DMNA A Prototypical Nitramine

Dimethylnitramine has been used as a simple model for cyclic nitramines because it is easier to deal with experimentally and theoretically than RDX and HMX, and it undergoes some of the same kinds of reactions N-NO2 bond fission, nitro-nitrite isomerization (which can be followed by NO elimination), and HONO elimination. However, even though it seemingly presents fewer problems than do the large nitramines, its decomposition mechanism is not fully resolved. 

Sumpter and Thompson [70] used DMNA as a prototypical nitramine in one of the earliest molecular dynamics simulations of gas-phase decompositions via competing pathways. The studies focused on practical aspects of simulating unimolecular reactions in large molecules (e.g., the influence of the details of the potential energy surface) and the fundamental dynamics (e.g., IVR) on the decomposition reactions. They carried out simulations using various models for the potential energy surfaces and for various initial energy distributions. 

The focus in the reaction dynamics studies was on the N02 elimination channel, but they also studied the HONO elimination reactions [70]. They based the potential energy surface on experimental data but performed some minimal basis set ab initio calculations to determine geometries, force fields, torsional potentials, and some information about the reaction paths. The representations of the global potential energy surfaces were based on valence force fields for equilibrium structures with arbitrary switching functions operating on the potential parameters to effect smooth and (assumed) proper behavior along the reaction paths. Based on the available experiments [71-73], they assumed that the primary decomposition reaction is simple N-N bond rupture to eliminate N02. 

One of the more significant results of the Sumpter and Thompson [70] study was that the N-N bond fission results were about the same for two quite different potential energy surfaces. Most of the calculations were done using a realistic potential in which the interactions for all the motions were represented as accurately as possible. The N-N bond energy was taken to be 46 kcal/mol and represented by a Morse function. The torsional motion of the nitro group (i.e., the CNNO dihedral) has a very low frequency and was treated as a free rotor, an approximation they checked by comparing IVR results with and without a barrier to the rotation. The 

The lack of accurate experimental and ab initio information about the reaction pathways at the time limited the Sumpter and Thompson [70] studies. Since then quantum chemistry calculations have better defined the energetics and reaction coordinates for the decomposition of DMNA. Politzer et al. [79] studied DMNA using DFT. They predicted the N-N bond energy to be 43.8 kcal/mol, in excellent agreement with the experimental value of 43.3 kcal/mol [80]. Harris and Lammertsma [81] computed the 

---