Synthesis of Patchouli Biomass Based (α-Fe₂O₃, γ-Fe₂O_3, Fe₃O₄)/Activated Biochar Composite for Water Environment Remediation

1. Introduction

Biochar is biomass based charcoal [1], categorized as porous black carbon [2] and produced by organic matter [3] or biomass pyrolysis under limited oxygen condition [4]. Patchouli biomass was used in this research as precursor due to one of Indonesian natural commodities. Its field is about 9600 hectare [5]. The other reason is that chemically it contains 39.41% of cellulose, 12.31% of hemicellulose, and 12.52 % of lignin [6]. Hemicellulose undergoes decomposition at 200-260 ◦C, cellulose at 240–350 ◦C, and lignin at 280–500 ◦C [7]. Pyrolysis is a process of thermochemical decomposition [8]. During the pyrolysis process, those three lignocellulosic components undergo various reactions including depolymerization, fragmentation and cross-linking. Products of the pyrolysis are char (solid), bio-oil, (liquid) and gaseous such as carbon dioxide, carbon monoxide, hydrogen and syngas of C₁-C₂ hydrocarbons [9,10]. Biochar contains about 65 to 90 per cent carbon. Increasing of pyrolysis temperature decreases the biochar yield but improves the carbon content [11].

Characteristics of biochar, including porosity and surface functional groups make it has good potency in many applications, such as for soil amendment by improvement of soil organic mater [12], reduction of the greenhouse gas emissions [13] such as nitrous oxide and methane from the soil [14], minimization of the heavy metals in soils [15] and immobilization of fertilizer’s anions [16]. Biochar enhances nutrient in the soil by increasing nitrogen mineralization or nitrification as consequence of biochars high ion exchange capacity and the improved microbial growth and activity [17]. The microbial activity in the soil was signed by absorption of oxygen gas and emission of carbon dioxide. More microorganism in the soil more microbial respiration [18]. The biochar reduces significantly soil bulk density and increase its porosity which improves water holding capacity of the soil [19]. In wastewater treatment, biochar handled various pollutants such as heavy metal [20,21] and organic [22,23,24]. The Biochar also has applications as animal food additives and electronic devices [7].

Adsorption is a favorable and feasible method for treatment of waste water due to its low cost, high efficiency, and ability to remove pollutant at low concentrations [25]. Modification of the biochar surface by nanomaterials such as AlOOH, ZnO/ZnS, Fe₃O₄, MgO, MnO_2, Ni_0.5 Zn_0.5 Fe₂O₄ can improve its physicochemical properties such as functional group, porosity, and active sites on the biochar surface which enhances the heavy metal adsorption [26]. For example, manganese-oxide/biochar (Mn/BC) composite raised the removal efficiency of lead (II) from 6.4 to 98.9 % at pH 5.00. This improvement is attributed to the increase of surface hydroxyls and the decrease of pH_ZPC (pH of Zero Point Charge) in carbon. The maximum monolayer adsorption of lead(II) on Mn/BC at 25 ^oC was five times that on BC [27]. Using Freundlich model and Langmuir model, modification of biochar by MgO improved adsorption of methylene blue from 132,65 to 156,35 mg/g and from 130.25 to 148.25 mg/g, respectively [28]. Magnetite impregnated biochar removed Cu(II) and Zn(II) higher than individual biochar or magnetite [29]. Fe₃O₄/biochar from different biomass removed methylene blue, Cr(VI), Tl(I), and U(IV) with removal eficiency more than 90%. The other adsorbates such as PO₄^3-, As(V), PFOA were eliminated about 100% [30]. NiFe₂O₄/biochar (NFO/BC) composites can adsorp tetracycline higher than biochar and NFO in percentage ratio of 94:56:49. These different adorptions are more related to chemical (functional groups and charge sites) than physical properties such as surface area and pore volume because both increased by sequence of biochar > NFO/biochar > NFO [31]. Composite of Ni-Fe LDH/biochar from pine wood increased adsorption capacity of arsenic anion from 0.2 mg/g to 4.4 mg/g due to improvement of hydroxyl for complexation on the surface [32].

Principally, composite of carbon-metal oxide is prepared by impregnation of carbon by addition of salt to produce carbon–salt composite and alcination of carbon-salt composite to get carbon–metal oxide composite [33]. Therefore, calcination temperature becomes an important parameter in the formation of impregnate structure which may give effect on the adsorption performance of the composite. The different temperature of calcination of carbon–Cu(NO₃)₂ in nitrogen gas streaming, i.e 400, 530, and 800 ^oC, have produced carbon–CuO, carbon–Cu₂O, and carbon–Cu, respectively. Similarly, calcination of carbon–Ni(NO₃)₂ in the nitrogen gas streaming at different temperatures, 550 and 800^oC, produced carbon–NiO and carbon–Ni, respectively. Adsorption test indicated that Cu(I) and Ni(II) impregnated carbons improved thiophenate adsorption of 40–53%. Carbon was prepared by pyrolisis of palm seed [24]. Adsorption of phosphorus by Mg/biochar improved by increasing of calcinations temperature from 300 to 600^oC [34] and adsorption of As(IV) on iron oxide impregnated rice husk biochar at temperature of 950^oC is higher than at 550^oC [23].

In this research FeCl₃–biochar composite was calcined at various calcination temperatures. Fe₂O₃ was predicted as product of FeCl₃ calcination with presence of oxy gasses emitted by biochar activation. The Fe₂O₃ has 3+ cation charge which has good affinity to Lewis base organic contaminant. On the other hand, as hematite structure (α – Fe₂O₃) this material is highly resistant to corrosion and can be used as catalyst, therefore it is greatly important in technological and industrial applications [35].

This research is aimed to study influence of the temperature on the change of structural properties of impregnate and adsorption performance of those composites, especially for adsorption of drug pollutant, such as paracetamol (acetaminophen). It is an analgesic drug type which can cause acute liver damage [36]. Paracetamol is not biodegradable pharmaceutical product, hence it will decompose difficultly. It will lead an health and environmental problem if it accumulates in wastewater and groundwater supply and eventually is consumed as the drinking water [37]. Removal of paracetamol from water environment was studied using adsorption method but using activated carbon [37,38,39]. Result of this research will be very useful to give image of biochar-Fe₂O₃ composite performance as adsorbent to handle the drug wastewater, especially containing paracetamol or drug structure which has physicochemistry (size, structure, shape, and polarity) similar to paracetamol.

Biochar which was prepared using the chemical activators was termed as biochar, such as using the activator of KOH [40,41], Na₂P₂O₇ [42], MgCl₂.6H₂O [43], NaOH [44], as activated biochar, for example by using activator of KOH [45], or as activated carbon such as using KOH, NaOH, H₃PO₄, ZnCl₂, FeCl₃ [46]. In this paper, the term of activated biochar is used because the patchouli biomas was pyrolyzed using including CoCl ₂ activator in pyrolysis process and FeCl₃ in modification process by calcination..

2. Materials and Methods

2.1. Materials of Research

This research used the same patchouli biomass (mixture of root and stem) as our previous researchs [47,48,49,50,51,52,53]]. CoCl₂.6H₂O (Merck, Darmstad, Germany), HCl 37% (Merck, Darmstad, Germany), FeCl₃ (Merck, Darmstad, Germany) and the distilled water were used as the chemicals and solvent, respectively.

An oven (Memmert, Büchenbach, Germany) and a conventional furnace (Carbolite, Derbyshire, UK) were used for drying and calcinating process, respectively.

X-ray diffractometer (PANalytical X’Pert PRO, Malvern, Germany), FTIR spectrometer (Shimadzu, Kyoto Japan), surface area analyzer (Quantachrome NovaWin2), and UV-Vis spectrophotometer (Shimadzu, Kyoto Japan.) were applied for characterization of the products and paracetamol analysis, respectively.

2.2. Preparation of Activated Biochar and Composite

Activated biochar preparation was performed in the same procedure as conducted by previous publication [40]. Patchouli biomass was washed using water and dried under sunrise. Then, the biomass size was conditioned at the range 60-100 mesh. The precursor (10 g) was mixed by CoCl₂.6H₂O salt (52 g), and distillated water (60 mL), then the mixture was evaporated at 100 ^oC for 4 h under stirring. The CoCl₂ impregnated biomass was pyrolized at 450 ^oC for 2 h in the tube furnace under nitrogen gas stream. The activator was removed from product by washing the product using 1 M HCl solution and distillated water, then dried at 130^oC (6 h) using oven and sieved for conditioning the size of 100-120 mesh. The activated biochar product was sieved to take the solid particle size of 100-120 mesh.

The FeCl₃/AB and Fe_xO_y/AB composites (AB = Activated Biochar) were prepared by using FeCl₃ salt as impregnant reactant [54]. The FeCl₃ solution (0.1 M, 100 mL) was added to the activated biochar (6 g) and shaked for 24 h at 200 rpm. The mixture was filtered and the composite was dried at 100 ^oC for 1 h, then calcined at 400, 600, and 800 ^oC for 1 h in the closed ceramic crucible. The uncalcined and calcined composites were characterized.

For XRD characterizations, the codes of the composite samples as written as sample identification in reports of analysis from LSUM as follows:

• Activated biochar by pyrolysis using CoCl₂ activator: Biiochar Aktivator CoCl₂ 100-120
• Composite by calcination of activated biochar-FeCl₃ at 400^oC: BIFe4
• Composite by calcination of activated biochar-FeCl₃ at 600^oC: BIFe6
• Composite by calcination of activated biochar-FeCl₃ at 800^oC: BIFe8

For FTIR spectrometry, the codes of the composite samples as written in report of analysis from Analysis Laboratory in Department of Chemistry, Brawijaya University as follows:

• Activated biochar by pyrolysis using CoCl₂ activator: Biochar Co 52_39
• Activated biochar – FeCl₃ before calcination: B.IFe
• Composite by calcination of activated biochar-FeCl₃ at 400^oC: B.IFe 400
• Composite by calcination of activated biochar-FeCl₃ at 600^oC: B.IFe 600
• Composite by calcination of activated biochar-FeCl₃ at 800^oC: B.IFe 800

By consideration of those different codes, both X-ray diffractograms and FTIR spectra in this paper were reported without the codes.

2.3. Characterization of Activated Biochar and Composites

Surface functional groups of biochars were determined by FTIR spectrophotometer (Shimadzu). Pellet kalium bromide (KBr) technique was applied by mixing the dried biochars and the oven-dried KBr (Merck). The measurements of those FTIR spectra were in Instrument Laboratory of Chemistry Department Brawijaya University.

Crystal structure of the biochar and composites were characterized using powder X-ray diffractometer in Lab Sentral Mineral dan Material Maju or Lab Sentral Universitas Negeri Malang (LSUM). All X-ray diffractograms were collected in 2θ range of 10–90° at a scan rate of 2° min by using Cu metal target (K_α = 1.54 Ǻ). Measurement conditions of analysis were 30 mA, 40 kV, and receiving slit size of 0.1 mm.

Porosity was determined using data of nitrogen gas adsorption using Surface Area Analyzer (SAA) in Materials Physics Laboratory, Department of Materials and Metallurgical Engineering, ITS Surabaya. Before adsorption, the composites were outgassed at 300 ^oC for 3 h. The adsorption process was performed at 77.5 K.

2.4. Adsorption Test

Adsorption test was conducted as the previous research [28] by using the composites which were produced at various calcination temperatures and uncalcined biochar as the control. Each paracetamol solution (10, 20, 30, 40, and 50 ppm) was added to the activated biochar and composites then were shaked for 24 h at 200 rpm. The drug solution concentration was analyzed using UV-Vis spectrophotometer at maximum wavelength of 243 nm. Concentration data was used to make isotherm adsorption curves based on Freundlich, Langmuir, and Dubinin–Raduskevich.

3. Results and Discussions

3.1. Crystal Structure of the Composites

Effects of calcination temperature on crystal structure of the composites were studied using X-ray diffraction method. Result of the characterization is reported in Figure 1. Diffractogram of the activated biochar was taken from product of PUPT 2015 research (product of the patchouli biomass pyrolysis using CoCl₂ activator). The others were based on PUPT 2016 (the composite products from the biochar from PUPT 2015 by using FeCl₃ as impregnant resources).

All diffractograms in Figure 1 are the redrawing results of all theactivated biochar and the composite diffractograms (Figure 2) from the characterizations of thesynthesized products in LSUM.

Pattern of the activated biochar X-raydiffractogram (Figure 1) shows broad peakaround 23 and 43^o which indicates turbostratic graphite structure [55]. Turbostratic is intermediate of graphite. Itis a mixture of amorphous and graphite structure [56].

The composite diffractogram which was produced bycalcination  at 400 ^oC showed no peak of the impregnantdiffractogram. Pattern of the diffractogram just shows pattern of the activatedbiochar’s diffractogram whereas FeCl₃ were added before thecalcination process. It is probably too low iron (III) oxide concentration inthe composite (compared to the activated biochar) which can be detectedsignificantly by the diffractometer as concequency of still little formation ofthe iron oxide at 400^oC. However, the diffractogram peak becomelower than the activated biochar which indicates thermal decomposition of theactivated biochar at 400^oC.

On the other side, the composite which was calcinedat 600 ^oC shows some peaks. Interpretation process was conducted forhematite and maghemite because both structures contain Fe(III) as contained inthe impregnant precursor (FeCl₃). Comparation of the diffractogramd-spacing to standard hematite (Table 1)indicates presence of hematite (α-Fe₂O₃), because itcontains relatively same or very closed 3 main peaks of standard hematite’sdiffractograms. All standard crystals which were used for interpretations ofthe product diffractograms are from American Mineralogyst Crystal StructureDatabase (AMCSD). This database is a very good choice, not only due to easy totake freely in google but also provide more than one choice, based on differentresearchs.

Figure 1 and Table 1 show that the composite which was obtained by calcination at 800 ^oC has higher peaks and more diffractograms than at 600^oC. It indicated the improvement of the α–Fe₂O₃ crystallization process which resulted in better crystallization. Theoritically, higher temperature makes more collisions of reactants which makes more effective chemical reaction. The reaction in calcination process is a solid state reaction which involves ions diffusion and reactions last on the surfaces of the reactants [57]. It means that more collisions in higher temperature may create better ion diffusion.

For other alternative, comparison to standard maghemite (γ-Fe₂O₃) was conducted in Table 2. Result of the interpretations gave conclusion that the products contain maghemite structure because both products which were obtained at 600 and 800^oC have 3 highest diffractogram peaks of the maghemite structure.

Another possibility is matchness of 3 main peaks for standar magnetite (α-Fe₃O₄) and 3 peaks of the products’ diffractogram (Table 3). It indicates that the composite also contained the magnetite. Magnetite is a spinel which can be written as FeFe₂O₄. Precursor of this metal oxide in this synthesis was FeCl₃ which contains iron charge of 3+, so that Fe(II) probably derived from patchouli biomass. Presence of Fe(II) may be related to content of Fe element in the biochar based on EDX analysis whereas no iron (II) substance was included in synthesis process, as reported in our previous research [51,52]. Magnetite contains both Fe(II) and Fe(III) ions in the crystal lattice, each half of the Fe³⁺ ions are located in tetrahedral and octahedral sites. Magnetite can be formulated as FeO·Fe₂O₃ [58].

Comparation to data of AMCSD standard FeCl₃ as precursor of metal oxides gave only the highest peak of 3 main FeCl₃ standard data (Table 4) which was closed to the data of products. It means that presence of FeCl₃ structure in the product is doubted or a weak conclusion.

FeCl₃ is a chemical activator which has a chemical role as the dehydrating agent in the activation reaction along calcination process. This salt has melting point of 300^oC so that it melted at the calcination temperatures of 600 and 800 ^oC [59]. In the biochar activation, the molten salt encapsulated the biochar and aromatize it [60]. As dehydrating agent, FeCl₃ provided the Lewis acid of Fe(III) to attract Lewis base of O atom in the H₂O molecules which was formed by thermal reactions of the oxy functional groups on surface of the biochar. This attraction improves the activation reaction effectivity. For example, the thermal reactions on surface of the biochar which has hydroxyl functional groups (C-OH) by presence of FeCl₃ can be predicted as follows:

The hydroxyl is only one example of the oxy functional groups on surface of the biochar. The other ones can be founded in discussion of section of 3.2.

3.2. Surface Functional Group of the Composites

Characterization of the functional groups using FTIR spectrometry has been performed and the result is reported in Figure 3. All the composite FTIR spectra were obtained using the products synthesized in PUPT 2016. The activated biochar spectrum was obtained from PUPT 2015.

Those overlay FTIR spectra (Figure 3) were based on results of measurements in Figure 4. The overlay is needed to make easier comparation of those spectra, especially the bands at 3500, 1750 and 1200 cm^-1.

Figure 3 showed that spectra of the composite before and after calcination till 400^oC are relatively same. It means that calcination process relatively gave no significant change of surface functional groups. However, the composite calcined at 600^oC provided the significant change of spectra pattern, especially related to a weakening band at about 1750 cm^-1 and 3500 cm^-1, but emerging 2 bands at about 600 and 490 cm^-1. According to the previous researches, the band at 1750 cm^-1 is band of C=O stretching vibration and about 3500 cm^-1 is OH of hydrate or phenolic groups [61]. The unchanged bands in the region around 1200 cm^-1 indicated the stretching of C-O bonds] [62]. By referring to the bands of Cu-O [63], the bands at 600 and 490 cm^-1 may be connected to the stretching vibration of Fe–O. The composite which was calcined at 800^oC showed the disappeared band of C=O at 1750 cm^-1 and sharper band at about 600 cm^-1. It indicated improvement of the iron oxide formation in the composites so that this characterization support presence of the iron oxide structures which were characterized by XRD method (Figure 1).

3.3. Porosity of the Composite

Pore and surface area were characterized using nitrogen gas adsorption method. Isotherm adsorptions of the biochar and the composite are reported in Figure 5. The isotherms were drawn based on raw data in Figure 6 and Figure 7. The composite calcined at 800 ^oC was chosen for this characterization due to its highest crystallinity (Figure 1).

Pattern of the isotherm can be used to identify pore size and porosity qualitatively, such as: 1) much higher position of the biochar curve means much higher nitrogen adsorption and indicates higher pore volume and surface area of the activated biochar than composite, 2) nitrogen adsorption at low P/P_o indicates presence of micropores, 3) hysteresis pattern indicates presence of mesopores, and 4) increasing curve at almost P/P_o = 1 indicates macropore or holes among particle solids which have hole size in the macropore area.

Quantitative data of surface are and pore volume for mesopore, micropore, and macropore areas were calculated and reported in Table 5. The specific surface area such as S_BET was result of adsorption data treatment using BET method. V_p values were calculated based on the adsorption of nitrogen gas volume at P/P_o = 0.99. The mesoporous and macroporous volumes and surface areas were calculated using Pierce Orr Dalla Vale (POD) method. The microporous volumes and surface areas were calculated from subtraction of V_p and S_BET by mesoporous volume and surface area respectively [64].

Data of Table 5 showed that all pore volumes and surface areas of the composite, including mesopore, macropore, and micropore, were lower than of the activated biochar, which indicates the reduction of porosity by impregnation and calcination. Physically, it can be caused by occupation of the impregnant in its pores. It is supported by presence of various iron oxide structures by XRD characterization (Figure 1) and by presence of M-O vibrations by FTIR spectrometry characterization (Figure 3). Chemically, the activation reaction of the biochar surface may reduce some functional groups such as C=O and -OH as also supported by FTIR spectra (Figure 3).

Table 5 confirms that modification of iron oxides on the activated biochar reduced mesoporous volume much lower than other areas in sequence of V_meso > V_macro > V_micro. It indicates that the impregnant especially occupied the mesopore. In other hand, the modification decreased the microporous surface area much lower than microporous and mesoporous areas. It indicates that the activation reaction on surface of the biochar especially affected damage of the microporous area much more than the others. The micropore is the dominant pores of the biochar. It is probably affected by usage of CoCl₂ as activator of carbonization and FeCl₃ as impregnant precursor. Physically salts can act as pore template.

The modification of the activated biochar by the iron oxides also changed the pore size distribution (PSD) as reported in Figure 8. The are various PSD curves, but the curve of V_p and average pore diameter (D) relation was chosen due to closer to the reality. The PSD curves were made based on columns of average V_p and D in the POD table. Figure 8 shows that the activated biochar has not only much higher but also has clearer pattern of PSD curve than the composite. The maximum D (at X axis) is at more than 500 Å which indicates presence of the macropores. The curve pattern shows higher curve for mesoporous area (20–500 Å) than the others and 2 peaks at about 150 and 300 Å which indicate mesopore dominations over the others. However, the modification reduced it much lower which signed occupation of the iron oxides in the mesopores. It is match with the change of mesoporous volume in Table 5.

3.4. Adsorption

Adsorption test has been conducted by paracetamol as adsorbate model of organic pollutant. As a model, the paracetamol has polar functional groups (-OH and N-H) andlow polar (-CH₃) also Lewis bases of N(at N-H) and O (at -OH and C-O). It makes interesting study of the adsorbate and adsorbent chemical interaction.

Three kinds of models (Freundlich, Langmuir, and Dubinin–Radushkevic) were used to treat adsorption data. The Langmuir model is based on an assumption that the adsorption occurs on a homogenous surface. Moreover, the Freundlich model employed assumption of heterogenous surface [65]. Dubinin–Radushkevic model describes the adsorption mechanism onto a heterogeneous surface by a Gaussian energy distribution [66]. The adsorption isotherms are reported in Figure 9. Parameters of adsorption which were calculated based on the isotherms are listed in Table 6

R_L, i.e Langmuir separation factor can be used to predict efficiency of adsorption and affinity of adsorbent–adsorbate, i.e R_L > 1 (unfavorable adsorption), R_L=1 (linier adsorption), 0 < R_L < 1 (favorable adsorption), R_L = 0 (irreversible adsorption) [65]. The R_L values in Table 6 are in the range of 0 < R_L < 1, so that it indicated the favourable adsorptions.

A Freundlich constant, n, is an indicator of adsorption intensity [67]. Value of n between 1 and 10 describes favourable adsorption [55]. The n constant is also an indicator of surface heterogeneity. Higher n value indicates more heterogeneity of the adsorbent [68]. Data of n in Table 6 are in the range between 1 and 10. It indicated favorable adsorption. All calcined composites had lower n values, means lower surface heterogeneities. These conditions were supported by damage of the biochar surface and improvement of the iron oxide structure formations based on X-ray diffractograms (Figure 1). Those favorable adsorptions can be explained by 2 different things. Performances of the composites before calcination and after calcination at 400^oC are probably caused by oxy functional groups of the activated biochar surface and Lewis acid of Fe(III) from FeCl₃. In other hand, performances of the composites after calcination at 600 and 800^oC may be caused by the Lewis acids such as Fe(III) and Fe(II) also Lewis base such as O atoms in the iron oxides.

Data in Table 6 showed that the correlation coefficient (R²) of both Freundlich and Langmuir isotherms decreased by calcination at 400^oC, but increased again at higher temperatures. Correlation coefficient is an indicator how fit a model for describing adsorption isotherm. Decreasing of the correlation coefficient of both Freundlich and Langmuir models may be caused by decreasing of the activated biochar porosity due to the decomposition of biochar structure by calcination as supported by reduction of X-ray diffractogram wide peak of the biochar in Figure 1. However, increasing of the correlation coefficients from 400 to 800 ^oC may be related to increasing of the iron oxide crystalinity, supported by increasing of their peak intensities. So that, for Langmuir model, more homogeneous surface condition was achieved by increasing of iron oxide formation and more damage of the biochar. In this case, the different iron oxides (hematite, maghemite, magnetite) relatively did not affect homogeneity of the adsorbent surface based on Langmuir model.

In other side, the suitable heterogenous surface for Freundlich model was still achieved by formation of various iron oxides. It means that different crystal structures affect heterogeneity of the surface based on the Freundlich model. Although based on the n values heterogeneity of the surface decreased due to damage of the biochar surface, but this condition was still available for Freundlich model with different reason, i.e various iron oxide structures. So that, what the interesting things are that the same tendency for the correlation coefficients of both Langmuir and Freundlich models can be explained in different reasons.

DR isotherm showed different tendency. The correlation coefficient was relatively constant by increasing of calcination temperature to 600^oC. It is probably because the composites in these conditions tended to have heterogenous surfaces, consisting of biochar – FeCl₃ and biochar Fe_xO_y. However, the correlation coefficient is decreased by calcination at 800 ^oC. It is probably affected by decreasing of the composite heterogeneity due to increasing of iron oxide crystalinity and more damage of biochar structure.

Physical/chemical adsorption process can be predicted from adsorption energy (E) which is calculated using K_D (Dubinin Radushkevich’s contant). Adsorption energy between 8 and 16 KJ/mol showed chemical adsorption, otherwise less than 8 Kj/mol indicated physical adsorption [67]. All data of energy in Table 6 are less than 8 KJ/mol, indicated that adsorption by composites before and after calcination is physical adsorption. However, the calcination made improvement of energy, so it indicates the improvement of adsorption strength by increasing of calcination temperatures. This tendency may be related to formation of the iron oxides which improved interaction strength of adsorbate and adsorbent.

Constant of q_m is a maximum monolayer coverage which is calculated using Langmuir model. In other side, q_s is a theoretical isotherm saturation capacity which is calculated using Dubinin–Radushkevich model [69]. Both parameters are used to determine adsorption performance of the composites. Data of q_m and q_s in Table 6 showed that based on Langmuir model, the highest capacity was achieved by calcination at 800 ^oC, whereas based on Dubinin–Radushkevic the highest capacity was obtained by calcination at 600 ^oC. Both capacities are higher than capacity obtained by composite before calcinations. The X–ray diffractogram patterns in Figure 1 showed that both calcined composites contain α-Fe₂O₃ impregnate structure. Based on data of porosity in Table 1, calcination caused lower porosity. Based on FTIR spectra in Figure 3, both calcinations reduce acid functional groups of biochar and improve formation of Fe–O bonding. Therefore, the increasing of capacity by calcination at 600 and 800 ^oC may be due to formation of α-Fe₂O₃ impregnate. Oxygen in α-Fe₂O₃ has ability to make H bond with polar functional groups of paracetamol which strengthen interaction of adsorbate–adsorbent. Therefore, α-Fe₂O₃ impregnate is more beneficial in the adsorption of paracetamol than FeCl₃.

Based on functional group characterization and molecular attraction force of both the activated biochar surface and hematite, maghemite, and magnetite with paracetamol molecules, the graphical abstract can be designed in Figure 9 by including the chemical structure of biochar [7].

Figure 9. Transformation of patchouli biomass to the activated biochar (AB) by pyrolysis and chemical interaction of biochar and hematite with paracetamol molecules.

Figure 9. Transformation of patchouli biomass to the activated biochar (AB) by pyrolysis and chemical interaction of biochar and hematite with paracetamol molecules.

5. Conclusions

This section is not mandatory but can be added to the manuscript if the discussion is unusually long or complex. Composite of FeCl₃/AB has been synthesized at various calcination temperatures to form (α-Fe₂O_3, γ-Fe₂O₃, Fe₃O₄)/AB composite. Some conclusions can be obtained as follows: Pore characterization indicated the reduction of pore volume and surface area of biochar by impregnation and calcinations. X-ray diffraction characterization showed the formation of Fe₂O₃/AB at 600 and 800 ^oC. FTIR spectrophotometry characterization assigned the significant change of spectra at 600 ^oC by indicating the surface functional groups of C=O, C-O, OH, and Fe-O. Adsorption test of paracetamol gave the highest capacity at 800 ^oC (using Langmuir model) by resulting adsorption capacity of 56.37mg/g (R² = 0.964) The highest adsorption energy was achieved at 800 ^oC, i.e 0.530 KJ/mol (physical adsorption) with R_L value of 0.861 (favorable adsorption), and n value of 1.58 (favorable adsorption).