Results

this result, we conclude, higher removal efficiencies can be achieved by adjustment of operating condition such as HRT.

Fig. 33 Continuous removals of EE2 under different HRTs

Fig. 34 Continuous treatment of EE2 by glassy carbon electrode reactor

5.4.1.2 Electric current

Figure 35demonstrates the removal performance of 1 µM EE2 by a granular Pt/Ti electrolytic reactor at the applied electric (0.1-10 mA). Observed results demonstrate that EE2 was effectively removed at 1mA within 2 days of operation. When the current changed (day 5), the effluent concentration of EE2 was increased up to 0.32 µM and

0 20 40 60 80 100 120

0 2 4 6 8 10

Concentration(µg/L)

Time (d)

EE2 Influent EE2 Effluent

HRT= 180 min

HRT= 90 min

HRT= 180 min

0.0 0.2 0.4 0.6 0.8 1.0 1.2

0 2 4 6 8 10 12 14 16

Concentration(µM)

Time (d)

EE2 Influent EE2 Effluent

HRT= 180 min HRT= 120 min HRT= 60 min

kept around 0.19 to 0.33 µM during this condition (0.1 mA). An increase of electric current 10mA results in a drop of effluent concentration to 0.12-0.16 µM, corresponding to 84%-88% removal efficiency. It was reported in the batch experiment (chapter 4) that removals of estrogens were governed by the mass transfer rate limiting step and little affected by the electric current. On the other hand, Nagata et al. (2005) reported the optimal electric current was 0.5 mA in batch treatment of EDCs for the similar reactor using granular Pt/Ti electrodes. This result suggests that an electric current from 0.5 mA to 1 mA could be optimal condition for the treatment of high load influent, which may contain complex chemicals in the treatment matrix.

Fig. 35 Continuous treatment of EE2 by Pt/Ti electrode reactor (1 µM= 296 µg/L)

5.4.2 Electrode material

Removals of 100 µg/L of E1 and E2 by Pt/Ti and GCE reactors were evaluated.

Figure 36 illustrates the removals of E1 and E2 at 1mA in the course of 18 days. The removal of E2 was achieved about 97%-99% meanwhile that of E1 was reached 90%-99% from day 2 of operation. At unsteady state and steady state, the observed data was in good agreement with the mathematical model (Eq. 10). The result suggests that E1 and E2 could be effectively removal by Pt/Ti electrolytic reactor. However, the cost for

0.0 0.2 0.4 0.6 0.8 1.0 1.2

0 3 6 9 12 15 18 21 24 27

Concentration(µM)

Time (d)

EE2 Influent EE2 Effluent

1 mA 0.1 mA 10 mA

fabrication of Pt/Ti is relatively high. Therefore, appropriate electrodes with high removal performance and low cost would be investigated.

Fig. 36 Continuous removals of E1 and E2 by Pt/Ti electrode reactor

(Experimental condition: Influent 100 µg/L)

Figure 37 demonstrates the removal of 100 µg/L E1 and E2 at 1 mA in 10 mM Na2SO4. It is clearly to see that 87% of E1 and 94% of E2 were removed within two days of operation. Noticeably, removal of E2 achieved 97.8% then kept stable for almost of the operation time. Simultaneously, the observed influent concentration remaining around 100 µg/L (initial concentration) suggests that the removal of E1 and E2 were occurred by electrochemical process. In contrast, the removals of E1 were not stable, varied in the range of 70%-90%. This may be due to the complex conversion among estrogens during oxidation process. However, the concentration of E1 in the effluent increased gradually from around day 21 and approached to influent concentration. The phenomena may be explained due to electrode passivation through electro-polymerization process. The same phenomena were reported elsewhere (Ngundi et al., 2003; Kuramitz et al., 2004). In this research, electrode regeneration was conducted on day 28 by injection of 2.7 mg/L of dissolved O3 followed by electrolytic reduction for 30 min. The result showed that removal of E1 was similar to its original state.

0 20 40 60 80 100 120

0 5 10 15 20

Concentraion(µg/L)

Time (d)

E1 Effluent E2 Effluent Eq (10)

Fig. 37 Continuous treatments of E1 and E2 at 1 mA

(Experimental condition: HRT: 3 hr; initial concentration: 100 µg/L)

5.4.3 Influence of co-substances

Synthetic estrogens containing 1 µg/L of E1, E2 or EE2 in 10 mM Na2SO4 were prepared referring to the real effluent of sewage treatment plants (STPs). The control experiment was evaluated for separated of each estrogen without presence of co-substances (Fig. 38). The removal performance of EE2 in presence of 1 mg/L-TOC is shown inFigure 39.

Figure 38illustrates experimental results of continuous treatments of 1 µg/L of E1, E2 or EE2, respectively. Figure 38 indicates an efficient removal of E1 and E2 at 0.5 mA and EE2 at 1 mA, corresponding to 2-5×10^-4 mA/cm², respectively. As shown, effluent concentration of E1, E2, and EE2 declined sharply within 3 days of operation, it reaches steady state in the course of 20 days. Referred to the current density in CV result, this result indicates that E1, E2 and EE2 were removed through electro-polymerization, not mineralization.

At unsteady and steady states, observed data were in good agreement with the mathematical model (solid line) regardless of different applied electric currents indicating that removals of E1, E2 and EE2 were governed by mass transfer rate. E1, E2, and EE2 were stably removed and achieved about 98% removal efficiency. In fact,

0 20 40 60 80 100 120

0 5 10 15 20 25 30 35 40

Concentration(µg/L)

Time (d)

E1 Influent E2 Influent E1 Effluent E2 Effluent

electrolytic reactor consisting of granular electrodes has enlarged the working surface of anode which increases the net mass transfer rates from bulk solution to electrode surface. In addition,Eq. (10) was applicable under trace EDCs conditions (i.e. 0.01 to 150 µg/L) in former study (Sakakibara et al., 2010); therefore similar performance or removal efficiencies are expected for the treatment of actual effluents of STPs.

Moreover, this study shows the possibility to treat trace estrogens at low current density (2 to 5×10^-4mA/cm²).

Fig. 38 Continuous treatments of E1, E2 and EE2 by GCE reactor

(Influent: 1 µg/L, HRT: 180 min, electrolyte: 10 mM Na2SO4; E1, E2 (0.5mA); EE2 (1mA)).

Figure 39demonstrates the removal performance of the electrolytic reactor without and with presence of humic acids. Result shows a stable removal of 1 µg/L EE2 in the courses of 52 days operation. After 52 days continuous treatment, 1 mg/L-TOC of co-substances were added in the feeding solution. It shows that the same removal performance was observed in 30 days even co-substances are existed in the reactor. In this study, electric current density was very low, thus, OH production might be negligible. Electricity was consumed for direct oxidation of EE2 on surface of electrode.

From day 85 of the treatment, observed residue of EE2 in the effluent start rising until day 90. We supposed that glassy carbon may be passivated during the treatment

0.0 0.2 0.4 0.6 0.8 1.0 1.2

0 5 10 15 20 25

Concentration(µg/L)

Time (days)

E1 Influent E2 Influent EE2 Influent Eq (10) E1 Effluent E2 Effluent EE2 Effluent

process. In batch treatment, we found a significant change of chemical elements on surface of Pt/Ti and GCEs before and after 4 hrs of treatment. Therefore, electrode morphology was investigated in this study.

Fig. 3939 Continuous treatment of EE2

(Influent: 1 µg/L, HRT: 180 min, electrolyte: 10 mM Na₂SOSO₄; E1, E2 (0.5 mA); EE2 (1 mA)).

5.

5.4.4 Morphology of GCE

The GCEs were taken after 93 days continuously treatment of 1µg/L of EE2.

Electrode morphology was analyzed by a scanning electron microscope (SEM).

Visualization analyses of the surface were carried out at 1000x and 10,000x of magnification. Figures 40 A-D demonstrate a change surface of electrode before and after 90 days of treatment with consideration of electro-polymerization effects. Figure 40Bshows a thin layer of substances which is not appeared in the new electrode (Fig.

40A). This thin polymerized layer may attribute to the ineffective removal of EE2 after 85 of operation. In addition, several small holes were observed after long time treatment, which may indicate a corrosion of electrode, occurred if there was any variation of electric currents. It was observed that removal performance of GCE is almost similar to that of Pt/Ti reactor; a life cycle assessment of GCE should be further investigated. Figure 40D illustrated a closed magnification of the polymerized products. As shown, the polymeric product was formed unequally with some area up to 1 µm.

0.0 0.2 0.4 0.6 0.8 1.0 1.2

0 2020 4040 6060 8080 100

@?46?EC2E:@? T8%

Time (day)

Influent (no HA) Effluent (no HA) Influent (with HA) Effluent (with HA)

A) New GCE B) GCE after 90 days treatment (M: 1,000x)

C) New GCE B) GCE after 90 days treatment (M: 10,000x)

Fig. 40 Glassy carbon surfaces before and after continuous treatment

5.4.5 Continuous treatment of EE2

EE2 is one of the strongest estrogenic compounds among estrogens, thus, elimination of its toxicity is great of concern. Continuous treatment of synthetic solution contained 100 µg/L in 10 mM Na2SO4was experimentally conducted using granular glassy carbon electrolytic reactor in the course of 135 days. The treatment performance is shown inFigure 41. Observed stable influent and effluent concentrations indicate the removal of EE2 was made by the electrochemical oxidation process. As shown in Figure 41, the removal of EE2 experienced three stages of which the first state shows a stable removal of EE2, and then second stage occurs with the increasing of effluent

1 µm

concentration approaching influent, the third stage dedicates to the steady state of no EE2 removal.

Upon the point of no removal, effluent concentrations were kept stable for 12 days.

This suggested no removal of EE2 was achieved once the electrode completely passivated. Electrode regenerations (Reg 1 and Reg 2) were conducted in presence of 1.8 mg/L and 3.0 mg/L dissolved ozone combined with change of polarization for 120 min and 30 min. It was reported that ozone decomposes of organic pollutants by whether its molecule as a strong oxidant or OH radicals through electrochemical reduction process. After regenerations, EE2 was removed effectively for a period of 50 days. This result has indicated two important points: 1) application of ozone for regeneration of electrolytic reactor for continuous treatment of estrogenic compounds through electropolymerized could be possible, and 2) periodically regeneration could be predicted with enhanced technique to quantify the mass deposition from a known influent concentration of pollutants.

In this study, the dissolved ozone was produced using a commercial ozone generator with supply of pure oxygen gas. It could be feasible to equip this process with a hybrid electrolytic reactor with specific anodes such as SnO2 for direct generating ozone gas by an appropriate electric current density.

Fig. 41 Continuous removal of EE2 by GCE reactor and electrode regeneration 0

20 40 60 80 100 120

0 20 40 60 80 100 120

Concentration(µg/L)

Time (d)

EE2 Influent (100µg/L) EE2 Effluent

Reg 1 Reg 2

5.

5.4.6 Energy consumption

Figure 4242 demonstrates the removal efficiency against energy consumption of electrochemical treatment and other processes. Electrochemical mineralization and ultrasonic irradiation achieved high removal efficiency (88%-92%) with energy consumption in the range of 1010- 200 kWh/m³. In this study, removal efficiencies of E1, E2 and EE2 were about 98%, but the energy consumption wawas 4 - 5 orders of magnitude lower. In recent years, different types of electrochemical treatments have gained increasing interest for removing a wide variety of pollutants in water and wastewaters such as refractory organic substances, heavy metals, nitrogen compounds with simple equipment and easy operation (Anglada et al., 2009; Chen, 2004; Mousavi et al., 2012).

Therefore, it is considered that electrochemical processes are suited for small-scale scatted wastewater plants.

Fig. 4242 Removal efficiency versus energy consumption of different processes.

5.