Modified State Enhanced Actinometry for measuring atomic oxygen density in a micro-scaled atmospheric pressure plasma jet

Plasma-produced O-, H- and N-species have high oxidative power and play major roles in biological functions. Plasmas containing water vapour are studied to produce reactive species such as OH, H₂O₂, HO₂ [1]. Even though the plasma chemistry is complex and not fully understood, atomic oxygen radicals (O) play a crucial role in these plasmas and measuring them is important to enhance our understanding and control.

Two-photon Absorption Laser Induced Fluorescence (TALIF) is widely used for the measurement of O densities, but is experimentally complex and often not compatible with in-situ process control. Actinometry, based on analysing spectral lines from Optical Emission Spectroscopy is an alternative that requires simple experimental hardware, is non-invasive but relies on more complex theoretical analysis of the experimental data. Advanced actinometry methods such as State Enhanced Actinometry (SEA) have recently been developed and are capable of matching TALIF measurements of O in a COST plasma jet.

Here we investigate a modified version of the SEA method by considering a Baysian method for matching experimental and theoretical data and including cascading effects in the theoretical SEA model. Furthermore, the effect of metastable states on the SEA emission lines is investigated by considering the He (3s ¹S => 2p ¹P), 728.1 nm, line is used instead of He (3s ³S => 2p ³P ), 706.5 nm, line. Atomic oxygen densities and mean electron energies were determined in a micro-scaled atmospheric pressure plasma jet (μAPPJ), operated with 1 slm of He and an 0.5% admixture of O₂ and 0.1% of Ar (for actinometry purposes). Plasma power was varied between 0.2 and 3.5 W.

It was found that the Baysian process of matching theory and experiment minimised the associated error to <1%. Both including cascading emission and considering the alternative He line, resulting in significant changes to the derived O densities, in the order of 25-50% for each process. However, because of the lack of accurate data from alternative methods, e.g. TALIF has an accuracy of ~30-50%, it is not possible to determine which model is most appropriate for SEA. It does highlight the currently achievable accuracy of advanced actinometry methods such as SEA.