*3.2. Soot Emitted by the Diesel Engine*

The soot emitted by the diesel engine used for the experimental tests was characterized by means of SEM and SEM-EDX analysis (Figures S1 and S2).

The SEM images evidenced the typical spherical aggregates of soot, whose primary constituents had dimensions in the order of 50–300 nanometers. The SEM-EDX analysis evidenced that the elements detected on soot trapped by the CDPF during the experimental tests were the typical elements present in the fuel, namely, S, Cr, Mn and Ni. Other detected elements (Fe, Cu, Si and O) are the constituents of both carrier and active species.

The SEM analysis was used to also characterize the catalytic filter after the soot deposition stage in order to investigate the soot–catalyst contact (Figure 4). In particular, the following images of catalyst pieces after a soot deposition step were obtained, where a load of about 5 g/L was reached, and so at the same load of soot relevant to the regeneration step.

The SEM images shown in Figure 4 evidenced the soot–catalyst contact features. In particular, in Figure 4a it is possible to observe that the filter surface is entirely covered by the trapped soot with a very homogeneous distribution, confirming the very high CDPF ability to store a high soot amount with an optimal filtration efficiency. Moreover, Figure 4b,c are related to the inner surfaces of the carrier walls. It can be observed that the

catalyst particles are in very strict contact with the trapped soot, allowing many contact points between the soot and the catalyst, resulting in higher catalytic activity of the CDPF.

The Raman spectrum of a sample of the spent CDPF (Figure 5) evidenced the typical peaks of the SiC carrier at 295, 784 and 964 cm−<sup>1</sup> [49]. Moreover, in Figure 5 the peaks relevant to soot are also present, and they have been attributed to bands G, D1, D3 and D4, as summarized in Table 1 [50].

**Figure 4.** SEM images at different magnitudes (**a**), 94 X; (**b**), 16.81 X; (**c**), 56.94 X of the catalytic filter after the soot deposition stage and with a soot load of 5 g/L.

**Figure 5.** Fitting of the curve representing the Raman spectrum of diesel soot with band identification.


**Table 1.** Raman bands and vibration modes reported for soot.

### *3.3. Fresh Catalytic Samples*

The XRD patterns of the prepared copper ferrite in comparison with cubic (database 77- 0010), tetragonal (database 34-0425) and commercial (Merck KGaA, Darmstadt, Germany) forms are reported in Figure S3. XRD analysis shows the presence in the prepared catalyst of the typical peaks of tetragonal and cubic forms of CuFe2O4, whose average crystallite dimensions, calculated using the Scherrer formula, are about 20 nm. Two minor peaks ascribed to low amounts of oxides (CuO and Fe2O3) are also present.

The optimized procedure for the deposition of the active species, coupled to the modified porosity of the SiC monoliths, resulted in a uniform and homogeneous distribution of the copper ferrite not only on the monolith walls, but also in the porosities. The comparison of the catalytic and the bare samples in terms of BET specific surface area evidenced that the deposition of the active species resulted in a decrease in this parameter from 2.20 to 0.40 m2/g. This result could be ascribed to the catalyst placement and the inside roughness caused by acid treatment. At this point, it is it is important to emphasize that this cannot be considered a negative result. In fact, these catalytic systems properly developed for the limitation of soot emissions are specifically applied to heterogeneous solid–solid–gas systems, in which the contact and the interactivity between the solid matrices (soot and catalyst) and the gas phase (mainly O2) are the main parameters influencing the feasibility of soot combustion. Therefore, the maximization of these features is mandatory. In the catalytic systems investigated in this work, an intensified soot–catalyst contact was realized, as reported in our previous works [44], highlighting the very good catalytic activity towards soot oxidation with the O2 present in the engine exhausts.

The SEM images shown in Figure 6 evidenced that the active species were homogeneously distributed, confirming previously reported results [23,44]. Moreover, no cracks formed due to preliminary acid treatment (Figure 6a), and the pores were not plugged after the catalyst deposition (Figure 6b). A deeper analysis of Figure 6 provides evidence that the catalyst aggregates formed a compact porous layer on the SiC granules (Figure 6c), which can be optimal for enhancing the soot–catalyst contact during the regeneration stage. Moreover, the catalyst deposition procedure allowed the deposition of the active species also in the pores of the DPF channel walls (Figure 6d), without their plugging.

The SEM-EDX analysis enabled the mapping of the elements present on the filter (Figure S4), and the detected elements were the structural ones (C, O and Si) and the catalyst active species (Cu and Fe).

The ultrasound adherence tests, shown in Figure S5 as weight loss (%) vs. number of cycles, evidenced weight losses much lower than those reported in the literature for washcoated supports [51]. These results proved that the preliminary calcination at 1000 ◦C for 48 h favored the adhesion of the active species on the SiC granules even in the absence of a washcoat, due to the formation of SiO2 streaks on the SiC granules, as previously reported [52].

The H2-TPR profile is shown in Figure 7 as H2 consumption vs. temperature.

**Figure 6.** SEM images of the 30 %wt CuFe2O4 loaded monolith at various magnitudes: (**a**) 190 KX, (**b**) 965 KX, (**c**) 10.00 KX and (**d**) 25.09 KX.

**Figure 7.** H2-TPR profile of a SiC monolith loaded with 30 %wt of CuFe2O4.

Two pronounced reduction peaks, at about 300 ◦C and 610 ◦C, were present, which were attributed to the reduction of copper ferrite at lower temperatures and of Fe3O4 to Fe at higher temperatures [53].

By considering the two reduction reactions, it is possible to calculate the total amount of H2 consumed for Cu moles (H2/Cu ratio). In this case, this value was about 4, which:


$$\text{CuFe}\_2\text{O}\_4 + 4\text{H}\_2 \to \text{Cu} + 2\text{Fe} + 4\text{H}\_2\text{O} \tag{1}$$
