- Email: [email protected]
Continuous electrochemical treatment of phenolic wastewater in a tubular reactor
Water Research 37 (2003) 1505–1514
Continuous electrochemical treatment of phenolic wastewater in a tubular reactor . Bahadır K. Korbahti, Abdurrahman Tanyolac-* Faculty of Engineering, Chemical Engineering Department, Hacettepe University, Beytepe 06532, Ankara, Turkey Received 18 July 2001; received in revised form 12 March 2002; accepted 10 October 2002
Abstract The electrochemical treatment of phenolic wastewater in a continuous tubular reactor, constructed from a stainless steel tube with a cylindrical carbon anode at the centre, was investigated in this study, being first in literature. The effects of residence time on phenol removal was studied at 251C, 120 g l1 electrolyte concentration for 450 and 3100 mg l1 phenol feed concentrations with 61.4 and 54.7 mA cm2 current densities, respectively. The change in phenol concentration and pH of the reaction medium was monitored in every run and GC/MS analyses were performed to determine the fate of intermediate products formed during the electrochemical reaction in a specified batch run. During the electrolysis mono, di- and tri-substituted chlorinated phenol products were initially formed and consumed along with phenol thereafter mainly by polymerization mechanism. For 10 and 20 min of residence time phenol removal was 56% and 78%, respectively, with 450 mg l1 phenol feed concentration and above 40 min of residence time all phenol was consumed within the column. For 1, 1.5, 2 and 3 h of residence time, phenol removal achieved was 42%, 71%, 81% and 98%, respectively, at 3100 mg l1 phenol feed concentration. It is noteworthy that more than 95% of the initial phenol was converted into a non-passivating polymer without hazardous end products in a comparatively fast and energy-efficient process, being a safe treatment. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: Electrochemical oxidation; Wastewater treatment; Phenol; Tubular reactor; Polymerization
1. Introduction Phenols are considered as primarily pollutant components in wastewater due to their high toxicity, high oxygen demand (theoretically, 2.4 mg O2 mg1 phenol) and low biodegradability [1]. Over 2 mg l1 phenol concentration is toxic to fish and concentrations between 10 and 100 mg l1 result in death of aquatic life within 96 h [1,2]. Activated carbon adsorption and solvent extraction processes have been used for phenol recovery, and biological and chemical oxidation treatment methods have been conventionally preferred for phenol destruc*Corresponding author. Tel: +90-312-2977404; fax: +90312-2992124. E-mail address: [email protected] (A. Tanyolac-).
tion [3]. Biological treatment is suitable for some phenolic wastewaters up to a maximum concentration of 50 mg l1 [4], after which inhibition of the biological treatment process takes over. On the other hand, electrochemical oxidation is becoming a new alternative for wastewater treatment and replacing the traditional processes, because many industrial processes produce toxic wastewaters, which are not easily biodegradable and requiring costly physical or physico-chemical pretreatment [5]. Many researchers had investigated the electrochemical oxidation of various types of wastewater including the phenolic wastewater [2,3,5–25]. During the electrochemical destruction of phenol, polymer products deposit on the electrodes as tars, forming a passivating film on the electrode [2,10,11,15,26–28]. The film causing electrode passivation interferes with the supply of fresh
0043-1354/03/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 3 - 1 3 5 4 ( 0 2 ) 0 0 5 2 3 - 7
1506
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 .
reactant, and the removal of products from the reaction zone, meanwhile decreasing the available potential to drive the reaction [15]. There are different configurations of electrochemical reactors employed for the electrochemical oxidation of phenol. Smith de Sucre and Watkinson [3] used lead dioxide packed-bed anodes for phenol oxidation. The authors reported that all the phenol in solution could be readily oxidized but complete organic carbon removal was more difficult. Chettiar and Watkinson [11] used an electrochemical system adapted from the previous work of Smith de Sucre and Watkinson [3] using a coated graphite plate with an electrodeposited PbO2 anode. According to their results the phenolics were largely destroyed, however, organic degradation products such as quinone and hydroquinone remained in the solution. Sharifian and Kirk [21] used a packed-bed reactor of PbO2 pellets with recirculating anolyte in order to follow the phenol loss and benzoquinone, maleic acid and CO2 formation. The authors stated that benzoquinone, an intermediate product, formation was favoured by high current and pH and by low phenol concentration, temperature, and dissolved oxygen at relatively short electrolysis times. Gattrell and Kirk [2] used a flow-by electrochemical cell with a peristaltic pump to feed the phenol solution through the anode in an upward direction. The authors reported that increasing cell voltage increased the amount of oxidized products along with increasing electrode corrosion and decreasing current efficiency. It was also concluded that higher phenol concentrations (over 1880 mg l1) showed a shift in product distribution to favour less oxidized, mostly insoluble products while elevated temperatures (about 501C and higher) showed a marked ability to reduce electrode passivation and increase the phenol oxidation . et al. [18] and Stucki et al. [22] used a benchrate. Kotz scale general-purpose reactor from ElectroCell AB equipped with up to five SnO2 anodes. The authors concluded that the oxidation efficiency was independent of the pH of the water and the oxidation of a wide range of organic compounds proceeded with an efficiency about 5 times higher than platinum anodes without any corrosion on cathode. In their work, oxidation efficiency for the process was between 30 and 40% with a power consumption of 40–50 kWh kg1 COD removed. Comninellis and Pulgarin [12] used a two-compartment cell with an anode made of SnO2 coated titanium and a cathode made of a platinum spiral enclosed in a porous porcelain pot. Stirring was provided with a magnetic bar in the reactor, there were only small amounts of aromatic intermediates formed at the anode, and aliphatic acids were rapidly oxidized. Comninellis and Nerini [13] used an electrochemical system adapted from the previous work of Comninellis and Pulgarin [12] using Ti/SnO2 and Ti/IrO2 anodes in the presence of NaCl to oxidize the phenol anodically. They noted that
the presence of NaCl catalysed the oxidation reaction of phenol at Ti/IrO2 anode, however did not affect Ti/SnO2 anode proving that electrochemical oxidation reaction was depending on both electrolyte type and anodic material. Boudenne et al. [9] used a flow-by electrochemical cell with a stack of 21 Pt/Ti grids consisting of a cooling tub and a pump. The authors concluded that electrochemical degradation of phenol with carbon black is much quicker than biological degradation and other electrochemical treatments using different catalysis such as TiO2 for photo oxidation. Lin et al. [20] used a batch system with four pieces of cast iron cathode and anode for phenol removal from saline wastewater. The authors reported that with a proper current input, the COD removal of the electrochemical treatment could be considered improved in comparison to that in biological treatment. The application of the tubular electrochemical reactor has been made only recently in literature. Israilides et al. [17] and Vlyssides et al. [25] used a recirculation reactor with a tubular electrolytic cell consisting of stainless steel cathode and titanium alloy anode placed at the centre of the reactor to treat olive oil wastewater and vinasse from beet molasses treatment, respectively. The authors operate the batch, laboratory-scale pilot plant system at 15 V DC and 100 A. Israilides et al. [17] reported that after 10 h of electrolysis at 0.26 A cm2 current density, total COD, TOC, VSS were reduced by 93%, 80.4% and 98.7%, respectively, with the energy consumption of 12.3 kWh kg1 COD removed, being an unfeasible process. In another work of the tubular reactor, Vlyssides et al. [40] used a similar reactor to that of Israilides et al. [17] for the oxidation of textile dye wastewater. The authors operated the lab-scale pilot plant recirculated reactor continuously at 20 V DC and 50 A. They concluded that the system could be used for effective textile dye wastewater oxidation or a feasible detoxification and colour removal with a mean energy consumption of 21 kWh kg1 COD removed. During the electrochemical conversion of phenol, an electrolyte is added to the medium to increase conductivity and in return current density. Various electrolytes such as Na2SO4, H2SO4 and their mixtures were used in literature [2,3,11,12,18,21] but very few investigators [13] made use of NaCl mainly due to toxic intermediates produced such as mono, di and trisubstituted chlorophenols when utilized [29,30]. When electrolysis of aqueous NaCl solution is attempted, Cl2 gas is discharged on the anode above 2.1872 V potential [31,32]. Following the discharge of Cl2 gas [14,20,25,31,32] HOCl formation occurs with hydrolysis reaction [14,20,33], which is consumed by the ionization reaction with the formation of OCl [14,33]. Meanwhile, HOCl is replenished in the medium with the electrochemical reaction of NaCl. Hypochlorous acid, HOCl, is a strong oxidant, which oxidizes the phenol in solution
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 .
while OCl is also consumed [13,17,25]. This represents the indirect oxidation of phenol, however direct oxidation occurs on the electrode [20], most likely with a different mechanism. On the cathode H2O molecules are reduced and H2 gas along with hydroxyl ions are produced. In this study, application of the continuous tubular reactor for phenol removal through electrochemical oxidation was realized for the first time in literature. In the work, the effect of residence time on electrochemical treatment of synthetic phenolic wastewater was investigated at low and high phenol feed concentrations. Other operational parameters were directly used in the runs with their optimum values determined previously [34]. Column outlet pH and phenol concentration and the energy consumed were monitored continuously, and periodic GC/MS analysis of the reaction mixture in a specified batch run was carried out to determine the fate of toxic intermediate products.
2. Materials and methods 2.1. Chemicals and materials Phenol and NaCl were obtained from Merck (99.5% pure) and double distilled water was used for the preparation of synthetic wastewater. All solvents used for the analysis were also obtained in extra pure
1507
condition from Merck. Carbon electrode (Meteor, Germany) used in the runs (cylindrical, D ¼ 1:32 cm) was originally designed and manufactured for DC electric motors to endure high voltage and current loads. Stainless steel 6R35 (Sandvik, Sweden) was used as cathode material (cylindrical, OD=8.89 cm and wall thickness=2 mm) with a weight percentage chemical composition of 0.045 C, 0.40 Si, 1.18 Mn, 0.026 P, 0.001 S, 17.45 Cr, 10.14 Ni, 0.46 Ti, and the rest Fe.
2.2. Experimental set-up A continuous electrochemical tubular reactor was designed in our laboratory with a net working volume of 1934 cm3 [35] (Fig. 1). The 35 cm tall reactor (OD=8.89 cm) was constructed from stainless steel (Sandvik, Sweden) with a heating/cooling coil around. The carbon electrode was used as the anode material and placed at the centre of the reactor. A Cole Parmer model peristaltic Masterflexs pump was used to pump the phenol feed in electrolyte solution from a 20-l reservoir. The reaction temperature was controlled with circulating water recycled from a temperature-controlled water bath (New Brunswick, G-86) and monitored with glass thermometers immersed in the exit and inlet of the column. The current was applied by a constant voltage/ current-controlled DC power source (NETES NPS1810D).
Fig. 1. The schematic view of continuous electrochemical tubular reactor ((1) carbon electrode, (2) plexiglas screws, (3) plexiglas reactor cover, (4) discharge line, (5) heating/cooling coil, (6) reactor support, (7) reservoir, (8) feed pump, (9) feed line, (10) connections, (11) DC power source).
1508
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 .
2.3. Experimental procedure Synthetic phenolic wastewater with the electrolyte was prepared in the wastewater reservoir at specified concentration and was fed to the column to fill it completely. An initial sample of 10 ml was taken from the reservoir to determine the initial phenol concentration prior to reaction. The reaction started with the application of the specified voltage and current and, the recycling water for temperature control was pumped through the reactor coil while feed stream was pumped to the column continuously. At appropriate time intervals, samples of 10 ml were taken from the reactor outlet and analysed to determine the intermediate properties of the reaction medium during electrochemical conversion. DC power source was turned off and the reaction was terminated after steady state reached.
column was used for the separation and oven temperature was programmed as constant 501C for 1 min, and then increased to 2501C with 201C min1 rate and finally 10 min at 2501C. Automatic injection system was kept at 250–2801C while mass detector temperature was held at 2801C. For identification of the products, standard samples and Wiley library search were used. Then, internal calibration was applied using a standard (dichlorophenol) with response factor 0.99 to calculate corresponding concentrations of compounds based on the area under the curve. For polymer, 4–5 peaks were detected with their respective mass number values and individual weight concentrations were calculated. The results were presented as percent of the initial phenol present in the wastewater.
3. Results and discussions 2.4. Reaction conditions The phenolic wastewater was used with 450 and 3100 mg l1 initial phenol concentrations in the runs at 251C reaction temperature with 120 g l1 electrolyte concentration. For 450 and 3100 mg l1 initial phenol concentrations, the applied current density was 61.4 and 54.7 mA cm2, respectively. The values of temperature, electrolyte concentration and current density applied were determined to be optimum through batch runs for maximum phenol removal rate in a previous work [34]. 2.5. Analysis The change in phenol concentration of the reaction medium was determined by means of a Cecil 1100 HPLC system. The system was equipped with a variable wavelength monitor (CE 11220) operating at 254 nm and a Cecil HPLC pump system (CE 1100). Chromatograms were obtained using a Hichrom-S5OS1 column (Chromosorb). Only one mobile phase, running at ambient temperature, was eluted and 20 ml sample was injected. Methanol/water (40/60, v/v) was used as the mobile phase and the flow rate was adjusted to 1 ml min1 under 28 MPa pressure. Before the analysis, the mobile phase was filtered and sonicated in order to remove dissolved gases. GC/MS analyses were performed with a 5972 MSD connected to a 5890 HPs GC system in order to determine the intermediate products formed during the electrochemical reaction in a specified batch run. For the analysis, 5 ml liquid sample was taken from the reactor at appropriate time intervals. The pH of the sample was adjusted to 3 with HNO3 and aqueous phase was saturated with NaCl. Then the sample was extracted for 20 min with 5 ml of diethyl ether, organic phase was separated in a separation funnel, and 1 ml of ether phase was injected into GC/MS system. A DB-1 capillary
Continuous tubular reactor experiments were carried out for the determination of the effects of the residence time on phenol conversion at optimized conditions determined previously through batch runs [34]. Phenol feed concentration was selected at low and high (450 and 3100 mg l1) concentrations with 61.4 and 54.7 mA cm2 applied current density for each residence time to visualize the change in phenol removal performance. Although both feed streams contained the same amount of electrolyte, the conductivity of the feed with lower phenol concentration was observed higher than that of higher phenol feed concentration. This is basically due to the cloud-like polymer formation around the anode at higher phenol concentrations in an unmixed flow system with low linear velocity (3oNRe o53), which diminished the current applied. The runs were carried out in doubles and average results were presented in figures. The data points for duplicate runs at any time were not different from each other more than 4%. The passivating polymeric formation on electrodes has not been detected in any run, which, in fact, was observed in literature studies. 3.1. The effect of residence time For 450 mg l1 initial phenol feed concentration, phenol concentration at the reactor outlet decreased with increasing residence time as shown in Fig. 2. At 10 and 20 min of residence time phenol removal was 56% and 78%, respectively. Above 40 min residence time, all phenol was eventually consumed in the column and 100% phenol removal was achieved. During the reactions, the colour of the solution at the outlet stream turned to light yellow from transparent then to dark brown due to increased residence time. At higher residence times, the formation of the suspended solid particles in reaction medium caused an opaque
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 .
1509
Phenol Outlet Concentration (mg l-1)
500 Residence Time
450
10 min 20 min 40 min 60 min 90 min
400 350 300 250 200 150 100 50 0 0
10
20
30
40
50
60
70
80
90 100 110 120
Elapsed Time (min)
Fig. 2. The change of outlet phenol concentration profiles by time in the range 10–90 min of residence time for 450 mg l1 phenol feed concentration (current density: 61.4 mA cm2; T: 251C; electrolyte concentration: 120 g l1).
14 12 10 8 pH
appearance and the colour of the solution at the reactor outlet was dark brown due to the formation of polymeric products. Easily separable solid polymeric materials were obtained at the outlet stream but the formation of polymers did not affect the performance of the column, mainly cell current density. Fig. 3 shows the pH change at the exit of the column for 450 mg l1 phenol concentration, where pH was not controlled. As a general trend, at the beginning reactor outlet pH rised up to pH 11, then began to drop gradually in proportion with the residence time in the tubular reactor, i.e. more residence time caused more pH drop after peak pH. Initial pH rise could be attributed to the production of hydroxyl ions on the cathode in the presence of NaCl and pH drop was caused by gradual consumption of hydroxyl anions as well as production of hydrogen cations in dissociation reactions of HOCl and OCl along with the formation reaction of HOCl in the treated waste stream. Short residence time ended up with higher final pH and increase in residence time caused lowered final pH values. Nevertheless, residence time greater than 20 min also caused instable pH values due to unestablished steady-state conditions in the reactor within the time range of runs as well as non-homogeneous character of the reactor. Stable outlet phenol concentrations were established approximately after 80, 70, 40, 40 and 30 min of elapsed time for the residence time of 10, 20, 40, 60 and 90 min, respectively, and the corresponding energy requirement per hour at steady state was 1.144, 2.077, 2.679, 3.057 and 1.410 kWh mol1 consumed phenol, respectively, as shown in Table 1. At steady state, experimental number of electrons consumed per hour was calculated as 0.33, 0.34, 0.35, 0.26, and 0.33 mol for 10, 20, 40, 60 and 90 min of residence time, respectively, in Table 1. Although electron consumption rate values were almost the same, phenol consumption rate decreased in accord
6
Residence Time 10 min 20 min 40 min 60 min 90 min
4 2 0 0
20
40
60
80
100
120
Elapsed Time (min)
Fig. 3. The change of pH profiles by time at the reactor outlet in the range 10–90 min of residence time for 450 mg l1 phenol feed concentration (current density: 61.4 mA cm2; T: 251C; electrolyte concentration: 120 g l1).
with increased residence time at steady state. Also, specific energy requirement increased proportionally with residence time except for 90 min. At 3100 mg l1 phenol feed concentration, the continuous column reactor reached steady state after 5 h as shown in Fig. 4. At 1, 1.5, 2 and 3 h of residence time, phenol removal was achieved 42%, 71%, 81% and 98%, respectively. The increase in residence time increased the phenol removal percentage; however, polymeric products formed started to block the reactor discharge at higher residence times due to low volumetric flow rate. Due to high phenol concentration, excessive amount of foam was produced and the colour of the solution was turned to light yellow from transparent and then to dark brown at the outlet stream. The variation of pH in the exit stream of the column for 3100 mg l1 phenol feed concentration was depicted in Fig. 5. Similar trends of
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 .
1510
Table 1 The energy requirement and electron consumption rate for 450 mg l1 phenol feed concentration (current density: 61.4 mA cm2; T: 251C; electrolyte concentration: 120 g l1) Residence time (min)
Time required to reach steady state (min) 80 70 40 40 30
Phenol Outlet Concentration (mg l-1)
10 20 40 60 90
Phenol consumption rate at steady state (mmol h1)
Specific energy requirement at steady state (kWh mol1 phenol consumed)
Electron consumption rate at steady state (mol e h1)
31.1071 21.6639 13.8871 9.2581 6.1720
1.144 2.077 2.679 3.057 1.410
0.33 0.34 0.35 0.26 0.33
Phenol removal (%) at steady state 56 78 100 100 100
Residence Time
3000
1h 1.5 h 2h 3h
2500 2000 1500 1000 500 0 0
2
1
3 4 5 Elapsed Time (h)
6
7
8
Fig. 4. The change of outlet phenol concentration profiles by time in the range 1–3 h of residence time for 3100 mg l1 phenol feed concentration (current density: 54.7 mA cm2; T: 251C; electrolyte concentration: 120 g l1).
14 12
pH
10 8 6
Residence Time
1h
4
1.5 h 2h
2
3h
0 0
1
2
3
4
5
6
7
8
Elapsed Time (h)
Fig. 5. The change of pH profiles by time at the reactor outlet in the range 1–3 h of residence time for 3100 mg l1 phenol feed concentration (current density: 54.7 mA cm2; T: 251C; electrolyte concentration: 120 g l1).
Fig. 3 were also observed in this figure. Except for the residence time of 1 h, pH profiles first reached as high as pH 11.5 and then dropped with an unstable manner as
explained for 450 mg l1 phenol feed concentration. The steady-state energy requirement per hour was calculated as 4.345, 4.670, 5.704 and 6.379 kWh mol1 consumed phenol for 1, 1.5, 2 and 3 of residence time, respectively, as shown in Table 2. At steady state, experimental number of electrons consumed per hour was calculated as 0.24, 0.29, 0.26 and 0.28 mol for 1, 1.5, 2 and 3 h of residence time, respectively, in Table 2. In Table 2, electron consumption rate was again almost the same like in Table 1 and phenol consumption rate decreased proportionally to residence time after 1.5 h of residence time. Although there has not been any work done yet with electrochemical treatment of phenol in literature using tubular reactor to compare with our results, there are other works carried out in different systems for electrochemical oxidation of phenol. Smith de Sucre and Watkinson [3] achieved complete phenol removal within 90 and 120 min reaction time for 525 and 1100 mg l1 initial phenol concentrations, respectively, at 10 A current and 8.5 V cell voltage in aqueous solutions of Na2SO4 and H2SO4 or NaOH using a lead
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 .
1511
Table 2 The energy requirement and electron consumption rate for 3100 mg l1 phenol feed concentration (current density: 54.7 mA cm2; T: 251C; electrolyte concentration: 120 g l1) Residence time (h)
1 1.5 2 3
Time required to reach steady state (h) 5 5 5 5
Phenol consumption rate at steady state (mmol h1)
Specific energy requirement at steady state (kWh mol1 phenol consumed)
Electron consumption rate at steady state (mol e h1)
26.7882 30.1899 25.8315 20.8353
4.345 4.670 5.704 6.379
0.24 0.29 0.26 0.28
dioxide anode packed-bed reactor. In their work, the energy requirement changed from 4.8 to 2.4 kWh mol1 phenol removed as the initial phenol concentration increased from 525 to 1100 mgl1 in batch runs. Chettiar and Watkinson [11] achieved 95.8% phenol removal for 1070 mg l1 initial phenol concentration after 5 h at 10 A with an electrochemical system adapted from the previous work of Smith de Sucre and Watkinson [3]. Sharifian and Kirk [21] achieved almost 100% phenol removal within 6 h for 1316 mg l1 initial phenol concentration in 1 M H2SO4 at a cell current of 2 A at 251C and 1 ml s1 flow rate with a packed-bed reactor of PbO2 pellets. Nevertheless, after 10 h of reaction time, the intermediates of maleic acid and benzoquinone were still in the medium. Gattrell and Kirk [2] achieved about 60%, 60% and 55% phenol removal for 470, 940 and 1880 mg l1 initial phenol concentrations, respectively, at 1.7 V in 8 h of electrolysis time with a flow-by electrochemical cell. For all three concentrations, the rate of phenol removal slowed down after the first hour due to formation of greater amounts of intermediates. The authors also reported that phenol passivated the electrode surface during the first hour of operation due . to the polymer forming reaction. Kotz et al. [18] achieved complete phenol removal after 2 A h charge passed for 1000 mg l1 phenol concentration by anodic oxidation using a SnO2 anode at 30 mA cm2 current density. The authors reported that after 1 A h charge passed; the phenol disappeared while the complete removal of TOC took about 8 A h and calculated the energy requirement as 1.88 kWh mol1 of phenol removed. Comninellis and Pulgarin [12] achieved 100% TOC removal at 70 Ah dm3 with SnO2 anodes for 1974 mg l1 initial phenol concentration, for 701C and 50 mA cm2 current density with a two-compartment electrochemical cell. Again, Comninellis and Nerini [13] achieved complete phenol removal at 15 A h dm3 for 940 mg l1 initial phenol concentration in the presence of NaCl, for 501C, 25.3 g l1 NaCl concentration and 100 mA cm2 current density using an electrochemical system adapted from the previous work of Comninellis and Pulgarin [12] with Ti/SnO2 anodes to oxidize phenol
Phenol removal (%) at steady state 42 71 81 98
anodically. Lin et al. [20] achieved 70% COD removal within 60 min for 1000 mg l1 initial phenol concentration with 60 mg l1 H2O2 addition at 2.5 A current and 1% of salinity using a batch system with four pieces of cast iron cathode and anode. The authors also reported that the COD removal without H2O2 addition at the same conditions was 34%. Alvarez-Gallegos and Pletcher [36] and Harrington and Pletcher [37] investigated the electrochemical removal of low-level organics including phenol. Alvarez-Gallegos and Pletcher [36] calculated the energy reqirement for 0.33 mM phenol solution as 1.25 kWh m3 to reduce the COD to o10 ppm. Harrington and Pletcher [37] reported that the energy consumptions for the treatment of 1 m3 of the 0.5 mM phenol solution assuming a cell voltage of 5 V lied in the range 1.5–2.5 kWh m3 which was low compared to other electrochemical alternatives for the removal of organics from wastewater. The authors also calculated the energy consumption for the teatment of 3 mM phenol solution around 14 kWh m3. After reaching steady state, our continuous tubular reactor residence time values for almost complete phenol removal, 40 min and 3 h for 450 and 3100 mg l1 phenol feed concentrations, respectively, are less than those of literature for electrochemical treatment, and the treatment also requires less specific energy, 2.679 and 6.379 kWh mol1 phenol consumed for 450 (40 min) and 3100 mg l1 (3 h) phenol feed concentrations, respectively; denoting a feasible alternative reactor configuration. This advantage is partially due to the reactor type and also due to a different reaction mechanism, which favoured the fast formation of nonpassivating polymeric end products rather than slow total oxidation of phenol. 3.2. Determination of conversion products GC/MS chromatograms and mass spectrums were obtained in order to identify the intermediate products formed in the liquid phase during the electrochemical conversion of phenol. The data in Figs. 6 and 7 were collected in batch runs carried out at 32.9 mA cm2,
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 .
1512
251C reaction temperature and 120 g l1 electrolyte concentration for 496 and 3505 mg l1 initial phenol concentrations. In Figs. 6 and 7, concentrations of all compounds were expressed as percentage weight of initial phenol present in the waste stream. For each batch run the exit stream was completely recycled with the flow rate of 6000 ml min1 and fed to the inlet of the column. From periodic analysis of the medium taken at the exit, it was understood that mainly mono, di-and trisubstituted chlorinated phenol products were formed during the electrolysis and consumed thereafter by either destruction or polymerization, or both as seen in Figs. 6 and 7.
120
Compound phenol
100
2-chlorophenol
Compound % (w/w)
4-chlorophenol 2,4 –dichlorophenol
80
2,4,6-trichlorophenol polymeric products
60 40 20 0 0
20
40
60
80
100
120
Elapsed Time (min)
Fig. 6. The weight percentage of conversion products based on initial phenol at 496 mg l1 initial phenol concentration in the batch run (32.9 mA cm2 current density, 251C reaction temperature, 120 g l1 electrolyte concentration).
Diverse from literature [2,3,9–12,14,15,21,23] except the work of Comninellis and Nerini [13], only chlorinated phenol products formed during the runs with Cl2 gas discharging on the anode; and all the chlorinated phenol derivatives were completely consumed either by destruction or polymerization at 496 and 3505 mg l1 initial phenol concentrations after 120 min and 11 h, respectively. As Lund and Iverson [38] previously stated, the oxidation of organic compounds at graphite anode often results in other products when considered with the oxidation at platinum anode, but in our work we had only four different intermediates, which did not turn into successive intermediate compounds in contrast to other studies of literature [2,3,9–12,14,15,21,23]. Firstly, 4-chlorophenol and 2,4,6-trichlorophenol formed dominantly during the electrolysis and followed by 2-chlorophenol and then 2,4-dichlorophenol for both initial phenol concentrations of 496 and 3505 mg l1. In Fig. 6 for 496 mg l1 initial phenol concentration, the weight percentage of 2,4,6-trichlorophenol increased rapidly up to 40 min, although 4-chlorophenol started to decrease after 20 min. The weight percentage of polymeric compounds increased significantly after 40 min with the disappearance of mono, di and trichlorophenols. This result showed that the chlorinated hydrocarbons contributed to the formation of polymeric structure at a certain extent, while part of them might be oxidized. Finally, after 120 min, no intermediates were left in the reaction medium at all. Due to the contribution of chlorine into the polymer structure for 450 mgl1 phenol feed concentration, the final weight percentage of polymeric compounds passed over 100% [39].
100
Compound 90
phenol 2-chlorophenol
80
Compound % (w/w)
4-chlorophenol 70
2,4 dichlorophenol
60
polymeric products
2,4,6-trichlorophenol
50 40 30 20 10 0 0
2
4
6
8
10
12
Elapsed Time (h)
Fig. 7. The weight percentage of conversion products based on initial phenol at 3505 mg l1 initial phenol concentration in the batch run (32.9 mA cm2 current density, 251C reaction temperature, 120 g l1 electrolyte concentration).
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 .
Similar results were observed in Fig. 7 for 3505 mg l1 initial phenol concentration; the weight percentage of 2,4,6-trichlorophenol increased rapidly up to 4 h, but pchlorophenol started to decrease after 1.5 h. The concentration of polymeric compounds increased noticeably after 4 h with the decreasing of mono, di and trichlorophenols. Finally after 10 h, no intermediates were left in the reaction medium. Although chlorinated phenol derivatives are more toxic than phenol itself [29,30], in our work all intermediate chlorophenols were converted into polymeric compounds or oxidized eventually and did not appear in the reaction solution, assuring a safe and non-toxic treatment. The insoluble polymer formed is lighter than the treated water and therefore can be easily removed by means of density difference.
4. Conclusion The electrochemical treatment of phenolic wastewater in a continuous tubular reactor was investigated. The effect of residence time on phenol removal was studied for 450 and 3100 mg l1 phenol feed concentrations at 251C, 120 g l1 electrolyte concentration, 61.4 and 54.7 mA cm2 current densities, respectively. For 450 mg l1 phenol feed concentration at steady state, above 40 min of residence time all phenol was consumed within the column and 100% phenol removal was achieved. At 3100 mg l1 phenol feed concentration, 98% of phenol removal was realized only at 3 h of residence time at the steady state. At steady state, residence times of 40 min and 3 h for 450 and 3100 mg l1 phenol feed concentrations, respectively, are comparatively shorter than those of literature for electrochemical treatment for phenol removal, and the process also requires less specific energy, 2.679 and 6.379 kWh mol1 phenol consumed for 450 and 3100 mg l1 phenol feed concentrations, respectively. Analysis of the reaction products has shown that during the electrolysis mono, di and tri substituted chlorinated phenol products were formed along with polymeric compounds and all chlorinated phenols were consumed thereafter by polymerization and destruction. The results indicate that this continuous tubular reactor may be a feasible and safe alternative system for the electrochemical treatment of phenolic wastewater.
Acknowledgements This project was supported by Hacettepe University Research Fund with Grant no: 98.02.602.004.
1513
References [1] Lanouette KH. Treatment of phenolic wastes. Chem Eng 1977;84(22):99–106. [2] Gattrell M, Kirk DW. The electrochemical oxidation of aqueous phenol at a glassy carbon electrode. Can J Chem Eng 1990;68:997–1003. [3] Smith de Sucre V, Watkinson AP. Anodic oxidation of phenol for waste water treatment. Can J Chem Eng 1981;59:52–9. [4] S- eker S- , Beyenal H, Salih B, Tanyolac- A. Multi-substrate growth kinetics of Pseudomonas putida for phenol removal. Appl Microbiol Biotechnol 1997;47:610–4. [5] Pulgarin C, Adler N, P!eringer P, Comninellis C. Electrochemical detoxification of a 1,4-benzoquinone solution in wastewater treatment. Water Res 1994;28(4): 887–93. [6] Awad YM, Abuzaid NS. Electrochemical treatment of phenolic wastewater: efficiency, design considerations and economic evaluation. J Environ Sci Health A 1997;32(5):1393–414. [7] Bockris JO’M, Kim J. Electrochemical treatment of lowlevel nuclear wastes. J Appl Electrochem 1997;27:623–34. [8] Boscoletto AB, Gottardi F, Milan L, Pannocchia P, Tartari V, Tavan M, Amadelli R, Battisti A, Barbieri A, Patracchini D, Battaglin G. Electrochemical treatment of bisphenol-A containing wastewaters. J Appl Electrochem 1994;24:1052–8. [9] Boudenne JL, Cerclier O, Gal!ea J, Van der Vlist E. Electrochemical oxidation of aqueous phenol at a carbon black slurry electrode. Appl Catal A: Gen 1996;143: 185–202. * * J, [10] Canizares P, Dom!ınguez JA, Rodrigo MA, Villasenor Rodr!ıguez J. Effect of the current intensity in the electrochemical oxidation of aqueous phenol wastes at an activated carbon and steel anode. Ind Eng Chem Res 1999;38:3779–85. [11] Chettiar M, Watkinson AP. Anodic oxidation of phenolics found in coal conversion effluents. Can J Chem Eng 1983;61:568–74. [12] Comninellis C, Pulgarin C. Electrochemical oxidation of phenol for wastewater treatment using SnO2 anodes. J Appl Electrochem 1993;23:108–12. [13] Comninellis C, Nerini A. Anodic oxidation of phenol in the presence of NaCl for wastewater treatment. J Appl Electrochem 1995;25:23–8. [14] Do JS, Yeh WC. Paired electrooxidative degradation of phenol with in situ electrogenerated hydrogen peroxide and hypochlorite. J Appl Electrochem 1996;26:673–8. [15] Gattrell M, Kirk DW. A study of electrode passivation during aqueous phenol electrolysis. J Appl Electrochem 1993;140(6):1534–40. [16] Genders JD, Hartsough D. Electrochemical reduction of nitrates and nitrites in alkaline nuclear waste solutions. J Appl Electrochem 1996;26:1–9. [17] Israilides CJ, Vlyssides AG, Mourafetti VN, Karvouni G. Olive oil wastewater treatment with the use of an electrolysis system. Biores Technol 1997;61:163–70. . R, Stucki S, Carcer B. Electrochemical waste water [18] Kotz treatment using high overvoltage anodes; Part I: physical
1514
[19]
[20]
[21] [22]
[23]
[24]
[25]
[26] [27]
[28]
[29]
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 . and electrochemical properties of SnO2 anodes. J Appl Electrochem 1991;21:14–20. Lin SH, Chen ML. Textile wastewater treatment by enhanced electrochemical method and ion exchange. Environ Technol 1997;18:739–46. Lin SH, Shyu CT, Sun MC. Saline wastewater treatment by electrochemical method. Water Res 1998;32(4): 1059–66. Sharifian H, Kirk D W. Electrochemical oxidation of phenol. J Electrochem Soc 1986;133(5):921–4. . R, Carcer B, Suter W. Electrochemical Stucki S, Kotz waste water treatment using high overvoltage anodes; Part II: anode performance and applications. J Appl Electrochem 1991;21:99–104. Tahar BN, Savall A. Electrochemical degradation of phenol in aqueous solution on bismuth doped lead dioxide: a comparison of the activities of various electrode formulations. J Appl Electrochem 1999;29:277–83. Tennakoon CLK, Bhardwaj RC, Bockris JO’M. Electrochemical treatment of human wastes in a packed bed reactor. J Appl Electrochem 1996;26:18–29. Vlyssides AG, Israilides CJ, Loizidou M, Karvouni G, Mourafetti V. Electrochemical of vinasse from beet molasses. Water Sci Technol 1997;36(2-3):271–8. Hedenburg JF, Freiser H. Anodic voltammetry of phenols. Anal Chem 1953;25:1355–8. Koile RC, Johnson DC. Electrochemical removal of phenolic films from a platinum anode. Anal Chem 1979;51(6):741–4. Wang J, Jiang M, Lu F. Electrochemical quartz crystal microbalance investigation of surface fouling due to phenol oxidation. J Electroanal Chem 1998;444: 127–32. US Department of Health and Human Services, Toxicological profile for phenol. Sciences International, Inc., Research Triangle Institute, USA, 1998.
[30] US Department of Health and Human Services, Toxicological profile for chlorophenols. Sciences International, Inc., Research Triangle Institute, USA, 1999. [31] Hearst JE, Ifft JB. Contemporary chemistry. San Fransisco, CA: Freeman, 1976. [32] Mortimer CE. Chemistry. Belmont, CA: California Wadsworth Publishers, 1986. [33] Metcalf & Eddy, Inc. In: Tchobanoglous G, Burton FL, editors. Wastewater engineering. Singapore: McGraw-Hill, 1991. . [34] Korbahti BK, Salih B, Tanyolac- A. Electrochemical conversion of phenolic wastewater on carbon electrodes in the presence of NaCl. J Chem Technol Biotechnol 2002;77(1):70–6. . [35] Korbahti BK. Continuous electrochemical treatment of phenolic wastewater in a tubular reactor. MS dissertation, Hacettepe University, Ankara, Turkey, 1999. [36] Alvarez-Gallegos A, Pletcher D. The removal of low-level organics via hydrogen peroxide formed in a reticulated vitreous carbon cathode cell. Part 2: the removal of phenols and related compounds from aqueous effluents. Electrochim Acta 1999;44:2483–92. [37] Harrington T, Pletcher D. The removal of low levels of organics from aqueous solutions using Fe(II) and hydrogen peroxide formed in situ at gas diffusion electrodes. J Electrochem Soc 1999;146(8):2983–9. [38] Lund H, Iversen P, Baizer MM, editors. Organic electrochemistry. New York: Marcel Dekker, 1973. . [39] Zareie MH, Korbahti BK, Tanyolac- A. Non-passivating polymeric structures in electrochemical conversion of phenol in the presence of NaCl. J Hazardous Mater B 2001;87:199–212. [40] Vlyssides AG, Loizidou M, Korlis PK, Zorpas AA, Papaioannou D. Electrochemical oxidation of a textile dye wastewater using a Pt/Ti electrode. J Hazardous Mater B 1999;70:41–52.
Continuous electrochemical treatment of phenolic wastewater in a tubular reactor . Bahadır K. Korbahti, Abdurrahman Tanyolac-* Faculty of Engineering, Chemical Engineering Department, Hacettepe University, Beytepe 06532, Ankara, Turkey Received 18 July 2001; received in revised form 12 March 2002; accepted 10 October 2002
Abstract The electrochemical treatment of phenolic wastewater in a continuous tubular reactor, constructed from a stainless steel tube with a cylindrical carbon anode at the centre, was investigated in this study, being first in literature. The effects of residence time on phenol removal was studied at 251C, 120 g l1 electrolyte concentration for 450 and 3100 mg l1 phenol feed concentrations with 61.4 and 54.7 mA cm2 current densities, respectively. The change in phenol concentration and pH of the reaction medium was monitored in every run and GC/MS analyses were performed to determine the fate of intermediate products formed during the electrochemical reaction in a specified batch run. During the electrolysis mono, di- and tri-substituted chlorinated phenol products were initially formed and consumed along with phenol thereafter mainly by polymerization mechanism. For 10 and 20 min of residence time phenol removal was 56% and 78%, respectively, with 450 mg l1 phenol feed concentration and above 40 min of residence time all phenol was consumed within the column. For 1, 1.5, 2 and 3 h of residence time, phenol removal achieved was 42%, 71%, 81% and 98%, respectively, at 3100 mg l1 phenol feed concentration. It is noteworthy that more than 95% of the initial phenol was converted into a non-passivating polymer without hazardous end products in a comparatively fast and energy-efficient process, being a safe treatment. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: Electrochemical oxidation; Wastewater treatment; Phenol; Tubular reactor; Polymerization
1. Introduction Phenols are considered as primarily pollutant components in wastewater due to their high toxicity, high oxygen demand (theoretically, 2.4 mg O2 mg1 phenol) and low biodegradability [1]. Over 2 mg l1 phenol concentration is toxic to fish and concentrations between 10 and 100 mg l1 result in death of aquatic life within 96 h [1,2]. Activated carbon adsorption and solvent extraction processes have been used for phenol recovery, and biological and chemical oxidation treatment methods have been conventionally preferred for phenol destruc*Corresponding author. Tel: +90-312-2977404; fax: +90312-2992124. E-mail address: [email protected] (A. Tanyolac-).
tion [3]. Biological treatment is suitable for some phenolic wastewaters up to a maximum concentration of 50 mg l1 [4], after which inhibition of the biological treatment process takes over. On the other hand, electrochemical oxidation is becoming a new alternative for wastewater treatment and replacing the traditional processes, because many industrial processes produce toxic wastewaters, which are not easily biodegradable and requiring costly physical or physico-chemical pretreatment [5]. Many researchers had investigated the electrochemical oxidation of various types of wastewater including the phenolic wastewater [2,3,5–25]. During the electrochemical destruction of phenol, polymer products deposit on the electrodes as tars, forming a passivating film on the electrode [2,10,11,15,26–28]. The film causing electrode passivation interferes with the supply of fresh
0043-1354/03/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 3 - 1 3 5 4 ( 0 2 ) 0 0 5 2 3 - 7
1506
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 .
reactant, and the removal of products from the reaction zone, meanwhile decreasing the available potential to drive the reaction [15]. There are different configurations of electrochemical reactors employed for the electrochemical oxidation of phenol. Smith de Sucre and Watkinson [3] used lead dioxide packed-bed anodes for phenol oxidation. The authors reported that all the phenol in solution could be readily oxidized but complete organic carbon removal was more difficult. Chettiar and Watkinson [11] used an electrochemical system adapted from the previous work of Smith de Sucre and Watkinson [3] using a coated graphite plate with an electrodeposited PbO2 anode. According to their results the phenolics were largely destroyed, however, organic degradation products such as quinone and hydroquinone remained in the solution. Sharifian and Kirk [21] used a packed-bed reactor of PbO2 pellets with recirculating anolyte in order to follow the phenol loss and benzoquinone, maleic acid and CO2 formation. The authors stated that benzoquinone, an intermediate product, formation was favoured by high current and pH and by low phenol concentration, temperature, and dissolved oxygen at relatively short electrolysis times. Gattrell and Kirk [2] used a flow-by electrochemical cell with a peristaltic pump to feed the phenol solution through the anode in an upward direction. The authors reported that increasing cell voltage increased the amount of oxidized products along with increasing electrode corrosion and decreasing current efficiency. It was also concluded that higher phenol concentrations (over 1880 mg l1) showed a shift in product distribution to favour less oxidized, mostly insoluble products while elevated temperatures (about 501C and higher) showed a marked ability to reduce electrode passivation and increase the phenol oxidation . et al. [18] and Stucki et al. [22] used a benchrate. Kotz scale general-purpose reactor from ElectroCell AB equipped with up to five SnO2 anodes. The authors concluded that the oxidation efficiency was independent of the pH of the water and the oxidation of a wide range of organic compounds proceeded with an efficiency about 5 times higher than platinum anodes without any corrosion on cathode. In their work, oxidation efficiency for the process was between 30 and 40% with a power consumption of 40–50 kWh kg1 COD removed. Comninellis and Pulgarin [12] used a two-compartment cell with an anode made of SnO2 coated titanium and a cathode made of a platinum spiral enclosed in a porous porcelain pot. Stirring was provided with a magnetic bar in the reactor, there were only small amounts of aromatic intermediates formed at the anode, and aliphatic acids were rapidly oxidized. Comninellis and Nerini [13] used an electrochemical system adapted from the previous work of Comninellis and Pulgarin [12] using Ti/SnO2 and Ti/IrO2 anodes in the presence of NaCl to oxidize the phenol anodically. They noted that
the presence of NaCl catalysed the oxidation reaction of phenol at Ti/IrO2 anode, however did not affect Ti/SnO2 anode proving that electrochemical oxidation reaction was depending on both electrolyte type and anodic material. Boudenne et al. [9] used a flow-by electrochemical cell with a stack of 21 Pt/Ti grids consisting of a cooling tub and a pump. The authors concluded that electrochemical degradation of phenol with carbon black is much quicker than biological degradation and other electrochemical treatments using different catalysis such as TiO2 for photo oxidation. Lin et al. [20] used a batch system with four pieces of cast iron cathode and anode for phenol removal from saline wastewater. The authors reported that with a proper current input, the COD removal of the electrochemical treatment could be considered improved in comparison to that in biological treatment. The application of the tubular electrochemical reactor has been made only recently in literature. Israilides et al. [17] and Vlyssides et al. [25] used a recirculation reactor with a tubular electrolytic cell consisting of stainless steel cathode and titanium alloy anode placed at the centre of the reactor to treat olive oil wastewater and vinasse from beet molasses treatment, respectively. The authors operate the batch, laboratory-scale pilot plant system at 15 V DC and 100 A. Israilides et al. [17] reported that after 10 h of electrolysis at 0.26 A cm2 current density, total COD, TOC, VSS were reduced by 93%, 80.4% and 98.7%, respectively, with the energy consumption of 12.3 kWh kg1 COD removed, being an unfeasible process. In another work of the tubular reactor, Vlyssides et al. [40] used a similar reactor to that of Israilides et al. [17] for the oxidation of textile dye wastewater. The authors operated the lab-scale pilot plant recirculated reactor continuously at 20 V DC and 50 A. They concluded that the system could be used for effective textile dye wastewater oxidation or a feasible detoxification and colour removal with a mean energy consumption of 21 kWh kg1 COD removed. During the electrochemical conversion of phenol, an electrolyte is added to the medium to increase conductivity and in return current density. Various electrolytes such as Na2SO4, H2SO4 and their mixtures were used in literature [2,3,11,12,18,21] but very few investigators [13] made use of NaCl mainly due to toxic intermediates produced such as mono, di and trisubstituted chlorophenols when utilized [29,30]. When electrolysis of aqueous NaCl solution is attempted, Cl2 gas is discharged on the anode above 2.1872 V potential [31,32]. Following the discharge of Cl2 gas [14,20,25,31,32] HOCl formation occurs with hydrolysis reaction [14,20,33], which is consumed by the ionization reaction with the formation of OCl [14,33]. Meanwhile, HOCl is replenished in the medium with the electrochemical reaction of NaCl. Hypochlorous acid, HOCl, is a strong oxidant, which oxidizes the phenol in solution
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 .
while OCl is also consumed [13,17,25]. This represents the indirect oxidation of phenol, however direct oxidation occurs on the electrode [20], most likely with a different mechanism. On the cathode H2O molecules are reduced and H2 gas along with hydroxyl ions are produced. In this study, application of the continuous tubular reactor for phenol removal through electrochemical oxidation was realized for the first time in literature. In the work, the effect of residence time on electrochemical treatment of synthetic phenolic wastewater was investigated at low and high phenol feed concentrations. Other operational parameters were directly used in the runs with their optimum values determined previously [34]. Column outlet pH and phenol concentration and the energy consumed were monitored continuously, and periodic GC/MS analysis of the reaction mixture in a specified batch run was carried out to determine the fate of toxic intermediate products.
2. Materials and methods 2.1. Chemicals and materials Phenol and NaCl were obtained from Merck (99.5% pure) and double distilled water was used for the preparation of synthetic wastewater. All solvents used for the analysis were also obtained in extra pure
1507
condition from Merck. Carbon electrode (Meteor, Germany) used in the runs (cylindrical, D ¼ 1:32 cm) was originally designed and manufactured for DC electric motors to endure high voltage and current loads. Stainless steel 6R35 (Sandvik, Sweden) was used as cathode material (cylindrical, OD=8.89 cm and wall thickness=2 mm) with a weight percentage chemical composition of 0.045 C, 0.40 Si, 1.18 Mn, 0.026 P, 0.001 S, 17.45 Cr, 10.14 Ni, 0.46 Ti, and the rest Fe.
2.2. Experimental set-up A continuous electrochemical tubular reactor was designed in our laboratory with a net working volume of 1934 cm3 [35] (Fig. 1). The 35 cm tall reactor (OD=8.89 cm) was constructed from stainless steel (Sandvik, Sweden) with a heating/cooling coil around. The carbon electrode was used as the anode material and placed at the centre of the reactor. A Cole Parmer model peristaltic Masterflexs pump was used to pump the phenol feed in electrolyte solution from a 20-l reservoir. The reaction temperature was controlled with circulating water recycled from a temperature-controlled water bath (New Brunswick, G-86) and monitored with glass thermometers immersed in the exit and inlet of the column. The current was applied by a constant voltage/ current-controlled DC power source (NETES NPS1810D).
Fig. 1. The schematic view of continuous electrochemical tubular reactor ((1) carbon electrode, (2) plexiglas screws, (3) plexiglas reactor cover, (4) discharge line, (5) heating/cooling coil, (6) reactor support, (7) reservoir, (8) feed pump, (9) feed line, (10) connections, (11) DC power source).
1508
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 .
2.3. Experimental procedure Synthetic phenolic wastewater with the electrolyte was prepared in the wastewater reservoir at specified concentration and was fed to the column to fill it completely. An initial sample of 10 ml was taken from the reservoir to determine the initial phenol concentration prior to reaction. The reaction started with the application of the specified voltage and current and, the recycling water for temperature control was pumped through the reactor coil while feed stream was pumped to the column continuously. At appropriate time intervals, samples of 10 ml were taken from the reactor outlet and analysed to determine the intermediate properties of the reaction medium during electrochemical conversion. DC power source was turned off and the reaction was terminated after steady state reached.
column was used for the separation and oven temperature was programmed as constant 501C for 1 min, and then increased to 2501C with 201C min1 rate and finally 10 min at 2501C. Automatic injection system was kept at 250–2801C while mass detector temperature was held at 2801C. For identification of the products, standard samples and Wiley library search were used. Then, internal calibration was applied using a standard (dichlorophenol) with response factor 0.99 to calculate corresponding concentrations of compounds based on the area under the curve. For polymer, 4–5 peaks were detected with their respective mass number values and individual weight concentrations were calculated. The results were presented as percent of the initial phenol present in the wastewater.
3. Results and discussions 2.4. Reaction conditions The phenolic wastewater was used with 450 and 3100 mg l1 initial phenol concentrations in the runs at 251C reaction temperature with 120 g l1 electrolyte concentration. For 450 and 3100 mg l1 initial phenol concentrations, the applied current density was 61.4 and 54.7 mA cm2, respectively. The values of temperature, electrolyte concentration and current density applied were determined to be optimum through batch runs for maximum phenol removal rate in a previous work [34]. 2.5. Analysis The change in phenol concentration of the reaction medium was determined by means of a Cecil 1100 HPLC system. The system was equipped with a variable wavelength monitor (CE 11220) operating at 254 nm and a Cecil HPLC pump system (CE 1100). Chromatograms were obtained using a Hichrom-S5OS1 column (Chromosorb). Only one mobile phase, running at ambient temperature, was eluted and 20 ml sample was injected. Methanol/water (40/60, v/v) was used as the mobile phase and the flow rate was adjusted to 1 ml min1 under 28 MPa pressure. Before the analysis, the mobile phase was filtered and sonicated in order to remove dissolved gases. GC/MS analyses were performed with a 5972 MSD connected to a 5890 HPs GC system in order to determine the intermediate products formed during the electrochemical reaction in a specified batch run. For the analysis, 5 ml liquid sample was taken from the reactor at appropriate time intervals. The pH of the sample was adjusted to 3 with HNO3 and aqueous phase was saturated with NaCl. Then the sample was extracted for 20 min with 5 ml of diethyl ether, organic phase was separated in a separation funnel, and 1 ml of ether phase was injected into GC/MS system. A DB-1 capillary
Continuous tubular reactor experiments were carried out for the determination of the effects of the residence time on phenol conversion at optimized conditions determined previously through batch runs [34]. Phenol feed concentration was selected at low and high (450 and 3100 mg l1) concentrations with 61.4 and 54.7 mA cm2 applied current density for each residence time to visualize the change in phenol removal performance. Although both feed streams contained the same amount of electrolyte, the conductivity of the feed with lower phenol concentration was observed higher than that of higher phenol feed concentration. This is basically due to the cloud-like polymer formation around the anode at higher phenol concentrations in an unmixed flow system with low linear velocity (3oNRe o53), which diminished the current applied. The runs were carried out in doubles and average results were presented in figures. The data points for duplicate runs at any time were not different from each other more than 4%. The passivating polymeric formation on electrodes has not been detected in any run, which, in fact, was observed in literature studies. 3.1. The effect of residence time For 450 mg l1 initial phenol feed concentration, phenol concentration at the reactor outlet decreased with increasing residence time as shown in Fig. 2. At 10 and 20 min of residence time phenol removal was 56% and 78%, respectively. Above 40 min residence time, all phenol was eventually consumed in the column and 100% phenol removal was achieved. During the reactions, the colour of the solution at the outlet stream turned to light yellow from transparent then to dark brown due to increased residence time. At higher residence times, the formation of the suspended solid particles in reaction medium caused an opaque
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 .
1509
Phenol Outlet Concentration (mg l-1)
500 Residence Time
450
10 min 20 min 40 min 60 min 90 min
400 350 300 250 200 150 100 50 0 0
10
20
30
40
50
60
70
80
90 100 110 120
Elapsed Time (min)
Fig. 2. The change of outlet phenol concentration profiles by time in the range 10–90 min of residence time for 450 mg l1 phenol feed concentration (current density: 61.4 mA cm2; T: 251C; electrolyte concentration: 120 g l1).
14 12 10 8 pH
appearance and the colour of the solution at the reactor outlet was dark brown due to the formation of polymeric products. Easily separable solid polymeric materials were obtained at the outlet stream but the formation of polymers did not affect the performance of the column, mainly cell current density. Fig. 3 shows the pH change at the exit of the column for 450 mg l1 phenol concentration, where pH was not controlled. As a general trend, at the beginning reactor outlet pH rised up to pH 11, then began to drop gradually in proportion with the residence time in the tubular reactor, i.e. more residence time caused more pH drop after peak pH. Initial pH rise could be attributed to the production of hydroxyl ions on the cathode in the presence of NaCl and pH drop was caused by gradual consumption of hydroxyl anions as well as production of hydrogen cations in dissociation reactions of HOCl and OCl along with the formation reaction of HOCl in the treated waste stream. Short residence time ended up with higher final pH and increase in residence time caused lowered final pH values. Nevertheless, residence time greater than 20 min also caused instable pH values due to unestablished steady-state conditions in the reactor within the time range of runs as well as non-homogeneous character of the reactor. Stable outlet phenol concentrations were established approximately after 80, 70, 40, 40 and 30 min of elapsed time for the residence time of 10, 20, 40, 60 and 90 min, respectively, and the corresponding energy requirement per hour at steady state was 1.144, 2.077, 2.679, 3.057 and 1.410 kWh mol1 consumed phenol, respectively, as shown in Table 1. At steady state, experimental number of electrons consumed per hour was calculated as 0.33, 0.34, 0.35, 0.26, and 0.33 mol for 10, 20, 40, 60 and 90 min of residence time, respectively, in Table 1. Although electron consumption rate values were almost the same, phenol consumption rate decreased in accord
6
Residence Time 10 min 20 min 40 min 60 min 90 min
4 2 0 0
20
40
60
80
100
120
Elapsed Time (min)
Fig. 3. The change of pH profiles by time at the reactor outlet in the range 10–90 min of residence time for 450 mg l1 phenol feed concentration (current density: 61.4 mA cm2; T: 251C; electrolyte concentration: 120 g l1).
with increased residence time at steady state. Also, specific energy requirement increased proportionally with residence time except for 90 min. At 3100 mg l1 phenol feed concentration, the continuous column reactor reached steady state after 5 h as shown in Fig. 4. At 1, 1.5, 2 and 3 h of residence time, phenol removal was achieved 42%, 71%, 81% and 98%, respectively. The increase in residence time increased the phenol removal percentage; however, polymeric products formed started to block the reactor discharge at higher residence times due to low volumetric flow rate. Due to high phenol concentration, excessive amount of foam was produced and the colour of the solution was turned to light yellow from transparent and then to dark brown at the outlet stream. The variation of pH in the exit stream of the column for 3100 mg l1 phenol feed concentration was depicted in Fig. 5. Similar trends of
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 .
1510
Table 1 The energy requirement and electron consumption rate for 450 mg l1 phenol feed concentration (current density: 61.4 mA cm2; T: 251C; electrolyte concentration: 120 g l1) Residence time (min)
Time required to reach steady state (min) 80 70 40 40 30
Phenol Outlet Concentration (mg l-1)
10 20 40 60 90
Phenol consumption rate at steady state (mmol h1)
Specific energy requirement at steady state (kWh mol1 phenol consumed)
Electron consumption rate at steady state (mol e h1)
31.1071 21.6639 13.8871 9.2581 6.1720
1.144 2.077 2.679 3.057 1.410
0.33 0.34 0.35 0.26 0.33
Phenol removal (%) at steady state 56 78 100 100 100
Residence Time
3000
1h 1.5 h 2h 3h
2500 2000 1500 1000 500 0 0
2
1
3 4 5 Elapsed Time (h)
6
7
8
Fig. 4. The change of outlet phenol concentration profiles by time in the range 1–3 h of residence time for 3100 mg l1 phenol feed concentration (current density: 54.7 mA cm2; T: 251C; electrolyte concentration: 120 g l1).
14 12
pH
10 8 6
Residence Time
1h
4
1.5 h 2h
2
3h
0 0
1
2
3
4
5
6
7
8
Elapsed Time (h)
Fig. 5. The change of pH profiles by time at the reactor outlet in the range 1–3 h of residence time for 3100 mg l1 phenol feed concentration (current density: 54.7 mA cm2; T: 251C; electrolyte concentration: 120 g l1).
Fig. 3 were also observed in this figure. Except for the residence time of 1 h, pH profiles first reached as high as pH 11.5 and then dropped with an unstable manner as
explained for 450 mg l1 phenol feed concentration. The steady-state energy requirement per hour was calculated as 4.345, 4.670, 5.704 and 6.379 kWh mol1 consumed phenol for 1, 1.5, 2 and 3 of residence time, respectively, as shown in Table 2. At steady state, experimental number of electrons consumed per hour was calculated as 0.24, 0.29, 0.26 and 0.28 mol for 1, 1.5, 2 and 3 h of residence time, respectively, in Table 2. In Table 2, electron consumption rate was again almost the same like in Table 1 and phenol consumption rate decreased proportionally to residence time after 1.5 h of residence time. Although there has not been any work done yet with electrochemical treatment of phenol in literature using tubular reactor to compare with our results, there are other works carried out in different systems for electrochemical oxidation of phenol. Smith de Sucre and Watkinson [3] achieved complete phenol removal within 90 and 120 min reaction time for 525 and 1100 mg l1 initial phenol concentrations, respectively, at 10 A current and 8.5 V cell voltage in aqueous solutions of Na2SO4 and H2SO4 or NaOH using a lead
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 .
1511
Table 2 The energy requirement and electron consumption rate for 3100 mg l1 phenol feed concentration (current density: 54.7 mA cm2; T: 251C; electrolyte concentration: 120 g l1) Residence time (h)
1 1.5 2 3
Time required to reach steady state (h) 5 5 5 5
Phenol consumption rate at steady state (mmol h1)
Specific energy requirement at steady state (kWh mol1 phenol consumed)
Electron consumption rate at steady state (mol e h1)
26.7882 30.1899 25.8315 20.8353
4.345 4.670 5.704 6.379
0.24 0.29 0.26 0.28
dioxide anode packed-bed reactor. In their work, the energy requirement changed from 4.8 to 2.4 kWh mol1 phenol removed as the initial phenol concentration increased from 525 to 1100 mgl1 in batch runs. Chettiar and Watkinson [11] achieved 95.8% phenol removal for 1070 mg l1 initial phenol concentration after 5 h at 10 A with an electrochemical system adapted from the previous work of Smith de Sucre and Watkinson [3]. Sharifian and Kirk [21] achieved almost 100% phenol removal within 6 h for 1316 mg l1 initial phenol concentration in 1 M H2SO4 at a cell current of 2 A at 251C and 1 ml s1 flow rate with a packed-bed reactor of PbO2 pellets. Nevertheless, after 10 h of reaction time, the intermediates of maleic acid and benzoquinone were still in the medium. Gattrell and Kirk [2] achieved about 60%, 60% and 55% phenol removal for 470, 940 and 1880 mg l1 initial phenol concentrations, respectively, at 1.7 V in 8 h of electrolysis time with a flow-by electrochemical cell. For all three concentrations, the rate of phenol removal slowed down after the first hour due to formation of greater amounts of intermediates. The authors also reported that phenol passivated the electrode surface during the first hour of operation due . to the polymer forming reaction. Kotz et al. [18] achieved complete phenol removal after 2 A h charge passed for 1000 mg l1 phenol concentration by anodic oxidation using a SnO2 anode at 30 mA cm2 current density. The authors reported that after 1 A h charge passed; the phenol disappeared while the complete removal of TOC took about 8 A h and calculated the energy requirement as 1.88 kWh mol1 of phenol removed. Comninellis and Pulgarin [12] achieved 100% TOC removal at 70 Ah dm3 with SnO2 anodes for 1974 mg l1 initial phenol concentration, for 701C and 50 mA cm2 current density with a two-compartment electrochemical cell. Again, Comninellis and Nerini [13] achieved complete phenol removal at 15 A h dm3 for 940 mg l1 initial phenol concentration in the presence of NaCl, for 501C, 25.3 g l1 NaCl concentration and 100 mA cm2 current density using an electrochemical system adapted from the previous work of Comninellis and Pulgarin [12] with Ti/SnO2 anodes to oxidize phenol
Phenol removal (%) at steady state 42 71 81 98
anodically. Lin et al. [20] achieved 70% COD removal within 60 min for 1000 mg l1 initial phenol concentration with 60 mg l1 H2O2 addition at 2.5 A current and 1% of salinity using a batch system with four pieces of cast iron cathode and anode. The authors also reported that the COD removal without H2O2 addition at the same conditions was 34%. Alvarez-Gallegos and Pletcher [36] and Harrington and Pletcher [37] investigated the electrochemical removal of low-level organics including phenol. Alvarez-Gallegos and Pletcher [36] calculated the energy reqirement for 0.33 mM phenol solution as 1.25 kWh m3 to reduce the COD to o10 ppm. Harrington and Pletcher [37] reported that the energy consumptions for the treatment of 1 m3 of the 0.5 mM phenol solution assuming a cell voltage of 5 V lied in the range 1.5–2.5 kWh m3 which was low compared to other electrochemical alternatives for the removal of organics from wastewater. The authors also calculated the energy consumption for the teatment of 3 mM phenol solution around 14 kWh m3. After reaching steady state, our continuous tubular reactor residence time values for almost complete phenol removal, 40 min and 3 h for 450 and 3100 mg l1 phenol feed concentrations, respectively, are less than those of literature for electrochemical treatment, and the treatment also requires less specific energy, 2.679 and 6.379 kWh mol1 phenol consumed for 450 (40 min) and 3100 mg l1 (3 h) phenol feed concentrations, respectively; denoting a feasible alternative reactor configuration. This advantage is partially due to the reactor type and also due to a different reaction mechanism, which favoured the fast formation of nonpassivating polymeric end products rather than slow total oxidation of phenol. 3.2. Determination of conversion products GC/MS chromatograms and mass spectrums were obtained in order to identify the intermediate products formed in the liquid phase during the electrochemical conversion of phenol. The data in Figs. 6 and 7 were collected in batch runs carried out at 32.9 mA cm2,
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 .
1512
251C reaction temperature and 120 g l1 electrolyte concentration for 496 and 3505 mg l1 initial phenol concentrations. In Figs. 6 and 7, concentrations of all compounds were expressed as percentage weight of initial phenol present in the waste stream. For each batch run the exit stream was completely recycled with the flow rate of 6000 ml min1 and fed to the inlet of the column. From periodic analysis of the medium taken at the exit, it was understood that mainly mono, di-and trisubstituted chlorinated phenol products were formed during the electrolysis and consumed thereafter by either destruction or polymerization, or both as seen in Figs. 6 and 7.
120
Compound phenol
100
2-chlorophenol
Compound % (w/w)
4-chlorophenol 2,4 –dichlorophenol
80
2,4,6-trichlorophenol polymeric products
60 40 20 0 0
20
40
60
80
100
120
Elapsed Time (min)
Fig. 6. The weight percentage of conversion products based on initial phenol at 496 mg l1 initial phenol concentration in the batch run (32.9 mA cm2 current density, 251C reaction temperature, 120 g l1 electrolyte concentration).
Diverse from literature [2,3,9–12,14,15,21,23] except the work of Comninellis and Nerini [13], only chlorinated phenol products formed during the runs with Cl2 gas discharging on the anode; and all the chlorinated phenol derivatives were completely consumed either by destruction or polymerization at 496 and 3505 mg l1 initial phenol concentrations after 120 min and 11 h, respectively. As Lund and Iverson [38] previously stated, the oxidation of organic compounds at graphite anode often results in other products when considered with the oxidation at platinum anode, but in our work we had only four different intermediates, which did not turn into successive intermediate compounds in contrast to other studies of literature [2,3,9–12,14,15,21,23]. Firstly, 4-chlorophenol and 2,4,6-trichlorophenol formed dominantly during the electrolysis and followed by 2-chlorophenol and then 2,4-dichlorophenol for both initial phenol concentrations of 496 and 3505 mg l1. In Fig. 6 for 496 mg l1 initial phenol concentration, the weight percentage of 2,4,6-trichlorophenol increased rapidly up to 40 min, although 4-chlorophenol started to decrease after 20 min. The weight percentage of polymeric compounds increased significantly after 40 min with the disappearance of mono, di and trichlorophenols. This result showed that the chlorinated hydrocarbons contributed to the formation of polymeric structure at a certain extent, while part of them might be oxidized. Finally, after 120 min, no intermediates were left in the reaction medium at all. Due to the contribution of chlorine into the polymer structure for 450 mgl1 phenol feed concentration, the final weight percentage of polymeric compounds passed over 100% [39].
100
Compound 90
phenol 2-chlorophenol
80
Compound % (w/w)
4-chlorophenol 70
2,4 dichlorophenol
60
polymeric products
2,4,6-trichlorophenol
50 40 30 20 10 0 0
2
4
6
8
10
12
Elapsed Time (h)
Fig. 7. The weight percentage of conversion products based on initial phenol at 3505 mg l1 initial phenol concentration in the batch run (32.9 mA cm2 current density, 251C reaction temperature, 120 g l1 electrolyte concentration).
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 .
Similar results were observed in Fig. 7 for 3505 mg l1 initial phenol concentration; the weight percentage of 2,4,6-trichlorophenol increased rapidly up to 4 h, but pchlorophenol started to decrease after 1.5 h. The concentration of polymeric compounds increased noticeably after 4 h with the decreasing of mono, di and trichlorophenols. Finally after 10 h, no intermediates were left in the reaction medium. Although chlorinated phenol derivatives are more toxic than phenol itself [29,30], in our work all intermediate chlorophenols were converted into polymeric compounds or oxidized eventually and did not appear in the reaction solution, assuring a safe and non-toxic treatment. The insoluble polymer formed is lighter than the treated water and therefore can be easily removed by means of density difference.
4. Conclusion The electrochemical treatment of phenolic wastewater in a continuous tubular reactor was investigated. The effect of residence time on phenol removal was studied for 450 and 3100 mg l1 phenol feed concentrations at 251C, 120 g l1 electrolyte concentration, 61.4 and 54.7 mA cm2 current densities, respectively. For 450 mg l1 phenol feed concentration at steady state, above 40 min of residence time all phenol was consumed within the column and 100% phenol removal was achieved. At 3100 mg l1 phenol feed concentration, 98% of phenol removal was realized only at 3 h of residence time at the steady state. At steady state, residence times of 40 min and 3 h for 450 and 3100 mg l1 phenol feed concentrations, respectively, are comparatively shorter than those of literature for electrochemical treatment for phenol removal, and the process also requires less specific energy, 2.679 and 6.379 kWh mol1 phenol consumed for 450 and 3100 mg l1 phenol feed concentrations, respectively. Analysis of the reaction products has shown that during the electrolysis mono, di and tri substituted chlorinated phenol products were formed along with polymeric compounds and all chlorinated phenols were consumed thereafter by polymerization and destruction. The results indicate that this continuous tubular reactor may be a feasible and safe alternative system for the electrochemical treatment of phenolic wastewater.
Acknowledgements This project was supported by Hacettepe University Research Fund with Grant no: 98.02.602.004.
1513
References [1] Lanouette KH. Treatment of phenolic wastes. Chem Eng 1977;84(22):99–106. [2] Gattrell M, Kirk DW. The electrochemical oxidation of aqueous phenol at a glassy carbon electrode. Can J Chem Eng 1990;68:997–1003. [3] Smith de Sucre V, Watkinson AP. Anodic oxidation of phenol for waste water treatment. Can J Chem Eng 1981;59:52–9. [4] S- eker S- , Beyenal H, Salih B, Tanyolac- A. Multi-substrate growth kinetics of Pseudomonas putida for phenol removal. Appl Microbiol Biotechnol 1997;47:610–4. [5] Pulgarin C, Adler N, P!eringer P, Comninellis C. Electrochemical detoxification of a 1,4-benzoquinone solution in wastewater treatment. Water Res 1994;28(4): 887–93. [6] Awad YM, Abuzaid NS. Electrochemical treatment of phenolic wastewater: efficiency, design considerations and economic evaluation. J Environ Sci Health A 1997;32(5):1393–414. [7] Bockris JO’M, Kim J. Electrochemical treatment of lowlevel nuclear wastes. J Appl Electrochem 1997;27:623–34. [8] Boscoletto AB, Gottardi F, Milan L, Pannocchia P, Tartari V, Tavan M, Amadelli R, Battisti A, Barbieri A, Patracchini D, Battaglin G. Electrochemical treatment of bisphenol-A containing wastewaters. J Appl Electrochem 1994;24:1052–8. [9] Boudenne JL, Cerclier O, Gal!ea J, Van der Vlist E. Electrochemical oxidation of aqueous phenol at a carbon black slurry electrode. Appl Catal A: Gen 1996;143: 185–202. * * J, [10] Canizares P, Dom!ınguez JA, Rodrigo MA, Villasenor Rodr!ıguez J. Effect of the current intensity in the electrochemical oxidation of aqueous phenol wastes at an activated carbon and steel anode. Ind Eng Chem Res 1999;38:3779–85. [11] Chettiar M, Watkinson AP. Anodic oxidation of phenolics found in coal conversion effluents. Can J Chem Eng 1983;61:568–74. [12] Comninellis C, Pulgarin C. Electrochemical oxidation of phenol for wastewater treatment using SnO2 anodes. J Appl Electrochem 1993;23:108–12. [13] Comninellis C, Nerini A. Anodic oxidation of phenol in the presence of NaCl for wastewater treatment. J Appl Electrochem 1995;25:23–8. [14] Do JS, Yeh WC. Paired electrooxidative degradation of phenol with in situ electrogenerated hydrogen peroxide and hypochlorite. J Appl Electrochem 1996;26:673–8. [15] Gattrell M, Kirk DW. A study of electrode passivation during aqueous phenol electrolysis. J Appl Electrochem 1993;140(6):1534–40. [16] Genders JD, Hartsough D. Electrochemical reduction of nitrates and nitrites in alkaline nuclear waste solutions. J Appl Electrochem 1996;26:1–9. [17] Israilides CJ, Vlyssides AG, Mourafetti VN, Karvouni G. Olive oil wastewater treatment with the use of an electrolysis system. Biores Technol 1997;61:163–70. . R, Stucki S, Carcer B. Electrochemical waste water [18] Kotz treatment using high overvoltage anodes; Part I: physical
1514
[19]
[20]
[21] [22]
[23]
[24]
[25]
[26] [27]
[28]
[29]
B.K. Korbahti, A. Tanyolac- / Water Research 37 (2003) 1505–1514 . and electrochemical properties of SnO2 anodes. J Appl Electrochem 1991;21:14–20. Lin SH, Chen ML. Textile wastewater treatment by enhanced electrochemical method and ion exchange. Environ Technol 1997;18:739–46. Lin SH, Shyu CT, Sun MC. Saline wastewater treatment by electrochemical method. Water Res 1998;32(4): 1059–66. Sharifian H, Kirk D W. Electrochemical oxidation of phenol. J Electrochem Soc 1986;133(5):921–4. . R, Carcer B, Suter W. Electrochemical Stucki S, Kotz waste water treatment using high overvoltage anodes; Part II: anode performance and applications. J Appl Electrochem 1991;21:99–104. Tahar BN, Savall A. Electrochemical degradation of phenol in aqueous solution on bismuth doped lead dioxide: a comparison of the activities of various electrode formulations. J Appl Electrochem 1999;29:277–83. Tennakoon CLK, Bhardwaj RC, Bockris JO’M. Electrochemical treatment of human wastes in a packed bed reactor. J Appl Electrochem 1996;26:18–29. Vlyssides AG, Israilides CJ, Loizidou M, Karvouni G, Mourafetti V. Electrochemical of vinasse from beet molasses. Water Sci Technol 1997;36(2-3):271–8. Hedenburg JF, Freiser H. Anodic voltammetry of phenols. Anal Chem 1953;25:1355–8. Koile RC, Johnson DC. Electrochemical removal of phenolic films from a platinum anode. Anal Chem 1979;51(6):741–4. Wang J, Jiang M, Lu F. Electrochemical quartz crystal microbalance investigation of surface fouling due to phenol oxidation. J Electroanal Chem 1998;444: 127–32. US Department of Health and Human Services, Toxicological profile for phenol. Sciences International, Inc., Research Triangle Institute, USA, 1998.
[30] US Department of Health and Human Services, Toxicological profile for chlorophenols. Sciences International, Inc., Research Triangle Institute, USA, 1999. [31] Hearst JE, Ifft JB. Contemporary chemistry. San Fransisco, CA: Freeman, 1976. [32] Mortimer CE. Chemistry. Belmont, CA: California Wadsworth Publishers, 1986. [33] Metcalf & Eddy, Inc. In: Tchobanoglous G, Burton FL, editors. Wastewater engineering. Singapore: McGraw-Hill, 1991. . [34] Korbahti BK, Salih B, Tanyolac- A. Electrochemical conversion of phenolic wastewater on carbon electrodes in the presence of NaCl. J Chem Technol Biotechnol 2002;77(1):70–6. . [35] Korbahti BK. Continuous electrochemical treatment of phenolic wastewater in a tubular reactor. MS dissertation, Hacettepe University, Ankara, Turkey, 1999. [36] Alvarez-Gallegos A, Pletcher D. The removal of low-level organics via hydrogen peroxide formed in a reticulated vitreous carbon cathode cell. Part 2: the removal of phenols and related compounds from aqueous effluents. Electrochim Acta 1999;44:2483–92. [37] Harrington T, Pletcher D. The removal of low levels of organics from aqueous solutions using Fe(II) and hydrogen peroxide formed in situ at gas diffusion electrodes. J Electrochem Soc 1999;146(8):2983–9. [38] Lund H, Iversen P, Baizer MM, editors. Organic electrochemistry. New York: Marcel Dekker, 1973. . [39] Zareie MH, Korbahti BK, Tanyolac- A. Non-passivating polymeric structures in electrochemical conversion of phenol in the presence of NaCl. J Hazardous Mater B 2001;87:199–212. [40] Vlyssides AG, Loizidou M, Korlis PK, Zorpas AA, Papaioannou D. Electrochemical oxidation of a textile dye wastewater using a Pt/Ti electrode. J Hazardous Mater B 1999;70:41–52.