In real world, the atmospheric air is used for engine combustion, so you may take the atmospheric air even its volume comparison to Nitrogen is less. The engine performance may be increased if it can be ionized before entering into the cylinder.
In any combustion, atmospheric air is filtered before entering the combustion chamber and oxygen will react with hydrocarbon diesel fuel to form carbon dioxide (CO2) and water (H2O) for complete combustion. In case of excess air condition, there will be O2 left and NOx in the product stream, otherwise, CO and unburned hydrocarbon will present in the product stream if burning under air deficiency. Nitrogen in air is considered as inert not active but inside the combustion chamber, flame temperature is very high, above 2000 deg. C, nitrogen will dissociate and react with oxygen in excess air environment.
With pure oxygen, you will have NOx in the product, which is good. But flame will be hot because with atmospheric air, which is 78% nitrogen, N2 act as cooling agent until to its oxidation state where NOx is formed. From practical point of view, it is better to use atm air because to use oxygen, extra process is required and involve additional cost.
Burning pure oxygen could theoretically increase the specific output of the engine by a large amount. But the temperatures of the burned gases would be too high for the metal components to survive it.
It's interesting to note that even hot-rodders, to increase performance temporarily, will inject nitrous oxide (N2O) into the intake air, although pure oxygen would be just as easy to inject.
Even though the ratio of nitrogen to oxygen is still 2:1 with N2O (vs. 5:1 in air), they have to be very careful in using it. If the engine is run lean with N2O, the excess oxygen will start "looking" for other things to react with. The first target is the aluminum in the pistons. So when they run N2O, they run rich, to avoid burning holes in pistons.
With pure oxygen, the problems would be still more severe. In fact, with pure oxygen, many things will burn that normally wouldn't, especially with the high pressures seen in an engine.
I agree with Mr. Randall on the aspect of O2 "looking" for other things to react as that being its inherent characteristic. On the other hand, if we can use an "oxygen-rich air gas" wherein the air would act as diluent rather than oxidizer, then we could significantly reduce NOx levels ('N' present only in air and nowhere else).
--> Such a reduced levels of NOx or
--> if possible removal of NOx through some arbitrary concept of controlled intake O2 and no air,
there would be a tradeoff between an additional cost of installing an O2 production plant and savings from the removal of aftertreatment systems (SCR)
I also agree with Mr. Randall but I would suggest if we want to use pure oxygen we should use rich mixture although it will increase the combustion chamber temperature. For that we could use EGR with intercooler and 6 stroke cycle (with two dummy stroke), so that we could keep the temprature in check and fully utilize engine power produced by pure oxygen and rich fuel mixture.
Briefly, pure oxygen is better. You get higher combustion rate and higher temperature, so does higher thermal efficiency, however, shorter engine duration is expected.
Anyway, oxygen production cost is much higher then the regular air.
I have thought about this solution many times. Replacing air with pure oxygen will reduce nitrogen oxide emissions to almost zero. In this case, the SCR system is not needed on the vehicle. The fuel burning rate should be higher with a correspondingly higher maximum temperature. Of course, this can be controlled by an appropriate fuel injection strategy.
It´s clear that pure oxigen has not Nitogen. In this case the combustion doesn´t emit NOx. Under this point of view, for the environment, pure oxygen is better than the atmospheric air