang, P. E. Eaton, and R. Gilardi, Angew. Chemie Int.
(5) R. L. Simpson, P. A. Urtiew, D. L. Ornellas, G. L. Moody, Ed. Engl. 39, 401 ± 404 (2000).
K. J. Scribner, and D. M. Hoffman, Propellants, Explosives, (26) The only certain example is the epsilon polymorph of CL-20
Pyrotechnics 22, 249 ± 255 (1997). with an X-ray structure density of 2.044 g ¥ cm 3 at room
(6) a) M. J. Kamlet, S. J. Jacobs, J. Chem. Phys. 48, 23 ± 35 (1968). temperature(4). Hexanitroethane×s density is reported to be
b) For example, the CHEETAH code; see, L. E. Fried, M. R. 2.075 g ¥ cm 3 by X-ray analysis, but this is at 128 8C. Its
Manaa, P. F. Pagoria, and R. L. Simpson, Ann. Rev. Mater. density at room temperature is controversial, with reports
Res. 31, 291 ± 321 (2001). ranging from 1.88 to 2.25 g ¥ cm 3, none backed up by an X-
(7) See, for example and comparison of methods: T. Matsunaga, ray structure analysis. Tetranitroglycolurile has a reported
Y. Nakyama, M. Iida, S. Oinuma, N. Ishikawa, and K. Tanaka, density of 2.01 g ¥ cm 3, but this was not determined by X-ray
Propellants, Explosives, Pyrotechnics 17, 63 ± 69 (1992). analysis. See: R. Meyer, ™Explosives∫ Verlag Chemie, New
(8) For an overview: J. R. Holden, Z. Du, and H. L. Ammon, J. York 1977.
Comput. Chem. 13, 422 ± 437 (1993). (27) B. Rice, private communication, July 24, 2001.
(9) a) L. T. Eremenko, D. A. Nesterenko, Chem. Phys. Reports (28) P. E. Eaton, K. Pramod, T. Emrick, and R. Gilardi J. Am.
16, 1675 ± 1683 (1997). Chem. Soc. 121, 4111 ± 4123 (1999).
b) A. M. Astakhov, R. S. Stepanov, and A. Yu. Babushkin, (29) a) D. A. Hrovat, W. T. Borden, P. E. Eaton, and B. Kahr, J.
Combustion, Explosion, and Shock Waves 34, 85 ± 87 (1998); Am. Chem. Soc. 123, 1289 ± 1293 (2001).
cf. J. Zhang, H. Xiao, and X. Gong, J. Phys. Org. Chem. 14, b) The correlated motion can be viewed on the net at http://
583 ± 588 (2001). depts.washington.edu/chemfac/kahr_movie.html
c) C. L. Mader, ™Numerical Modeling of Explosives and (30) The packing of ONC in the crystal is described in P. E. Eaton,
Propellants∫, 2nd ed., CRC Press, New York 1998. M.-X. Zhang, and R. Gilardi, Adv. Mat. 12, 1143 ± 1148 (2000).
(10) R. D. Gilardi, J. Karle in: G. A. Olah, D. A. Squire (eds.), (31) P. Politzer, P. Lane, and J. J. M. Wiener in: K. Laali (ed.),
™Chemistry of Energetic Materials∫, Academic Press, San ™Carbocyclic and Heterocyclic Cage Compounds and Their
Building Blocks∫, JAI Press, Stamford, CN, 1999, p 73 ± 85.
Diego 1991, p 2.
(11) A. T. Nielson (ed.), ™Nitrocarbons∫, VCH-Verlagsgesell-
schaft, Weinheim 1995.
(12) a) J. Alster, O. Sandus, R. Gentner, N. Slagg, J. P. Ritchie, and Acknowledgements
M. J. S. Dewar, ™Calculation of Molecular Properties for
Polynitrohedranes Molecules∫, Working Group Meeting on This work was supported at Chicago by the United States Army
High-Energy Compounds, Hilton Head, SC, 1981. Armament Research Development and Engineering Center (AR-
b) J. Alster, S. Iyer, O. Sandus in Ref. 2, pp 641 ± 652. DEC) and at the Naval Research Laboratory (NRL) by the Office
c) G. P. Sollott, J. Alster, E. E. Gilbert, and N. Slagg, J. Energ. of Naval Research, Mechanics Division. We are grateful to Bransby
Mat. 4, 5 ± 28 (1986). Bushey and Joseph Lannon of ARDEC for their continuing interest
(13) H. Ammon, Private Communication, June 10, 1999. and encouragement.
(14) A much lower value (95 kJ ¥ mol 1) is given in Ref. 7.
(15) Ref. 11, Table 7, p 97. (Received September 26, 2001; Ms 2001/019)