Combined solar electrocoagulation and adsorption processes for Pb(II) removal from aqueous solution

Combined solar electrocoagulation and adsorption processes for Pb(II) removal from aqueous solution

Chemical Engineering & Processing: Process Intensification 143 (2019) 107619 Contents lists available at ScienceDirect Chemical Engineering & Proces...

3MB Sizes 0 Downloads 10 Views

Chemical Engineering & Processing: Process Intensification 143 (2019) 107619

Contents lists available at ScienceDirect

Chemical Engineering & Processing: Process Intensification journal homepage: www.elsevier.com/locate/cep

Combined solar electrocoagulation and adsorption processes for Pb(II) removal from aqueous solution

T

Farihahusnah Hussina, , Mohamed Kheireddine Arouaa,b, Małgorzata Szlachtac ⁎

a

Research Centre for Carbon Dioxide Capture and Utilisation (CCDCU), School of Science and Technology, Sunway University, Jalan Universiti, Bandar Sunway, 47500 Selangor Darul Ehsan, Malaysia b Department of Engineering, Lancaster University, Lancaster, LA1 4YW, UK c Faculty of Environmental Engineering, Wrocław University of Science and Technology, Wybrzeże Wyspiańskiego 27, 50-370 Wrocław, Poland

ARTICLE INFO

ABSTRACT

Keywords: Electrocoagulation Adsorption Combined treatment Response surface methodology Central composite design

A combination of electrocoagulation with other methods seems to have garnered much attention in the research area for the past decade to eliminate heavy metal ions from the synthetic and real wastewater effluents. Combining two various methods into a single system appears to be an efficient and promising approach for heavy metal removal, mainly due to their cost-effectiveness, simple operation and suitability for industrial applications. Solar photovoltaic systems have gained much attention because they make use of clean, renewable energy and make the treatment method cost-effective. In this regard, it is imperative to explore the potential of solar photovoltaic systems to remove heavy metals. A response surface methodology based on the central composite design (CCD) was employed to examine the effects of three independent variables such as pH, initial Pb(II) concentration and adsorbent dosage. The results indicated that the highest Pb(II) removal efficiency up to 99.88% can be achieved using the CCD model with the following optimum conditions: (1) pH: 6.01, (2) initial Pb (II) concentration: 15.00 mg/L and (3) adsorbent dosage: 2.50 g/L. Based on the results, the combined system offered an attractive alternative over the single electrocoagulation and adsorption treatment systems as it can produce high Pb(II) removal efficiency.

1. Introduction Water pollution is a serious issue that needs an immediate solution because of its detrimental impact on the environment. At present, about 80% of the world’s wastewater is discharged to the environment without adequate treatment. Preventing the discharge of untreated wastewater (especially from industrial activities) will not only create healthy ecosystems, but it can also advance sustainable growth [1]. The emission of heavy metal ions from factories into water reservoirs is the main contributor to water pollution. In general, metals are used to manufacture a broad range of products. Despite the important role of metals in various processes, some of these metals are discharged to the environment and they contribute a significant portion of hazardous wastes in effluent streams [2,3]. For this reason, it is crucial to reduce heavy metal concentrations in industrial wastewater up to the acceptable limits before being discharged to the environment. The Agency for Toxic Substances and Disease Registry (ATSDR) and the U.S. Environmental Protection Agency (USEPA) are the two federal public agencies in the United States of America that play a vital role in



protecting the people and the environment from diseases due to exposure towards hazardous substances [4]. The ATSDR has consolidated a comprehensive list on the toxicological profiles of each substance and placed second in the list of top 20 hazardous substances based on its frequency, toxicity and potential for human exposure [4]. Lead exposure has a variety of adverse effects on human health. For instance, lead can cause problems in the central and peripheral nervous systems and these effects are more pronounced on foetuses, pregnant women and young children [5,6]. Lead exposure can also cause headaches, loss of memory, dullness, poor attention span and hallucinations. These neurological effects are more pronounced in children than adults because children have immature immune systems, making them vulnerable to diseases. Moreover, prolonged lead exposure can cause undesirable side effects, including delirium, lack of coordination, paralysis and coma [7]. Lead exposure is relatively higher in some parts of the world, particularly in underdeveloped countries. The rapid growth of industrialisation has contributed to water pollution due to the discharge of wastewater streams containing lead ions. In recent years, a few small-scale industries are identified to be the major contributors of

Corresponding author. E-mail address: [email protected] (F. Hussin).

https://doi.org/10.1016/j.cep.2019.107619 Received 6 March 2019; Received in revised form 26 June 2019; Accepted 26 July 2019 Available online 02 August 2019 0255-2701/ © 2019 Elsevier B.V. All rights reserved.

Chemical Engineering & Processing: Process Intensification 143 (2019) 107619

F. Hussin, et al.

wastewater. This poses a significant hazard because there is insufficient control over wastewater discharge and treatment in these small-scale industries. Hence, it is essential to enforce stringent regulations on wastewater discharge and educate the public on environmental issues. Various methods have been established to eliminate lead ions from aqueous solutions, for instance, chemical precipitation, coagulationflocculation, electrocoagulation, reverse osmosis, adsorption, ion exchange and membrane filtration [8,9]. The main challenge in industrial wastewater treatment is to achieve a good trade-off between treatment efficiency and operating costs. Among the aforementioned techniques, electrocoagulation and adsorption have received much attention in the academic research community in the past decade owing to the benefits of these techniques. Electrocoagulation offers the following advantages: (1) simplicity of operation, (2) rapid sedimentation, (3) low sludge production and (4) environmental compatibility [10]. However, electrocoagulation using conventional power systems are not economically feasible because of their high capital investments and operating costs. In addition, electrocoagulation systems can be powered by renewable energy-based systems such as photovoltaic panels, which will then convert solar energy into electricity [11]. Solar photovoltaic (PV) panels are favourable compared to the conventional electrical power supplies because they make use of clean, renewable energy and make the treatment method cost-effective [12]. The adsorption technique is also proven to be effective in eliminating heavy metal ions. Adsorption is a promising technique to remove lead because of its low cost and high effectiveness, and more importantly, the process does not produce secondary pollutants [13]. Even though much progress has been made in the field of electrocoagulation and adsorption, more research is needed to boost the efficiency of wastewater treatment. Electrocoagulation and adsorption processes are typically carried out independently and therefore, combining two or more methods of treatment can promote lead removal efficiency. Combining different methods of wastewater treatment is also more economically viable. Numerous studies have been carried out over the years to combine electrocoagulation with other methods such as electro-flotation, electro-oxidation, ozone, photo-Fenton and adsorption [14,15]. A higher removal efficiency can be obtained when two treatment methods are combined to treat effluents compared to a single treatment method. A few studies that combine the electrocoagulation with other treatment techniques were reported in the literature and these findings are summarised in Table 1. Nevertheless, none of the studies are focused on combining electrocoagulation and adsorption processes to eliminate Pb(II) ions from aqueous solutions. It is believed that the benefits of the electrocoagulation and adsorption processes can be leveraged by combining them into a single system, which will significantly boost the lead removal efficiency compared with a single process. With this in mind, the objective of this study was to assess the performance of a novel wastewater treatment system in which solar-powered electrocoagulation was combined with adsorption to eliminate Pb(II) ions from aqueous

solutions. In consideration to investigate the interaction between the parameters and to optimise the treatment process, the use of a suitable experimental design is essential in finding the best operating conditions. Design of Experiments (DOE) is a practical and effective mathematical approach that uses the statistical methodology to analyse data and predict a product that can be applied in a variety of experimental conditions. DOE is an efficient problem-solving tool, which can enhance or optimise the performance of the product and process design with significant achievement in experimental time and cost [23]. There are many available techniques used in DOE and one of them is the Response Surface Methodology (RSM), which enables multi-variable optimisation [24]. RSM is a fast and cost-effective method for gathering research results and can identify important interactions that may be overlooked when experimenting with the classic one-factor-at-a-time (OFAT) approach [25]. The OFAT method requires large numbers of experiments and is inefficient in estimating the interaction between the parameters, especially when compared with the concurrent changing factor levels [26]. The main aim of the RSM utilisation is to simultaneously optimise the levels of these variables to achieve the best system performance. Central Composite Design (CCD) and Box-Behnken Design (BBD) are the experimental designs that have been used extensively among the various RSM methods [27]. BBD is a round, rotatable (or closely rotatable) second-order design. It is a three-level fractional factorial design consisting of center points at the edges of the cube [28,29]. CCD is suitable to fit both the linear and quadratic models and it can be used to identify the significant parameters that affect the response with a minimum number of trials [30]. CCD can also be used to examine the interaction effects of the parameters [27,31]. CCDs are generally rotatable or spherical, which offers advantages such as better suitability for blocking. CCD consists of factorial design (either full or fractional) with center points along with star points that extend the cuboidal region of the original factorial design. The extra points permit the estimation of curvature of the response across different factor levels [32]. Furthermore, the presence of these points provides the rotational workability, high accuracy, good predictability and enables the gathering of more data with lower number of experiments. CCD models becomes effective and popular for parameters optimisation among researchers [33,34]. In contrast, BBDs do not involve an embedded factorial design and the experimental levels which are placed at the midpoints of the hypercube [32]. For example, Ngan et al. [35], compared the performance of CCD and BBD methods to find the optimum formulation for fullerene loaded Nano-emulsions for cosmeceutical application. The authors reported that the CCD model predicts better responses that are closer to the actual values and produces better performance compared with the BBD model. The uniqueness of the current study is in the application of combined solar electrocoagulation and adsorption processes carried out at a low current density to attain high treatment efficiency in a cost-effective manner. In addition, CCD was applied to investigate the interaction

Table 1 Electrocoagulation process combined with other methods. Combination of processes

Effluent

Results

Ref 2

Electrocoagulation + Electrooxidation

Electroplating wastewater

Electrocoagulation + Adsorption

Textile wastewater

Electrocoagulation + Electrooxidation

Dairy wastewater

Electrocoagulation + Adsorption

Pistachio wastewater

Electrocoagulation + Ozone Electrocoagulation + Fenton/Photo-Fenton + Adsorption Electrocoagulation + Flotation

Textile wastewater Textile wastewater Aqueous solutions (Humic acid removal)

2

Optimum conditions (current density: 30 mA/m , reaction time: 20 min, COD (93.6 %)) Optimum conditions (current density: 71.4 A/m2, reaction time: 90 min, colour removal (96-98%)) Optimum conditions (current density: 3 A/m2, reaction time: 21 min, COD (66.4%)) Optimum conditions (current density: 100 A/m2, electrode distance: 6 cm, reaction time: 293 min, COD (43.7%)) Optimum conditions (reaction time: 18 min, Colour removal (95%)) Optimum conditions (reaction time: 45 min, COD (76%)) Optimum conditions (reaction time: 60 min, removal efficiency (81.2%))

[16] [17] [18] [19] [20] [21] [22]

Chemical Engineering & Processing: Process Intensification 143 (2019) 107619

F. Hussin, et al.

between the parameters and optimise the treatment process. Pb(II) removal efficiency was chosen as the response parameter and the pH, initial Pb(II) concentration and adsorbent dosage were selected as process variables. The optimal conditions for a complete Pb(II) removal were also determined by the CCD model.

data from our preliminary study (data not shown in this paper). This part of the study was conducted to investigate the suitable pH value to be applied for the adsorption kinetics. Statsoft Statistica software version 12 was used in data analysis. The nonlinear regression approach was the most viable method to select the optimum isotherm. This method gave minimum error distributions between predicted isotherms and the experimental data [42].

2. Materials and methods 2.1. Raw materials

2.3.2. Adsorption kinetic The adsorption kinetics tests were conducted for three hours at different Pb(II) concentrations (10, 30 and 50 mg/L) and temperatures (18, 28 and 38 °C). The range of process parameters was selected based on our preliminary study and literature review [37–39,42]. The experiments were performed with an initial pH of 6 (selected from the adsorption isotherm) using 4 g/L of adsorbent. In the kinetics study, the agitation speed used to stir Pb(II) solution was 250 rpm. In order to obtain a stable temperature throughout the process, a temperaturecontrolled water bath was utilised.

Granular oil palm shell activated carbon (OPSAC) was bought from Bravo Green Sdn. Bhd., Malaysia. The OPSAC was produced using the physical activation process. The activated carbon (as adsorbent) was grounded and then sieved, resulting in a fine adsorbent with a particle size of 700–850 μm. The adsorbent was subsequently washed with distilled water in order to eliminate impurities and then left to dry in a laboratory oven at a temperature of 110 °C overnight to eliminate moisture. 2.2. Chemicals

2.3.3. Desorption experiment To obtain the most effective desorption solution, 0.06 g of OPSAC was introduced into Erlenmeyer flasks that contained 0.05 L of 0.1 M hydrochloric acid (HCl), nitric acid (HNO3) and distilled water (H2O). Samples were shaken for three hours using an incubator shaker at a temperature of 27 °C and a speed of 180 rpm without any pH adjustment. Then, the mixture was filtered and analysed by using inductively coupled plasma-optical emission spectrometer (ICP-OES, Optima 7000 DV, PerkinElmer Inc., USA).

Pb(II) ions stock solution (1000 mg/L) was prepared by dissolving lead (II) nitrate, Pb(NO3)2, (Sigma-Aldrich, Malaysia) in the distilled water. Sodium hydroxide (NaOH), sulphuric acid (H2SO4, 95–97%), nitric acid (HNO3, 65%) and hydrochloric acid (HCl, 37%) were bought from Merck KGaA, Germany. The resulting solution’s pH was regulated by adding 0.01 M of H2SO4 or 0.01 M of NaOH. H2SO4 was preferred for pH adjustment due to its low cost, high efficiencies and high reaction rate [36]. The pH of the solutions was determined using a portable pH meter (Model: HI991001, Hanna Instruments Inc., USA).

2.4. Experimental procedure – combined treatment system

2.3. Experimental procedure – adsorption process

In this work, the electrocoagulation and adsorption processes were combined into a single system. The combined system was developed based on a bipolar configuration. In this study, combined system was performed at a fixed current density at 0.484 mA/cm2 for all experiments. The perforated Zn was used as the electrodes for the combined treatment process with electrode sizes of 7 cm × 5 cm and 0.5 cm holes were then connected to a solar PV source by a DC-DC charge controller. The distance between electrodes was maintained at 1.0 cm in all experiments. The solutions were agitated at an agitation speed of 250 rpm with a constant temperature of 27 ± 1 °C. The adsorbent was added into the combined system at various dosages (2–4 g/L). Electrocoagulation (arrangement of electrodes-occurred in the upper

2.3.1. Adsorption isotherm Batch adsorption experiments were carried out in centrifuge tubes whereby each tube contained 30 mL of Pb(II) ions solution with different initial concentrations (10–130 mg/L) and fixed adsorbent dose. The required amount of adsorbent was pre-selected based on literature studies [37–39] as well as experimental data from our preliminary work. The samples were stirred at 180 rpm via an orbital shaker with the temperature being controlled at 27 ± 1 °C and constant pH of 6, 8 or 10 until a state of equilibrium was reached (96 h). Varying of pH was designated based on literature studies [40,41] as well as experimental

Fig. 1. Experimental flowchart of the combined treatment system: (1) perforated anode and cathode electrodes; (2) plane electrode (without any electrical connection-bipolar configuration); 3) mesh; 4) activated carbon; 5) charge controller; 6) photovoltaic module; 7) regulator. 3

Chemical Engineering & Processing: Process Intensification 143 (2019) 107619

F. Hussin, et al.

Table 2 Optimum process parameters for the single electrocoagulation and adsorption systems.

Table 3 Range of process parameters chosen for the combined solar-powered electrocoagulation and adsorption system.

Parameter

Electrocoagulation

Adsorption

Parameters

Range

Initial Pb(II) concentration (mg/L) pH Adsorbent dosage (g/L) Current density (mA/cm2) Pb(II) removal efficiency (%)

10 7 – 0.484 78.5

10 6 4 – 79.3

Initial Pb(II) concentration (mg/L) pH Adsorbent dosage (g/L)

15–35 5–7 2–4

Note: These process parameters were used for further study using CCD model.

part of the reactor) while adsorption (using adsorbent-reaction happened from the bottom of the reactor) were carried out simultaneously in the same reactor (Fig. 1). After the experiments, the residual Pb(II) ions concentration was filtered using a filter with a pore size of 0.45 μm and the samples were analysed immediately using ICP-OES. Each analysis was carried out in triplicate. All the experiments were carried out at room temperature. The experimental flowchart of a combined treatment system is depicted in Fig. 1. The combined system was applied to maximise the Pb(II) ions removal efficiency, minimise power consumption by utilising solar energy and reduce the amount of adsorbent used to eliminate Pb(II) ions from the aqueous solution. The Zn concentration was evaluated after the treatment to ensure the Zn electrodes were suitable to be used for further studies without causing detrimental effects to the environment and it was confirmed that the Zn was completely removed after the treatment. According to the environmental laws in Malaysia, the permissible limit of Zn in the effluent is 1.0 mg/L [43]. Based on the ICPOES results, the average Zn content in the aqueous solution was very low, which is approximately 0.00011 mg/L. Therefore, it can be deduced that the use of Zn electrodes does not give a detrimental impact to the environment and these electrodes are suitable to remove Pb(II) ions.

factors and ranges as the reference data for further study on optimisation using CCD. The screening experiment consists of trial runs at the low- and high-level combinations of a range of research variables. The information data obtained in the screening study were used to set the range of process parameters in the Design Expert software for optimisation study. In the later stages of the experimental work, the goal shifts from screening to optimisation process. Once the important factors are obtained, the next stage of the investigation focuses on the relationship between the response and the factors using CCD model. The details information related to results of the adsorption isotherms and adsorption kinetic (for single adsorption process) can refer to the supplementary data (Fig. S1 and Table S1 for adsorption isotherm) and (Fig. S2 and Table S2 for adsorption kinetic). Table 2 shows the optimum process parameters for the single electrocoagulation and adsorption processes after three hours of treatment. According to these results, the pH (5–7), initial Pb(II) concentration (15–35 mg/L), and adsorbent dosage (2–4 g/L) were studied using the combined solar electrocoagulation and adsorption system (Table 3). The current density, the distance between electrodes and agitation speed were kept constant during the experiments using the combined system. 2.8. Response surface methodology (RSM)

2.5. Analytical methods

As mentioned earlier, RSM is a powerful statistical and mathematical technique used to model and analyse processes. As the name implies, the main aim of RSM is to optimise the response surface, which is influenced by several process parameters. RSM offers the advantage of reducing the number of trials associated with conventional experiments [45]. This technique can be used to optimise various laboratory-scale or industrial-scale processes within a reasonable amount of time. RSM is a faster, simpler, and cost-effective approach to conduct experiments compared to the conventional one-factor-at-a-time approach (OFAT). In case of OFAT approach, it is not possible to evaluate the interaction effects between the process parameters and it is very time-consuming to examine the effects of all factors [46]. The optimisation process using RSM consists of the following steps: (1) a series of experiments is carried out in order to obtain a sufficient and dependable dimension of the response of interest; (2) the experimental data are fitted to the most appropriate response surface model; (3) the optimum process parameters that maximises or minimises the response are determined; (4) the interaction effects of the process parameters are presented in the form of three-dimensional response surface plots [47]. The CCD is a two-level factorial design with factorial points (2n), axial points (2n) which correspond to the low and high levels of the factors, and center points (nc ) which correspond to the intermediate levels of the factors [48,49]. The center points are used to determine the experimental errors as well as reproduce data when the model does not provide a good fit. Second-order polynomial models (Eq. (2)) are typically used in RSM [48,50]:

The Pb(II) concentrations were examined using ICP-OES (Model: Optima 7000 DV, PerkinElmer Inc., USA). The Pb(II) removal efficiency was determined by the subsequent equation [44]:

Removal efficiency (%) =

C0

Ct C0

100

(1)

2.6. Characterisation of adsorbent A field emission scanning electron microscope (FESEM) cross-beam workstation (SU8000, Hitachi, Japan) connected with an energy dispersive X-ray spectroscopy (EDS) detector system was used to analyse the surface morphology of the adsorbent. The adsorption and desorption measurements were performed in liquid nitrogen at 77 K using a surface area and porosity analyser (Tristar II 3020, Micromeritics Instrument Corp., USA). The Brunauer-Emmett-Teller (BET) surface area was determined based on the nitrogen adsorption isotherms and the average pore diameter was calculated based on the Barrett-JoynerHalenda (BJH) method. The particle size distribution was analysed using a particle size analyser (Mastersizer 2000, Malvern Panalytical Ltd., UK). 2.7. Screening experiment Prior to the optimization process, a screening experiment was conducted to select a suitable range for each process parameter. In the screening experiments, the electrocoagulation and adsorption processes were conducted independently to determine the optimal values for the pH, initial Pb(II) concentration, current density, and adsorbent dosage. The goal of the screening study was to identify the correct variables/

k

y=

0

+

k i xi

i=1

2 ii x i

+ i=1

+

ii x i x j i=1 j>1

+

(2)

Here, y represents the predicted response (Pb(II) removal efficiency), Xi 4

Chemical Engineering & Processing: Process Intensification 143 (2019) 107619

F. Hussin, et al.

Fig. 2. Flow chart of the process optimisation process using RSM.

and Xj represent the independent variables, i , j , and ij represent the linear coefficient, quadratic coefficient, and cross-product coefficient, respectively, and represents the residual term. In the CCD, each factor is assessed at five dissimilar points: −1, +1, 0, -α and +α. The value of α is dependent on the number of factors used in the factorial design. The value of α can be determined using the following equation [51,52]: 1

interaction (2FI), quadratic, and cubic polynomial models in order to attain the regression equations [55]. The F-values, adjusted R2 and predicted R2 are compared for all models and the highest-order polynomial with significant terms is typically chosen as the regression model for further analysis. It showed that the experimental designs fit well with the quadratic polynomial model with high adjusted and predicted R2. Based on these findings, the adjusted and predicted R2 values were found to be 0.9984 and 0.9949, this appearing in a very good agreement between the predicted and actual data. Then, analysis of variance (ANOVA) was carried out to validate the significance and adequacy of the regression models [56]. There are 20 experimental runs for CCD, with six replicates at the center points to eliminate errors and curvature. The CCD consists of axial points (2n), the number of independent variables (2n) and center points (nc). Thus, the CCD in this study consists of (2n = 2(3) =6), (2n = 23 = 8) and nc = 6, resulting in 20 experiments (Eq. 4) [51]. The experimental data of Pb(II) removal efficiencies obtained from the CCD models are also shown in Table 5.

(3)

= [2n] 4

where n is the number of factors such as pH, initial Pb (II) concentration, and adsorbent dosage. Since, the number of factors used in the experimental design is 3, α is 2. Hence, the coded levels for the CCD are −1, +1, 0, −2 and +2. Design-Expert® software version 11 (Stat-Ease, Inc. USA) was used for this purpose as well as the presentation of data. The steps involved to optimize the process parameters using RSM are shown in Fig. 2. Each experimental design was evaluated independently in order to examine the effects of pH, initial Pb(II) concentration, and adsorbent dosage on the Pb(II) removal efficiency. There are six center points for the CCD model. Generally, center points that are located in the center of the cube are used to estimate the experimental error and the duplicability of the data [51]. This is also used to measure the process stability/variability, as well as to check for curvature of the response surface, where it contributes to the estimation of coefficients of the quadratic terms [53]. Six center points were chosen due to the variance of a predicted value. When fitting a response surface, the response function in this design region is to be estimated, as the study aims to achieve maximum removal efficiency at optimum operating conditions. Most importantly, it needs the prediction to be reliable throughout the region, and mainly near the center and the optimum will be in the central region. By selecting six center points, the variance in the middle is almost the same as the variance at the edge. The more replicates there are, the more precise the error estimation will be. If only two or three center points would be selected, this will result in less precision in the middle, than it would have at the edge. It is important to balance the precision at the edge of the design relative to the middle which can be done using six center points [54]. Three process parameters were coded as X1 (pH), X2 (initial Pb (II) concentration (mg/L)), and X3 (adsorbent dosage (g/L)), respectively. The coded levels of the factors for CCD models are shown in Table 4, respectively. The experimental data were fitted using linear, two-factor

Desirability function is commonly used to determine the optimum process parameters to maximise the removal of Pb(II) ions [57,58]. Desirability is an objective function (D) that desirable ranges are from zero to one at the goal [59]. The desirability function method is one of the most popular for multiple response problems. The desirability function was introduced by Harrington [57]. If the product characteristic is in an unacceptable range, the desirability value is zero (0) and if the product characteristic is at the optimum value, the desirability Table 5 Experimental design for CCD. Run

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Table 4 Selected factors and their corresponding coded levels used in CCD. Factors

pH Initial Pb(II) concentration Adsorbent dosage

Coded variable

X1 X2 X3

Unit

mg/L g/L

Coded levels −2

−1

0

+1

+2

5 15 2

5.5 20 2.5

6 25 3

6.5 30 3.5

7 35 4

(4)

N = 2n + 2n + nc

5

Coded variables

Pb(II) removal efficiency (%)

X1

X2 (mg/L)

X3 (g/L)

Experimental

+1 0 −1 0 0 0 +2 0 0 0 −2 +2 0 −2 −2 0 +2 0 −2 +2

0 0 0 +1 0 0 −2 0 0 −1 +2 −2 0 −2 −2 0 +2 0 +2 +2

0 +1 0 0 0 0 +2 0 0 0 +2 −2 −1 −2 +2 0 +2 0 −2 −2

87.10 85.80 89.10 82.10 92.10 92.40 86.30 91.30 91.50 98.70 56.60 94.20 91.30 81.60 85.30 91.70 32.50 91.40 68.10 57.30

Chemical Engineering & Processing: Process Intensification 143 (2019) 107619

F. Hussin, et al.

value is one (1). The desirability function is generally within a range of 0-1. The optimum process parameters with the desirability of 1 will provide a more accurate response. Following this, Derringer and Suich [60] developed another version to determine the optimum process conditions, in which the response (Pb(II) removal efficiency) can be maximised, minimised, or set at a target value. He proposed three response types such as nominal-the-better (NTB), smaller-the-better (STB) and larger-the-better (LTB) types of response variables. Three response types as follows (Eqs. 5–7):

silica are the minor constituents in the adsorbent. Their features can be observed at 0.5–2.0 keV, as shown in Fig. 4. The nitrogen adsorption-desorption isotherms of the OPSAC are shown in Fig. 5. Based on the International Union of Pure and Applied Chemistry (IUPAC) classification, the OPSAC is corresponding to Type I and IV isotherms [61]. It can be observed that there is a steep adsorption of nitrogen which was corresponding to the Type I adsorption isotherm at the low relative pressure region (< 0.05). This is due to the presence of micropores. A long and slender hysteresis loop was observed within a relative pressure of 0.3-0.6 which was corresponded to the Type IV isotherm. This hysteresis loop was due to the presence of micropores and mesopores. Based on the IUPAC classification, the hysteresis loop of OPSAC is classified as the Type H4 [62]. Type H4 hysteresis loops are characterised by parallel adsorption and desorption isotherms and the branches are almost horizontal. The isotherms are always related to narrow slit-like pores, which include pores in the micropore region. The textural properties of the OPSAC are presented in Table 6. The specific BET surface area of the OPSAC was found to be 1092.90 m2/g. The high specific surface area of the adsorbent provides more adsorption sites, which will enhance the removal of Pb (II) ions [63]. The inset of Fig. 5 shows that the pore width distribution was 12 Å (1.2 nm), which proves that the OPSAC mainly consists of micropores. The wide range of pore widths is sufficient to provide an effective active surface for adsorption, which facilitates in removing Pb (II) ions from aqueous solutions. Based on the t-plot analysis, the OPSAC was primarily composed of micropores with an area of 889.91 m²/g, as shown in Table 6. The micropore area contributes ˜77% of the total pore area. The Pb (II) ions have large ionic radius (0.132 nm) and high atomic weight (207.2 g/mol) and therefore, the attraction force increases between Pb (II) ions and binding sites of the adsorbent [64,65].

a) Nominal-the-better (NTB)

0 y T y T

d=

LSL LSL USL USL

s

t

LSL y T T y USL y < LSL or y > USL

,

(5)

0

where the exponents s and t are the shape constants of the desirability function, and LSL and USL are respectively the lower and upper specification limit for the NTB type response variable with a target value T. In general, the shape constants are chosen in the range from 0.01 to 10. a) smaller-the-better (STB)

d=

1.0 y a 0

USL USL

t

a

,

y USL y > USL

(6)

where a is a smallest possible value for response Y. a) Larger-the-better (LTB)

3.2. Development of the regression models

0 d=

y LSL USL LSL 1.0

t

,

LSL y USL y < LSL y > USL

The correlations between the process parameters on the Pb(II) removal efficiency were evaluated by using the CCD. Runs 5, 6, 8, 9, 16, and 18 were selected to evaluate the experimental error for CCD. Based on the results shown in Table 5, the maximum Pb(II) removal efficiency was 98.7% for the experimental designs. Regression analysis was conducted to fit the response function to the Pb(II) removal efficiency. The regression model for CCD is given by Eq. 8, respectively, where Y represents the Pb(II) removal efficiency, and X1, X2 and X3 represent the coded variables for pH, initial Pb(II) concentration, and adsorbent dosage, respectively. The response variable (Pb(II) removal efficiency) was fitted by a second-order polynomial in order to evaluate the experimental data. The regression model used to predict the Pb(II) removal efficiency in the coded values for the CCD is shown below:

(7)

3. Results and discussion 3.1. Properties of the oil palm shell activated carbon Fig. 3 shows the FESEM image of the OPSAC. It can be observed from the surface morphological features that the OPSAC is a highly porous material, which is characterised by spongy and irregular-shaped particles. The pores range from micropores (< 2 nm) to macropores (> 50 nm). It is evident that carbon is the major element whereas oxygen and

Y = 91.26

2.62X1

4.01X2 X3

16.61X2

10.69X12

5.09X3 8.89X32

6.06X1 X2

3.11X1 X3 (8)

Multiple regression analysis was used to estimate the coefficients to enhance a better understanding of the interactions between the dependent and independent variables [51,66]. The fitness of the regression models was measured based on their correlation coefficient (R) and coefficient of determination (R2). The t-test was used to measure the precision in the estimation of regression coefficients while the p-value was used to investigate the impact of the regression coefficients, as shown in Table S4 in supplementary data. The t-values were obtained by dividing each factor with its standard error. According to Rengadurai et al. [67], a larger t-value and smaller p-value indicate that the regression coefficient is significant. In this work, it was found that the linear factors ( X1, X2 , and X3), interaction factors ( X1 X2 , X1 X3, and X2 X3 ), and quadratic factors ( X12 and X32 ) were the significant model terms. However, X22 was not statistically significant, where the p-values

Fig. 3. FESEM image of the OPSAC. 6

Chemical Engineering & Processing: Process Intensification 143 (2019) 107619

F. Hussin, et al.

Fig. 4. Energy-dispersive X-ray spectrum of the OPSAC.

Fig. 5. Nitrogen adsorption-desorption isotherms of the OPSAC.

is considered significant [51]. Based on the results shown in Table 7, X1 (pH), X2 (initial Pb(II) concentration), X3 (adsorbent dosage), X1X2, X1X3 , and X2X3 (interaction terms), as well as X12 and X32 (quadratic terms) were significant model terms. For the CCD model, the effect of the independent variables on the Pb(II) removal efficiency was ranked based on the F-value from highest to lowest: X2 : initial Pb(II) concentration > X3: adsorbent dosage > X1: pH. Based on the ANOVA results, it can be assumed that the initial Pb(II) concentration and adsorbent dosage were the most significant independent variables influencing the Pb(II) removal efficiency. The reliability of the regression models was assessed based on the R2 value. The adjusted R2 is a statistical measurement of how close the experimental data fall within the fitted regression line, while the predicted R2 is a statistical measure of how well the model predicts the response (Pb (II) removal efficiency) [69]. The R2 value should be within a range of 0–1. In general, a higher R2 value is desirable because this indicates that the variability of the response is described adequately by the regression model. In other words, higher R2 value indicates that there is a good agreement between the predicted and experimental data. The R2 value was found to be 0.9991. The models have high R2 value (0.9991) due to the significant F-value, insignificant lackof-fit P-value and low standard deviation and coefficient of variance. The insignificant lack-of-fit shows that the model is valid for the present work. These results indicate high precision in predicting the Pb(II)

Table 6 Textural properties of the OPSAC. Parameter

Value

BET surface area (m2/g) Micropore area (m²/g) Pore volume (cm3/g) Average pore diameter (nm) Particle size (μm) pHPZC

1092.90 889.91 0.354 1.815 765.65 8.0

were found to be more than 0.05. This indicates that this factor does not have a significant effect on the Pb (II) removal efficiency. 3.3. Statistical analysis Statistical analysis was conducted to obtain well-fitted regression models for the Pb(II) removal efficiency. The significance of the regression models was determined from ANOVA. The F-values of the regression models are 1294.25, respectively, as shown in Table 7. The pvalues are less than 0.0001, which indicate that the regression model terms are significant [68]. Based on the p-values, there is only 0.01% chance that the large F-values of the regression model is due to extraneous noise. The significance of each model term was assessed based on the p-value. If a model term has a p-value less than 0.05, the model term 7

Chemical Engineering & Processing: Process Intensification 143 (2019) 107619

F. Hussin, et al.

Table 7 ANOVA results for the reduced quadratic regression model (CCD). Source

Sum of squares

Degree(s) of freedom

Mean Square

F-value

p-value

Remarks

Model X1 X2 X3 X1 X2 X1 X3 X2 X3

9 1 1 1 1 1 1 1

566.25 58.50 2345.58 220.07 294.03 77.50 128.80 21.32

1294.25 133.72 5361.14 502.99 672.05 177.14 294.39 48.74

< 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001

Significant

X12

5096.28 58.50 2345.58 220.07 294.03 77.50 128.80 21.32

X32 Lack of fit Std. dev.a Mean C.V. %c Cor total

14.75

1

14.75

33.70

0.0002

3.44 0.6614 82.32 0.80 5100.65

5

0.6884 R2 Adj. R2 b Pred. R2 d Adeq. Prec.

3.69 0.9991 0.9984 0.9949 142.2065

0.0892

a b c d e

19

e

Not significant

Standard deviation. Adjusted R2. Coefficient of variation. Predicted R2. Adequate precision.

removal efficiency. This indicates that there is a good agreement between the values predicted by the regression models and the experimental data. It was evident that the experimental values fall within the proximity of the predicted values, indicating that the regression models can adequately describe the interaction between the variables and response is strongly influenced by the independent variables [51]. The adequate precision (Eqs. 9 and 10) was compared with the range of the actual values at the design points to the average prediction error. The adequate precision value was found to be 142.21. The value was greater than 4, which indicates that the regression model is adequate to navigate through the design space and the model is capable of predicting the Pb(II) removal efficiency. Adequate precision can be calculated based on the information from ANOVA table or the adequate precision can be determined by Eqs. 9 and 10 [70]. In this study, the adequate precision value was found from AVONA table in the CCD model.

max (Yˆ )

min (Yˆ ) >4 V¯Yˆ

pˆ2 V¯Yˆ = n

3.4. Effects of process parameters on the Pb(II) removal efficiency RSM was used to assess the effects of three process parameters (pH, initial Pb(II) concentration, and adsorbent dosage) on the Pb(II) removal efficiency and the resulting three-dimensional response surface plots were plotted. In the response surface plots, pH, initial Pb(II) concentration, and adsorbent dosage were coded as A, B, and C, respectively. The three-dimensional response surface plots will provide insight into the interactions between the three independent variables. Fig. 6(A) shows the interaction effects of initial Pb(II) concentration and adsorbent dosage on the Pb(II) removal efficiency for the CCD, at a constant pH of 6. It can be observed that the Pb(II) removal efficiency increased with a decreased in the initial Pb(II) concentration. In addition, the Pb(II) removal efficiency increased with an increased in the adsorbent dosage from 2.0 to 3.5 g/L. This agrees well with the hypothesis whereby the active sites of the activated carbon can adsorb more Pb(II) ions from the aqueous solution. This is because the adsorbent dosage is increasing with the adsorption capacities [38]. The enhancement of Pb(II) removal efficiency at a higher adsorbent dosage is due to the higher number of available active adsorption binding sites and the availability of a larger surface area for adsorption of Pb(II) [38]. However, the Pb(II) removal efficiency slightly decreased when the adsorbent dosage increased beyond 3.5 g/L. This indicates that the adsorption of Pb(II) diminished when the adsorbent dosage exceeded this value. This is due to the overlapping of active sites at a higher adsorbent mass, which reduced the effective surface area needed for adsorption [72]. Both the adsorbent dosage and initial Pb(II) concentration influence the Pb(II) removal efficiency and the interaction effects of these independent variables are obvious based on their Fvalues. For the CCD model, the F-values were 5361.14, 502.99, and 133.72 for the initial Pb(II) concentration, adsorbent dosage, and pH, respectively. Therefore, a small amount of adsorbent was selected to reduce the cost of the adsorbent material and treatment process. The effects of the initial Pb(II) concentration and pH on the Pb(II) removal efficiency at a constant adsorbent dosage of 3 g/L obtained from the CCD are shown in Fig. 6(B). As expected, the Pb(II) removal efficiency was greater at lower initial Pb(II) concentrations. This is likely because the metal ions quickly adhere to the adsorption sites, which resulted in higher adsorption efficiency [73]. The increased in the initial Pb(II) concentration and pH enhance the Pb(II) removal efficiency. However, the elimination rate of Pb(II) ions slightly decreased when its initial Pb(II) concentration was beyond 30 mg/L due to the slower rate of coagulation. In addition, the settling process triggers a

(9)

(10)

Yˆ = the predictions at the run settings p ˆ 2 = the residual mean square from ANOVA table p = the number of terms in the model n = the number of runs in the design The coefficients of variation (C.V) was 0.80. The value is less than 2, demonstrating that the experimental value has high precision and good reliability. The lack of fit F-value is determined from the total sum of squared deviations between the mean response at each factor level and the corresponding fitted value. It can be seen from Table 7 that the lackof-fit F-value was 3.69 and its p-value was 0.0892, indicating 8.92% chance that the lack-of-fit F-value was due to noise and the lack-of-fit of the regression model did not correlate with the pure error. In the CCD model, if the p-value of lack-of-fit is p > 0.05 (insignificant), it suggests that the proposed model fit the experimental data and the independent variables or parameters have considerable effects on the response. If the model does not fit well with the data, the lack-of-fit will be significant (p < 0.05 significant). In this model, we want the p-value of lack-of-fit to be insignificant to produce a model predicted data that fits the actual experimental response data [71].

8

Chemical Engineering & Processing: Process Intensification 143 (2019) 107619

F. Hussin, et al.

the restricted mass transfer of the liquid solution to the outer surface of the adsorbent as well as the deliberate mass transfer within the activated carbon particles [37,74]. The combined effects of the adsorbent dosage (2–4 g/L) and pH (5–7) on the Pb(II) removal efficiency at a constant initial Pb(II) concentration of 25 mg/L are shown in Fig. 6(C). The Pb(II) removal efficiency was also dependent on pH because the Pb(II) precipitated in solutions with a pH value of more than 7. The Pb(II) removal efficiency increased as the pH increased from 5.0 to 6.5. It can be seen that the maximum Pb(II) removal efficiency occurred at pH 6 and the Pb(II) removal efficiency decreased gradually beyond pH 6.5. This is due to competition between the Pb(II) and H+ ions for active sites on the adsorbent surface [38,75]. The formation of Pb(OH)2 and Pb(OH)+ took place at higher pH values [76]. In addition, precipitation occurred at pH > 7.0 due to the association of the Pb(II) ions with the OH– ions in the basic solution, forming lead hydroxides. This decreased the rate of adsorption, which boosted the Pb(II) removal efficiency [74,77]. Trace amounts of other Pb species such as Pb(NO3)2 (aq), Pb4(OH)+, and Pb2(OH)3+ are also present in the solution. Since the Pb(II) ions have a tendency to precipitate at pH > 7, pH 6 is considered to be the most appropriate pH. 3.5. Model validation In order to validate the regression models, a series of experiments was conducted using the optimum process parameters obtained from the response surface methodology with the desirability of 1. The optimum process parameters with the highest desirability of 1 generated more accurate results. The final regression models were validated to determine the adequacy of the predicted response values by selecting the optimum process parameters that maximised the efficiency of Pb(II) ions removal. Experiments were then conducted based on the optimum process parameters and the Pb(II) removal efficiencies were compared with the values predicted by the regression models by computing the residual standard error [35] as given in Eq.11. Based on the results, the optimum initial Pb(II) concentration, pH, and adsorbent dosage were 15.00 mg/L, 6.01, and 2.50 g/L, respectively. The difference between the predicted and experimental results was less than 0.1% for experimental model, which indicates that the predicted and experimental Pb (II) removal efficiencies are similar.

Standard error (%) =

Experimental value Predicted value x 100 Predicted value

(11)

3.6. Comparison between single and combined treatment systems Table 8 shows the comparison between single and combined methods. Based on the results, it can be seen that the highest Pb(II) removal efficiency can be achieved at the following optimum process parameters: (1) pH 6.01, (2) initial Pb(II) concentration 15.00 mg/L, and (3) adsorbent dosage 2.50 g/L. Moreover, the kinetic analysis of the pseudo first-order model fit the experimental data. The kinetic rate constant was calculated using the first-order model [78]. This showed that Pb(II) more quickly adsorbed and removed from aqueous solution in the combined treatment process than the single processes. The results indicate that the combined solar electrocoagulation and adsorption system gives superior performance compared to the single method where the processes are carried out separately.

Fig. 6. Three-dimensional response surface plot of the interaction effects of (A) adsorbent dosage and initial Pb(II) concentration, (B) initial Pb(II) concentration and pH, (C) adsorbent dosage and pH on the Pb(II) removal efficiency from the CCD.

3.7. Comparison of current work with previously reported studies

reduction in Pb(II) removal efficiency. This indicates that more time is needed to attain equilibrium if higher amounts of Pb(II) are present in the solution. This is likely because the adsorption process is quicker when the number of active sites available is superior to the adsorption rate of metal ions [74]. In addition, equilibrium was achieved due to

The highest Pb(II) ions removal efficiency using the combined treatment system was compared with other systems reported in the literature. To the best of our knowledge, only two literature studies used the combined treatment system for the removal of Pb(II) ions. For instance, Orescanin et al. [79] proposed a new electroplating 9

Chemical Engineering & Processing: Process Intensification 143 (2019) 107619

F. Hussin, et al.

Table 8 Comparison of the Pb (II) removal efficiencies between the single solar electrocoagulation and adsorption treatment systems and combined treatment system. Parameters

*

Adsorption

*

pH Adsorbent dosage (g/L) Initial Pb (II) concentration (mg/L) Kinetic rate constant k (min−1) (Pseudo 1st order kinetic) Pb (II) removal efficiency (%)

7 – 10 3.905 78.5

6 4 10 3.945 79.3

6.01 2.50 15 4.971 99.88

Electrocoagulation

Combined system

* Note: Current density was kept constant during the experiments using the single and combined methods.

wastewater treatment system by combining electrocoagulation using iron and aluminium electrodes with ozonation methods which resulted in 96.81% Pb(II) ions removal efficiency. Meanwhile, Ferniza-García et al. [80] studied the removal efficiency of heavy metals from the simulated mining water through a combined electrocoagulation process using aluminum electrodes with phytoremediation (Typha latifolia L.) powered by conventional power supply. The results showed that the removal efficiency during the 180 min of the combined treatment achieved a 99.4% removal of Pb(II) ions at a pH of 5.27. The removal efficiency of Pb(II) ions (99.88%) obtained in this work are better than those obtained using the other combined treatment processes from the literature studies due to the usage of clean energy, low current density and low amount of adsorbent dosage. The experiments demonstrated that the combined system is sustainable, economical and effective in treating waters that contain heavy metals. Moreover, application of combined system allows to fulfil the standard regulations for Pb(II) ion removal, defined by the Department of Environment (DOE), Malaysia under the Malaysia Environmental Quality (Sewage and Industrial Effluents) Regulations 2009 for discharge of industrial effluent for mixed effluent of standard B (maximum permissible concentration of Pb(II) is 0.5 mg/L) [81].

recovery of Pb(II) ions can be ranked in the order of HCl > HNO3 > distilled water, from the highest to lowest recovery rate. As observed, the maximum desorption of Pb(II) ions was 98% using 0.1 M of HCl, while HNO3 and distilled water exhibited approximately 78% and 50% of the adsorbed lead ions, respectively. Similarly, Shekinah et al. [86] proved that the maximum desorption of Pb(II) was achieved using 0.025 M to 0.175 M of HCl as the desorption agent. It can be concluded that, the palm shell activated carbon is possible to be regenerated and recycled up to five times. 5. Conclusion This study is believed to be the first report on the removal of Pb(II) from aqueous solutions using a combined solar electrocoagulation and adsorption treatment system. The combined process was performed at low current condition to produce a cost-effective treatment system. The combined system used in this study was effective to reduce the Pb(II) concentration to below the allowable limit stipulated by the Department of Environment in Malaysia, where the maximum permissible limits for Pb(II) in industrial effluents is 0.5 mg/L for downstream discharge. Based on the results, it can be seen that the highest Pb(II) removal efficiency (99.88%) can be achieved at the following optimum process parameters: (1) pH of 6.01, (2) initial Pb(II) concentration of 15.00 mg/L, and (3) adsorbent dosage of 2.50 g/L. The combined system offers an attractive alternative over the single electrocoagulation and adsorption treatment systems due to high Pb(II) removal efficiency. In addition, the combined treatment system is cost-effective due to the low adsorbent dosage required for the process. The experiments demonstrated that the combined process can be regarded as a robust and effective solution in treating waters and wastewaters that contain heavy metal ions due to its efficient, economical, and sustainable system (solar/renewable energy system). Application of the combined treatment system using solar PV in a continuous mode has a great potential for large-scale processing, where continuous operation is preferred. Further studies should be carried out to enhance the performance and upscale the process from laboratory to pilot scale and assess its feasibility for industrial applications.

3.8. Consideration on energy consumption Energy consumption is one of the most important economical parameters in the combined treatment processes. Energy consumption can be calculated using the following method, Ghosh et al. [44]:

E(kWh/m3) =

UxIxt V

(12) 3

Where E is the energy consumption (kWh/m ), U is the voltage (V), I is the current (A), t is the reaction time (h) and V is the volume of Pb(II) solution (m3). The energy supplied by solar PV to remove Pb(II) ions by combined solar electrocoagulation and adsorption system was 0.045 kW h/m3. The proposed combined treatment method powered by PV is sustainable and cost effective with high removal efficiency. This is in agreement with the work of Palahouane at al. [82] who stated that solar electrocoagulation process is suitable for wastewater treatment because it provides high removal efficiency with low electrical energy consumption of 3.43 kW h/m3. Nawarkar and Salkar [83] also reported that energy consumption using solar electrocoagulation at the optimum operating condition of 2.27 kW h/m3 and it indicates that solar electrocoagulation system produce cost -effective treatment method. Garcia-Garcia et al. [84] compared two treatment systems between solar electrocoagulation and solar electro-oxidation. They stated that energy consumption for solar electrocoagulation (0.031 kW h/m3) is lower compared to and solar electro-oxidation (0.112 kW h/m3).

Declaration of Competing Interest There are no conflicts to declare. Acknowledgements We are grateful for financial support provided by High Impact Research Grant, University of Malaya (UM.C/HIR/MOHE/ENG/43) and Internal Research Grant, Sunway University (INT-2019-SST-CCDCU-01)

4. Desorption experiment

Appendix A. Supplementary data

Numerous categories of desorbing agents, for instance, HCl, HNO3 and distilled water were applied for desorption of adsorbed Pb(II) ions from aqueous solution [85]. Based on the experimental results, the

Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.cep.2019.107619. 10

F. Hussin, et al.

Chemical Engineering & Processing: Process Intensification 143 (2019) 107619

References

E.G.P. da Silva, L.A. Portugal, P.S. dos Reis, A.S. Souza, W.N.L. dos Santos, BoxBehnken design: an alternative for the optimization of analytical methods, Anal. Chim. Acta (2007) 179–186. O. Ocholi, M. Menkiti, M. Auta, I. Ezemagu, Optimization of the operating parameters for the extractive synthesis of biolubricant from sesame seed oil via response surface methodology, Egypt. J. Pet. 27 (2018) 265–275. I.A.W. Tan, B.H. Hameed, A.L. Ahmad, Optimization of preparation conditions for activated carbons from coconut husk using response surface methodology, Chem. Eng. J. 137 (2008) 462–470. G.A. Lujan-Moreno, P.R. Howard, O.G. Rojas, D.C. Montgomery, Design of experiments and response surface methodology to tune machine learning hyperparameters, with a random forest case-study, Expert Syst. Appl. 109 (2018) 195–205. Y.L. Yu, Y.N. Li, Y. Zhang, R.N. Sun, J.S. Tu, Y. Shen, Optimization and characterization of deoxypodophyllotoxin loaded mPEG-PDLLA micelles by central composite design with response surface methodology, Chin. J. Nat. Med. 16 (2018) 471–480, https://doi.org/10.1016/S1875-5364(18)30081-5. M.D. Turan, Statistical Approach to Mineral Engineering and Optimization, Contributions to Mineralization, A.I Al-Juboury, (2018), https://doi.org/10.5772/ intechopen.71607 https://www.intechopen.com/books/contributions-tomineralization/statistical-approach-to-mineral-engineering-and-optimization. C.L. Ngan, M. Basri, F.F. Lye, H.R. Fard Masoumi, M. Tripathy, R. Abedi Karjiban, E. Abdul-Malek, Comparison of Box–Behnken and central composite designs in optimization of fullerene loaded palm-based nano-emulsions for cosmeceutical application, Ind. Crop Prod. 59 (2014) 309–317. H. Chen, Biotechnology of Lignocellulose: Theory and Practice, Springer, Netherlands, 2014. J. Acharya, J.N. Sahu, C.R. Mohanty, B.C. Meikap, Removal of lead(II) from wastewater by activated carbon developed from Tamarind wood by zinc chloride activation, Chem. Eng. J. 149 (2009) 249–262. D.S.S. Rama Raju, G.A. Ravi Kiran, V.N. Rao, Comparison studies on biosorption of lead (II) from an aqueous solution using anacardium occidentale and carica papaya leaves powder, Int. J. Emerging. Trends. Eng. Dev. 1 (2013) 273–283. T. Van Thuan, B.T.P. Quynh, T.D. Nguyen, V.T.T. Ho, L.G. Bach, Response surface methodology approach for optimization of Cu2+, Ni2+ and Pb2+ adsorption using KOH-activated carbon from banana peel, Surf. Interface 6 (2017) 209–217. G. Wang, S. Zhang, P. Yao, Y. Chen, X. Xu, T. Li, G. Gong, Removal of Pb(II) from aqueous solutions by Phytolacca americana L. biomass as a low cost biosorbent, Arab. J. Chem. 11 (2018) 99–110. A.M. Soliman, H.M. Elwy, T. Thiemann, Y. Majedi, F.T. Labata, N.A.F. AlRawashdeh, Removal of Pb(II) ions from aqueous solutions by sulphuric acidtreated palm tree leaves, J. Taiwan. Inst. Chem. Eng. 58 (2016) 264–273. M. Szlachta, N. Chubar, The application of Fe–Mn hydrous oxides based adsorbent for removing selenium species from water, Chem. Eng. J. 217 (2013) 159–168. IWK, Effluent standards, Malaysia, (2016) (Accessed January 2018, https://www. iwk.com.my/do-you-know/effluent-standards. D. Ghosh, C.R. Medhi, M.K. Purkait, Treatment of fluoride containing drinking water by electrocoagulation using monopolar and bipolar electrode connections, Chemosphere 73 (2008) 1393–1400. R.H. Myers, D.C. Montgomery, C.M. Anderson-Cook, Response Surface Methodology: Process and Product Optimization Using Designed Experiments, 4th edn., Wiley, United States, 2002. M.A. Alim, J.H. Lee, C.C. Akoh, M.S. Choi, M.S. Jeon, J.A. Shin, K.T. Lee, Enzymatic transesterification of fractionated rice bran oil with conjugated linoleic acid: optimization by response surface methodology, LWT-Food. Sci. Technol. 41 (2008) 764–770. V. Gunaraj, N. Murugan, Application of response surface methodologies for predicting weld base quality in submerged arc welding of pipes, J. Mater. Process. Technol. 8 (1999) 266–275. P. Asaithambi, A.R.A. Aziz, W.M.A.W. Daud, Integrated ozone—electrocoagulation process for the removal of pollutant from industrial effluent: optimization through response surface methodology, Chem. Eng. Process. 105 (2016) 92–102. A. Asfaram, M. Ghaedi, S. Hajati, A. Goudarzi, A.A. Bazrafshan, Simultaneous ultrasound-assisted ternary adsorption of dyes onto copper-doped zinc sulfide nanoparticles loaded on activated carbon: optimization by response surface methodology, Spectrochim. Acta A. Mol. Biomol. Spectrosc. 145 (2015) 203–212. A.I. Khuri, S. Mukhopadhyay, Response surface methodology, Wiley Interdiscip. Rev. Comput. Stat. 2 (2010) 128–149. A. Arami-Niya, W.M.A. Wan Daud, S. Mjalli, F. Abnisa, M.S. Shafeeyan, Production of microporous palm shell based activated carbon for methane adsorption: modeling and optimization using response surface methodology, Chem. Eng. Res. Des. 90 (2012) 776–784. G.E.P. Box, K.B. Wilson, On the experimental attainment of optimum conditions, J. R. Stat. Soc. Ser. B 13 (1951) 1–45. M. Dutka, M. Ditaranto, T. Løvås, Application of a central composite design for the study of NOx emission performance of a low NOx burner, Energies 8 (2015) 3606–3627, https://doi.org/10.3390/en805360. 503 STAT, Lesson 11: Response Surface Methods and Designs, 11.2.1 Central Composite Designs (Accessed June 2019) from, https://newonlinecourses.science. psu.edu/stat503/node/57/. S.P. Chai, K.Y. Lee, S. Ichikawa, A.R. Mohamed, Synthesis of carbon nanotubes by methane decomposition over Co–Mo/Al2O3: process study and optimization using response surface methodology, Appl. Catal. A Gen. 396 (2011) 52–58. A.R. Esfahani, A.F. Firouzi, G. Sayyad, A. Kiasat, L. Alidokht, A.R. Khataee, Pb(II) removal from aqueous solution by polyacrylic acid stabilized zero-valent iron nanoparticles: process optimization using response surface methodology, Res. Chem. Intermediat. 40 (2014) 431–445.

[1] UNESCO, UN World Water Development Report, Wastewater: The Untapped Resource, World Water Assessment Programme (WWAP), UNESCO, Paris, France, 2017, p. 2017 (Accessed June 2018, http://www.unesco.org/new/en/unesco/ events/prizes-and-celebrations/celebrations/international-days/world-water-day2017/. [2] H. Maarof, W.A.W. Daud, M.K. Aroua, Recent trends in removal and recovery of heavy metals from wastewater by electrochemical technologies, Int. Rev. Chem. Eng. 33 (2017) 359. [3] M. Siahkamari, A. Jamali, A. Sabzevari, A. Shakeri, Removal of lead(II) ions from aqueous solutions using biocompatible polymeric nano-adsorbents: a comparative study, Carbohydr. Polym. 157 (2017) 1180–1189. [4] ATSDR, The Priority List of Hazardous Substances Atlanta, U.S. Department of Health and Human Services, Georgia, 2017, p. 2017 (Accessed March 2018), https://www.atsdr.cdc.gov/spl/. [5] D.C. Bellinger, Lead, Pediatrics 113 (2004) 1016–1022. [6] S. Vasudevan, J. Lakshmi, G. Sozhan, Optimization of electrocoagulation process for the simultaneous removal of mercury, lead, and nickel from contaminated water, Environ. Sci. Pollut. Res. 19 (2012) 2734–2744. [7] H. Needleman, Lead poisoning, Annu. Rev. Med. 55 (2004) 209–222. [8] S. Vasudevan, J. Lakshmi, Effects of alternating and direct current in electrocoagulation process on the removal of cadmium from water – a novel approach, Sep. Purif. Technol. 80 (2011) 643–651. [9] S. Vasudevan, M.A. Oturan, Electrochemistry: as cause and cure in water pollution—an overview, Environ. Chem. Lett. 12 (2014) 97–108. [10] R. Kamaraj, P. Ganesan, S. Vasudevan, Removal of lead from aqueous solutions by electrocoagulation: isotherm, kinetics and thermodynamic studies, Int. J. Environ. Sci. Technol. 12 (2015) 683–692. [11] F. Hussin, F. Abnisa, G. Issabayeva, M.K. Aroua, Removal of lead by solar-photovoltaic electrocoagulation using novel perforated zinc electrode, J. Clean. Prod. 147 (2017) 206–216. [12] G. Sharma, J. Choi, H.K. Shon, S. Phuntsho, Solar-powered electrocoagulation system for water and wastewater treatment, Desalin. Water Treat. 32 (2011) 381–388. [13] C. Tang, Y. Shu, R. Zhang, X. Li, J. Song, B. Li, Y. Zhang, D. Ou, Comparison of the removal and adsorption mechanisms of cadmium and lead from aqueous solution by activated carbons prepared from Typha angustifolia and Salix matsudana, RSC Adv. 7 (2017) 16092–16103. [14] V. Khandegar, A.K. Saroha, Electrocoagulation for the treatment of textile industry effluent – a review, J. Environ. Manage. 128 (2013) 949–963. [15] O. Sahu, B. Mazumdar, P.K. Chaudhari, Treatment of wastewater by electrocoagulation: a review, Environ. Sci. Pollut. Res. 21 (2014) 2397–2413. [16] K. Dermentzis, D. Stergiopoulos, P. Giannakoudakis, A. Moumtzakis, Removal of copper and COD from electroplating effluents by photovoltaic electrocoagulation/ electrooxidation process, Water. Utility. J. 14 (2016) 55–62. [17] J. Castañeda-Díaz, T. Pavón-Silva, E. Gutiérrez-Segura, A. Colín-Cruz, Electrocoagulation-adsorption to remove anionic and cationic dyes from aqueous solution by pv-energy, J. Chem. 2017 (2017) 14. [18] I. Chakchouk, N. Elloumi, C. Belaid, S. Mseddi, L. Chaari, M. Kallel, A combined electrocoagulation-electrooxidation treatment for dairy wastewater, Braz. J. Chem. Eng. 34 (2017) 109–117. [19] Y. Ozay, E.K. Ünşar, Z. Işık, F. Yılmaz, N. Dizge, N.A. Perendeci, M.A. Mazmanci, M. Yalvac, Optimization of electrocoagulation process and combination of anaerobic digestion for the treatment of pistachio processing wastewater, J. Clean. Prod. 196 (2018) 42–50. [20] L. Bilińska, K. Blus, M. Gmurek, S. Ledakowicz, Coupling of electrocoagulation and ozone treatment for textile wastewater reuse, Chem. Eng. J. 358 (2019) 992–1001. [21] E. GilPavas, I. Dobrosz-Gómez, M.A. Gómez-García, Optimization and toxicity assessment of a combined electrocoagulation, H2O2/Fe2+/UV and activated carbon adsorption for textile wastewater treatment, Sci. Total Environ. 651 (2019) 551–560. [22] G. Hasani, A. Maleki, H. Daraei, R. Ghanbari, M. Safari, G. McKay, K. Yetilmezsoy, F. Ilhan, N. Marzban, A comparative optimization and performance analysis of four different electrocoagulation-flotation processes for humic acid removal from aqueous solutions, Process Saf. Environ. Prot. 121 (2019) 103–117. [23] A. Tsimpliaraki, S. Svinterikos, I. Zuburtikudis, S.I. Marras, C. Panayiotou, Nanofibrous structure of nonwoven mats of electrospun biodegradable polymer nanocomposites—a design of experiments (DOE) study, Ind. Eng. Chem. Res. 48 (9) (2009) 4365–4374. [24] D.O. Aksoy, E. Sagol, Application of central composite design method to coal flotation: modelling, optimization and verification, Fuel 183 (2016) 609–616. [25] D. Krishnaiah, A. Bono, R. Sarbatly, R. Nithyanandam, S.M. Anisuzzaman, Optimisation of spray drying operating conditions of Morinda citrifolia L. fruit extract using response surface methodology, J. King Saud Univ. Eng. Sci. 27 (2015) 26–36. [26] S.S. Ranade, P. Thiagarajan, Selection of a design for response surface, IOP Conf. Ser. Mater. Sci. Eng. 263 (2017) 022043, , https://doi.org/10.1088/1757-899X/ 263/2/022043. [27] J. Zolgharnein, A. Shahmoradi, J.B. Ghasemi, Comparative study of box–behnken, central composite, and doehlert matrix for multivariate optimization of Pb (II) adsorption onto Robinia tree leaves, J. Chemometr. 27 (2013) 12–20. [28] G.E.P. Box, D.W. Behnken, Simplex-sum design: a class of second order rotatable designs derivable from those of first order, Ann. Math. Statist. 31 (1960) 838–864. [29] S.L.C. Ferreira, R.E. Bruns, H.S. Ferreira, G.D. Matos, J.M. David, G.C. Brandão,

[30] [31] [32] [33]

[34]

[35]

[36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46]

[47] [48] [49]

[50] [51]

[52] [53] [54] [55] [56]

11

Chemical Engineering & Processing: Process Intensification 143 (2019) 107619

F. Hussin, et al. [57] E.C. Harrington, The desirability function, Ind. Qual. Control. 21 (1965) 494–498. [58] M. Mourabet, A. El Rhilassi, H. El Boujaady, M. Bennani-Ziatni, A. Taitai, Use of response surface methodology for optimization of fluoride adsorption in an aqueous solution by Brushite, Arab. J. Chem. 10 (2017) S3292–S3302. [59] A. Wicks, Structural Health Monitoring, Volume 5 in Proceedings of the 32nd IMAC, A Conference and Exposition on Structural Dynamics, Springer International Publishing, 2014, https://doi.org/10.1007/978-3-319-04570-259. [60] G. Derringer, R. Suich, Simultaneous optimization of several response variables, Int. J. Qual. Assur. Eng. Technol. Educ. 12 (1980) 214–219. [61] J. Ren, W.Y. Li, K.C. Xie, Surface properties and reactivity of iron-doped titanium oxides catalysts in oxidative dehydrogenation of ethylbenzene with CO2, Petrol. Sci. Technol. 24 (2006) 963–972. [62] F. Silveira, C.F. Petry, D. Pozebon, S.B. Pergher, C. Detoni, F.C. Stedile, J.H.Z. dos Santos, Supported metallocene on mesoporous materials, Appl. Catal. A Gen. 333 (2007) 96–106. [63] Z.P. Hu, Z.M. Gao, X. Liu, Z.Y. Yuan, High surface area activated red mud for efficient removal of methylene blue from wastewater, Adsorp. Sci. Technol. (2017) 1–18. [64] A. Kamari, S.N.M. Yusoff, F. Abdullah, W.P. Putra, Biosorptive removal of Cu(II), Ni (II) and Pb(II) ions from aqueous solutions using coconut dregs residue: adsorption and characterisation studies, J. Environ. Chem. Eng. 2 (2014) 1912–1919. [65] W.P. Putra, A. Kamari, M.S.N. Yusoff, C.F. Ishak, A. Mohamed, N. Hashim, I.M. Isa, Biosorption of cu(II), pb(II) and zn(II) ions from aqueous solutions using selected waste materials: adsorption and characterisation studies, J. Encapsulation. Adsorpt. Sci. 4 (2014) 25–35. [66] K. Yetilmezsoy, S. Demirel, R.J. Vanderbei, Response surface modeling of Pb(II) removal from aqueous solution by Pistacia vera L.: Box–Behnken experimental design, J. Hazard. Mater. 171 (2009) 551–562. [67] S. Rengadurai, B. Preetha, T. Viruthagiri, Response surface technique for optimization of parameters for decolorization of reactive red BS using trametes hiruta, Int. J. Chemtech Res. 4 (2012) 21–28. [68] Y. Liu, Q. Zhao, G. Cheng, H. Xu, Exploring the mechanism of lead(II) adsorption from aqueous solution on ammonium citrate modified spent Lentinus edodes, Chem. Eng. J. 173 (2011) 792–800. [69] E. Kacan, Optimum BET surface areas for activated carbon produced from textile sewage sludges and its application as dye removal, J. Environ. Manage. 166 (2016) 116–123. [70] Stat-ease, Adequate Precision, (2019) (Accessed June 2019), https://www.statease. com/docs/v11/navigation/adequate-precision/. [71] T.K. Trinh, L.S. Kang, Application of response surface method as an experimental design to optimize coagulation tests, Environ. Eng. Res. 15 (2010) 063–070. [72] G. Asgari, B. Roshani, G. Ghanizadeh, The investigation of kinetic and isotherm of fluoride adsorption onto functionalize pumice stone, J. Hazard. Mater. 217–218 (2012) 123–132.

[73] S. Abbaszadeh, S.R. Wan Alwi, C. Webb, N. Ghasemi, I.I. Muhamad, Treatment of lead-contaminated water using activated carbon adsorbent from locally available papaya peel biowaste, J. Clean. Prod. 118 (2016) 210–222. [74] M.A.P. Cechinel, S.M.A.G. Ulson de Souza, A.A. Ulson de Souza, Study of lead (II) adsorption onto activated carbon originating from cow bone, J. Clean. Prod. 65 (2014) 342–349. [75] V. Hernández-Morales, R. Nava, Y.J. Acosta-Silva, S.A. Macías-Sánchez, J.J. PérezBueno, B. Pawelec, Adsorption of lead (II) on SBA-15 mesoporous molecular sieve functionalized with –NH2 groups, Microporous Mesoporous Mater. 160 (2012) 133–142. [76] M. Kobya, E. Demirbas, E. Senturk, M. Ince, Adsorption of heavy metal ions from aqueous solutions by activated carbon prepared from apricot stone, Bioresour. Technol. Rep. 96 (2005) 1518–1521. [77] N. Bouchelkia, L. Mouni, L. Belkhiri, A. Bouzaza, J.C. Bollinger, K. Madani, F. Dahmoune, Removal of lead(II) from water using activated carbon developed from jujube stones, a low-cost sorbent, Sep. Sci. Technol. 51 (2016) 1645–1653. [78] M. Al-Shannag, Z. Al-Qodah, K. Bani-Melhem, M.R. Qtaishat, M. Alkasrawi, Heavy metal ions removal from metal plating wastewater using electrocoagulation: kinetic study and process performance, Chem. Eng. J. 260 (2015) 749–756. [79] V. Orescanin, R. Kollar, K. Nad, The application of the ozonation/electrocoagulation treatment process of the boat pressure washing wastewater, J. Environ. Sci. Health. A. Tox. Hazard. Subst. Environ. Eng. 46 (12) (2011) 1338–1345. [80] F. Ferniza-García, A. Amaya-Chávez, G. Roa-Morales, C.E. Barrera-Díaz, Removal of pb, cu, cd, and zn present in aqueous solution using coupled electrocoagulationphytoremediation treatment, Int. J. Electrochem. Sc. 11 (2017). [81] DOE, Department of Environment, Malaysia Under the Malaysia Environmental Quality (Sewage and Industrial Effluents) Regulations, (2009) http://www.doe.gov. my/eia/wp-content/uploads/2012/03/A-Guide-For-Investors1.pdf. [82] B. Palahouane, N. Drouiche, S. Aoudj, K. Bensadok, Cost-effective electrocoagulation process for the remediation of fluoride from pretreated photovoltaic wastewater, Ind. Eng. Chem. Res. 22 (2015) 127–131. [83] C.J. Nawarkar, V.D. Salkar, Solar powered electrocoagulation system for municipal wastewater treatment, Fuel 237 (2019) 222–226. [84] A. García-García, V. Martínez-Miranda, I.G. Martínez-Cienfuegos, P.T. AlmazánSánchez, M. Castañeda-Juárez, I. Linares-Hernández, Industrial wastewater treatment by electrocoagulation–electrooxidation processes powered by solar cells, Fuel 149 (2015) 46–54. [85] M.H. Karaoğlu, İ. Kula, M. Uğurlu, Adsorption kinetic and equilibrium studies on removal of lead(II) onto glutamic acid/sepiolite, Clean – Soil. Air Water 41 (2013) 548–556. [86] P. Shekinah, K. Kadirvelu, P. Kanmani, P. Senthilkumar, V. Subburam, Adsorption of lead(II) from aqueous solution by activated carbon prepared from Eichhornia, J. Chem. Technol. Biot. 77 (2002) 458–464.

12