Is magnesium nitride formed in the combustion of magnesium?

Sometimes errors appear in textbooks and are propagated like memes. It is unfortunately common for textbook writers to assume that previous texts are correct and that the facts no longer need to be checked. In the case of the combustion of magnesium ribbon in air, the “fact” that a mixture of magnesium oxide and magnesium nitride is formed has become “common knowledge” and appears never to be questioned.

We have performed and supervised this experiment many times and have never seen any evidence of nitride formation. A search of the literature and some experiments indicate that no significant amount of nitride is formed.


The determination of the composition of MgO by heating Mg ribbon is a popular chemistry experiment both in senior high school and first-year university. A known amount of Mg ribbon is heated in a partially covered crucible over a gas flame. The Mg reacts with O2 in the air and a white residue of MgO is left. The partially open crucible cover slows the rate at which the Mg burns and reduces the loss of MgO as smoke. Knowing the masses of the Mg and MgO allows the stoichiometry of the reaction to be determined.

In every description of this experiment that we have seen, the crucible is heated until the reaction seems to be complete, about 5 – 10 minutes, and is allowed to cool. A few drops of water are added to hydrolyse any magnesium nitride, Mg3N2, that may have formed and the crucible is reheated to remove the water and decompose any Mg(OH)2 to MgO and H2O. After cooling again, the crucible is weighed. These are the reactions that are proposed to take place:

2Mg + O2 → 2MgO                                                                            (1)

3Mg + N2 → Mg3N2                                                                          (2)

Mg3N2 + 6H2O → 3Mg(OH)2 + 2NH3                                          (3)

Mg(OH)2 → MgO + H2O                                                                  (4)

The cooling – hydration – reheating step adds considerably to the difficulty of the experiment. Even university students can lose product or break crucibles by adding the water too soon or by heating the crucible too strongly while it still contains wet product. Removing this step would make the experiment more robust and easier to perform.

Magnesium nitride is commonly described as yellow and neither of us had ever seen a yellow residue after the first heating. Students commonly notice the lack of yellow product and either ask about it or silently skip the second step with no apparent degradation of results. The question then arises whether any nitride is ever formed.

In researching this, we found that there is disagreement even in the advanced level text books about some of the physical properties of Mg3N2. For example, it is variously described as colourless,1,2 yellowish green,3 yellow-green to yellow-orange4 and yellow5 while Bickley and Greg6 describe both grey and yellow forms. Further, various melting or decomposition points are given. Belin7 notes that Mg reacts preferentially with traces of O2 during the reaction of Mg in N2.

There is agreement, however, that Mg3N2 can be synthesised by direct reaction of the elements and the consensus is that the reaction on a lab scale completes in four to five hours at 800 – 850oC. Interestingly, Durrant2 notes that Mg in liquid N2 “burns brilliantly”. (Perhaps this is the seed of a great demonstration!)


Several experiments were conceived and performed to answer our question.


This replicates the experiment as it is usually performed but without the hydration step.

About 1 g of Mg ribbon was weighed and heated gently at first then strongly in a partially covered crucible for 15 minutes and cooled. No yellow or greenish colour could be seen in the product and weighing the product gave a result consistent with the pure MgO being the product.

Burning Mg ribbon in free air

In the texts, it is commonly asserted that MgO which is formed by the combustion of Mg in air may be contaminated with Mg3N2. This is not unreasonable given that the temperature of burning Mg is much greater than the 850oC noted above and although Mg3N2 burns in O2 to form MgO, the rapid cooling of the ash might preserve any nitride present.

To test this, about 20 cm of Mg ribbon was ignited and burned over an asbestos pad while being held by crucible tongs. Some of the reaction product was lost as white smoke but much solid ash was collected in a pile on the pad. The ash that fell to the pad and cooled was heterogeneous, being mostly pure white but with some dark grey material also present. It was wetted thoroughly with a few drops of water and an attempt was made to detect ammonia (eq. 3), something the nose is quite good at. None was found. Further, the dark matter remained, apparently unchanged.

Heating Mg ribbon in a closed crucible

In order to see if an O2 depleted (and therefore N2 enriched) atmosphere might give some evidence for the production of Mg3N2 an experiment was performed as described above except that the crucible was completely closed by the lid, and the crucible was heated for one hour. It was expected that as the oxygen was consumed by reaction with the Mg that the atmosphere in the crucible would asymptotically become richer in nitrogen and any reaction between the Mg and N2 would be evident.

There was almost no reaction except for a small amount of white ash which had no apparent reaction with water. Clearly, O2 was effectively excluded from the crucible, and the remaining N2 did not react with the Mg ribbon.


Given that no evidence of Mg3N2 formation could be found, it appears that the hydration step is not necessary and only makes the experiment more difficult. Not only could no ammonia be detected by smell; within the precision of the electronic scale (0.01 g) the results were consistent with pure MgO being the product.


  1. F.A. Cotton and G. Wilkinson, “Advanced Inorganic Chemistry”, 4th edition, Interscience, 1980.
  2. P.J. Durrant, “Introduction to Advanced Inorganic Chemistry”, Wiley, 1962.
  3. G.D. Parkes, ed., “Mellor's Modern Inorganic Chemistry, New edition Revised and Edited”, Longmans, 1961.
  4. G. Brauer, ed., “Handbook of Preparative Inorganic Chemistry”, Academic Press, 1963.
  5. John L. Porter, “Preparation of magnesium nitride”, United States Patent 2487474.
  6. R.I. Bickley and S.J. Greg, Journal of the Chemical Society. A, 1849-1854, 1966.
  7. P. Belin, Comptes Rendus Chimie, 254, 3538, 1962.

The author

Tim Allman is currently looking for a classroom. He has a Ph.D. in inorganic chemistry, University of Guelph, Guelph ON.

Thanks to David Greisman, John F. Ross Collegiate and Vocational Institute, Guelph ON for raising the question about the formation of magnesium nitride and suggesting a closer look.