*3.2. PM Mass Emission Indices*

The PM mass emission indices (EIM), under various test conditions, were calculated from the EEPS measurements. Figure 2 shows the effect of rotation speed and engine torque on the EIM. Similar to particle number emission indices, the mass emission indices in comparison to other rotation speeds are the highest for 1800 rpm rotation speed in the entire range of engine torque. For 1800 rpm, the average EIM decreased from 1.14 × <sup>10</sup><sup>3</sup> to 3.0 × 10 milligrams per kilogram of fuel as the engine torque increased from 15 to 55 Nm. After that point, it increased to the highest value of 1.23 × <sup>10</sup><sup>3</sup> for 95 Nm. For 2100 rpm rotation speed, the EIM decreased from 6.06 × 10 to 7.73 milligrams per kilogram of fuel as the engine torque increased from 15 to 75 Nm. From this point, it increased to 2.94 × <sup>10</sup><sup>2</sup> for 144 Nm, while the regeneration of DPF started. PM mass emission indices for rotation speed

2400 rpm were the lowest from all tested rotation speeds because of the DPF regeneration. The average value of EIM decreased from 6.75 to 2.91 milligrams per kilogram of fuel as the engine torque increased from 15 to 135 Nm.

**Figure 1.** Particle number emissions index (EIN) versus torque. Error bars represent single standard deviation.

**Figure 2.** PM mass emission index (EIM) versus torque. Error bars represent one standard deviation.

## *3.3. Particle Size Distribution*

The data recorded by the EEPS was averaged for the same engine torque and then converted to differential number-based (dEIN/dlogDp) and differential volume-based (dEIV/dlogDp) particle size distributions. Figure 3 represents example plots of dEIN/dlogDp

and dEIV/dlogDp under different engine torques for rotation speed 1800 rpm (Figure 3a,b), 2100 rpm (Figure 3c,d), and for 2400 rpm (Figure 3e,f).

**Figure 3.** Differential particle number emission index (EIN) and differential particle volume emission index (EIV) PSDs: (**a**) EIN PSD for 1800 rpm; (**b**) EIV PSD for 1800 rpm; (**c**) EIN PSD for 2100 rpm; (**d**) EIV PSD for 2100 rpm; (**e**) EIN PSD for 2400 rpm; (**f**) EIV PSD for 2400 rpm.

For rotation speed 1800 rpm, the particles formed in the accumulation mode dominate; thus, number-based PSD exhibited a single mode log-normal distribution. EIN and EIV

was the highest for the last engine torques (95, 115, and 135 Nm), but also for the first tested engine torque (15 Nm).

For 2100 rpm, mostly, the particles formed in the accumulation mode dominate. When the torque reached the point of 144 Nm, regeneration of DPF started; thus, for 148 Nm, the particles formed in nucleation mode dominate. Particle number emission index before regeneration of DPF is the highest for first tested torque (15 Nm) and for the last tested engine torque before DPF regeneration (115 Nm). After DPF regeneration started, the EIN is the highest for 148 Nm. Particle volume emission index before DPF regeneration is the highest for 15 Nm, and, for 144 Nm, when the regeneration happened.

PSD for 2400 rpm shows number emission index and volume emission index after regeneration of DPF. Number-based PSD exhibited a multimodal log-normal distribution. The particles formed in accumulation mode dominated, but there were also particles formed in nucleation mode. In addition, peak of accumulation mode was higher than nucleation mode. The highest EIN and EIV for almost all particles diameters was for the three first tested engine torques (15, 35, and 55 Nm). The center of nucleation and accumulation peaks mostly decreased as engine torque increased. The median of change in average particle number and mass emission compared to 1800 rpm rotational speed was, respectively, 99.0% and 99.4%.

Figures 4–6 show the measurements results of particulate matter from the scooter in two phases: idling and under an average load on a chassis dynamometer. Figure 4 shows the dimensional distribution of particles. During idling, particles with a diameter of about 10 nm dominate, reaching the maximum value of 4.5 × 105. The engine load caused the dimensional distribution to shift to the right, towards larger particles. Thus, the dominant diameter is approximately 80 nm. In addition to increasing the mean particle size, an increase in the concentration of the particles number was noted, reaching 1.3 × 106 for the diameter dominating in the distribution.

**Figure 4.** Particle number distribution obtained for scooter.

Figure 5 shows the mass distribution of particles. During idling, a trimodal mass distribution was noted. The largest share in the emitted mass particles is represented by particles with diameters greater than 50 nm. A significant share is attributed to particles with a diameter of approximately 10 nm, which results from their dominant number in the dimensional distribution. The engine load significantly changed the mass distribution to the unimodal one. It shows that over 95% of the emitted mass of particles were caused by particles with a diameter of 30–200 nm. Additionally, a significant increase in mass concentration was noted.

**Figure 5.** Particle mass distribution obtained for scooter.

Figure 6 shows the volume distribution of particles. When the engine is idling, the majority of the emitted volume of particles correspond to particles with a diameter of about 10 nm, the number of which is the largest. Particles with diameters of 30–200 nm also have a significant share in the emitted volume. In the case of engine operation under load, it was found that particles with diameters of 30–200 nm are responsible for the vast majority of the generated volume of particles.

**Figure 6.** Particle volume distribution obtained for scooter.

#### *3.4. FTIR ATR of PM Adsorbed on Filter*

ATR spectra of PM adsorbed on the filter are presented in Figure 7.

All adsorbed PMs collected from both engines (Diesel and petrol) possess carbon black: a baseline sloping down to the right (%T display) as carbon black displays absorption over the entire region from 4000 to 400 cm−1, and, when using the ATR technique, the effect of carbon black becomes greater, with deeper light penetration at the long wavelength (low wavenumber) end. The largest amount of carbon black is present at PMs emitted from Diesel engine, both working with and without load and during cold start. Slightly less carbon black is emitted in the case of an engine without load than with load. There

are no evidence signals from alkanes or aromatic compounds; however, the weak signals at ~2900 cm−<sup>1</sup> at the range 1650–1400 cm−<sup>1</sup> might be seen. It is larger for PM from the Diesel engine than petrol engine. The signals at ~2300 cm−<sup>1</sup> and ~2100 cm−<sup>1</sup> can be from nitryl compounds or alkanes, and they are stronger in the case of PMs from the Diesel engine than petrol engine. During the cold start of engine, more carbon black is emitted that is stronger for the Diesel engine than petrol engine. To summarize, the PMs from different engines, and working under the different conditions, emitted PMs that differ in the chemical composition.

**Figure 7.** FTIR ATR spectra of PM adsorbed on filter from Diesel and petrol engine.

FTIR technique showed similarity and difference of the PM chemistry coming from the different engines and during their different work; however, this does not give an exact answer about the chemical composition of PM. More detailed and sophisticated methods must be applied. For this reason, in this study, the EGA technique was used. The aforementioned techniques were used in this research to investigate the thermal behavior of the filters and, at the same time, to examine the gasses released during thermal treatment. As a reference sample, the pure filter was investigated, and the results are shown in Figure 8.
