Preparation of activated charcoal adsorbent from pitombeira seeds (Talisia esculenta) and its application for Ca²⁺ ions removal

Raw water or untreated water for human consumption can cause several types of endemic diseases. The health risks related to water regard two types of agents: biological agents, by direct contact or vectors, and chemical agents (Fu & Wang 2011). In Brazil, about 6.93 million households live without access to water through distribution systems, and a common strategy to overcome the lack of a supply network is using artesian wells. However, in most cases, the water that comes from them presents a high degree of salinization (hard water) and does not undergo any type of treatment before consumption (Brasil, Ministério da Saúde 2006). This situation can pose several problems, both for the population's health and for industrial use. Water with a high concentration of dissolved calcium and magnesium mineral ions is classified as hard water. These ions are derived from underground deposits, such as limestone or dolomite, and an excessive amount of these is added to the water composition in the form of bicarbonates, nitrates, chlorides, and sulfates (Piratoba et al. 2017). When water passes through the soil, the carbon dioxide in the water dissolves the limestone, making groundwater harder than surface water (Memon et al. 2011).

Regarding water hardness, the Brazilian Ministry of Health establishes for potable water standards the maximum permissible value of 500 mg/L for the sum of calcium and magnesium concentrations. Generally, when water contains less than 50 mg/L of CaCO₃, it is considered as soft, up to 150 mg/L is classified as semi-hard, with less than 300 mg/L is labeled as hard and when it presents values above 300 mg/L it is classified as very hard (Brasil, Ministério da Saúde 2006; da Silva et al. 2016). As a comparison, European countries, such as Portugal, use the 150–500 mg/L range as a reference value for total hardness. The increased concentration of these metals in artesian wells may pose a serious threat to human health (Kozisek 2020). Calcium, although present in the human body as an essential element, both because it performs a series of functions related to cellular metabolism, as well as bone and tooth maintenance, when exceeded can bring a series of complications and diseases such as renal and bladder lithiasis, which are mainly associated with the intake of calcium and magnesium highly concentrated in water, and hypercalcemia, which is when the calcium level in blood is above normal (Kozisek 2020). Other problems related to an excess of calcium in the blood are renal failure, soft tissue calcification, headaches, and interference with the absorption of minerals such as magnesium, phosphorus, iron, and zinc (Whiting et al. 1997).

The means used to control and/or remove metal ion excess in aquatic systems are quite diverse. For instance, there are chemical precipitation (Liu et al. 2020) and reverse osmosis (Wang et al. 2020), as well as electrochemical methods (Xie et al. 2018). Adsorption is also a very versatile technique and, depending on the adsorbent type, is economically feasible. Among the many types of adsorbents available in the market for metal adsorption in aqueous media, the most important are the activated carbons, which may be of both mineral and vegetable origin (Santoso et al. 2020). Activated carbons obtained from carbonaceous materials are characterized by high porosity and surface area, and are able to adsorb ions or molecules, both in the liquid and gas phase. There are several raw materials used for their preparation, such as wood, coconut husks, seeds, peat, sawdust, residues of fertilizers, and rubbers, among others (Nor et al. 2013). The use of lignocellulosic biomass for activated carbon production is an important approach as a strategy to take advantage of vegetal waste. The conversion of these wastes to value-added products, such as activated carbon, can solve environmental problems such as waste accumulation and air and water pollution (Tan et al. 2017).

The species Talisia esculenta, which is popularly known in Brazil as pitombeira and is abundant in northern states of the Brazilian northeast, mainly between January and April, may be an alternative for the production of activated carbon, since the consumption of its fruit, the pitomba, is restricted to its pulp. Excess, such as bark and stone, is discarded into nature as residue because it does not have a specific use (Guarin Neto et al. 2003). Therefore, the use of pitomba pits for the production of activated carbon can provide an alternative adsorptive material, of low cost and with a similar efficiency to that of biochars already available in the literature for the removal of different contaminants. The raw material type plays a principal role in the properties of the activated carbon, influencing its quality and overall characteristics (Abechi et al. 2013).

Thus, in this work, the adsorptive potential of phosphoric-acid-impregnated activated carbon originated from pitomba pits against calcium ions (Ca²⁺) in an aqueous medium was investigated. The quantification of adsorbed quantities was carried out using the complexometric titration technique, using standard solutions of the metal cation titrated with ethylenediaminetetraacetic acid (EDTA) in an aqueous medium. The experimental results were adjusted to isothermal and kinetic models (Sarabadan et al. 2019).

The pitombas were purchased at the fairs of São Luis city/MA and their cores were separated from the pulp by hand. The chemical activator used was phosphoric acid (H₃PO₄) (SYNTH), the complexing agent was EDTA (SYNTH), Eriochrome black T indicator and calcium ions were generated from aqueous solutions of its carbonate (CaCO₃) (SYNTH). Deionized water was applied to prepare the solution in all experiments.

The pitomba pits were oven-dried and crushed in a cyclone rotor mill (TE-651/2). The studies of thermal behavior and characterization were performed using a thermobalance, model TGA-51 (Shimadzu), in the following experimental conditions: (a) flow rate of 100 mL/min for synthetic air and 50 mL/min for argon; (b) alumina crucibles; (c) heating rate of 10 °C/min; (d) sample mass of 22.00 ± 0.50 mg; (e) temperature range of 25–600 °C. The pyrolysis was carried out in a Microprocessor Based Programmable Muffle Furnace 2000-C.

The moisture and ash content were performed according to the Technical Standards ASTM D 2867-09 (2014) and 2866-94 (2004) respectively. The specific surface area was determined by Quantachrome NovaWin, a surface area and pore size analyzer, using N₂(g) at a temperature of 77.35 K. The identification of functional groups present on the surface of the activated carbon was done by the Boehm method. Infrared absorption spectroscopy was performed using a Bruker Equinox 55 IR spectrometer. Infrared spectra were recorded at 4 cm⁻¹ resolution over a wavenumber region of 400–4,000 cm⁻¹ and 100 scans per sample on KBr pellets. The morphology of the materials was studied by scanning electron microscopy (SEM) analysis with 500-fold Phenom Pro X equipment. The surface pH at the point of zero charges of PAC was determined by isoelectric point titration as a function of pH using a HANNA HI-2315 digital pH meter.

The pitomba pits were dried at 100 °C for 24 hours. The ground powder was put together with the activating agent (H₃PO₄) 100 mg/L, in a 1:1 ratio for 24 hours. To identify the best burning temperature of the material for the formation of the coal, a thermogravimetric analysis of the phosphoric-acid-impregnated pit was carried out. Based on the results of the thermogravimetric analysis, the burning temperature was set at 350 °C. After the burning step, the activated coal of the pitomba pits, called PAC, was obtained. PAC was washed with distilled water until the pH of the supernatant reached the pH of the distilled water used (pH = 5.9). After the washing process, the coal was oven-dried for one hour at 100 °C, crushed in a mortar, and set for granulometry (≤150 μm).

The pH at the point of zero charges (pH_PZC) was determined using the batch equilibrium method. Amounts of 100 mg of the adsorbent and 10 mL of the 100 mg/L NaCl solution were added with pH values adjusted from 1 to 11 with 100 mg/L NaOH and HCl solutions. The mixtures were shaken at 25 °C for 24 hours at 99 rpm, and after this contact time they were filtered and the pH values (pH_final) were recorded.

All adsorption tests were performed in batch, from a standard solution of CaCO₃, 400 mg/L. The adsorbed quantities were determined by complexometric titration with EDTA 400 mg/L in aqueous medium using the Eriochrome black T indicator.

The kinetics study was conducted in a batch system. To each 100 mg of PAC was added 100 mL of Ca²⁺ ion solution at a concentration of 400 mg/L at pH 7. The system was under constant stirring at time intervals of 5–60 min. For each time interval, aliquots of 25 mL were removed and the Ca²⁺ ion content was quantified by complexometric titration.

To verify the pH influence on Ca²⁺ adsorption, the studies were conducted in batches at 25 °C, with a coal mass of 0.1 g added in Erlenmeyers, with the same volume of solution (100 mL) and the same standard solution of CaCO₃, 400 mg/L. The pH influence on the adsorption process was verified for pH ranges 2–3, 7, and 11–12, with 5 min contact and constant stirring. For pH correction, solutions of NaOH and HCl 100 mg/L were used.

The behavior of lignocellulosic materials as a function of temperature can be observed by thermal analysis. The selection of the carbonization temperature for the raw material used in this activated carbon production was based on the thermogravimetric analysis (TGA) curves. Figure 1 shows the thermal behavior of each sample of in natura pitomba core.

The behavior of the observed curve shows a constant mass in thermal degradations of the sample at 550 °C or higher, indicating that at this temperature all organic matter has already been decomposed. First losses refer to the moisture present and some volatile compounds (0–100 °C). The first loss of significant mass at a temperature range between 100 and 300 °C refers to the degradation of hemicellulose and cellulose that occurs at lower temperatures, and the second one corresponds to degradation of lignin and small amounts of cellulose, with lignin being dominant in high conversions (Bridgeman et al. 2008). Lignin is more thermally stable when compared with cellulose and hemicellulose (Branca & Di Blasi 2003). From the results of thermal degradation of the samples against temperature, it is noted that the samples' mass loss at 300 °C or higher is quite similar, so the samples' pyrolysis temperature was set at 350 °C, with 90 minutes heating ramp and plateau, summing three hours of total burning time, which is enough time for the degradation of volatile materials and hemicellulose and cellulose components (dos Santos et al. 2011).

The determination of the coal's moisture content was carried out to verify the adsorbent's affinity for water molecules. A high moisture content means the adsorbent can have its adsorptive capacity compromised by the incorporation of water molecules on its surface (Yakout et al. 2015). On the other hand, the ash content is a parameter used to characterize the coal after the pyrolysis process. Ashes generally consist of iron, aluminum, calcium, magnesium, sodium, and potassium oxides, sulfates or carbonates (Borges et al. 2015). Table 1 shows the percentages obtained for yield, moisture content, and ash of PAC and other activated carbons of the same nature.

The ash content was 2.9%, which is a very satisfactory result, seeing as a low ash content indicates a high amount of organic matter and a low content of inorganic material, favoring the adsorptive potential of the coal. The high content of inorganic matter may lead to changes in the solution pH when coal is present, since mineral matter causes a dielectric effect in the adsorption process, preferentially adsorbing water, due to its hydrophilic character (Momčilović et al. 2011).

The surface properties of the PAC were evaluated according to the original classification proposed by BET. The PAC presented a specific surface area of 141.6 m²/g and a mean radius of 1.3 nm, classifying the coal as microporous according to IUPAC (Bayne et al. 2013). The Boehm titration used to evaluate the functional groups present on the surface of the PAC showed a predominance of acidic groups, which is justified by the use of the activating agent (H₃PO₄). Phenolic groups (1.7 mEq/g), carboxylic (0.6 mEq/g) and lactonic (0.57 mEq/g) occur in greater quantity.

The FT-IR spectrum indicated the presence of C = O bands at 1,544–1,600 cm⁻¹ (carboxylic, lactones), and there are also many overlapping peaks appear that form an absorption band in the 1,300–1,000 cm⁻¹ region, what can be attributed to carboxylic acid CO, phenolic structures and esters (Anirudhan & Sreekumari 2011). There are no peaks in the range of 3,600–3,200 cm⁻¹ indicating the disappearance of water molecules (Nono et al. 2016).

The morphology of the PAC (Figure 2), analyzed from the coal's surface micrographs, shows a porous structure, irregular open pores, and recesses that may be characteristic of the presence of transport pores (Gregg & Sing 1982).

The micrographs show the effects of the activation and pyrolysis steps of the precursor material, since they are meant to infer functional groups on the surface of the adsorbent and to promote the increase of the porosity of the material. PAC presents an irregular porous morphology and the analysis of the surface area indicates that the pore radius is on the order of 1.3 nm, within the range of microporous materials (Dąbrowski 2001).

The importance of this variable in the adsorption process is that the adsorbate and adsorbent charges must be opposing so that there is the greater electrostatic interaction between both, because, if the charges are equal, the adsorption process will be impaired, and electrostatic repulsion will happen (Bautista-Toledo et al. 2005). The adsorbent may present negative charges when the pH of the medium is higher than pH_PZC and positive ones when the pH value is below pH_PZC (Ding et al. 2014). A graph of ΔpH vs pH was plotted and the point of intersection of the two graphs represents the pH_PZC of the PAC (Figure 3).

The pH at the point of zero charge for the PAC matrix was 2.75. In this way, PAC tends to present negative charges in solutions with a pH greater than 2.75. Therefore, higher adsorption of calcium ions is expected to occur at pH values greater than 2.75. Nono et al. (2016) produced phosphoric acid activated carbons of coffee husk and corncob, obtaining pH_PZC values of 3.27 and 3.00, respectively.

Several factors, such as adsorbent type, pore shape, and the surface area beyond the type of metal involved, may influence the contact time required for equilibrium between the phases (AL-Othman et al. 2012). Figure 4 shows the effect of the adsorbed amount as a function of contact time.

The curve behavior indicates a maximum adsorption peak at 10 min. During the adsorption process, a decrease in the amount adsorbed by PAC can be observed, characterizing a desorption process of Ca²⁺ ions. The adsorption peak presented may occur due to the higher availability of vacant active sites on the surface of the adsorbent at the initial stage. At the moment that the sites are being occupied, availability decreases and the adsorption process becomes slower due to repulsive forces between adsorbed molecules and free molecules in the solution. Similar results were observed for other metal adsorptions, such as Pb²⁺ and Cd²⁺ on activated carbon derived from bamboo, obtaining equilibrium time after 30 minutes (Stevens et al. 2014).

The pH of the medium in which the adsorption occurs determines the degree of distribution of the chemical species. The pH at the point of zero charge (pH_PZC) is the pH below which the surface of the adsorbent is positive, being negative above this value (Islam et al. 2015). The adsorption of Ca²⁺ ions in an aqueous medium with different pH values is shown in Figure 5. The importance of this variable in the adsorption is that the adsorbate and adsorbent charges must be opposite to have greater electrostatic interaction between them, because if the charges are equal the adsorption process will be impaired, since there will be electrostatic repulsion (Bautista-Toledo et al. 2005).

The adsorption underwent interference at acidic pH (pH value = 2). At this pH value, the availability of negatively charged groups on the PAC surface is required for the adsorption of the Ca²⁺ cation. The adsorption at pH 2 is lower since the carbon's surface is potentially altered due to the predominance of H⁺ and H₃O⁺ ions. In this system, the H⁺ ions compete with the metal cation, resulting in protonation of the active sites preventing the removal of the Ca²⁺ ion that binds itself to the surface of the adsorbent. This means that, at higher H⁺ concentration, the surface of the PAC becomes more positively charged, thus reducing the attraction between the Ca²⁺ metal cations. In contrast, as the pH increases, a portion of the biochar surface becomes more negatively charged, increasing the availability of adsorption sites and electrostatic interaction between the surface of the PAC and the positively charged Ca²⁺ cations, thus favoring greater adsorption of the cations. Therefore, the metal ion adsorption tends to increase significantly by increasing the pH (Rao & Khan 2009).

The determination of the adsorptive capacity of an adsorbent in a constant temperature analysis represents an important parameter for understanding the efficiency and characteristics of the adsorbent. The isothermal models proposed by Langmuir and Freundlich are often used as parameters to characterize some adsorptive properties of porous materials. Normally, the Langmuir isotherm model suggests monolayer adsorption and a surface containing a finite number of adsorption sites. In contrast, the Freundlich isotherm model assumes that the adsorption process is related to a heterogeneous surface (Zhou et al. 2017). Figure 6 shows the adsorption isotherm obtained in adsorptive assays from solutions with different concentrations of Ca²⁺ ions.

The adsorption equilibrium was obtained graphically from the experimental amount of adsorbed Ca²⁺ ions, q_e (mg/g), as a function of the metal cation concentration, C_e (mg/L), under equilibrium conditions. The adsorption equilibrium was later evaluated using two isothermal models, namely Langmuir and Freundlich. The isotherm obtained experimentally corresponds to an isotherm that is characteristic of microporous solids (Muttakin et al. 2018). Figure 7 shows the adjustments of the experimental data to the Langmuir and Freundlich isotherm models.

The parameters of the Langmuir and Freundlich isotherms obtained from their models' equations applied to the Ca²⁺ adsorption data for the PAC sample at 25 °C are shown in Table 2.

The results show that the adsorption process is better adjusted to the Freundlich model (R = 0.98). Even though the K_F parameter of the Freundlich model is related to the adsorption capacity, it does not provide the maximum removal capacity, since the Freundlich model does not predict adsorbent saturation (Khan et al. 2016). The value of the Freundlich constant obtained was 19.05, and the value of n (6.6) indicates that the adsorption is favorable. Similar results were obtained for the removal of Pb²⁺, Hg²⁺ and Cu²⁺ metal ions adsorbed by activated carbon from coconut residues, presenting K_F values of 23.81 for Pb²⁺, 14.96 for Hg²⁺ and 11.32 for Cu²⁺. The best fit of the experimental data to the Freundlich isotherm model may be due to the heterogeneous distribution of active sites on the adsorbent surface (Anirudhan & Sreekumari 2011). Thus, the activated carbon obtained from pitomba seed proved to be a potential adsorbent for the adsorption of Ca²⁺ ions in aqueous solution, removing about 92% of the ions in solution. Kurniasari et al. (2020) used activated charcoal obtained from sugarcane bagasse (Saccharum officinarum L.) to remove Ca²⁺ ions from well water, with a removal efficiency of 14.4%.

The adsorption kinetics can be defined as the rate of removal of the adsorbate, present in the fluid phase, over time. It therefore involves the mass transfer of one or more dissolved components to the adsorbent particle, migrating through the macropores to the innermost regions of the absorbent particles (Karthikeyan et al. 2005). The adsorption kinetics of Ca²⁺ ions in PAC was investigated by applying three kinetic models to the experimental data of the present study. These models included the pseudo-first-order and pseudo-second-order kinetic models. The adsorption mechanism was also investigated by applying the intraparticle diffusion model. Figure 8 presents the adequacy of the experimental data in the kinetic models for PAC.

The kinetic study data for the pseudo-first-order model were further away from linearity when compared with the pseudo-second-order model and intraparticle diffusion. Table 3 presents the values of speed constants, correlation coefficients for the pseudo-first-order kinetic and pseudo-second-order kinetic models, and the intraparticle diffusion model.

The adjustment of the experimental data is in agreement with the kinetic model of pseudo-second order (0.996) and is also favorable for the intraparticle diffusion model (0.998), but the intercept does not cross the origin, implying that intraparticle diffusion is not the main diffusion mechanism (Dural et al. 2011). The first step may be related to surface adsorption where the adsorption rate is maximum at five minutes and the adsorption occurs on the outer surface of the adsorbent, whereas the second stage must correspond to adsorption on the internal surfaces of the pores of the coal. After 30 minutes we have the equilibrium process, and it is then suggested that the calcium ion adsorption involved more than one process (Juang et al. 2002). Da Silva et al. (2016) studied the adsorption of Ca²⁺ ions in lignocellulosic material, and the results of the kinetic study showed that the pseudo-second-order model was a better fit to the experimental data.

Pitomba pits have proven to be a good precursor material for activated carbon production, presenting a low content of ashes and moisture, significant adsorption potential for adsorption of Ca²⁺ ions in aqueous solution, with a removal of about 92% of the ions in solution with suitable pH. The activating agent was able to produce active sites on the surface of the coal, which was evident by the results of the Boehm tests, which showed the predominance of acidic groups on the surface of the PAC, with emphasis on phenolic groups. The pH_PZC showed that coal has favorable adsorption for Ca²⁺ ions at pH above 3.

In the adsorption studies, it was observed that equilibrium occurs after about 5 min of the reaction and the most effective pH for the adsorption of calcium ions in the solution is the basic one. The adsorption isotherms were more adequate for the Freundlich model, with a correlation coefficient of 0.98 and maximum adsorption of 500 mg·g⁻¹. The analysis of kinetic models showed that the adsorption kinetics of the activated carbon of the pitomba core was more adequate for the kinetic model of pseudo-second order, with the predominance of chemisorption.

The results showed that the carbon produced can be a potential adsorbent for cations in aqueous solution. The results presented in this work were favorable for the use of activated carbon produced from pitomba pits as an adsorbent in the removal of Ca²⁺ ions in an aqueous medium, which can significantly contribute to the reduction of environmental impacts.

The authors are grateful to the Chemistry Graduate Program of the Federal Institute of Education, Science and Technology of Maranhão – IFMA, and to FAPEMA, for financial assistance, and to IFMA/Campus Codó.

All relevant data are included in the paper or its Supplementary Information.