Structure and Physicochemical Properties of Water Treated under Carbon Dioxide with Low-Temperature Low-Pressure Glow Plasma of Low Frequency

: Treatment of water saturated with CO 2 with low-temperature, low-pressure glow plasma of low-frequency (GP) produced a series of liquids. Their temperature and intensity of thermal e ﬀ ects non-linearly depended on the treatment time. However, the Raman spectra patterns of the treated water pointed to a speciﬁc structure of the water treated for 30 min. The spectra of control, non-treated water saturated with CO 2 , and such water treated for 15, 60, 90, and 120 min showed that their macrostructure was built mainly by a single donor, and single hydrogen bonded arrangements accompanied, to a certain extent, with free water molecules. The macrostructure of the water treated for 30 min consisted chieﬂy of tetrahedral and deformed tetrahedral structural units. That water contained long-living free radicals of discussed structure, stabilized in such macrostructure.


Introduction
Our recent papers have presented macrostructure and properties of water treated with low-temperature and low-pressure glow plasma of low-frequency (GP) in the air [1] as well as under pure nitrogen [2], ammonia [3], and methane [4]. The GP-treated water, generally called nanowater, exhibits several interesting functional properties. They depend on the mode of preparation. Nanowater stimulates various fermentative microorganisms, biopesticides and drastically influences the production of essential oils by herbs, control breeding plants, and animals. Such applications have been recently reviewed in our recent paper [5]. Physicochemical studies [1][2][3][4][5] have revealed that in every case, the treatment of water with GP results in declusterization of the water macrostructure, and molecules of gases dissolved in the treated water are excited. In such states, they are closed inside aqueous clathrates. In every case, the extent to which water is declusterized and the structure of resulting clathrates, as well as their excited guest molecules, depend on the treatment time and The equilibrium is determined by the three equilibrium constants (Equations (2)-(4)) (2) and by the total concentration of CO 2 (c tot ) present in solution given in Equation (5 where f i is the activity coefficient. Activities of the uncharged species CO 2 and H 2 CO 3 are commonly taken to be equal to their concentrations, and the activity of water is unity. Although the full evidence for this scheme is not yet available, it is clearly more general than the "classical" reaction scheme [8].
There are several papers that have described the hydration of CO 2 [9][10][11][12][13][14] and dehydration of H 2 CO 3 [15,16]. Interestingly, CO 2 forms clathrates with water. Usually, low temperature and elevated pressure are required for their formation [17]. Clathrate hydrates are cage-like structures of water molecules that house guest gas species. Less readily, such clathrates are formed in the presence of mineral compounds, for instance, MgSO 4 [18]. Such properties may be utilized for the storage of CO 2 as well as the evolution of methane from its aqueous clathrates [17,19]. Such direction of the application of aqueous CO 2 clathrates additionally underlined the importance of our studies presented in this paper.

Treating Water with GP
Distilled water (200 mL) in a glass flask was placed in the chamber of the reactor [20]. Through the deionized water, a stream of pure (100%) CO 2 (AIR-PRODUCTS, Warsaw, Poland) was bubbled and exposed to interaction with plasma for 5, 15, 30, 60, and 90 min. The plasma of 38 • C was generated at 5 × 10 −3 mbar, 600 V, 50 mA, and 10 kHz frequency. The produced water was stored at ambient temperature in 100 mL closed Teflon TM containers.

pH
pH was taken using a laboratory multifunction meter Cx-505 (Elmetron, Zabrze, Poland) equipped with a combined calomel-glass electrode. The resolution of the instrument was 0.01 pH. Measurements were performed at 25 • C. The estimations were run in triplicates.

Differential Scanning Calorimetry (DSC)
Differential scanning calorimeter DSC 204F1 Phoenix (Netzsch, Selb, Germany) was used. Multipoint calibration (Hg, In, Sn, Bi, Zn, CsCl) was involved. Twelve milligrams of samples were analyzed by heating them from 20 to 140 • C at the rate of 5 • C/min. Characteristic temperatures and enthalpy of transitions (T onset , T mid , T end , and ∆H) were determined with Proteus Analysis (Netzsch, Selb, Germany) software.

Refractive Index
For measurements, a lab refractometer RL from PZO Warsaw, Poland, was used. Its resolution was 0.0002. The measurements were run in triplicates.

Fourier Transformation Infrared-Attenuated Total Reflectance (FTIR-ATR) Spectra
The FTIR-ATR spectra of the film were recorded in the range of 700-4000 cm −1 at a resolution of 4 cm −1 using a Mattson 3000 FT-IR (Madison, WI, USA) spectrophotometer. That instrument was equipped with a 30SPEC 30 • reflectance adapter fitted with the MIRacle ATR accessory from PIKE Technologies Inc., Madison, WI, USA.

Electron Spin Resonance (ESR) Spectra
The spectra were recorded at room temperature, employing an instrument constructed at Wroclaw Technical University. The range of the X-band (ν = 9.5 GHz, λ = 3.158 cm) was employed at the 20-25 dB attenuation and 2048 s swiping time. An ESR controller software designed for taking the ESR spectra was applied and recorded in the form of the first derivative.

Ultraviolet/Visible (UV-VIS) Spectra
The spectra were recorded with a Thermo Scientific Evolution Spectrometer 220 (Thermo Fisher Scientific, Waltham, MA, USA) spectrophotometer in the wavelength range of 190-1100 nm in a quartz gas-tight cell of 10 mm path length.

Raman Spectra
The spectra were taken with a Perkin-Elmer MPF44A Fluorescence Spectrophotometer (Waltham, MA, USA) equipped with a xenon lamp and 4 mL quartz cell. The sample was excited at 330 ± 1 nm at 20 • C, and both slits were set on 2 nm. The spectra were recorded in wavelength scale and then recalculated to wavenumber scale shifted to value excitation energy.

Results and Discussion
Treatment of water saturated with CO 2 with GP produced a series of liquids. Their onset temperature and intensity of thermal effects depended on the treatment time, as shown by Differential Scanning Calorimetry (DSC) ( Table 1). These parameters varied non-linearly against the treatment time. Likely, the number of nucleation centers decreased against temperature increase. Saturation of water with CO 2 resulted in a decrease in the onset temperature solely by 0.2 • C and increase in peak maximum by 0.7 • C. The treatment of water with GP for initial 5 min increased temperature of the onset temperature and the peak maximum by further 2.1 and 1.3 • C, respectively. GP treatment prolonged to 30 min increased relevant temperatures by 5.7 and 5.0 • C, respectively. Further GP treatment increased the values of those parameters practically linearly with respect to one another. However, the values of those parameters elevated nonlinearly against the treatment time in the orders: Water treated for 15 and 30 min demonstrated the highest onset and peak temperatures. These changes were associated with changes in enthalpy, entropy, and the value of the accompanying thermal effect measured for particular samples. They changed in the following orders (see Table 1 for notation): ∆H: WCO 2 120 < WCO 2 5 < WCO 2 90 < WCO 2 0 <W < WCO 2 60 < WCO 2 15 < WCO 2 30 ∆S: WCO 2 30 < WCO 2 60 < WCO 2 15 < WCO 2 90 < W < WCO 2 0 < WCO 2 120 < WCO 2 5 MHF: WCO 2 30 < WCO 2 15 < W < WCO 2 90 < WCO 2 5 < WCO 2 60 < WCO 2 0 < WCO 2 120 PA: WCO 2 0 < WCO 2 30 < WCO 2 15 < WCO 2 5 < W = WCO 2 120 < WCO 2 90 < WCO 2 60 These orders were linear neither with respect to one another nor against the treatment time. Again, samples of water treated for 30 and 15 min closed the H order and opened the S and MHF orders.
Initially, the homogenous macrostructure of water reflected by a high value of PA was perturbed by the saturation of water with CO 2 . Several niches were formed to store the CO 2 molecules.
The introduced disorder was reflected by the lowest PA value. The GP treatment evoked a continuous transformation from one macrostructure into another, and the condition of the thermodynamic stability was not the ruling factor. It is known [21][22][23][24][25][26] that the CO 2 molecule is linear in the free state but bend in the first electronic excited state in order to decrease the total energy of the molecule. This phenomenon could contribute to these irregularities as the given way of involvement of CO 2 in constructing macrostructure could selectively favor excitation of the molecule. The GP treatment led to the homogenous distribution of niches holding the CO 2 molecules as indicated by increasing PA values. Finally, very compact, polymeric macrostructure was formed as indicated by a gradually increasing energy of evaporation.
Saturation of deionized water with CO 2 resulted in a slight decrease in pH of the solution, which was accompanied by a considerable increase in its conductivity. The original pH of non-treated water saturated with CO 2 was practically insensitive to the GP treatment (Table 2). After an initial Water 2020, 12, 1920 5 of 12 increase in conductivity, the treatment with GP provided practically linear subtle-linear against the treatment time-increase in that parameter (Table 2). These results suggested that prolonged GP treatment provided a gradual surrounding of the CO 2 molecules with water, which inhibited their hydration and dissociation. An increase in conductivity might suggest that some ions, for instance, either hydronium ions or the hydroxyl groups, were not engaged in specific interactions, decreasing their mobility. Besides, some residual ions present in the deionized water could be involved.
In the region of 200-350 nm, the UV-VIS spectra (Figure 1) of the deionized water, non-saturated with CO 2 , and prior to the treatment with GP revealed solely a low shoulder (A = 0.05 a.u at 200 nm) of a peak located below 200 nm. After the GP treatment, regardless of the treatment time (0, 5, 15, 30, 60, 90, 120 min), the samples saturated with CO 2 developed absorption band at 200 nm. The intensity of that band, originating from the n → π* transitions in the solvated CO 2 molecule, was independent of the treatment time reaching A = 1.35 ± 0.05 a.u. at 200 nm. Thus, the concentration of CO 2 in particular samples was independent of the GP treatment time. Besides, the measured refractive index (n 20 D = 1.3330) was independent of the GP treatment time, indicating no changes in electron concentration (stability of chemical concentration, the stability of density) of the liquid phase.
Water 2020, 12, x FOR PEER REVIEW 5 of 12 These results suggested that prolonged GP treatment provided a gradual surrounding of the CO2 molecules with water, which inhibited their hydration and dissociation. An increase in conductivity might suggest that some ions, for instance, either hydronium ions or the hydroxyl groups, were not engaged in specific interactions, decreasing their mobility. Besides, some residual ions present in the deionized water could be involved.
In the region of 200-350 nm, the UV-VIS spectra (Figure 1) of the deionized water, non-saturated with CO2, and prior to the treatment with GP revealed solely a low shoulder (A = 0.05 a.u at 200 nm) of a peak located below 200 nm. After the GP treatment, regardless of the treatment time (0, 5, 15, 30, 60, 90, 120 min), the samples saturated with CO2 developed absorption band at 200 nm. The intensity of that band, originating from the n  π* transitions in the solvated CO2 molecule, was independent of the treatment time reaching A = 1.35 ± 0.05 a.u. at 200 nm. Thus, the concentration of CO2 in particular samples was independent of the GP treatment time. Besides, the measured refractive index (n 20 D = 1.3330) was independent of the GP treatment time, indicating no changes in electron concentration (stability of chemical concentration, the stability of density) of the liquid phase. The FTIR-ATR reflectance spectrum of deionized water free of CO2 and saturated with CO2 is presented in Figure 2. The absorbance of the bands at around 1600 (νs C=O), 1300 (νs C-O), and 1100 (δOH) cm −1 in the spectra of particular GP-treated samples showed that the carbonic acid dissociation depended on the GP treatment time in a non-linear manner. The FTIR-ATR reflectance spectrum of deionized water free of CO 2 and saturated with CO 2 is presented in Figure 2. The absorbance of the bands at around 1600 (ν s C=O ), 1300 (ν s C-O ), and 1100 (δ OH ) cm −1 in the spectra of particular GP-treated samples showed that the carbonic acid dissociation depended on the GP treatment time in a non-linear manner.  Table 3 for assigning band intensities to the given sample. Table 3. The absorbance of the OH stretching (bands A and B), νOH, bending (band C), δOH, modes in the FTIR spectra of water saturated with CO2 taken in the region of 4000-1500 cm −1 . The strong peak in the region of 3000-3550 cm −1 represented νOH band composed, in fact, of several overlapped bands belonging to symmetric and asymmetric OH vibrations of water and carbonic acid. However, about 1% concentration of carbonic acid was not essential for attempts of separation of that band into components belonging to asymmetric and symmetric stretching vibrations belonging to water. For that purpose, following Chaplin [27], the Gaussian separation of that band into two bands was performed, representing asymmetric (band A) and symmetric (band B) components ( Figure 3).  The FTIR reflectance spectra of deionized water and such water saturated with CO 2 and treated with GP for 0-120 min: (a) the fundamental spectrum, (b-d) particular sectors of the spectra separately enlarged. Consult Table 3 for assigning band intensities to the given sample. Table 3. The absorbance of the OH stretching (bands A and B), ν OH , bending (band C), δ OH , modes in the FTIR spectra of water saturated with CO 2 taken in the region of 1500-4000 cm −1 . The strong peak in the region of 3000-3550 cm −1 represented ν OH band composed, in fact, of several overlapped bands belonging to symmetric and asymmetric OH vibrations of water and carbonic acid. However, about 1% concentration of carbonic acid was not essential for attempts of separation of that band into components belonging to asymmetric and symmetric stretching vibrations belonging to water. For that purpose, following Chaplin [27], the Gaussian separation of that band into two bands was performed, representing asymmetric (band A) and symmetric (band B) components ( Figure 3).

Water Plasming Time (min) A(3480) B(3260) C(1635) A/B A/C B/C
Water 2020, 12, x FOR PEER REVIEW 7 of 12 (d) Figure 2. The FTIR reflectance spectra of deionized water and such water saturated with CO2 and treated with GP for 0-120 min: (a) the fundamental spectrum, (b-d) particular sectors of the spectra separately enlarged. Consult Table 3 for assigning band intensities to the given sample.  The strong peak in the region of 3000-3550 cm −1 represented νOH band composed, in fact, of several overlapped bands belonging to symmetric and asymmetric OH vibrations of water and carbonic acid. However, about 1% concentration of carbonic acid was not essential for attempts of separation of that band into components belonging to asymmetric and symmetric stretching vibrations belonging to water. For that purpose, following Chaplin [27], the Gaussian separation of that band into two bands was performed, representing asymmetric (band A) and symmetric (band B) components (Figure 3).  The A-and B-bands represented water molecules vibrating asymmetrically and symmetrically, respectively. The C-band corresponded to the bending modes of the OH-group (δ OH ). In the spectra of water treated with GP in the air, after the 5 min treatment, the number of asymmetrically vibrating molecules was the highest. It added weak hydrophobic properties to that water [1]. In contrast to that, the GP treatment of water saturated with CO 2 , regardless of the treatment time, always reduced the number of asymmetrically vibrating water molecules. Again, the extent of that reduction non-linearly depended on the treatment time ( Table 3). The treatment time had no distinct effect on the pattern of the spectrum, but it influenced the intensity ration of the corresponding bands (A/B, A/C, and B/C). It spoke in favor of the influence of the treatment time on the water macrostructure and the equilibria (2)-(4). Additionally, the effect of the GP treatment upon the structure of CO 2 should be taken into account. The peak around 2900 cm −1 represented an artifact resulting from the imprecise subtraction of the background in our single-beam spectrophotometer. A low band at around 2343 cm −1 reflected vibrations in dissolved CO 2 . The intensity of that band was identical for all samples under study.

Water Plasming Time (min) A(3480) B(3260) C(1635) A/B A/C B/C
In our recent papers [1][2][3], the formation of clathrated singlet oxygen [1], triplet nitrogen [2], and ammonia [3] in water was postulated as the result of the treatment of water in contact with the air, under nitrogen, and under ammonia, respectively. Such a conclusion was drawn based on the pattern of the Raman spectra. The relevant spectra always contained two essential intensive bands in the region 300-450 nm (the wavelengths represented the Raman scattering on excitation at 330 nm).
The Raman spectrum (Figure 4) of original water (1st control) demonstrated a strong peak centered at 3260 cm −1 with a shoulder on its long wavelength 3480 cm −1 . It was fully adequate with the data presented in Figure 3. The main peak could be assigned [28] to single donor single hydrogen-bonded water and free water, whereas the more energetic shoulder corresponded to deformed tetrahedrally arranged structures. The spectrum of the 2nd control water being the original tap water saturated with CO 2 was significantly less intensive. This effect could partly result from a lower number of water molecules compared to the number of water molecules in 1st control in a given volume of the investigated sample. However, the pattern of that spectrum spoke for a substantial reconstruction of the macrostructure in the sample. The low maximum of broadband was located in the region, corresponding to the spectrum of 1st control to the shoulder, representing tetrahedrally arranged structures. In this spectrum, the band representing symmetrically vibrating single donor single hydrogen-bonded water and free water was represented by a low, flat shoulder. It could also be observed as a shoulder in the spectrum of control water prior to its saturation with CO 2 .
Only the shape of the spectrum of the water saturated with CO 2 treated for 30 min did not exhibit the low energetic shoulder. The other spectra of the water saturated with CO 2 treated for 5-120 min resembled the spectrum of 2nd control. However, the intensity of those spectra varied non-linearly against the treatment time. This irregularity fitted the regularities observed in UV and FTIR spectra.
The ESR spectra of the samples revealed a very specific character of the sample treated for 15 and particularly for 30 min. Only these samples developed a signal in the ESR spectrum ( Figure 5). The spectrum pointed to a free radical with unpaired spin localized either on the carbon or oxygen atoms.
For an extremely short lifetime, any free radical developed from the single CO 2 molecule could not be observed with the ESR technique. That instability resulted from free rotation around the σ-σ bond. On such rotation by 180 • , antibonding orbital turned that orbital into bonding, which would be equivalent to ceasing unpaired electron, i.e., ceasing free radical. However, trapping short-living radicals by another CO 2 molecule would lead to a stable structure, as presented in Figure 5.
Alternative consideration took into account a possibility of the formation of niches, hosting two CO 2 molecules on the GP treatment for 15 and 30 min. Under such circumstances, these two guest molecules could be excited, developing diradical with unpaired spins localized on the oxygen atoms. Such diradical could be stabilized by reaction with one water molecule, providing a unit with its hydrophilic head and hydrophobic tail with two O-diradicals and one likely singlet carbene. That unit could be stabilized by its incorporation into a wall of the niche. Prolonged GP treatment ruined these Water 2020, 12, 1920 9 of 12 structures, separated CO 2 molecules, and in consequence, observed radicals ceased, but it resulted in the development of a more uniform macrostructure ( Figure 6). tetrahedrally arranged structures. The spectrum of the 2nd control water being the original tap water saturated with CO2 was significantly less intensive. This effect could partly result from a lower number of water molecules compared to the number of water molecules in 1st control in a given volume of the investigated sample. However, the pattern of that spectrum spoke for a substantial reconstruction of the macrostructure in the sample. The low maximum of broadband was located in the region, corresponding to the spectrum of 1st control to the shoulder, representing tetrahedrally arranged structures. In this spectrum, the band representing symmetrically vibrating single donor single hydrogen-bonded water and free water was represented by a low, flat shoulder. It could also be observed as a shoulder in the spectrum of control water prior to its saturation with CO2. Figure 4. The Raman spectrum (excitation 330 nm) of the water treated with GP for 0 to 120 min. Notation: 1-pure water (primary control); 2-water saturated with CO 2 , non-treated (secondary control); 3-7-water treated for 15, 30 60, 90, and 120 min, respectively.  Only the shape of the spectrum of the water saturated with CO2 treated for 30 min did not exhibit the low energetic shoulder. The other spectra of the water saturated with CO2 treated for 5-120 min resembled the spectrum of 2nd control. However, the intensity of those spectra varied non-linearly against the treatment time. This irregularity fitted the regularities observed in UV and FTIR spectra.
The ESR spectra of the samples revealed a very specific character of the sample treated for 15 and particularly for 30 min. Only these samples developed a signal in the ESR spectrum ( Figure 5). The spectrum pointed to a free radical with unpaired spin localized either on the carbon or oxygen atoms. For an extremely short lifetime, any free radical developed from the single CO2 molecule could not be observed with the ESR technique. That instability resulted from free rotation around the σ-σ bond. On such rotation by 180°, antibonding orbital turned that orbital into bonding, which would be equivalent to ceasing unpaired electron, i.e., ceasing free radical. However, trapping short-living radicals by another CO2 molecule would lead to a stable structure, as presented in Figure 5.
Alternative consideration took into account a possibility of the formation of niches, hosting two CO2 molecules on the GP treatment for 15 and 30 min. Under such circumstances, these two guest molecules could be excited, developing diradical with unpaired spins localized on the oxygen atoms.

Conclusions
Initially, the homogenous macrostructure of water was perturbed by the saturated water with CO2. Several niches (clusters) in the macrostructure were formed to store the CO2 molecules. Transformations of those macrostructures took place on prolonged GP treatment. They involved the varying role of the tetrahedral, deformed tetrahedral, symmetrically vibrating single donor single hydrogen-bonded water, and free water in particular developed macrostructures. Some niches could host two CO2 molecules. Under such circumstances, the excitation could produce triplet CO2, efficiently trapped by a non-excited CO2 molecule. The oxygen diradical formed in such a manner could be stabilized, interacting with the surface of the cluster. Thus, non-treated water saturated with CO2 and such water treated with GP for 15 and 30 min formed a specific macrostructure with incorporated free radicals. It consisted chiefly of tetrahedral and deformed tetrahedral structural units, including O-free radicals formed of carbon dioxide stabilized in the macrostructure by bonding to the walls of niches. Prolonged GP treatment ruined these structures. Particular CO2 molecules hosted in isolated niches and dimeric radical decomposed into two singlet CO2 molecules. As a consequence, observed radicals ceased, but simultaneously, the macrostructure turned more homogenous.