Introduction

Conserving and maintaining the quality of water resources represent one of the worldwide challenges. Therefore, water resources are exposed to pollution principally originating from human activities (Schwarzenbach et al. 2010), development of industrial activities (Hettige et al. 1995), intensive agriculture, and landfills of household and industrial wastes (Bensadok et al. 2011; Burkholder et al. 2007; Killebrew and Wolff 2010). A high amount of organic and inorganic pollutants are continuously released into receiving medium via direct discharge of wastewater (Zhao et al. 2015). However, rivers in the arid countries, such as the south Mediterranean region, present a reduced water flow which limit pollution dilution and had an impact on the natural biologic degradation capacity of the river (Gasith and Resh 1999).

Considering the composition of municipal wastewater, 75 % of the nitrogen load to a conventional wastewater treatment plant (WWTP) originates from urine (Kuntke et al. 2012; Larsen and Gujer 1996). However, urine contributes only by 1 % of the volume of the wastewater (Larsen and Gujer, 1996). In fact, urea is considered as the main sources of nitrogen compounds in the municipal wastewater (MWW) and it could be hydrolyzed to ammonia by the urease enzyme produced by microorganisms (Ali et al. 2011). Therefore, the discharge of ammonia increases the soil pH and contributes to water eutrophication. Consequently, the use of conventional technologies in the municipal wastewater treatment plant (MWWTP) for the removal of organic and inorganic matter could be related with the irreversible environmental damage (Belhaj et al. 2015; Gatica and Cytryn 2013).

Moreover, most of organic pollutants are persistent and refractory to the conventional wastewater treatment processes and pass through treatment plants without any modification due to limited removal of conventional treatment processes (Bahri 1998; Huebsch et al. 2013, Jellali et al. 2011). The diffusion of pollution to the soil, surface, and ground water via draining systems causes potential risks for the ecosystem, animals, and humans’ health.

Therefore, the improvement and the development of new treatment technologies and the continuous control of the quality of treated wastewater are the key issues to protect our water resources. In fact, it becomes important to develop more suitable treatment than conventional ones. These treatments have to be effective and ecologically compatible for the total removal of toxic and biorecalcitrant organic pollutants as well as inorganic pollutants from wastewaters. To overcome these problems, electrochemical methods have been widely studied, currently developed, and successfully applied to the treatment of synthetic solutions (Abdessamad et al. 2013a; Dbira et al. 2015; Guelfi et al. 2016; Li et al. 2015). Those methods were also successfully used for the treatment of real wastewater (Abdessamad et al. 2015; Daghrir et al. 2014; Fernandes et al. 2014; Garcia-Segura et al. 2015; Monica et al. 1980; Wang et al. 2015). Anodic oxidation and cathodic reduction processes are considered non-selective with high removal of most pollutants (organic and inorganic) up to their complete degradation (Brinzila et al. 2012; Fatima et al. 2007; Garcia-Segura et al. 2015; Li et al. 2010; Panizza and Cerisola 2009; Pomati et al. 2008; Shao-ting et al. 2007; Ventura et al. 2008). Among several used electrode materials, Boron-Doped Diamond (BDD) anode is considered more powerful compared to others electrodes like O2-Dimensionally Stable Anodes (DSA), Ti/PbO2, and Ti/TiRu-SnO2 (Zhou et al. 2011; Panizza and Martinez-Huitle 2013). It is especially efficient for recalcitrant organic pollutants due to the important production of hydroxyl radicals weakly adsorbed on the electrode surface and their strong unselective oxidative power (E° (V/SHE) = 2.80). BDD electrodes are also considered as the most powerful for simultaneous oxidation and reduction of pollutants (Abdessamad et al. 2013a, 2013b, 2015; Daghrir et al. 2014; Fernandes et al. 2014; Ghazouani et al. 2015; Li et al. 2015; Panizza and Cerisola 2006; Zhu et al. 2009). These properties were related to their electrochemical properties such as wide working potential window, low and stable voltametric background current, high corrosion stability even in strongly acidic media, and high overpotential for oxygen and hydrogen evolution in aqueous electrolytes (Hupert et al. 2003). The last property improves the ability of BDD to electro-generate hydroxyl radicals by direct electro-oxidation, in addition to other strong oxidants such as active chlorine (Cl2, HClO, and ClO−) (Akrout and Bousselmi 2012), peroxodisulphate (\( {\mathrm{S}}_2{\mathrm{O}}_8^{2\hbox{-} } \)) (Akrout and Bousselmi 2012), and peroxophosphate (\( {\mathrm{P}}_2{\mathrm{O}}_8^{4\hbox{-} } \)) (Canizares et al. 2009). Mineralization of organic molecule can be reached, leading to CO2 and inorganic ions as final products according to the following reactions:

$$ BDD+{H}_2O\to BDD\left(H{O}^{\bullet}\right)+{H}^{+}+{e}^{-} $$
(1)
$$ \mathrm{Organics}+BDD\left(H{O}^{\bullet}\right)\to BDD+C{O}_2+{H}^{+}+\mathrm{inorganics}+{e}^{-} $$
(2)

The indirect electro-oxidation by means of chloride salts using BDD as anode material was also mentioned in many studies (Akrout and Bousselmi, 2012; Huang et al. 2012). In fact, the direct oxidation of chloride ions on BDD anode leads to the electro-generation of active chlorine species (Cl2, HClO/ClO−) according to reactions (3, 4, 5, 6, 7, 8) which facilitate the oxidation of organic matters and ammonia molecule (Akrout and Bousselmi, 2012; Huang et al. 2012).

$$ 2C{l}^{-}\to C{l}_2+2{e}^{-} $$
(3)
$$ C{l}_2+{H}_2O\to HOCl+{H}^{+}+C{l}^{-} $$
(4)
$$ HOCl\to {H}^{+}+OC{l}^{-} $$
(5)
$$ HClO+\mathrm{organic}\to C{O}_2+{H}_2O+C{l}^{-}+{H}^{+} $$
(6)
$$ 2N{H}_3+2OC{l}^{-}\to {N}_2+2HCl+2{H}_2O+2{e}^{-} $$
(7)
$$ N{H}_3+4OC{l}^{-}\to N{O}_3^{-}+{H}_2O+{H}^{+}+4C{l}^{-} $$
(8)

Furthermore, researchers have focused on nitrate removal on BDD cathode as well as the oxidation of electro-generated by-products and the effect of salts on removal rates and mechanisms. From previous studies (Ghazouani et al. 2015; Xing et al. 2011; Zhu et al. 2009), the typical mechanisms taking place on BDD anode/cathode are given in the following reactions:

$$ N{O}_3^{-}+{H}_2O+2{e}^{-}\to N{O}_2^{-}+2O{H}^{-} $$
(9)
$$ N{O}_3^{-}+6{H}_2O+8{e}^{-}\to N{H}_3+9O{H}^{-} $$
(10)
$$ N{O}_3^{-}+3{H}_2O+5{e}^{-}\to \frac{1}{2}{N}_2+6O{H}^{-} $$
(11)
$$ BDD+O{H}^{-}\to BDD\left(H{O}^{\bullet}\right)+{e}^{-} $$
(12)
$$ 3H{O}^{\bullet }+N{H}_3\to \frac{1}{2}{N}_2+3{H}_2O+3{e}^{-} $$
(13)

The nitrite electro-generated is an intermediary by-product, and it can react to give ammonia, nitrogen gaseous, and also nitrate according to the reactions below (Ghazouani et al. 2015; Xing et al. 2011; Zhu et al. 2009):

$$ N{O}_2^{-}+5{H}_2O+6{e}^{-}\to N{H}_3+7O{H}^{-} $$
(14)
$$ 2N{O}_2^{-}+4{H}_2O+6{e}^{-}\to {N}_2+8O{H}^{-} $$
(15)
$$ N{O}_2^{-}+2O{H}^{-}\to N{O}_3^{-}+{H}_2O+2{e}^{-} $$
(16)

When pH values are between 7 and 11.5, equilibrium between ammonia and ammonium can be established as follows (Couto et al. 2016):

$$ N{H}_4^{+}+O{H}^{-}\to N{H}_3+{H}_2O $$
(17)

At acidic pH, the reduction of nitrate to ammonium is possible (Couto et al. 2016):

$$ N{O}_3^{-}+{H}_2O+6{H}^{+}+3{e}^{-}\to N{H}_4^{+}+4O{H}^{-} $$
(18)

At acidic medium and in the presence of chloride ions, the indirect oxidation of ammonium by HOCl could be considered according to reactions below (Fernandes et al. 2014):

$$ \frac{2}{3}N{H}_4^{+}+ HOCl\to \frac{1}{3}{N}_2+{H}_2O+\frac{5}{3}{H}^{+}+C{l}^{-} $$
(19)
$$ N{H}_4^{+}+4 HOCl\to N{O}_3^{-}+{H}_2O+6{H}^{+}+4C{l}^{-} $$
(20)

However, most previous studies on electrochemical oxidation focused on treating synthetic polluted wastewaters by adding salts as supporting electrolyte (Abdessmad et al., 2013b; Akrout and Bousselmi 2012; Lacasa et al. 2012). Nevertheless, only a few studies have been attempted to treat raw wastewater (e.g., landfill leachate (Fernandes et al. 2014), olive oil wastewaters (Panizza and Cerisola 2006), coking wastewater (Zhu et al. 2009), domestic sewage, (Monica et al., 1980), tannery wastewater (Szpyrkowicz et al. 2005) and secondary effluent (e.g., textile effluents (Abdessamad et al. 2015), MWW (Garcia-Segura et al. 2015), and reverse osmosis concentrate (Bagastyo et al. 2012; Zhou et al. 2011). Generally, raw wastewater had a good conductivity in fact, it contained high amount of salts especially chloride ions. Therefore, there is no need to supply additional electrolytes to ensure the conductivity of the medium.

This paper investigated the simultaneous removal of nitrogen compounds and organic matter (OM) using BDD as anode/cathode materials of four real effluents with different level of pollution. MWW and the receiving natural system (Wadi El Bey) were considered for this study as well as human urine (HU) and slaughterhouse wastewater (SHWW) in order to understand mechanisms involved in the reduction of nitrate and the oxidation of ammonia and organic carbon compounds.

Experimental

Wastewaters

The city of Grombalia is located at Nabeul prefecture (North East of Tunisia) with an estimated population of 18,663 (National Institute of Statistics, Tunisian Statistics 2014). MWWTP of Nabeul is using biological process (activated sludge–extended aeration) after pretreatment. Several industrial units (dairy unit, textile industry, slaughterhouse industry, agri-food, etc.,) are connected to the sewage system. However, pretreatment of the wastewater is undertaken in the industrial unit before discharge in the sewage. The secondary treated MWW is discharged in small river “Wadi El Bey”. Its watershed is approximately 475 km2, and it receives 60 % of the treated wastewaters of five cities and several industrial units as well as agriculture discharge which damage the environment and water quality of the river (Fiche descriptive sur les zones humides Ramsar (FDR) 2007; Khadhar et al. 2013). The secondary treated MWW, discharged into environment and loaded with OM, nitrogen and phosphorus, exceeds the assimilative capacity of water bodies causing eutrophication and proliferation of aquatic weeds.

SHWW of industrial unit was selected as an example of industry discarded in Wadi El Bey. The SHWW contained mainly organic solid wastes in addition to nutrients generated during meat processing and wastewater from washing at various steps of the process. Different sampling sites were presented in Fig. 1.

Fig. 1
figure 1

Sampling plan

Full size image

Sampling and effluent characterization

The wastewater samples were collected from the outlet of the MWW, the SHWW after decantation step, and in the Wadi after discharge. All samples were collected and preserved at 4 °C. The pHs of the collected wastewaters are presented in Table 1, and all treatment were monitored using the indicated parameters. Chemical analysis of the four wastewaters was performed using standard protocols (Rodier et al. 2009). The oxidation of OM was deduced by the COD removal of the treated effluent, determined by open reflux, dichromate titrimetric method as described in standard methods (Rodier et al. 2009). Nitrate, nitrite, chloride, phosphate, and sulfate anion concentrations were determined by the ion chromatography using a Metrohm Ion Chromatography System fitted with an anion-exchange column (AS4A-SC, 150 mm × 4 mm) and coupled with a conductivity detector under the control of an IC NET software. Total Kjeldhal Nitrogen (TKN) was analyzed by the Kjeldhal method (VELP UDK). Ammonium concentration was monitored by the Nessler method using UV–Visible spectroscopy (Thermo Spectronic model) at wavelength of 420 nm. The conductivity was measured through a conductive meter (Consort C561). A pH meter (HENNA instruments HI 2210) was employed to follow up the pH in the treated solution. The initial physicochemical characteristics of the four sampled wastewaters are given in Table 1; human urine sample was included for more investigation.

Table 1 Physicochemical characteristics of the considered real effluents
Full size table

All effluents were complex matrixes additionally contained several organic, nitrogen, and phosphorus pollutants with relatively high conductivity between 2.6 and 5.2 mS cm−1.

Electrochemical cell

An electrochemical cell contains bipolar BDD film supported on a Silicon substrate (Si/BDD, manufactured by the Adamant Technologies, Switzerland) was used. It was composed by three equal compartments auto-polarized, with an active surface area of 70 cm2 (each one) and an electrode gap of 1 mm (Fig. 2). A feed volume of 1 L recirculated through the electrolytic cell by peristaltic pump (Invertek Drives Optidrive, Tunisia) working in recycling mode (Fig. 2) with flow rate fixed at 194.4 L h−1 and the temperature at 25 ± 5 °C. A constant current density value (j = 35.7 mA cm−2) was applied by a direct-current power supply (Adamant Technologies).

Fig. 2
figure 2

Experimental setup: (1) back of solution, (2) pump, (3) electrochemical cell, and (4) power supply

Full size image

Analytical procedures

Average current efficiency (ACE) and energy consumption (EC)

The ACE of each effluent oxidation is calculated using the COD values (in g L−1), using the following relationship (Saez et al. 2007):

$$ \mathrm{ACE}\left(\%\right)=\frac{\left({\mathrm{COD}}_0\hbox{-} {\mathrm{COD}}_t\right)\times F\times {V}_{\mathrm{s}}}{8 It}\times 100 $$
(21)

where COD0 and CODt are the initial COD and the obtained COD after a certain time t, respectively, F is the Faraday constant, V s is the solution volume (L), 8 is the oxygen equivalent mass (g eq−1), I is the applied current (A), and t is the electrolysis time (s).

The EC expression per unit COD mass (kWh (g COD)−1) at a time t, is calculated as following (Xing et al. 2011):

$$ EC\left(kWh\ {\left(\mathrm{g}\ \mathrm{COD}\right)}^{\hbox{-} 1}\right)=\frac{IEt}{\Delta \mathrm{COD}\times {V}_{\mathrm{s}}} $$
(22)

where I is the applied current (A), E is the average cell voltage (V), t is the electrolysis time (h), ∆COD is the decay in COD (mg L−1), and V s is the solution volume (L).

Phytotoxicity study

Toxicity was evaluated before and after treatment by means of the germination test of lettuce seeds (Lactuca sativa). Twenty seeds were placed in Petri dishes and covered with filter papers. In each dishes, 5 mL of the considered wastewater solution was added. After that, Petri dishes were subsequently incubated in the dark room at 26–29 °C for 5 days. Distilled water was used as a control. All experiments, including the controls, were done in duplicate. The Germination Index (GI) was calculated by counting the number of germinated seeds and the length of the root mean observed in each sample compared to the control (Komilisa et al., 2005). Final results were determined using formula below and expressed as a percentage of the control sample results:

$$ GI\left(\%\right)=\frac{\mathrm{number}\ \mathrm{of}\ \mathrm{grown}\ \mathrm{seeds}\ \mathrm{in}\ \mathrm{sample}}{\mathrm{number}\ \mathrm{of}\ \mathrm{grown}\ \mathrm{seeds}\ \mathrm{in}\ \mathrm{control}}\times \frac{\mathrm{average}\ \mathrm{sum}\ \mathrm{of}\ \mathrm{root}\ \mathrm{lenghts}\ \mathrm{in}\ \mathrm{sample}}{\mathrm{average}\ \mathrm{sum}\ \mathrm{of}\ \mathrm{root}\ \mathrm{lenghts}\ \mathrm{in}\ \mathrm{control}}\times 100 $$
(23)

A seed was considered grown when its root length exceeded 5 mm. Lengths of root less than 5 mm were considered equal to 0, and seed was not considered germinated. The average sum of root lengths comprised the sum of the lengths of all the germinated seeds in a Petri dish divided by their total number. An effluent is considered nontoxic if the GI exceeds 50 % (Komilisa et al., 2005).

Results and discussions

Nitrogen compounds evolution and organic matter removal

The performance of BDD used as cathode/anode material on the degradation of organic pollutants and simultaneous removal of nitrogen compounds from four real wastewaters was investigated. MWW, as shown Table 1, is characterized initially by a relatively high amount of TKN (427.49 mg L−1). Ammonium/ammonia represents the quarter of TKN amount, in contrast of nitrate that was at low level (8.18 mg-N L−1). Figure 3 presents the time course of the nitrate, ammonium/ammonia, nitrite concentrations, as well as the COD removal rate and the Cl− variation obtained during the experiments performed with MWW effluent.

Fig. 3
figure 3

Performance of the Si/BDD electrode for a N-NO3 − (square), N-NH4 +/NH3 (circle), N-NO2 − (triangle), inorganic N (star) evolution (mg L−1), initial (dished pink line) and final (dished green line) TKN (mg L−1), and pH (dished red line) evolution. b COD removal (circle), inorganic N removal (triangle), and chloride evolution (mg L−1) (square) in MWW effluent

Full size image

As shown in Fig. 3a, a decrease of the initial \( \mathrm{N}-{NH}_4^{+}/{NH}_3 \) was observed during the first hour of treatment to reach a stable value until the second hour of experiment. This decrease was due to the oxidation of \( \mathrm{N}-{NH}_4^{+}/{NH}_3 \) on BDD anode into nitrate and nitrogen gaseous by means of hydroxyl radicals and hypochlorous species according to reactions (7, 8 and 13). The stability may be attributed to the equality of production rate and oxidation rate of ammonium/ammonia. In fact, ammonium/ammonia was formed by reduction of nitrate on BDD cathode (reaction 10) and also by oxidation of nitrogenous organic matter on BDD anode. Thus, electro-generated ammonium/ammonia was oxidized and formed simultaneously with an equivalent rate. Moreover, a slow raise on the nitrate concentration was obtained to note a maximum after 1 h. Formed nitrate was also reduced on BDD cathode providing nitrogen gas and ammonium/ammonia as by-product, which justify the slightly increase recorded for \( \mathrm{N}-{NH}_4^{+}/{NH}_3 \) concentration and the decrease of nitrate concentration during the last hour of treatment. Similar behavior was recorded by Wang et al. (2015) by using stainless steel (SS) as cathode material and BDD as anode material.

Simultaneously, BDD anode led as well to the OM oxidation. Therefore, for an initial COD load of 920 mg O2 L−1 (MWW effluent), a removal rate about 91 % was obtained after 3 h of treatment to reach 80 mg O2 L−1. Figure 3b shows that COD removal was fast during the first 1.30 h of treatment and reached 80 %. Whereas, during the last 1.30 h, the removal rate was decreased and the oxidation became slow. The COD degradation was fitting the pseudo first order with a constant kinetic value equal to 8.64 × 10−1 h−1 (Table 2) which was reported by Huang et al. (2012).

Table 2 Kinetic constant value of COD removal
Full size table

The removal of inorganic nitrogen compounds (nitrate, nitrite, and ammonium/ammonia) was likewise rapid and similar to COD removal during the first 30 min of treatment. Afterwards, the removal rate became stable until the end of treatment time. By the end of treatment, inorganic nitrogen matter was mainly presented as ammonium/ammonia. Thus, the oxidation of organic nitrogen matter in BDD anode led to the electro-generation ammonium/ammonia and nitrate which were simultaneously and rapidly degraded on electrode surface. Thereby, an important TKN removal (67 %) was recorded by the end of treatment process. By examining the Cl− evolution (Fig. 3b), the behavior was similar to inorganic nitrogen evolution and especially to ammonium/ammonia one. Same variation was obtained for COD removal. The decrease of the COD removal rate and ammonium/ammonia rate could be attributed to the drop of the oxidants chloride species in treated solution. However, the removal of ammonium/ammonia and OM took place through direct oxidation by means of HO• and an indirect oxidation route by HOCl/OCl− formed by oxidation of chloride on BDD anode (Akrout and Bousselmi 2012; Garcia-Segura et al. 2015; Huang et al. 2012; Katsoni et al. 2014). Thus, efficiency of these oxidants was underlined in previous studies (Akrout and Bousselmi 2012) and their production during the anodic oxidation on BDD anode was confirmed.

The treated MWW led to a final COD value of 80 mg O2 L−1 after 3 h of treatment which is in agreement with the Tunisian discharge standard in the natural environment (COD = 90 mg L−1 (NT 106.02 1989)). Whereas, TKN value was higher than the discharge limit into the natural environment according to the Tunisian Legislation standards (\( {\mathrm{NO}}_3^{\hbox{-} } \) = 90 mg L−1 (NT 106.02 1989)).

The HU is the main contributor to nitrogen pollution in the MWW. Concentrations of \( \mathrm{N}-{NH}_4^{+}/{NH}_3 \), COD, and Cl− in HU were, respectively 5, 8, 5 times higher than those in MWW. These high values are expected to emphasize oxidation/reduction behavior and bring more insights on the simultaneous degradation of nitrogen and organic pollutions. The degradation behavior of HU on BDD electrode was presented in Fig. 4. Nitrogen compounds, pH variation, as well as COD removal, and chlorides evolution are followed upon the treatment time.

Fig. 4
figure 4

Performance of the Si/BDD electrode for a N-NO3 − (square), N-NH4 +/NH3 (circle), N-NO2 − (triangle), inorganic N (star) evolution (mg L−1), initial (dished pink line) and final (dished green line) TKN (mg L−1), and pH (dished red line) evolution. b COD removal (circle) and chloride evolution (mg L−1) (square) in HU effluent

Full size image

The behavior was slightly different from MWW effluent case. Figure 4a shows that nitrate and ammonium (acidic pHs were observed) existed in treated solution with high amount from the beginning (\( \mathrm{N}\hbox{-} {\mathrm{NO}}_3^{\hbox{-} } \) = 6.18 mg L−1 and \( {\mathrm{N}\hbox{-} NH}_4^{+} \) = 508.21 mg L−1). In fact, the direct and indirect oxidation due to the high amount of chloride ions (2474 mg L−1) in the treated medium enhance the oxidation of OM and led to the mineralization and formation of nitrogen compounds (ammonium and nitrate). Moreover, electro-generated amount of nitrate and ammonium exceeded the reduced quantity of nitrate and that removed of ammonium. Several competitive reactions can be processed at the same time on active surface of BDD anode as well on BDD cathode. Therefore, a decrease of pH by one unit (from 5.85 to 4.73) was recorded during the first 30 min of treatment accompanied by an increase of the amount of ammonium. This result can be contributed to the mineralization of existent OM by means of active species produced on BDD anode. Moreover, this oxidation is faster than that of ammonium, thus the electro-generated amount of ammonium was higher than the oxidized one. The increase of pH can be also attributed to the oxidation of urea (CH4N2O) and creatinine (C4H7N3O) (Fig. 4b), which are the principal components of urine (Dbira et al. 2015; Li et al. 2015; Yaroshenko et al. 2015), according to reaction (2). Therefore, the reaction of mineralization contributed to the formation of H+ and \( {NH}_4^{+} \) as final products which acidify the medium. Results (Fig. 6) proved the total mineralization which was confirmed by high TKN removal (85 %). Moreover, Fig. 4a shows that the organic nitrogen removal was about 100 % with production of \( {NH}_4^{+} \) (increase around 1/4 of initial amount) and \( {\mathrm{NO}}_3^{\hbox{-} } \) (increase more than three times) during this mineralization.

Afterwards, a slight increase of produced nitrate and ammonium was observed accompanied with a raise of pH. Moreover, COD removal (Fig. 4b) was slightly rapid during the first hour reaching around 50 % of removal rate with a kinetic constant value equal to 4.56 × 10−1 h−1, two times faster than the removal that was obtained with MWW. During the last 30 min of treatment, a drop of pH from 6.33 to 3.43 was accompanied by a prompt elimination of ammonium by contrast of nitrate that was increased. Same pH evolution was obtained by Li et al. (2015) when treated artificial area by using BDD as anode material. Therefore, the removal efficiency of ammonium was high under acidic conditions due to the easy oxidation of this element and its greater sensitivity to acidic pH (Huang et al. 2012; Li et al. 2015). Moreover, indirect oxidation of ammonium/ammonia mediated with active chlorine species was expected to be faster in the acidic media because of the higher standard reduction potential of Cl2(aq) (E° (V/SHE) = 1.36) and HClO (E° (V/SHE) = 1.49) than ClO− (E° (V/SHE) = 0.89) (Garcia-Segura et al. 2015). As in the case of ammonium oxidation, the decrease of pH (case of HU sample) led to the increase of OM degradation rate which is in accordance with Garcia-Segura et al. (2015)).

Especially, we focused on the degradation of nitrogen pollution and COD at lower levels (four and three times lower, respectively, compared to MWW, Table 1) and always in a natural water matrix like the one of “Wadi el Bey” where the municipal as well the industrial wastewaters were discharged after treatment.

As shown in Fig. 5a, a decrease of initial nitrate amount due to the nitrate reduction on BDD cathode was observed during the first 5 min accompanied by an increase of pH. A raise of nitrate content was detected until 1 h of treatment owing to the degradation of OM and ammonium/ammonia. Afterwards, the reduction behavior of nitrate on BDD cathode started again and the amount declined to obtain 0.30 mg L−1 at the end of electrolysis step. Simultaneously, a rapid drop was recorded for \( \mathrm{N}-{NH}_4^{+}/{NH}_3 \) to reach 0 mg L−1 from the 1.30 h of treatment, but a slight increase was obtained at the last 30 min. A trace of nitrite was also formed and reached a first maximum at 30 min and a second maximum at 120 min. Results underlined also the key role of chlorides, existing at high level, on the oxidation of ammonium/ammonia and OM (Fig. 5b).

Fig. 5
figure 5

Performance of the Si/BDD electrode for a N-NO3 − (square), N-NH4 +/NH3 (circle), N-NO2 − (triangle), inorganic N (star) evolution (mg L−1), initial (dished pink line) and final (dished green line) TKN (mg L−1), and pH (dished red line) evolution. b COD removal (circle), inorganic N removal (triangle), and chloride evolution (mg L−1) in Wadi sample

Full size image

Despite the complete removal of the detected amount of nitrate, nitrite, and ammonia after 3 h of treatment, a small TKN removal (10 %) was observed (Fig. 6). Obtained results showed the impact of discharges on surface water quality in Wadi El Bey. Moreover, the Wadi represented the receiving environment for different discharges collecting many types of wastewaters from five cities (MWW), several industrial units (textile, agri-food), as well as agriculture pollution. All discharged wastewaters contributed to bring many refractory and recalcitrant pollutants that were persistent and not easily degradable. Therefore, a low organic nitrogen removal was recorded by the end of the treatment, While, the inorganic nitrogen removal was around 100 %.

Fig. 6
figure 6

Performance of the Si/BDD electrode for TKN removal (%) in four studied effluents

Full size image

As shown in Fig. 7, same behavior was obtained for the studied industrial wastewater (SHWW of El Mazraa). A single noticed difference was that SHWW contained initially high amount of nitrite (53.5 mgL−1), by contrast of other studied effluents (Table 1). Initial nitrite content was removed rapidly to obtain only 14.5 mg L−1 after 30 min to become traces at the end of treatment step (Fig. 7a). A high TKN removal rate (64 %) was also recorded (Fig. 6), and the amount of TKN compounds removed was more than that electro-generated. This indicated that an important amount was reduced into nitrogen gaseous according to reactions (11, 13, and 15).

Fig. 7
figure 7

Performance of the Si/BDD electrode for a N-NO3 − (square), N-NH4 +/NH3 (circle), N-NO2 − (triangle), inorganic N (star) evolution (mg L−1), initial (dished pink line) and final (dished green line) TKN (mg L−1), and pH (dished red line) evolution. b COD removal (circle), inorganic N removal (triangle), and chloride evolution (square) in SHWW effluent

Full size image

By examining the nitrogen compounds and the COD profiles (Fig. 7) during the anodic oxidation and cathodic reduction runs for BDD electrode, the inorganic nitrogen compounds removal occurred at a slightly faster rate than OM removal.

The oxidation and reduction reactions occurred simultaneously affecting the removal rates of different pollutants due to competitive reactions at the electrode interface. Thus, the oxidation of OM contributes to the electro-generation of nitrogen compounds (ammonium, nitrate, …). However, at the same time, ammonium and nitrate removal will be replaced immediately by other electro-generated ones affecting the total removal behavior. This result was not in accordance with those obtained by Cossu et al. (1998) study using a Ti/Pt anode, Fernandes et al. (2014) using Ti/Pt/PbO2, Ti/Pt/SnO2-Sb2O4, and Si/BDD anodes, and Li et al. (2015) using BDD and IrO2 anodes. According to these researches, the removal rate of nitrogen compounds was lower than that of OM at the initial stage of the electro-oxidation. Then, nitrogen compounds were substantially removed in the subsequent electrochemical oxidation stage when indirect oxidation became prevalent.

The final treated wastewater cannot be reused or discharged into natural environment because the final COD (360 mg O2 L−1) and ammonia values (55.3 mg-N L−1) were higher than the limit fixed by the Tunisia Legislation standards (NT 106.02 1989). Such values could be obtained by increasing the electrolysis time.

By examining the kinetic constant values (Table 2) of COD removal for four studied wastewaters and pH evolution with treatment time, we could deduce that the COD removal was fast at acidic pH. Moreover, the increase of pH solution was accompanied by the decrease of kinetic constant of degradation, which indicated the higher treatment efficiency in terms of COD removal at acidic pH than in neutral and alkaline medium, this is due to the formation of higher amounts of oxidant (BDD(HO•)) at acidic pH (Garcia-Segura et al. 2015). Furthermore, electro-generation of other competitive oxidants at higher pH with less oxidant potential such as \( {\mathrm{O}}_2^{\hbox{-} \bullet } \) which decreases the oxidation ability of anode material. Also, the indirect oxidation with active chlorine species was expected to be faster at acidic pH because of the higher standard oxidation potential of Cl2(aq) and HClO than ClO− (Garcia-Segura et al. 2015). In addition, chlorine generation reaction is preferred in the acid condition than parasite reaction such as O2 evolution (Scialdone et al. 2009)).

Similar effect was found for ammonium/ammonia oxidation; the removal efficiency under acid conditions (HU effluent) was better than under alkali (MWW, Wadi, and SHWW). This result was in accordance with Huang et al. studies (2012). In fact, the effluent nature affects well the treatment behavior. Thus, the pH variation of each effluent was rather different. Previous studies conducted by Huang et al. (2012), Wang et al. (2010), and Samet et al. (2006) noted that a better oxidation efficiency was obtained in acidic condition for both organic and ammonia oxidation, while Panizza and Cerisola (2004) mentioned that the alkaline condition was more conducive to the treatment of synthetic tannery waste water. However, initial pH had no obvious effect on geosmin degradation in Xue et al. studies (2011).

Energy evaluation: ACE and EC per amount of COD oxidized

The performance of electrochemical treatment by means of anodic oxidation and cathodic reduction cannot overshadow the high operating costs because of the high EC required by this process. To evaluate its efficiency, an evaluation in terms of current efficiency and EC per amount of COD oxidized for different studied wastewaters was done.

Figure 8 highlights the current efficiency and EC contribution issued by BDD electrodes for the treatment of different studied wastewaters. The EC values showed that the increase of COD content involved an important decrease in the EC. As instance, the EC value was decreased from 0.65 to 0.03 kWh (g COD)−1 when the COD was increased from 320 to 7300 mg O2 L−1. Furthermore, the ACE dropped clearly with the increase of COD content.

Fig. 8
figure 8

a Average current efficiency (ACE) and b energy consumption (EC) per amount of COD oxidized for the studied wastewaters

Full size image

For HU, the ACE had a high value (252 %), larger than 100 %. This result was due to the low adsorption at the BDD electrodes. Indeed, hydroxyl radicals are very weakly adsorbed and consequently they are very reactive toward oxidation of organics (Abdessamad et al. 2013a; Gherardini et al. 2001). In this case BDD is behaving like an ideal anode for electrocombustion of organics.

Phytotoxicity of studied wastewaters: germination test

Figure 9 shows the evolution of the GI before and after treatment of the five studied wastewaters on BDD electrode.

Fig. 9
figure 9

Evolution of the GI before and after treatment of four studied wastewaters on BDD electrode, natural pH, j = 35.7 mA cm−2, control: distilled water

Full size image

All studied wastewaters are considered toxic in the raw state with a GI equal to 29.7, 0, 30.6, and 22.8 % (less than 50 %) for the MWW, HU, Wadi, and SHWW, respectively. After oxidation, the treated wastewater was less toxic in the case of MWW and SHWW, which could be explained by the fact that the GI was equal to 58.8 and 72.2 %, respectively, while, the toxicity was persistent in the case of the Wadi and HU with a GI equal of 31.8 and 0 %, respectively. Despite the lower level of contained pollutants and an organic degradation more than 90 % after 3 h of treatment in the case of the Wadi, the toxicity persisted. This fact might be caused by the existence of persistent toxic pollutants until the end of electrolysis step such as refractory organic by-products. Tiquia et al. (1996) proved the negative effect of ammonium on the relative seed germination of many species of seeds which confirmed our results for the case of HU. The seeds were also more sensitive to the high amount of salt and heavy metals having a toxic effect on the seed germination (case of the Wadi and HU) (Tiquia et al. 1996; Wtmdram et al. 1996).

Conclusion

The viability of electrochemical oxidation/reduction with a BDD electrode was assessed to remove a considerable rate of TKN and organic contaminants from the studied real effluents during a short time (3 h). OM oxidation was accomplished to obtain 91, 74, 87, and 84 % of COD removal from MWW, HU, Wadi, and SHWW, respectively, after 3 h of treatment. The initial amount of chloride in HU and the decrease of pH solution to value 3 promoted the oxidation of OM and ammonium.

In the current study, we found that the nature of the effluent had a noticeable influence on the method efficiency and the electrical energy consumption required to treat a given wastewater. Thus, the lowest EC (0.03 kWh (g COD)−1) was obtained for the more concentrated effluents (7300 mg O2 L−1 for the HU).

It is clear that the anodic oxidation and simultaneously cathodic reduction on BDD electrode were an effective and ecological treatment processes. From an economical point of view, electrochemical oxidation process is fairly expensive in comparison with conventional treatment that is why this method should not be considered as a single treatment for a real waste but as a finishing stage in a combined process or as an auxiliary unit able to work in emergency situations for large-scale wastewater treatment.