**4. Discussion**

Spraying of blends containing large powder particles has the disadvantage that only one component can be sprayed with a truly optimized parameter set. For ternary blends, this is naturally even more di fficult as for binary ones. By this reason, for this study, an identical parameter set was used, which represents a compromise for the processability of all oxides. In case of the blends of the Al2O3-Cr2O3-TiO2 system, two high melting components (Al2O3 and Cr2O3) are combined with one lower melting oxide TiO*<sup>x</sup>*. Thus, the liquid phase occurs at lower temperatures compared to TiO2 and the addition of TiO*x* to Cr2O3 and Al2O3 decreases the coating porosity. For Al2O3-rich and Cr2O3-rich powder blends (ACT and CAT), a part of the large particle fraction is not melted to the extent that it can contribute to coating build up due to the higher melting temperature. As indicated by the coating thickness after 10 passes (see Figure 4), the highest deposition e fficiency is found for the TAC coating.

The XRD investigations show that the typical phase transformation from α-Al2O3 to γ-Al2O3 as described in Section 1 occurs for all powder blends. The additions of Cr2O3 and TiO2 for the given spray conditions do not have an e ffect on this phase transformation, which aligns with the literature data [2]. In addition, reduced peak intensities related to sub-stoichiometric phases and an increase of the rutile peaks intensity at 2θ = 27.4◦ (110) and 2θ = 54.2◦ (211) can be an indication for oxygen gain during the processing and formation of stoichiometric TiO2 or near stoichiometric TiO*x* as a result of the thermal spray process.

Recently, a surprising high dissolution of titanium atoms in γ-Al2O3 in APS coatings obtained from a commercial Al2O3-40% TiO2 powder blend was observed [19]. Surprisingly, this was not observed in this study for any of the ternary blends.

In the literature, the formation of a Cr2O3-rich solid solution (Cr,Ti)2O3 is described for binary suspensions [14], while, for blends of large particles, contradictory results are reported [17,20]. In this study, the EDS measurements have shown the presence of small amounts of titanium in the chromium oxide lamellae. The titanium content in the chromia splats of the ACT and TAC coatings was below 2 at % but is slightly higher in the TiO*x*-rich TAC coating. The higher titania content increases the number of interfaces between titanium and chromium oxide splats. Therefore, more titanium di ffuses into the chromium oxide lamellae.

Since the hardness of the Cr2O3 coating was the highest, it is not surprising that the hardness is increased with a rising content of Cr2O3. The Cr2O3-rich CAT coating shows the highest hardness for the ternary blends. As expected, the Al2O3-rich ACT coating and the plain Al2O3 coating show higher hardness values than TiO2-rich coatings. However, the titanium oxide-rich TAC coating has a similar hardness compared to the ACT coating. The lower porosity of the TAC coating is assumed to be responsible for this e ffect.

The sliding wear rate of Al2O3 coatings can be significantly reduced by adding TiO*x* and Cr2O3. Both coatings have a higher wear resistance than the plain Al2O3 coating. Possible reasons could be the improved toughness of the coating expected by adding TiO*x* [21] and/or addition of Cr2O3, which has, by far, the highest wear resistance. Except for the Al2O3 coating, coating hardness and sliding wear rates show some correlation especially for coatings from the blends. Whereas ACT and TAC coatings have similar hardness values and wear rates, the higher content of Cr2O3 in the CAT coating leads to significantly higher hardness values and a reduction in the wear rate. It should also be taken into account that the sliding wear rates of the plain oxide could be influenced by the identical spray parameter set, which is applied to the deposition of all coatings in this study.

All investigations reveal that the individual properties of the plain oxide strongly influence the properties of the coating from the blends. Therefore, for example, increasing the TiO*x* content leads to denser coatings and higher deposition rates as well as higher as-sprayed roughness. The increase of the Cr2O3 content positively influences the sliding wear resistance.

#### **5. Summary and Conclusions**

It was shown that firmly adhering coatings can be produced by atmospheric plasma spraying from powder blends containing Al2O3, Cr2O3, and TiO2 using an identical spray parameter set. Investigations with XRD have shown that the phase transformation from α-Al2O3 to γ-Al2O3 also occurs in the powder blends and is not influenced by the addition of Cr2O3 or TiO*<sup>x</sup>*. Furthermore, for titania, a gain of oxygen content was found. EDS measurements have shown the existence of small amounts of titanium in the chromium oxide lamellaes. The investigations have shown that the respective dominant single oxide has a significant influence on the coating properties. Whereas a high TiO*x* content leads to higher deposition rates, higher as-sprayed roughness, and low porosities, the hardness and wear resistance of the coatings can be improved by increasing the Cr2O3 content. By using powder blends, the sliding wear rate can be improved when compared to plain Al2O3 coatings. Thus, the use of powder blends presents a promising approach to adapt or extend the property profile of plain oxide coatings. The reactivity between the materials involved needs to be further investigated in order to exploit further improvement potentials.

**Author Contributions:** M.G., S.C., and G.P. conceived, designed, and performed the experiments. M.G., L.-M.B. and T.L. (Thomas Lindner) analyzed the data and wrote the paper. T.L. (Thomas Lampke) directed the research and contributed to the discussion and interpretation of the results. All authors have read and agreed to the published version of the manuscript.

**Funding:** This project was funded under contracts 100310631/100310633 via Sächsische Aufbaubank by the European Structural Fonds EFRE and by the Free State of Saxony. The German Research Foundation/DFG-392676956 and the Technische Universität Chemnitz in the funding program Open Access Publishing funded the publication costs of this paper.

**Acknowledgments:** The authors would like to thank Marc Pügner for the XRD measurements and his support in interpreting them, Frank Trommer for supporting the coating process and Kerstin Sempf for EDS analyses.

**Conflicts of Interest:** The authors declare no conflict of interest.
