Sodium Amide, Imide, and Nitride

Sodium metal and dry ammonia gas, when heated to 320-350degC, react to form sodium amide and hydrogen. The reaction proceeds around 5 times faster by bubbling the anhydrous NH3 gas into the molten sodium (heated between 300-350degC) for two hours. Interestingly, below 250 degC, an unknown substance with a lower proportion of nitrogen is obtained. After complete reaction of the sodium, the unknown product did not produce hydrogen gas on reaction with water. With the reaction conducted at 200degC, the product contained 71.2% sodium by weight and 23.4% nitrogen. This possibly suggests formation of some Na2NH. However, this observation was published in the early part of the century (without modern analytical equipment) and there does not seem to be any other published detailed investigations into the unknown product of this reaction.

The "Concise encyclopedia chemistry", by Mary Eagleson, claims that, "sodium amide, m.p. 210C, sublimable in vacuum, decomposing above 500C to sodium imide, Na2NH, and then to sodium nitride, Na3N, and ammonia".*

It also says that sodium amide “is produced…by allowing the blue solution of sodium in liquid ammonia to stand. This slowly loses color and develops hydrogen (the reaction is more rapid in the presence of catalytic amounts of iron)…”

So apparently a catalyst is not absolutely required for sodium to react with cold liquid ammonia, the reaction is just very slow.

*However, I have severe doubts about the accuracy of the "sodium imide" and "sodium nitride".
A recent article (Synthesis and characterization of amide–borohydrides: New complex light hydrides for potential hydrogen storage, Philip A. Chater, 2007) claims that "Sodium imide is unknown". Sodium nitride is very unstable, decomposing into its elements at only 87 °C. Sodium metal cannot be made to react directly with nitrogen, and thermal decomposition of sodium azide only produces elemental sodium and N2.

Investigations by Titherly, C.A. Kraus, and William Argo showed that molten fused sodium amide, even when free from traces of sodium hydroxide, both very slowly dissolves metallic platinum, and catalyzes the partial decomposition of molten sodium amide. Using platinum black and fused sodium amide, the platinum slowly dissolved, and the color turned first red, then brown, and finally black, with apparent increased viscosity. The evolution of gas from the reaction was slow at 210C but increased rapidly with increased temperature. After heating for 14 hours at 300C, the gas evolved was almost entirely NH3. The solid product remaining was found to contain 1.658 moles of sodium for every mole of nitrogen present in the compound (sodium amide contains 1 mole of sodium for every 1 mole of nitrogen).

The fact that no nitrogen was evolved during the partial decomposition, and that the reaction product only produced NH3 when later reacted with water, suggested the formation of either some sodium imide and/or sodium nitride, although neither of these were isolated.