1 Introduction

Cocoa (Theobroma cacao) is an economically important crop because it has been used mainly in chocolate manufacturing. During the cocoa bean-producing process, the ripe pods are harvested from the cocoa trees and then broke open for removing wet bean trees. After fermentation of surrounding pulp for draining cocoa pulp juice, the beans are further dried and bagged, constituting the commodity cocoa bean. Cocoa pod husk (CPH), accounting for over 50% of the weight of the cacao fruit [1], is originated from the leftover pod material of the matured cocoa fruit. It was often discarded as waste or residue during the processing of raw cocoa beans on the farm, causing environmental health problems such as greenhouse gas emissions, foul odor, and disease inoculum source due to its lignocellulosic constituents and biological decomposition [2]. Therefore, the cocoa farmers were encouraged to utilize CPH for animal feed and as a source of alkali for soap making aside from its use as a potash (potassium-rich) fertilizer on the cocoa plantation or other crop fields [2,3,4,5,6,7,8,9,10].

Because about 10 tons of wet CPH are generated from one ton of dry cocoa beans produced, it represents million tons of CPH being disposed of without available utilization every year. More significantly, CPH contains some valuable components such as carbohydrate, protein, fiber, lipid, soluble phenolics, tannin, theobromine (one of the alkaloids), potassium-rich ash, and pectin [2, 11,12,13]. In recent years, it was further considered as a potential food source of natural components such as dietary fibers with significant antioxidant activity [11], pectins used as gelling/thickening/stabilizing agents [1, 2, 14, 15], and enzyme lipase for biodiesel production [16, 17]. On the other hand, CPH is comprised of lignocellulosic constituents (i.e., cellulose, hemicellulose, and lignin), showing that it can be served as reinforcement materials or fillers for polymer-based matrices like bioplastics and biocomposites [18,19,20,21,22]. Furthermore, it is rich in carbon source, suggesting that it can be used as an energy source by combustion [23], converted into useful chemicals by pyrolysis [24,25,26], or directly utilized as an adsorbent for removal of organic compounds and metal ions from aqueous solution [27,28,29,30,31,32].

Activated carbon (AC) is the most common adsorbent due to its excellent pore properties like high surface area and wide pore size distribution. Regarding its production process by chemical activation method, potassium hydroxide (KOH) is an environment-friendly reagent for the preparation of high-surface-area activated carbon [33]. By contrast, the commonly used chemicals zinc chloride (ZnCl2) and phosphoric acid (H3PO4) may generate environmental pollution problems as a result of the releases of chemical activation compounds. As studied by Leite et al. [34], however, it was demonstrated that the contents of Zn leached out with HCl are practically removed from the carbon material. In addition, the KOH activation provides a cleaner process to produce highly microporous carbon material because the solution after the acid washing of crude AC can be reused as a liquid fertilizer.

Since the early 1970s, the uses of cheap agricultural byproducts and environmentally chemical agents have been a focus of research efforts. Due to its high polysaccharides, cocoa-derived residues (i.e., CPH and cocoa bean husk) have recently been utilized as a low-cost precursor for producing AC and its application for the removal of organic compounds from the aqueous solution [35,36,37,38,39,40,41,42,43,44,45]. As reported by Ahmad et al. [35,36,37], CPH-based AC was produced by carbonization at 800 °C followed by physical activation at 850 °C in a CO2 flow, showing that the largest surface area of the resulting AC was recorded at about 560 m2/g. In the similar study by Bello et al. [38], CPH-based AC was produced by carbonization at 500 °C with a heating rate of 10 °C/min followed by CO2 activation at 700 °C for 2 h, indicating that the surface area of the resulting AC was obtained at 503 m2/g. In the study by Pereira et al. [39], the resulting ACs were produced by using H3PO4 and ZnCl2 at 500 °C with a heating rate of 5 °C/min, showing their surface areas of 1077 and 642 m2/g, respectively. In the preliminary investigation by Ribas et al. [40], inorganic components, consisting of red mud, lime, KOH, Al(NO3)3, and Na2SO4, were used as chemical-activating agents for producing ACs at 20 °C/min up to 800 °C for 30 min. As studied by Saucier et al. [41], they further used the microwave-induced chemical activation process with similar inorganics for producing ACs, which indicated the BET surface areas ranging from 541 to 619 m2/g. In the previous study [42], the authors used the acid-leached CPH as a precursor for preparing AC by physical activation at a temperature range of 650–850 °C, showing an increasing BET surface areas from 1 to 638 m2/g as the temperature increased.

As reviewed above, many studies have reported the utilization of CPH as a precursor for producing AC [35,36,37,38,39,40,41,42,43,44,45]. However, these studies focused on traditional activation processes for producing ACs, which only possess a low-medium surface area (i.e., less than 700 m2/g). Previously, the authors thus used potassium salts (i.e., KOH and K2CO3) for producing high-surface-area ACs from corn cob [46], convincing that activation temperature is the most important process parameter. Using a small amount of potassium hydroxide (KOH) as chemical activation agent, this study aims to produce a low-cost and high-surface-area AC from the CPH waste, which is rich in potassium salts. Therefore, a series of KOH activation tests were carried out in the production of ACs from CPH. Furthermore, the resulting ACs were performed to observe their variations on the pore properties by means of nitrogen adsorption-desorption isotherms, helium-displacement true densities, elemental analyses, and scanning electron microscopy (SEM). Finally, we evaluated the adsorption kinetic performance for the removal of methylene blue (MB, cationic dye) from aqueous solution using the CPH-based AC in comparison with commercial AC.

2 Materials and methods

2.1 Materials

The CPH used as a precursor for AC preparation was obtained from a local cocoa plantation farm at Neipu Township (Pingtung, Taiwan). Firstly, the collected CPH was crushed to remove attached residues and subsequently sieved to fractions with particle size 0.63 mm on average. Prior to the impregnation pretreatment with KOH, the dried CPH has been preserved in a desiccator. The proximate properties (i.e., volatile, ashes, and fixed carbon) of the dried CPH have been determined by using the American Society for Testing and Materials (ASTM) Standard Test Methods (i.e., D-3172). Table 1 listed the thermochemical properties of the feedstock, showing that the as-received CPH consists largely of volatile matter (i.e., 61.17 wt%), but the content of ash in the sample reaches as high as 8.14 wt%. These results are similar to those by the previous study [42] and other studies [11, 47]. The cationic dye used as the target adsorbate in the present study is methylene blue (MB), which was purchased from Merck Co. (Germany). A commercial AC (FILTRASORB 300, named as Calgon-300 in this work) from Calgon Carbon Co. (USA) was used as an adsorbent for removal of MB at the same adsorption conditions. Its main pore properties are given below: BET surface area 660 m2/g, total pore volume 0.38 cm3/g, and average pore diameter (by the analytical data on BET surface area and total pore volume) 2.32 nm.

Table 1 Proximate analysis and ultimate analysis of cocoa pod husk (CPH)
Full size table

2.2 Chemical activation experiments

In the previous study [48], the temperature of 500 °C was found to be optimal for preparing high-surface-area AC from corn cob by ZnCl2 activation. Prior to KOH activation experiments at 500 °C in this work, 10 g of the dried CPH was well mixed with 150 mL deionized water and 2.5–10.0 g KOH (Merck Co.; Assay ≥ 85.0%) in 250 mL Erlenmeyer flasks using sealing film. Impregnation was carried out on a thermostat agitator at about 30 °C for 1 h. Afterward, the impregnated sample solution was placed into a heated air circulation oven to leave KOH-impregnated CPH by removal of water. Then, the influences of the impregnation ratios (i.e., 0.25, 0.50, 0.75, and 1.00 g KOH/g CPH) at a fixed holding time (i.e., 30 min) and holding times (i.e., 15, 30, 45, and 60 min) at a fixed impregnation ratio (i.e., 0.50 g KOH/g CPH) on the pore properties of the resulting ACs were preliminarily studied. The data on the BET surface areas of the resulting ACs, ranging from 458.4 to 534.4 m2/g, varied without being consistent with the impregnation ratio and holding time. Those instances will suggest that these activation parameters are not significant factors that affect the pore properties of the AC products.

In the subsequent KOH activation experiments, the activation temperature was only adopted in the production of AC because it has the greatest influence on chemical activation. The AC products were prepared from the pyrolysis of the KOH-impregnated CPH sample (4 g, in each set of experiments) as reported previously [48]. For all tests, the CPH sample was heated to 400–900°C in the presence of nitrogen gas (500 cm3/min) at about 10 °C/min, after which the temperature was held for 30 min before cooling to room temperature. The crude solid products after KOH activation were then rinsed with 3 N HCl and deionized water 3–5 times to remove potassium ions and other residues remained. Finally, the resulting products were dried overnight in an oven at about 100 °C. The resulting ACs prepared at 400, 500, 600, 700, and 800 °C were thus referred to as CPHAC-400, CPHAC-500, CPHAC-600, CPHAC-700, and CPHAC-800, respectively. In order to obtain AC with a higher yield, another resulting AC (CPHAC) was prepared at 800 °C for 0.5 h under the impregnation ratio of 0.75 g KOH/g CPH. This AC product was used to test its adsorption performance for removal of cationic dye (i.e., MB) from aqueous solution in comparison with commercial AC.

2.3 Characterization of CPH-based activated carbons

Based on the measurement using a surface area and porosity analyzer (model no.: ASAP 2020; Micromeritics Co., USA), the pore properties of CPH-based ACs can be determined. The specific surface area of the resulting AC was based on the Brunauer-Emmett-Teller (BET) standard method, giving a straight line in the relative pressure (P/P0) range of 0.05–0.30. The total pore volume was obtained from the liquid nitrogen amount adsorbed at P/P0 of about 0.95 [49]. Furthermore, the mean pore size can be estimated by using the measured BET and pore volume [50]. Using the Barrett-Joyner-Halenda (BJH) method at desorption branch, their pore size distributions were further determined in the range of 2.0–25.0 nm.

The true density (also called absolute density or helium-displacement density) of the resulting AC product was obtained by a gas pycnometer (model no.: AccuPyc 1340; Micromeritics Co., USA), which is based on the principle of helium displacement. The instrument features high-resolution pressure transducers and a thermoelectric device to deliver temperature stability. The data on true density will be further used to calculate its particle density and porosity [50].

Due to the pore properties of CPH-based ACs associated with their contents of carbon and other organic elements, the elemental analyses of these resulting ACs (about 3 mg for each analysis) were conducted using an elemental analyzer (model: vario EL III; Elementar Co., Germany). In terms of its main elements of carbon (C), hydrogen (H), nitrogen (N), and sulfur (S), the oxygen content was estimated by the difference (i.e., O = 100%−C−H−N−S).

The surface textures of CPH and CPH-based ACs were carried out using a scanning electron microscopy (model no.: S-3000N; Hitachi Co., Japan). These observations were operated at an accelerating potential of 15 kV. Prior to the image work, a thin film of gold (Au) was coated on the sample to form an electrically conductive surface using an ion sputter (model no.: E1010; Hitachi Co., Japan).

2.4 Adsorption performance test of CPH-based activated carbon

In the present study, the adsorption performance of the CPH-based AC (i.e., CPHAC) was carried out in a 3-L stirred batch adsorption apparatus with four baffles in comparison with the commercial AC (Calgon-300). The adsorption kinetic behaviors of methylene blue (MB) from aqueous solution were obtained under the initial MB concentration of 10 mg/L (or 3.13 × 10−2 mol/L), adsorbent dosage of 0.2 g/2 L, agitation speed of 200 rpm, and temperature of 25 °C. The solution sample (about 10 cm3) was taken at a specified time up to 1 h and then filtrated with fiber membrane (cat. no.: A045A025A; ADVANTEC MFS, Inc., Japan). The MB concentration analysis of the filtrate solution was immediately measured with UV/Visible spectrophotometer (model: U-2001; Hitachi Co., Japan) at 661-nm wavelengths. The amount of MB adsorbed (qt, mg/g) was thus determined by the solution volume (i.e., 2 L), adsorbent mass, and the difference of initial MB concentration (C0) and MB concentration at adsorption time (Ct).

3 Results and discussion

3.1 Elemental analyses of CPH-based activated carbons

Table 2 provided the data on the elemental analysis of the resulting ACs produced from CPH. It is obvious that the CPH-based AC products contained high contents of carbon (C) and oxygen (O), which should be attributed to the lignocellulosic composition and ash content of CPH. In comparison, the elemental content of a typical AC was found to be 88 wt% C [51]. As described in Table 1, the content of C in the CPH precursor was around 41.6 wt%, which was smaller than those of CPH-based ACs (i.e., 58.57–78.86 wt%). It should be noted that the contents of oxygen (O) in the resulting ACs indicated an increasing trend, which should be attributed to the fact that more potassium carbonates and oxides were likely remained at higher activation temperature. Furthermore, more oxygen-containing functional groups could be present on the surface of the resulting AC, thus indicating a negatively charged surface. On the other hand, there is no correlation between hydrogen content (i.e., 1.05–2.18 wt%) and activation temperature.

Table 2 Ultimate analyses of CPH-based ACs at different activation temperatures (400–800 °C) held for 30 min under the fixed impregnation ratio of 0.5 with KOH
Full size table

3.2 True densities of CPH-based activated carbons

The true density (ρs) is related to its hardness or compressive strength because the increased order of carbon molecules indicated higher resistance strength than its starting material. The increase in rigid and porous structures was caused by the significant loss of volatiles and condensable compounds from the original feedstock. Consequently, the data on true density of CPH-based AC indicated an increasing trend with activation temperature, ranging from 1.430 g/cm3 for CPHAC-400 to 3.495 g/cm3 for CPHAC-800 as shown in Table 3. In this respect, the true density of CPH-based AC can be used as an appropriate indicator for its extents of condensation and shrinkage during the activation process. Furthermore, it seemed that there exists a close connection between the true density and pore property.

Table 3 Pore properties and densities of CPH-based ACs at different activation temperatures (400–800 °C) held for 30 min under the fixed impregnation ratio of 0.5 with KOH
Full size table

3.3 Pore properties of CPH-based activated carbons

Table 3 also indicated the variations of pore properties of CPH-based AC with activation temperature. Obviously, the BET surface area of CPH-based AC significantly increased to 483.0 m2/g (CPHAC-500) after KOH activation was performed at a temperature range from 400 to 500 °C. The pore properties of the resulting ACs continuously increased in line with the activation temperature. The findings were in accordance with those of studying KOH activation for olive stone [52], cherry stone [53], and grape seed [54]. Porosity in carbons activated at 400 °C is very low, whereas temperatures of 700 and 800 °C lead to activated carbons with increasing porosity, thus suggesting a critical temperature range of KOH activation for producing high-surface-area ACs greater than 1000 m2/g. Also, a reproducibility test in the chemical activation experiments was examined at the carbon products of CPHAC-400 and CPHAC-700 in duplicate and in triplicate, respectively. The finding showed that their BET surface areas were very consistent based on the values of 6.9 ± 1.3 m2/g and 1154.6 ± 7.9 m2/g, respectively.

The nitrogen (N2) adsorption-desorption isotherm is a powerful tool for the characterization of porous materials like AC. Figures 1 and 2 showed the N2 adsorption-desorption isotherms (at − 196 °C) and pore size distribution curves of CPH-based AC products, respectively. Obviously, the resulting ACs possessed type I isotherms because they are often encountered in the case of physical adsorption on microporous materials with the pore width less than 2 nm [49]. The micropore filling was observed at relatively low relative pressures (less than 0.05) because of the high adsorption potential [49]. Therefore, the adsorbed values significantly increased at above 700 °C, which should be attributable to the rigorous reaction, creating more fine pores in the resulting carbon products (i.e., CPHAC-700 and CPHAC-800). However, beyond 800 °C, there could be less increase in the development of pores due to pore collapse, not favorable for their pore properties [55]. To assure the assumption in the KOH activation experiments, the CPHAC-850 and CPHAC-900 (not listed in Table 3) were produced at 850 and 900 °C, respectively. It showed that the BET surface area of the former increased slightly to about 1850 m2/g, but the latter decreased significantly due to serious burn-off. Li et al. [56] also found similar results in the preparation of ACs from poplar wood by KOH activation at a higher temperature (˃ 750 °C).

Fig. 1
figure 1

N2 adsorption-desorption isotherms of cocoa pod husk–derived activated carbon (CPHAC) products (i.e., CPHAC-400, CPHAC-500, CPHAC-600, CPHAC-700, and CPHAC-800)

Full size image
Fig. 2
figure 2

Pore size distributions of cocoa pod husk–derived activated carbon (CPHAC) products (i.e., CPHAC-400, CPHAC-500, CPHAC-600, CPHAC-700, and CPHAC-800)

Full size image

3.4 SEM observations of CPH-based activated carbons

Figure 3 showed the SEM micrograph of CPH (× 1000) for the observation of the surface texture of the biomass sample. As observed in the sample of dried CPH, it exhibited a heterogeneous surface as it is expected of fibrous agricultural residue. In contrast, the CPHAC-700 and CPHAC-800 products in Fig. 4 displayed porous or pitted surface textures at the scale of a macropore with > 50 nm, indicating that rigorous activation reaction with KOH took place at these carbon materials. In brief, the production conditions at the higher activation temperature will develop more fine pores in the carbon products, thus resulting in the increasing trends of pore properties and true densities as listed in Table 3.

Fig. 3
figure 3

SEM image of cocoa pod husk (CPH, × 1000)

Full size image
Fig. 4
figure 4

SEM images of CPHAC product (a) (i.e., CPHAC-700, × 10,000) and optimal CPHAC product (b) (i.e., CPHAC-800, × 10,000)

Full size image

3.5 Adsorption performance of CPH-based activated carbon

Figure 5 showed the adsorption kinetic profiles of two adsorbents (i.e., CPHAC and Calgon-300) for removal of MB from aqueous solution at initial MB concentration of 10 mg/L (or 3.13 × 10−2 mol/L), adsorbent dosage of 0.2 g/2 L, agitation speed of 200 rpm, and temperature of 25 °C. Obviously, the fast decrease in residual MB concentration (Ct/C0) can be seen at a short contact time, especially in the adsorbent CPHAC. It implied that the strong interaction should exist between the cationic dye and the negatively charged surface. Therefore, a simple model (i.e., pseudo-second-order equation) was used to fit the adsorption kinetic data in this work [57]. Its linear form was given below:

$$ t/{q}_t=1/\left(k\times {q_e}^2\right)+\left(1/{q}_e\right)\times t $$
(1)

where qt is the amount of MB removed at time t (mg/g), qe is the amount of MB adsorbed at equilibrium (mg/g), k is the pseudo-second-order rate constant (g/(mg min)), and t is the contact time (min). The fitted values of parameters for the adsorption system have been listed in Table 4, indicating that the correlation between the experimental and fitting results was great. As observed in Fig. 5, the adsorbent CPHAC exhibited a better adsorption performance for MB removal as compared with the commercial AC (Calgon-300) in spite of their similar pore properties. This difference in adsorption capacity can be also attributable to the interaction between the cationic adsorbate and the adsorbent on the surface nature [58]. Obviously, the adsorbent CPHAC has a stronger interaction in comparison with the adsorbent Calgon-300.

Fig. 5
figure 5

Plots of dimensionless concentration (Ct/C0) vs. time for the resulting AC (CPHAC) and commercial AC (Calgon-300); Adsorption conditions: initial MB concentration = 10 mg/L (or 3.13 × 10−2 mol/L), adsorbent dosage = 0.2 g/2 L, adsorption temperature = 25 °C, and agitation speed = 200 rpm; symbols: experimental data, full lines: calculated from Eq. (1) and Table 4

Full size image
Table 4 Kinetic parameters for adsorption of methylene blue (MB) onto the resulting AC (CPHAC) and commercial AC (Calgon-300) based on pseudo-second-order modela
Full size table

4 Conclusions

A series of AC products were prepared from dried cocoa pod husk (CPH) using KOH activation process. According to the chemical, pore, and adsorption properties of the CPH-based ACs, the following conclusions below can be drawn:

  • The pore development of the resulting AC was mainly determined by the process temperature. The true densities of the CPH-based ACs indicated an increasing trend with activation temperature, thus suggesting a more rigid and porous carbon material produced at higher activation temperatures.

  • The optimal AC product produced at around 800 °C showed its BET surface area of over 1800 m2/g and the total pore volume of about 1.0 cm3/g.

  • It has been shown that the microporous features (i.e., pore width is less than 2.0 nm) of the resulting ACs were based on their nitrogen adsorption-desorption isotherms.

  • The CPH-based AC can be an effective low-cost adsorbent from aqueous solution in comparison with commercial AC based on the fitting results of pseudo-second-order model for adsorption removal of cationic dye (i.e., methylene blue).