XKCD #1162 piqued my curiosity. Obviously, there's nothing we generally use as a chemical fuel that can match the energy density of uranium fission. But I wondered what it would take, hypothetically, to match it mole-for-mole?
Looking at the bond enthalpy charts, I see that the carbon-oxygen triple bond has the highest enthalpy of formation (the energy required to break it and form something else, and by extension the energy produced when this bond is formed, minus the energy required to break the previous ones) at $\pu{1072 kJ mol-1}$. However, the carbon-oxygen single bond has an EoF of $\pu{358 kJ mol-1}$, and other bonds likely to be involved ($\ce{H-C}$: $\pu{413 kJ mol-1}$, $\ce{H-O}$: $\pu{463 kJ mol-1}$, etc.) have higher bond enthalpies, meaning the formation of carbon monoxide by combustion of an organic molecule is very likely a losing proposition.
Nitrogen, on the other hand, has an infamously high bond energy ratio; breaking a single nitrogen bond costs only $\pu{163 kJ mol-1}$, while forming a triple bond releases $\pu{941 kJ mol-1}$. Nitrogen single bonds, especially in high percentages of the total molecular mass, are notoriously explosive; the azide anion, covalently bonded, forms the backbone of the eminent Dr. Lowe's Things I Won't Work With. The nitro group ($\ce{RNO2}$), well-known in explosives, can be made even more potent by bonding it to another nitrogen to form a nitramine, and these are central to the well-known explosive cyclotrimethylenetrinitramine (better known in pure form as RDX, and with added plasticizers as C-4) and its big brother, HMX (cyclotetramethylenetetranitranime).
So, let's consider a theoretical molecule composed of only single-bonded nitrogen (eek). $\ce{N4}$, tetranitrogen, has been synthesized, but in a form theorized to be more like two pairs of double-bonded nitrogens that are joined by two single bonds (kind of hard to tell, since it only exists at very high temperature and for less than a microsecond at a time). Let's instead bond each nitrogen to each other one, forming an isohedron. When that structure reverts to diatomic nitrogen (and it will, oh, you betcha), we can expect a heat of detonation of
$$4\cdot\pu{163 kJ mol-1} - 2\cdot\pu{941 kJ mol-1} = \pu{-1230 kJ mol-1}$$
That is a pretty aggressive exotherm; RDX has a detonation energy of 1118 kJ/mol, and its molecular mass is four times that of our theoretical "isohedral tetranitrogen". RDX is as heavy as it is for stability; high molecular weights sacrifice pure energy density for a structure that won't detonate given the slightest excuse (or even no excuse at all).
However, bring in the champion:
$$\ce{^{235}_{92}U + ^1_0n -> ^{142}_{56}Ba + ^{91}_{36}Kr + 3 ^1_0n}$$
(in the atomic chaos of a nuclear reaction, a lot of other products are possible, but this is one of the most common). The masses of the reactants total $\pu{236.0526 Da}$, and the masses of the products, $\pu{235.8658 Da}$, for a delta of mass of $\pu{0.1868 Da}$ ($\pu{3.102e-28 kg}$). $E = mc^2$, so the energy produced by one uranium atom splitting this way is
$$E = \pu{3.102e-28 kg}\cdot(\pu{3.00e8 m s-1})^2 = \pu{2.79e-12 J}$$
which when multiplied by Avogadro's number and converted to kilojoules, gives us
$$\pu{2.79e-12 J}\cdot\pu{6.022e23 mol-1} = \pu{1.68e8 kJ mol-1}$$
That's 168 million kilojoules per mole available by fissioning uranium-235.
Assuming that double the nitrogen equals double the heat of reaction, in order to have the same heat of decomposition, mole-for-mole, our nitrogen molecule would instead have to be an isometric "sheet" of single-bonded nitrogen, or a geodesic dome (buckyazide?), consisting of $546340$ nitrogen atoms single-bonded to each other (a concept that, if ever seriously proposed, should have any student of chemistry falling out of their chair either laughing or horrified). One mole of this material would weigh $\pu{14.0067 g mol-1}\cdot 546340 = \pu{7652.42 kg mol-1} \approx \pu{8.4 t}$. One mole of U-235 weighs $\pu{235 g}$, just a little more than half a pound.
Long live the king.
The question
(thank you for reading this far)
Does the math look right? I've done it three times and come up with three different answers for the heat of detonation of my hypothetical isohetral tetranitrogen, and thus the size and weight of the resulting molecule, possibly based on the use of different enthalpy figures (I've seen the single bond energy of formation of nitrogen to be as little as $\pu{158 kJ mol-1}$, which produces a heat of detonation of $\pu{1250 kJ mol-1}$, meaning, all other things being equal, a required molecule of only $537600$ nitrogen atoms).
Also, can anyone come up with something that would be more energetic to go up against the champ? There are molecules that have been successfully created in the lab, such as octanitrocubane, that are more powerful than RDX while being usably shock-insensitive, but are prohibitively expensive to manufacture.