Abstract
Plasma-liquid interaction is a critical area of plasma science and a knowledge bottleneck for many promising applications. In this paper, the interaction between a surface air discharge and its downstream sample of deionized water is studied with a system-level computational model, which has previously reached good agreement with experimental results. Our computational results reveal that the plasma-induced aqueous species are mainly H(+), nitrate, nitrite, H2O2 and O3. In addition, various short-lived aqueous species are also induced, regardless whether they are generated in the gas phase first. The production/loss pathways for aqueous species are quantified for an air gap width ranging from 0.1 to 2 cm, of which heterogeneous mass transfer and liquid chemistry are found to play a dominant role. The short-lived reactive oxygen species (ROS) and reactive nitrogen species (RNS) are strongly coupled in liquid-phase reactions: NO3 is an important precursor for short-lived ROS, and in turn OH, O2(-) and HO2 play a crucial role for the production of short-lived RNS. Also, heterogeneous mass transfer depends strongly on the air gap width, resulting in two distinct scenarios separated by a critical air gap of 0.5 cm. The liquid chemistry is significantly different in these two scenarios.
Introduction
Cold atmospheric-pressure plasmas are a versatile tool for modern science, showing massive potential in medicine, wastewater treatment, and agriculture. However, a major knowledge bottleneck remains: while the gas-phase chemistry of plasma is relatively well-understood, the transition of reactive species into and their behavior within liquids is not fully characterized.
This is particularly puzzling for short-lived species like atomic oxygen (O) or hydroxyl radicals (OH). Due to their high reactivity, these species should theoretically vanish before reaching a water surface just a few centimeters away. Yet, they are consistently detected in treated water. This study seeks to quantify the mass transfer and chemical pathways that allow these species to exist and interact in the aqueous phase.
Methods and Experimental Validation
The researchers employed a multi-scale, system-level computational model consisting of three coupled modules: a 0D plasma region simulating surface discharge in humid air (53 species, 624 reactions), a 1D air gap region modeling neutral species diffusion, and a 1D liquid region describing the behavior of 33 species within deionized water using Henry’s law for the gas-liquid interface.
To validate the model, results were compared against experimental data using a surface micro-discharge (SMD) setup. The team measured pH, $H_2O_2$, and nitrate/nitrite concentrations, and utilized Electron Spin Resonance (ESR) with spin-trapping reagents to confirm the presence of short-lived radicals like OH and NO in the water.
Key Insights: The 0.5 cm Critical Gap
A central finding of this research is the discovery of two distinct chemical scenarios determined by the air gap width ($L_g$). In Scenario 1 ($L_g < 0.5$ cm), short-lived species like atomic oxygen can diffuse directly across the gap and dissolve into the water. In Scenario 2 ($L_g > 0.5$ cm), the gap is too wide for direct transfer; instead, aqueous species are generated primarily through liquid-phase chemistry triggered by long-lived precursors like ozone ($O_3$).
ROS and RNS Coupling
The study highlights a strong coupling between oxygen and nitrogen species. Aqueous $NO_3$ acts as a vital precursor for short-lived ROS, while species like $OH$ and $HO_2$ are essential for producing short-lived RNS like peroxynitrite ($ONOO^-$). Furthermore, internal "reaction loops" (e.g., $OH \rightarrow HO_2 \rightarrow HO_3 \rightarrow OH$) ensure that while individual species react rapidly, the net concentration of short-lived ROS remains relatively stable.
Results Summary
- Density Peaks: The densities of aqueous $O_3$ and $H_2O_2$ peak at an air gap of approximately 0.5 cm before declining.
- Nitrite vs. Nitrate: Concentrations consistently decrease as the air gap increases, as their gaseous precursors ($HNO_2$, $HNO_3$) have less chance to reach the surface.
- Model Accuracy: Numerical results generally stayed within a 1.2 to 7.2-fold range of experimental measurements, demonstrating strong agreement given the complexity of the chemistry.
Limitations and Future Work
The study fixed the discharge power at $0.05 W/cm^2$. At higher powers, the system might enter an "ozone poisoning" mode where chemistry changes drastically. Additionally, the use of deionized water is a simplification; real-world applications involving organic molecules would likely see faster reactions that could shield deeper liquid layers. Future research will investigate varying discharge powers and more complex liquid compositions, such as buffered solutions or organic-rich media.
Summary updated