Effects of Chemical Activation Conditions on Hierarchical Porous Carbon via Oxytetracycline Adsorption

Abstract

This study investigated the characteristics and oxytetracycline (OTC) adsorption of hierarchical porous carbons (HPCs) synthesized under different activation conditions using biomass wheat flour (WF) and NaHCO₃ as an activator. It was found that the characteristics of the HPCs, such as specific surface area, total and mesopore volume, defects, and crystallinity, were highly dependent on the activation temperature, the dose of the activator, and activation time. In particular, the adsorption mechanisms and adsorption capacity were dominantly governed by the degree of defects in graphitic structures and mesoporosity, which are largely determined by the dose of the activator and the activation temperature. The best OTC adsorption was achieved with the HPC of the most abundant mesopores, which was prepared at 900 °C, a WF:NaHCO₃ mass ratio of 1:1, and a duration of 2 h (HPC900_R1.0_2H). Based on the results of the experiments concerning adsorption kinetics, equilibrium adsorption, and the effects of pH, OTC adsorption onto HPC900_R1.0_2H was monolayer in type, homogeneous, governed by the combination of diffusion and chemisorption, and largely attributed to π–π electron–donor–acceptor interactions and electrostatic interactions. The thermodynamic parameters suggest that it is spontaneous and endothermic. These findings provide valuable information about the design and synthesis of HPCs from biomass, which exhibit optimized properties for the adsorption of refractory organic pollutants.

1. Introduction

Oxytetracycline (OTC) is a readily available and efficient antibiotic that is used for disease prevention and the growth promotion of livestock and aquaculture [1]. For a significant period of time, a high amount of OTC has been discharged into the environment, resulting in negative consequences on aquatic ecosystems, including detrimental changes in microbial community structures and the spread of antibiotic-resistant bacteria and genes [2,3]. Moreover, conventional wastewater treatment plants (WWTPs) cannot effectively remove OTC from water due to its recalcitrant characteristics [1,2,4].

Several alternatives have been tried in attempts to remove OTC in aqueous environments, such as photolysis, advanced oxidation processes (AOPs), hydrolysis, and ozonation. However, they have several drawbacks: photolysis only degrades compounds that are photosensitive, and the reaction is relatively slow at environmental temperatures; AOPs have poor selectivity and the metals employed to achieve high removal efficiency can be leached, thus causing secondary pollution; hydrolysis is easily affected by the pH and temperature of the solution; and more toxic byproducts can be formed by ozonation [5,6,7,8].

It has been established that the safest and most efficient method to remove OTC from water is through adsorption using carbon-based adsorbents [9]. Adsorption has been extensively employed in the treatment of antibiotic wastewater due to the benefits of convenient operation, low cost, high efficiency, and the absence of toxic byproducts, without a reaction between an adsorbate (i.e., pollutant) and an adsorbent [10]. Carbon-based functional materials have recently been given great attention for a variety of applications, such as the removal of water pollutants, application as a catalyst or electrode, and the development of fuel cells [10,11,12,13]. Their high potential for the adsorptive removal of water pollutants relies on the well-developed pore structure, high specific surface area, chemical and thermal stability, and abundant functional groups, i.e., graphitic structures, C=O, C–O–C, –OH, and –NH₂ [9,12]. Moreover, the properties can be modified via chemical and physical methods to achieve target pollutants with high adsorption capacity, which can be used extensively in water treatment [12,14].

Some of them have been studied for the adsorption of OTC organic contaminants such as multi-walled carbon nanotubes [15], graphene oxide [16], activated carbon [9,17], biochar [18,19], and biological sludge [20]. Their advantages include low cost, the abundance of raw materials, chemical stability, and high surface area, which is mainly attributed to the presence of micropores and mesopores [21].

However, conventional carbon-based adsorbents have irregular and defective pore structures with limited interconnectivity, which inhibits mass transfer and diffusion, thus restricting molecular access to the adsorbent surfaces [22,23,24]. In this regard, hierarchically porous carbon (HPC) can be an efficient alternative to conventional carbon-based materials. HPC is characterized by a well-defined pore structure consisting of both micropores and mesopores, with a pore size distribution that spans a range of less than 2 nm to 50 nm [25,26].

HPCs are generally prepared using templates, such as nanoscale spheres of SiO₂, resorcinol formaldehyde resin, and CaCO₃ [25,27]. The use of templates results in complex synthesis procedures and the generation of acidic wastewater. On the other hand, the drawbacks can be ruled out using the activation of pyrolysis, which is one of the template-free methods. In general, chemical activation can partially oxidize/gasify the components of carbon-based materials, which leads to a significant increase in the number of micropores and mesopores and the transformation of surface functional groups [28,29].

The relative abundance of micropores and mesopores in HPC generated by chemical activation depends on several factors, including the sources of the material and the activation conditions, such as the type of activating agent, its concentration, temperature, and time [23]. For example, insufficient and excess use of an activation agent results in poor and amorphous pore structures and pore coalescence or collapse, respectively [30,31]. The creation of pores and increased surface area are promoted with increasing temperature, but it is accompanied by the destruction of micropores, increases in fixed carbon and ash content, and the excessive volatilization of functional groups [23,31]. Therefore, the establishment of activation conditions is essential to the development of required pore structures. However, it also suggests that the pore structures in HPCs can be controlled by the activation conditions, which increases the variety of potential applications of HPCs.

Other than the activation conditions, the selection of a carbonaceous raw material is also very important when aiming to prepare an HPC of a specific purpose, such as an adsorbent or a supercapacitor [25,32]. It was reported that the pores and microscopic structures of carbonized biomasses, i.e., biochars, are greatly influenced by feedstock structures, as demonstrated by the biochars derived from rice straw, swine manure, Douglas fir (Pseudotsuga menziessii), and hybrid poplar (Populous deltoids) [33,34]. In general, the raw organic materials used as precursors of HPC need to satisfy a couple of requirements: first, they must be capable of producing a substantial amount of HPC following heat treatment, and second, they must be able to fabricate carbons with a disordered microstructure (non-graphitizable carbon). The carbon sources that meet both requirements are biomass-derived products and certain types of polymers [11,23,35].

Meanwhile, chemical activation involves the mixing of an activating agent, such as H₃PO₄, ZnCl₂, KOH, and NaHCO₃, with a raw material, followed by high-temperature treatment in an inert gas atmosphere [36,37]. The most widely used chemical activator is KOH, but it has a critical drawback in that it is an extremely corrosive substance according to NFPA 704 hazard labels [35], limiting the use of it at an industrial level. Thus, the use of more benign activating agents (e.g., sodium bicarbonate (NaHCO₃)) that lead to a greener activation process is required [35,38].

Regarding the above, in this study, wheat flour (WF)-based HPC was fabricated with NaHCO₃ as the activator under various activation conditions. WF has been used in attempts to prepare supercapacitors [10,11]. However, the activation conditions must be optimized for the resultant material to have the optimum property for the adsorption of organic pollutants, such as OTC. For example, the adsorption is largely attributable to the defects, mesopores, and the degree of graphitization for a better transport and for the π–π electron–donor–acceptor (EDA) interactions [9,12], while the capacitance is more related to the micropores and O-containing groups [10,11]. Therefore, the HPCs were characterized using various techniques such as N₂ adsorption/desorption isotherms, scanning electron microscopy (SEM), X-ray diffraction (XRD), Fourier Transform Infrared (FTIR) spectroscopy, and Raman spectroscopy to analyze the pore structure and surface properties. The adsorption of OTC on to the HPCs were evaluated using the kinetics and equilibrium of OTC adsorption. In addition, the activation condition, characteristics, and the adsorption performance were discussed comprehensively.

2. Materials and Methods

2.1. Materials and Reagents

WF was purchased from Homeplus Co., Ltd. (Seoul, Republic of Korea). NaHCO₃ and ethanol (C₂H₅OH, ≥99.9%) were acquired from Samchun (Seoul, Republic of Korea). Oxytetracycline hydrochloride (C₂₂H₂₄N₂O₉·HCl, OTC·HCl), hydrochloric acid (HCl, 37%), and sodium hydrate (NaOH, ≥97%) were acquired from Merck KGaA (Darmstadt, Germany). The chemicals were all of analytical grade and utilized as they were received. The deionized water was obtained from the water purification system, Aquapuri 551 system (Younglin, Anyang, Republic of Korea).

2.2. Preparations of HPCs

HPC was prepared via the pyrolysis of WF and NaHCO₃ mixture according to previous work with some necessary modification [38]. Typically, 24 g of wheat powder was mixed with a different amount of NaHCO₃ (0 g, 12 g, 24 g) and poured into 300 mL of distilled water to produce a paste under stirring. Then, the paste was carbonized at different temperatures (600, 800, and 900 °C) with a 5 °C/min heating rate for a different activation time (1, 2, and 3 h) under N₂ atmosphere. The temperature was selected by considering previous studies, which state that an organic material is carbonized over 600 °C [39]. The resulting black powder was rinsed sequentially with diluted HCl solution (1 M), deionized water, and ethanol. It was dried for 12 h at 80 °C and then ground. The final product was named HPCX_RY_ZH, where X is the activation temperature, Y is the mass ratio between activator and biomass (NaHCO₃/WF, g/g), and Z is the activation time.

2.3. Materials Characterization

The pore structure and the specific surface area were analyzed using a BELSORP-mini (Microtrac BEL, Osaka, Japan). The N₂ adsorption/desorption isotherms of the HPCs were obtained at 77K after degassing at 110 °C for 2 h under N₂ gas. The Brunauer–Emmett–Teller (BET) method was used to calculate the specific surface area (SSA), and the total pore volume (PV) was evaluated at a relative pressure (P/P₀) of 0.99. The t-plot method was used to estimate micropore volume. The HPCs were sputtered with Pt, and the microscopic images were taken using a field emission SEM (FE-SEM; Carl Zeiss, Oberkochen, Germany). Raman spectra were obtained using an in Via Raman microspectrometer (Renishaw, Wotton-under-Edge, UK) with a 514 nm excitation wavelength and 4 cm⁻¹ spectral resolution. FTIR spectra were listed at 4000–400 cm⁻¹ with the pellets prepared from the mixture of 0.2 mg HPC and 200 mg of KBr, using an FTIR spectrophotometer (Spectrum One System, Perkin-Elmer, Waltham, MA, USA). XRD patterns were obtained using a DB Advance X-ray diffractometer (Bruker, MA, USA) in the 2θ range of 5–69.9925°. The zeta potential was analyzed in the pH range of 2–11 using a zeta potential analyzer (ZetaPlus, Brookhaven Instruments, New York, NY, USA), and with a suspension of 5 g/L HPCs in a 0.01M KCl aqueous solution.

2.4. Adsorption Experiments

Batch adsorption kinetic experiments were conducted in a 250 mL amber bottle with continuous mixing. The HPCs were dispersed in DIW, followed by the introduction of OTC stock solution. Samples were taken at predetermined times. The equilibrium adsorption isotherm experiments were performed in 50 mL amber glass tubes. The 0.025 g of the HPCs were dispersed in 40 mL OTC solution of 10–600 mg/L. The tubes were shaken in a shaking water bath (BS-31, Lab Companion, JEIO Tech., Daejeon, Republic of Korea) for 4 h at 298, 308, and 318 °K. All samples were filtered through a 0.45 μm polyvinylidene fluoride (PVDF) filter, and then the concentration of OTC in the filtrates was measured by a UV–visible spectrophotometer (UV-1280, Shimadzu, Kyoto, Japan) at 358 nm. All experiments were conducted in triplicate. The pH did not change during the experiments and was 5.60 ± 0.25 for all experiments, unless noted otherwise.

2.5. Modeling of Adsorption

The results of adsorption kinetic experiments were fitted to a pseudo-first-order adsorption kinetic model (Equation (1)), a pseudo-second-order adsorption kinetic model (Equation (2)), an intra-particle diffusion model (Equation (3)), and an Elovich equation (Equation (4)):

$$\frac{d{q}_t}{dt}=k_{a1}\left(q_e-q_t\right),$$

(1)

$$\frac{d{q}_t}{dt}=k_{a2}{\left(q_e-q_t\right)}^2,$$

(2)

$$q_t=k_{id}{\left(t\right)}^a,$$

(3)

$$\frac{d{q}_t}{dt}={\alpha }_{El}\mathit{exp}\left(-{\beta }_{El}{q}_t\right),$$

(4)

where, q_t is the amount of OTC adsorbed at time t (min) (mg/g), q_e is the equilibrium adsorption amount (mg/g), k_a1 is the pseudo-first-order adsorption rate constant (min⁻¹), k_a2 is the pseudo-second-order adsorption rate constant (g/mg‧min), k_id is the rate constant of intra-particle diffusion, α_El is the initial adsorption rate (mg/g‧min), and β_El is a constant related to the surface coverage and activation energy of adsorption (g/mg). The experimental results were fitted to the models via nonlinear fitting, using MATLAB (R2021a, The MathWork, Inc., Natick, MA, USA).

The results of the equilibrium adsorption isotherm experiments were described using Langmuir (Equation (5)), Freundlich (Equation (6)), and Temkin (Equation (7)) isotherms:

$$q_e={q\frac{K_L{C}_e}{1+K_L{C}_e}}_{max},$$

(5)

$$q_e=K_F{C_e}^{1/n},$$

(6)

$$q_e=\frac{RT}{b_T}\mathit{ln}\left(A_T{C}_e\right),$$

(7)

where, q_e is the equilibrium adsorption amount (mg/g), q_max is the maximum adsorption capacity (mg/g), C_e is the equilibrium concentration of OTC (mg/L), K_L is the Langmuir adsorption constant (L/mg), K_F is the Freundlich constant ((mg/g)‧(L/mg)^1/n), 1/n is a constant related to the adsorption intensity, R is the gas constant (8.314 J/mol‧°K), T is the absolute temperature (K), b_T is the Temkin constant associated with the heat of adsorption (J/mol), and A_T is the Temkin equilibrium constant (L/mg).

3. Results and Discussion

3.1. Characterization of HPCs

3.1.1. Microstructure and Pore Structure

The FE-SEM images of all HPCs showed the co-existence of macropores, mesopores, and micropores, indicating that the hierarchical porous structures have been successfully fabricated (Figure 1). The carbonization of WF activated by NaHCO₃ led to form a 3D scaffolding framework [38]. The surfaces of HPC900_R0.5_2H (Figure 1B) were rougher and more aggregates of smaller particles were found, suggesting more developed pore structures than those of HPC600_R0.5_2H (Figure 1A). The effects of NaHCO₃ dose are shown in Figure 1B–D for the ratio of NaHCO₃ to WF of 0, 0.5, and 1.0, respectively. The HPC900_R0.0_2H showed a negligible development of surface pores. However, surface pores increased as the activator dose increased, because of the increase in the release of gaseous products and the expansion of the graphitic lattices via the intercalates generated by the metallic Na [40]. The effects of activation time are shown in Figure 1B,E,F, showing the surfaces of the HPCs activated for 1, 2, and 3 h, respectively. The surface of HPC900_R0.5_1H displayed macropores on its surface, and they increased with increasing activation time (HPC900_R0.5_2H and HPC900_R0.5_3H). In addition, small pores were found on the surface of HPC900_R0.5_3H, which can be assigned to the exposed mesopores (the inset of 1(F)).

These clearly suggest the detrimental effects of NaHCO₃ dose, temperature, and activation time on the pore structure of HPCs, which led to the enhanced hierarchical pore structure in the carbonized WF, as they increased, under the conditions in this study [30,35,38]. However, the inner-sphere pores could not be properly characterized by the images in Figure 1. Therefore, the details of the effects of the activation conditions on the pore structures were further investigated with specific surface area (SSA), total pore volume (V_total), mesopore volume (V_meso), micropore volume (V_micro), and average pore diameter (d_a) of the HPCs (

Table 1. Pore characteristics of HPCs as a function of pyrolysis temperature.

	Specific Surface Area (SSA)	Total Pore Volume (Vtotal)	Micropore Volume (Vmicro)	Mesopore Volume (Vmeso)	Vmeso/ Vtotal	Average Pore Diameter (da)
	(m2/g)	(cm3/g)	(cm3/g)	(cm3/g)		(nm)
HPC600_R0.5_2H	460.54	0.2281	0.1779	0.0366	0.1605	1.9811
HPC800_R0.5_2H	622.29	0.2879	0.2513	0.0369	0.1282	1.8505
HPC900_R0.5_2H	648.13	0.2908	0.2539	0.0502	0.1726	1.7947

,

Table 2. Pore characteristics of HPCs as a function of the ratio between WF and NaHCO₃.

	Specific Surface Area (SSA)	Total Pore Volume (Vtotal)	Micropore Volume (Vmicro)	Mesopore Volume (Vmeso)	Vmeso/ Vtotal	Average Pore Diameter (da)
	(m2/g)	(cm3/g)	(cm3/g)	(cm3/g)		(nm)
HPC900_R0.0_2H	662.75	0.2750	0.2611	0.0139	0.0505	1.6595
HPC900_R0.5_2H	648.13	0.2908	0.2539	0.0369	0.1269	1.7947
HPC900_R1.0_2H	633.30	0.4382	0.3070	0.1312	0.2994	2.7678

and

Table 3. Pore characteristics of HPCs as a function of activation time.

	Specific Surface Area (SSA)	Total Pore Volume (Vtotal)	Micropore Volume (Vmicro)	Mesopore Volume (Vmeso)	Vmeso/ Vtotal	Average Pore Diameter (da)
	(m2/g)	(cm3/g)	(cm3/g)	(cm3/g)		(nm)
HPC900_R0.5_1H	259.01	0.1169	0.0975	0.0194	0.1659	1.8054
HPC900_R0.5_2H	648.13	0.2908	0.2539	0.0369	0.1269	1.7947
HPC900_R0.5_3H	720.81	0.3720	0.3044	0.0676	0.1817	2.0643

), using N₂ adsorption/desorption isotherms.

The SSA and V_total of the HPCs increased greatly as the pyrolysis temperature was elevated from 600 °C to 800 °C, but slightly when it was further raised to 900 °C. Table 1 shows that the increased SSA and V_total were attributed to the formation and decrease in micropores and mesopores, respectively, resulting in the decrease in d_a. It was reported that Na₂CO₃ and the C in biomass are thermally decomposed to form Na₂CO₃, H₂O, CO, and CO₂ (Equations (8) and (9)), under relatively low temperatures [38]. An increase in temperature near to or over the fusion point of Na₂CO₃, i.e., 851 °C, Na₂CO₃ and its byproduct (NaO) are decomposed to Na and CO, generating micropores [23,38]. In addition, mesopores of carbonaceous materials would be broken into smaller pores under high temperature [30,38].

$$2NaHC{O}_3\to N{a}_{2}C{O}_3+C{O}_2+H_{2}O,$$

(8)

$$C+C{O}_2\to 2CO,$$

(9)

$$N{a}_{2}C{O}_3\to N{a}_{2}O+C{O}_2,$$

(10)

$$N{a}_{2}C{O}_3+2C\to 2Na+3CO,$$

(11)

$$N{a}_{2}O+C\to 2Na+CO,$$

(12)

Table 2 shows that the SSA slightly decreased, while the V_total significantly increased as the ratio of NaHCO₃ to WF increased from 0 (no activator) to 1.0. It is thought that the increase in V_total is largely attributed to the increase in mesopores, considering the slight and great increase in V_micro and V_meso, respectively (Table 2). This indicates that the release of CO, CO₂, and H₂O (Equations (8)–(12)) results in the formation of mesopores, rather than micropores, as well as in the widening of existing micropores via their destruction [30,31,41]. On the other hand, the SSA was the highest when no NaHCO₃ was added, i.e., HPC900_R0.0_2H, and the decrease in the SSA was not as significant as the increase in the SSA with increasing temperature (Table 1). Table 3 shows that the SSA, V_total, and V_micro greatly increased when the activation time increased from 1 to 2 h, but less significantly with a further increase to 3 h. The V_micro showed a linear correlation with the activation time (Pearson correlation coefficient of 0.988).

Considering the results presented in Table 1, Table 2 and Table 3, it is suggested that the SSA and V_micro in the HPCs are more dependent on the activation time rather than temperature and NaHCO₃ dose, while V_meso is the most significantly affected by NaHCO₃ dose. However, the dominance of micropores was found for all HPCs, ranging from 70.1 (HPC900_R1.0_2H) to 95.0 (HPC900_R0.0_2H) %. The plot of adsorption–desorption isotherm and the pore size distribution can be found on Figures S1 and S2, respectively.

3.1.2. FTIR Spectroscopy

Figure 2 displays the FTIR spectra of the HPCs, showing the common bands, corresponding to –OH stretching (at 3430 cm⁻¹), aliphatic C-H_n stretching (at 2920 and 2850 cm⁻¹), adsorbed CO₂ (at 2360 and 2335 cm⁻¹), and aromatic C=C and/or conjugated C=O stretching (at 1630 cm⁻¹) [30,42].

The intensities of the –OH and CO₂ bands decreased, while those of the bands of aliphatic C–H_n and C=C/C=O did not exhibit any notable changes, as the temperature increased from 600 to 900 °C (Figure 2A). This indicates that more –OH and CO₂ was decomposed and/or volatilized over 600 °C, while the aliphatic C–H_n and C=C/C=O were formed at 600 °C and are thermally stable at the temperature up to 900 °C [31,43].

The intensities of the –OH, C–H_n, and CO₂ bands decreased with increasing NaHCO₃ amount (Figure 2B). The band of C=C/C=O decreased, and the noise of the band also decreased, as the ratio increased from 0 to 0.5, indicating the formation of more rigid structures with a higher degree of graphitization when an activator (NaHCO₃) was used [43]. An increase in the injection counts of the NaHCO₃ (HPC900_R1.0_2H) resulted in increased intensity and noises of the C=C/C=O band, suggesting more amorphous aromatic structures.

The intensity of the CO₂ band decreased with increasing activation time, while that of –OH, C–H_n, and C=C/C=O decreased and then increased, as the time increased from 1, 2, and then 3 h (Figure 3C). This indicates an increase in the abundance of the functional groups on the HPC surfaces via intermolecular interactions under a longer reaction time [44]. This is also attributable to the etching effect of NaHCO₃ on the porous carbon during a longer activation process [31].

3.1.3. XRD

The XRD patterns of all of the HPCs exhibit two peaks at the 2θ of ~24° and ~44°, which correspond to the (002) plane of the interlayer spacing and (100) planes of the graphitic structures, respectively, as demonstrated in the literature (Figure S3) [11,38]. Also, a sharp peak at 2θ of ~14°, which corresponds to the signal of the native starch from wheat flour, was found [45,46].

Figure 3A displays that the XRD spectra of HPC600_R0.5_2H showed some small peaks at 2θ of ~23°, ~27°, and ~33°, indicating the presence of Na₂CO₃ formed from NaHCO₃ decomposition (Equation (8)) [38]. For HPC900_R0.5_2H, the peaks of (002) and (100) planes were more intense and wider than those in HPC600_R0.5_2H, and the peaks of Na₂CO₃ disappeared and were replaced by two (2) peaks near ~22°, which may correspond to the decomposition of Na₂CO₃ to Na₂O (Equation (9)). These changes in the XRD patterns by increasing the temperature condition suggest an increase in the degree of graphitization, due to the existence of more pores and defects in carbon material [10,38,44].

Figure 3B shows that the XRD pattern of the HPC prepared without NaHCO₃ (HPC900_R0.0_2H) consists of (002) and (100) planes of graphitic structures, as well as a peak of starch, but no notable peaks of Na₂CO₃ (2θ of ~23°, ~27°, and ~33°) or Na₂O (2θ of ~22°). The peaks of (002) and (100) planes became more intensive and sharp peaks of Na₂O were detected as the amount of NaHCO₃ increased, i.e., HPC900_R0.5_2H, demonstrating the activator’s evident impact on the formation of more crystalline graphitic structures. On the other hand, the pattern of HPC900_R1.0_2H shows that the intensity of the peaks of (002) and (100) planes was depleted, while that of the peak at 2θ of ~22° became stronger. This suggests that this could be explained due to the abundant availability of Na₂O (Equations (8) and (9)), leading to an increase in the formation of pores and defects, but resulting in a decreased degree of graphitization and crystallinity [38].

The XRD pattern of the HPC, prepared at 900 °C and 1 h, consists of the peaks of (002) and (100) planes, as well as of Na₂CO₃ (Figure 3C). An additional peak of Na₂O was found in that of HPC900_R0.5_2H, and the intensities of the peaks of (002) and (100) planes increased, indicating more developed graphitic structures. However, with the additional one hour of activation time, i.e., HPC900_R0.5_3H, the peaks of Na₂CO₃ and Na₂O disappeared and the (002) plane weakened, indicating the complete decomposition of NaHCO₃, Na₂CO₃, and Na₂O, which can lead to more pores created as well as more defects. It has been reported that a longer thermal treatment under a higher temperature is responsible for the depletion of certain minerals, such as Na₂O and quartz [44,47].

3.1.4. Raman Spectroscopy

The Raman spectra of HPCs are illustrated in Figure 4A, and they were deconvoluted into seven (7) bands, as illustrated in Figure 4B for HPC900_R0.5_3H. The bands were at 1210, 1350–1351, 1540–1541, 1580–1581, 1610, 2900–2923, and 2626–2670 cm⁻¹, which are assigned to D4, D, D3 (D′), G, D2 (D′), 2D (G′), and 2D+G (D+D′) bands, respectively.

The D4 band is associated with disordered graphitic lattice, polyenes, and ionic impurities. The bands of D, D3 (D′), G, and D2 (D′) are associated with disordered graphitic lattice induced by the sp³ hybridization such as graphene layer edges, amorphous carbon, ideal graphitic lattice, and disordered graphitic lattice in surface graphene layers, respectively. The intensity of the D band was substantial for all HPCs, suggesting that all of them possess a high degree of defects and porosity [48]. The 2D and D+G bands are associated with single-layer graphene and a combination of the D and G bands, respectively [49]. Therefore, the ratio of the area of D band to G band (I_D/I_G) can be used to evaluate the degree of structural defects and disorder, as well as the sizes of the graphite nanocrystallites, in carbon materials OTC [50]. The ratio of 2D and to G band (I_2D/I_G) is an indicator of the relative abundance of single-layer graphene in a graphitic structure, while the types of the defects can be suggested by that of D band to D2 band (I_D/I_D2).

Table 4. Result of Raman spectroscopy of HPCs with different activation temperatures.

		D4	D	D3	G	D2	D+G	2D	ID/IG	I2D/IG	ID/ID2
HPC600_	Center (cm−1)	1210	1350	1540	1580	1610	2900	2660	4.528	1.579	15.700
R0.5_2H	Fraction (%)	0.126	0.584	0.124	0.129	0.037	0.599	0.401
HPC900_	Center (cm−1)	1210	1350	1541	1580	1610	2906	2670	7.887	1.766	11.883
R0.5_2H	Fraction (%)	0.097	0.637	0.131	0.081	0.054	0.567	0.433

shows that the I_D/I_G increased from 4.528 to 7.887 as the temperature increased from 600 °C (HPC600_R0.5_2H) to 900 °C (HPC900_R0.5_2H). This points out the increase in disordered carbon structures or defects in the HPCs, while the size of the carbon crystallites decreased, with increasing temperature [10,38,48]. These suggest a higher capacity of surface chemical reactions and adsorption via accelerated electron transfer and stronger π–π interactions [50,51] for HPC900_R0.5_2H than for HPC600_R0.5_2H. The I_2D/I_G also increased from 1.579 (HPC600_R0.5_2H) to 1.776 (HPC900_R0.5_2H), which agrees with the increase in I_D/I_G. On the other hand, the I_D/I_2D decreased from 15.700 to 11.883 for HPC600_R0.5_2H and HPC900_R0.5_2H, respectively. This indicates that the defects in HPC600_R0.5_2H and HPC900_R0.5_2H were more associated with sp³–C-related defects and hopping defects, respectively [41].

The I_D/I_G and I_2D/I_G increased from 4.821 to 7.959 and from 1.638 to 2.818, respectively, while the I_D/I_2D decreased from 12.607 to 10.765, as the NaHCO₃ amount increased (

Table 5. Result of Raman spectroscopy of HPCs with different activator dose.

		D4	D	D3	G	D2	D+G	2D	ID/IG	I2D/IG	ID/ID2
HPC900_	Center (cm−1)	1210	1350	1540	1580	1610	2915	2670	4.821	1.638	12.607
R0.0_2H	Fraction (%)	0.114	0.592	0.124	0.123	0.047	0.410	0.590
HPC900_	Center (cm−1)	1210	1350	1541	1580	1610	2906	2670	7.887	1.766	11.883
R0.5_2H	Fraction (%)	0.097	0.637	0.131	0.081	0.054	0.567	0.433
HPC900_	Center (cm−1)	1210	1351	1540	1580	1610	2911	2670	7.959	2.818	10.765
R1.0_2H	Fraction (%)	0.105	0.625	0.133	0.079	0.058	0.448	0.552

). In addition, the results in

Table 6. Result of Raman spectroscopy of HPCs with different activation time.

		D4	D	D3	G	D2	D+G	2D	ID/IG	I2D/IG	ID/ID2
HPC900_	Center (cm−1)	1210	1351	1541	1580	1610	2911	2670	5.729	1.603	19.451
R0.5_1H	Fraction (%)	0.107	0.621	0.132	0.108	0.032	0.493	0.507
HPC900_	Center (cm−1)	1210	1350	1541	1580	1610	2906	2670	7.887	1.766	11.883
R0.5_2H	Fraction (%)	0.097	0.637	0.131	0.081	0.054	0.567	0.433
HPC900_	Center (cm−1)	1210	1350	1540	1580	1610	2900	2660	7.888	1.953	9.764
R0.5_3H	Fraction (%)	0.088	0.647	0.117	0.082	0.066	0.542	0.458

show an increase in the I_D/I_G and I_2D/I_G, from 5.729 to 7.888 and from 1.603 to 1.953, respectively, and a decrease in the I_D/I_2D, from 19.451 to 9.746, with increasing activation time.

The results in Table 4, Table 5 and Table 6 strongly suggest that an increase in activator dose, elevating temperature, and a prolonged activation time commonly induced similar changes in the HPCs, such as increasing the degree of defects, especially hopping defects, and the relative number of single-layer graphenes. However, the most and the least significant factors for the I_D/I_G were temperature and activation time, respectively; for the I_2D/I_G, they were activator amount and temperature, respectively; and for the I_D/I_2D, they were activation time and activator amount, respectively. Therefore, it is suggested that the degree of defects, formation of graphene layers, and the relative abundance of hopping defects are mostly governed by temperature, activator amount, and activation time, respectively.

3.2. Adsorption of OTC

3.2.1. Effect of Activation Conditions of HPCs on OTC Adsorption Kinetics

The effects of the activation conditions on the OTC adsorption kinetics by the HPCs are presented in Figure 5 and

Table 7. Result of OTC adsorption kinetics using HPCs prepared with different temperatures, activator amount, and activation time.

Model		Temperature Effect			Activator Ratio Effect			Activation Time Effect
		HPC600_ R0.5_2H	HPC800_ R0.5_2H	HPC900_ R0.5_2H	HPC900_ R0.0_2H	HPC900_ R0.5_2H	HPC900_ R1.0_2H	HPC900_ R0.5_1H	HPC900_ R0.5_2H	HPC900_ R0.5_3H
	ka1 (min−1)	0.7164	0.0603	0.1746	0.0789	0.1746	0.2472	0.0938	0.1746	0.0981
Pseudo-first-order	qe1 (mg/g)	7.3002	10.5649	31.0715	7.9271	31.0715	79.9331	3.1866	31.0715	42.8964
	r2	0.9715	0.9356	0.8836	0.9172	0.8836	0.9252	0.8746	0.8836	0.8985
	SSE	1.36	6.68	112.04	5.21	112.04	463.43	1.27	112.04	194.00
	ka2 (g/mg·min)	0.1752	0.0898	0.0079	0.0120	0.0079	0.0044	0.0399	0.0079	0.0029
Pseudo-second-order	qe2 (mg/g)	7.6021	11.1119	33.5996	9.1863	33.5996	85.8779	3.4861	33.5996	47.4636
	r2	0.9930	0.9716	0.9545	0.9550	0.9545	0.9812	0.9234	0.9545	0.9551
	SSE	0.33	2.95	43.79	1.05	43.79	116.59	0.78	43.79	85.86
Elovich	αEl (mg/g)	16,253.980	1600.437	32.7913	2.5162	32.7913	202.7550	1.4056	32.7913	35.1593
Equation	βEl (g/mg)	1.9062	1.0417	0.1878	0.5795	0.1879	0.0843	1.6265	0.1879	0.1135
	r2	0.9575	0.8793	0.9964	0.9595	0.9964	0.9805	0.9433	0.9964	0.9929
	SSE	1.36	6.68	111.98	5.21	111.98	463.53	1.27	111.98	194.00

Note: SSE: Sum-Squared Error.

. As presented in Figure 5A, HPC900_0.5R_2H exhibits the highest adsorption rate and capacity compared to HPC600_0.5R_2H and HPC800_0.5R_2H. The pseudo-second-order kinetic model fits better for HPC600_0.5R_2H and HPC800_0.5R_2H, demonstrating that the OTC adsorption onto them is mainly attributable to chemical mechanisms and that the adsorption was affected by the OTC amount on the surfaces [9,13,44]. Meanwhile, the HPC900_R0.5_2H condition is better fitted with the Elovich equation model (r² > 0.99), demonstrating that the OTC adsorption onto HPC900_R0.5_2H is the combination of bulk and surface diffusion, as well as the chemisorption, on energetically heterogeneous surfaces [9,52].

Figure 5B presents that OTC adsorption was greatly enhanced as the NaHCO₃ amount increased. It should be noted that HPC900_R0.0_2H showed the lowest OTC adsorption, even though it has the highest SSA (662.75 m²/g). This could be explained by the poor development in mesopores and a higher fraction of micropores (Table 2), where the OTC diffusion is limited [23]. The adsorption was enhanced as the activation time increased (Figure 5C). It seems reasonable that this is attributable to the increased SSA and mesopores as given in Table 3.

Meanwhile, the Elovich equation model provided good fits to the OTC adsorption onto the HPCs prepared at 900 °C (HPC900_0.0R_2H, HPC900_0.5R_2H, HPC900_1.0R_2H, HPC900_0.5R_1H, and HPC900_0.5R_3H), regardless of activator amount and activation time. This strongly suggests that the temperature influenced the adsorption mechanisms more than the amount of activator or activation time. It was also supported by the more significant effect of the temperature on SSA and I_D/I_G, than activator amount and activation time (Table 1, Table 2, Table 3, Table 4, Table 5 and Table 6). However, the performance, i.e., equilibrium adsorption amount and adsorption rate, was more significantly affected by the activator amount (Figure 5), by which the V_meso was the most influenced (Table 1, Table 2 and Table 3).

The OTC adsorption was further studied hereinafter, using HPC900_R1.0_2H, which showed the best performance.

3.2.2. Effect of pH

Figure 6A shows that the OTC removal was not notably affected in a pH range of 2–8, but significantly decreased as the pH further increased to 11. The pH would significantly affect adsorption, when it is largely governed by electrostatic interactions, influencing the charge of an adsorbent and an adsorbate. In particular, OTC has three (3) pK_as because of its multiple ionizable functional groups, making it an amphoteric molecule [53]. OTC dominantly exits as OTC⁺ at pH < 3.4, OTC⁰ at 3.4 < pH < 7.6, OTC⁻ at 7.6 < pH < 9.7, and OTC²⁻ at pH > 9.7 [54]. On the other hand, the pH_PZC of HPC900_R1.0_2H was 3.9, making the adsorbent renders positively and negatively charged, when the solution pH is under and over the pH_PZC, respectively (Figure 6B). However, the result in Figure 6A suggests that the OTC adsorption onto HPC900_R1.0_2H was not solely affected by the electrostatic interactions, but also significantly influenced by π–π EDA interactions and hydrophobic effects [55,56].

As the pH is decreased, the π–electron-accepting ability of OTC increases; however, electrostatic repulsion increases because both HPC900_R1.0_2H and OTC are positively charged, and hydrophobicity decreases by protonation [13,56]. Under the pH 3.4 < pH < 7.6, OTC is apparently not charged, having the highest hydrophobicity. Therefore, the stable OTC adsorption at acidic-to-neutral pHs is attributable to the strong π–π EDA interactions, overwhelming the others, between OTC molecules, which accept π–electrons via the conjugated ionic structures, and the surfaces of HPC900_R1.0_2H, which donate π–electrons because of the C=O group and well-crystallized graphitic C–C [57]. On the other hand, the electrostatic repulsion increases because of the dominance of OTC⁻/OTC²⁻ and the increase in negative charge on HPC900_R1.0_2H as the pH increased. In addition, the π–π EDA interactions are suppressed, and the hydrophobicity decreases because of deprotonation and ionization under alkaline conditions [26,56].

3.3. Adsorption Isotherm and Thermodynamic Study

The equilibrium adsorption amount (q_e) of OTC on HPC900_R1.0_2H gradually increased as the temperature increased from 298 to 318 °K (Figure 7). The three (3) widely used isotherm models (Langmuir, Freundlich, and Temkin) were selected to analyze the results [44,58]. Langmuir isotherm provided the best fit (

Table 8. Result of adsorption isotherm.

Isotherm			298 °K	308 °K	318 °K
Langmuir	qmax	(mg/g)	87.184	128.370	162.707
	KL	(L/g)	0.052	0.025	0.030
	r2		0.999	0.995	0.996
	SSE		67.38	110.44	303.40
Freundlich	KF	((mg/g)∙(L/g)1/n)	13.596	14.912	17.194
	n		3.082	2.763	2.571
	r2		0.936	0.973	0.953
	SSE		901.64	683.14	1609.76
Temkin	bT	(J/mol)	0.006	0.008	0.010
	AT	(L/mg)	1.274	0.883	0.877
	r2		0.936	0.939	0.951
	SSE		380.57	69.22	231.37

), suggesting that the adsorption is homogeneous and monolayer, where the surface coverage does not influence the adsorption enthalpy [58]. The comparison of the OTC adsorption performance of different adsorbents can be found in Table S1. The HPC material has a higher adsorption capacity than most conventional biochars.

Equations (13) and (14) were used to calculate the Gibbs free energy change (ΔG⁰) and the changes of enthalpy (ΔH⁰) and entropy (ΔS⁰) of the adsorption [9,59,60]:

$$∆G^0=-RT \ln \left(K^0\right),$$

(13)

$$\ln \left(K^0\right)=-\frac{∆H^0}{RT}+\frac{∆S^0}{R}$$

(14)

where, K⁰ is the equilibrium constant, calculated with the correlation of the OTC mass on the surface of the adsorbent (m_s) and that in the liquid phase (m_e), R is the ideal gas constant (8.314 J/mol·K), T is the absolute temperature (K), ΔG⁰ is the standard Gibbs free energy change, ΔH⁰ is the change in the standard Gibbs free energy change, and ΔS⁰ is the change in the entropy. The correlation of ln (K⁰) vs. 1/T can be found on Figure S4.

Table 9. Thermodynamic parameters of OTC adsorption by HPC900_R1.0_2H.

		298 K	308 K	318 K
ΔG⁰	(kJ/mol)	−3.093	−4.109	−6.028
ΔH⁰	(kJ/mol)	40.788
ΔS⁰	(kJ/mol·°K)	0.147

shows that ΔG⁰ was negative and decreased with increasing temperature, implying that the adsorption is spontaneous. The ΔH⁰ and ΔS⁰ were positive, suggesting that the adsorption is endothermic and the increase in the randomness after adsorption, respectively. It is thought that the adsorption is dominated by chemisorption, considering that the ΔH⁰ was over 40 kJ/mol [9,44,60,61].

3.4. Discussion about the Stability and Regeneration

The stability of the HPCs for the adsorptive removal of organic pollutants, such as OTC, may not be important because adsorbents are generally wasted or regenerated after use due to the accumulated pollutants. However, the stability of catalysts is of high importance for the practical applications. The deactivation of catalysts at various degrees has been reported in a number of works found in the literature, for a biochar derived from wood shavings and thiourea–urea [62], Cu@Fe₃O₄ [63], GAC [64], Fe@NS-C-2-12 [65], nitrogen and phosphorus co-doped porous carbon [50], which are mainly attributable to the accumulation of the pollutants, and their byproducts, as well as the changes in the property of the catalysts.

The adsorption capacity of adsorbents generally decreases via the occupation of adsorption sites, the change of surface chemistry, and pore blockage [62,63,65]. It was reported that the adsorption capacity of DCF and VFX onto CS7 was decreased [25], while the adsorption of OTC onto a poplar-leaf-derived biochar was decreased [66], as they were used repeatedly. However, the carbonaceous adsorbent can be successfully regenerated. The adsorption capacity of a biochar was recovered by pyrolysis at 800 °C for 2 h under nitrogen flow [66]. Thermal treatment can also improve the performance by increasing the extent of graphitization, which is advantageous for better structural stability and performance [9]. Therefore, it seems reasonable that carbonaceous adsorbents have a high potential of practical and economic applications.

4. Conclusions

In this study, NaHCO₃-activated WF-based HPCs were prepared under various conditions to investigate the conditions affecting their properties and OTC adsorption. It was demonstrated that the crystallinity of graphitic structures, degree of defects, the number of hopping defects, the stacking graphene-like layers, and pore structures were largely dependent on the activation conditions, while the functional groups were relatively less influenced by them. The activation temperature most significantly affected the adsorption mechanisms and the degree of defects, as evidenced by the increase in the I_D/I_G from 4.528 to 7.959, with increasing temperature from 600 to 900 °C, while the V_meso changed the most significantly from 0.1169 to 0.4382 cm³/g, when the dose of the activator (NaHCO₃) increased from 0 to 100% of the WF mass. In addition, SSA increased greatly from 529.01 to 720.81 m²/g, as the activation time increased from 1 to 3 h.

The best OTC adsorption capacity of 87.184–162.707 mg/g at 298–318 °K was achieved by the HPC prepared at 900 °C for 2 h, at the WF:NaHCO₃ ratio of 1:1 (w:w), i.e., HPC900_R1.0_2H, with the most abundant mesopores, suggesting the detrimental role of mesopores for the OTC adsorption onto HPCs.

The kinetics, equilibrium, and thermodynamics of OTC adsorption onto HPC900_1.0R_2H were good fits to the Elovich equation model and Langmuir isotherm, suggesting that the adsorption is the result of the combination of the diffusion and chemisorption, homogeneous, monolayer, spontaneous, and endothermic. No notable change in adsorption capacity was found under pH of 2–8, but it decreased substantially as the pH continued to increase, implying the crucial role of π–π EDA interactions and electrostatic attraction.

It is strongly suggested by the results in this study, that the HPCs derived from WF could be one of the excellent adsorbents for OTC removal, and that the performance of the HPCs can be controlled by the manipulation of the properties via changing activation conditions, which opens the opportunity of HPCs to a variety of applications.