By “oxygen” do you mean the resource that made the ozone you’re smelling, or the locus of the recipe that used itself, the other two oxygens and an electricity source to make the scent? Using and being used aren’t the same, even where one uses his own substance. This account is opposed to the sort of particle-as-pure-foundation account that does not see it as essentially resource and recipe.
So should we give a plan-resource account of the oxygen that makes ozone or a ontological monistic account of it? It is pointless to try to solve the matter by hard-nosed realism. In both cases the particles are “all there is”, but on the first account “to be a particle” is to exist on two ontological axes whereas in the second there is only one. The most significant difference is that on the first account it is possible to allow for the substantial existence of both oxygen and ozone whereas in the second case only oxygen can be a substance while ozone has to be an accidental or emergent form.
It is difficult to find the right word for the account of matter opposed to the resource-recipe account. “Mechanical” account comes close, but it’s really more like the “pellet-cloud” account of matter, where particles form compounds like water droplets form cloud-shapes. There is no empirical difference between the two accounts, and on both accounts the particle is as substantial as it can ever be. The essential difference, again, is whether anything other than the particle can be substantial: molecule, protein, cell, tissue, animal… The second difference is that the pellet-cloud account of the particle makes it entirely driven a tergo as opposed to being either a determinate or indeterminate recipe using whatever resource it needs to establish its existence.