Adsorption of Cu(II) and Zn(II) onto Activated Carbon from Oil Palm Empty Fruit Bunch Prepared by Two-Step Acid Treatment and Microwave-Assisted Pyrolysis

This study explores the feasibility of biochar-based activated carbon derived from oil palm empty fruit bunch (EFB) as a potential precursor for the preparation of activated carbon via 2-step H3PO4 activation under microwave-assisted pyrolysis (2ACEFB). The characterization of EFB and 2ACEFB was observed by FTIR and BET, and chemical composition was determined using proximate and elemental analysis data. The adsorptive removal of Cu(II) and Zn(II) from an aqueous solution was studied, and the effects of metal concentration and solution pH were also investigated. The pseudo-second-order equation was properly described, providing the best fit to the observed experimental data. The adsorption capacities of Cu(II) and Zn(II) onto the EFB were 20.28 and 18.06 mg/g, respectively, and improved by 2.04- and 1.89-fold onto the 2ACEFB. The potential of 2ACEFB was also proved by adsorbent reusability with five consecutive circles of the batch experiment without regeneration or treatment. This study demonstrated that 2ACEFB is an efficient adsorbent for eliminating heavy metals from aqueous solutions.


Introduction
The amount of heavy metals, harmful to humans and most animal species, is released from chemical industries such as the production of electric batteries, mining, and glass manufacturing factories. Several techniques, e.g., chemical reduction/precipitation, ion exchange, membrane separation, and adsorption, have been developed to eliminate heavy metal ions from wastewater. Among these techniques, adsorption is a widely used method for the removal of heavy metal ions due to more benefits that have been recognized, including high efficiency, easy handling, economic effectiveness, and the availability of many types of adsorbents [1][2][3].
In biomass adsorbents, activated carbons are excellent adsorbents because of their high performance with high sur-face area and a high degree of microporosity. However, the efficiency expected of the activated carbons dramatically depends on their precursors, activators, and treatment/production techniques. So, relatively expensive, nonrenewable precursors, including long-time heating production methods are a significant shortcoming for commercial activated carbon [2,4]. Numerous researches have focused on different sources of adsorbent and found that the use of biological material is one of the most effective ways to use as precursors. The biological precursors (agricultural waste, biomass, etc.) have been considered an alternative and high-potential material for activated carbons due to their plenty and renewable nature [5]. Either with or without activation treatment, the initial process of activated carbon is required. Carbonization is the most widely used method, being a low-cost technique in which organic matter can be transferred to elemental carbons by either conventional or microwave-assisted pyrolysis.
Microwave-assisted pyrolysis becomes advantageous because it provides a fast and selective heating mechanism [6] which can overcome many restrictions obtained from conventional heating regulation. Furthermore, during the microwave-assisted pyrolysis process, the media molecules hold a high frequency of vibration and frictional collision, resulting in a sudden increase in temperature. Thus, the treatment time can be decreased due to microwave energy. Numerous studies have shown that microwave-assisted pyrolysis can produce activated carbon with a higher specific surface area compared to conventional pyrolysis methods. This is due to the more rapid heating rates, which can lead to more efficient carbonization and activation of the material, resulting in a higher surface area and more uniform pore size distribution which can enhance its adsorption capacity "Microwave-assisted pyrolysis of coconut shell for production of activated carbon: Optimization using response surface methodology" [7], An et al. 2016 "Microwave-assisted pyrolysis of biomass for production of high-quality biooil and biochar with catalyst recycling by using activated carbon [8].
Biochar-based activated carbons have been used in several applications, providing high efficiency for sorption [9][10][11]. Acid activation is a simple and effective process, generally performed in one-step activation, for enhancing the adsorptive properties, functionalities, specific surface area, and selectivity of different adsorbates. Previously, twostage activation was a process used in the preparation of activated carbon, providing more suitable activated carbon with a highly porous structure and useful functional groups for the removal of heavy metal ions from aqueous media (a novel route for the preparation of chemically activated carbon from pistachio wood for highly efficient Pb(II) sorption, activated carbon from wood wastes for the removal of uranium and thorium ions through modification with mineral acid). However, it is necessary to prove the efficiency of the produced activated carbon with different adsorbates. Therefore, this study attempted to examine the potential of activated carbons derived from palm oil empty fruit bunch via 2-step H 3 PO 4 activation. The effects of different response parameters such as solution pH, metal ion concentration, and contact time were studied. The adsorption capacity of activated carbon as an adsorbent onto Cu(II) and Zn(II) was evaluated by a two-and three-parameter adsorption isotherm.
Heavy metals are toxic to the human body. Whether in the form of pure elements, organic compounds, and inorganic compounds, heavy metal's toxicity manifests itself when sufficient amounts are accumulated in the body. Copper and zinc are used mainly in the automotive and industrial paint industries. In the study of the removal of Cu(II) and Zn(II), the purpose was to use the synthesized activated carbon in the paint industry effluent treatment before releasing it into the natural water source.

Experimental and Methodology
2.1. Material and Chemicals. The oil palm empty fruit bunches utilized as raw materials in this investigation were collected in Surat Thani, South of Thailand. All of the substances used in this study were of analytical grade. Heavy metal salts CuSO 4 ·5H 2 O and Zn(NO 3 ) 2 , as well as H 3 PO 4 , purchased from Ajax Finechem and Loba Chemie, respectively, were utilized to prepare Cu(II) and Zn(II) stock solutions. Copper(II) sulfate pentahydrate (CuSO 4 ·5H 2 O) for analysis in EMSURE® and zinc standard solution traceable to SRM from NIST Zn(NO 3 ) 2 in HNO 3 0.5 mol/l 1000 mg/ l Zn Certipur®.

Adsorbent Preparation.
Oil palm empty fruit bunches were gathered in Surat Thani, Southern Thailand, for usage as a raw material in this study. The activated carbon was made in the following processes: the raw materials were crushed and sieved into tiny sizes ranging from 1.0 to 2.0 mm and then pyrolyzed in a fixed bed reactor with flowing N 2 at 450°C for 1.5 hours, yielding biochar (hereafter named EFB) [12,13]. The biochar was soaked in a phosphoric acid solution with a biochar/H 3 PO 4 impregnation ratio of 1 : 1.75 (wt%) and periodic stirring [14]. The activation process was carried out at room temperature overnight. The resultant-activated carbon was obtained and then placed in the microwave-induced reactor for 10 min at 600 W and then washed with DI water until it reached pH 6-7. Overall activation processes were performed as duplicates, hereafter named 2ACEFB, in order to improve the adsorbent surface and adsorption efficiency, providing more efficacious [15].

Batch Adsorption.
A series of batch experiments were conducted to evaluate the performance of 2ACEFB for the removal of Cu(II) and Zn(II) in an aqueous solution. Stock solutions were initially prepared by dissolving 3.929 g of CuSO 4 ·5H 2 O and 2.896 g of Zn(NO 3 ) 2 in 1 L of distilled water. Standard metal solutions ranged 50-800 mg/L were then prepared appropriately by diluting the stock solution. Eventually, the pH of the metal solutions was adjusted by 0.1 M NaOH or HCl.

Adsorption Science & Technology
In a batch adsorption study, 50 mL of prepared solution was placed in a 250 mL conical flask, and 1 g of adsorbent was added to the flask placing it in an incubator under 150 rpm at a controlled temperature (298 K). Factors affecting the adsorption efficiency such as pH solution (2-6), contact time (0, 10, 20, 40, 60, 90, 120, 180, 240, and 360 min), and initial metal concentration (12.5, 25, 50, 100, 150, 200, 300, 400, 600, and 800 mg/L) were intentionally investigated to determine the optimal condition, qualifying the adsorption and reusability performance of the adsorbent. The amount of adsorption adsorbed can be calculated as follow: where q e is the amount of metal ions adsorbed per mass of adsorbent at equilibrium (mg/g), V is the volume (L) of solution used for batch experiments, m is the mass (g) of adsorbent, and C o and C e are the initial concentration of metal ions and the equilibrium concentration of the solution (mg/g), respectively.

Instrumentation. The concentrations of Cu(II) and
Zn(II) were measured by atomic absorption spectrophotometry (AAS) using AAnalyst 100 Spectrometer, PerkinElmer, Norwalk, CT/USA. EFB and 2ACEFB were characterized with an X-ray fluorescence spectrometer, PW2400, Philips, Netherlands. The FTIR spectra of the biochar EFB and 2ACEFB were recorded from KBr pellets with a Fourier transform infrared spectrometer, Vertex 70, Bruker, Germany. Surface area and porosity were determined using BET-technique, ASAP2460, Micromeritics, USA.

Result and Discussion
3.1. Chemical Composition and BET. The proximate analysis, in terms of moisture content, ash content, volatile matter, and fixed carbon, was carried out according to the ASTM standard. The result revealed that the composition of biochar EFB was 4.82 wt% moisture, 28.70 wt% volatile, 60.54 wt% fixed carbon, and 5.95 wt% ash (Table 1). It is clear that EFB contains high fixed carbon and low ash content, making it possible to be used as an activated carbon precursor [16]. After the 2-step H 3 PO 4 treatment with a 1.75 impregnation ratio, the proximate composition of the 2ACEFB was 6.15 wt% moisture, 25.50 wt% volatile, 64.49 wt% fixed carbon, and 3.87 wt% ash.
The BET result showed that the surface area of 2ACEFB reached 816.68 m 2 /g, an increase over biochar EFB, as shown in Table 1. This result is similar to those previously reported in the comprehensive literature [18,19]. The pore volume of the adsorbent increased corresponding with a high surface area which can imply a large number of active sites, and the high mesopore volume enhances the mass transfer rate [19]. This behavior can be described as phosphoric acid acting as a template for creating microporosity during the activation stage [20].

Surface Functionality and Morphology.
Carbon matric is a complex nature that not only consists of the carbon atom but is also formed by others such as oxygen, nitrogen, halogen, sulfur, and phosphorus. These heteroatoms are bonded to the carbon surface layer at the edges [18]. The FTIR spectra of biochar (EFB) and 2-step H 3 PO 4 -activated carbon (2ACEFB) are shown in Figure 1. The broadband at the peak of 3400 cm -1 corresponds to the -OH groups, hydroxyl groups from carboxyls, phenols, and adsorbed water [21,22]. The band at 2922 cm -1 is due to the C-H stretching of methyl and methylene groups [21]. The region of the spectrum 1300 and 900 cm -1 consists of different bands being assigned to the C-O stretching in acids, alcohols, phenols, ethers, and esters found in oxidized carbons. However, it is noteworthy to mention that the phosphorous and phosphorous carbonaceous compounds existing in the phosphoric acid-activated carbon are presented in the band region between 900 and 1300 cm -1 [22].

Adsorption Science & Technology
The morphology of empty fruit bunches (EFB) demonstrates a surface that is heterogeneous, containing varying amounts of debris which arise as a consequence of the pyrolysis process. The imaging utilizing energy-dispersive X-ray spectroscopy (EDS) indicated the existence of several organic components, including phosphorus, which has the potential to be incorporated in the formation of a phosphate skin on the surface of EFB after the treatment with phosphate acid (Figure 2).

Adsorption Kinetics.
To examine the controlling mechanism of the adsorption process, the adsorption data were evaluated in terms of pseudo-first-and second-order-kinetic models. [23] The model expressions in terms of nonlinear and linearized forms are shown in Table 2.
The results revealed that models were relevant to the experimental data with a coefficient of determination (R 2 ) of data with a coefficient of determination (R 2 ) was more than 0.95. However, the kinetics of Cu(II) and Zn(II) onto EFB and 2ACEFB were adequately described by the pseudosecond-order model, in which the kinetic parameters determined are tabulated in Table 3. It can be seen from Figure 3, a nonlinear plot with the kinetic parameters from Table 3, that a relationship between the adsorption uptake (q e ) from the experiments (q e,exp ) and the calculated value (q e,cal ) was a good fit with R 2 ranged from 0.998 to 0.999. This means that the     Freundlich q e = K F C 1/n e ln q e = ln K F + 1 n ln C e K F , n

Adsorption Science & Technology
Temkin  Figure 6: Nonlinear data fitting of isotherm models for Cu(II) and Zn(II) adsorption onto EFB and 2ACEFB under optimal condition for (a, b) a two-parameter model and for (c, d) three parameters. 6 Adsorption Science & Technology pseudo-second-order model can be appropriately used to describe the adsorption kinetics. The amount of uptake for all experiments reached 84% by 60 min and gradually increased to the equilibrium time of 240 min. However, to ensure equilibrium uptake, 360 min equilibration time was used when determining the adsorption isotherms in the next section. Figure 4 illustrates the relationship of metal ion concentrations for Cu(II) and Zn(II) removals at equilibrium. The overall response shapes are similar across all cases, but 2ACEFB was a better adsorbent than EFB at high concentrations of heavy metals. The maximal improvement in removal by adsorption on comparing 2ACEFB to EFB was 40.80 and 37.04% for Cu(II) and Zn(II), respectively. This is due to the fact that high metal concentration provides a higher driving force for the transfer process to overcome the mass transfer resistance [24][25][26].

Effect of Concentration.
3.5. Effect of Solution pH. Solution pH is an essential factor that can affect the adsorption process. In fact, chemical precipitation became a dominant process for metal removal at pH 7 [27]. This condition was not solely considered as adsorption because the metal precipitation may lead to a misinterpretation of adsorption capacity. The effect of pH on metal Cu(II) and Zn(II) adsorption onto EFB and 2ACEFB was shown in Figure 5. The result revealed that the maximal adsorption of Cu(II) and Zn(II) onto the adsorbents was at pH 5.0. This result had a similar trend with adsorption study on several metal ions reported by numerous studies [26,28,29]. Therefore, the optimal condition used to determine the adsorption capacity of EFB and 2ACEFB was as follows: 1 g of adsorbent dose, pH 5, contact time of 360 min, and room temperature.
3.6. Adsorption Isotherm. Adsorption data were evaluated by both two-parameter (Langmuir, Freundlich, and Temkin) and three-parameter (Redlich-Peterson, Sips model, and Toth model) isotherm models. Each model explains a specific feature of the adsorption process, and the model expressions and linearized forms are presented in Table 4. To examine the adsorption data, the mean absolute percentage error (MAPE) was used to quantify the experimental results compared with these isotherm models, which can be expressed as follows: where q e,exp and q e,pre are the experimental and predicted uptake capacity (mg/g) and N is the number of data points. The adsorption study results are illustrated in Figure 6, and associated parameters are tabulated in Table 5. Linear regression analysis was used to evaluate two-parameter isotherm models, providing the parameters used for nonlinear fitting. The experimental data fitted well with the Langmuir isotherm better than others, whose correlation coefficients were 0.996-0.997, suggesting that the mechanism was monolayer adsorption. The assumption of the Langmuir model involves an account of the formation of a monolayer of adsorbate molecules on the adsorbent surface. This is consistent with the assumption that the adsorption occurs at specific sites on the adsorbent surface [30].
The selected three-parameter models were fitted with the same experimental data (see Table 6). These threeparameter models provided good fits which MAPE ranged 2.41-7.97% was better than for the two-parameter isotherms (MAPE ranged 3.80-75.52%) for all experiments. It is also noteworthy that the parameters g, m S , and m T in the Redlich-Peterson, Sips, and Toth models, respec-tively, were approximately close to unity. This study result suggested that the adsorption of Cu(II) and Zn(II) onto 2ACEFB and EFB can be properly described by the Langmuir model with a homogenous surface.

Maximal Adsorption Capacity.
For comparison propose of available model types, the three-parameter models obviously fit with the experimental data better than the twoparameter type in the range of initial concentration used and also supported the two-parameter Langmuir isotherm appropriately. It has been clearly shown that the adsorptive potential of Cu(II) was stronger than that of Zn(II). The maximal adsorption capacity of EFB onto Cu(II) and Zn(II) was observed as 20.01-20.28 and 17.80-18.06 mg/g, respectively. When EFB was activated by H 3 PO 4 , the adsorption capacity improved 2.04-and 1.886-fold to 40.67-41.66 and 33.58-36.00 mg/g over EFB ( Table 5). The Cu(II) and Zn(II) uptake capacities obtained from this study are compared to relevant prior studies in the literature presented in Table 6. Therefore, it is concluded that the two-step acid treatment by H 3 PO 4 improved both the physical and chemical properties, which can become more wealth capacity in Cu(II) and Zn(II) removal over untreated biochar from EFB.

Adsorption
Mechanism. The surface functional groups of an adsorbent play a vital role in establishing a connection between the adsorbent and the adsorbate [31,32]. Upon conducting FTIR analysis, it was observed that the process of phosphorylation resulted in an increase in porosity, carbon surface area, and the presence of carboxyl (-COOH) groups. Moreover, the introduction of the phosphoric acid group in 2ACEFB expanded the pore structure during the adsorption of Cu(II) and Zn(II), thus enhancing the  [34,35]. Figure 7 illustrated the scheme of Cu(II) and Zn(II) adsorption mechanism using phosphorylated sorbents in an aqueous solution, which can be summarized as follows.
3.9. Adsorbent Reusability. Adsorbent reusability is considered to be a significant advantage in determining the adsorbent's feasibility for practical application [51]. The reusability of the adsorbent in this study was accessed using a series of adsorption reactions under the similar conditions mentioned above (25°C temperature, 6 hrs, 1 g of adsorbent, pH 5, concentration 200 mg/L) by five consecutive reactions. After each reaction was completed, the spent adsorbents were collected from an aqueous solution by filtration and were dried at 105°C. It is revealed that the removal of Cu(II) and Zn(II) by 2ACEFB was obviously decreased to 79.26 and 72.41%, as shown in Figure 8, respectively, after five consecutive reactions, which indicated inferior reusability of the prepared adsorbents. This result was repeatedly confirmed by the FTIR, as shown in Figure 9. The phosphate groups (band P-OH 919 cm -1 and amino phosphonic acid functional group 1191 cm -1 ) have not been significantly changed in band shedding (1st-5th reuse). There was a slight decrease    9 Adsorption Science & Technology in the removal of Cu(II) and Zn(II) which may be due to the inferior reusability of the prepared adsorbent [52].

Conclusions
In this study, activated carbon was prepared via 2-step H 3 PO 4 activation under microwave-assisted pyrolysis in order to examine its feasibility for the adsorption of Cu(II) and Zn(II) in an aqueous solution. A series of batch experiments was successfully done, and the conclusions can be drawn.
(1) The 2-step H 3 PO 4 -activated carbon (namely, 2ACEFB) derived by oil palm empty fruit bunch (EFB) was successfully performed in this study, totally proved by FTIR, BET, and proximate analysis. The specific surface area and pore volume of the adsorbent increase due to the H 3 PO 4 activation (2) Adsorption equilibrium reached 84% in approximately 60 min and developed to equilibrium time in 240 min for all cases. The pseudo-second-order model was the best description, with the coefficient of determination (R 2 ) ranged 0.998-0.999 (3) The 2ACEFB was clearly a better adsorbent than EFB, especially at high initial concentration, and improved the removal of Cu(II) and Zn(II) over EFB by 40.80 and 37.04% at pH 7, respectively. The maximal adsorption capacity of 2ACEFB was approximately 2.04-and 1.89-fold related to EFB, respectively, as described adequately by the Langmuir model with monolayer and homogenous surface adsorption (4) The 2ACEFB showed good ability in adsorption reusability by five consecutive reactions, with the removal efficiency reduced to 79.26 and 72.41% for Cu(II) and Zn(II), respectively. Therefore, it is implied that the 2ACEFB acts as a promising adsorbent for metal ion adsorption

Data Availability
The data that support the findings of this study are available from the corresponding author, C.S., upon reasonable request.