- Email: [email protected]
Arsenic removal from groundwater by electrocoagulation in a pre-pilot-scale continuous filter press reactor
Chemical Engineering Science 97 (2013) 1–6
Contents lists available at SciVerse ScienceDirect
Chemical Engineering Science journal homepage: www.elsevier.com/locate/ces
Arsenic removal from groundwater by electrocoagulation in a pre-pilot-scale continuous filter press reactor Omar J. Flores a, José L. Nava a,n, Gilberto Carreño a, Enrique Elorza b, Francisco Martínez a a
Departamento de Ingeniería Geomática e Hidráulica, Universidad de Guanajuato, Av. Juárez 77, Zona Centro, C.P. 36000 Guanajuato, Guanajuato, Mexico Departamento de Ingeniería en Minas Metalurgia y Geología, Universidad de Guanajuato, ExHacienda San Matías S/N, C.P. 36020 Guanajuato, Guanajuato, Mexico
b
H I G H L I G H T S
Arsenic removal from groundwater by electrocoagulation. Aluminum as sacrificial anodes. Electrolyzes at different flow rates and current densities. Electrocoagulation led to 100% arsenic removal (50 μg L−1) with 3.9 kWh m−3. The influence of hypochlorite on arsenic removal efficiency was analyzed.
art ic l e i nf o
a b s t r a c t
Article history: Received 19 November 2012 Received in revised form 6 April 2013 Accepted 15 April 2013 Available online 24 April 2013
We investigated arsenic removal from groundwater by electrocoagulation (EC) using aluminum as the sacrificial anode in a pre-pilot-scale continuous filter press reactor. The groundwater was collected at a depth of 320 m in the Bajío region in central Mexico (arsenic 50 μg L–1, carbonates 40 mg L–1, hardness 80 mg L–1, pH 7.5 and conductivity 150 μS cm–1). The influence of current density, mean linear flow and hypochlorite addition on the As removal efficiency was analyzed. Poor removal of total arsenic (o 60 %) in the absence of hypochlorite is due to a mixture of arsenite (HAsO2(aq) and H3AsO3(aq)) and arsenate (HAsO42−). Arsenic removal is more efficient when arsenite is oxidized to arsenate by addition of hypochlorite at a concentration typically used for disinfection (1 mg L–1). Arsenate removal by EC might involve adsorption on aluminum hydroxides generated in the process. Complete arsenate removal by EC was satisfactory at a current density of 5 mA cm–2 and mean linear flow of 0.91 cm s–1, with electrolytic energy consumption of 3.9 kWh m3. & 2013 Elsevier Ltd. All rights reserved.
Keywords: Arsenic removal Aluminum sacrificial anode Dissolution Electrolysis Coagulation Precipitation
1. Introduction Mining activities have historically led to environmental pollution. The main elements in groundwater contamination are heavy metals and metalloids such as lead, copper, zinc, iron, and arsenic produced by acid rock drainage (Ramos et al., 2011). Dissolved arsenic (As) was recently found in groundwater in the Bajío region (Guanajuato, central Mexico) at concentrations of 50–155 μg L–1. This problem was caused by arsenic mobilization in mine tailings dams (Hernández et al., 2010) and natural weathering of arseniccontaining rocks (Ramos et al., 2011). Mining was once the predominant economic activity in the region.
n
Corresponding author. Tel.: +52 4731020100x2289; fax: +52 4731020100x2209. E-mail addresses: [email protected], [email protected] (J.L. Nava).
0009-2509/$ - see front matter & 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ces.2013.04.029
Arsenic-contaminated water is a significant problem owing to its toxicity. Long-term exposure to arsenic leads to chronic health problems such as hyperpigmentation and keratosis of the hands and feet; it also causes bladder, lung, skin, kidney, liver, and prostate cancer (Smith et al., 2006). Considering the high toxicity of arsenic, the World Health Organization (WHO) and the US Environmental Protection Agency have set a maximum acceptable level of 10 μg L–1 for arsenic in drinking water, which differs from the recommendation of the Mexican authorities (25 μg L–1). Removal of arsenic from large volumes of water is generally performed by adding chemical coagulants such as aluminum or iron sulfate. However, owing to low concentrations of dissolved As, this process produces large amounts of sludge because SO4– anions consume 50% of the coagulant (Rios et al., 2005). This is the main reason why electrocoagulation (EC) is typically used, in which aluminum or iron, when dissolved electrolytically, supports the coagulation process and decreases the amount of sludge. EC is
2
O.J. Flores et al. / Chemical Engineering Science 97 (2013) 1–6
effective in destabilizing dispersed fine particles and anions contained in water (Chen et al., 2002). Flocs of metallic hydroxides are formed and remove dissolved species (Chen et al., 2002; Shen et al., 2003; Hernández et al., 2010; Mohora et al., 2012). Arsenic species in water from deep wells (pH∼7.5) include innocuous arsenite (HAsO2 and H3AsO3, oxidation state III) and arsenate anions (HAsO42–, oxidation state V) (Pourbaix, 1974; Smedley and Kinniburgh, 2002). Arsenate at neutral pH is more susceptible to EC removal (Hansen et al., 2007; Hernández et al., 2010; Wan et al., 2011; Mohora et al., 2012; Balasubramanian et al., 2009) because arsenate anions adsorb onto flocs of iron (Hansen et al., 2007, 2008; Kobya et al., 2011; Wan et al., 2011) and aluminum hydroxides (Hansen et al., 2008; Hernández et al., 2010; Kobya et al., 2011; Mohora et al., 2012). Martínez-Villafañe et al. (2009) and Kobya et al. (2011) reported that arsenate removal occurs via ligand exchange with ferric hydroxides. According to van Genuchten et al. (2011), As(V) forms inner-sphere complexes with the surface of iron oxides. However, reports on the mechanism of As removal by aluminum flocs are rather limited. Hansen et al. (2007) reported that oxidation of As(III) to As(V) by ferric ions and/or oxygen bubbling is convenient. The electrode materials commonly used for As removal by EC are mainly iron (Parga et al., 2005; Hansen et al., 2007, 2008; García-Lara and Montero-Ocampo, 2010; Kobya et al., 2011; Wan et al., 2011) and aluminum (Hernández et al., 2010; Kobya et al., 2011; Mohora et al., 2012). Kobya et al. (2011) demonstrated that aluminum was better than iron electrodes at removing arsenic from drinking water. We obtained similar results in a previous study (Hernández et al., 2010). There are literature reports on EC removal of As using Al electrodes from synthetically prepared water (Hernández et al., 2010; Kobya et al., 2011) and groundwater (Mohora et al., 2012) containing high arsenic concentrations. The results differ because a great number of species that are present in groundwater interfere with arsenic removal, besides reactor operating parameters such as current density, flow rate and geometry. The aim of this study was to decrease the arsenic concentration in groundwater from the Bajío region by EC using aluminum as the sacrificial anode in a pre-pilot-scale continuous filter press reactor. The influence of current density, mean linear flow rate and hypochlorite addition on the As removal efficiency was analyzed. The energy consumption for electrolysis was also estimated.
2. Electrochemical and chemical processes during As removal by EC using aluminum electrodes EC involves in situ generation of coagulants by electrodissolution of aluminum electrodes. Aluminum cations are generated at the anode and hydrogen gas is evolved at the cathode. This process generates aluminum hydroxide, which is believed to adsorb As (Hernández et al., 2010; Kobya et al., 2011; Mohora et al., 2012). The electrode and chemical reactions in neutral solution (pH∼7) are shown below. At the anode, electrodissolution generates aluminum ions (Al3+), which are transformed to aluminum hydroxide and aluminum oxide in the bulk Al(s)-Al3++3e– 3+
Al +3H2O-Al(OH)3(s)+3H
(1) +
2Al3++3H2O-Al2O3(s)+6H+
Al(OH)3 and Al2O3 flocs are believed to adsorb HAsO2− 4 (Hernández et al., 2010; Kobya et al., 2011; Mohora et al., 2012) 2– Al(OH)3(s)+HAsO2– 4 -[Al(OH)3 HAsO4 ](s)
(4)
2– Al2O3(s)+HAsO2– 4 -[Al2O3 HAsO4 ](s)
(5)
We previously reported that arsenite removal may occur by sweep coagulation, in which interaction occurs between As(III) and Al(OH)3 and Al2O3 flocs (Hernández et al., 2010). However, reports on the mechanism of As removal by aluminum flocs are rather limited. A deeper understanding of the As interaction with aluminum precipitates requires additional research. At the aluminum cathode, hydrogen gas is released 3H2O+3e–-1.5H2+3OH–
(6)
The solution at the cathode interface becomes alkaline over time. OH– ions migrate and diffuse away from the cathode to the anode, thus favoring water formation by reaction with protons from reaction (2) to (3) OH–+H+-H2O
(7)
The major problem with aluminum anodes is passivation due to Al2O3 precipitation, which leads to high anode and cell potentials and increases the energy consumption and cost of EC (Kobya et al., 2011; Mohora et al., 2012). Passivation can be controlled at low current densities in combination with convection (turbulent flow conditions), which favors Al3+ transport away from the surface to the bulk solution. In addition, cathodes of the same material can be used to electrodissolve Al2O3 by periodic current reversal (Mohora et al., 2012), which allows even consumption of the aluminum electrodes during the process.
3. Experimental 3.1. Solutions Groundwater for electrochemical treatment was collected from deep wells (320 m) located in the Bajío region (Guanajuato, México). Table 1 lists the characteristics of the sample. The groundwater contained 50 μg L–1 total As, which exceeds the Mexican standard for arsenic in water (25 μg L–1). It is believed that As(III) is predominant in this anaerobic groundwater. Water samples were obtained from the wells before the disinfection process, which usually involves hypochlorite addition. 3.2. Pre-pilot-scale filter press cell Fig. 1 shows a schematic of the experimental set-up. The system consists of a continuous pre-pilot-scale filter press cell in which the coagulant is produced. The resulting solution
Table 1 Characteristics of the groundwater samples collected at a depth of 320 m in the Bajío region (Guanajuato, México).
(2)
Parameter
Fresh
After hypochlorite additiona
(3)
Total arsenic (μg L–3) Carbonates (mg L–3) Hardness (mg L–1) pH Conductivity (μS cm–1)
59 40 80 7.5 150
59 40 80 8.1 392
The oxidation states of As in water are As(III) and As(V). As(V) is typically dominant in aerobic surface waters, while As(III) is found in anaerobic groundwater. At neutral pH, the predominant As (V) species has a net negative charge (HAsO42−), while As(III) species generally have no net charge (HAsO2 and H3AsO3) (Pourbaix, 1974).
a Hypochlorite was added at 1 mg L–1, the concentration typically used for disinfection of groundwater.
O.J. Flores et al. / Chemical Engineering Science 97 (2013) 1–6
3
electrode (SSE; Radiometer model XR200), 0.615 V vs. SHE/V. The potential of this electrode was determined using a high-impedance multimeter (Agilent model 34401A). All electrode potentials are reported with respect to SHE. 3.3. Methodology EC studies were carried out using the experimental set-up shown in Fig. 1. EC was performed under different hydrodynamic conditions with a mean linear flow rate of 0.91–4.55 cm s–1 and current density of 4, 5 and 6 mA cm–2. Each resulting solution was immediately passed to the test jar and mixed at slow speed (30 rpm) for 15 min for aggregate growth. The aggregates were allowed to precipitate in static solution for approximately 1 h. Arsenic was analyzed in the resulting clarified solution. After dissolution of the floc, aluminum was also analyzed. 3.4. Analytical procedure
Fig. 1. Electrical and flow circuit for the EC process in a filter press reactor.
Table 2 Dimensions of the electrocoagulation reactor. Reactor volume, Vr (cm3) Height, B (cm) Channel length, L (cm) Chanel width, S (cm) Total length of six channels, LT (cm) Number of channels Anode area in each channel in contact with solution (cm2) Cathode area in each channel in contact with solution (cm2) Cross-sectional area, AT (cm2)
88.94 3.05 8.1 0.6 48.6 6 24.7 24.7 1.83
(mixture of water and coagulant) is passed to a test jar to induce flocculation and adsorption of arsenic on aluminum flocs. The arsenic is then precipitated and the clarified solution is analyzed. The reactor was switched in monopole configuration. Three aluminum electrodes (3.05 cm 8.10 cm 0.30 cm) were used as anodes and four similar electrodes were used as cathodes. The electrodes were spaced at 0.60 cm using propylene separators. The reactor dimensions are listed in Table 2. This pre-pilot-scale reactor was developed in our laboratory for removal of arsenic from synthetic solutions (Hernández et al., 2010). The serpentine array induces fluid turbulence, enhancing mass transport of the coagulant from the anode to the bulk and avoiding both alumina precipitation and anode passivation. The reactor was connected to a hydraulic system consisting of a centrifuge pump (model MDX-MT-3) of 0.25 HP and a flow meter (model F-44250LH-8) with a capacity of 0–1 L min–1. The PVC pipes connecting the reservoir to the pump were 0.3 in. in diameter. The valves and connections were also made of PVC. A BK Precision power supply (model 1090) was used for electrolysis tests. The cell potential was measured using the power supply. The aluminum electrode potential was measured as the difference between aluminum and a saturated mercurous sulfate reference
Arsenic concentrations in the samples were measured using a standard method (APHA, 1995) with an atomic absorption spectrometer (Perkin-Elmer AAnalist 200) equipped with a manual hydride generator at 188.9 nm. The As detection limit was 0.1 μg L–1 and the error for triplicate analyses was within 2%. The kinetics of aluminum dissolution was followed by dissolving the sludge at pH 2 and then quantifying aluminum ions by AA. Carbonates and hardness were analyzed according to standard methods (APHA, 1995). Conductivity and pH measurements were carried out on a waterproof instrument (HANNA model HI 991300). All chemical reagents were of analytical grade. Each individual experiment was performed at least three times and the results were averaged.
4. Analysis of results and discussion 4.1. Arsenic removal from fresh groundwater Fig. 2 shows the residual concentration of arsenic (CAs) in fresh groundwater after EC as a function of the mean linear flow rate (u) at 4, 5 and 6 mA cm–2. The experimental and theoretical aluminum doses are also shown. At 4 mA cm–2, CAs linearly increased between 23 and 50 μg L–1 as a function of u in the interval 0.91–3.64 cm s–1 (Fig. 2a), indicating that arsenic removal decreases as u increases. This is due to depletion of the experimental Al(III) dose from 42 to 5 mg L–1. The quasi-constant CAs at u ≥ 3.64 cm s–1 is due to the low dose of aluminum (CAl(III) o4 mg L–1). The experimental aluminum dosage, obtained by analysis of the sludge, was lower than the theoretical amount added, calculated as jLMW C AlðIIIÞðNÞ ¼ ð1 106 ÞðNÞ ð8Þ nFSu where CAl(III) is the aluminum concentration (mg L–1), j is the current density (A cm–2), L is the length of one channel (cm), MW is the molecular weight of aluminum (26.98 g mol–1), n is the number of electrons exchanged (n ¼3), F is the Faraday constant (96485 C mol–1), S is the channel width (cm), u is the mean linear flow rate (cm s–1), N is the number of channels, and 1 106 is a conversion factor used to obtain the aluminum concentration in mg L–1. The difference between the experimental and theoretical values can probably be explained, on one hand, by oxygen evolution according to Eq. (9), which typically occurs simultaneously with Eq. (1) (Kobya et al., 2011), and, on the other hand, to
4
O.J. Flores et al. / Chemical Engineering Science 97 (2013) 1–6
1991). The arsenite species can be oxidized to HAsO42− by addition of hypochlorite. The effectiveness of this oxidation is well known (Ghurye and Clifford, 2001). The experimental aluminum dose decreased as the current density increased (Fig. 2a–c), contrary to what was expected. This could be related, on one hand, to enhanced oxygen evolution and, on the other hand, to simultaneous precipitation of Al2O3 at the anode (Kobya et al., 2011; Mohora et al., 2012), even under such hydrodynamic conditions. The change in shape for the experimental aluminum dose curve in Fig. 2a–c may be attributable to the formation of a layer of Al2O3 at the anode, which may passivate the anode against progressive electrodissolution. However, this analysis may be inaccurate. X-Ray studies would be helpful in identifying the Al2O3 layer, but this was beyond the scope of the present study. 4.2. Arsenic removal from groundwater after hypochlorite addition Fig. 3 shows CAs in pre-conditioned drinking water after EC as a function of u at 4, 5 and 6 mA cm–2. The experimental and theoretical aluminum doses are also shown.
Fig. 2. Influence of the mean linear flow rate on the residual arsenic concentration and aluminum dose for fresh groundwater (initial parameters: arsenic 50 μg L–1, carbonates 40 mg L–1, hardness 80 mg L–1, pH 7.5 and conductivity 150 μS cm–1) at a current density of (a) 4, (b) 5 and (c) 6 mA cm–2. The anode area for six channels was 148.2 cm2.
anode passivation due to Al2O3 precipitation (Kobya et al., 2011; Mohora et al., 2012). H2O-0.5O2+2H++2e–
(9)
At 5 mA cm–2, CAs increased from 23 to 50 μg L–1 as a function of u in the interval 1.82–4.55 cm s–1 (Fig. 2b), indicating that arsenic removal also decreases as u increases at this current density. Again, the experimental aluminum dose was lower than the theoretical one. At 6 mA cm–2, the behavior of CAs as a function of u differed for the ranges 1.82–2.73, 2.73–3.64 and 3.64–4.55 cm s–1. (Fig. 2c). However, the increase in CAs from 20 to 42 μg L–1 at 1.82– 4.55 cm s–1 was significant. The experimental aluminum dose was lower than the theoretical one. The poor EC removal of arsenic can be attributed to the mixture of HAsO2, H3AsO3 and HAsO2− 4 . The third species is probably more susceptible to removal by adsorption on Al(OH)3 and Al2O3 flocs (Hernández et al., 2010; Kobya et al., 2011; Mohora et al., 2012), since these flocs have a positive zeta potential (Amirtharajah et al.,
Fig. 3. Influence of the mean linear flow rate on the residual arsenic concentration and aluminum dose for groundwater after addition of hypochlorite (initial parameters: arsenic 50 μg L–1, carbonates 40 mg L–1, hardness 80 mg L–1, hypochlorite 1 mg L–1, pH 8.1 and conductivity 392 μS cm–1) at a current density of (a) 4, (b) 5 and (c) 6 mA cm–2. The anode area for six channels was 148.2 cm2.
O.J. Flores et al. / Chemical Engineering Science 97 (2013) 1–6
At 4 mA cm–2, CAs increased linearly between 3 and 38 μg L–1 as a function of u in the interval 0.91–4.55 cm –1 due to a decrease in the experimental aluminum dose from 17 to 1 mg L–1 (Fig. 3a). However, the CAs values are lower than those obtained for fresh groundwater at 4 mA cm–2 (Fig. 2a). This confirms the hypothesis that HAsO42− is probably removed by adsorption on Al(OH)3 and Al2O3 flocs (Hernández et al., 2010; Kobya et al., 2011; Mohora et al., 2012). X-Ray absorption spectroscopy and X-ray photoelectron spectroscopy would be helpful in investigating the local environment of Al and As in EC flocs generated in Bajío groundwater, but this was beyond the scope of the present study. It should be noted that CAs satisfies the WHO limit of ≤10 μg L–1 at u ¼0.91 cm s–1 and the Mexican standard of ≤25 μg L–1 at u ¼0.91– 1.82 cm s–1. At 5 mA cm–2, the behavior of CAs as a function of u differed for the ranges 0.91–1.81, 1.81–2.73 and 2.73–4.55 cm s–1 (Fig. 3b). CAs is close to zero at u¼0.91 cm s–1, 22 μg L–1 at u¼ 1.81 cm s–1 and 23 μg L–1 at u¼ 2.73 cm s–1. The latter slight variation in CAs can be attributed to the almost constant experimental aluminum dose of 6–7 mg L–1. CAs then increased linearly from 23 to 33 μg L–1 at 2.73–4.55 cm s–1 due to depletion of the experimental aluminum dose. At 6 mA cm–2, CAs as a function of u showed different behavior at 0.91–3.64 and 3.64–4.55 cm s–1 (Fig. 3c). The increase in CAs from 11 to 33 μg L−1 at 0.91–3.64 cm s–1 was significant, and CAs remained constant at 33 μg L–1 with a further increase to u¼4.55 cm s–1. In the presence of hypochlorite, the experimental dose of aluminum was lower than the corresponding theoretical dose (Fig. 3a–c). CAs after EC was lower in the presence of hypochlorite than without hypochlorite, indicating that arsenic removal is more efficient when arsenite is oxidized to arsenate. Arsenate removal by EC might involve adsorption on aluminum hydroxides generated in the process. Although the initial groundwater As concentration of 50 μg L–1 complies with safety standards in many countries, the Mexican authorities want to reduce the maximum acceptable level of arsenic in drinking water from 25 to 10 μg L–1 because of chronic health problems in the population. Experiments were performed at 4 and 5 mA cm–2 and 0.81 cm s–1 (Table 3) with the aim of obtaining residual arsenic concentrations of o10 μg L–1. Table 3 summarizes EC results that satisfy the Mexican arsenic standard of ≤25 μg L–1, including the cell potential (Ecell) and electrolytic energy consumption, evaluated as Econs ¼
Ecell I 3:6 B S u
ð10Þ
where I is the current intensity during electrolysis (C s–1), B is the channel height (cm), S is the channel width (cm) and 3.6 is a conversion factor used to obtain Econs in units of kWh m–3. Table 3 reveals that EC at 4 mA cm–2 satisfied the Mexican standard for arsenic at flow rates between 0.91 and 1.82 cm s–1, Table 3 Residual arsenic concentrations after hypochlorite addition satisfying the Mexican standard (CAs≤25 μg L–1), as well as the aluminum dose (CAl(III)) and dose rate (QAl(III) ¼CAl(III)u/LT), cell potential and electrolytic energy consumption. Current Density
u (cm s–1)
CAs (μg L–1)
CAl(III) (mg L–1)
QAl(III) (mg L–1 s–1)
Ecell (V)
Econs (kWh m–3)
4 mA cm−2
0.91 1.82 2.73 0.91 1.82 2.73 0.91 1.82 2.73
3 17 – 0 22 23 11 22 25
17 9 – 31 7 6 16 10 7
0.32 0.34 – 0.58 0.26 0.34 0.30 0.37 0.39
8.4 7.9 – 10.6 8.9 8.9 11.6 11.2 10.8
2.5 1.2 – 3.9 1.6 1.2 5.3 2.5 1.6
5 mA cm−2
6 mA cm−2
5
corresponding to an aluminum dose of 9–17 mg L–1 and an aluminum dose rate of QAl(III) ¼0.32–0.34 mg L–1 s–1, evaluated as Q AlðIIIÞ ¼
C AlðIIIÞ u LN
ð11Þ
EC at 5 mA cm–2 satisfied the Mexican As standard at flow rates between 0.91 and 2.73 cm s–1, corresponding to aluminum doses of 6–31 mg L–1and aluminum dose rates of 0.34–0.58 mg L–1 s–1. EC at 6 mA cm–2 satisfied the Mexican As standard at flow rates of 0.91–2.73 cm s–1, corresponding to aluminum doses of 7–16 mg L–1 and aluminum dose rates of 0.30–0.39 mg L–1 s–1. It should be noted that the cell potential decreased as a function of u at each current density, indicating that aluminum ions are transported away from the surface to the bulk solution under such turbulent flow conditions, decreasing Al2O3 precipitation on the anode. The best EC in terms of energy consumption was obtained at 4 mA cm–2 and u¼1.82 cm s−1, for which energy consumption of 1.2 kWh m–3 reduced arsenic from 50 to 17 μg L–1. Data for carbonates, hardness, pH and conductivity are not presented, since these parameters remained almost constant. Studies have shown that several dissolved species present in water samples can interfere with EC arsenic removal. For example, Wan et al. (2011) reported that phosphates in synthetic groundwater interfere with EC arsenic removal using Fe as sacrificial anodes, because phosphate and As species compete with surface iron oxide sites. The authors also showed that silicates and sulfates did not affect As removal. However, reports regarding As removal by EC using Al electrodes are rather limited. Mohora et al. (2012) reported that natural organic matter is simultaneously removed with As from groundwater by EC using Al anodes. It should be noted that we did not measure concentrations of phosphates, silicates, sulfates and natural organic matter, and hence their influence on As removal is not discussed. However, this analysis should serve as a useful starting point and these interferences may be investigated in future research. The experimental aluminum dose in the presence of hypochlorite did significantly improve when the current density was increased from 4 to 5 mA cm–2 and at 4 mA cm–2 it did not significantly differ in the absence and presence of hypochlorite. These random discrepancies are typically observed in electrodissolution processes. The experimental aluminum dose and aluminum dose rate in the presence and absence of hypochlorite at 6 mA cm–2 were lower than those obtained at 4 and 5 mA cm–2 due to massive generation of electrolytic gases and anode passivation. EC at 46 mA cm–2 did not lead to further removal of As (data not shown). The random discrepancies for the experimental dose and dose rate of aluminum at each flow rate did not systematically improve with increasing current density (Table 3); these discrepancies indicate a lack of correlation between As removal and the current density and flow rate.
5. Conclusions We systematically investigated arsenic removal from Bajío groundwater by EC using aluminum as a sacrificial anode in a pre-pilot-scale continuous filter press cell. The influence of current density, mean linear flow rate and hypochlorite addition on the As removal efficiency was analyzed. Poor removal of total As by EC in the absence of hypochlorite at current densities of 4, 5 and 6 mA cm–2 and mean linear flow rates of 0.91–4.55 cm s–1 was due to a mixture of arsenite and arsenate. Arsenic removal is more efficient when arsenite is oxidized to arsenate by addition of hypochlorite at a concentration typically used for disinfection. Arsenate removal by EC might involve adsorption on aluminum hydroxides generated in the process.
6
O.J. Flores et al. / Chemical Engineering Science 97 (2013) 1–6
The resulting sample meets the Mexican arsenic standard of ≤ 25 μg L–1. EC investigations revealed that mean linear flow rates between 0.91 and 2.73 cm s–1 at current densities of 4–6 mA cm–2 yielded samples that met the Mexican standard for arsenic in water. Under these conditions, aluminum doses and dose rates of 6–31 mg L–1 and 0.26–0.58 mg L–1 s–1, respectively, were suitable. Complete arsenic removal by EC was achieved at 5 mA cm–2 and 0.91 cm s–1, with electrolytic energy consumption of 3.9 kWh m–3. The best EC in terms of energy consumption was obtained at 4 mA cm–2 and 1.82 cm s–1, with energy consumption of 1.2 kWh m–3 to decrease arsenic from 50 to 17 μg L–3. An increase in current density to 46 mA cm–2 did not improve the EC process any further. This is attributed to massive generation of electrolytic gases and anode passivation. Acknowledgments We are grateful to CONACYT and CONCYTEG for financial support under the FOMIX GTO-2011-C03-162757 project. We thank Cynthia Montaudon for her help in revising the manuscript. We would also like to thank Universidad de Guanajuato for financial support. References American Public Health Association (APHA), 1995. Standard Methods for the Examination of Water and Wastewater, 19th ed. American Public Health Association/American Water Works Association/Water Pollution Control Federation, New York. Amirtharajah, A., Clark, M.M, Trussell, R.R., 1991. Mixing in Coagulation and Flocculation. , American Water Works Association Research FoundationDenver. Balasubramanian, N., Kojima, T., Ahmed Basha, C., Srinivasakannan, C., 2009. Removal of arsenic from aqueous solution using electrocoagulation. J. Hazard. Mater. 167, 966–969. Chen, X., Chen, G., Yue, P.L., 2002. Investigation on the electrolysis voltage of electrocoagulation. Chem. Eng. Sci. 57, 2449–2455. García-Lara, A.M., Montero-Ocampo, C., 2010. Improvement of arsenic electroremoval from underground water by lowering the interference of other ions. Water Air Soil Pollut. 205, 237–244.
Ghurye, G., Clifford, D., 2001. Laboratory Study on the Oxidation of Arsenic III to Arsenic V. US EPA, Cincinnati. Hansen, H.K., Henrik, K., Nunez, P., Jil, C., 2008. Removal of arsenic from wastewaters by airlift electrocoagulation Part 2: Continuous reactor experiments. Sep. Sci. Technol. 43, 3663–3675. Hansen, H.K., Nuñez, P., Raboy, D., Schippacasse, I., Grandon, R., 2007. Electrocoagulation in wastewater containing arsenic: comparing different process designs. Electrochim. Acta 52, 3464–3470. Hernández, J.R., Nava, J.L., Carreño, G., Martínez, F., 2010. Removal of arsenic (V) by electro-coagulation process, in: Birkle, P., Torres-Alvarado, I.S. (Eds.), Proceedings of the 13th International Symposium on Water–Rock Interaction (WRI-13), CRC Press/Taylor and Francis, London, 2010, pp. 583–585. Kobya, M., Gebeloglu, U., Ulu, F., Oncel, S., Demirbas, E., 2011. Removal of arsenic from drinking water by the electrocoagulation using Fe and Al electrodes. Electrochim. Acta 56, 5060–5070. Martínez-Villfañe, J.F., Montero-Ocampo, C., García-Lara, A.M., 2009. Energy and electrode consumption analysis of electrocoagulation for the removal of arsenic from groundwater. J. Hazard. Mater. 172, 1617–1622. Mohora, E., Roncevic, S., Dalmacija, B., Agbaba, J., Watson, M., Karlovic, E., Dalmacija, M., 2012. Removal of natural organic matter and arsenic from underwater electrocoagulation/flotation continuous flow reactor. J. Hazard. Mater. 235–236, 257–264. Parga, J.R., Cocke, D.L., Valenzuela, J.L., Gomes, J.A., Kesmez, M., Irwin, G., Moreno, H., Michael, W., 2005. Arsenic removal via electrocoagulation from heavy metal contaminated groundwater in La Comarca, Lagunera, Mexico. J. Hazard. Mater. 124, 247–254. Pourbaix, M., 1974. Atlas of Electrochemical Equilibria in Aqueous Solutions. NACEHouston. Ramos, Y., Ramírez, F., Martínez, J., Muñoz, D., Rodríguez, I., Carreño, G., 2011. Affectations of the drainage of mines abandoned to the supply of drinking water in Guanajuato, Mexico. Macla 15, 175–176. Rios, G.B., Almeraya, F., Herrera, M.T.A., 2005. Electrode passivation in the electrocoagulation process. Port. Electrochim. Acta 23, 17–34. Shen, F., Chen, X., Gao, P., Chen, G., 2003. Electrochemical removal of fluoride ions from industrial wastewater. Chem. Eng. Sci. 58, 987–993. Smedley, P.L, Kinniburgh, D.G., 2002. A review of the source, behaviour and distribution of arsenic in natural waters. Appl. Geochem. 17, 517–568. Smith, A.H., Marshall, G., Yuan, Y., Ferreccio, C., Liaw, J., von Ehrenstein, O., Steinmaus, C., Bates, M.N., Selvin, S., 2006. Increased mortality from lung cancer and bronchiectasis in young adults after exposure to arsenic in utero and in early childhood. Environ. Health Perspect. 114, 1293–1296. van Genuchten, C.M., Addy, S.E.A., Peña, J., Gadjil, A.J., 2011. Removing arsenic from synthetic groundwater with iron electrocoagulation: an Fe and As K-edge EXAFS study. Environ. Sci. Technol. 46, 986–994. Wan, W., Pepping, T.J., Banerji, T., Chaudhari, S., Giammar, D.E., 2011. Effects of water chemistry on arsenic removal from drinking water by electrocoagulation. Water Res. 45, 384–392.
Contents lists available at SciVerse ScienceDirect
Chemical Engineering Science journal homepage: www.elsevier.com/locate/ces
Arsenic removal from groundwater by electrocoagulation in a pre-pilot-scale continuous filter press reactor Omar J. Flores a, José L. Nava a,n, Gilberto Carreño a, Enrique Elorza b, Francisco Martínez a a
Departamento de Ingeniería Geomática e Hidráulica, Universidad de Guanajuato, Av. Juárez 77, Zona Centro, C.P. 36000 Guanajuato, Guanajuato, Mexico Departamento de Ingeniería en Minas Metalurgia y Geología, Universidad de Guanajuato, ExHacienda San Matías S/N, C.P. 36020 Guanajuato, Guanajuato, Mexico
b
H I G H L I G H T S
Arsenic removal from groundwater by electrocoagulation. Aluminum as sacrificial anodes. Electrolyzes at different flow rates and current densities. Electrocoagulation led to 100% arsenic removal (50 μg L−1) with 3.9 kWh m−3. The influence of hypochlorite on arsenic removal efficiency was analyzed.
art ic l e i nf o
a b s t r a c t
Article history: Received 19 November 2012 Received in revised form 6 April 2013 Accepted 15 April 2013 Available online 24 April 2013
We investigated arsenic removal from groundwater by electrocoagulation (EC) using aluminum as the sacrificial anode in a pre-pilot-scale continuous filter press reactor. The groundwater was collected at a depth of 320 m in the Bajío region in central Mexico (arsenic 50 μg L–1, carbonates 40 mg L–1, hardness 80 mg L–1, pH 7.5 and conductivity 150 μS cm–1). The influence of current density, mean linear flow and hypochlorite addition on the As removal efficiency was analyzed. Poor removal of total arsenic (o 60 %) in the absence of hypochlorite is due to a mixture of arsenite (HAsO2(aq) and H3AsO3(aq)) and arsenate (HAsO42−). Arsenic removal is more efficient when arsenite is oxidized to arsenate by addition of hypochlorite at a concentration typically used for disinfection (1 mg L–1). Arsenate removal by EC might involve adsorption on aluminum hydroxides generated in the process. Complete arsenate removal by EC was satisfactory at a current density of 5 mA cm–2 and mean linear flow of 0.91 cm s–1, with electrolytic energy consumption of 3.9 kWh m3. & 2013 Elsevier Ltd. All rights reserved.
Keywords: Arsenic removal Aluminum sacrificial anode Dissolution Electrolysis Coagulation Precipitation
1. Introduction Mining activities have historically led to environmental pollution. The main elements in groundwater contamination are heavy metals and metalloids such as lead, copper, zinc, iron, and arsenic produced by acid rock drainage (Ramos et al., 2011). Dissolved arsenic (As) was recently found in groundwater in the Bajío region (Guanajuato, central Mexico) at concentrations of 50–155 μg L–1. This problem was caused by arsenic mobilization in mine tailings dams (Hernández et al., 2010) and natural weathering of arseniccontaining rocks (Ramos et al., 2011). Mining was once the predominant economic activity in the region.
n
Corresponding author. Tel.: +52 4731020100x2289; fax: +52 4731020100x2209. E-mail addresses: [email protected], [email protected] (J.L. Nava).
0009-2509/$ - see front matter & 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ces.2013.04.029
Arsenic-contaminated water is a significant problem owing to its toxicity. Long-term exposure to arsenic leads to chronic health problems such as hyperpigmentation and keratosis of the hands and feet; it also causes bladder, lung, skin, kidney, liver, and prostate cancer (Smith et al., 2006). Considering the high toxicity of arsenic, the World Health Organization (WHO) and the US Environmental Protection Agency have set a maximum acceptable level of 10 μg L–1 for arsenic in drinking water, which differs from the recommendation of the Mexican authorities (25 μg L–1). Removal of arsenic from large volumes of water is generally performed by adding chemical coagulants such as aluminum or iron sulfate. However, owing to low concentrations of dissolved As, this process produces large amounts of sludge because SO4– anions consume 50% of the coagulant (Rios et al., 2005). This is the main reason why electrocoagulation (EC) is typically used, in which aluminum or iron, when dissolved electrolytically, supports the coagulation process and decreases the amount of sludge. EC is
2
O.J. Flores et al. / Chemical Engineering Science 97 (2013) 1–6
effective in destabilizing dispersed fine particles and anions contained in water (Chen et al., 2002). Flocs of metallic hydroxides are formed and remove dissolved species (Chen et al., 2002; Shen et al., 2003; Hernández et al., 2010; Mohora et al., 2012). Arsenic species in water from deep wells (pH∼7.5) include innocuous arsenite (HAsO2 and H3AsO3, oxidation state III) and arsenate anions (HAsO42–, oxidation state V) (Pourbaix, 1974; Smedley and Kinniburgh, 2002). Arsenate at neutral pH is more susceptible to EC removal (Hansen et al., 2007; Hernández et al., 2010; Wan et al., 2011; Mohora et al., 2012; Balasubramanian et al., 2009) because arsenate anions adsorb onto flocs of iron (Hansen et al., 2007, 2008; Kobya et al., 2011; Wan et al., 2011) and aluminum hydroxides (Hansen et al., 2008; Hernández et al., 2010; Kobya et al., 2011; Mohora et al., 2012). Martínez-Villafañe et al. (2009) and Kobya et al. (2011) reported that arsenate removal occurs via ligand exchange with ferric hydroxides. According to van Genuchten et al. (2011), As(V) forms inner-sphere complexes with the surface of iron oxides. However, reports on the mechanism of As removal by aluminum flocs are rather limited. Hansen et al. (2007) reported that oxidation of As(III) to As(V) by ferric ions and/or oxygen bubbling is convenient. The electrode materials commonly used for As removal by EC are mainly iron (Parga et al., 2005; Hansen et al., 2007, 2008; García-Lara and Montero-Ocampo, 2010; Kobya et al., 2011; Wan et al., 2011) and aluminum (Hernández et al., 2010; Kobya et al., 2011; Mohora et al., 2012). Kobya et al. (2011) demonstrated that aluminum was better than iron electrodes at removing arsenic from drinking water. We obtained similar results in a previous study (Hernández et al., 2010). There are literature reports on EC removal of As using Al electrodes from synthetically prepared water (Hernández et al., 2010; Kobya et al., 2011) and groundwater (Mohora et al., 2012) containing high arsenic concentrations. The results differ because a great number of species that are present in groundwater interfere with arsenic removal, besides reactor operating parameters such as current density, flow rate and geometry. The aim of this study was to decrease the arsenic concentration in groundwater from the Bajío region by EC using aluminum as the sacrificial anode in a pre-pilot-scale continuous filter press reactor. The influence of current density, mean linear flow rate and hypochlorite addition on the As removal efficiency was analyzed. The energy consumption for electrolysis was also estimated.
2. Electrochemical and chemical processes during As removal by EC using aluminum electrodes EC involves in situ generation of coagulants by electrodissolution of aluminum electrodes. Aluminum cations are generated at the anode and hydrogen gas is evolved at the cathode. This process generates aluminum hydroxide, which is believed to adsorb As (Hernández et al., 2010; Kobya et al., 2011; Mohora et al., 2012). The electrode and chemical reactions in neutral solution (pH∼7) are shown below. At the anode, electrodissolution generates aluminum ions (Al3+), which are transformed to aluminum hydroxide and aluminum oxide in the bulk Al(s)-Al3++3e– 3+
Al +3H2O-Al(OH)3(s)+3H
(1) +
2Al3++3H2O-Al2O3(s)+6H+
Al(OH)3 and Al2O3 flocs are believed to adsorb HAsO2− 4 (Hernández et al., 2010; Kobya et al., 2011; Mohora et al., 2012) 2– Al(OH)3(s)+HAsO2– 4 -[Al(OH)3 HAsO4 ](s)
(4)
2– Al2O3(s)+HAsO2– 4 -[Al2O3 HAsO4 ](s)
(5)
We previously reported that arsenite removal may occur by sweep coagulation, in which interaction occurs between As(III) and Al(OH)3 and Al2O3 flocs (Hernández et al., 2010). However, reports on the mechanism of As removal by aluminum flocs are rather limited. A deeper understanding of the As interaction with aluminum precipitates requires additional research. At the aluminum cathode, hydrogen gas is released 3H2O+3e–-1.5H2+3OH–
(6)
The solution at the cathode interface becomes alkaline over time. OH– ions migrate and diffuse away from the cathode to the anode, thus favoring water formation by reaction with protons from reaction (2) to (3) OH–+H+-H2O
(7)
The major problem with aluminum anodes is passivation due to Al2O3 precipitation, which leads to high anode and cell potentials and increases the energy consumption and cost of EC (Kobya et al., 2011; Mohora et al., 2012). Passivation can be controlled at low current densities in combination with convection (turbulent flow conditions), which favors Al3+ transport away from the surface to the bulk solution. In addition, cathodes of the same material can be used to electrodissolve Al2O3 by periodic current reversal (Mohora et al., 2012), which allows even consumption of the aluminum electrodes during the process.
3. Experimental 3.1. Solutions Groundwater for electrochemical treatment was collected from deep wells (320 m) located in the Bajío region (Guanajuato, México). Table 1 lists the characteristics of the sample. The groundwater contained 50 μg L–1 total As, which exceeds the Mexican standard for arsenic in water (25 μg L–1). It is believed that As(III) is predominant in this anaerobic groundwater. Water samples were obtained from the wells before the disinfection process, which usually involves hypochlorite addition. 3.2. Pre-pilot-scale filter press cell Fig. 1 shows a schematic of the experimental set-up. The system consists of a continuous pre-pilot-scale filter press cell in which the coagulant is produced. The resulting solution
Table 1 Characteristics of the groundwater samples collected at a depth of 320 m in the Bajío region (Guanajuato, México).
(2)
Parameter
Fresh
After hypochlorite additiona
(3)
Total arsenic (μg L–3) Carbonates (mg L–3) Hardness (mg L–1) pH Conductivity (μS cm–1)
59 40 80 7.5 150
59 40 80 8.1 392
The oxidation states of As in water are As(III) and As(V). As(V) is typically dominant in aerobic surface waters, while As(III) is found in anaerobic groundwater. At neutral pH, the predominant As (V) species has a net negative charge (HAsO42−), while As(III) species generally have no net charge (HAsO2 and H3AsO3) (Pourbaix, 1974).
a Hypochlorite was added at 1 mg L–1, the concentration typically used for disinfection of groundwater.
O.J. Flores et al. / Chemical Engineering Science 97 (2013) 1–6
3
electrode (SSE; Radiometer model XR200), 0.615 V vs. SHE/V. The potential of this electrode was determined using a high-impedance multimeter (Agilent model 34401A). All electrode potentials are reported with respect to SHE. 3.3. Methodology EC studies were carried out using the experimental set-up shown in Fig. 1. EC was performed under different hydrodynamic conditions with a mean linear flow rate of 0.91–4.55 cm s–1 and current density of 4, 5 and 6 mA cm–2. Each resulting solution was immediately passed to the test jar and mixed at slow speed (30 rpm) for 15 min for aggregate growth. The aggregates were allowed to precipitate in static solution for approximately 1 h. Arsenic was analyzed in the resulting clarified solution. After dissolution of the floc, aluminum was also analyzed. 3.4. Analytical procedure
Fig. 1. Electrical and flow circuit for the EC process in a filter press reactor.
Table 2 Dimensions of the electrocoagulation reactor. Reactor volume, Vr (cm3) Height, B (cm) Channel length, L (cm) Chanel width, S (cm) Total length of six channels, LT (cm) Number of channels Anode area in each channel in contact with solution (cm2) Cathode area in each channel in contact with solution (cm2) Cross-sectional area, AT (cm2)
88.94 3.05 8.1 0.6 48.6 6 24.7 24.7 1.83
(mixture of water and coagulant) is passed to a test jar to induce flocculation and adsorption of arsenic on aluminum flocs. The arsenic is then precipitated and the clarified solution is analyzed. The reactor was switched in monopole configuration. Three aluminum electrodes (3.05 cm 8.10 cm 0.30 cm) were used as anodes and four similar electrodes were used as cathodes. The electrodes were spaced at 0.60 cm using propylene separators. The reactor dimensions are listed in Table 2. This pre-pilot-scale reactor was developed in our laboratory for removal of arsenic from synthetic solutions (Hernández et al., 2010). The serpentine array induces fluid turbulence, enhancing mass transport of the coagulant from the anode to the bulk and avoiding both alumina precipitation and anode passivation. The reactor was connected to a hydraulic system consisting of a centrifuge pump (model MDX-MT-3) of 0.25 HP and a flow meter (model F-44250LH-8) with a capacity of 0–1 L min–1. The PVC pipes connecting the reservoir to the pump were 0.3 in. in diameter. The valves and connections were also made of PVC. A BK Precision power supply (model 1090) was used for electrolysis tests. The cell potential was measured using the power supply. The aluminum electrode potential was measured as the difference between aluminum and a saturated mercurous sulfate reference
Arsenic concentrations in the samples were measured using a standard method (APHA, 1995) with an atomic absorption spectrometer (Perkin-Elmer AAnalist 200) equipped with a manual hydride generator at 188.9 nm. The As detection limit was 0.1 μg L–1 and the error for triplicate analyses was within 2%. The kinetics of aluminum dissolution was followed by dissolving the sludge at pH 2 and then quantifying aluminum ions by AA. Carbonates and hardness were analyzed according to standard methods (APHA, 1995). Conductivity and pH measurements were carried out on a waterproof instrument (HANNA model HI 991300). All chemical reagents were of analytical grade. Each individual experiment was performed at least three times and the results were averaged.
4. Analysis of results and discussion 4.1. Arsenic removal from fresh groundwater Fig. 2 shows the residual concentration of arsenic (CAs) in fresh groundwater after EC as a function of the mean linear flow rate (u) at 4, 5 and 6 mA cm–2. The experimental and theoretical aluminum doses are also shown. At 4 mA cm–2, CAs linearly increased between 23 and 50 μg L–1 as a function of u in the interval 0.91–3.64 cm s–1 (Fig. 2a), indicating that arsenic removal decreases as u increases. This is due to depletion of the experimental Al(III) dose from 42 to 5 mg L–1. The quasi-constant CAs at u ≥ 3.64 cm s–1 is due to the low dose of aluminum (CAl(III) o4 mg L–1). The experimental aluminum dosage, obtained by analysis of the sludge, was lower than the theoretical amount added, calculated as jLMW C AlðIIIÞðNÞ ¼ ð1 106 ÞðNÞ ð8Þ nFSu where CAl(III) is the aluminum concentration (mg L–1), j is the current density (A cm–2), L is the length of one channel (cm), MW is the molecular weight of aluminum (26.98 g mol–1), n is the number of electrons exchanged (n ¼3), F is the Faraday constant (96485 C mol–1), S is the channel width (cm), u is the mean linear flow rate (cm s–1), N is the number of channels, and 1 106 is a conversion factor used to obtain the aluminum concentration in mg L–1. The difference between the experimental and theoretical values can probably be explained, on one hand, by oxygen evolution according to Eq. (9), which typically occurs simultaneously with Eq. (1) (Kobya et al., 2011), and, on the other hand, to
4
O.J. Flores et al. / Chemical Engineering Science 97 (2013) 1–6
1991). The arsenite species can be oxidized to HAsO42− by addition of hypochlorite. The effectiveness of this oxidation is well known (Ghurye and Clifford, 2001). The experimental aluminum dose decreased as the current density increased (Fig. 2a–c), contrary to what was expected. This could be related, on one hand, to enhanced oxygen evolution and, on the other hand, to simultaneous precipitation of Al2O3 at the anode (Kobya et al., 2011; Mohora et al., 2012), even under such hydrodynamic conditions. The change in shape for the experimental aluminum dose curve in Fig. 2a–c may be attributable to the formation of a layer of Al2O3 at the anode, which may passivate the anode against progressive electrodissolution. However, this analysis may be inaccurate. X-Ray studies would be helpful in identifying the Al2O3 layer, but this was beyond the scope of the present study. 4.2. Arsenic removal from groundwater after hypochlorite addition Fig. 3 shows CAs in pre-conditioned drinking water after EC as a function of u at 4, 5 and 6 mA cm–2. The experimental and theoretical aluminum doses are also shown.
Fig. 2. Influence of the mean linear flow rate on the residual arsenic concentration and aluminum dose for fresh groundwater (initial parameters: arsenic 50 μg L–1, carbonates 40 mg L–1, hardness 80 mg L–1, pH 7.5 and conductivity 150 μS cm–1) at a current density of (a) 4, (b) 5 and (c) 6 mA cm–2. The anode area for six channels was 148.2 cm2.
anode passivation due to Al2O3 precipitation (Kobya et al., 2011; Mohora et al., 2012). H2O-0.5O2+2H++2e–
(9)
At 5 mA cm–2, CAs increased from 23 to 50 μg L–1 as a function of u in the interval 1.82–4.55 cm s–1 (Fig. 2b), indicating that arsenic removal also decreases as u increases at this current density. Again, the experimental aluminum dose was lower than the theoretical one. At 6 mA cm–2, the behavior of CAs as a function of u differed for the ranges 1.82–2.73, 2.73–3.64 and 3.64–4.55 cm s–1. (Fig. 2c). However, the increase in CAs from 20 to 42 μg L–1 at 1.82– 4.55 cm s–1 was significant. The experimental aluminum dose was lower than the theoretical one. The poor EC removal of arsenic can be attributed to the mixture of HAsO2, H3AsO3 and HAsO2− 4 . The third species is probably more susceptible to removal by adsorption on Al(OH)3 and Al2O3 flocs (Hernández et al., 2010; Kobya et al., 2011; Mohora et al., 2012), since these flocs have a positive zeta potential (Amirtharajah et al.,
Fig. 3. Influence of the mean linear flow rate on the residual arsenic concentration and aluminum dose for groundwater after addition of hypochlorite (initial parameters: arsenic 50 μg L–1, carbonates 40 mg L–1, hardness 80 mg L–1, hypochlorite 1 mg L–1, pH 8.1 and conductivity 392 μS cm–1) at a current density of (a) 4, (b) 5 and (c) 6 mA cm–2. The anode area for six channels was 148.2 cm2.
O.J. Flores et al. / Chemical Engineering Science 97 (2013) 1–6
At 4 mA cm–2, CAs increased linearly between 3 and 38 μg L–1 as a function of u in the interval 0.91–4.55 cm –1 due to a decrease in the experimental aluminum dose from 17 to 1 mg L–1 (Fig. 3a). However, the CAs values are lower than those obtained for fresh groundwater at 4 mA cm–2 (Fig. 2a). This confirms the hypothesis that HAsO42− is probably removed by adsorption on Al(OH)3 and Al2O3 flocs (Hernández et al., 2010; Kobya et al., 2011; Mohora et al., 2012). X-Ray absorption spectroscopy and X-ray photoelectron spectroscopy would be helpful in investigating the local environment of Al and As in EC flocs generated in Bajío groundwater, but this was beyond the scope of the present study. It should be noted that CAs satisfies the WHO limit of ≤10 μg L–1 at u ¼0.91 cm s–1 and the Mexican standard of ≤25 μg L–1 at u ¼0.91– 1.82 cm s–1. At 5 mA cm–2, the behavior of CAs as a function of u differed for the ranges 0.91–1.81, 1.81–2.73 and 2.73–4.55 cm s–1 (Fig. 3b). CAs is close to zero at u¼0.91 cm s–1, 22 μg L–1 at u¼ 1.81 cm s–1 and 23 μg L–1 at u¼ 2.73 cm s–1. The latter slight variation in CAs can be attributed to the almost constant experimental aluminum dose of 6–7 mg L–1. CAs then increased linearly from 23 to 33 μg L–1 at 2.73–4.55 cm s–1 due to depletion of the experimental aluminum dose. At 6 mA cm–2, CAs as a function of u showed different behavior at 0.91–3.64 and 3.64–4.55 cm s–1 (Fig. 3c). The increase in CAs from 11 to 33 μg L−1 at 0.91–3.64 cm s–1 was significant, and CAs remained constant at 33 μg L–1 with a further increase to u¼4.55 cm s–1. In the presence of hypochlorite, the experimental dose of aluminum was lower than the corresponding theoretical dose (Fig. 3a–c). CAs after EC was lower in the presence of hypochlorite than without hypochlorite, indicating that arsenic removal is more efficient when arsenite is oxidized to arsenate. Arsenate removal by EC might involve adsorption on aluminum hydroxides generated in the process. Although the initial groundwater As concentration of 50 μg L–1 complies with safety standards in many countries, the Mexican authorities want to reduce the maximum acceptable level of arsenic in drinking water from 25 to 10 μg L–1 because of chronic health problems in the population. Experiments were performed at 4 and 5 mA cm–2 and 0.81 cm s–1 (Table 3) with the aim of obtaining residual arsenic concentrations of o10 μg L–1. Table 3 summarizes EC results that satisfy the Mexican arsenic standard of ≤25 μg L–1, including the cell potential (Ecell) and electrolytic energy consumption, evaluated as Econs ¼
Ecell I 3:6 B S u
ð10Þ
where I is the current intensity during electrolysis (C s–1), B is the channel height (cm), S is the channel width (cm) and 3.6 is a conversion factor used to obtain Econs in units of kWh m–3. Table 3 reveals that EC at 4 mA cm–2 satisfied the Mexican standard for arsenic at flow rates between 0.91 and 1.82 cm s–1, Table 3 Residual arsenic concentrations after hypochlorite addition satisfying the Mexican standard (CAs≤25 μg L–1), as well as the aluminum dose (CAl(III)) and dose rate (QAl(III) ¼CAl(III)u/LT), cell potential and electrolytic energy consumption. Current Density
u (cm s–1)
CAs (μg L–1)
CAl(III) (mg L–1)
QAl(III) (mg L–1 s–1)
Ecell (V)
Econs (kWh m–3)
4 mA cm−2
0.91 1.82 2.73 0.91 1.82 2.73 0.91 1.82 2.73
3 17 – 0 22 23 11 22 25
17 9 – 31 7 6 16 10 7
0.32 0.34 – 0.58 0.26 0.34 0.30 0.37 0.39
8.4 7.9 – 10.6 8.9 8.9 11.6 11.2 10.8
2.5 1.2 – 3.9 1.6 1.2 5.3 2.5 1.6
5 mA cm−2
6 mA cm−2
5
corresponding to an aluminum dose of 9–17 mg L–1 and an aluminum dose rate of QAl(III) ¼0.32–0.34 mg L–1 s–1, evaluated as Q AlðIIIÞ ¼
C AlðIIIÞ u LN
ð11Þ
EC at 5 mA cm–2 satisfied the Mexican As standard at flow rates between 0.91 and 2.73 cm s–1, corresponding to aluminum doses of 6–31 mg L–1and aluminum dose rates of 0.34–0.58 mg L–1 s–1. EC at 6 mA cm–2 satisfied the Mexican As standard at flow rates of 0.91–2.73 cm s–1, corresponding to aluminum doses of 7–16 mg L–1 and aluminum dose rates of 0.30–0.39 mg L–1 s–1. It should be noted that the cell potential decreased as a function of u at each current density, indicating that aluminum ions are transported away from the surface to the bulk solution under such turbulent flow conditions, decreasing Al2O3 precipitation on the anode. The best EC in terms of energy consumption was obtained at 4 mA cm–2 and u¼1.82 cm s−1, for which energy consumption of 1.2 kWh m–3 reduced arsenic from 50 to 17 μg L–1. Data for carbonates, hardness, pH and conductivity are not presented, since these parameters remained almost constant. Studies have shown that several dissolved species present in water samples can interfere with EC arsenic removal. For example, Wan et al. (2011) reported that phosphates in synthetic groundwater interfere with EC arsenic removal using Fe as sacrificial anodes, because phosphate and As species compete with surface iron oxide sites. The authors also showed that silicates and sulfates did not affect As removal. However, reports regarding As removal by EC using Al electrodes are rather limited. Mohora et al. (2012) reported that natural organic matter is simultaneously removed with As from groundwater by EC using Al anodes. It should be noted that we did not measure concentrations of phosphates, silicates, sulfates and natural organic matter, and hence their influence on As removal is not discussed. However, this analysis should serve as a useful starting point and these interferences may be investigated in future research. The experimental aluminum dose in the presence of hypochlorite did significantly improve when the current density was increased from 4 to 5 mA cm–2 and at 4 mA cm–2 it did not significantly differ in the absence and presence of hypochlorite. These random discrepancies are typically observed in electrodissolution processes. The experimental aluminum dose and aluminum dose rate in the presence and absence of hypochlorite at 6 mA cm–2 were lower than those obtained at 4 and 5 mA cm–2 due to massive generation of electrolytic gases and anode passivation. EC at 46 mA cm–2 did not lead to further removal of As (data not shown). The random discrepancies for the experimental dose and dose rate of aluminum at each flow rate did not systematically improve with increasing current density (Table 3); these discrepancies indicate a lack of correlation between As removal and the current density and flow rate.
5. Conclusions We systematically investigated arsenic removal from Bajío groundwater by EC using aluminum as a sacrificial anode in a pre-pilot-scale continuous filter press cell. The influence of current density, mean linear flow rate and hypochlorite addition on the As removal efficiency was analyzed. Poor removal of total As by EC in the absence of hypochlorite at current densities of 4, 5 and 6 mA cm–2 and mean linear flow rates of 0.91–4.55 cm s–1 was due to a mixture of arsenite and arsenate. Arsenic removal is more efficient when arsenite is oxidized to arsenate by addition of hypochlorite at a concentration typically used for disinfection. Arsenate removal by EC might involve adsorption on aluminum hydroxides generated in the process.
6
O.J. Flores et al. / Chemical Engineering Science 97 (2013) 1–6
The resulting sample meets the Mexican arsenic standard of ≤ 25 μg L–1. EC investigations revealed that mean linear flow rates between 0.91 and 2.73 cm s–1 at current densities of 4–6 mA cm–2 yielded samples that met the Mexican standard for arsenic in water. Under these conditions, aluminum doses and dose rates of 6–31 mg L–1 and 0.26–0.58 mg L–1 s–1, respectively, were suitable. Complete arsenic removal by EC was achieved at 5 mA cm–2 and 0.91 cm s–1, with electrolytic energy consumption of 3.9 kWh m–3. The best EC in terms of energy consumption was obtained at 4 mA cm–2 and 1.82 cm s–1, with energy consumption of 1.2 kWh m–3 to decrease arsenic from 50 to 17 μg L–3. An increase in current density to 46 mA cm–2 did not improve the EC process any further. This is attributed to massive generation of electrolytic gases and anode passivation. Acknowledgments We are grateful to CONACYT and CONCYTEG for financial support under the FOMIX GTO-2011-C03-162757 project. We thank Cynthia Montaudon for her help in revising the manuscript. We would also like to thank Universidad de Guanajuato for financial support. References American Public Health Association (APHA), 1995. Standard Methods for the Examination of Water and Wastewater, 19th ed. American Public Health Association/American Water Works Association/Water Pollution Control Federation, New York. Amirtharajah, A., Clark, M.M, Trussell, R.R., 1991. Mixing in Coagulation and Flocculation. , American Water Works Association Research FoundationDenver. Balasubramanian, N., Kojima, T., Ahmed Basha, C., Srinivasakannan, C., 2009. Removal of arsenic from aqueous solution using electrocoagulation. J. Hazard. Mater. 167, 966–969. Chen, X., Chen, G., Yue, P.L., 2002. Investigation on the electrolysis voltage of electrocoagulation. Chem. Eng. Sci. 57, 2449–2455. García-Lara, A.M., Montero-Ocampo, C., 2010. Improvement of arsenic electroremoval from underground water by lowering the interference of other ions. Water Air Soil Pollut. 205, 237–244.
Ghurye, G., Clifford, D., 2001. Laboratory Study on the Oxidation of Arsenic III to Arsenic V. US EPA, Cincinnati. Hansen, H.K., Henrik, K., Nunez, P., Jil, C., 2008. Removal of arsenic from wastewaters by airlift electrocoagulation Part 2: Continuous reactor experiments. Sep. Sci. Technol. 43, 3663–3675. Hansen, H.K., Nuñez, P., Raboy, D., Schippacasse, I., Grandon, R., 2007. Electrocoagulation in wastewater containing arsenic: comparing different process designs. Electrochim. Acta 52, 3464–3470. Hernández, J.R., Nava, J.L., Carreño, G., Martínez, F., 2010. Removal of arsenic (V) by electro-coagulation process, in: Birkle, P., Torres-Alvarado, I.S. (Eds.), Proceedings of the 13th International Symposium on Water–Rock Interaction (WRI-13), CRC Press/Taylor and Francis, London, 2010, pp. 583–585. Kobya, M., Gebeloglu, U., Ulu, F., Oncel, S., Demirbas, E., 2011. Removal of arsenic from drinking water by the electrocoagulation using Fe and Al electrodes. Electrochim. Acta 56, 5060–5070. Martínez-Villfañe, J.F., Montero-Ocampo, C., García-Lara, A.M., 2009. Energy and electrode consumption analysis of electrocoagulation for the removal of arsenic from groundwater. J. Hazard. Mater. 172, 1617–1622. Mohora, E., Roncevic, S., Dalmacija, B., Agbaba, J., Watson, M., Karlovic, E., Dalmacija, M., 2012. Removal of natural organic matter and arsenic from underwater electrocoagulation/flotation continuous flow reactor. J. Hazard. Mater. 235–236, 257–264. Parga, J.R., Cocke, D.L., Valenzuela, J.L., Gomes, J.A., Kesmez, M., Irwin, G., Moreno, H., Michael, W., 2005. Arsenic removal via electrocoagulation from heavy metal contaminated groundwater in La Comarca, Lagunera, Mexico. J. Hazard. Mater. 124, 247–254. Pourbaix, M., 1974. Atlas of Electrochemical Equilibria in Aqueous Solutions. NACEHouston. Ramos, Y., Ramírez, F., Martínez, J., Muñoz, D., Rodríguez, I., Carreño, G., 2011. Affectations of the drainage of mines abandoned to the supply of drinking water in Guanajuato, Mexico. Macla 15, 175–176. Rios, G.B., Almeraya, F., Herrera, M.T.A., 2005. Electrode passivation in the electrocoagulation process. Port. Electrochim. Acta 23, 17–34. Shen, F., Chen, X., Gao, P., Chen, G., 2003. Electrochemical removal of fluoride ions from industrial wastewater. Chem. Eng. Sci. 58, 987–993. Smedley, P.L, Kinniburgh, D.G., 2002. A review of the source, behaviour and distribution of arsenic in natural waters. Appl. Geochem. 17, 517–568. Smith, A.H., Marshall, G., Yuan, Y., Ferreccio, C., Liaw, J., von Ehrenstein, O., Steinmaus, C., Bates, M.N., Selvin, S., 2006. Increased mortality from lung cancer and bronchiectasis in young adults after exposure to arsenic in utero and in early childhood. Environ. Health Perspect. 114, 1293–1296. van Genuchten, C.M., Addy, S.E.A., Peña, J., Gadjil, A.J., 2011. Removing arsenic from synthetic groundwater with iron electrocoagulation: an Fe and As K-edge EXAFS study. Environ. Sci. Technol. 46, 986–994. Wan, W., Pepping, T.J., Banerji, T., Chaudhari, S., Giammar, D.E., 2011. Effects of water chemistry on arsenic removal from drinking water by electrocoagulation. Water Res. 45, 384–392.