Adsorption of Lead from Aqueous Solution Using Activated Carbon Derived from Rice Husk Modified with Lemon Juice

Abstract

In the present work, activated carbon (RHAC-LJ) was synthesized utilizing rice husk as a carbon source that is activated with lemon juice. Moreover, the adsorptive capacity of RHAC-LJ in removing Pb(II) from aqueous solution was investigated. FTIR analysis revealed the presence of amides, alkenes, carboxyl and hydroxyl groups in RHAC-LJ. SEM micrographs illustrate that activation with lemon juice resulted in high pore volume and greater pore diameter. Activation using acid from lemon juice can remove impurities from the adsorbent surface. The surface area and pore volume of RHAC-LJ were determined to be 112.87 m²·g⁻¹ and 0.107 cm³/g, respectively. Adsorption kinetics followed the pseudo-second-order equation (R² = 0.9941) with a rate constant of 3.3697 g/mg·min for Pb (II), which indicates chemisorption to be the rate-determining step of the process. The BBD model using RSM was applied in studying the effects of pH, stirring speed and adsorbent dosage and their interactions on the removal efficiency of RHAC-LJ. Analysis of variance was used to examine the significance of the model, independent parameters, and their interactions. Moreover, a removal efficiency of 98.49% can be attained using the following optimal conditions: 197 rpm, pH 5.49, and adsorbent dosage of 0.3487 g. Overall, the present work illustrates RHAC-LJ as a potential low-cost adsorbent for the removal of Pb(II) from synthetic wastewater.

1. Introduction

Lead, Pb(II), is a heavy metal with a bluish-grey color that is classified as a class one toxic contaminant [1]. It is naturally found in the environment as cerussite, sulfide and galena [2]. There are several anthropogenic sources of Pb(II), including explosive manufacturing, battery production and recycling, fossil fuel combustion, pigment printing and photography [3]. In an aqueous environment, Pb(II) is typically found in oxidation states II and IV [4]. Exposure to high concentrations of Pb(II) in the environment is known to have deleterious effects on human health, including anemia, damage to the nervous system and liver, reproductive disorders in men, high blood pressure and mental illness [5]. In children, Pb(II) can cause lower IQ, learning impairment, social difficulties, and cognitive problems [6]. Numerous countries have implemented regulations to monitor Pb(II) levels in air, water and soil. The Philippine National Standard for Drinking Water enacted in 2017 has set the maximum allowable concentration for Pb(II) at 0.01 mg/L. At present, the US Environmental Protection Agency has set the maximum allowable concentration of Pb(II) to be zero for drinking water. To safeguard the environment and public health, it is mandatory to treat and remove Pb(II) from wastewater before discharge into nearby water bodies.

There are numerous treatment technologies applied in the removal of Pb(II) from water, including membrane filtration, flocculation, ion exchange, chemical precipitation and reverse osmosis. These methods are known to have several shortcomings, such as high demand for energy, costly to operate and maintain, and production of sludge that requires special handling [7]. Adsorption presents an attractive alternative since it is cost-effective, easy to operate and control [8]. The adsorption of Pb(II) from water and wastewater is affected by multiple parameters, including pH, contact time, stirring speed, adsorbent dosage, and initial concentration of contaminant. Most of the previous studies on adsorption have applied the one-factor-at-a-time method, which determines an inaccurate optimal point caused by disregarding the interaction between independent factors [9]. Response surface methodology (RSM) can be integrated with the experimental method for the prediction of nonlinear data and optimization of the variables that could result in better treatment capacity. RSM is a statistical, mathematical tool used in the optimization of complicated processes and assessment of inaccuracies in variables [10]. RSM has numerous applications in biological and industrial procedures, applied chemistry, analytical chemistry, and environmental treatment technologies. Design of experiments (DoE) is a useful mathematical technique that decreases the number of experimental runs without affecting the results’ reproducibility and accuracy [11]. The application of DoE would mean decreased levels to assess the independent variables and their interactions. There are several types of DoE, including the Doehlert Matrix design, Fractional Factorial design, Box-Behnken design (BBD) and Central Composite design. BBD is a multilevel design that is commonly used in RSM due to its decreased number of experimental runs, which proves to be low-cost and practical [10]. BBD lowers the number of combinations of experimental runs by avoiding experiments at their domain extremities, where factor variations usually provide an unacceptable outcome [12].

The utilization of activated carbon in adsorption technologies has gained importance as an adsorbent in the treatment of industrial wastewater derived from chemical manufacturing, food, textiles, and pharmaceuticals. Activated carbon, which is a type of carbonized material, has a variety of applications attributed to its extensive porosity, improved chemical stability, high mechanical strength, and high surface area of greater than 400 m² g⁻¹ [13]. The adsorption characteristics of activated carbon are correlated to the numerous functional groups found on its surface, including quinone, lactone, phenol, carbonyl and carboxyl [14]. However, there are major limitations in the commercial application of activated carbon, such as its costly synthesis [15], expensive regeneration method and use of nonrenewable carbon sources including lignite, anthracite, coal, peat and hydrogen gas [16,17]. To promote a green circular economy, research on low-cost, renewable carbon resources, including agricultural waste, as alternative materials has been investigated. The use of biomass feedstocks that include forestry and agricultural wastes and by-products offers a diverse source for the production of activated carbon. There are several advantages, such as the sustainable and renewable nature of biomass, and less dependence on nonrenewable resources [18,19]. Previous studies have examined sugarcane bagasse ash [20], coconut shells [21], orange peels [22], rice husks [23,24,25], rapeseed [26], and sludge [27] as alternative carbon sources in the synthesis of activated carbon.

Rice production worldwide is estimated at 698 million metric tons in 2021 [28]. Asian countries considered major rice producers include Japan, Thailand, Vietnam, Bangladesh, Philippines, China and Indonesia, which account for 90% of the global rice production [29,30]. Rice husks refer to the outer part of the rice grain. It comprises about 20–33% of the total weight of rice, which implies about 139.54 to 230.24 million tons of rice husks are generated. It is a ligno-cellulosic substance that is composed of 5% ash, 15% carbon, 60% volatile matter and 20% silica [4]. Rice husk has attracted attention as a precursor for activated carbon due to its affordability, nontoxicity and abundance. The Philippines is ranked eighth in the world as a rice-producing country and generates about 2.68 tons of rice husk, which is estimated to increase by 76.2 ktons every year [31]. The conversion of agricultural solid waste, such as rice husks, into value-added products promotes environmental preservation and solid waste reduction. In addition, this also reduces the cost of solid waste disposal and the need for incineration. The reuse and recycling of rice husks mitigate air and soil contamination that is caused by open dumping and burning of agricultural wastes [32]. Numerous adsorbents, including rice hull ash [33], rice husk ash [34] and biochar derived from rice husks [35,36] were investigated in the removal of Pb(II) from wastewater and water. Naiya et al., (2009) examined rice husk ash acquired from local millers and determined its adsorption capacity to be 91.74 mg/g for Pb(II) removal [34]. Xu et al., (2013) used rice husk-derived biochar for the adsorption of Cu(II), Cd(II), Pb(II) and Zn(II) via complexation [35]. Shi et al., (2019) determined that rice husk biochar produced at 700 °C provided a high adsorption capacity of 26.7 mg/g for Pb(II) [36].

Activated carbon is comprised of an irregular group of carbon atoms. It is activated using physical and chemical activation. Physical activation is a method that carbonizes rice husks under an inert atmosphere at temperatures ranging from 400 to 600 °C [37]. However, the biochar produced has blocked pores, resulting in low adsorption capacity. Hence, there is a need to undergo a second stage of carbonization that would expose the biochar to even higher temperatures (600 to 700 °C) in the presence of carbon dioxide, steam, or air. This would result in the removal of tarry substances, development of new pores, increase in pore diameter, higher pore volume and large surface area [37,38]. On the other hand, chemical activation is a single-step method where activation and carbonization of rice husks proceed simultaneously. The chemical reagents are known to have greater surface contact and deeper penetration within the carbon structure, which can lead to smaller pores developing. Chemical activation has several advantages over physical activation, which include greater product yield and low operating temperature. The process typically employs activating agents, including KOH [39,40], phytic acid [41], K₂CO₃ [42], NaOH [43,44], H₃PO₄ [45], H₂SO₄ [46], citric acid [47,48,49] and ZnCl₂ [2]. Activating agents of the acidic nature were determined to be more efficient in improving the physicochemical properties of activated carbon derived from rice husks [50,51,52]. However, activation often utilizes commercial reagents that are costly, toxic in nature, and involve a complex method for its handling and preparation [47]. Citric acid is an organic, triprotic, weak acid with a variety of applications in food preservation [47]. International and national food regulatory agencies accept citric acid as a safe substance to be applied in food applications [53]. It is a hydroxy-tricarboxylic acid with the capability to form polydentate and polynuclear complexes with actinides and metal ions [54]. Citric acid has been widely applied for surface modification by adding carboxyl functional groups on the adsorbent surface [49]. Several organic acids are contained in lemon juice, including malic acid, succinic acid, ascorbic acid, oxalic acid and citric acid. Lemon juice mainly contains citric acid, which is the prevalent acid with a high concentration of 48 g/L and contributes 95% of the total acidity [55]. Lemon juice can serve as an organic substitute for commercially available citric acid. Lemon juice, which is a potential source of citric acid, has been utilized in organic synthesis as an acidic catalyst [56]. A literature review reveals a research gap in the synthesis of activated carbon using rice husk as a renewable carbon source and citric acid from lemon juice as an activating agent. Furthermore, optimization studies using RSM will be utilized to investigate the possibility of rice husk-derived activated carbon (RHAC-LJ) as a potential adsorbent of Pb(II) from water.

In the present work, an ecofriendly activated carbon will be developed via carbonization using locally available rice husk with lemon juice as an activating agent. The basic physicochemical properties of RHAC-LJ will be determined using scanning electron microscopy (SEM), Brunauer–Emmett–Teller (BET) analysis and Fourier Transform Infrared spectroscopy (FTIR). The effect of contact time on the feasibility of RHAC-LJ in removing Pb(II) from aqueous solution will be investigated. Kinetic studies will be carried out to evaluate the adsorption rate and adsorption efficiency. Optimization assessment will be performed to model and examine various factors (pH, adsorbent dosage, stirring speed) and their effect on the removal efficiency of Pb(II) using RSM with BBD.

2. Materials and Methods

2.1. Chemicals and Materials

Rice husk and lemon juice were obtained from Quezon Province, Philippines. Lead nitrate, Pb(NO₃)₂ was purchased from RTC Laboratory Services and Supply House (Quezon City, Philippines), while HCl (37% fuming) and NaOH (98% purity) pellets were acquired from Merck (Kenilworth, NJ, USA). All reagents and chemicals used in the study are of analytical grade.

2.2. Preparation of Rice Husk Derived from Activated Carbon

Rice husks were washed several times with deionized water to remove dirt and debris. Then, 100 g rice husk was dried in the oven (Biobase ANS-20, Calamba, Philippines) at 110 °C for 24 h. About 18 g rice husk were placed in a 30-mL crucible and was carbonized in a muffle furnace (Biobase, Philippines) for 2 h at 500 °C. Then, the carbonized rice husk was stirred in with 100 mL lemon juice for 4 h at 100 rpm. Then, the mixture was allowed to impregnate in an oven (Biobase ANS-20, Philippines) at 150 °C, after which the treated rice husk was washed with deionized water several times until a neutral pH was attained. The RHAC-LJ was dried in an oven at 110 °C for 48 h and was cooled in a desiccator. Finally, RHAC-LJ was crushed and passed through a 40-mesh and 70-mesh sieve screen (ASTM Standard Sieve Series, West Conshohocken, PA, USA), where RHAC-LJ with particle diameters ranging from 0.210 to 0.430 mm were utilized in the study.

2.3. Characterization Analysis

Scanning electron microscopy (SEM model JSM-6490LV JEOL Ltd., Tokyo, Japan) analyzed the physical and surface properties of RHAC-LJ using 10 µm for working distance at 3 kV accelerating voltage. Fourier transform infrared spectroscopy (FTIR Affinity 1S Model Shimadzu, Kyoto, Japan) was used to analyze the functional groups on the adsorbent with 20 scans and 4 cm⁻¹ resolution within a range of 4000 to 400 cm⁻¹. KBr pellets were prepared from 250 mg KBr and 1 mg RHAC-LJ for the IR mounting. The BET surface analysis (Micrometrics 2020) was used to measure the pore volume and specific surface area of RHAC-LJ. Solid addition experiments were carried out to determine the point of zero charge of RHAC-LJ. About 25 mL of 0.1 M NaCl solution was prepared and added with a predetermined dosage of RHAC-LJ. The mixture was shaken for 48 h and the final pH was measured.

2.4. Kinetic Study

Batch experiments were performed by placing 0.4 g RHAC-LJ in a 100 mL solution at pH 6. The solution was agitated at 200 rpm under various contact times (1, 3, 5, 10, 20, 30, 60, 120, 180 and 240 min). After each run, the solution was filtered and residual Pb(II) concentration was measured using atomic absorption spectroscopy (AAS Shimadzu AA6300, Japan). Triplicate runs were carried out.

The adsorption capacity (q_t) was calculated using Equation (1):

$$q_t=\frac{\left(C_0-C_t\right)V}{m}$$

(1)

where m is the mass of adsorbent (g), V is the volume of the solution (L), C₀ and C_t refer to the initial concentration and concentration of Pb(II) at any time t (min), respectively.

The removal efficiency (RE) is computed using Equation (2):

$$RE=\left(\frac{C_0-C_e}{C_e}\right)\times 100$$

(2)

where C_e refers to the concentration of Pb(II) at equilibrium.

2.5. Experimental Design

The DoE for the adsorption of Pb(II) using RHAC-LJ was carried out using BBD for RSM. The response variable was assigned to be the removal efficiency (RE) and the three independent variables are pH, adsorbent dosage and stirring speed (

Table 1. Different levels of the independent process factors applied in the adsorption of Pb(II) using RHAC-LJ.

Operating Factor	Unit	Range and Levels
		−1	0	1
pH		4	5	6
Adsorbent dosage	g	0.2	0.4	0.6
Stirring speed	rpm	100	150	200

). Design expert software (v 13.1.0 55413 Minneapolis, MN, USA) was used in data analysis and statistical design.

The relationship between the independent variables and responses was investigated using a quadratic response surface equation (Equation (3)):

$$Y=b_0+{\sum }_{i=1}^k{b}_i{X}_i+{\sum }_{i=1}^k{b}_{ii}{X}_i^2+{\sum }_{i<j}^k{b}_{ij}{X}_i{X}_j+D_r$$

(3)

where b₀ is the coefficient constant, Y is the response variable, D_r refers to the error from the model, X_j and X_i refer to input variables of the adsorption system, and b_i, b_ii, and b_ij are interaction coefficients.

The interaction between the responses and their input process parameters, as well as the adequacy of the quadratic model, was assessed using the Fisher’s F-test (95% confidence level) and analysis of variance (ANOVA). The adjusted R² and coefficient of determination (R²) determined the prediction ability of the proposed models.

3. Results and Discussion

3.1. Characterization of Materials

Figure 1 illustrates the SEM micrographs of RHAC and RHAC-LJ under various magnifications. Figure 1a,b show a relatively smooth surface of RHAC with the presence of a few fine pores on the surface. After activation with lemon juice, the morphology of RHAC-LJ (Figure 1c,d) showed a higher number of vacant cavities and pores with bigger diameters. The surface area, pore diameter and pore volume of the carbonized rice husk were determined to be 22.79 m²·g⁻¹, 3.835 nm and 0.043 cm³/g, respectively. An increase in surface area, pore diameter and pore volume were observed after activation with lemon juice. The pore volume, pore diameter and surface area of RHAC-LJ were measured to be 112.87 m²·g⁻¹, 3.947 nm and 0.1074 cm³/g, respectively. Based on the pore diameter values, both carbonized rice husk and RHAC-LJ are classified as mesoporous materials (2 nm < d < 50 nm). The results of the BET analysis agreed with the results of SEM micrographs. The addition of lemon juice as an activating agent in RHAC suggests that the impurities present were washed off and solubilized, which would expose voids and pores. Similar results were observed by Xu et al., (2016) and Chen et al., (2003) where improved pore volume and pore radius were attained after activation with citric acid [48,57].

Figure 2 shows the FTIR spectrum of RHAC-LJ where several peaks were observed that can be attributed to activated carbon. The sharp peaks at 1740 cm⁻¹ and 1367 cm⁻¹ are due to the C-O stretching vibration of the carboxyl group [48]. The peak at 1219 cm⁻¹ indicates the presence of C-O stretching of aliphatic ether [58]. The band at 1079 cm⁻¹ refers to asymmetrical stretching of C-O-C functional group, while the band at 809 cm⁻¹ refers to C-H bending [58]. Results show the presence of oxygen-containing functional groups in RHAC-LJ could be attributed to the low-carbonization temperature of 500 °C used in the present work. The previous study by Ahmad et al., (2012) illustrates that higher pyrolysis temperature of 700 °C would cause the elimination of functional groups such as O-H, C-H and C-O in activated carbon [59]. After adsorption of Pb(II), the peak was observed to have shifted from 1740 to 1842 cm⁻¹, which implies the involvement of C-O of the carboxyl group in the removal of Pb(II). A new peak is also observed at 3015 that can be attributed to the specific interactions between Pb(II) and C-H of the alkane functional group during adsorption.

3.2. Effect of Contact Time

Figure 3 illustrates the effect of contact time on the treatment performance of RHAC-LJ for the removal of Pb(II). A similar trend was observed where increasing the contact time would result in a corresponding increase in both removal efficiency (Figure 3a) and q_t (Figure 3b). As the contact time increases from 1 to 240 min, the removal efficiency was observed to increase from 41.24% to 88.33%, respectively. Meanwhile, q_t after 1 min was observed to be 1.18 mg/g and increased to 2.54 mg/g at 240 min.

Adsorption of Pb(II) was observed to occur rapidly in the first few minutes, which is attributed to the availability of a high number of vacant binding sites on the adsorbent surface. After 5 min, a sharp increase in removal (68.95%) was attained, which suggests a high adsorption rate. From 10 to 180 min, a gradual increase in both removal efficiency and q_t was observed due to the lower number of active sites available for adsorption. This indicates that the binding sites are slowly reaching saturation [4]. After 180 min, equilibrium was attained where there are no significant changes observed in the removal efficiency and q_t.

3.3. Kinetic Study

Adsorption kinetics assists in understanding the efficacy of the adsorbent material. Kinetic models, such as pseudo-first-order, pseudo-second-order and intraparticle diffusion equations, were utilized to analyze the experimental data. Equations (4)–(6) illustrate the linear form of the pseudo-first-order [60], pseudo-second-order [61] and intraparticle diffusion model [62]:

$$log\left(q_e-q_t\right)=log{q}_e-\frac{k_1}{2.303}t$$

(4)

$$\frac{t}{q_t}=\frac{1}{k_2{q}_e^2}+\frac{t}{q_e}$$

(5)

$$q_t=k_{id}{t}^{0.5}+C$$

(6)

where q_e refers to adsorption capacity at equilibrium (mg/g), C refers to the boundary layer thickness, k_id (g/mg·min^0.5) refers to rate constant of intraparticle diffusion model, k₂ (g/mg·min) and k₁ (min⁻¹) refer to the rate constant of the pseudo-second-order and pseudo-first-order equations, respectively.

Results in

Table 2. Adsorption kinetic constants for the removal of Pb(II) using RHAC-LJ.

Kinetic Model	Parameters
Pseudo-first order	qe,theoretical (mg/g)	0.9499
	k1 (min−1)	0.000852
	R2	0.8292
Pseudo-second order	qe,theoretical (mg/g)	2.5439
	k2 (g/mg·min)	3.3697
	R2	0.9941
Intraparticle diffusion	C	1.6330
	kid (g/mg·min0.5)	0.0635
	R2	0.6980

show that pseudo-first-order and intraparticle diffusion equations have low values of the coefficient of determination (R²), which implies these equations do not best describe the kinetic data. On the other hand, the pseudo-second-order equation provided the highest R² (0.9941). Figure 4 shows that there is a good agreement observed between the theoretical plot and experimental data generated by the pseudo-second-order equation. In addition, the theoretical q_e (2.5439 mg/g) provided by the pseudo-second-order equation and experimental q_e (2.687 mg/g) have similar values. Overall, results further confirm that Pb(II) adsorption using RHAC-LJ follows a pseudo-second-order kinetic, which indicates that the rate-determining step is chemisorption.

Table 3. Comparison of the adsorption performance of various biochar and activated carbon for the removal of Pb(II).

Adsorbent	Activating Agent	Surface Area (m2/g)	Pore Diameter (nm)	Kinetic Rate Constant (g/mg·min)	Reference
Rice husk ash	Not applicable	57.50	not available	0.267	[34]
Activated carbon from black cumin waste	K2CO3	2211	2.80	0.00151	[63]
Activated carbon from coffee residue	ZnCl2	890	not available	0.123	[64]
Biochar from rice husk	Not applicable	27.8	not available	not available	[35]
Rice husk biochar	Not applicable	193.15	6.80	0.0179	[36]
RHAC-LJ	Lemon juice	112.87	3.95	3.369	Present work

shows the characteristics and kinetic rates of various biochar and activated carbon derived from biomass feedstock that have been previously reported. The use of commercial activating agents such as ZnCl₂ and K₂CO₃ resulted in high surface area and pore diameter when compared to RHAC-LJ. However, the kinetic rate constant for RHAC-LJ is the highest among the adsorbents, which implies lemon juice can provide a satisfactory surface area, high pore diameter and high adsorption rate.

3.4. Optimization Using RSM Modelling

3.4.1. Establishment of the Mathematical Regression Model

Statistical modelling via BBD under RSM was used to investigate the effect of pH, adsorbent dosage and stirring speed and examine their corresponding interactions. In addition, RSM-BBD was applied to determine the optimal parameters for the treatment efficiency of Pb(II) using RHAC-LJ. The quadratic regression model is provided to infer a response when given a set of parameters and is given in Equation (7):

$$\begin{array}{l}y=98.05+27.33A-0.2800B-0.5237C-2.46AB-0.4600AC+0.4325BC\\ -26.58A^2-2.91B^2+3.40C^2\end{array}$$

(7)

where y refers to the removal efficiency, A refers to the pH, B refers to the adsorbent dosage, C refers to the contact time, AC, BC, A², B² and C² refer to model terms. In the equation, the presence of synergy is indicated by positive coefficients, while negative coefficients imply a decreasing effect on the removal efficiency. Equation (7) shows that interactive model term BC has a positive, significant impact on the removal of Pb(II).

Table 4. Results of the design of experiments using RSM.

Run	Independent Variables
	pH	Adsorbent Dosage (g)	Stirring Speed (rpm)	Removal Efficiency (%)
1	5	0.2	200	99.99
2	5	0.4	150	98.32
3	5	0.6	200	98.22
4	5	0.6	100	96.24
5	4	0.4	200	50.73
6	6	0.4	100	99.93
7	5	0.2	100	99.74
8	4	0.2	150	35.78
9	6	0.6	150	96.43
10	6	0.2	150	99.83
11	4	0.6	150	42.21
12	5	0.4	150	97.66
13	4	0.4	100	48.83
14	5	0.4	150	97.75
15	6	0.4	200	99.99
16	5	0.4	150	97.94
17	5	0.4	150	98.60

displays the results and the experimental design for the 17 experimental runs. A high value for the coefficient of determination (R² = 0.9876) implies a good fit and any disparity in the actual values can be validated by the model.

3.4.2. Statistical Analysis and Interaction of Variables

The goodness of fit of the model was determined using analysis of variance (ANOVA), while the significance and acceptability of model terms were determined using the F-value and p-value (<0.0001). As shown in

Table 5. Results of the ANOVA analysis for the quadratic model.

Source	Sum of Squares	Degrees of Freedom	Mean Square	F-Value	p-Value
Model	9070.63	9	1007.85	142.97	<0.0001
A	5974.88	1	5974.88	847.57	<0.0001
B	0.6272	1	0.6272	0.0890	0.7741
C	2.19	1	2.19	0.3113	0.5943
AB	24.16	1	24.16	3.43	0.1066
AC	0.8464	1	0.8464	0.1201	0.7391
BC	0.7482	1	0.7482	0.1061	0.7541
A2	2975.73	1	2975.73	422.12	<0.0001
B2	35.58	1	35.58	5.05	0.0595
C2	48.69	1	48.69	6.91	0.0340
Residual	49.35	7	7.05
Lack of Fit	48.72	3	16.24	103.18	0.0003
Pure Error	0.6295	4	0.1574
Cor Total	9119.98	16

, the model is significant based on its low p-value (<0.0001) and high F-value (142.97). Model terms such as A and A² are highly significant and can affect the removal process of Pb(II) using RHAC-LJ. The rest of the model terms (B, C, AB, BC, AC, B², C²) are not significant (p > 0.05). The pH of the solution has a significant effect on the removal efficiency of Pb(II). As seen in Table 2, pH 5 to 6 provided the highest removal efficiency, ranging from 96.24% to 99.99% regardless of the adsorbent dosage and stirring speed. The point of zero charge (pH_zpc) of RHAC-LJ was determined to be pH 4.58. The protonated surface functional groups of RHAC-LJ (O-H, C-O) occur at a more acidic pH of less than 4.0. Therefore, electrostatic repulsion occurs between Pb(II) and the positive surface of RHAC-LJ. When the pH of the solution was increased from pH 5 to 6, the adsorption of Pb(II) increased as well. This is attributed to the electrostatic attraction between the negatively charged surface of RHAC-LJ and positive Pb(II) ions.

Figure 5a–c illustrate the 3D contour plots that show the removal efficiency in response to interactions of the three independent variables. Figure 5a shows a higher removal efficiency as pH increases from 4 to 6 while varying the adsorbent dosage has insignificant effect. The same trend is observed for Figure 5b where increasing the pH has a significant increase in the removal efficiency while stirring speed has no observed effect. Lastly, the influence of stirring speed and adsorbent dosage has no considerable effect on the removal efficiency of the adsorbent (Figure 5c). Results in Figure 5d show a good agreement between the predicted and actual values on the removal efficiency of Pb(II). This implies that the experimental results and statistical predicted values have a good correlation.

3.4.3. Optimization and Validation of the Model

The regression equation was determined and the surface response plots were assessed to attain the optimized conditions. The optimal conditions were determined to be pH 5.49, adsorbent dosage of 0.3487 g and 197 rpm that provided a predicted removal of Pb(II) of 98.49%. To examine the accuracy of the response model, a total of three runs for the validation tests were conducted using the optimum conditions, as seen in

Table 6. Confirmatory runs using optimum conditions for the adsorption of Pb(II) using RHAC-LJ.

		Removal Efficiency (%)
Predicted		98.49%
Observed	Run 1	97.05%
	Run 2	96.32%
	Run 3	97.17%

. Results show that the observed values were in good agreement with the predicted values derived from the model. This demonstrates the effectiveness and adequacy of the model.

4. Conclusions

The present work developed activated carbon (RHAC-LJ) via carbonization using rice husk as a carbon resource and lemon juice as an activating agent. SEM micrographs showed that more pores with bigger diameters were developed after using lemon juice as an acidic agent for activation. Adsorption kinetics show that the pseudo-second-order model with its R² value of 0.9941 can adequately describe the Pb(II) adsorption using RHAC-LJ, which indicates that chemisorption is the rate-limiting step of the process. The effects of pH, stirring speed and adsorbent dosage were examined using the BBD matrix with RSM. The model predictability and significance of model terms were assessed using ANOVA. Among the independent variables, pH (p < 0.0001) was determined to be the most significant factor in enhancing the removal efficiency of Pb(II). The predicted removal efficiency of 98.49% can be attained using the optimal conditions of pH 5.49, adsorbent dosage of 0.3487 g and stirring speed of 197 rpm. In general, a high removal efficiency of Pb(II) can be attained using RHAC-LJ, which implies its potential as an adsorbent for wastewater treatment.