Electrochemical removal of naphthalene from contaminated waters using carbon electrodes, and viability for environmental deployment

Journal Pre-proof Electrochemical removal of naphthalene from contaminated waters using carbon electrodes, and viability for environmental deployment Rebecca V. McQuillan, Geoffrey W. Stevens, Kathryn A. Mumford

PII:

S0304-3894(19)31198-7

DOI:

https://doi.org/10.1016/j.jhazmat.2019.121244

Reference:

HAZMAT 121244

To appear in:

Journal of Hazardous Materials

Received Date:

15 July 2019

Revised Date:

12 September 2019

Accepted Date:

16 September 2019

Please cite this article as: McQuillan RV, Stevens GW, Mumford KA, Electrochemical removal of naphthalene from contaminated waters using carbon electrodes, and viability for environmental deployment, Journal of Hazardous Materials (2019), doi: https://doi.org/10.1016/j.jhazmat.2019.121244

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier.

Title: Electrochemical removal of naphthalene from contaminated waters using carbon electrodes, and viability for environmental deployment

Author Names and Affiliations: Rebecca V. McQuillan, Geoffrey W. Stevens, Kathryn A. Mumford*

Department of Chemical Engineering, The University of Melbourne, Parkville VIC 3010, Australia

ro of

*Corresponding Author. Building 165, The University of Melbourne, Parkville VIC 3010, Australia Tel.: +61 3 8344 0048, Email: [email protected]

lP

re

-p

Graphical Abstract:

ur

Bulk electrochemical oxidation is seemingly beneficial for petroleum spill clean-up Active chlorine is effective for the removal of naphthalene and its derivatives Carbon electrodes are inexpensive and competent for bulk removal of organics A dynamic kinetic model is proposed that can accurately predict treatment outcomes Energy consumption is low enough for use in remote environmental remediation

Jo

    

na

Highlights:

Abstract

This work assesses the potential of electrochemical technologies for the treatment of groundwaters contaminated with petroleum hydrocarbons. Specific consideration was given to deployment in Antarctic regions where numerous fuel spills have occurred over the last two centuries, and resources and manual labour for remediation efforts are limited. The polycyclic aromatic hydrocarbon, naphthalene, was a used as a model contaminant and was treated with low-cost, active carbon electrodes to promote the active chlorine degradation pathway. Results showed that 20 mg/L naphthalene solutions could be treated to sufficient standards in less than 3 hours of treatment, and that the formation of toxic and chlorinated by-products is not an issue of concern if

the appropriate timeframes are used (4 hours of treatment). The effects of the applied current (0 – 160 mA) and electrolyte concentration (0.01 – 0.1 M NaCl) were evaluated and a dynamic kinetic model proposed and found to be in good agreement with the experimental results. The energy consumption is an important limitation in remote environmental regions where resources are scarce. It was found that an energy usage of 104 kWh/kg of naphthalene removed could be achieved. Keywords: Wastewater treatment, active chlorine, active electrode, environmental remediation

-p

ro of

1 Introduction Approximately 25,000 tonnes of crude oil are spilled into the environment each year [1], and of the seven continents of the world none have managed to escape this statistic. This includes Antarctica where the constant transport of oils for human use has led to numerous petroleum hydrocarbon spills over the past two centuries. It is currently estimated that there is between 1 – 10 million m3 of contaminated soils on the Antarctic continent [2] that seasonally leach toxic fuels and metals through rock and groundwater and affect sensitive environmental receptors. Unfortunately, the environmental remediation methods that are often used in temperate climates are not fit for purpose in the Antarctic. The large demand for manual labour and resources associated with ex situ dig-and-haul and incineration methods are not practical for use in such remote areas, whilst natural attenuation and enhanced biological methods are slow and difficult to control in the extreme environments faced. Scientific and engineering intervention is now required to develop feasible in situ and low maintenance methods that are applicable for the treatment of contaminant sites in Polar Regions.

na

lP

re

In the last decade, the treatment of contaminated waters in temperate regions has undergone a shift towards using electrochemical technologies to remove impurities. It’s use in environmental applications is favoured as it can be conducted in situ and is considered to be a ‘green’ technique as the only reagent is the electron. No chemical additives or extensive manual labour are required. Other advantages include its versatility, high energy efficiency, ability to be automated, and its costeffectiveness [3]. Regarding its versatility, various electrochemical reactions can be promoted that target the removal of specific and high priority pollutant compounds, including oxidation and reduction reactions, electrocoagulation, and electrodialysis processes.

Jo

ur

For application in coastal regions, where contaminated groundwaters present with a high salinity, it is possible to promote the ‘active chlorine’ electrochemical pathway, known to aid in the oxidation and breakdown of several aqueous organic components [4–9]. This pathway takes place via three sequential reactions as described by Equations (1) - (3): Chloride ions (Cl-) in the bulk solution first migrate to the anode surface and undergo electrochemical transformation to form solubilized chlorine (Cl2(aq)). Second, the generated chlorine re-enters the bulk solution where it is rapidly transformed into either hypochlorous acid (HOCl) or hypochlorite ions (OCl-) depending on the pH of the aqueous phase. Cl2(aq) dominates at a pH of 2 and lower, HOCl between a pH of 3 and 8, and OClin more alkaline conditions. 2𝐶𝑙 − → 𝐶𝑙2(𝑎𝑞) + 2𝑒 −

(1)

𝐶𝑙2(𝑎𝑞) + 𝐻2 𝑂 → 𝐻𝑂𝐶𝑙 + 𝐻 + + 𝐶𝑙 −

(2)

𝐻𝑂𝐶𝑙 ↔ 𝐻 + + 𝑂𝐶𝑙 −

(3)

𝐴𝑐𝑡𝑖𝑣𝑒 𝐶ℎ𝑙𝑜𝑟𝑖𝑛𝑒 + 𝑂𝑟𝑔𝑎𝑛𝑖𝑐𝑠 → 𝐼𝑛𝑡𝑒𝑟𝑚𝑒𝑑𝑖𝑎𝑡𝑒𝑠 → 𝐶𝑙 − + 𝐶𝑂2 + 𝐻2 𝑂

(4)

The resultant oxidizing mixture of Cl2, HOCl, and OCl- is generally referred to as ‘active chlorine’, where all three species are capable of oxidizing and breaking down organic species, shown by Equation (4). HOCl, however, is deemed to be the most oxidizing and useful product in wastewater treatment processes, making it beneficial to operate at acidic and neutral conditions [3,7]. This reaction pathway has shown to be effective for the treatment of wastewaters contaminated with several organic and inorganic compounds such as textile effluents [10–12], nitrates and ammonium [13,14], pharmaceuticals [15] and phenolics [16,17], both independently and simultaneously to electrochemical treatments involving hydroxyl radicals or direct anodic oxidation methods [18].

-p

ro of

Utilizing electrochemical technology to produce these reagents is advantageous over introducing them manually as it avoids the need for handling strong oxidizers and the active chlorine can be continuously produced at the anode surface until sufficient contaminant removal is achieved. Despite the benefits of this treatment, its leading limitation is the cost required to scale-up the treatment process, mainly attributed to the anodic materials used. Boron-doped diamond (BDD), platinum (Pt), and mixed-metal oxides (e.g. RuO2, PbO2, IrO2) are the most commonly used electrode materials due to their high efficiency, noncorrosive nature, and selectivity, all of which are met at an increased cost [3]. Pilot-scale treatment volumes of 4 L of phenol wastewater [19], 5 L of petrochemical waste [20], 200 L of rubber manufacture wastewater [21], and 230 L of landfill leachate [22] have had promising results, yet employ BDD anodes limiting further scale-up for field deployment applications. It is thus desirable to reduce costs such that further scaling becomes more tangible.

Jo

ur

na

lP

re

Therefore in this study, the use of less efficient yet readily available and inexpensive carbon electrodes was investigated for the bulk oxidation of a petroleum hydrocarbon contaminated water source. This would ultimately lower capital costs and make field deployment easier. The active chlorine pathway was chosen for the electrochemical pathway, and focus was placed on treating compounds commonly seen in Antarctic contaminated sites. In this region, diesels are generally modified via removing the heavy and waxy end components such that it has improved performance in the cold climates faced [23]. This vastly changes the composition of the fuel in comparison to those commonly used in temperature regions, with benzene and toluene components having limited presence [24]. In agreement with temperate regions, however, aromatic compounds are still believed to have the greatest toxicity and are the least susceptible to biodegradation, thus, one of the more prevalent aromatic compounds, naphthalene, was chosen as a model contaminant. Naphthalene is also an interesting compound to investigate, as apart from Muff et al. [25,26] who used Pt electrode materials, the electrochemical removal of naphthalene has not been explored in the literature. Coinciding with the use of low-cost carbon electrodes, a kinetic model is proposed that can accurately predict treatment outcomes when experimental conditions such as applied electric current, hydrodynamic conditions, and saline concentrations are altered. This is to ultimately make scale-up more conceivable. 2 Experimental 2.1 Reagents The analytical reagents sodium chloride (NaCl), sodium sulphate (Na2SO4), and potassium ferricyanide (K3Fe(CN)6) were purchased from Chem-Supply and used as received. Potassium ferrocyanide (K4Fe(CN)6) and solid naphthalene (98 %) were purchased from Sigma-Aldrich. Reverse osmosis (RO) water was used for the preparation of all solutions, except for naphthalene in which a concentrated stock solution was prepared in ethanol (EtOH) due to its low solubility in water.

Adequate dilution of the EtOH stock solution was performed in RO water, leaving all final synthetic naphthalene solutions with an alcohol concentration of less than 0.2 % v/v.

lP

re

-p

ro of

2.2 Electrochemical Reactor Electrochemical experiments were carried out in an undivided glass cell, Figure 1, having a capacity of 250 ml. Commercially available cylindrical carbon rods (8 mm Ø, Jaycar) were used as single anode and cathode, spaced 20 mm apart and vertically parallel to one another. A MP3087 (PowerTech) DC regulated power supply (0 – 32 V, 0 – 3 A) was connected to the electrodes and operated under galvanostatic conditions at room temperature (23 °C); during experiments, temperature was found to fluctuate ± 1 °C. The solution was continually agitated with a magnetic stirrer to ensure ample mixing and guarantee that sampling events were unaffected by concentration gradients within solution. For each experimental run, 200 ml of a synthetic 20 mg/L naphthalene solution within a NaCl electrolyte at various concentrations (0.01, 0.025, 0.05 and 0.1 M) was introduced into the cell, giving each electrode a submerged geometrical surface area of 10.3 cm2. The reactor was covered with a lid to minimize the effects of naphthalene volatilization, with a small opening to allow for sampling and equilibration with atmospheric pressure. Upon application of an electric current (40, 70, 100 or 160 mA), 150 μl aqueous samples were taken at various time points and analysed for naphthalene concentration against time. The resulting by-products formed, pH, ORP, and cell voltage were also examined. All experimental runs were conducted in duplicate to test for repeatability.

na

Figure 1: Schematic of the electrochemical cell used.

Jo

ur

2.3 Mass Transfer Characterization To characterize the mass transfer behaviour of the hydrodynamic conditions within the reactor, the well-established limiting-current technique of the potassium ferro/ferricyanide couple was utilized [27,28]. This type of study gives an estimate of the diffusion rate of electroactive species migrating from the bulk solution to the electrode surface where it undergoes electrochemical transformation. A 1 mM solution of K4Fe(CN)6 was oxidized in the presence of excess 0.1 M K3Fe(CN)6 and 0.5 M Na2SO4 to ensure that the cathodic reduction never became the limiting reaction. The same electrochemical reactor detailed in Figure 1 and 200 ml of solution were utilized, and a potential sweep ranging from 0 – 5 V was applied. For each applied voltage, the electrical current was given ample time to stabilize and the steady-state readings were recorded. The resulting current-potential (I-E) curve allowed for the determination of the limiting current, 𝑖𝑙𝑖𝑚 (A), and in turn, the mass transfer coefficient, 𝑘𝑚 (m/s) of the electrochemical cell as detailed by Equation (5) [29]. 𝑖𝑙𝑖𝑚 = 𝑧𝐹𝐴𝑘𝑚 𝐶 𝑏 Where 𝑧 is the number of electrons involved in the anodic oxidation reaction, 𝐹 is Faraday’s

(5)

constant (96,485 C/mol), 𝐴 is the area of the electrode (m2), and 𝐶 𝑏 is the bulk concentration of the reactant (mol/m3). 2.4 Analytical Methods 2.4.1 Naphthalene Concentration Naphthalene concentration was monitored using High Performance Liquid Chromatography (HPLC). Analysis was performed using a ZORBAX Eclipse XBD-C18, 3.5 μm, 3.0 x 150 mm column under the following chromatographic conditions: an acetonitrile and water mixture with a ratio of 60:40 v/v and a mobile phase flowrate of 1.0 ml/min, column temperature of 20 °C, sample injection volume of 20 μl, isocratic elution mode and a run time of 6 minutes. UV-Visible detection was performed with a diode array detector (DAD) at a wavelength of 275 nm, with naphthalene eluting at 3.1 minutes.

ro of

The HPLC was calibrated by preparing naphthalene standard solutions with a concentration of 30 mg/L in acetonitrile. Sample injection volumes between 0.1 and 25 μl were analysed on the HPLC to form a calibration curve consisting of the mass of naphthalene (x 106 g) versus the integrated peak area on the chromatogram (x 105). The calibration curve was performed in duplicate and resulted in an R2 value greater than 0.99.

-p

2.4.2 pH and Oxidation-Reduction Potential The pH and oxidation-reduction potential (ORP) of aqueous solutions were measured using an InLab Versatile pH and InLab Redox probe respectively (Mettler Toledo). Prior to use, the pH probe was calibrated with pH 4, 7, and 10 buffer solutions, while the ORP probe was calibrated with a 240 mV standard buffer solution.

ur

na

lP

re

2.4.3 By-Product Identification An Agilent gas chromatograph (GC 6890) coupled with mass spectroscopy (MS 5973N) was utilized to detect and identify by-product species generated during electrochemical treatments. 5 ml aqueous phase samples were taken from the electrochemical reactor at various timepoints and extracted for 5 minutes via vigorous shaking with 5 ml of hexane in a 40 ml headspace vial. Following extraction, 2 µl samples of the organic layer were injected into the GC with the following chromatographic conditions: helium carrier gas operating at 1.3 ml/min at a split injection ratio of 1:20 and temperature of 300 °C, initial oven temperature of 50 °C that was held for 3 minutes and then ramped at a rate of 18 °C until a temperature of 300 °C was reached and held for 10 minutes. Under these conditions, naphthalene was found to elute at 8.66 minutes, with all new by-product peaks being identified with the Wiley Spectral MS Library.

Jo

3 Results and Discussion 3.1 Unenhanced Anodic Oxidation Solubilized organic compounds can be electrochemical degraded via three main pathways: i) direct electron transfer at the anode, ii) electrogenerated hydroxyl radicals at the anode, and iii) indirection oxidation in the bulk solution due to generated oxidizing species such as active chlorine and peroxides [3,6,9,18]. Direct oxidation is not known to play a significant role as it has slow kinetics towards organic compounds [30,31], whilst indirect oxidation with hydroxyl radicals is considered to be the most effective. Such radicals are formed at the anode during the dissociation of water and are capable of fully mineralizing a variety of organic compounds [32]. However, expensive “non-active” anodic materials are generally needed to avoid the chemisorption of radicals onto the anodic surface [9,18] which makes them unavailable for organic destruction.

To test the capability for which the “active” carbon electrodes used in this work promote degradation pathways i) and ii), an inert 0.1 M Na2SO4 electrolyte was tried. Sodium sulphate is electrochemically inert in this system which results in no bulk oxidizing agents (e.g. persulphate compounds S2O42-) being generated; non-active anodes with high oxygen overpotentials such as BDD and PbO2 [9,25] are required for their production. Similarly, the formation of hydrogen peroxides via the electrochemical transformation of dissolved oxygen at the cathode surface was investigated via the TiSO4 method, and none was produced at any of the experimental conditions trialled [33,34]. Thus, any naphthalene removal under these experimental conditions would be due to direct oxidation or reactions with hydroxyl radicals at the anodic surface.

ro of

The cell was first operated under no applied electric currents in the presence of electrodes to act as volatilization and adsorptive controls. As naphthalene is one of the more volatile polycyclic aromatic compounds [35], its evaporation needs to be quantified to fully understand its removal from the system. This detail has not previously been addressed in the literature regarding the electrochemical treatment of other organic compounds. Similarly, carbon is a known adsorptive material and any naphthalene adsorption onto the carbon electrode surface may affect its removal rate; this was measured via a second control conducted in a closed vial with no headspace to prohibit any volatilization from occurring. Following the control runs, the cell was operated at 40 and 100 mA to test for hydroxyl radical formation at a range of electric currents. The results are shown in Figure 2 where the experimental data is fitted to an exponential first-order fitting.

lP

re

-p

Over a 48 hour period (data not shown), insignificant amounts of naphthalene adsorption were seen to occur on the electrode surfaces. This is believed to be due to the unidirectionally-oriented carbon fibres and binding agent used to hold the carbon electrodes together, limiting the amount of micropores available for organic adsorption. In contrast, naphthalene volatilization was measured to occur at a rate of 7.3 × 10-5 s-1, which was slightly enhanced upon application of an electric current. A constant current of 40 mA was seen to increase the overall removal rate to 1.7 × 10-4 s-1, removing 70 % of the naphthalene in 2 hours of treatment, likely due to small amounts of hydroxyl radicals being produced at the active anodic surface. HPLC analysis was used to confirm this and determine what electrochemical transformations the solubilized naphthalene was undergoing.

0.6

na ur

0.8

Jo

Naphthalene Concentration [C/Co]

1.0

0.4

0.2

0 mA 40 mA Oxidation

0.0 0

2000

4000 Time [seconds]

6000

8000

Figure 2: Naphthalene removal with time during electrolysis at 0 and 40 mA in a 0.1 M Na2SO4; initial concentration is 20 mg/L. Points are experimental data and solid lines a first-order exponential fitting; the dashed line is naphthalene losses due to anodic oxidation alone.

re

-p

ro of

HPLC chromatograms presented one by-product peak that eluted at 1.2 minutes that continued to increase in size simultaneously to naphthalene disappearance. The combined rates of naphthalene volatilization and the growing size of the by-product peak area amounted to the overall naphthalene removal rate that was observed, indicative of a direct electrochemical transformation process. The anodic oxidation rate itself was calculated by taking the difference between the removal rates of the volatilization and 40 mA experimental trials, and was found to be 9.3 × 10-5 s-1; approximately 1.3 times faster than the evaporative rate. This rate, presented by the dashed line in Figure 2, is critical to know when considering field deployment. In environmental applications where contaminated groundwaters present at subsurface levels of 10 cm and greater, naphthalene volatilization is known to be significantly slowed (nearly 13 fold) [36] and will have negligible effects on overall removal rates. This indicates that unenhanced electrochemical treatments for environmental remediation efforts would rely solely on radical formation at the anode surface, without the aid of volatilization. Although this is feasible, extremely long treatment times would be required to obtain sufficient removal of naphthalene species. The extrapolated data from Figure 2 suggests timeframes over 800 hours of treatment would be needed to reach contaminant levels low enough to protect 95 % of marine species as denoted by the ‘Australian and New Zealand Guidelines for Fresh and Marine Water Quality’ [37]. Operating the cell at higher currents such as 100 mA (data not shown) did not increase naphthalene removal rates in comparison to the 40 mA treatment, likely due to the occurrence of parasitic side reactions occurring on the electrode surfaces, such as oxygen and hydrogen evolution, rather than hydroxyl formation at the increased cell potentials [10]. These results demonstrate the inabilities of the active carbon anode to electrogenerate hydroxyl radicals for the purposes of naphthalene degradation, illuminating why expensive anodic materials such as BDD are generally chosen for electrochemical treatment methods.

ur

na

lP

HPLC analysis identified the by-product being formed as 1-naphthol, a hydroxylated naphthalene species known to be a product formed when naphthalene reacts with hydroxyl radicals [38]. This compound is toxic to aquatic life and would not be an acceptable result if this treatment were to be applied in environmental applications. Furthermore, increasing treatment times did not result in naphthol being oxidized further, and its concentration was not seen to decrease, indicating that naphthalene can only be partially oxidized by the active electrodes used in this work. Partial oxidation is commonly seen in the literature when using active electrodes, again exemplifying why non-active electrodes are more beneficial for this type of reaction pathway; they can achieve full combustion of organic species.

Jo

These results show that with the use of active carbon electrodes, an enhanced treatment process using indirect bulk oxidation methods is required to obtain efficient, quick removal rates, as well as naphthalene conversion into a lesser toxic form. A new set of experiments was conducted to address whether electrogenerated active chlorine species on carbon anodes enable naphthalene degradation in a more efficient manner. Several experimental criteria were assessed such as degree of naphthalene removal, associated kinetics, energy consumption and efficiency, and by-products formed during electrochemical treatment. 3.2

Enhanced Active Chlorine Pathway

3.2.1 Effect of Applied Current Enhanced bulk oxidation experiments were conducted utilizing a 0.1 M NaCl supporting electrolyte at a series of currents ranging 40 – 220 mA. The resulting normalized naphthalene concentration

ro of

(C/C0) as a function of time and applied current is shown in Figure 3, where it is seen that naphthalene removal was significantly increased in comparison to the unenhanced trials. In all cases, removal commenced immediately and continued until fully removed from the system (within three hours of treatment). Full removal was detected at timeframes nearing 1, 1.5, 2, and 2.5 hours for applied electric currents of 160, 100, 70, and 40 mA respectively, and removal rates were up to 12 and 9 times greater than those observed in the volatilization control and oxidation processes alone. The Australian regulations for marine water contaminant concentration were reached in all cases [37], demonstrating the effectiveness for which electrogenerated active chlorine species attack solubilized naphthalene when electrochemically treated in the presence of chloride ions. Such results are due to the electrogeneration of active chlorine species on the carbon anodic surface, such as hypochlorous acid in Equations (1) - (4), that quickly attack and aid in the breakdown of organic compounds in the bulk solution. Thus, although the active carbon electrodes used in this work are ineffective for unenhanced anodic oxidation processes, they are effective for bulk oxidation processes such as the active chlorine mediated pathway.

re

-p

Results in the inset of Figure 3 agree with Faraday’s law of electrolysis, stating the rate at which a substance reacts at an electrode surface is directly proportional to the amount of current passed. Removal rates linearly increased with increases in the applied current until a current of 160 mA was reached. This is due to the increased rates at which Reactions (1) - (3) occur, and with a greater quantity of active chlorine species produced, naphthalene removal is enhanced. Operating the electrochemical cell at currents greater than 160 mA (data not presented) did not further increase naphthalene removal rates. This phenomena is commonly observed in electrochemical systems when the overall kinetics become limited by the mass transfer of chloride ions diffusing from the bulk electrolyte to the anode surface, indicative of a mass transfer controlled process [29]. Further increases in current will not enhance reaction rates as there is limited supply of solute at the electrode which can undergo electrochemical reaction.

Jo

ur

na

lP

The limiting removal rate observed at 160 mA equates to an applied current density of 15.5 mA/cm2 of external electrode surface area and is the maximum rate per square unit at which active chlorine species can be produced at the carbon anode used in this work. Further increases in naphthalene removal rates within this system would only occur via increases in the electrode area, as the 15.5 mA/cm2 limit would remain if the hydrodynamics and bulk concentration of the electrolyte are kept constant [29,39]. This is a key finding, as it would aid in sizing electrode areas and predicting treatment outcomes when scaling up the treatment process for field deployment scenarios.

1.0 4

0.8

3

Ln(C/C0)

Naphthalene Concentration [C/C0]

3.5

0.6

2.5 2 1.5 1 0.5

0.4

0 0

2000

4000

6000

8000

Time [seconds]

0.0 0

2000

4000

6000 Time [seconds]

40 mA

70 mA

100 mA

160 mA

ro of

0.2

8000

10000

-p

Figure 3: Naphthalene losses with time at varying applied currents in 0.1 M NaCl. Points are experimental data and lines are a first-order exponential fitting; the subplot is the linearized data and first-order model.

Jo

ur

na

lP

re

At currents greater than 160 mA, parasitic side reactions, such as solvent breakdown into hydrogen and oxygen gases, will also start to occur and limit the amount of effective current passed for active chlorine production [30,40]. Oxygen evolution at the anode (1.23 V/Standard Hydrogen Electrode (SHE)) becomes thermodynamically favoured over the oxidation of chloride ions (1.36 V/SHE) when operating at higher cell potentials. Table 1 summarizes the increasing cell potentials observed at the varying applied currents within this work, resulting in a less efficient use of energy as oxygen production becomes favoured over naphthalene destruction. This may also be why the hydroxylated naphthalene products observed during the Na2SO4 electrolyte experiments were not detected during the active chlorine studies; the discharge of oxygen and presence of chloride ions at the anode inhibit the hydroxylated reactions of hydroxyl radicals from occurring [8,11,41] as oxygen and active chlorine production are thermodynamically favoured on the active anode used in this work. Other possibilities could be that the naphthalene underwent reaction with the quickly electrogenerated active chlorine species prior to reaching the anodic surface containing hydroxyl radicals, or that higher applied currents beyond 160 mA result in increased amounts of active chlorine species migrating to and being reduced at the cathode surface prior to reacting with naphthalene, further limiting the rate of organic removal [42,43]. At all applied currents, naphthalene concentrations were seen to follow an exponential decrease with time as is widely observed in the literature [11–13,44]. A pseudo-first order kinetic analysis was conducted and the exponential and linearized models were fit to the data in Figure 3. The first order assumption is described by Equation (6) where it is assumed that the concentration of active chlorine species (OX) is much greater than the concentration of organics, making [OX] independent of time and an apparent rate constant, 𝑘𝑎𝑝𝑝 , is used. 𝑟 = −𝑘𝑂𝑋 [𝑂𝑋][𝑂𝑟𝑔𝑎𝑛𝑖𝑐] ≅ −𝑘𝑎𝑝𝑝 [𝑂𝑟𝑔𝑎𝑛𝑖𝑐]

(6)

The first-order assumption was seen to fit the data well, summarized in Table 1, however it became

evident that the initial timepoints are overpredicted, whilst the later timepoints are underpredicted. This is likely due to the model being unable to account for the time dependence of OX concentration, and assuming a constant removal rate of 𝑘𝑎𝑝𝑝 ; the rate of naphthalene removal will instead vary as the concentration of OX does. The first-order model also makes it difficult to predict naphthalene removal rates at current densities and experimental conditions not studied, as 𝑘𝑎𝑝𝑝 is known to fluctuate with reactor configuration, hydrodynamic conditions, and operating parameters [11,13,14,21,45,46]. Thus, it would be beneficial for a kinetic model to be able to incorporate such variations.

-p

ro of

3.2.2 Kinetic Model Theory It is assumed that naphthalene removal occurs via two main mechanisms: a first-order volatilization and a second-order reaction with electrogenerated active chlorine species. This is described by Equation (7), where 𝑘𝑣𝑜𝑙 is the volatilization rate (s-1), [𝑁] is the naphthalene concentration (M), 𝑘𝑟𝑥𝑛 is the second-order reaction rate between naphthalene and active chlorine species (M-1 s-1), and [𝑂𝑋] is the concentration of active chlorine species (M). The first-order volatilization rate was experimentally found from the volatilization control run presented in Figure 2 and is considered to be constant at 7.30 X 10-5 s-1. Although the temperature of the electrochemical cell was seen to fluctuate ± 1 °C from the room temperature of 23 °C during experimental trails, this variation was assumed to have negligible effects on the volatilization rate. No other temperatures were studied in this work, however previously published works have shown that temperatures ranging a span of 35 °C had insignificant effects on the abatement of organics (e.g. the 𝑘𝑟𝑥𝑛 [𝑁][𝑂𝑋] term) when utilizing the active chlorine pathway [10]. (7)

re

𝑑[𝑁] = −𝑘𝑣𝑜𝑙 [𝑁] − 𝑘𝑟𝑥𝑛 [𝑁][𝑂𝑋] 𝑑𝑡

Jo

ur

na

lP

The time dependence of OX concentration is described by its generation and reduction at the anodic and cathodic surfaces, and its second-order reaction with naphthalene. These three terms are represented on the right-hand side of Equation (8) respectively where 𝑖 is the applied current (C/mol), 𝑛 is the number of electrons required for OX production on the anode surface, 𝐹 is Faraday’s constant (96,485 C/mol), 𝑉 is the volume of solution within the reactor (L), 𝑎 is the specific area of the electrode, A/𝑉 (m-1), and 𝑘𝑚 is the mass transfer coefficient within the electrochemical reactor (m/s). For simplification of the model, the production of OX is assumed to be a single-step reaction encompassing Equations (1) - (3) that is dependent on the applied electric current, 𝑖 [12,16,45,47].and the reduction of OX on the cathode surface is strictly mass transfer limited (i.e. dependent on the diffusion of OX species from the anode to the cathode surface, represented by the 𝑎𝑘𝑚 term) [16]. The reactions of OX species with by-products formed during the treatment are also not considered. 𝑑[𝑂𝑋] 𝑖 = − 𝑎𝑘𝑚 [𝑂𝑋] − 𝑘𝑟𝑥𝑛 [𝑁][𝑂𝑋] 𝑑𝑡 𝑛𝐹𝑉

(8)

The mass transfer coefficient of the electrochemical cell, 𝑘𝑚 , is determined via the limiting-current technique of the widely used potassium ferro/ferricyanide redox couple [27,28]. The cell voltage was scanned from 0 - 5 V, and the resulting I-E plot is depicted in Figure 4.

500

Current [mA]

400

300

200

100

0 0.5

1

1.5

2

2.5

3

Voltage [V]

3.5

4

4.5

5

ro of

0

Figure 4: Current-potential curve obtained for the electrochemical cell when scanning from 0 – 5 V in a solution of 1 mM K4Fe(CN)6, 0.1 M K3Fe(CN)6, and 0.5 M Na2SO4.

ur

na

lP

re

-p

No evident current is observed in the potential range of 0 – 0.92 V, meaning electrochemical reactions are not occurring at the electrode surfaces. In the range of 0.92 – 1.34 V, the current slowly increases as Faradaic reactions of the ferri/ferro redox couple commence. The current is low enough in this range that the concentrations of the redox couple near the electrode surfaces are not significantly different from the bulk solution (e.g. greater than 95 % of the bulk concentration), and the rate of the electrochemical reaction is strictly limited by the amount of current being passed through the system, a charge transfer controlled process [29,48]. With further increases in cell potential, the current rises until a stabilized current is reached in the 2.85 – 3.5 V range, where negligible increases in current occur in relation to changes in the applied potential. This plateau, observed in Figure 4 at 212 mA, is representative of the limiting current, 𝑖𝑙𝑖𝑚 , at which the oxidation of K4Fe(CN)6 becomes strictly dependent on its diffusive mass transport from the bulk solution to the anode surface, a mass transfer controlled process [29,48]. At this current, the concentration of K4Fe(CN)6 is effectively zero at the anode surface and increasing the current has no effect on reaction rates. The determined limiting current is used in Equation (5) to determine the mass transport coefficient within the cell, 𝑘𝑚 , calculated to be 0.0021 m/s for the electrochemical reactor used in this work. It is noted that the current increases observed nearing 4.0 V and beyond are due to formation of hydrogen gas on the cathodic surface.

Jo

The 𝑖, 𝑛, 𝐹, 𝑉, and 𝑎 parameters in Equations (7) and (8) of the kinetic model are known and constant, whilst parameters 𝑘𝑣𝑜𝑙 and 𝑘𝑚 are determined experimentally and vary with temperature and hydrodynamics of the reactor respectively; all parameters are summarized in Table 2. The specific reaction rate between naphthalene and active chlorine, 𝑘𝑟𝑥𝑛 , becomes the only fitting parameter of the kinetic model and should be constant for all experimental conditions. The ODE15S multistep solver of MATLAB® is used to solve the system of ordinary differential equations (ODEs), and the NLINFIT tool is used to minimize the difference between the experimental data and model equations using a least squares method estimation. The average 𝑘𝑟𝑥𝑛 was determined to be 2.2 ± 0.4 M-1 s-1, and the modelled results are depicted in Figure 5 for all experimental runs described.

0.8 40 mA 70 mA

0.6

100 mA 160 mA

0.4

0.2

0.0 0

2000

4000

6000 Time [seconds]

ro of

Naphthalene Concentration [C/C0]

1.0

8000

10000

Figure 5: The kinetic model (solid lines) fit to the experimental data (points) using a 𝑘𝑟𝑥𝑛 value of 2.2 M-1s-1, to describe naphthalene removal with time during electrolysis in 0.1 M NaCl at varying electric currents.

na

lP

re

-p

The kinetic model is seen to fit the experimental data with the second-order reaction rate of 𝑘𝑟𝑥𝑛 = 2.2 M-1 s-1, and the comparison is summarized in Table 3. The dynamic model shows similar, if not greater, levels of accuracy in comparison to the first-order analysis yet has the added benefit of being able to accurately predict naphthalene removal despite changes in applied electric current. This is largely a result of the model predicting variations in OX concentration as a function of the applied current and time. This is shown in Figure 6 where the model predicts OX concentrations gradually increasing from an initial concentration of zero until a maximum value is reached due to the combined effects of its generation, being dependent on the applied electric current, and its removal from the system due to both reaction with naphthalene and reduction at the cathode surface as described by Equation (8). As seen, increasing the current up to 160 mA is beneficial for increased organic destruction, as equilibrium OX concentrations are maximized.

0.2

ur

0.3

160 mA

100 mA

Jo

OX Concentration [mM]

0.4

70 mA 40 mA

0.1

0 0

1000

2000

3000

Time [seconds]

Figure 6: OX concentrations with time as predicted by the kinetic model at varying electric currents in 0.1 M NaCl.

Ignoring the effects of electrogenerated by-products scavenging OX species is likely to result in the kinetic model overestimating the naphthalene removal rates achieved, whilst assuming all species within the system have the same mass transfer rate may result in an underestimation. However, the resultant fit of the kinetic model and data was concluded to be acceptable for predicting treatment outcomes, whilst simplifying the mathematical requirements of solving the system of ODEs. 3.2.3 Effect of NaCl Concentration A second set of experiments were conducted to investigate how electrolyte concentration affects resultant OX concentration and subsequent naphthalene removal rates. It is expected that less chloride in the system would lower electrogenerated OX concentrations as there is less available solute at the electrode surface to react, and naphthene removal rates would decrease. To investigate this, NaCl concentrations were varied between 0.01, 0.025, 0.05, and 0.1 M whilst the applied electric current was held constant at 100 mA in all cases.

-p

ro of

The results shown in Figure 7 show that naphthalene removal rates were greatly affected by the starting electrolyte concentration, more so than to changes in the applied current. First-order removal rates were seen to increase 1.6-fold with increasing NaCl concentration until a maximum rate was reached around 0.05 M. The rates increased from 2.1 × 10-4 to 5.2 × 10-4 s-1 when increasing the NaCl concentration from 0.01 to 0.05 M, highlighting the significance that chloride ions have on the removal process. Simultaneous to this, increasing the ionic conductivity of the electrolyte resulted in a significant reduction in the required cell potential, tabulated in Table 4. Thus, higher NaCl has the added benefits of both increased removal rates and reduced energy consumption, a valuable result for application in environmental scenarios. 4

re

1.0

-Ln[C/C0]

3

lP

1

na

0.6

0.4

Jo

0.2

2

0 0

2000

4000 6000 Time [seconds]

8000

ur

Naphthalene Concentration [C/C0]

0.8

0.0

0

2000 0.01 M NaCl

4000 Time [seconds] 0.025 M NaCl 0.05 M NaCl

6000

8000

0.1 M NaCl

Figure 7: Naphthalene removal with time at varying NaCl concentrations when operating at 100 mA. Points present experimental data and lines a fitted first-order exponential model; the subplot is the linearized data and first-order model.

The observed increases in removal rate are expected as the rate of reaction at an electrode is dependent on the concentration of electroactive species at its surface. Further increases to 0.1,

Figure 7, and 0.2 M NaCl (data not shown) had negligible effects on the system as the electrode began operating within a charge-transfer controlled process [29,48]. This observation is commonly seen in the literature where a maximum electrolyte concentration is ultimately reached [10,12,49] and will no longer effect electrochemical removal rates. Despite having no increased effect on removal, the energy consumption is still greatly reduced in these cases. The increased energy consumption observed at the lesser concentrated electrolytes is not only due to the decreased amount of solute at the electrode surface that can undergo electrochemical conversion to OX species, but also oxygen evolution becomes more competitive at the higher cell potentials required [12] in the less conductive solution, decreasing the effective current passed for OX production.

re

lP

6.0

y = 78.17x + 1.26 R² = 1.00

na

4.0

2.0

ur

First-Order Rate Constant, kapp X 104 [s-1]

-p

ro of

To account for these variations in the kinetic model, a new efficiency term, ℰ, is introduced into the OX generation term to describe its production at the anode surface. Oftentimes in the literature, an efficiency term is presented as a fitting parameter of the kinetic model to describe the fraction of current being utilized for chlorine production, as opposed to oxygen evolution, on the anodic surface [13,14,16,22,43,46,50,51]. However, this work found there to be a direct and linear correlation between the current efficiency and the starting chloride concentration. As depicted in Figure 8 and summarised in Table 4, the first-order naphthalene removal rate linearly increases with NaCl concentration until a concentration of 0.05 M is reached. Any higher concentrations (e.g. 0.1 and 0.2 M) having negligible effects on the system and the efficiency is not increased. Based on this, ℰ in the kinetic model becomes a known parameter representing the fraction of maximum chloride concentration utilized during treatment (CCl/Cmax, where Cmax is 0.05 M); this corresponds to ℰ values of 1.0, 1.0, 0.5, and 0.2 for NaCl concentrations of 0.1, 0.05, 0.025, and 0.01 M respectively. The addition of this term into the kinetic model is shown in Equation (9).

0.0

Jo

0

0.025

0.05

0.075

0.1

NaCl Concentration [M]

Figure 8: Variations in first-order removal rates of naphthalene as a function of NaCl concentration.

𝑑[𝑂𝑋] 𝜀𝑖 = − 𝑎𝑘𝑚 [𝑂𝑋] − 𝑘𝑟𝑥𝑛 [𝑁][𝑂𝑋] 𝑑𝑡 𝑛𝐹𝑉

(9)

It is assumed that the efficiency term is dependent only on the initial chloride concentration, and not time, as its concentration is expected to remain constant during treatment timeframes. This was deemed as an acceptable assumption as although chloride ions are used in the production of OX species, they are also regenerated via OX reduction at the cathode, Equations (10) and (11), and OX reaction with naphthalene species, Equation (4).

𝐻𝐶𝑙𝑂 + 2𝑒 − → 𝐶𝑙 − + 𝑂𝐻 −

(10)

(11) 𝐶𝑙𝑂− + 𝐻2 𝑂 + 2𝑒 − → 𝐶𝑙 − + 2𝑂𝐻 − The experimental results were replotted with the kinetic model containing the efficiency term, and the results are shown in Figure 9. In agreement with the results in Section 3.2.2 predicting naphthalene removal as a function of applied current, the 𝑘𝑟𝑥𝑛 value was still found to be 2.2 M-1 s-1, validating the model equations described within this work. The comparison of the experimental data and kinetic model are tabulated in Table 5, further providing evidence that the kinetic model can accurately describe the system despite changes in experimental conditions. In all cases, R2 values greater than 0.9 are achieved.

ro of

1

0.01 M

0.6

0.025 M

-p

0.05 M

0.4

0.1 M

re

Naphthalene [C/C0]

0.8

lP

0.2

0 0

2000

4000

6000

8000

na

Time [seconds]

Figure 9: The kinetic model (solid lines) fit to the experimental data (points) using a 𝑘𝑟𝑥𝑛 value of 2.2 M-1s-1, to describe naphthalene removal with time during electrolysis at 100 mA in varying NaCl concentrations.

Jo

ur

The slight discrepancy observed between the kinetic fit and data is likely due to the modelling assumptions made. Ignoring the oxidation of generated by-product species will result in overestimations of naphthalene removal, whereas the large quantities of oxygen and hydrogen gases produced due to operating the cell at 17.7 V (as observed in the 0.01 M NaCl trial) is likely to enhance the naphthalene volatilization rates [52] and result in an underestimation. However, the error detected with the kinetic fit was concluded to be small enough to still accurately predict treatment outcomes within the 10-fold range of 0.01 to 0.1 M NaCl. 3.2.4 Energy Requirements One of the main inhibitors of utilizing electrochemical technologies for the treatment of contaminant wastewaters, particularly in environmental applications, is the practicality for which it can be scaled. Large treatment volumes require large electrode areas and increased energy consumption, both of which may be impractical when using expensive anodic materials or treating waters in remote regions with limited resources. This work firstly aimed to reduce capital

expenditure via investigating the use of cheap carbon electrodes, but operational costs still need to be assessed to ensure energy efficiency is not compromised when utilizing a less effective material. The specific energy consumption, 𝐸𝑠𝑝 (kWh/kg), was calculated for each experimental trial to determine the energy usage per unit of naphthalene removed. This was done using Equation (12) where 𝐸𝑐𝑒𝑙𝑙 is the average cell voltage (V) which remained mostly constant, 𝐼 is the applied current (A), 𝑡 is the treatment time (hr), 𝑉 is the volume of solution treated (L), and 𝛥𝑁 is the quantity of naphthalene removed from the system (g/L). 𝐸𝑠𝑝 =

𝐸𝑐𝑒𝑙𝑙 𝐼𝑡 𝑉𝛥𝑁

(12)

-p

ro of

The results shown in Figure 10 clearly indicate that the energy consumed is highly dependent on both the applied electric current and electrolyte concentrations. Increasing the electric current to achieve quicker naphthalene removal rates consumed more energy per unit of naphthalene degraded. When operating the cell at 40 and 160 mA, 104 and 303 kWh/kg of energy were consumed to reach full removal, whilst removal rates of 4.3 × 10-4 and 8.9 × 10-4 s-1 were achieved. There exists a clear trade-off between required treatment times and energy consumption, as at higher applied currents, energy is wasted on solvent breakdown (e.g. the conversion of water into oxygen and hydrogen gases) rather than OX species production. As discussed previously, this is largely due to increased competition between the two reactions at the higher voltages applied.

na

lP

re

In contrast, increases in chloride concentration greatly reduce energy consumption whilst simultaneously increasing removal rates. Energy usage was reduced 7-fold, from 1502 to 219 kWh/kg, when increasing the NaCl concentration from 0.01 to 0.1 M, while the removal increased from 2.1 × 10-4 to 5.5 × 10-4 s-1 respectively. This is expected as more ions within the system will increase the conductivity of the electrolyte and reduce the electrical resistance of the cell. This results in a lowered applied voltage (17.7 vs 5.3 V for 0.01 and 0.1 M NaCl) needed to drive the desired current through the cell, and ultimately saves on energy consumption. The combined effects of lowered potentials and increases in chloride ions at the anodic surface also reduce the competition between oxygen and OX formation at the anode, allowing for a greater effective current to be passed per unit of naphthalene removed.

Jo

ur

It is interesting to note that although increasing the NaCl concentration beyond 0.05 M was found to have insignificant effects on naphthalene removal rates, it still aids in reducing energy consumption. Increasing the NaCl concentration from 0.05 to 0.1 M had no effect on naphthalene removal yet utilized half the amount of energy due to the decreased electrical resistance within the cell. This is a beneficial finding for the environmental remediation of coastal regions and other high saline environments such as the Antarctic, as seawater is comprised of a chloride concentration nearing 0.6 M. In this instance, no external chemicals would be required, and energy consumption would be minimized. Not considering capital expenditure, this data from this work estimates that 104 kWh of energy can remove one kg of solubilized naphthalene when operating at 40 mA; this corresponds to an operating cost nearing AUD $1.05 per m3 of a naphthalene saturated water supply (assuming Tasmanian electricity prices of 32.59 ₵/kWh and a naphthalene saturation level of 31 mg/L). It is difficult to compare these results to other authors who have utilized electrochemical methods to treat wastewaters due to differences in experimental conditions, yet the energy usage in this work has utilized less than one-tenth of the energy consumed in other works utilizing BDD anodes for aqueous organic destruction [10,53]. Although non-active BDD anodes are considered to be the

most effective for hydroxyl radical formation and wastewater treatment, their energy consumption can be outperformed via the chloride-mediated oxidation pathway on the active electrodes used in this work. The low energy requirements and quick treatment times make this treatment highly favourable for use in environmental remediation efforts. It is presumed that such low energy requirements could be supplied by either diesel or solar power in remote environmental applications, such as those presented in the Antarctic and other comparable Polar Regions. 1600

250

150

1100

600

100

50 40

70

100

ro of

Specific Energy Consumption [kWh/kg]

Specific Energy Consumption [kWh/kg]

350

0.01

160

0.025

0.05

0.1

NaCl Concentration [M]

-p

Applied Current [mA]

re

Figure 10: Specific energy consumption at varying electric currents in 0.1 M NaCl (left) and NaCl concentrations at 100 mA of current (right).

lP

3.2.5 By-Product Identification A prevalent concern with utilizing active chlorine degradation techniques is the possibility of generating chlorinated by-products that may be more toxic than the starting compound. To investigate this, HPLC and GC/MS were utilized to determine what electrochemical transformations the solubilized naphthalene was undergoing during treatment.

Jo

ur

na

HPLC analysis in Figure 11 shows the naphthalene chromatogram peak (I) eluting at 3.1 minutes and exponentially decreasing with treatment time. The simultaneous nature at which three main byproduct peaks evolve (II-IV) with the disappearance of naphthalene is highly indicative of a direct electrochemical conversion process. Unlike the results of the inert sodium sulphate electrolyte, no naphthol peaks were detected as all of the effective current was going towards active chlorine and oxygen production. GC/MS was able to identify the by-products being formed and the summarized results are tabulated in Table 6. It was revealed that the first and main reaction that naphthalene undergoes is its mono-chlorination to form 1-chloro- and 2-chloronaphthalene species, associated with the shoulder peak of compounds II and III. This is shown in Figure 11B and is consistent with another work in the literature [26] describing naphthalene degradation. These by-products are due to the highly oxidizing HOCl species undergoing electrophilic substitution on the electron dense aromatic naphthalene structure [7], illustrating the effectiveness of the active chlorine pathway. The concentration of the mono-chlorinated species continued to increase until a maximum concentration was reached nearing the 60-minute mark whereby it accounted for 31 % of the initial naphthalene concentration. Chlorinated naphthalene compounds are known to be toxic to aquatic organisms and their generation is a cause for concern, especially in environmental applications. However, prolonged treatment time (refer to Figure 11C – D) showed that their concentration began to decrease around

lP

re

-p

ro of

60 minutes and continued to do so until they were fully removed from the system within 180 minutes of treatment. This is a crucial finding as even though toxic by-products are being generated, they are also being further degraded and removed from the system. This is in agreement with other works in the literature that show up to 95 % of the chemical oxygen demand in an organic contaminated water being removed when using the electrochemical chlorine mediated pathway [10].

na

Figure 11: Evolution of naphthalene and by-products on HPLC-DAD chromatograms (detection at 275 nm) when treated in 0.1 M NaCl at 100 mA. (A) is pre-treatment, (B) at 30 minutes, (C) at 90 minutes, and (D) at 180 minutes of treatment.

Jo

ur

During the reduction and ultimate removal of the mono-chlorinated naphthalene species, a new chromatographic peak (IV) formed and eluted at 3.4 minutes. The broadened chromatogram peak shown in Figure 11D was found to be comprised of a variety of by-products including oxygenated naphthalene species such as naphthoquinone, 1,4-napthoquinone-2,3-oxide, and small quantities of 1,2-dichloronaphthalene. The presence of oxygenated compounds suggests that hypochlorous acid may attack naphthalene in a single electron transfer process to produce a radical cation, which then undergoes nucleophilic substitution from water to form oxygenated naphthalene compounds such as those described [54]. A second possible reaction pathway would be the presence of small quantities of hydroxyl radicals present at the anodic surface with which the chlorinated naphthalene species could react with; this would produce only a limited number of by-products, in agreement with Figure 11, as the carbon electrodes used in this work were shown to produce limited amounts of radicals vial the Na2SO4 electrolyte trial discussed above. These by-products were only seen at low concentrations, and all detectable naphthalene had been removed from the system. Interestingly, no tri-chloronaphthalene species were detected. The presence of toxic naphthoquinone and dichloro-naphthalene species is not considered an appropriate result for treatment in environmental regions. However, additional degradation and

oxidization evidenced by the presence of 1,4-naphthoquinone-2,3-oxide vastly decreases the toxicity of by-products formed and gives promise to the treatment method. With further treatment time, chromatographic peaks continued to decrease until no detectable peaks were seen at 240 minutes and beyond on either HPLC of GC/MS instrumentation. It is assumed that the oxygenated naphthalene compounds underwent further ring opening and degradation to form compounds such as carboxylic acids and alcohols [55,56], however the equipment utilized in this work do not allow for their detection; this will be addressed in future research. However, it is concluded that if sufficient timeframes are used (up to 4 hours of treatment) the active chlorine pathway promoted on active electrodes such as carbon may be a promising technique for use in environmental remediation efforts, as no toxic products are seen to remain. Conclusions

na

lP

re

-p

ro of

This work has identified that electrochemical methods are a deployable technology for the treatment of contaminated groundwaters in Polar Regions, such as the Antarctic, where resources and manual labour are limited. It was specifically shown that the active chlorine pathway can aid in the remediation of petroleum hydrocarbon spills, where the high priority pollutant of removal naphthalene can be fully degraded and removed from a contaminant water supply within three hours of treatment. Although toxic by-products are formed during this treatment, increased treatment times of up to 4 hours were found to subsequently remove all by-products from the system. Such results were achieved when using inexpensive carbon materials as electrodes, providing evidence that they are an effective alternative to the expensive anodic materials generally used in electrochemical processing, such as BDD and PbO2. Moreover, a kinetic model was proposed that can accurately describe naphthalene removal rates despite changes in operational conditions, such as saline concentration and applied electric current. Although the work herein focused on the remediation of extreme Polar environments, the outcomes of this research extend far beyond such regions. This green technology is believed to be beneficial for any environmental application as it has low energy requirements, requires no external chemicals, and achieves sufficient removal outcomes in short timeframes. Future work will focus on how this treatment applies to complex mixtures of diesel components, further aiding in its environmental application.

Funding

ur

Declarations of interest: none

This work was supported by the Australian Antarctic Science Project [4036].

Jo

References [1]

R. McDonald, O.G.G. Knox, Cold region bioremediation of hydrocarbon contaminated soils: Do we know enough?, Environ. Sci. Technol. 48 (2014) 9980–9981.

[2]

I. Snape, M.J. Riddle, J.S. Stark, C.M. Cole, C.K. King, S. Duquesne, D.B. Gore, Management and remediation of contaminated sites at Casey Station, Antarctica, Polar Rec. (Gr. Brit). 37 (2001) 199–214.

[3]

Y. Feng, L. Yang, J. Liu, B.E. Logan, Electrochemical technologies for wastewater treatment and resource reclamation, Environ. Sci. Water Res. Technol. 2 (2016) 800–831.

[4]

E. Brillas, C.A. Martínez-Huitle, Decontamination of wastewaters containing synthetic organic

dyes by electrochemical methods. An updated review, Appl. Catal. B Environ. 166–167 (2015) 603–643. C.A. Martínez-Huitle, S. Ferro, Electrochemical oxidation of organic pollutants for the wastewater treatment: direct and indirect processes, Chem. Soc. Rev. 35 (2006) 1324–1340.

[6]

F.C. Moreira, R.A.R. Boaventura, E. Brillas, V.J.P. Vilar, Electrochemical advanced oxidation processes: A review on their application to synthetic and real wastewaters, Appl. Catal. B Environ. 202 (2017) 217–261.

[7]

M. Deborde, U. von Gunten, Reactions of chlorine with inorganic and organic compounds during water treatment-Kinetics and mechanisms: A critical review, Water Res. 42 (2008) 13– 51.

[8]

C.A. Martínez-Huitle, M.A. Rodrigo, I. Sirés, O. Scialdone, Single and Coupled Electrochemical Processes and Reactors for the Abatement of Organic Water Pollutants : A Critical Review, Chem. Rev. 115 (2015) 13362–13407.

[9]

M. Panizza, G. Cerisola, Direct And Mediated Anodic Oxidation of Organic Pollutants, Chem. Rev. 109 (2009) 6541–6569.

[10]

A.M. Sales Solano, C.K. Costa de Araújo, J. Vieira de Melo, J.M. Peralta-Hernandez, D. Ribeiro da Silva, C.A. Martínez-Huitle, Decontamination of real textile industrial effluent by strong oxidant species electrogenerated on diamond electrode: Viability and disadvantages of this electrochemical technology, Appl. Catal. B Environ. 130–131 (2013) 112–120.

[11]

J. Zou, X. Peng, M. Li, Y. Xiong, B. Wang, F. Dong, B. Wang, Electrochemical oxidation of COD from real textile wastewaters: Kinetic study and energy consumption, Chemosphere. 171 (2017) 332–338.

[12]

N. Mohan, N. Balasubramanian, C.A. Basha, Electrochemical oxidation of textile wastewater and its reuse, J. Hazard. Mater. 147 (2007) 644–651.

[13]

A. Urtiaga, I. Ortiz, A. Anglada, D. Mantzavinos, E. Diamadopoulos, Kinetic modeling of the electrochemical removal of ammonium and COD from landfill leachates, J. Appl. Electrochem. 42 (2012) 779–786.

[14]

Á. Anglada, A. Urtiaga, I. Ortiz, D. Mantzavinos, E. Diamadopoulos, Boron-doped diamond anodic treatment of landfill leachate: Evaluation of operating variables and formation of oxidation by-products, Water Res. 45 (2011) 828–838.

[15]

A. V. Moreno-Palacios, R.E. Palma-Goyes, J. Vazquez-Arenas, R.A. Torres-Palma, Bench-scale reactor for Cefadroxil oxidation and elimination of its antibiotic activity using electrogenerated active chlorine, J. Environ. Chem. Eng. 7 (2019) 103173.

[16]

M. Mascia, A. Vacca, A.M. Polcaro, S. Palmas, J.R. Ruiz, A. Da Pozzo, Electrochemical treatment of phenolic waters in presence of chloride with boron-doped diamond (BDD) anodes: Experimental study and mathematical model, J. Hazard. Mater. 174 (2010) 314–322.

[17]

A.S. Fajardo, H.F. Seca, R.C. Martins, V.N. Corceiro, I.F. Freitas, M.E. Quinta-Ferreira, R.M. Quinta-Ferreira, Electrochemical oxidation of phenolic wastewaters using a batch-stirred reactor with NaCl electrolyte and Ti/RuO2anodes, J. Electroanal. Chem. 785 (2017) 180–189.

[18]

R. V. McQuillan, G.W. Stevens, K.A. Mumford, The electrochemical regeneration of granular activated carbons: A review, J. Hazard. Mater. 355 (2018) 34–49.

[19]

X. Zhu, J. Ni, J. Wei, P. Chen, Scale-up of B-doped diamond anode system for electrochemical oxidation of phenol simulated wastewater in batch mode, Electrochim. Acta. 56 (2011) 9439–

Jo

ur

na

lP

re

-p

ro of

[5]

9447. E.V. dos Santos, S.F.M. Sena, D.R. da Silva, S. Ferro, A. De Battisti, C.A. Martínez-Huitle, Scaleup of electrochemical oxidation system for treatment of produced water generated by Brazilian petrochemical industry, Environ. Sci. Pollut. Res. 21 (2014) 8466–8475.

[21]

A. Urtiaga, P. Gómez, A. Arruti, I. Ortiz, Electrochemical removal of tetrahydrofuran from industrial wastewaters: Anode selection and process scale-up, J. Chem. Technol. Biotechnol. 89 (2014) 1243–1250.

[22]

A. Anglada, A. Urtiaga, I. Ortiz, Pilot scale performance of the electro-oxidation of landfill leachate at boron-doped diamond anodes, Environ. Sci. Technol. 43 (2009) 2035–2040.

[23]

J. Wasley, T.J. Mooney, C.K. King, Soil invertebrate community change over fuelcontaminated sites on a subantarctic island: An ecological field-based line of evidence for site risk assessment, Integr. Environ. Assess. Manag. 12 (2016) 306–314.

[24]

J.L. Rayner, I. Snape, J.L. Walworth, P.M. Harvey, S.H. Ferguson, Petroleum-hydrocarbon contamination and remediation by microbioventing at sub-Antarctic Macquarie Island, Cold Reg. Sci. Technol. 48 (2007) 139–153.

[25]

J. Muff, E.G. Søgaard, Electrochemical degradation of PAH compounds in process water: A kinetic study on model solutions and a proof of concept study on runoff water from harbour sediment purification, Water Sci. Technol. 61 (2010) 2043–2051.

[26]

J. Muff, E.G. Søgaard, Identification and fate of halogenated PAHs formed during electrochemical treatment of saline aqueous solutions, J. Hazard. Mater. 186 (2011) 1993– 2000.

[27]

D.A. Szánto, S. Cleghorn, C. Ponce-De-León, F.C. Walsh, The Limiting Current for Reduction of Ferricyanide Ion at Nickel: The Importance of Experimental Conditions, AIChE J. 54 (2008) 802–810.

[28]

C. Ponce-De-León, C.T.J. Low, G. Kear, F.C. Walsh, Strategies for the determination of the convective-diffusion limiting current from steady state linear sweep voltammetry, J. Appl. Electrochem. 37 (2007) 1261–1270.

[29]

A.J. Bard, L.R. Faulkner, Electrochemical methods : fundamentals and applications., Wiley, New York, 2001.

[30]

C. Comninellis, G. Chen, Electrochemistry for the Environment, Springer, London, 2010.

[31]

J.L. Wang, L.J. Xu, Advanced Oxidation Processes for Wastewater Treatment: Formation of Hydroxyl Radical and Application, Crit. Rev. Environ. Sci. Technol. 42 (2012) 251–325.

[32]

C. Comninellis, Electrocatalysis in the Electrochemical Conversion / Combustion of Organic Pollutants, Electrochim. Acta. 39 (1994) 1857–1862.

Jo

ur

na

lP

re

-p

ro of

[20]

[33]

Y. Xiao, J.M. Hill, Mechanistic insights for the electro-Fenton regeneration of carbon materials saturated with methyl orange: Dominance of electrodesorption, J. Hazard. Mater. 367 (2019) 59–67.

[34]

W. Zhou, L. Rajic, L. Chen, K. Kou, Y. Ding, X. Meng, Y. Wang, B. Mulaw, J. Gao, Y. Qin, A.N. Alshawabkeh, Activated carbon as effective cathode material in iron-free Electro-Fenton process: Integrated H 2 O 2 electrogeneration, activation, and pollutants adsorption, Electrochim. Acta. 296 (2019) 317–326.

[35]

H.I. Abdel-Shafy, M.S.M. Mansour, A review on polycyclic aromatic hydrocarbons: Source,

environmental impact, effect on human health and remediation, Egypt. J. Pet. 25 (2016) 107– 123. W. Jury, W. Spencer, W. Farmer, Behavior Assessment Model for Trace Organics in Soil: III. Application of Screening Model, J. Environ. Qual. 13 (1984) 573–579.

[37]

A. and N.Z.E. and C. Council, A. and R.M.C. of A. and N. Zealand, eds., Australian and New Zealand guidelines for fresh and marine water quality. Volume 1, The Guidelines / Australian and New Zealand Environment and Conservation Council, Agriculture and Resource Management Council of Australia and New Zealand, Australian and New Zealand Environment and Conservation Council and Agriculture and Resource Management Council of Australia and New Zealand, Canberra, 2000.

[38]

N.J. Bunce, L. Liu, J. Zhu, D.A. Lane, Reaction of naphthalene and its derivatives with hydroxyl radicals in the gas phase, Environ. Sci. Technol. 31 (1997) 2252–2259.

[39]

K. Scott, Reactor modelling for electrochemical processes, J. Chem. Technol. Biotechnol. 54 (1992) 257–266.

[40]

A. Ozcan, M.A. Oturan, Degradation of picloram by the electro-Fenton process, J. Hazard. Mater. 153 (2008) 718–727.

[41]

P. V. Nidheesh, R. Gandhimathi, Trends in electro-Fenton process for water and wastewater treatment: An overview, Desalination. 299 (2012) 1–15.

[42]

S.J. George, R. Gandhimathi, P.V. Nidheesh, S.T. Ramesh, Electro-Fenton Method Oxidation of Salicylic Acid in Aqueous Solution with Graphite Electrodes, Environ. Eng. Sci. 30 (2013) 750– 756.

[43]

A.M. Polcaro, M. Mascia, S.P. Palmas, A. Vacca, Kinetic study on the removal of organic pollutants by an electrochemical oxidation process, Ind. Eng. Chem. Res. 41 (2002) 2874– 2881.

[44]

D. Rajkumar, K. Palanivelu, Electrochemical treatment of industrial wastewater, J. Hazard. Mater. 113 (2004) 123–129.

[45]

M.R. Cruz-Díaz, E.P. Rivero, F.A. Rodríguez, R. Domínguez-Bautista, Experimental study and mathematical modeling of the electrochemical degradation of dyeing wastewaters in presence of chloride ion with dimensional stable anodes (DSA) of expanded meshes in a FM01-LC reactor, Electrochim. Acta. 260 (2018) 726–737.

[46]

V. Díaz, R. Ibàñez, P. Gómez, A.M. Urtiaga, I. Ortiz, Kinetics of electro-oxidation of ammoniaN, nitrites and COD from a recirculating aquaculture saline water system using BDD anodes, Water Res. 45 (2011) 125–134.

[47]

A.M. Polcaro, A. Vacca, M. Mascia, S. Palmas, J. Rodiguez Ruiz, Electrochemical treatment of waters with BDD anodes: Kinetics of the reactions involving chlorides, J. Appl. Electrochem. 39 (2009) 2083–2092.

[48]

Instrumental methods in electrochemistry., Chichester : E. Horwood ; New York : Halsted Press, 1985., 1985.

[49]

G. Panizza, M; Bocca, C.; Cerisola, Electrochemical treatment of wastewater containing polyaromatic organic pollutants, Water Res. 34 (2000) 2601–2605.

[50]

P. Canizares, J. Garcia-Gomez, J. Lobato, M. a Rodrigo, Modeling of wastewater electrooxidation processes part II. Application to active electrodes, Ind. Eng. Chem. Res. 43 (2004) 1923–1931.

Jo

ur

na

lP

re

-p

ro of

[36]

F. Walsh, G. Reade, Design and Performance of Electrochemical Reactors for Efficient Synthesis and Environmental Treatment Part 1. Electrode Geometry and Figures of Merit, Analyst. 119 (1994) 791–796.

[52]

R.L. Johnson, P.C. Johnson, D.B. McWhorter, R.E. Hinchee, I. Goodman, An Overview of In Situ Air Sparging, Groundw. Monit. Remediat. 13 (1993) 127–135.

[53]

J.M. Aquino, G.F. Pereira, R.C. Rocha-Filho, N. Bocchi, S.R. Biaggio, Electrochemical degradation of a real textile effluent using boron-doped diamond or β-PbO 2 as anode, J. Hazard. Mater. 192 (2011) 1275–1282.

[54]

X. Xu, R. Xiao, D.D. Dionysiou, R. Spinney, T. Fu, Q. Li, Z. Wang, D. Wang, Z. Wei, Kinetics and mechanisms of the formation of chlorinated and oxygenated polycyclic aromatic hydrocarbons during chlorination, Chem. Eng. J. 351 (2018) 248–257.

[55]

S.D. Boyce, J.F. Hornig, Reaction Pathways of Trihalomethane Formation from the Halogenation of Dihydroxyaromatic Model Compounds for Humic Acid, Environ. Sci. Technol. 17 (1983) 202–211.

[56]

L. Tran, P. Drogui, G. Mercier, J.-F. Blais, Comparison between Fenton oxidation process and electrochemical oxidation for PAH removal from an amphoteric surfactant solution, J. Appl. Electrochem. 40 (2010) 1493–1510.

Jo

ur

na

lP

re

-p

ro of

[51]

Table 1: Pseudo-first order removal rates of naphthalene in 0.1 M NaCl at varying applied currents. The coefficient of determination between the first-order analysis and data, and the average voltage is tabulated.

kapp [s-1] 7.3 × 10-5 4.3 × 10-4 4.8 × 10-4 5.5 × 10-4 8.9 × 10-4

Current [mA] 0 40 70 100 160

R2 [-] 0.982 0.977 0.957 0.997 0.988

Average Voltage [V] 0.0 4.2 4.7 5.3 6.4

Applied Current, i (C/s) 0.04, 0.07, 0.10, or 0.16

Electrons, n (-) 2

Faraday’s Constant, F (C/mol) 96,485

Reactor Volume, V (L) 0.2

Specific Electrode Area, a (m-1) 5.15

Mass Transfer Coefficient, km (m/s) 0.0021

-p

Volatilization Rate, kvol (s-1) 7.30 × 10-5

ro of

Table 2: Known and experimentally determined parameters used within the proposed mathematical model.

R2 [-] 0.978 0.983 0.990 0.983

na

lP

Current [mA] 40 70 100 160

re

Table 3: Mathematical fit between experimental data and proposed kinetic model at various applied currents.

RMSE 0.049 0.040 0.029 0.040

Table 4: Pseudo-first order removal rates of naphthalene in varying concentrations of NaCl at 100 mA. The coefficient of determination between the first-order analysis and data, and the average voltage is tabulated.

Jo

ur

NaCl [M] 0.01 0.025 0.05 0.1

kapp [s-1] 2.1 × 10-4 3.3 × 10-4 5.2 × 10-4 5.5 × 10-4

R2 [-] 0.993 0.991 0.987 0.997

Average Voltage [V] 17.7 9.9 7.0 5.3

Table 5: Mathematical fit between experimental data and proposed kinetic model at various NaCl concentrations.

R2 [-] 0.948 0.983 0.961 0.990

NaCl Concentration [M] 0.01 0.025 0.05 0.1

RMSE 0.053 0.034 0.057 0.029

Table 6: Mass spectrophotometry results detailing by-products formed when electrochemically treating naphthalene solutions in 0.1 M NaCl at 100 mA.

HPLC Retention Time (min)

GC/MS Retention Time (min)

Molecular Ion, M+

Identification

I

3.1

8.65

128

Naphthalene

II / III

5.1 / 5.3

10.30

162

IV

3.4

10.5 / 11.2 / 11.6

158 / 174 / 196

ro of

Peak

Jo

ur

na

lP

re

-p

1-chloronaphthalene / 2chloronaphthalene Naphthoquinone / 1,4-naphthoquinone 2,3-oxide / 1,2-dichloronaphthalene