Temperature measurements.
In order to acquire the temperature as a function of distance, a fiber-optical temperature (FOT) probe (0.2 K precision, 0.55 mm out diameter, model TS5, Weidmann, Germany) was placed in front of each channel 1 to 4 in succession, oriented to face away from the gas flow direction. The distance was controlled either with a motorized stage (MTS50/M-Z8 in the XYZ configuration with TSC 001 motor control, Thorlabs, USA) controlled with Kinesis (Version 1.14.37, Thorlabs, USA) for PJAS I-1 or manually with a linear stage for PJAS I-2 and II-2. The FOT probe was placed in front of each nozzle to measure the core of the outlet flow as well as sideways for radial profile measurements. Despite previous measurement routines, the present detection system was operated at constant movement and constant data acquisition of the FOT probe. Thereby, a more detailed profile of the temperature as a function of distance could be resolved [29]. For the temperature profile over distance, the motorized stage moves at a constant velocity of 0.15 mm s−1 with the FOT probe moving either constantly away or toward the capillary orifice. The profile for each channel of the PJAS as well as the PJ was acquired. For the overall temperature map of the PJAS, the FOT probe also moved continuously sideways
over the full array of channels with a velocity of 0.5 mm s−1. The distance to the outlet was stepwise increased at 1 mm resolution.

Patient leakage current.
In order to determine the PLC, a similar operation of the motorized stages was used. A copper strip on a plastic surface serves as a counter electrode. The counter electrode was connected via an RC-circuit (according to DIN EN 60 601-1-6 via UNIMET® 800ST, Bender GmbH & Co. KG, Gruenberg, Germany) to ground [30, 31]. The motorized stage moves the grounded electrode toward the capillary orifice at a velocity of 0.1 mm s−1 from a distance of 10 mm toward 1 mm. The measurements were performed for all four channels simultaneously and for individual channels as well. For the latter setting, the grounded electrode is shielded with a plastic part except for an access area of 1 × 1 cm−2 in front of the investigated channel. Data acquisition for temperature and PLC measurements
was performed with the homemade Python code PlaDinSpec (Python 3.8).

Optical emission spectroscopy.
OES was employed to identify the emitting species as a function of distance by the PJAS. All OES measurements were performed with the PJ impinging on a grounded metallic mesh over a quartz plate similar to a setup described previously [32]. A spectrometer (Avantes, AvaSpec 3648, 200–1100 nm range, ∼2 nm resolution, 300 line mm−1 grating, blazed at 300 nm, deep-UV-detector coated CCD linear array) with an optical fiber connected to a cosine corrector was employed for OES measurements. The optical fiber was placed in front of each PJAS channel in succession at 80 ms integration time and 100 averages. The OES system was calibrated for absolute light intensity measurements. The distance between the plasma outlet and target was varied from 3 mm to 25 mm.

UV absorption spectroscopy
Long-lasting reactive species in the gas phase were measured using a commercial gas detector system (model APOA-360 for O3 and APNA-370 for NO2, Horiba, Japan). The PJAS was placed in front of the detection tube (end on, 0◦) at a distance of up to 100 cm. In addition, measurements were performed with the PJAS at an 90◦ impact angle on a dielectric plate at 10 mm distance and the detection tube placed at an angle of 45◦ or 90◦ toward the PJAS at a distance of up to 100 cm away (see further [28]).