Water Modification by Cold Plasma Jet with Respect to Physical and Chemical Properties

Abstract

This work is devoted to unbuffered and buffered water treatment by means of atmospheric pressure cold plasma of electrical discharges. The interest in the activation of these two liquids by plasma-induced, gaseous-phase chemistry ranges over a wide area of potential applications and interdisciplinary scientific fields. These include biology, medicine, sanitation, environmental restoration, agriculture, etc. Atmospheric pressure cold plasma is here produced in the form of a plasma jet and set into physical contact with the liquid specimens. The operational window of the treatment, in terms of plasma reactivity, is determined by means of UV-NIR optical emission spectroscopy, and the treated liquids are probed in a variety of respects. Evaporation rate, temperature, acidity and basicity, resistivity, and oxidation-reduction potential are measured as a function of the treatment time, either in-situ or ex-situ. The formation of principal reactive oxygen species, i.e., ^•OH, H₂O₂ and ${\mathrm{O}}_2^{•-}$, with a plasma jet mean power lower than 400 mW, is eventually demonstrated and their concentration is measured with original methods borrowed from the biology field. The experimental results are linked to reports published over the last ten years, which are compiled in a brief but meaningful review.

1. Introduction

Interest continues to grow worldwide in weakly ionized, low temperature, atmospheric pressure plasmas [1]. Plasmas generated and maintained at atmospheric pressure enjoyed a renaissance in the last decades mostly driven by the employment of dielectric-barrier discharges (DBDs) and plasma jets (DBD-based or not) as sources of high-density, reactive species. Such plasmas avoid the restrictions imposed by costly, cumbersome vacuum chambers and by destructively high temperatures, extending their applicability in the fields of defense [2], electronics [3], material science [4], environmental health [5], aviation [6], biomedicine [7], agriculture [8], food industry [9], etc. However, despite their apparent engineering simplicity, these systems vastly complicate the plasma physics and chemistry involved, particularly during their incidence on liquid-phase targets. The prominent interest in the latter case is prompted by attempts to interpret fundamental physicochemical phenomena along the plasma–liquid interface [10,11,12,13] and to properly exploit the properties of the plasma activated liquids for tailored applications [14,15,16].

Among the various liquids, a special focus has been given on the water treatment by atmospheric pressure cold plasmas with respect to remediation [17], sterilization [18], fertilizer alternatives [19], dentistry [20], food preservation [21], nanoscience [22], etc., and the concept has already been commercialized (e.g., VitalFluid BV). Although it is widely accepted that, in most applications, the plasma induced effects are mainly mediated via reactive oxygen and nitrogen species (ROS and RNS, respectively; cumulatively designated RONS), influencing redox regulated processes [23], detailed consideration of the physical and chemical impacts of the plasma species on the liquids is a prerequisite for any targeted application. This need is yet more pressing when living specimens are involved. Such dedicated, worthy representative studies, unveiling the potential plasma-induced modifications of water (and other liquids too), are laid out in Appendix A in chronological order over the last ten years. The works cited therein denote the wide interest in the topic of plasma-based modification of water, as well as the divergence of setups and diagnostic methods used in various case studies.

However, despite the intensive research investments during the recent decennia, the field is plagued by controversies and gaps in knowledge, which might restrict further progress [24]. Within this context, the present work is an attempt to enrich the available data by communicating experimental results acquired with one of the most popular plasma types employed in liquid treatment. Accordingly, an atmospheric pressure plasma jet, consuming mean power less than 400 mW, is optimized in terms of reactive species formation when the carrier gas is argon (Ar), and this optimal operational window is applied to the activation of demineralized, doubled-distilled water (DDW) and phosphate-buffered saline (PBS). The dynamic evolution of the evaporation rate, temperature, acidity and basicity, resistivity, and oxidation-reduction potential are recorded. Finally, the formation of principal ROS, i.e., ^•OH, H₂O₂ and ${\mathrm{O}}_2^{•-}$, is demonstrated and their concentration is measured by means of strict protocols that have not previously tried on plasma-treated liquids.

2. Experimental Setup and Materials

2.1. Plasma Setup and Diagnostics

Figure 1 depicts the concept of the experimental setup used for the plasma production and characterization. In the caption of the same figure, the distinct components are described.

The design of the plasma jet reactor is based on a coaxial DBD configuration. This reactor has been efficiently employed in the past for bacterium [25] and cancer cell [26] inactivation, human skin treatment [27], liposome disruption [28], and other studies in the field of biomedicine. It comprises: (i) a capillary alumina tube (Ø_out = 2.5 mm; Ø_in = 1.14 mm; length of 76 mm) acting as the dielectric barrier; (ii) a filamentary internal electrode (made of tungsten; Ø = 0.125 mm) inserted into the tube and terminated 20 mm behind the tube orifice; (iii) a cylindrical hollow external electrode (made of brass; Ø_out = 10 mm; length of 10 mm) tightly fixed around the tube. Its outer edge is placed 20 mm behind the tube orifice forming thus a 10 mm in length coaxial DBD with the internal electrode. The internal electrode is the driven one and the external is grounded directly. The power supply is designed by PlasmaHTec^® (plasmahtec.com, accessed on 20 November 2022), and it consists of a signal generator (based on direct signal synthesis; DDS), a power amplifier of low harmonic distortion, and a step-up, ferrite-core, high voltage transformer of special design. The specifications refer to sinusoidal waveform, rated voltage 12 kV peak-to-peak, rated power 210 W, and frequency range 1–20 kHz. The working gas is high purity argon (Ar; 99.999%) and it is introduced into the dielectric tube through a digital mass flow controller (0–5 slm).

The voltage and current waveforms are continuously monitored during the experiments for consistency reasons (voltage divider: 1000:1, DC – 75 MHz; current monitor: 400 Hz–250 MHz; oscilloscope: 400 MHz, 5 GSamples s⁻¹). Two identical current transformers are used for recording both the DBD and the plasma jet currents, as it is shown in Figure 1. At the same time, the mean power delivered to the system is measured as the mean value of the voltage-to-current product over successive periods. Thus, the DBD and the plasma jet consumed mean powers are distinguished and quantified. More concretely, the instantaneous power $v\left(t\right)\times i\left(t\right)$ is numerically integrated over 3 voltage cycles T and 30 such samples are averaged. The current $i\left(t\right)$ stands for either the DBD current $i_{\mathrm{DBD}}\left(t\right)$ or the jet current $i_{\mathrm{JET}}\left(t\right)$. Formulas (1) and (2) summarize the process.

$$\overline{P_{\mathrm{DBD}}}=\frac{1}{30}{\displaystyle \sum }_1^{30}\left[\frac{1}{3T}\underset{0}{\overset{3T}{{\displaystyle \int }}}v\left(t\right)\times i_{\mathrm{DBD}}\left(t\right)\mathrm{d}t\right]$$

(1)

$$\overline{P_{\mathrm{JET}}}=\frac{1}{30}{\displaystyle \sum }_1^{30}\left[\frac{1}{3T}\underset{0}{\overset{3T}{{\displaystyle \int }}}v\left(t\right)\times i_{\mathrm{JET}}\left(t\right)\mathrm{d}t\right]$$

(2)

Figure 2 provides representative waveforms of the corresponding quantities. The physical interpretation of these signals has been reported previously [29].

The optical emission of the plasma is probed with a 0.25-m imaging spectrograph equipped with two motorized gratings; one holographic (2400 grooves mm⁻¹; blaze wavelength 250 nm; 180–700 nm) and one ruled (600 grooves mm⁻¹; blaze wavelength 400 nm; 250–1300 nm). The photodetector is a linear charge-coupled device (200–1100 nm). A high-grade fused silica optical fiber is mounted on an axial translator for side-on observations. A fused silica lens, 19 mm focal length and F/1.7, collects light along an about 15 mm zone downstream of the reactor orifice. Alternatively, high resolution optical emission spectra are acquired with a 1-m monochromator equipped with a holographic grating (170–750 nm; 2400 grooves mm⁻¹) and a photoelectron multiplier tube (185–900 nm). In both cases, the optical fiber is optimally aligned with the spectrometers using suitable F-number matchers. The assembly of the optical components is calibrated in terms of absolute wavelength and relative spectral efficiency, using Hg(Ar) and quartz–tungsten–halogen lamps, respectively. Rotational distributions of probe molecules are numerically constructed and fitted to the experimental ones with a home-built software [30].

2.2. Basic Characterization of the Liquids

The probed liquids in this work are lab-prepared, demineralized, double distilled water (DDW) and phosphate-buffered saline (PBS). Four different physicochemical parameters are studied. (i) Temperature: it is measured in-situ (during treatment) or ex-situ (before/after treatment) by employing a non-invasive technique, i.e., a GaAs-crystal based, optical thermometer (LumaSense Technologies Inc.; Santa Clara, CA, USA) [31]. The data are collected by a datalogger, with recommended standard 232 communication, and time resolution of 0.5 s. (ii) Conductivity: it is measured using a bench meter (Mettler–Toledo; FP30) equipped with a probe with an integrated temperature sensor (LE703); (iii) pH: it is measured using a bench meter (Mettler–Toledo; FP20) equipped with a pH micro-probe (LE422). (iv) Redox potential: it is measured by employing the same bench meter as for pH, but with an appropriate redox electrode (Mettler–Toledo; LE501). Quantities (ii)–(iv) are reduced to the actual temperature of the liquid, with the latter to be measured within the first 10 s before/after plasma treatment. Conductivity and pH meters are both calibrated before any series of experiments, using standard liquids provided by Mettler–Toledo (solution of σ = 1413 μS cm⁻¹; buffers of pH = 4.0, 7.0, and 9.2).

Evaporation tests are realized with micro-balance measurements (Highland; HCB 123; readability 0.001 g; repeatability 0.002 g). The initial, i.e., the before plasma treatment, quantity of the water is fixed at 4.5 ± 0.007 g. As will be seen in the results below, liquid evaporation indeed takes place during plasma treatment; however, even for a prolonged time (30 min), it does not exceed a percentage of about 15%. Accordingly, the present treatment protocol makes a compromise and does not include any refill which could disturb the experiment.

It is noted that all experimental points in this work are obtained cumulatively, i.e., for each series of experiments, a water specimen of 4.5 ± 0.007 g is used and subjected to the process for successive time intervals. At the end of each interval, the liquid is probed ex-situ and then brought back to the process immediately. In this way, the study of the cumulative action of the plasma on a specific specimen is secured for each series of experiments. The liquid is inserted in a standard 10 mL beaker, while the distance between the orifice of the plasma reactor and the surface of the liquid is set at about 25 mm. The argon physical flow and the electro–hydrodynamic effects [32] induced by the plasma jet itself provide adequate agitation of the liquid specimen (see inset in Figure 1).

2.3. ROS Detection in the Liquids

Apart from the basic liquid characterization, before and after the plasma treatment, various reagents, standard, and special solutions are used for the detection of principal chemical radicals in both the DDW and the PBS. The materials and the procedures are given below in detail.

Reagents

-

Hydroethidine (HE; Fluka, Buchs, Switzerland; cat. no. 37291).

-

Dimethyl sulfoxide (DMSO, 99.9% or 14 M; ChemLab, Zedelgem, Belgium; cat. no. CL00.0422).

-

Hydrochloric acid (HCl, 37% w/w or 12 M; ChemLab, Zedelgem, Belgium; cat. no. CL00.0310).

-

2-hydroxy terephthalate (2-OH-TPA, 98%; TCI, Portland, USA; cat. no. H1385).

-

2-hydroxy ethidium (2-OH-E⁺, synthesized as described elsewhere [33]).

-

Sodium hydroxide (NaOH, pellets; Merck, New Jersey, USA; cat. no. 106498).

-

Phosphate Buffered Saline (PBS powder, for preparing 1 L, pH 7.4; Sigma-Aldrich, Burlington, MA, USA; P3813).

-

Terephthalic acid (TPA, 98+%; Alfa Aesar, Kandel, Germany; cat. no. A12527).

-

DDW, purified by a Milli-Q system (Millipore, Massachusetts, USA).

-

Acetonitrile (ACN, HPLC grade; Sigma-Aldrich, Burlington, MA, USA; cat. no. 34851).

-

Chloroform (CHCl₃; Merck, New Jersey, USA; cat. no. 1.02445).

-

Methanol (MetOH, ≥99.9%; Merck, New Jersey, USA; cat. no. 106009).

Standard solutions

-

5 mM TPA, pH 7: Dissolve 4.15 mg TPA in 1 mL 0.1 M NaOH and ddH₂O is added to 5 mL final volume. Adjust pH to 7, by dropwise addition of 0.5 M HCl. Long storage at −20 °C.

-

1 M DMSO: prepare 1 mL by mixing 71.4 μL 14 M DMSO with 928.6 μL DDW.

-

0.4 M HCl: prepare 2 mL by mixing 66.7 μL 12 M HCl with 1933.3 μL DDW.

-

5 M NaOH: dissolve 0.4 g in DDW to final 2 mL.

-

20 mM HE: dissolve 6.3 mg HE in 1 mL 0.4 M HCl. Long storage at −20 °C.

-

PBS, pH 7: mix the content of a PBS bag with 1 L DDW.

-

1 M DMSO in PBS, pH 7: prepare 1 mL by mixing 71.4 μL 14 M DMSO with 928.6 μL PBS.

-

5 mM 2-OH-TPA, pH 7: dissolve 1.4 mg HTPA in 0.3 mL 0.1 M NaOH and add DDW to final 1.5 mL. The pH is neutralized to 7, by addition of about 10 μL 1 M HCl. Long storage at −20 °C.

-

0.2 mM 2-OH-E⁺ (in 0.4 M HCl) prepared from synthetic 2-OH-E⁺ as described elsewhere [33]. Long storage at −20 °C.

-

50 mM Pi buffer, pH 7.8: dissolve 0.45 g Νa₂HPO₄.2H₂O in about 48 mL DDW, final 50 mL DDW, adjust to pH 7.8 (dropwise with 1 M HCl) and bring to final 50 mL with DDW.

TPA-based ROS detection solutions: for ^•OH and H₂O₂

-

100 μM TPA, pH 7: prepare 8 mL by mixing 0.16 mL 5 mM TPA with 7.84 mL DDW, and if needed adjust pH to 7.

-

100 μM TPA + 50 mM DMSO, pH 7: prepare 8 mL by mixing 0.16 mL 5 mM TPA with 0.4 mL 1 M DMSO and 7.44 mL DDW and, if needed, adjust pH to 7.

-

100 μM TPA in PBS, pH 7: prepare 8 mL for each experiment by mixing 0.16 mL 5 mM TPA with 7.84 mL PBS and, if needed, adjust pH to 7.

-

100 μM TPA + 50 mM DMSO in PBS, pH 7: prepare 8 mL for each experiment by mixing 0.16 mL 5 mM TPA with 0.4 mL 1 M DMSO and 7.44 mL PBS and, if needed, adjust pH to 7.

HE-based ROS detection solutions: for ${\mathrm{O}}_2^{•-}$

-

100 μM HE, pH 7: prepare 8 mL for each experiment by mixing 40 μL 20 mM HE with 7.96 mL DDW. Adjust pH to 7 by dropwise addition of 5 M NaOH.

-

100 μM HE in PBS, pH 7: prepare 8 mL for each experiment by mixing 40 μL 20 mM HE with 7.96 mL PBS and, if needed, adjust pH to 7.

Detection of ^•OH and H₂O₂

^•OH and H₂O₂ are measured in the 8-mL ROS (detecting) solutions containing TPA (specific probe for ^•OH), TPA+DMSO (DMSO is specific scavenger for ^•OH; serves as plasma-generated ^•OH elimination control), TPA in PBS, and TPA+DMSO in PBS, which are each placed in 10-mL glass beakers. After plasma exposure, ^•OH formation in these solutions is measured by the fluorescence units (FU) of 2-OH-TPA (resulting from the reaction of 1 mole ^•OH with 1 mole TPA; DMSO is used as negative control showing that it eliminates ^•OH and cancels FU from 2-OH-TPA not formed due to the unavailability of ^•OH). FU are measured at ex/em 315/425 nm (using a Shimadzu spectrofluorometer, model RF-1501, set at a 10 nm slit width and at low sensitivity) [34,35]. Specifically, in 0.3 mL fractions of each undiluted TPA-based ROS solution, 2-OH-TPA FUs are converted to molarity against the FU of 1 μM synthetic 2-OH-TPA. The TPA-containing ROS solutions are also tested (diluted at 4,000- to 16,000-fold) for H₂O₂ formation by a sensitive ferrous oxidation–xylenol orange (FOX) photometric assay, described elsewhere [36].

Detection of ${\mathrm{O}}_2^{•-}$

${\mathrm{O}}_2^{•-}$ is measured in the 8-mL ROS solutions containing HE and HE in PBS, which are each placed in 10-mL glass beakers. After plasma exposure, the content of the beakers is tested for ${\mathrm{O}}_2^{•-}$ formation via the quantification of 2-OH-E⁺ (generated by the reaction of 2 moles ${\mathrm{O}}_2^{•-}$ with 1 mole HE). 2-OH-E⁺ is extracted from 1 mL of each HE-based ROS solution by a mixture of 333 μL 100% ACN and 1.333 mL CHCl₃, and vacuum dried in centrifugal vacuum concentrator (CHRIST; model RVC 2–18), connected to a vacuum pump (KNF; N 820.3 FT.18). The resulting red pellet is solubilized in 50 μL CHCl₃:MetOH 70%:30% and 2 μL are placed in an HPTLC silica gel plate (Merck 1.05644.0001), as described elsewhere [37], in order to verify and quantify 2-OH-E⁺ production. Following another independent methodology for ${\mathrm{O}}_2^{•-}$ quantification, the remaining 48 μL are vacuum-dried and resolubilized in 300 μL 50 mM Pi buffer, pH 7.8, and its FU are measured at ex/em 480/575 nm (using a Shimadzu spectrofluorometer, model RF-1501, set at a 10 nm slit width and at high sensitivity) as described elsewhere [33]. 2-OH-E⁺ is quantified via its fluorescence quenching after ± treatment with the horseradish peroxidase (HRP)/H₂O₂ system. 2-OH-E⁺ FU (after plasma exposure) are converted to molarity against the FU of a known concentration of synthetic 2-OH-E⁺.

3. Results and Discussion

3.1. Plasma Jet Operational Window

The present DBD-based plasma jet setup has three adjustable operating parameters, namely: the applied voltage amplitude, the applied voltage frequency, and the gas flow rate. These parameters are here varied over the range 8–12 kV peak-to-peak, 5–20 kHz, and 0.5–3 slm, respectively, and the generating species emitting in the UV-NIR region are probed. Figure 3a,b unveil the principal emissive species detected downstream of the plasma jet. The species identification with the corresponding transitions are summarized in

Table 1. Principal emissive species in the argon plasma jet.

Species	λ (nm)	Vibrational Transition [38] v′–v″ (Δv)	Lower Level El (eV)	Upper Level Eu (eV)	Radiative Lifetime τ (ns)
$\mathrm{OH}:{\mathsf{A}}^2{\mathsf{\Sigma }}^{+}\to$ X2Π	309 [38]	0–0(0)	-	4 [38]	690 [38]
${\mathrm{N}}_2\left(\mathrm{SPS}\right):{\mathrm{C}}^3{\mathsf{\Pi }}_{\mathrm{u}}\to$ B3Πg	337.13 [38]	0–0(0)	-		36.6–42 ± 2 [38]
${\mathrm{N}}_2\left(\mathrm{SPS}\right):{\mathrm{C}}^3{\mathsf{\Pi }}_{\mathrm{u}}\to$ B3Πg	357.69 [38]	0–1(−1)	-		-
${\mathrm{N}}_2\left(\mathrm{SPS}\right):{\mathrm{C}}^3{\mathsf{\Pi }}_{\mathrm{u}}\to$ B3Πg	380.49 [38]	0–2(−2)	-		-
Ar I: (2P01/2)4p → (2P03/2)4s	696.54 [39]	-	11.54 [39]	13.32 [39]	-
Ar I: (2P01/2)4p → (2P03/2)4s	706.72 [39]	-	11.54 [39]	13.30 [39]	29 [38,40]
Ar I: (2P01/2)4p → (2P03/2)4s	727.29 [39]	-	11.62 [39]	13.32 [39]	28.3 [38,40]
Ar I: (2P01/2)4p → (2P03/2)4s	738.39 [39]	-	11.62 [39]	13.30 [39]	29 [38,40]
Ar I: (2P03/2)4p → (2P03/2)4s	751.46 [39]	-	11.62 [39]	13.27 [39]	24.4 [38,40]
Ar I: (2P03/2)4p → (2P03/2)4s	763.51 [39]	-	11.54 [39]	13.17 [39]	29.4 [38,40]
Ar I: (2P01/2)4p → (2P01/2)4s	772.42 [39]	-	11.72 [39]	13.32 [39]	28.3 [38,40]
Ar I: (2P01/2)4p → (2P01/2)4s	794.81 [39]	-	11.72 [39]	13.28 [39]	29.3 [38,40]
Ar I: (2P03/2)4p → (2P03/2)4s	801.47 [39]	-	11.54 [39]	13.09 [39]	30.6 [38,40]
Ar I: (2P03/2)4p → (2P03/2)4s	811.53 [39]	-	11.54 [39]	13.07 [39]	30.7 [38,40]
Ar I: (2P01/2)4p → (2P01/2)4s	826.45 [39]	-	11.82 [39]	13.32 [39]	28.3 [38,40]
Ar I: (2P01/2)4p → (2P01/2)4s	840.82 [39]	-	11.82 [39]	13.30 [39]	29 [38,40]
Ar I: (2P03/2)4p → (2P03/2)4s	842.46 [39]	-	11.62 [39]	13.09 [39]	30.6 [38,40]
Ar I: (2P01/2)4p → (2P01/2)4s	852.14 [39]	-	11.82 [39]	13.28 [39]	29.3 [38,40]
Ar I: (2P03/2)4p → (2P03/2)4s	912.29 [39,40]	-	11.54 [39,40]	12.90 [39,40]	-
Ar I: (2P03/2)4p → (2P01/2)4s	922.45 [39,40]	-	11.82 [39,40]	13.17 [39,40]	29.4 [40]

[38,39,40]. Excited molecular nitrogen, hydroxyl, and abundant argon excited atoms are seen. Figure 3c presents both the experimentally recorded and the numerically reproduced rotational distribution of the hydroxyl molecule, OH(A), which has been accepted as a reliable molecule for rotational temperature measurements [41]. Hence, following the curve fitting shown in Figure 3c, an average gas temperature of around 350 ± 40 K is evaluated along the plasma jet. The data of Figure 3 refer to the optimal operational window, in terms of reactivity, which is determined as follows.

Figure 4, Figure 5 and Figure 6 provide the relative spectral intensity of the abovementioned species, as a function of the voltage amplitude, voltage frequency, and gas flow rate, respectively. In the case of amplitude variation, i.e., Figure 4, species population increases for higher externally applied electric field. This is due to intense ionizing waves, propagating along the argon channel, which outplace and penetrate the atmospheric air [29,42]. Thus, considering the power supply limit of 12 kV peak-to-peak and a rational safety factor, the voltage of 11 kV peak-to-peak is used hereafter.

In the case of frequency variation, i.e., Figure 5, the OES intensity peaks around 17.5 kHz for most of the species. In other words, while an increasing frequency promotes species production initially, possibly due to the higher power injected into the system, elevated frequency leads to adverse results eventually. This fact may be attributed to the transition from laminar to turbulent gas flow profile (see discussion on the frequency role in reference [42]), which provokes increased argon-air mixing and disturbs the ionization wave propagation. Thus, the value of 17.5 kHz is applied hereafter.

Regarding the gas flow rate, i.e., Figure 6, the intermediate value of 1.5 slm appears favorable to produce most species. The existence of an optimal gas flow is directly related with the breakdown of the laminar flow into turbulent flow for increased gas flow rates. This electro–hydrodynamic interplay between the plasma species and the gas jet has been studied extensively by our group [32,43,44]. Consequently, the value of 1.5 slm is considered henceforth.

Two comments should be made on the above spectroscopic analysis: (i) The emission which is related to the carrier gas (i.e., argon atomic lines) has consistent peaks and patterns against all the operating parameters. Similar patterns and peak locations are observed for the hydroxyl molecule too. However, the emission intensity which is related to the surrounding air (i.e., nitrogen molecule bands) deviates from the above patterns when either the frequency or the gas flow rate is being swept. Thus, it peaks at around 12.5 kHz and 1 slm (i.e., it then decays) and then it somewhat recovers when the frequency and gas flow rate get values close to those corresponding to the peaks of all other species (see N₂ curves in Figure 5a for >17.5 kHz and Figure 6a for >2 slm; the peaks in Figure 5b and Figure 6b are around 17.5 kHz and 1.5 slm, respectively). These observations mirror once more the displacement of the nitrogen molecules by the argon atoms and vice versa, depending on the frequency and flow rate values contextually. (ii) The results of Figure 3, Figure 4, Figure 5 and Figure 6 refer to the free running plasma jet, i.e., without liquid target which could modify plasma features [45]. In the present work, comparative experimental series during months with and without DDW/PBS, showed that DDW/PBS does not promote the production of new emissive species. Furthermore, any variation in the relative spectral intensity of the above species when DDW/PBS is being treated remains within the typical standard deviations obtained in the case of the free running plasma jet. Finally, the patterns and the peak locations of Figure 4 to Figure 6 remain unchanged.

These observations are possibly related to the following facts: there is quite a long distance between the reactor orifice and the liquid surface (about 25 mm), resulting in weak immersion of the plasma jet into the liquid; the distance between the liquid surface and the beaker bottom is quite long (about 20 mm); between the liquid and the grounded holder, the glass dielectric bottom of the beaker is inserted (Figure 1).

However, the bar charts of Figure 7 picture the influence of the DDW and PBS on the DBD and the plasma jet in terms of power consumption. The system consumes a low power (<2.5 W) with the main fraction of this power being delivered always to the DBD, but the ratio of the plasma jet to the DBD power varies slightly. It is around 2%, 9%, and 11% in the case of the free plasma jet, the DDW, and the PBS target, respectively. Nevertheless, as the optimum operational window is unaffected (i.e., 11 kV, 17.5 kHz, and 1.5 slm), any further concern on the liquid-induced plasma modifications remains beyond the scope of the present study. This study focuses on the plasma-induced liquid modifications.

3.2. Liquid Treatment—Basic Characteristics

Hereafter, the specimens are treated at the optimized conditions found previously. Firstly, the liquid mass variation versus the time is given in Figure 8 for three different cases, i.e., the specimens are settled in the laboratory air, the specimens are subjected to the argon flow (plasma off) downstream of the reactor orifice, and the specimens are subjected to the argon plasma. For the prolonged time of 30 min, in the first case, the mass of both liquids remains constant, in the second case, a mass reduction of 7% in DDW and 8.5% in PBS is observed, and in the last case, the corresponding percentages are 14.5% and 15.6%.

Figure 9 gives the liquid temperature variation (%) at the corresponding cases, following in-situ measurements. The representation of the temperature relative variation instead of absolute values is preferable since it offsets the different ambient temperature on different dates. The given error bars indicate the highest (worst case) deviations found between two series of experiments. Both liquids slightly cool down when they are left in the laboratory air, whereas an important cooling is noted when they are subjected to argon flow downstream of the capillary tube. Interestingly, this cooling is intercepted when the plasma is activated. The effects are more intense in the case of unbuffered water.

It is known that when the molecules of a liquid collide, they transfer energy to each other based on how they collide with each other. When a molecule near the surface absorbs enough energy to overcome the vapor pressure, it escapes and enters the surrounding air as a gas. When evaporation occurs, the energy removed from the vaporized liquid reduces the temperature of the liquid, resulting in evaporative cooling. The evaporation continues until an equilibrium is reached when the evaporation of the liquid is equal to its condensation.

Based on the present experimental data, plasma results in a higher evaporation rate than does argon gas itself. At the same time, plasma cools the liquids less than does the argon gas. Thus, a simplified evaporative cooling mechanism is not consistent with these observations. Evaporation should be attributed to momentum and energy transfer from the plasma species to the liquid molecules. More importantly, it is well established that plasma jets increase the gas velocity and its mixing with the surrounding air due to electro–hydrodynamic forces [32,43,44]. Thus, above the liquid surface, the non-saturated air is renewed more efficiently, and the concentration of the liquid vapors goes down when the plasma is activated. This fact encourages faster evaporation. The more intense cooling effect induced by the argon gas is likely due to an “adiabatic-like” cooling that takes place as the gas is leaving the compressed gas reservoir in a short time. On the other hand, small amounts of heat induced by the plasma (see reference [46] and Figure 3c) compensate for this “adiabatic-like” cooling and the temperature variation appears lower.

Apart from the evaporation and the temperature variations, the DDW pH is exponentially decreases versus the treatment time (Figure 10). Within 30 min, a decrease of about two in the pH scale (ΔpH = −1.98) is obtained. Since the pH scale is defined as the decimal logarithm of the reciprocal of the hydrogen ion activity (at that level, this is essentially the number of moles of hydrogen ions per liter of solution), it is deduced that plasma induces an increase of the hydrogen ion concentration in the water. It should be underlined that the initial low pH (around 6.0) is not surprising since, if pure water is exposed to air, it becomes mildly acidic. This happens because water absorbs carbon dioxide from the air which is then slowly converted into carbonic acid (the chemical reaction is reversible). In any case, the observed plasma-induced acidification of water is a crucial result which can account for the deactivation of different biospecimens. On the other hand, PBS pH remains almost constant (ΔpH = −0.18), as was expected for an isotonic solution.

Furthermore, another quantity related to pH and the disinfection efficiency of a liquid is the redox potential, E_h. Redox potential is an indicator of the ability of a solution to oxidize and is related to the oxidizer concentration. Substances like dissolved oxygen will contribute to higher oxidation and higher redox. This substance has higher electronegativity (oxidizing agent) and will normally pull electrons from other substances like unwanted contaminants and bacteria. Under the present conditions, the redox potential of the DDW changes from about 360 mV to 430 mV (ΔE_h = +60.5 mV), in an exponential manner, according to Figure 11a. Consistently with the pH measurements, the PBS redox is less affected even after 30 min of treatment (ΔE_h = +10 mV); Figure 11b.

Finally, Figure 12 bears testament to the ionic species increase in both DDW and PBS as a function of the treatment time. In the case of DDW, an almost seven-fold conductivity is achieved within 30 min of plasma action (Δσ = +30.33 μS cm⁻¹), whereas in the case of PBS, an increase Δσ = +1.34 mS cm⁻¹ is obtained within 20 min. The dashed line in Figure 12b corresponds to a numerical extrapolation of the experimental data, because the remaining PBS in the 10-mL beaker after 20 min of treatment was not sufficient to cover the immersed probe and any measurement taken could not be considered as reliable. This extrapolation suggests an increase Δσ = +3.1 mS cm⁻¹ for the 30-min treatment.

3.3. Liquid Treatment—ROS Detection

Following the above observations on the plasma induced modifications of the water, the next step is their interrogation in terms of ROS formation. Thus, here follows the identification and quantification of the ROS components ^•OH, H₂O₂ and ${\mathrm{O}}_2^{•-}$ generated by the plasma jet in both DDW and PBS.

Table 2. Plasma generated ^•OH and H₂O₂ in DDW and PBS; values are averages from 5 repeats, with standard deviation around 5%.

	•OH (2-OH-TPA) (nM)	H2O2 (μM)	ΔpH
Plasma-exposed: TPA in DDW	565 (565) ± 27	359 ± 17	2 ± 0.02
Plasma-exposed: TPA + DMSO (•OH scavenger) in DDW	0 (0)	336 ± 15	2 ± 0.01
Plasma-non-exposed: TPA in DDW	0 (0)	0 (0)	0 ± 0.02
Plasma-exposed: TPA in PBS	481 (481) ± 23	255 ± 11	0 ± 0.02
Plasma-exposed: TPA + DMSO (•OH scavenger) in PBS	0 (0)	263 ± 13	0 ± 0.02
Plasma-non-exposed: TPA in PBS	0 (0)	0 (0)	0 ± 0.01

and

Table 3. Plasma generated ${\mathrm{O}}_2^{•-}$ in DDW and PBS; values are averages from 5 repeats, with standard deviation around 5%.

	$${\mathbf{O}}_2^{•-}(2-\mathbf{OH}-{\mathbf{E}}^{+})\left(\mathbf{nM}\right)$$	ΔpH
Plasma-exposed: HE in DDW	873 ± 40 (1745 ± 80)	3 ± 0.02
Plasma-non-exposed: HE in DDW	0 (0)	0 ± 0.02
Plasma-exposed: HE in PBS	80 ± 4 (160 ± 8)	0 ± 0.02
Plasma-non-exposed: HE in PBS	0 (0)	0 ± 0.01

summarize the results for the treatment time of 30 min.

The radical ^•OH is identified by its specific probe TPA against the control conditions (i) of non-exposure to the plasma and (ii) exposure in the presence of its specific scavenger DMSO. As ^•OH is being produced in DDW or PBS due to the plasma exposure, in the presence of excess TPA, the concentration of 2-OH-TPA (thus, of ^•OH) increases linearly over time, reaching about 550 nM in DDW at 30 min. H₂O₂ identification and quantification employ an established specific method [36], using as a control condition the non-exposure to the plasma. In addition, the use of the ^•OH-specific scavenger DMSO [47,48] as an additional control condition ensures that H₂O₂ produced in the plasma-treated liquid is not produced by the recombination of 2 moles ^•OH. This control translates to a concentration of about 350 μM of H₂O₂ in DDW after 30 min of plasma exposure. A similar approach to that of the ^•OH is followed for the detection of ${\mathrm{O}}_2^{•-}$ by its specific probe HE with two independent innovative methodologies [33,37]. The non-exposure control condition is also used, and the produced concentration is around 850 nM in DDW after 30 min of plasma exposure. The corresponding concentrations in PBS are about 450 nM, 250 μM, and 80 nM (^•OH, H₂O₂, and ${\mathrm{O}}_2^{•-}$, respectively).

However, the recorded ^•OH and ${\mathrm{O}}_2^{•-}$ levels are considered minimum. The reasons being: (i) these radicals are only those being trapped by their respective probes TPA and HE during 30-min plasma exposure; (ii) their quantification is made by the respective 2-OH-TPA and 2-OH-E⁺ (i.e., products of the reaction of ^•OH and ${\mathrm{O}}_2^{•-}$ with their respective probes TPA and HE) that survived plasma oxidation; and (iii) only part of HE forms 2-OH-E⁺ with ${\mathrm{O}}_2^{•-}$ while the other part is converted to non-specific oxidation products by 1 or 2 electron-oxidation [49,50,51]. Regarding H₂O₂, its concentration is similar in both the DDW and the PBS (about 25% lower in PBS). Another factor adding to the difficulty in obtaining absolute production values for ^•OH, ${\mathrm{O}}_2^{•-}$ and H₂O₂ is their participation in numerous side reactions involving both their generation and destruction [12,14,15,16,52,53,54,55,56,57,58].

The following reactions are given as representative. Apart from the generation/destruction paths of the above-determined radicals, these reactions demonstrate the complicated interplay between the plasma induced species (electrons, photons, radicals, etc.) and the dissolved in the liquid species, which ultimately affect the ion concentration (pH and σ) and the oxidizer concentration (redox) in the liquid. However, the above results are coherent, since a decrease in pH is accompanied by an increase in E_h and σ, with a parallel increase in ROS.

H2O + hf	→	H2O+ + e−	(01)
H2O+	→	•OH + H+	(02)
UV + H2O	→	H2O•	(03)
UV + H2O•	→	H+ + OH−	(04)
OH−	→	•OH + e−	(05)
e− + O2	→	${\mathrm{O}}_2^{•-}$	(06)
e− + O2 + H+	→	${\mathrm{HO}}_2^{•}$	(07)
e− + H2O	→	•OH + •H + e−	(08)
e− + H2O	→	•OH + H+ + 2e−	(09)
e− + H2O	→	•H + •OH	(10)
•H + O2	→	${\mathrm{HO}}_2^{•}$	(11)
${\mathrm{O}}_2^{•-}$$+{\mathrm{O}}_2^{•-}$ + 2H+	→	H2O2 + O2	(12)
${\mathrm{HO}}_2^{•}$ $+{\mathrm{O}}_2^{•-}$ + H2O	→	O2 + H2O2 + OH−	(13)
${\mathrm{HO}}_2^{•}$$+{\mathrm{HO}}_2^{•}$	→	H2O2 + O2	(14)
•$\mathrm{O}\mathsf{H}+{\mathrm{HO}}_2^{•}$	→	H2O + 1O2	(15)
•$\mathrm{O}\mathsf{H}+{\mathrm{O}}_2^{•-}$	→	OH− + O2	(16)
H2O2 + •H	→	•OH + H2O	(17)
•OH + H2O2	→	${\mathrm{HO}}_2^{•}$ + H2O	(18)
•OH + •OH	→	H2O2	(19)

Finally, it is noted that participation in the various side reactions of either HO₂^• or ${\mathrm{O}}_2^{•-}$ or both, depends on their pKa (= 4.8 [59]) and on the pH of the unbuffered water, where their molar ratio is determined by the Henderson–Hasselbalch equation, i.e., pH = pKa + log([${\mathrm{O}}_2^{•-}$]/[HO₂^•]). Thus, in the unbuffered water, for a pH drop from, e.g., 7 to 5, the molar ratio [${\mathrm{O}}_2^{•-}$]/[HO₂^•] changes from about 158 to 1.58.

4. Conclusions

In the present report, buffered or unbuffered water was exposed to the species produced by an atmospheric pressure cold plasma in the form of a plasma jet, operating with argon, driven by audio frequency sinusoidal high voltage, and consuming less than 400 mW mean power. It was shown that radicals are produced in the gaseous phase and their interaction with the liquid targets leads to (i) evaporation, (ii) acidification, (iii) increased oxidizer concentrations, and (iv) increased ionic species, in agreement with previous reports. ^•OH, H₂O₂, and ${\mathrm{O}}_2^{•-}$ were specifically identified and quantified by innovative methodologies.

Analytically, the above-mentioned quantities for 30 min plasma treatment of the unbuffered water were found to be: mass reduction 14.5%, pH decrease 1.98, redox potential increase 60.5 mV, conductivity increase 30.33 μS cm⁻¹, ^•OH concentration 550 nM, H₂O₂ concentration 350 μM, and ${\mathrm{O}}_2^{•-}$ concentration 850 nM. The corresponding values for the buffered solution were found to be: 15.6%, 0.18, 10 mV, 3.1 mS cm⁻¹, 450 nM, 250 μM, and 80 nM. It was demonstrated that these physicochemical changes take place while both liquids maintain a low temperature. These combined features of the plasma activated water have important implications, both to biomedical and environmental fields.