*2.9. Sterility of the Experimental Set-Up (Positive Control)*

To prove the sterility of the different experimental set-ups, the air-powder-water spray was directed to blood agar plates (two per group, six in total). The sterility of the nutritional solution was proved by breeding it at 37 ◦C for 24 h, and then spreading it on a blood agar plate. The agar plates were incubated at 37 ◦C for 24 h, and then CFUs were counted.

#### *2.10. Quality of Biofilm (Negative Control)*

To prove the quality of the biofilm, two test implants from each group (12 in total) served as the negative control and were not cleaned. After biofilm formation, the nutritional solution was rinsed away with sterile water. One implant per group (six in total) was rinsed with sterile water (Aqua ad injectabilia, Braun, Melsungen, Germany) and dried in air for 24 h. The usual fixation in ethanol was not done, to avoid washing-out of the matrix. The samples were gold-coated and examined using a Philips SE XL30 (LaB6 cathode) scanning electron micrograph (SEM) with 20 kV power and a spot size 4 to check biofilm formation and quality. One implant per group (six in total) was incubated in an Eppendorf tube with 1.5 mL nutritional solution and bred at 37 ◦C for 24 h. In anticipation of floating bacteria, the nutritional solution was spread on blood agar and bred at 37 ◦C for 24 h. Colony forming units (CFUs) were then counted (Figure 5).

**Figure 5.** Bacterial growth with the nutritional solution of the negative control (not treated). Dilution grade 1:1,000,000.

### *2.11. Vitality of Bacteria after Cleaning*

To prove the vitality, the waste from both cleaning methods (electrolytic and PSS) was spread on blood agar and bred at 37 ◦C for 24 h. Then CFUs were counted.

#### *2.12. Counting Colonies*

The number of colonies that had grown on the blood agar plates were manually counted using the ImageJ (version 1.51u) software.

#### *2.13. Statistics*

The number of colonies that had grown on the blood agar plates after a dilution to 1:1,000,000 where manually counted and processed by the ImageJ (version 1.51u) software.

Continuous variables are reported as mean ± standard deviation. All groups were tested for normality by the Shapiro-Wilk test. Comparisons between the two cleaning methods were performed with the paired sample *t*-test. Comparisons between the surface and material groups for AirFlow were performed with the Kruskal-Wallis test.

A two-sided *p*-value of ≤0.05 was considered statistically significant. Statistical analysis was performed with R (R Foundation for Statistical Computing, Vienna, Austria).
