Advertisement

Journal of Solid State Electrochemistry

, Volume 22, Issue 5, pp 1349–1363 | Cite as

Degradation of paracetamol in a bubble column reactor with ozone generated in electrolyte-free water using a solid polymer electrolyte filter-press electrochemical reactor

  • Lindomar G. De Sousa
  • José Geraldo M. C. Júnior
  • Rodrigo M. Verly
  • Manoel J. M. Pires
  • Débora V. Franco
  • Leonardo M. Da Silva
Original Paper
  • 199 Downloads

Abstract

A porous anode composed of β-PbO2 was electrochemically deposited onto a carbon cloth substrate (e.g., CC/β-PbO2) aiming for the electrochemical ozone production (EOP) in electrolyte-free water using a solid polymer electrolyte (SPE) filter-press reactor. Scanning electron microscopy (SEM) images revealed the presence of a three-dimensional oxide structure necessary to obtain a fluid-permeable anode. X-ray analysis showed the predominance of the β-PbO2 phase. The maximum current efficiency for the EOP was 9.5% with an ozone production rate of 1.40 g h−1. Using a constant ozone production rate of 0.5 g h−1, the oxidative degradation of paracetamol (PCT) dissolved in water was accomplished as a function of the PCT concentration (20, 30, and 50 mg L−1) and the pH (acid, natural (without adjustment), and alkaline). The UV-Vis spectrophotometric analysis showed that the degradation process is more pronounced in alkaline media with a strong reduction in the electrical energy per order (E EO). A reduction of the chemical oxygen demand (COD) of up to 80% was observed. A linear correlation between data referring to COD and HPLC measurements with the UV absorbance measured at 243 nm (UV243) was verified indicating that these different techniques can be complementary to each other. The nuclear magnetic resonance (NMR) study of the ozonation by-products revealed that the oxidation of PCT occurred through the rupture of the aromatic ring. The major part of phenol’s ring was oxidized to CO3 2− while no reaction occurs in the acetamide group of paracetamol during the ozonation reaction.

Keywords

Porous lead dioxide layer SPE filter-press reactor Electrochemical ozone production Oxidative degradation of paracetamol Bubble column reactor NMR characterization 

Introduction

Ozone is a powerful oxidizing agent, capable of participating in a large number of reactions with organic and inorganic compounds [1]. Thus, in recent years, studies have been intensified in order to improve its production as well as its application in the mineralization of different organic pollutants. Among the pharmaceuticals (e.g., emerging pollutants) used in recent studies involving the ozonation, a special attention was given to paracetamol, a drug with analgesic, antipyretic, and mild anti-inflammatory effects. This drug was already found in aquatic environments since approximately 90% of the drug is excreted by the human body without being metabolized [2]. Therefore, the main route for the environmental contamination by several types of pharmaceuticals is the domestic sewage. In addition, several drugs found in polluted waters cannot be adequately treated by the conventional biological treatment methods [1, 2].

Therefore, several treatment technologies based on the heterogeneous photocatalysis, ozone/Fe2+/UVA, electrochemistry, ozonation, H2O2/UV, etc. were investigated to find alternative technologies for the degradation of several drugs, as is the case of paracetamol, aiming for minimization of the deleterious impact of the drugs in the environment [3, 4, 5, 6, 7, 8]. In some cases, the drug oxidation was characterized by rupture of the aromatic ring, with a partial conversion of the organic carbon content into carbon dioxide [6, 7]. In this sense, in the last years, a special attention has been given to treatment methods that are capable of generating “in situ” the hydroxyl radicals (HO·E o = 2.80 V), as is the case of the ozonation [1, 9, 10].

Ozone applications in wastewater treatment plants contribute to the decontamination process in at least two important aspects [9, 10]: (i) increase in biodegradability of the dissolved organic matter and (ii) introduction of a considerable amount of oxygen in water, thus creating excellent conditions for the biological process implemented in a separated treatment stage. In addition, the ozone degradation in water can lead to the “in situ” formation of hydroxyl radicals [1, 10].

From the above considerations, a special attention was given in the last three decades to the “electrochemical ozone production” (EOP), owing to the high concentration of the dissolved ozone obtained in this technology [1, 9, 10]. In this sense, lead dioxide is the most used anode for the EOP since this electrode material is inexpensive and highly stable at high anodic potentials [11, 12, 13, 14, 15], i.e., the contamination of water by Pb2+ ions can be disregarded when the anode is used in conjunction with a solid polymer electrolyte [9, 12]. Therefore, there is great interest in improving the properties of the lead dioxide anode, PbO2, as is the case of the fabrication of a mechanically stable three-dimensional anode for applications in SPE filter-press cells that operate using electrolyte-free water.

The lead dioxide anodes prepared by electrodeposition commonly exhibits a compact surface morphology which is stable to wear (erosion and/or corrosion) during intense gas evolution (e.g., oxygen evolution reaction, OER) [14]. Depending on the conditions employed during the electrodeposition, the lead dioxide layer can be obtained in two crystalline forms: α-PbO2 (orthorhombic structure), originated from alkaline solutions, and β-PbO2 (tetragonal structure), originated from acidic solutions [15]. The β-PbO2 phase is the preferred one in several electrochemical applications due to its higher overpotential for the OER [16] and a better electronic conductivity when compared to the α-PbO2 phase [15, 17].

When hydrated, PbO2 is classified as a metallic conductor (high electron density, 1020–1021 cm−3) [18]. This conductivity is attributed to the non-stoichiometry of the oxide, due to its oxygen vacancies [19] or the substitution of surface oxygen for hydroxyl radicals as a result of the film hydration [20], represented by PbO2−x (yH2O), where “2−x” is the deviation from the ideal stoichiometry and “yH2O” represents the amount of water present in the oxide structure.

A survey of the literature revealed that the PbO2 layer was already electrochemically deposited onto different conducting substrates, planar or porous, such as Ti, Pt, Ti-Pt, Al, Ta, Cu, Pb, Au, carbon cloth, and stainless-steel [9, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30]. However, so far only a few attempts were given to the application of the porous β-PbO2 anode in SPE filter-press cells aiming for the environmental applications of ozone involving the water and wastewater treatments [1, 25].

In the present study, the porous β-PbO2 layer was electrochemically deposited onto a carbon cloth substrate (serge type) in order to obtain a fluid-permeable anode for application in an SPE filter-press reactor. A systematic investigation of the EOP process was conducted using electrolyte-free water. The electrochemically generated ozone was applied on the degradation of aqueous solutions containing the drug paracetamol under semi-batch conditions using a bubble column reactor. The ozonated solutions were characterized using different techniques, including the nuclear magnetic resonance (NMR).

Experimental

Chemicals

Paracetamol (C8H9NO2) (Acetaminophen—CAS: 103-90-2, purity > 99%) was purchased from Sigma-Aldrich. The pH control was made by adding some drops of a sulfuric acid solution (Merck) or by the addition of some drops of a sodium hydroxide solution (Sigma-Aldrich). Organic solvents and other chemicals employed were of high performance liquid chromatography (HPLC) grade (Fluka). Ultra-pure water used for preparing the solutions was obtained using a model MS 2000 purification system from Gehaka (Brazil) with a resistivity of 18.2 MΩ cm at 25 °C.

Study of the OER/EOP processes using quasi-stationary polarization curves

Quasi-stationary potentiostatic polarization curves were recorded in electrolyte-free water at a scan rate (ν) of 1.0 mV s−1 from the equilibrium potential (OCP value) until a potential value corresponding to a maximum apparent current density of 150 mA cm−2. Experiments were carried out in triplicate using a model PGSTAT 128N Potentiostat from AUTOLAB (The Netherlands). The SPE cell used in the present work was previously reported by Costa and Da Silva [24]. Ultra-pure water (e.g., electrolyte-free water) was obtained using a model MS 2000 Purification System from Gehaka (Brazil) with a resistivity of 18.2 MΩ cm at 25 °C.

Fabrication of the β-PbO2 layer for the SPE filter-press reactor

Lead dioxide layers (β-PbO2) presenting a three-dimensional structure supported on fibers of a carbon cloth substrate were prepared by electrodeposition in a 0.1 dm3 single-compartment cell equipped with two graphite bars (counter electrodes) presenting each a geometric area of 16 cm2 [24]. The solution was magnetically stirred during electrodeposition. The carbon cloth CCS200 (serge type and ε = 0.34 mm) furnished by Maxepoxi Co. (Brazil) used as the substrate was previously cleaned using isopropanol and then rinsed with water. A fine insulating layer (ε ≈ 0.1 mm) was applied on the edges of the substrate using silicon glue (Tekbond, Brazil) in order to avoid damages for the serge-type weaving during the cutting process using a scissor. In addition, this procedure minimized the “edge effect” that leads to a preferential electrodeposition of the lead dioxide on the edges of the substrate. A mold made of Teflon (ε = 0.5 mm and 4.0 cm × 5.0 cm) was used during application of the insulating layer in order to delimitate the electrodeposition area for each side of the substrate [24]. After the careful removal of the Teflon mold, the sample was left in air for 24 h at 25 °C. This procedure was repeated for the other side of the substrate. After that, the portion of the substrate devoted to electrodeposition (4.0 cm × 5.0 cm) was totally immersed in the electrodeposition solution. β-PbO2 layers presenting a geometric area for each side of 20 cm2 was prepared by electrodeposition using a 0.2 mol dm−3 Pb(NO3)2 + 0.01 mol dm−3 HNO3 solution at 50 °C, applying a constant apparent current density of 40 mA cm−2 for 30 min. After the electrodeposition, the uncovered part of the carbon cloth used for the electric contact was removed resulting in the final configuration of the anode used in the SPE cell. Three samples were prepared in order to check the reproducibility of the experimental findings. Electrodes were rinsed thoroughly with deionized water to remove any traces of Pb2+ ions and then air-dried and stored appropriately. The average mass obtained for the β-PbO2 electrodes was 100 ± 3 mg cm−2. Electrodeposition was carried out using a power source (3 A/30 V) from ICEL (Brazil). Vetec (Brazil) “purum p.a.” products were used throughout.

Cathode preparation for the SPE filter-press reactor

A carbon cloth (CCS200: serge type; A G = 20 cm2 and ε = 0.34 mm) supplied by Maxepoxi Co. (Brazil) was used as the cathode, which was previously cleaned using isopropanol and then rinsed with deionized water.

Membrane electrode assembly and the SPE filter-press reactor

A system comprising the electrodes (anode and cathode), solid polymer electrolyte (SPE), and perforated current collectors was assembled using a specially designed cell housing made of acrylic, in which the electrodes were pressed against the SPE (Nafion® 324—Teflon fabric reinforced) membrane from Dupont (Brazil) using a clamping system [9]. Fluid manifolds (water distribution channels) were machined into the intermediate acrylic plates to facilitate the supply of water to the SPE/electrode interface (active zones for electrolysis).

To obtain the desired configuration for the SPE cell, a stainless-steel mesh (AISI-304: A G = 20 cm2, ∅ = 0.2 mm × 0.2 mm, and ε = 0.2 mm) was placed between the electrodes (anode and cathode) and the perforated current collectors, which was also made of stainless steel (AISI-304). These auxiliary stainless-steel meshes were used to propitiate a uniform distribution of pressure on the membrane electrode assembly (MEA), since the “sandwich” (collector/mesh/anode/SPE/cathode/mesh/collector) was compressed through springs fixed at the edges of the perforated current collectors [9]. A pressure of 0.5 kgf cm−2 was applied by fastening spring-loaded screws (a clamping system) that were fixed in the perforated current collectors in order to promote an adequate mechanical/electrical contact at the electrode/SPE interface. This procedure ensured adequate compression of the SPE, providing the necessary conditions for the zero-gap configuration [24]. The general scheme of the elements composing the SPE filter-press reactor used in the present work was previously described by some of the present authors [9].

Reinforced Nafion® 324 (Dupont, Brazil) was pre-treated by immersing it into boiling 50 v/v% HNO3 solution for 30 min and then in boiling deionized water for 2 h to provide adequate hydration of the membrane [24].

The volume of the anodic and cathodic reservoirs was 2.0 dm3 (see Fig. 1). The electrolysis of electrolyte-free water was carried out by recirculation through a plug-flow batch reactor at a volumetric flow rate (Q) of 23.6 cm3 s−1. The hydraulic behavior of the anodic fluid inside the manifolds containing S-shaped distribution channels (dimensions: cross-sectional area (A SE) = 0.30 cm2, length (l) = 80 cm, and channel volume (V r) = 24 cm3) pressed against the perforated current collector (zero-gap condition) was characterized by a residence time value, τr = (V r/Q), of 1.02 s and a fluid velocity, u r = (Q/A SE), of 78.7 cm s−1. In all cases, the reactor was powered using a power source (100 A/12 V) from AMZ (Brazil). All experiments were carried out using electrolyte-free water (ρ = 18.2 MΩ cm at 25 °C).
Fig. 1

Flow diagram representing the experimental setup used for the electrochemical ozone production and ozonation of the paracetamol solutions using a bubble column reactor

Electrochemical ozone production: characterization of the SPE filter-press reactor

The gas mixture (O2/O3) that leaves the anodic compartment was separated from the electrolyte-free water that flows through a gas separator flask and directed to the spectrophotometer for the UV analysis at 254 nm (see Fig. 1).

The ozone concentration in the gas phase was analyzed by measuring the UV absorbance at 254 nm, according to the methodology described by De Sousa et al. [25]. The partial current I EOP and the current efficiency Φ EOP for EOP were calculated using the following equations [25]:
$$ {I}_{\mathrm{EOP}}\left(\mathrm{A}\right)=\frac{AGzF}{\upvarepsilon l} $$
(1)
$$ {\varPhi}_{\mathrm{EOP}}\left(\mathrm{wt}\%\right)=\frac{I_{\mathrm{EOP}}}{I_{\mathrm{T}}}, $$
(2)
where A = absorbance at 254 nm; G = volumetric flow rate of the gas mixture (O2 + O3) (dm3 s−1); z = number of electrons (= 6); ε = ozone molar absorptivity at 254 nm (3024 cm−1 mol−1 dm3 [12]; l = optical path length (1 cm); I T = total current (A); and F = 96,485 C mol−1.
The specific electric energy consumption, P EOP, was calculated using Eq. (3) [25]:
$$ {P}_{\mathrm{EOP}}\left(\mathrm{Wh}\ {\mathrm{g}}^{-1}\right)=\frac{UzF}{1.73\times {10}^5{\varPhi}_{\mathrm{EOP}}}, $$
(3)
where U is the cell voltage.
The ozone production rate, ν EOP, was calculated using Eq. (4) [25]:
$$ {\nu}_{\mathrm{EOP}}\left(\mathrm{g}\ {\mathrm{h}}^{-1}\right)=\frac{3600{I}_{\mathrm{EOP}}M}{zF}, $$
(4)
where M is the molecular weight of O3 (48 g mol−1).
The figure-of-merit denoted as “mass gain of O3 per total energy consumption,” ϑ EOP, representing the overall reactor performance was calculated using the Eq. (5) [25]:
$$ {\vartheta}_{\mathrm{EOP}}\left(\mathrm{g}\ {\mathrm{W}}^{-1}\ {\mathrm{h}}^{-1}\right)=\frac{\nu_{\mathrm{EOP}}}{I_{\mathrm{T}}U} $$
(5)

Ozonation of the paracetamol solutions using the bubble column reactor

For the degradation of paracetamol, the electric current applied to the reactor and the temperature of the circulating electrolyte-free water were both adjusted to provide a constant ozone production rate of 0.5 g h−1. This value was chosen based on the geometrical properties of the bubble column reactor (e.g., liquid height, volume, and the diameter of the sintered porous glass plate) in order to maximize the ozone utilization efficiency (OUE) during the ozonation of paracetamol solutions.

The oxidative degradation of this drug during the ozonation process was studied under semi-batch conditions (V = 1.0 dm3) using a bubble column reactor (see Fig. 1). The gas mixture (O2/O3) was bubbled through a porous plate diffuser made of sintered glass (Schott #2 and A = 3.5 cm2) placed at the bottom of the bubble column reactor. In addition to the evaluation of the influence of the initial concentration of the drug (20, 30, and 50 mg dm−3), the influence of the pH (2, natural (pH 6.3), and 10) on the degradation process was also studied. The pH was adjusted using a syringe containing one of the following solutions: 0.1 mol dm−3 H2SO4 or 0.1 mol dm−3 NaOH. The pH was monitored using a glass electrode connected to a model DM-22 pH meter from Digimed (Brazil).

To evaluate the degradation process, samples (V = 3 cm3) were withdrawn every 5 min for the UV-Vis analysis until the total ozonation time of 60 min. In this particular study, all ozonated samples were returned to the bubble column reactor after the measurement of the absorbance. Additional samples (V = 1 cm3) were withdrawn for the chromatographic (HPLC) analysis.

The ozonated samples were analyzed using a model Cary 50 spectrophotometer from Varian and a model ProStat 315 HPLC from Varian, where a model LC 018 C 18 column from Supel Casil™ was used in the reverse mode. The UV detector was set at 243 nm to monitor the concentration of paracetamol.

Also, to evaluate the degradation process, the chemical oxygen demand (COD) analysis was carried out according to the methodology reported in the standard methods of analysis [31]. In this case, COD values were determined for the treated samples and recorded as a function of the ozonation time at regular intervals of 10 min. COD is a measure of the oxygen equivalent of the organic matter content of the sample that is susceptible to oxidation by a strong chemical oxidant (e.g., potassium dichromate). In this sense, silver sulfate was used as a catalyst to promote the oxidation of certain classes of organic compounds (e.g., oxidation by-products). The sample vials (V = 10 cm3) with cap were heated at 150 °C for 2 h using the COD reactor block heating from HACH (Brazil). After cooling the samples, the COD values were determined by titration of the amount of unreacted dichromate with ferrous ammonium sulfate using ferroin as the indicator. The potassium hydrogen phthalate was used as a standard. All experiments were carried out in triplicate.

The chemical structure of paracetamol is shown in Fig. 2.
Fig. 2

Chemical structure of paracetamol

Characterization of the oxidation by-products using NMR spectroscopy

The NMR spectra of paracetamol and its oxidation by-products were recorded using a model FOURRIER 300 NMR spectrometer from Bruker (Switzerland) operating at 300.18 MHz for H1 and 75.48 MHz for C13, following standard protocols [32]. In this particular study, the oxidation by-products was obtained after the ozonation carried out over a period of 240 min (conditions: v EOP = 0.5 g h−1, V = 250 mL, [paracetamol]0 = 250 mg dm−3, and pH 10).

The paracetamol and its oxidation by-products were dissolved in a mixture of H2O/D2O (90/10, v/v) containing 5 μmol dm−3 of tetramethyl-silapentane and using 10 mmol dm−3 of TMS (internal reference). In all cases, a one-dimensional (1D) NMR spectra were recorded using a 1H/13C standard 5 mm probe at 20 °C. All experiments were acquired and processed using the Topspin 3.1 software developed by Bruker BioSpin. 1H spectra were acquired with water suppression by using pre-saturation techniques with zgcppr pulse sequence. The spectral width was 6103.52 Hz for both samples, 512 numbers of scan, 2.0 number of dummy scans, and 16,384 of time domain data size. Decoupled 13C and DEPT-135 13C spectra were acquired using zgpg30 and dept135 pulse sequence, respectively. 13C spectra were acquired for both samples with spectral width of 24,414.06 Hz, 4096 numbers of scan, 4 number of dummy scans, and 65,536 of time domain data size. DEPT-135 13C spectra were acquired with spectral width of 24,414.06 Hz, 2048 numbers of scan, 4 number of dummy scans, and 65,536 of time domain data size.

Results and discussion

Ex-situ characterization of the β-PbO2 electrode

SEM analysis

Figure 3 shows the scanning electron microscopy (SEM) images obtained for the porous PbO2 layer electrochemically deposited on the carbon cloth (serge type) substrate.
Fig. 3

SEM images of the β-PbO2 layer electrochemically deposited onto the carbon cloth substrate

As can be seen, the electrodeposition process resulted in a good coverage of fibers of the carbon cloth by the oxide layer. In addition, the porous structure of the carbon substrate was preserved resulting in a porous PbO2 layer that exhibits a three-dimensional structure. A comparison with the literature [33, 34] revealed that the absence of a metal to support the oxide layer did not affect the morphological characteristics of the individual PbO2 crystals. However, the use of carbon cloth lead to the formation of isolated grains, that is, there was the formation of a three-dimensional porous structure that can allow the water flow and transport of the gaseous products (O2 and O3) through the microstructure of the MEA. As a result, the electrochemically deposited porous PbO2 layer can be used as a fluid-permeable anode (FPA) in SPE filter-press reactors [9].

XRD analysis

Figure 4 shows the X-ray diffractogram obtained for the PbO2 layer supported on the carbon cloth substrate.
Fig. 4

X-ray diffractogram obtained for the β-PbO2 layer electrochemically deposited onto the carbon cloth substrate

The XRD spectrum shows well defined narrow peaks, which indicate that the lead dioxide layer has good crystallinity. In addition, it was verified that the β-PbO2 phase is predominant over the α-PbO2 phase. These observations are in accordance with the literature [26, 35]. An average size of the crystallites of 54 nm was obtained using the well-known Debye-Scherrer equation [36].

Kinetic study of the OER and OER/EOP processes

The electrode mechanism for the OER can consists of different consecutive steps, being that each one can be the rate-determining step (RDS) [37]. The rate of each step can be influenced by the electrode composition, adsorption energy of the reaction intermediates, geometric arrangement of the atoms on the surface, etc. [18]. From the current-potential curve obtained under quasi-stationary conditions (low scan rate), the Tafel plot can be obtained permitting to identify the RDS.

At least in principle, a “complete” kinetic investigation for a particular electrode reaction also requires the determination of the kinetic parameters denoted as “stoichiometric number” and “reaction order,” the latter being relative to the stable species that participate in the reaction (e.g., H+ or OH ions in case of the OER) [38]. Despite these considerations, the RDS was determined in the present work based only on the Tafel slope since the use of an SPE immersed in electrolyte-free water does not allow the determination of the other aforementioned kinetic parameters. In this sense, since the values obtained for the Tafel slope in the present study were high (b ≥ 120 mV dec−1 (25 °C)), it is likely to consider the RDS as being the primary water discharge step (e.g., the first electron transfer) [37] (see further discussion in this section).

Following the theoretical approach proposed by Da Silva et al. [39, 40], the experimental values of “b” were used to obtain the apparent charge-transfer coefficient (α ap) for the primary water discharge step. The ideal value of the charge-transfer coefficient (α) at 25 °C is 0.50 for the hypothetical case of a perfectly flat electrode and in the absence of the specific adsorption [37, 38, 39, 40]. Therefore, an ideal value of 120 mV dec−1 is predicted for the Tafel slope (b) in the case where the primary water discharge is the RDS [37, 39].

According to Böckris [37], initially, the water molecule is oxidized at low current densities originating the adsorbed HO· radicals (e.g., primary water discharge step, b ideal = 120 mV dec−1). In the next step, the HO· radicals are oxidized originating the adsorbed O· radicals which react between them leading to the formation of O2. As proposed by Da Silva et al. [39, 40], the O2 molecule can stand temporarily adsorbed on the electrode surface thus favoring the ozone formation reaction (e.g., O· (ads) + O2(ads) → O3(ads)) and/or it can agglomerate on the electrode surface permitting the coalescence with the release of gas bubbles.

In fact, the high surface concentration of the oxygenated species (HO· (ads.), O· (ads), and O2(ads)) can propitiate the occurrence of the EOP. Different electrode mechanisms proposed to represent the OER/EOP processes, taking into account different reaction intermediates, can be found in the literature [1, 11, 16]. However, these mechanisms do not permit a direct correlation between kinetic, surface adsorbed species, and current efficiency for EOP. A comprehensive kinetic study regarding the OER/EOP processes was previously reported by Da Silva et al. [39, 40]. In these studies, an electrode mechanism was proposed in order to describe the theoretical current efficiencies for the OER and EOP processes as a function of the surface coverages of the reaction intermediates. Also, it was verified for the β-PbO2 electrode that the first electron transfer (primary water discharge step) present in the electrode mechanism referring to the OER/EOP processes is the RDS [39, 40]. The electrode mechanism proposed by Da Silva et al. [39, 40] is presented as follows:

Electrochemical steps: Kinetic control
$$ {\left({\mathrm{H}}_2\mathrm{O}\right)}_{\mathrm{ads}}\to {\left({\mathrm{H}\mathrm{O}}^{\cdotp}\right)}_{\mathrm{ads}}+{\mathrm{H}}^{+}+{\mathrm{e}}^{-}\kern12em \mathrm{RDS} $$
(6a)
$$ {\left({\mathrm{H}\mathrm{O}}^{\cdotp}\right)}_{\mathrm{ads}}\to {\left({\mathrm{O}}^{\cdotp}\right)}_{\mathrm{ads}}+{\mathrm{H}}^{+}+{\mathrm{e}}^{-} $$
(6b)
Chemical steps: Efficiency control
$$ {\left({\mathrm{O}}^{\cdotp}\right)}_{\mathrm{ads}}\to \left[1-\theta \right]{\left({\mathrm{O}}^{\cdotp}\right)}_{\mathrm{ads}}+\theta \left({\mathrm{O}}^{\cdotp}\right){\ast}_{\mathrm{ads}},\kern4em \left(0<\theta <1\right) $$
(6c)
$$ \left[1-\theta \right]{\left(2{\mathrm{O}}^{\cdotp}\right)}_{\mathrm{ads}}\to \left[1-\theta \right]{\left({\mathrm{O}}_2\right)}_{\mathrm{ads}} $$
(6d)
$$ \left[1-\theta \right]{\left({\mathrm{O}}_2\right)}_{\mathrm{ads}}\to \left[1-\beta \right]\cdot \left[1-\theta \right]{\left({\mathrm{O}}_2\right)}_{\mathrm{ads}}+\beta \left[1-\theta \right]\left({\mathrm{O}}_2\right){\ast}_{\mathrm{ads}}\kern3em \left(0<\beta <1\right) $$
(6e)
Oxygen evolution:
$$ \left[1-\beta \right]\cdot \left[1-\theta \right]{\left({\mathrm{O}}_2\right)}_{\mathrm{ads}}\to {\mathrm{O}}_2 $$
(6f)
Ozone formation:
$$ \theta \left({\mathrm{O}}^{\cdotp}\right){\ast}_{\mathrm{ads}}+\beta \left[1-\theta \right]\left({\mathrm{O}}_2\right){\ast}_{\mathrm{ads}}\to \left[\theta +\beta \left(1-\theta \right)\right]{\left({\mathrm{O}}_3\right)}_{\mathrm{ads}} $$
(6g)
$$ \left[\theta +\beta \left(1-\theta \right)\right]{\left({\mathrm{O}}_3\right)}_{\mathrm{ads}}\to {\mathrm{O}}_3 $$
(6h)
where “θ” and “β” represent the surface coverages by oxygen species while “*” represents the fractional surface coverage leading to the ozone formation (e.g., EOP).
As proposed by Da Silva et al. [40], the theoretical current efficiencies with respect to OER (Φ OER) and EOP (Φ EOP) processes are a function of the θ and β coverages as follows:
$$ {\varPhi}_{\mathrm{OER}}=\left[1-\beta \right]\cdot \left[1-\theta \right] $$
(7)
$$ {\varPhi}_{\mathrm{EOP}}=\left[\theta +\beta \left(1-\theta \right)\right] $$
(8)

Theoretical calculations [40] showed that a maximum Φ EOP value is obtained for θ and β values tending to unity. Under these conditions, the ozone formation (steps (6 g) and (6 h)) is favored over the oxygen evolution (step (6f)). As a result, under these conditions the generation of oxygen is minimum, serving only as a source of adsorbed O2-species necessary for the ozone formation on the electrode surface.

Figure 5 shows the Tafel plot obtained at 1.0 mV s−1 for the OER (low overpotential domain) and OER/EOP (high overpotential domain) on the CC/β-PbO2 electrode covering the current density interval of ~ 0.1 to 150 mA cm−2. It is worth mentioning that according to the spectrophotometric analysis carried out at 254 nm (see Fig. 1) when j ap ≥ 50 mA cm−2, the EOP occurs together with the OER.
Fig. 5

Tafel plot obtained for the OER and OER/EOP processes on the CC/β-PbO2 electrode using electrolyte-free water (ν = 1.0 mV s−1 and T = 24 °C): (a) before and (b) after the correction for the ohmic drop

As can be seen, the Tafel plot, not corrected for the ohmic drop, exhibited a straight line at low current densities and an ascending curvature at higher current densities. According to the literature [39, 40, 41, 42, 43], this deviation from the linearity is due to the sum of the uncompensated ohmic resistances, R Ω, which may be given in the present case by R Ω = R SPE + R film. However, due to the low resistivity of β-PbO2 (ρ ≅ 0.95 × 10−4 Ω cm) [15, 18] associated with the reduced film thickness (ε ~ 40 μm), one can argue that R SPE > > R film, hence R Ω ≅ R SPE. After the correction for the ohmic drop (R Ω = 5.8 Ω cm2, a value obtained using the classical impedance method where R Ω = Z real(ω → ∞)), two Tafel slopes were observed: b 1 (low overpotentials) and b 2 (high overpotentials).

According to the literature [11, 15, 39, 43, 44], there are many reasons for the existence of two slopes in the Tafel plots for the gas-evolving reactions (e.g., OER and OER/EOP). However, in the present case, it is more likely to consider that the change in the Tafel slope is associated with a change in the effective electrode surface area due to the competitive reactions (formation and accumulation of oxygen and ozone) occurring simultaneously on the electrode surface. While these species reduced the electrode surface area for the oxidation of water molecules, they also increased the surface resistance and thus the electron transfer from water to the electrode would become slower (reduced rate).

As previously discussed by Costa and Da Silva [24], the HO· radicals can remain strongly attached to the gel-layer formed on surface of the lead dioxide electrode while the O· radicals are free to move over the electrode surface. Thus, assuming that the primary water discharge step (e.g., first electron transfer) is the RDS, the current density (j) for the OER or OER/EOP processes in the high-field approximation (η > 0.1 V) is given as follows [45]:
$$ j=4F{k}_{ap}^o{\left[{H}_2O\right]}_{MEA}\exp \left[\frac{\alpha_{ap}F\upeta}{RT}\right] $$
(9)
where k o ap is the apparent rate constant, [H2O]MEA is the water concentration inside the MEA, and α ap is the apparent charge-transfer coefficient. The other symbols have their usual meaning. It is worth mentioning that Eq. (9) can be indeed applied considering the case when the hydroxyl radicals (HO·) are present on the electrode surface at a low coverage, i.e., the HO· radical is formed in the slow step and rapidly consumed in the following rapid step [45].
Since the Tafel slope is defined by b = (∂η/∂log(j))T, the expression for α ap is obtained as follows [24, 45]:
$$ {\alpha}_{ap}=1.985\times {10}^{-4}{K}^{-1}\mathrm{V}\left(\frac{T}{b}\right), $$
(10)

where b is the experimental Tafel slope. In this sense, the experimental α ap value incorporates the deviations from the ideal case commonly found for the electron transfer occurring on solid electrodes [39, 45].

The kinetic parameters (b, α ap, j o(ap), and k o ap) obtained for the OER or OER/EOP processes on the CC/β-PbO2 electrode during electrolysis of the electrolyte-free water are gathered in Table 1.
Table 1

Kinetic parameters obtained from the Tafel plot for the OER and OER/EOP processes on the CC/β-PbO2 electrode in electrolyte-free water (T = 24 °C)

η domain

b/mV dec−1

α ap

j o(ap)/A cm−2

k o ap/cm s−1

Low

411

0.144

7.6 × 10−5

3.5 × 10−9

High

287

0.206

3.4 × 10−6

1.6 × 10−10

As can be seen, the OER and OER/EOP processes were characterized by Tafel slopes of 411 and 287 mV dec−1, respectively. As a result, α ap values of 0.144 and 0.206 were obtained, respectively. The deviation from the ideal case (α ap ≠ 0.5), frequently found for solid electrodes, may be attributed to the influence of the non-uniform distribution of the electric field on the rough electrode surface and/or the adsorption of gas bubbles [39, 46]. The present kinetic findings are in agreement with those reported by Costa and Da Silva [24]. Other works [39, 40, 46] also reported high Tafel slopes for the OER on the PbO2 electrode in acidic solutions. For instance, Ho and Hwang [46] found a Tafel slope of 256 mV dec−1, an α ap of 0.30, and an apparent exchange current density of 5.48 × 10−6 A cm−2.

Electrochemical ozone production and characterization of the SPE filter-press reactor in electrolyte-free water

In principle, in the case of electrochemical ozonizers using planar electrodes immersed in liquid electrolytes, all regions of the electrode surface can be electrochemically active [47]. On the contrary, in the specific case of the filter-press SPE reactors that use electrolyte-free water, the active surface sites (reaction centers) present in the three-dimensional (porous) electrodes are restricted to the “triple-contact” regions (e.g., water/SPE/electrode) [9]. The mechanisms for the electron and proton conductions in the SPE filter-press reactors using electrolyte-free water were previously discussed in the literature [1, 9, 24, 25].

According to Da Silva et al. [39], the current efficiency for EOP is a function of the surface concentration of the active centers, which in turn depends on the surface coverage of the atomic oxygen, O· (ads), and the interaction of the latter with the oxygen molecule adsorbed on the electrode surface, O2(ads). In this sense, it can be proposed in the case of SPE reactors, using porous electrodes and electrolyte-free water, that a good current efficiency for EOP can be obtained as a result of a high surface concentration of the active centers available for the surface reaction (e.g., O· (ads) + O2(ads) → O3(ads)) which formation is favored at high current densities [39, 40].

It is worth mentioning that the use of the electrochemical technology for the “in situ” ozone generation has an important advantage over the well-established corona (silent electric discharge) process, since in the former case it is possible to accomplish the direct dissolution of a high concentration of ozone directly into the water stream, thus eliminating several drawbacks associated with the mass-transfer of ozone to the liquid phase [48].

Figure 6 shows the dependence of the cell voltage, U, the EOP current efficiency, Φ EOP, the ozone production rate, v EOP, and the overall reactor performance represented by ϑ EOP, as function of the applied current, I.
Fig. 6

The cell voltage (U), the current efficiency for EOP (Φ EOP), the ozone production rate (v EOP), and the mass gain of O3 per total energy consumption (ϑ PEO) as a function of the applied current (I). Q = 23.6 cm3 s−1. Electrolyte-free water at 24 °C

It can be noted that there is a great change (≈ 2.7 V) in the cell voltage as a function of the applied current. In the present case, U values incorporate the uncompensated ohmic components (e.g., SPE resistance and the resistance imposed by gas bubbles confined inside the porous structure of MEA), and the cathodic (hydrogen evolution reaction (HER)) and anodic (OER + EOP) overpotentials. Thus, the specific electric energy consumption for EOP (data not shown) can be affected by different factors. Nishiki et al. [49] reported the characterization of an SPE filter-press electrochemical ozonizer that also used the Nafion® N324 membrane as the SPE and a boron-doped diamond (BDD) as the anode. However, it was necessary in this case a very high cell voltage of 16 V in order to obtain a low current of 0.8 A. In principle, these findings indicate a lack of good electric contact between the electrode material (e.g., BDD) and the SPE, decreasing the surface concentration of the active centers for EOP and resulting in large ohmic losses.

The analysis of the EOP current efficiency revealed a maximum value of 9.5%, which is in good agreement with other studies using the lead dioxide electrode supported on different substrates [9, 12, 13, 35, 39, 40, 44, 47, 50]. In addition, it was verified that the ozone production rate increased almost linearly with the applied current. In principle, these findings indicate that the heat dissipation in the “triple-contact” regions is quite efficient, avoiding the ozone degradation induced by thermal effects [1, 12]. In fact, in some cases, the ozone production rate passes through a maximum at ≈ 1.5 A cm−2 [12] as a result of the ozone decomposition incurred by an inefficient heat removal in the active regions of the electrode material.

The above findings reveal the importance of using highly stable anode materials since they can permit to adjust the ozone production rate in a very good manner by means of changing the applied current. Therefore, a pre-requisite for the EOP process is the choice of a stable anode since in most of the cases the electrochemical reactor must be operated covering a wide current density range (ca. 0.2−1.5 A cm−2) in order to obtain the desired mass (or concentration) of ozone necessary for a particular application, as is the case of the water treatment for different purposes, as well as the wastewater treatment process [1, 9, 10, 12].

The dependence of ϑ EOP with the applied current revealed a maximum reactor performance at ≈ I > 30 A. Therefore, great advantages can be obtained when the reactor is operated at high current values.

The CC/β-PbO2 electrode was submitted to an endurance test under galvanostatic conditions using the electrolyte-free water (e.g., I = 23 A; U = 7.3 ± 0.2 V, T = 24 ± 2 °C, and t = 24 h) in order to verify the electrode stability for the EOP process. According to this study, the CC/β-PbO2 electrode remained highly stable with a constant ozone production rate of 0.50 ± 0.01 g h−1.

Ozonation of paracetamol under semi-batch conditions using a bubble column reactor and the spectrophotometric analysis of the treated samples

The ozonation of organic compounds are dependent on the pH, the chemical reaction (intrinsic kinetics), mass-transfer of ozone, and the ozone load introduced in the aqueous phase [10, 51, 52]. Therefore, the ozonation process is affected by the type of the gas dispersion system which controls the rate of mass-transfer from the gas to the liquid phase. Also, the driving force for the mass-transfer of ozone is given by the difference in ozone concentration in the distinct phases, i.e., inside the gas phase (bubbles) and in the liquid phase (dissolved ozone) [1, 10, 47]. In fact, as previously discussed by Da Silva et al. [1, 10], the ozonation of contaminated waters (or effluents) strongly depends on the partial pressure of ozone present in the gas phase (e.g., O2 + O3) and also on the mass-transfer of ozone. In the former case, the performance of the ozonation process is governed by the ozone production rate exhibited by the ozone generator system (e.g., corona), while in the latter case it depends on the ozone solubilization in the liquid phase which is promoted by using an efficient gas-liquid contactor system.

When the ozone bubbles are in contact with the water phase there is the formation of a “gas-liquid” interface, where gas absorption (solubilization) is followed by the irreversible oxidation reaction (e.g., ozonation process). As a result, two steps can control the overall ozonation process: (i) the ozone mass-transfer from the gas phase to the liquid phase and/or (ii) the chemical reaction occurring at the gas/liquid interface. According to the “film theory” [1, 10], considering that the gas is sparingly soluble in water, one has that no mass-transfer limitation is observed within the gas phase and, therefore, only the mass-transfer resistance in the liquid phase is considered in practical situations. As a result, it is very likely the formation of a “liquid film” of an average thickness, d L, which is established between the “liquid bulk” and the “gas-liquid” interface [10]. In addition, the mass-transfer of ozone is in the steady-state within the film in the absence of mass accumulation [1, 10].

From the above considerations, when the gas mixture exhibiting a high concentration (partial pressure) of ozone is introduced into the contaminated water, the oxidative degradation of paracetamol can be mainly controlled by the elementary kinetic processes occurring in aqueous solution that results in the rupture of the aromatic ring, decreasing the absorbance at 243 nm, and originating several oxidation by-products that can remain stable in solution (e.g., persistent oxidation by-products) [52]. This degradation process comprises only the initial step of an exhaustive advanced oxidation process that can result in a good degradation of the dissolved organic matter (e.g., moderate/high reduction of COD values). However, from a practical point of view, the use of a not exhaustive (partial) degradation process can be preferred in some cases since it requires a reduced amount of ozone, i.e., the partial degradation of the target pollutant may be effective for the removal of the toxicity and, therefore, the pre-treated effluent can be directed to a final treatment process using the conventional biological treatment [10].

Experimentally, the pseudo first order kinetic condition for the ozonation reaction can be ensured by applying a constant ozone concentration in the aqueous phase under semi-batch conditions, where the oxygen/ozone gas mixture is constantly supplied at the bottom of the bubble column reactor at a fixed volumetric flow rate [1, 10, 52]. Therefore, from the theoretical point of view, the ozonation of paracetamol can be described using the following equation [1, 8, 10, 52]:
$$ \ln \left(\frac{\left[C\right]}{{\left[C\right]}_0}\right)=-{k}_{ap}t, $$
(11)
where k ap is the apparent rate constant (in min−1) and k ap = k[O3(aq)]α[HO· (aq)]γ, where α and γ are empirical constants. [C] and [C]0 are the instantaneous and initial concentrations of paracetamol, respectively. Considering that the concentration can be represented by the absorbance (e.g., Lambert-Beer’s law), the kinetic equation can be presented as follows:
$$ \ln \left(\frac{A}{A_0}\right)=-{k}_{ap}t, $$
(12)
where the A/A o ratio represents the normalized absorbance obtained at the wavelength of maximum absorption (λ max = 243 nm).
UV-Vis spectra for the aqueous solutions containing paracetamol were recorded as a function of the ozonation time. As seen in Fig. 7 (acidic solution), the spectrum of paracetamol (t = 0) was characterized by an intense band at 243 nm which was used to monitor the degradation of the dissolved organic matter (paracetamol + oxidation by-products) as a function of the ozonation time. Different drug concentrations (20, 30, and 50 mg dm−3) and pHs (2, natural (6.3), and 10) were considered in this study. Figure 7 clearly showed that the absorbance was considerably reduced during the ozonation in acidic conditions. Similar findings were obtained for the other pH values (data not shown).
Fig. 7

Influence of the ozonation time on the absorption spectra of paracetamol in aqueous solution. Conditions: pH 2; ν EOP = 0.5 g h−1, T = 24 °C, and [paracetamol]o = 50 mg dm−3

In principle, the absorbance removal in the UV region indicates the occurrence of an electrophilic attack on the aromatic ring promoted by the oxidizing agents: O3 (acidic and neutral solutions) and HO· (alkaline solution) [7, 10]. Usually, the deprotonated species formed in this reaction react rapidly with the ozone in an electrophilic attack since they are strong nucleophiles [51, 52].

Figure 8 shows the influence of the pH on the degradation of paracetamol.
Fig. 8

Dependence of the absorbance (λ = 243 nm) on the ozonation time. Conditions: pH 2; ν EOP = 0.5 g h−1, T = 24 °C, and [paracetamol]o = 50 mg dm−3

As can be seen, the removal of the absorbance was more efficient in alkaline solution, followed by the acidic and neutral solutions, respectively. In the case of the alkaline solution, after 45 min of ozonation, there was stabilization of the degradation process. Thus, the indirect (less selective) oxidation process which is mainly associated with the electrophilic attack promoted by the hydroxyl radicals on the target substance resulted in a more effective absorbance removal.

Andreozzi et al. [6] reported similar findings for the degradation of paracetamol using the ozonation and the H2O2/UV system. The reaction mechanism proposed by these authors indicates that the electrophilic attack (O3 and/or HO·) mainly occurs on the aromatic ring with the formation of glyoxal and oxalic acid (acidic conditions) and formic acid (alkaline conditions).

As shown in Fig. 9, a good linear behavior was observed in all cases (r 2 > 0.980) thus supporting the assumption of the pseudo first order conditions. The values of k ap were gathered in Table 2.
Fig. 9

Pseudo first order kinetic profiles obtained as a function of the initial concentration of paracetamol. Conditions: pH 10, T = 24 °C, and ν EOP = 0.5 g h−1

Table 2

Electrical energy per order (E EO) consumed during ozonation and k ap values obtained as a function of the operating variables

[paracetamol]o/mg dm−3

pH

k ap/min−1

E EO/kW h m−3 order−1

20

2

0.109

38

30

2

0.068

61

50

2

0.028

148

20

nat

0.069

60

30

nat

0.073

57

50

nat

0.020

207

20

10

0.159

26

30

10

0.111

37

50

10

0.088

47

Total rated power: 108 W

As can be seen, there was a progressive reduction in k ap values as a function of the concentration of paracetamol. In principle, this behavior is expected since for a fixed amount of the oxidizing agent supplied to the gas/solution interface there is an increase in the amount of the target substances for the electrophilic attack propitiated by the O3 and/or HO· oxidizing agents. More precisely, in acidic or neutral solutions, the degradation of paracetamol is mainly governed by the direct ozonation since the concentration of the hydroxyl radicals can be quite low in this case [1, 7].

The figure-of-merit denoted as “electrical energy per order” (E EO) (expressed in W h m−3 order−1) was calculated using the following relationship [53]:
$$ {E}_{\mathrm{EO}}=\frac{38.4P}{V{k}_{\mathrm{ap}}}, $$
(13)
where P is the rated power of the electrochemical ozonizer (in W), V is the effluent volume (in m3), and k ap is the apparent rate constant (in h−1).

Table 2 shows the E EO values consumed for treating the samples (V = 1.0 dm3) as a function of the operating variables (e.g., paracetamol concentration and pH) for an ozonation time of 60 min.

The analysis of data presented in Table 2 (0.020 min−1 < k ap < 0.159 min−1) revealed that k ap values are low in comparison with the experimental value of the volumetric mass-transfer coefficient of ozone (k La = 0.41 min−1), which was determined in the present study following the procedure previously described in the literature [54]. These findings indicate that the chemical reaction comprising the oxidative degradation of paracetamol is the slow process when compared to the ozone mass-transfer in the bubble column reactor. Therefore, the ozonation process accomplished in the bubble column reactor is mainly governed by the oxidative process (e.g., direct (O3) and/or indirect (HO·)) occurring in the solution phase.

In addition, the analysis of the experimental findings obtained in alkaline media, where the formation of the hydroxyl radicals is very likely from the ozone degradation [10], revealed that the oxidative degradation reaction promoted by the indirect oxidation pathway is characterized by higher values of k ap (see Table 2). Analysis of Table 2 also revealed that the E EO values varied covering the wide range of 26 kW h m−3 order−1E EO < 207 kW h m−3 order−1. It was verified that the lowest energy consumption was obtained for the ozonation process carried out in alkaline solutions. The E EO values obtained in the present work are in good agreement with those reported in the literature. For instance, Mehrjouei et al. [55] reported E EO values covering the wide interval of 12 kW ;h m−3 order−1 < E EO < 500 kW h m−3 order−1.

As will be shown in the following section of this work, all data shown above, which were based on the spectrophotometric analysis, are in good agreement with data obtained from the HPLC technique.

Reduction of COD during the ozonation process and the HPLC analysis

COD measurements are necessary in order to evaluate the degree of degradation of the dissolved organic matter (e.g., paracetamol and its oxidation by-products) as a function of the ozonation time, since the spectrophotometric analysis alone might not be sufficient to accomplish this end. However, when the absorbance at 243 nm is strongly reduced this may be an indication that the organic matter is being degraded [56].

COD is a parameter used to evaluate the amount of oxygen necessary to convert the organic matter in mineral substances in the presence of a powerful oxidizing agent [57]. Thus, a reduction in COD values verified as a function of the ozonation time can be associated with the degree of degradation of the dissolved organic matter.

As previously discussed by Thomas et al. [58], the absorbance measured in the UV region (e.g., 254 nm) and TOC (or even COD) measurements can yield complementary information about the investigated system. In fact, the UV measurements at ≈ 254 nm are indicative of a number of total aromatic compounds present in water as a function of the ozonation time, while the reduction in COD values during the ozonation indicates an increase in the susceptibility of the dissolved organic matter to chemical oxidation with formation of mineral substances.

Figure 10 shows the reduction of COD as a function of the ozonation time obtained for different paracetamol concentrations and pHs.
Fig. 10

COD reduction as a function of the pH for different ozonation times: a 20 mg dm−3; b 30 mg dm−3, and c 50 mg dm−3. Conditions: ν EOP = 0.5 g h−1 and T = 24 °C

As can be seen, there was a strong reduction in COD values mainly for the first 30 min of ozonation. In addition, the rate of COD reduction was greatly affected by the pH, especially at higher concentrations of paracetamol. This is due to the fact that when molecular ozone enters in contact with the hydroxyl anion present in aqueous solutions its catalyzed degradation rapidly initiates leading to the formation of several radical species including the hydroxyl radical [10]. Since this radical exhibits a higher oxidizing power (E o = 2.80 V) when compared to molecular ozone (E o = 2.07 V), the degradation of the dissolved organic matter can be more pronounced in alkaline conditions.

Also, Fig. 10 revealed a greater reduction in COD values obtained for the ozonation carried out without the pH adjustment (e.g., natural, pH 6.3) when compared to the case involving the acidic solution (pH 2). These findings were not verified in the previous study of the current work where the ozonated samples were analyzed based only on the spectrophotometric analysis. According to Andreozzi et al. [6], the ozone consumption in almost neutral media (pH ~ 7) is higher than in acidic solutions since in this case there is the formation of the hydrogen peroxide, which in turn can result in the formation of the hydroxyl radicals.

Figure 11 shows the linear correlation of COD and HPLC with UV243 obtained for the ozonation of paracetamol.
Fig. 11

Correlation of COD and HPLC with UV243 as a function of the ozonation time. Conditions: ν EOP = 0.5 g h−1 and T = 24 °C

As seen, the COD values exhibited a good linear correlation with the absorbance values obtained from the spectrophotometric analysis conducted at 243 nm (e.g., UV243). Similar correlations were previously reported by Mrkva [56]. In addition, it was also possible to verify a very good linear correlation between the HPLC response (e.g., the UV detection at 243 nm) and UV243 values obtained from spectrophotometry. In principle, these findings reveal the existence of a complementarity between the different experimental techniques used to characterize the ozonated samples.

NMR characterization of ozonated samples

NMR spectroscopy was used for structural analysis of paracetamol (parental substance) and its oxidation by-products formed during the ozonation process. For this particular study, a high amount of the oxidation by-products was obtained after the ozonation of 250 mL of a concentrated paracetamol solution (250 mg dm−3 and pH 10) carried out over a period of 240 min (v EOP = 0.5 g h−1). After that, the ozonated sample was freeze-dried and solubilized in a mixture of H2O/D2O (95:5, v/v) for 1H, decoupled 13C, and 13C DEPT-135 unidimensional experiments.

Unequivocal 1H resonance assignments (see Table 3) of paracetamol can be performed based on the standard chemical shifts observed in the literature [59]. The intense singlet at 1.94 ppm can be assigned to the methyl protons and the two duplets about 6.69 and 7.03 ppm were related to the aromatic proton. In addition, a chemical shift of 9.42 ppm observed in the 1H spectra of paracetamol is in accordance with amidic proton resonance. 13C spectrum presented all characteristic peaks of the paracetamol structure. However, several differences could be noticed in the 1H and 13C spectra obtained for the oxidation by-products. Whereas the aromatic peaks were not observed, the amidic proton is in a lesser chemical shift when compared to the 1H paracetamol spectrum. Clearly, the different chemical environment of amidic hydrogen, due to the absence of local anisotropic effect from the aromatic ring, indicates a C-N cleavage between acetamide group and phenol followed by degradation of the aromatic ring. Moreover, no aromatic chemical shifts could be observed in both decoupled 13C and DEPT-135 spectra. At the same time, an intense and no hydrogenated peak was verified at 162.69 ppm. On the other hand, no significant changes were noticed to the carbonyl (173.09 ppm) and methyl (23.23 ppm) carbons of acetamide group. In general, these findings indicate that the major part of phenol’s ring was oxidized to CO3 2− [60] while no reaction occurs in the acetamide group of paracetamol during the ozonation reaction.
Table 3

1H and 13C chemical shifts of paracetamol and its oxidation by-products obtained at 20 °C in a H2O/D2O (95:5, v/v) mixture (300 MHz)

Nucleus/group

Paracetamol

Oxidation by-products

1H (ppm)

13C (ppm)

1H (ppm)

13C (ppm)

Methyl

1.94 (s)

22.41

1.91

23.23

Amidic

9.42 (s)

8.44

Aromatic

6.69 (d)

7.03 (d)

115.64

124.65

129.25

153.30

Carbonyl

172.99

173.09

CO3 2−

162.69

Conclusions

The EOP accomplished in electrolyte-free water using the three-dimensional lead dioxide anode electrochemically deposited onto the porous carbon cloth substrate (CC/β-PbO2) showed promising findings, since using a volumetric water flow rate (Q) of 23.6 cm3 s−1 (T = 24 °C), it was possible to obtain a current efficiency for EOP of up to 9.5% with an ozone production rate of up to 1.40 g h−1 (I = 50 A).

Using a fixed ozone production rate of 0.5 g h−1 supplied to the bubble column reactor, the ozonation of aqueous solutions containing paracetamol was accomplished. This study clearly revealed that the degradation of the dissolved organic matter (e.g., paracetamol and its oxidation by-products) occurs more efficiently in alkaline media due to the important electrophilic attack propitiated by the hydroxyl radicals. For the best experimental conditions, a degradation degree of 80% was obtained.

The ozonated samples were analyzed using different experimental techniques (UV243, COD, and HPLC). It was verified good linear correlations between the experimental findings obtained using these different experimental techniques as a function of the ozonation time. Therefore, a complementarity exists between the different techniques used to characterize the ozonated samples.

NMR spectroscopy was used to characterize the oxidation by-products obtained after the ozonation of a concentrated paracetamol solution (V = 250 mL; [paracetamol] = 250 mg dm−3; pH 10; t = 240 min, and v EOP = 0.5 g h−1). A thorough analysis was provided in this study to elucidate the major chemical events that can occur after the oxidative degradation of paracetamol. It was verified that the major part of phenol’s ring was oxidized to CO3 2− while no reaction occurs in the acetamide group of paracetamol during the ozonation reaction.

Notes

Funding information

L.M. Da Silva wishes to thank the “Fundação ao Amparo à Pesquisa do Estado de Minas Gerais—FAPEMIG” (Projects CEX-APQ-1181-14 and CEX-112-10), “Secretaria de Estado de Ciência, Tecnologia e Ensino Superior de Minas Gerais - SECTES/MG” (Support for the LMMA Laboratory), and “Conselho Nacional de Desenvolvimento Científico e Tecnológico—CNPq” (PQ-2 grant). This work is a collaborative research project of members of the “Rede Mineira de Química” (RQ-MG) supported by FAPEMIG (Project: CEX - RED-00010-14).

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.

References

  1. 1.
    Da Silva LM, Franco DV, Gonçalves IC, De Sousa LG (2009) In: Gertsen N, Sonderby L (ed) Water purification. Nova Science, New YorkGoogle Scholar
  2. 2.
    Correia B, Freitas R, Figueira E, Soares AMVM, Nunes B (2016) Oxidative effects of the pharmaceutical drug paracetamol on the edible clam Ruditapes philippinarum under different salinities. Comp Biochem Physiol Part C: Toxicol Pharmacol 179:116–124Google Scholar
  3. 3.
    Mameri Y, Debbache N, Benacherine MEM, Seraghni N, Sehili T (2016) Heterogeneous photodegradation of paracetamol using goethite/H2O2 and goethite/oxalic acid systems under artificial and natural light. J Photochem Photobiol A Chem 315:129–137CrossRefGoogle Scholar
  4. 4.
    Yang L, LE Y, Ray MB (2008) Degradation of paracetamol in aqueous solutions by TiO2 photocatalysis. Water Res 42:3480–3488CrossRefGoogle Scholar
  5. 5.
    Arredondo-Valdez HC, García-Jiménez G, Gutiérrez-Granados S, de León PC (2012) Degradation of paracetamol by advance oxidation processes using modified reticulated vitreous carbon electrodes with TiO2 and CuO/TiO2/Al2O3. Chemosphere 89:1195–1201CrossRefGoogle Scholar
  6. 6.
    Andreozzi R, Caprio V, Marotta R, Vogna D (2003) Paracetamol oxidation from aqueous solutions by means of ozonation and H2O2/UV system. Water Res 37:993–1004CrossRefGoogle Scholar
  7. 7.
    Hamdi El Najjar N, Touffet A, Deborde M, Journel R, Leitner KV (2014) Kinetics of paracetamol oxidation by ozone and hydroxyl radicals, formation of transformation products and toxicity. Sep Purif Technol 136:137–143CrossRefGoogle Scholar
  8. 8.
    Skoumal M, Cabot PL, Centellas F, Arias C, Rodríguez RM, Garrido JA, Brillas E (2006) Mineralization of paracetamol by ozonation catalyzed with Fe2+, Cu2+ and UVA light. Appl Catal B Environ 66:228–240CrossRefGoogle Scholar
  9. 9.
    Da Silva LM, Franco DV, De Sousa LG, Gonçalves IC (2010) Characterization of an electrochemical reactor for the ozone production in electrolyte-free water. J Appl Electrochem 40:855–864CrossRefGoogle Scholar
  10. 10.
    Da Silva LM, Jardim WF (2006) Trends and strategies of ozone application in environmental problems. Quím Nova 29:310–317CrossRefGoogle Scholar
  11. 11.
    Babak AA, Amadelli R, De Battisti A, Fateev VN (1994) Influence of anions on oxygen/ozone evolution on PbO2/spe and PbO2/Ti electrodes in neutral pH media. Electrochim Acta 39:1597–1602CrossRefGoogle Scholar
  12. 12.
    Stucki S, Baumann H, Christen HJ, Kötz R (1987) Performance of a pressurized electrochemical ozone generator. J Appl Electrochem 17:773–778CrossRefGoogle Scholar
  13. 13.
    Velichenko AB, Girenko DV, Kovalyov SV, Gnatenko AN, Amadelli R, Danilov FI (1998) Lead dioxide electrodeposition and its application: influence of fluoride and iron ions. J Electroanal Chem 454:203–208CrossRefGoogle Scholar
  14. 14.
    Li X, Pletcher D, Walsh FC (2011) Electrodeposited lead dioxide coatings. Chem Soc Rev 40:3879–3894CrossRefGoogle Scholar
  15. 15.
    Ho JCK, Tremiliosi-Filho G, Simpraga R, Conway BE (1994) Structure influence on electrocatalysis and adsorption of intermediates in the anodic O2 evolution at dimorphic α- and β-PbO2. J Electroanal Chem 366:147–162CrossRefGoogle Scholar
  16. 16.
    Sang-Do H, Kim JD, Myung KS, Rana RK, Singh KC (2006) Electro-chemical production of ozone using water electrolysis cell of solid polymer electrolyte (SPE). Indian J Chem Technol 13:156–161Google Scholar
  17. 17.
    Foller PC, Tobias CW (1982) The anodic evolution of ozone. J Electrochem Soc 129:506–515CrossRefGoogle Scholar
  18. 18.
    Trasatti S (1981) Electrodes of conductive metallic oxides. Elsevier, AmsterdamGoogle Scholar
  19. 19.
    Abaci S, Pekmez K, Yildiz A (2005) The influence of nonstoichiometry on the electrocatalytic activity of PbO2 for oxygen evolution in acidic media. Electrochem Commun 7:328–332CrossRefGoogle Scholar
  20. 20.
    Devilliers D, Dinh-Thi MT, Mahé E, Dauriac V, Lequeux N (2004) Electroanalytical investigations on electrodeposited lead dioxide. J Electroanal Chem 573:227–239CrossRefGoogle Scholar
  21. 21.
    Andrade LS, Ruotolo LAM, Rocha-Filho RC, Bocchi N, Biaggio SR, Iniesta J, García-Garcia V, Montiel V (2007) On the performance of Fe and Fe, F doped Ti–Pt/PbO2 electrodes in the electrooxidation of the blue reactive 19 dye in simulated textile wastewater. Chemosphere 66:2035–2043CrossRefGoogle Scholar
  22. 22.
    Awad MI, Saleh MM (2010) Electrochemical generation of ozone at PbO2-loaded platinum screens. J Solid State Electrochem 14:1877–1883CrossRefGoogle Scholar
  23. 23.
    Chen B, Guo Z, Huang H, Yang X, Cao Y (2009) Effect of the current density on electrodepositing alpha-lead dioxide coating on aluminum substrate. Acta Metall Sin 22:373–382CrossRefGoogle Scholar
  24. 24.
    Costa FR, Da Silva LM (2012) Fabrication and characterization of a porous gas-evolving anode constituted of lead dioxide microfibers electroformed on a carbon cloth substrate. Electrochim Acta 70:365–374CrossRefGoogle Scholar
  25. 25.
    De Sousa LG, Franco DV, Da Silva LM (2016) Electrochemical ozone production using electrolyte-free water for environmental applications. J Environ Chem Eng 4:418–427CrossRefGoogle Scholar
  26. 26.
    Mahalingam T, Velumani S, Raja M, Thanikaikarasan S, Chu JP, Wang SF, Kim YD (2007) Electrosynthesis and characterization of lead oxide thin films. Mater Charact 58:817–822CrossRefGoogle Scholar
  27. 27.
    Shen PK, Wei XL (2003) Morphologic study of electrochemically formed lead dioxide. Electrochim Acta 48:1743–1747CrossRefGoogle Scholar
  28. 28.
    Yeo IH, Kim S, Jacobson R, Johnson DC (1989) Electrocatalysis of anodic oxygen transfer reactions: comparison of structural data with electrocatalytic phenomena for bismuth-doped lead dioxide. J Electrochem Soc 136:1395–1401CrossRefGoogle Scholar
  29. 29.
    Zhou M, Dai Q, Lei L, Ma CA, Wang D (2005) Long life modified lead dioxide anode for organic wastewater treatment: electrochemical characteristics and degradation mechanism. Environ Sci Technol 39:363–370CrossRefGoogle Scholar
  30. 30.
    Amadelli R, Velichenko AB (2001) Lead dioxide electrodes for high potential anodic processes. J Serb Chem Soc 66:835–845Google Scholar
  31. 31.
    Rice EW et al. (ed), (2012) Standard methods for the examination of water and wastewater, 22 ed. American Public Health Association, New YorkGoogle Scholar
  32. 32.
    Resende JM, Moraes CM, Prates MV, Cesar A, Almeida FCL, Mundim N, Valente AP, Bemquerer MP, Piló-Veloso D, Bechinger B (2008) Solution NMR structures of the antimicrobial peptides phylloseptin-1,-2, and-3 and biological activity: the role of charges and hydrogen bonding interactions in stabilizing helix conformations. Peptides 29:1633–1644CrossRefGoogle Scholar
  33. 33.
    Amadelli R, Maldotti A, Molinari A, Danilov FI, Velichenko AB (2002) Influence of the electrode history and effects of the electrolyte composition and temperature on O2 evolution at β-PbO2 anodes in acid media. J Electroanal Chem 534:1–12CrossRefGoogle Scholar
  34. 34.
    Mohd Y, Pletcher D (2005) The influence of deposition conditions and dopant ions on the structure, activity, and stability of lead dioxide anode coatings. J Electrochem Soc 152:D97–D102CrossRefGoogle Scholar
  35. 35.
    Velichenko AB, Amadelli R, Benedetti A, Girenko DV, Kovalyov SV, Danilov FI (2002) Electrosynthesis and physicochemical properties of PbO2 films. J Electrochem Soc 149:C445–C449CrossRefGoogle Scholar
  36. 36.
    Mikrajuddin A, Khairurrijal K (2008) Derivation of Scherrer relation using an approach in basic physics course. Jurnal Nanosains & Nanoteknologi 1:28–32Google Scholar
  37. 37.
    Böckris JOM (1956) Kinetics of activation controlled consecutive electrochemical reactions: anodic evolution of oxygen. J Chem Phys 24:817–827CrossRefGoogle Scholar
  38. 38.
    Vetter KJ (1961) Elektrochemische kinetik. Springer-Verlag, BerlimCrossRefGoogle Scholar
  39. 39.
    Da Silva LM, De Faria LA, Boodts JFC (2003) Electrochemical ozone production: influence of the supporting electrolyte on kinetics and current efficiency. Electrochim Acta 48:699–709CrossRefGoogle Scholar
  40. 40.
    Da Silva LM, De Faria LA, Boodts JFC (2001) Green processes for environmental application. Electrochemical ozone production. Pure Appl Chem 73:1871–1884CrossRefGoogle Scholar
  41. 41.
    da Silva LA, Alves VA, da Silva MAP, Trasatti S, Boodts JFC (1997) Morphological, chemical, and electrochemical properties of Ti/(TiO2 + IrO2) electrodes. Can J Chem 75:1483–1493CrossRefGoogle Scholar
  42. 42.
    De Faria LA, Boodts JFC, Trasatti S (1996) Electrocatalytic properties of ternary oxide mixtures of composition Ru0.3Ti(0.7-x)CexO2: oxygen evolution from acidic solution. J Appl Electrochem 26:1195–1199CrossRefGoogle Scholar
  43. 43.
    Da Silva LM, Boodts JFC, De Faria LA (2001) Oxygen evolution at RuO2(x)+Co3O4(1-x) electrodes from acid solution. Electrochim Acta 46:1369–1375Google Scholar
  44. 44.
    Kötz ER, Stucki S (1987) Ozone and oxygen evolution on PbO2 electrodes in acid solution. J Electroanal Chem 228:407–415CrossRefGoogle Scholar
  45. 45.
    Gileadi E (2011) Physical electrochemistry: fundamentals, techniques and applications. Wiley-VCH, WeinheimGoogle Scholar
  46. 46.
    Ho CN, Hwang BJ (1994) Effect of hydrophobicity on the hydrophobic-modified polytetrafluoroethylene/PbO2 electrode towards oxygen evolution. J Electroanal Chem 377:177–190CrossRefGoogle Scholar
  47. 47.
    Da Silva LM, Franco DV, Forti JC, Jardim WF, Boodts JFC (2006) Characterisation of a laboratory electrochemical ozonation system and its application in advanced oxidation processes. J Appl Electrochem 36:523–530CrossRefGoogle Scholar
  48. 48.
    Foller PC, Kelsall GH (1993) Ozone generation via the electrolysis of fluoboric acid using glassy carbon anodes and air depolarized cathodes. J Appl Electrochem 23:996–1010CrossRefGoogle Scholar
  49. 49.
    Nishiki Y, Kitaori N, Nakamuro K (2011) Performances of small-sized generator of ozone-dissolved water using boron-doped diamond electrodes. Ozone Sci Eng 33:114–120CrossRefGoogle Scholar
  50. 50.
    Feng J, Johnson DC, Lowery SN, Carey JJ (1994) Electrocatalysis of anodic oxygen-transfer reactions: evolution of ozone. J Electrochem Soc 141:2708–2711CrossRefGoogle Scholar
  51. 51.
    Huber MM, Canonica S, Park GY, Von Gunten U (2003) Oxidation of pharmaceuticals during ozonation and advanced oxidation processes. Environ Sci Technol 37:1016–1024CrossRefGoogle Scholar
  52. 52.
    Franco DV, Jardim WF, Boodts JFC, Da Silva LM (2008) Electrochemical ozone production as an environmentally friendly technology for water treatment. Clean 36:34–44Google Scholar
  53. 53.
    Bolton JR, Bircher KG, Tumas W, Tolman CA (2001) Figures-of-merit for the technical development and application of advanced oxidation technologies for both electric- and solar-driven systems. Pure Appl Chem 73:627–637CrossRefGoogle Scholar
  54. 54.
    Santana MHP, Da Silva LM, Freitas AC, Boodts JFC, Fernandes KC, De Faria LA (2009) Application of electrochemically generated ozone to the discoloration and degradation of solutions containing the dye Reactive Orange 122. J Hazard Mater 164:10–17CrossRefGoogle Scholar
  55. 55.
    Mehrjouei M, Müller S, Möller D (2014) Catalytic and photocatalytic ozonation of tert-butyl alcohol in water by means of falling film reactor: kinetic and cost-effectiveness study. Chem Eng J 248:184–190CrossRefGoogle Scholar
  56. 56.
    Mrkva M (1983) Evaluation of correlations between absorbance at 254 nm and COD of river waters. Water Res 17:231–235CrossRefGoogle Scholar
  57. 57.
    Von Sperling M (1996) Introdução à qualidade das águas e ao tratamento de esgotos. Editora UFMG, Belo HorizonteGoogle Scholar
  58. 58.
    Thomas O, El Khorassani H, Touraud E, Bitar H (1999) TOC versus UV spectrophotometry for wastewater quality monitoring. Talanta 50:743–749CrossRefGoogle Scholar
  59. 59.
    Wishart DS, Knox C, Guo AC, Eisner R, Young N, Gautam B, Hau DD, Psychogios N, Dong E, Bouatra S et al (2009) HMDB: a knowledgebase for the human metabolome. Nucleic Acids Res 37:D603–D610CrossRefGoogle Scholar
  60. 60.
    Moret S, Dyson PJ, Laurenczy G (2013) Direct, in situ determination of pH and solute concentrations in formic acid dehydrogenation and CO2 hydrogenation in pressurised aqueous solutions using 1H and 13C NMR spectroscopy. Dalton Trans 42:4353–4356CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany 2017

Authors and Affiliations

  • Lindomar G. De Sousa
    • 1
  • José Geraldo M. C. Júnior
    • 1
  • Rodrigo M. Verly
    • 1
  • Manoel J. M. Pires
    • 2
  • Débora V. Franco
    • 2
  • Leonardo M. Da Silva
    • 1
  1. 1.Departamento de QuímicaUniversidade Federal dos Vales do Jequitinhonha e MucuriDiamantinaBrazil
  2. 2.Instituto de Ciência e Tecnologia, Faculdade de Engenharia QuímicaUniversidade Federal dos Vales do Jequitinhonha e MucuriDiamantinaBrazil

Personalised recommendations