3.1.2. Ag Doped CeO2

The film thickness after deposition with different CeO2:Ag ratios was measured by spectroscopic ellipsometry. The total number of CeO2 cycles was chosen to be 1500 while the number of Ag cycles was varied from 50 up to 150 according to the supercycle configuration. Growth rate of the films per cycle (GPC) over the total number of ALD cycles is shown in Figure 4 as a function of Ag dopant fraction in CeO2 film. The reduction of GPC with increasing Ag dose observed in Figure 4 could be the result of a slight etching of CeO2 by (hfac)Ag(PMe3), but it could also be due to nucleation delay and the inhibition of CeO2 growth after (hfac)Ag(PMe3) pulsing.

**Figure 4.** The effect of Ag concentration on the growth rate of Ag doped CeO2 thin films at 200 ◦C reactor temperature.

The comparison of the two calculated and experimental silver doping concentrations inside the Ag doped CeO2 films is shown in Figure 5. The calculated value is determined by the ratio of the number of doping ALD cycles divided by the total number of ALD cycles in one supercycle (1/(*n* + 1)). The experimental value is obtained from XPS measurements. It is worth emphasizing that the concentrations of dopant do not relate directly to the doping efficiency. Some of the dopant silver atoms might have formed silver oxide or alloy clusters rather than only doping the film, as will be discussed in Section 3.2.

**Figure 5.** Relationship between CeO2:Ag pulsing ratio vs. silver molar percent in the corresponding Ag doped CeO2 thin films.

The hypothetical deposition of Ag doped CeO2 by ALD can be explained as follows: (a) Ce(thd)4 adsorbs on nucleation sites (–OH) and dissociates into the attached fraction Ce(thd)*x*\*, where \* designates surface species, with *x* varying from 1 to 3 depending on the number of OH sites it bonds to. Most probably, some –OH groups remain unreacted, due to steric hindrance; (b) Ozone regenerates O\* groups which can act as nucleation sites during subsequent (hfac)Ag(PMe3) exposures. The O3 half-cycle also probably results in the formation of OH groups because of decomposition of the precursor ligand; (c) (hfac)Ag(PMe3) may adsorb on Ce–O\* or Ce–OH nucleation sites and on unreacted –OH groups, and dissociates into Ce(hfac)\* and Ag(hfac)\* species; (d) ozone may react with Ce(hfac)\* and Ag(hfac)\* species regenerating O\* groups for further Ce(thd)4 exposures. We propose that during Ce(thd)4 treatments, not all –OH or regenerated O\* groups can react with the precursor or, as mentioned above, hfac ligands remain bound to the surface and cannot be completely removed by ozone. This statement is supported by XPS measurements, where a high level of impurities was noticed, and is discussed in detail in Section 3.2. It could be the reason for formation of Ag2O clusters in the films with higher concentration of Ag doping, as will be considered during AFM and SEM analysis. With more CeO2 ALD cycles, more nucleation sites are generated and this facilitates the growth of CeO2 thin film with Ag as a dopant.
