*2.1. Catalyst Preparation*

The deposition of Ag doped cerium oxide was carried out using an F-120 ALD reactor (ASM Microchemistry Ltd., Espoo, Finland). Thin films were deposited with different doping concentrations at a reaction temperature of 200 ◦C. The cyclic nature of ALD means that pulses of dopant can easily be incorporated into the main process. The desired composition of catalytic thin film can be achieved by depositing *n* cycles of the base CeO2 material (where *n* can be varied to suit the required doping level) with one cycle of the doping material inserted (Figure 1). The supercycle (*n* + 1), which consists of two individual ALD processes, was repeated *x* times until the required film thickness was achieved. The process for CeO-based material contains two half-cycles using 2,2,6,6-tetramethyl-3,5-heptadionatecerium Ce(C11H19O2)4 (Ce(thd)4 for brevity) and O3 as precursors. The Ag doping material also comprises two half-cycles of Ag precursor (see below) and O3.

**Figure 1.** A schematic representation of the atomic layer deposition (ALD) supercycle used to deposit Ag doped CeO2 catalytic thin films.

Ce(thd)4 (Volatec, Porvoo, Finland) and trimethylphosphine (hexafluoroacetylacetonalo)-silver Ag(CF3COCHCOCF3)P(CH3)3 ((hfac)Ag(PMe3), 99%; Strem Chemicals, Newburyport, MA, USA) were used as Ce and Ag precursors respectively. Ozone O3 was used as the co-reactant in both cases and was generated by an ozone generator (Wedeco Modular 4HC Lab, Herford, Germany) from a pure oxygen (>99.999%) source. Ozone concentration was 120 g/m3. Nitrogen (>99.999%) was used as a carrier and purge gas between precursor pulses. The pressure in the reactor was approximately 1 mbar. Ce(thd)4 and (hfac)Ag(PMe3) were evaporated at 160 and 80 ◦C, respectively to achieve sufficient vapour pressure. The saturated deposition rate in the ALD supercycle should be obtained when the two individual ALD processes are in saturation. We used the previously optimized CeO2 ALD process parameters: 1.5 s Ce(thd)4 dose, 2.5 s purge, 2.5 s O3 dose, 2.5 s purge [28]. The pulse time for (hfac)Ag(PMe3) was varied from 0.5 to 4 s in 0.5 s steps while keeping the O3 pulse time constant at 2.5 s. After finding the optimal pulse time for (hfac)Ag(PMe3) the pulse time for O3 was determined with the same method with 2.5 s of the purge time.

In order to achieve doping of CeO2 with Ag, one supercycle consisted of *n* CeO2 cycles, with *n* equal to 10, 20, and 30, and 1 cycle of Ag. The supercycle was repeated 150, 75, and 50 times for CeO2:Ag ratios 10:1, 20:1, and 30:1 respectively in order to achieve comparable film thicknesses.

Silicon substrates <100> (Si-Mat, Kaufering, Germany) were used for process development while stainless steel foil AISI 316 with a thickness of 0.025 mm (Goodfellow Cambridge Ltd., London, UK) was used as a substrate for soot burning tests. Stainless steel foil was chosen because of its relatively low weight, which reduced the error during weighing of samples to determine the amount of soot oxidation. The substrates were cut in pieces of 20 mm × 10 mm and cleaned using an ultrasonic bath with acetone, isopropanol, and deionized water consecutively, each with a time of 5 min and thereafter dried using compressed air.
