Fig. 12 illustrates the concentration of Compound 56 over the time period of experiments. Ozone concentration is very low at a voltage level of 17 kVPP, regardless of the applied repetition rates, which implies no ozone production by NTP at these levels. When the applied voltage approaches 17 kVPP, ozone concentration increases significantly for all repetition rates. The ozone concentration at 17 kVPP is almost constant over the period of study for 2.5 and 5 kHz. However, it increases slightly at 7.5 kHz and more significantly at 10 kHz. For 19 kVPP, ozone concentration increases at 2.5, 5 and 7.5 kHz with different gradients over the period of study. On the other hand, in the same period of time, ozone concentration has been decreased continuously at 19 kVPP and 10 kHz. This trend is in contrast with the PM concentrating trend obtained from Fig. 11 at the same operating points. Thus, decreasing the ozone concentration, a lesser amount of PM can be removed by the DBD plasma reactor at this state. The same trade-off can be observed between ozone and PM at 21 kVPP for repetition rates of 7.5 and 10 kHz as well. Furthermore, as can be observed in Fig. 10, CO2 concentration at 21 kVPP and 10 kHz will be constant and even increase to some extent at the end of the time period of the experiment. This can be due to the reduction of ozone and also PM oxidation at this state. Therefore, ozone is found to be a key parameter for PM removal from diesel exhaust gas, which should be considered in all plasma emission treatment applications.