Nitrogen dioxide is made by the catalyticoxidation of ammonia: steam is used as a diluent to reduce the combustion temperature. Most of the water is condensed out, and the gases are further cooled; the nitric oxide which was produced is oxidised to nitrogen dioxide, and the remainder of the water is removed as nitric acid. The gas is essentially pure nitrogen tetroxide, which is condensed in a brine-cooled liquefier.
Nitrogen tetroxide is a brownish yellow liquid which is easily vaporized. It is a powerful oxidizer, and is highly toxic and corrosive. However, it is not affected by mechanical shock and does not react with air. Dinitrogen tetroxide is always in equalibrium with Nitrogen_Dioxide (NO2) and some Nitrogen Dioxide will be present in any quntity of Nitrogen tetroxide, higher temperatures push the equilibrium towards Nitrogen Dioxide.
If you want nitrogen, just decomposing hydrazine (no N2O4 involved) is a fairly reasonable way to get it; this has been mentioned as a way to store nitrogen conveniently for stations.
The only real issue, apart from the heat generated, is that you must be absolutely sure there is *no* leftover hydrazine in the output stream, because hydrazine is highly poisonous, and carcinogenic to boot.