Chlorphenamine adsorption on commercial activated carbons: Effect of Operating Conditions and Surface Chemistry

Chlorphenamine (CPA) adsorption onto three activated carbons (ACs), namely, Megapol M (MM), Micro 10 (M10), and GAMA B (GB), was studied in this work. The textural properties, concentrations of active sites, surface charge and point of zero charge of the ACs were assessed. The surface areas (SBET) of MM, GB and M10 were 1107, 812 and 766 m2/g, respectively. The MM surface had an acidic character, while the surfaces of M10 and GB were basic. The adsorption capacity of MM, M10, and GB towards CPA was studied at pH 7 and 11, and the adsorption capacity decreased in the order MM > M10 ≈ GB, which was ascribed to the magnitude of SBET and the concentration of acidic sites. The solution pH significantly increased the adsorption capacity of MM towards CPA by raising the solution pH from 5 to 9, and this behavior was attributed to the electrostatic attraction between the negatively charged surface of MM and the cationic species of CPA. The maximum uptake of CPA adsorbed on MM was 574.6 mg/g at pH = 11 and T = 25 °C. The adsorption capacity of MM was slightly raised by incrementing the temperature. Lastly, the zeta potential measurements of pristine MM and MM saturated with CPA confirmed that the electrostatic attraction predominated in the pH range of 5–9, and the π-π stacking interactions were the principal mechanism of CPA adsorption on MM at pH 11.


Introduction
Emerging pollutants, like pharmaceutical and personal care products, in water resources have prompted great awareness because they can affect the environment and human health due to their adverse effects (Geissen et al. 2015;Hena et al. 2021;Tang et al. 2019).Pharmaceutical compounds are vital because they are prescribed to humans and animals for medical purposes.Extensive research has been conducted on the adsorption of various organic compounds on activated carbon; the most common organic compounds studied are dyes (Kannaujiya et al. 2023), pesticides (Mohammad et al. 2020) and pharmaceutical compounds (Carrales-Alvarado et al. 2014).
Chlorphenamine (CPA) is a nonprescription pharmaceutical compound for treating allergic conditions and respiratory infections (Lin et al. 2019).CPA is considered a seasonal drug because of its high concentrations detected during summer (Roberts et al. 2016) and influences human health (Hoang et al. 2022).Although CPA is frequently detected in wastewater treatment plants (WWTPs) effluents, evidence of its effects on aquatic environments is limited (Roberts et al. 2016).CPA has also been identified as a potential precursor of N-nitroso dimethylamine (NDMA), which is hazardous for its carcinogenic potential (Sgroi et al. 2018;White 2019).Hence, removing CPA from water is crucial to preserving the environment and human health.
Responsible Editor: Tito Roberto Cadaval Jr Some processes have been developed for removing CPA from wastewater because wastewater treatment plants (WWTP) are inefficient in removing pharmaceutical compounds.The most frequently studied treatments are advanced oxidation and ozonation.However, these methods can generate more toxic byproducts than the original pollutant.For example, the Fenton degradation of CPA was carried out by a nanoscale zero-valent iron; although CPA was degraded completely, NDMA was detected as a degradation subproduct of CPA (Wang et al. 2016).The formation of NDMA has also been confirmed in the ozonation of CPA, restraining the feasible application of ozonation (Lv et al. 2015).Thereby, alternative technological processes must be utilized to eliminate pollutants efficiently from water without endangering the environment or human health by generating more toxic subproducts.In this regard, adsorption can be adequately applied to CPA removal due to its effectiveness and straightforward operation (Li et al. 2011(Li et al. , 2017;;Ali et al. 2019).
The clay mineral montmorillonite was tested for CPA removal and had an adsorption capacity of 0.72 mmol g −1 (~ 238 mg/g) (Li et al. 2011).The results also showed that cation exchange governed the CPA uptake, and the solution pH did not significantly influence the adsorption capacity of the clay.CPA can be effectively adsorbed on graphene oxide-Fe 3 O 4 composites (Li et al. 2017;Lin et al. 2019), and the CPA adsorption on graphene oxide-Fe 3 O 4 particles was mainly ascribed to electrostatic interactions, resulting in a CPA uptake of 470 mg/g (Li et al. 2017).When a graphene oxide-Fe 3 O 4 suspension was used for CPA removal from wastewater, the CPA uptake from wastewater was improved by deionized water due to ionic strength variations (Lin et al. 2019).Nevertheless, the adsorption and degradation of CPA occurred concurrently on the graphene oxide-Fe 3 O 4 surface, leading to NDMA formation.These studies demonstrated that CPA could be effectively adsorbed on various materials.Up to now, activated carbon (AC) has been the most efficient and cost-effective adsorbent for organic compounds.ACs have outstanding characteristics, like large surface area, sufficient functional groups, suitable pore size distribution and void fraction, and affinity towards organic compounds (Leite et al. 2018).ACs can be produced from various precursors, including wood, coal, peat, lignite and agricultural wastes (Tan et al. 2017).
CPA adsorption on different ACs has been investigated in various works.For instance, the adsorption of CPA on an AC synthesized from Cornulaca-monacantha stems, an agronomical waste, was studied and presented an adsorption capacity of 72.41 mg/g, and the CPA adsorption depended on the solution pH (Sharma et al. 2018).In another study, an AC prepared from date palm leaflets and modified with ethylamine to produce hydrophobic surfaces was used for CPA adsorption (Ali et al. 2019).This modification enhanced the AC adsorption capacity towards CPA (455 mg/g) due to hydrophobic interaction forces controlling the process, highlighting the role of surface chemistry.Similarly, a lignocellulosic granular AC exhibited a maximum uptake for CPA of 193 mg/g (García-Reyes et al. 2021).This AC showed an acidic character due to the predominance of acidic sites that play an essential role in electrostatic interactions.The influence of solution pH was analyzed at different pH (5, 7, and 9).At pH = 9 (neutral CPA), the adsorption capacity was higher than at low pH values.This behavior was attributed to π-π stacking interactions, resulting in a high adsorption capacity.All these studies verified that ACs are suitable for CPA adsorption.
This research aimed to investigate CPA adsorption onto three different commercial ACs.First, the commercial ACs were characterized.The adsorption capacity dependence on the following experimental conditions was studied in detail: solution pH, temperature, and ionic strength.The surface chemistry was examined by Boehm's titration method and was related to AC adsorption capacity towards CPA.Lastly, the adsorption mechanisms were elucidated using zeta potential measurements and determining surface chemical characteristics.

Adsorbents
The following three different commercial ACs were studied: Megapol M (powdered wood-based AC), Micro 10 (granular coconut shell AC), and GAMA B (granular bituminous carbon) labeled as MM, M10 and GB, respectively.The surface nature of the former is acidic, and those of the other two are basic.Carbotecnia, Mexico, supplied the ACs.

Chemical reagents
CPA maleate salt was purchased from Sigma-Aldrich, USA.Table 1 displays the CPA physicochemical characteristics, and its chemical structure and speciation diagram are depicted in Fig. 1a and b).The CPA speciation diagram shows three distinctive CPA species in an aqueous solution.CPA 2+ is the predominant cationic species at pH < 4, CPA + cationic species predominates at pH between 5 and 7, and neutral CPA species prevails at pH > 9 (Lv et al. 2014).

Characterization of activated carbons and saturated with CPA
The textural properties of MM, M10, and GB were evaluated by analyzing the N 2 adsorption-desorption isotherms at 77 K, procured in an N 2 -physisorption analyzer, Micromeritics (ASAP 2020).Previous to the analysis, the samples were degassed in vacuum conditions.Surface area (S BET ) and micropore volume (V mic ) were computed using Brunauer-Emmett-Teller and Dubinin-Radushkevich procedures, correspondingly (Lowell et al. 2004).The Barret-Joyner-Halenda method was implemented to assess the pore size distribution of MM (Barrett et al. 1951), and that of M10 and GB was examined by density functional theory calculations implemented in the physisorption analyzer software.The average pore diameter (D p ) was computed using Eq.(1): where V p is the total pore volume (Rouquerol et al. 2014).
The scanning electron microscope (SEM), JEOL (JSM-6610LV), was utilized to examine the surface morphologies of MM, M10 and GB.The distinctive functional groups of CPA, activated carbons with and without CPA, were identified by FTIR spectroscopy using the attenuated total reflectance technique.The FTIR analysis was accomplished with an FTIR spectrometer, Thermo Scientific, Nicolet iS10. (1) The concentrations of active sites in MM, M10, and GB were quantified by Boehm's titration method (Schönherr et al. 2018) using a previously and thoroughly described procedure (Salazar-Rabago and Leyva-Ramos 2016).In brief, 0.4 g of MM, M10, or GB was placed in a centrifuge tube set in a thermostatic bath for 7 days and agitated for 15 min daily.Correspondingly, the total acidic and basic sites were assessed by titrating using 0.01 N NaOH or HCl solutions.Carboxylic and lactonic sites were titrated using NaHCO 3 and Na 2 CO 3 neutralizing solutions.The concentration of phenolic sites was evaluated by subtracting the summation of the lactonic and carboxylic sites from the total acidic sites.The point of zero charge (pH PZC ) for MM, M10 and GB was measured to assess the electrostatic interactions between CPA species in the water solution and the AC surface.Potentiometric curves were obtained as follows.In brief, 23 neutralizing solutions were fixed by mixing varying aliquots of 0.1 N HCl or NaOH in volumetric flasks (50 mL) and filling up to the mark using 0.1 N NaCl solution.Afterward, 0.05 g of MM, M10, or GB was added to 50 mL centrifuge  tubes, followed by supplementing 40 mL of the neutralizing solutions.The remaining volumes of the solution were used as blank.The tubes were placed in a thermostatic bath for 7 days and stirred every day with an orbital shaker.The final pH of blank solutions and solutions with the ACs was plotted against the corresponding volume of the titrating solution.Surface charge was computed with a previously reported relationship (Leyva-Ramos et al. 2011).
The zeta potential of pristine MM and MM saturated with CPA (MM-CPA) in water solution was determined using a Malvern instrument (Zetasizer model).First, MM dispersions with 0.1 g/20 mL were prepared by mixing the MM with constant ionic strength solutions (0.01 N) in a pH range of 2-11.The pH values were maintained constant for 7 days by supplementing daily drops of 0.01 N HCl and NaOH.The supplemented volume consistently accounted for less than 2.0% of the initial volume, so it was considered constant.The tubes were placed in a thermostatic water bath at 25 °C and agitated with an orbital shaker.The zeta potential of MM-CPA was determined at pH of 3, 5, 7, 9 and 11, and various initial CPA concentrations (100, 500, and 750 mg/L).After six days, aliquots of 1 mL were transferred to the sample cuvette to determine the zeta potential of pristine MM and MM-CPA.The mass of CPA adsorbed on MM was also assessed, as described later.

Quantification of CPA in aqueous solution
The CPA concentration in an aqueous sample was quantified by UV-Vis spectrophotometry.The sample absorbance was appraised employing a UV-Vis spectrophotometer, Shimadzu 2600, at a particular wavelength of 261 nm.The CPA concentration was computed from calibration curves fixed using CPA standard solutions having concentrations varying from 0.5 to 50 mg/L and at solution pH of 5, 7, 9 and 11.

Adsorption equilibrium experiments
A CPA stock solution (1500 mg/L) was made by dissolving a specific amount of CPA maleate (2.133 g) with 0.01 N constant ionic strength solutions and pH adjusted to 5, 7, 9, and 11 by combining suitable volumes of 0.01 N HCl and NaOH solutions.
CPA adsorption experiments were conducted in batch adsorbers (50 mL centrifuge tubes).CPA solutions with initial concentrations ranging from 50 to 1200 mg/L were prepared in volumetric flasks (50 mL) by adding an aliquot of CPA stock solution and diluting using constant ionic strength solutions with a specific pH.Afterward, 0.05 g of MM, M10, or GB was placed into the batch adsorbers, followed by adding 40 mL of the CPA solutions (50-1200 mg/L).The batch adsorbers were set up in a thermostatic water bath under a constant temperature for 7 days and agitated daily for 15 min.The pH was recorded daily and maintained at its initial value by supplementing drops of 0.01 N HCl or NaOH solutions.The added volume was always less than 2.0% of the initial solution, so the initial volume was considered constant.
Once equilibrium was reached within 7 days, the dispersions in the adsorber were centrifuged at 2500 rpm for 15 min.Afterward, the sample was filtered to obtain 10 mL aliquots and analyzed to quantify the CPA concentration at equilibrium.Lastly, the uptake adsorbed of CPA was appraised with the ensuing equation: where C e and C 0 stand for equilibrium and initial concentrations of CPA (mg/L), respectively; m denotes the amount of AC (g); q represents the mass adsorbed of CPA (mg/g); and V designates the CPA solution volume inside the adsorber (L).

Activated carbons properties
Table 2 summarizes the textural properties of MM, M10, and GB.Additionally, Figure S.1 depicts the N 2 adsorption-desorption isotherms of MM, M10 and GB.Under IUPAC classification, M10 and GB show a type Ib isotherm, and MM exhibits a type IIb isotherm.The type Ib isotherm of M10 and GB is reversible and typical of microporous adsorbents and shows a very narrow type H4 hysteresis loop.The isotherm type IIb of MM presents a hysteresis loop type H4 related to the filling of slit-shaped pores in mainly microporous activated carbons (Rouquerol et al. 2014;Thommes et al. 2015).
The S BET of MM, GB and M10 were 1107, 812 and 766 m 2 g −1 , respectively.The micropore volume (V mic ) contributed significantly to the total pore volume, comprising 38%, 70%, and 71% for MM, M10 and GB, respectively.Regarding the mean pore diameter (d p ) and mean micropore width (L 0 ), the diminishing order is MM > M10 > GB and MM > GB > M10, respectively.This finding indicated that CPA was not restricted to diffuse freely inside the pores due to its molecular dimensions.In addition, the area of micropores (S mic , m 2 /g) had a significant contribution to the surface area in MM (47%), M10 (86%), and GB (71%).
Figure S.2 depicts the pore size distribution and cumulative pore volume percentage for MM, M10, and GB.As shown in Figure S.2(a), MM was mainly composed of micropores of 1.24 nm size and with some narrow and wide mesopores, and its micropore volume accounted for 82% of its total pore volume.M10 mainly comprised micropores and a few narrow mesopores, and the percentage of its Figure 2 shows that the pH PZC of MM is 2.9, indicating the acidic character of its surface.In other words, the MM surface was positively charged at pH < 2.9 since its basic sites were protonated but negatively charged at pH > 2.9 because its acidic sites were deprotonated.MM showed a substantial decrease in surface charge, increasing pH.However, this behavior was only observed up to pH 7.
Meanwhile, M10 and GB had pH PZC of 9.5 and 10.2, revealing the basic character of their surface.
Table 3 exhibits the active sites of MM, M10, and GB quantified by Boehm's titration method.The total acidic sites of MM (0.626 meq/g) were 313-fold larger than those of the total basic sites (0.002 meq/g).The predominant acidic sites were the carboxylic ones (0.456 meq/g), and the contribution of lactonic and phenolic sites was 0.086 and 0.083 meq/g, respectively.These results suggested that MM has an acidic surface character, which concurred with its pH PZC .For M10, the total basic sites were 2.98-fold more than the acidic ones.The concentrations of carboxylic, lactonic, and phenolic sites were correspondingly 0.002, 0.071, and 0.099 meq/g.For GB, the total basic sites (0.294 meq/g) were 2.67 times more than the acidic ones (0.110 meq/g).The concentrations of carboxylic and lactonic sites were 0.090 and 0.020 meq/g, and phenolic sites were undetected.The predominance of total basic sites in M10 and GB confirmed that these adsorbents had a basic surface character consistent with their pH PZC .
The SEM images of MM, M10, and GB are depicted in Figures S.3   (Masood et al. 2020).The CPA spectrum exhibits 3 characteristic peaks at 1583 cm −1 and 1469 cm −1 for stretching C = C aromatic and at 1352 cm −1 for bending the C-N bond (Jelvehgari et al. 2014).Besides, the signal at 3014 cm −1 corresponds to C-H aromatic, 2968 cm −1 to C-H aliphatic, 1702 cm −1 to C = N, and 1149 and 1088 cm −1 to C-N stretching vibrations.The bands at 850-800 cm −1 are assigned to the bending out plane of C-H aromatic (Masood et al. 2020).
FTIR is an analytical technique used to identify the functional groups present in organic molecules when they are subjected to infrared radiation.Under these conditions, the bonds lengthen, shorten or flex, causing absorptions, which have different intensities depending on the polarity of the bond and the number of bonds responsible for the absorption.If the CPA molecule is adsorbed on the surface of the ACs, the CPA molecule loses mobility and therefore, the definition of the spectrum is lost, and there is no longer differentiation between the signals, obtaining only a continuous band in the range from 1570 to 1160 cm −1 .This result confirms that CPA is adsorbed on the ACs.

Adsorption isotherms of CPA on the activated carbons
The Freundlich, Langmuir, and Radke-Prausnitz (R-P) isotherms represented by the succeeding relationships (Terdputtakun et al. 2017) interpreted the adsorption equilibrium data of CPA on the ACs: where k F (mg 1−1/n L 1/n /g) represents the Freundlich isotherm constant; n is an exponent associated with the adsorption intensity; q m (mg/g) refers to the maximum adsorption capacity; K L (L/mg) represents the Langmuir isotherm parameter; and a (L/g), b (L 1−β /mg 1−β ), and β are constants of the R-P equation.
The preceding parameters were optimized by a nonlinear least-square procedure and the Levenberg-Marquardt algorithm implemented in Origin software.The average percentage deviation was then estimated by the succeeding equation: where N designates the number of experimental data, q i,pred (mg/g) represents the uptake adsorbed of CPA estimated with the adsorption isotherms, and q i,exp (mg/g) denotes the experimental uptake adsorbed of CPA.
The constants for Freundlich, Langmuir and R-P isotherms and %D are included in Table 4, and the model best interpreting the experimental data was selected as the isotherm model with the lowest %D.The %D values indicated that the R-P, Langmuir and Freundlich isotherms resulted in lower %D for 6, 3 and 1 out of 10 experimental conditions, respectively.Thus, the R-P model better fitted the adsorption equilibrium data and was chosen to represent the adsorption data of CPA in this work.

Comparison of capacities of ACs for adsorbing CPA
At pH = 7 and 25 °C, the adsorption isotherms of CPA on MM, M10 and GB are depicted in Fig. 3(a), and the cationic CPA 1+ is the predominant species at pH = 7.For an equilibrium concentration of CPA of 450 mg/L, the uptake of CPA adsorbed (Q 450 ) on MM, M10 and GB was 336.5, 214.2 and 206.5 mg/g, correspondingly.The MM exhibited the highest adsorption capacity toward CPA, which was 1.57-and 1.63fold larger than that of M10 and GB, respectively.
The variation in the adsorption capacity of the ACs could be ascribed to the S BET and the concentration of acidic sites on their surfaces where CPA adsorption occurred (Carrales-Alvarado et al. 2014;García-Reyes et al. 2021).The concentration of acidic sites diminished as follows: MM > M10 > GB; such a trend coincided with their adsorption capacity.In addition, the adsorption capacity of MM was favored at pH 7 because there was electrostatic attraction between CPA 1+ in the solution and the negatively charged surface of MM.In contrast, the M10 and GB had positively charged surfaces, which led to electrostatic repulsion towards CPA 1+ .Figure 3(b) illustrates the adsorption isotherm of CPA on MM, M10 and GB at pH 11, but at this pH, all ACs had negatively charged surfaces, and the predominant CPA species is neutral, so there is no × 100 % electrostatic attraction.For a CPA equilibrium concentration of 450 mg/L, the mass of CPA adsorbed (Q 450 ) on MM, M10 and GB was 574.6, 208.2 and 207.1 mg/g, respectively.The MM showed the largest adsorption capacity, and M10 and GB had similar adsorption capacities.
The surface area occupied by molecules of CPA adsorbed at pH = 11 was computed with the succeeding mathematical relationship (Carrales-Alvarado et al. 2020): where A p represents the projected area of a CPA molecule ( 7.78 × 10 −19 m 2 molecule −1 ), M CPA is the CPA molecular weight (g mol −1 ), and N A denotes the Avogadro number ( 6.022 × 10 23 molecules mol −1 ).The projected area was assessed by multiplying the X and Y dimensions listed in Table 1.In addition, the percentage of surface area occupied by adsorbed CPA molecules was calculated subsequently: The results of S Oc, CPA and %S are displayed in Table 5.The area occupied by CPA adsorbed decreased in the subsequent order MM > M10 ≈ GB, with values ranging from 88 to 44%.In the MM, more S BET was occupied than in the other ACS because the MM presented a larger D p , so the S BET of MM was more accessible to the CPA molecules.

Effect of solution pH on the adsorption capacity of MM towards CPA
The dependence of the adsorption capacity of MM on the solution pH was investigated because MM presented the highest capacity towards CPA. Figure 4 displays the adsorption isotherms   of CPA at pH 5, 7, 9 and 11, noticing that the adsorption capacity increased by raising the pH.The uptake Q 450 on MM was 230.9, 336.5, 504.2 and 574.6 mg/g at pH 5,7, 9 and 11, respectively.These results revealed that the adsorption capacity of MM increased 2.5-fold, incrementing the pH from 5 to 11.
In the pH range of 5-9, the adsorption capacity increased 2.2 times.In agreement with the CPA speciation diagram (Fig. 1(b)), the distribution of CPA 1+ was 90% and 100% at pH 5 and 7; hence the predominant species is CPA 1+ .Besides, the MM surface was negatively charged due to the deprotonation of carboxylic groups, suggesting that electrostatic interactions govern the adsorption.The adsorption capacity incremented because the surface charge became more negative when the pH increased from 5 to 9.However, at pH = 9, the distribution of CPA species is CPA 1+ (60%) and neutral CPA (40%), implying that adsorption is due to electrostatic attraction and other mechanisms.
For concentrations of CPA at equilibrium less than 50 mg/L, the pH solution affected the adsorption capacity of carbon MM, and its capacity decreased concerning pH in the following order: pH = 9 > pH = 7 > pH = 11 > pH = 5.The maximum adsorption capacity of MM was noted at pH 11 for CPA concentrations at equilibrium higher than 120 mg/L, and the highest CPA uptake of 574.6 mg/g was achieved.At this pH, the predominant species is the neutral CPA, and the surface charge of MM is negative (See Fig. 1); therefore, no electrostatic interactions occurred between CPA and the MM surface at pH 11, and the enhancement of adsorption capacity was due to other mechanisms.One possible explanation for this phenomenon is that π-π stacking interactions became the principal driving force for CPA adsorption, resulting in a high adsorption capacity.

Effect of temperature on the capacity of MM for adsorbing CPA
Figure 5 presents the dependence of the adsorption of CPA on MM upon the temperature, and the findings show a slight enhancement in the adsorption capacity by increasing temperature.The adsorption capacity of MM at 25 °C increased by 2.20% at 35 °C but declined by 9.35% at 15 °C; the temperature did not noticeably influence the adsorption capacity.This behavior implied that the adsorption of CPA on MM is endothermic.Furthermore, the isosteric heat of adsorption was computed employing the next mathematical relationship (Do 1998): where R designates the ideal gas constant (8.314J mol −1 K −1 ); C e1 and C e2 (mg/L) denote the CPA equilibrium concentrations for temperatures T 1 and T 2 (K), respectively, and are at the same mass of CPA adsorbed (mg/g); and ΔH ads q (J/mol) (10)  represents the isosteric heat of adsorption.The value of ΔH ads q corroborates whether the adsorption occurs through physical (< 40 kJ/mol) or chemical (> 40 kJ/mol) interactions (Do 1998;Ortiz-Ramos et al. 2022).

Fig
The following data were used for calculating ΔH ads q : a con- stant mass of CPA adsorbed (q) of around 339 mg/g; CPA equilibrium concentrations C e1 = 354.35mg/L and C e2 = 313.44mg/L for temperatures T 1 = 298.15K and T 2 = 308.15K.The ΔH ads q was 9.4 kJ/mol, proving that the CPA adsorption was endothermic and occurred by physical interactions.

Effect of Operating Conditions on the Removal of CPA by Adsorption on MM
The removal percentage of CPA from a water solution by adsorption on carbon MM can be estimated using the subsequent mathematical expression: The C e was computed by solving simultaneously the mass balance represented by Eq. ( 2) and the experimental R-P adsorption isotherm model, Eq. ( 5).
Figure 6 depicts the effect of the solution pH upon the removal percentage of CPA for a carbon dosage of 1.25 g/L and various initial concentrations of CPA (C 0 ).Except for C 0 = 800 mg/L, the optimal pH for removing CPA is pH = 9 since the maximum values of %R were achieved at this pH, and the %R ranged from 92 to 99%, whereas the %R was 69% for C 0 = 800 mg/L.This result can be explained by recalling that at a CPA equilibrium concentration below 120 mg/L, the adsorption capacity of the MM is higher at pH = 9 than at pH = 11, as shown in Fig. 6(a).
The pH of 9 was chosen to investigate the carbon dosage (V/m) on the %R because the carbon MM exhibited the maximum %R at this pH.The effect of MM dosage upon the %R is shown in Fig. 6(b), and as expected, the %R increased by raising the MM dosage because more adsorption sites are available by incrementing the mass of MM.The %R increased slightly by increasing the MM dosage for C 0 of 100 and 200 mg/L.For example, the %R were 98.1, 99.5 and 99.8% for C O = 200 mg/L and m/V of 1.25, 2.5 and 5 g/L, respectively.Thus, the satisfactory dosage is 1.25 g/L for C 0 ≤ 200 mg/L since less mass of MM carbon was used.Besides, for C 0 CPA between 200 and 800 mg/L, a satisfactory MM dosage was 2.5 g/L because the %R varied from 95 to 99%.

CPA adsorption mechanisms on MM
The results in "Comparison of capacities of ACs for adsorbing CPA" section proved that CPA adsorption on MM (9) % R = C 0 − C e C 0 × 100 % depended significantly on solution pH.Different interactions were responsible for CPA adsorption.For instance, in the pH range of 5-7, electrostatic attractions governed the adsorption because the acidic sites were deprotonated, resulting in a negatively charged surface of MM, and CPA 1+ species prevailed in the water solution.Nonetheless, the adsorption capacity increased in the pH range of 9-11, but electrostatic interactions no longer supported this behavior.Therefore, π-π stacking is an essential mechanism in CPA adsorption at pH = 11.Zeta potential measurements were performed to corroborate that electrostatic attraction and π-π interactions are the principal mechanisms for CPA adsorption.Figure 7 shows that MM had an isoelectric point (pH IEP ) around 2.8, and its zeta potential diminished with the decreasing pH; such behavior was also observed for the surface charge of MM (See Fig. 2).The zeta potential of MM-CPA with varying initial concentrations of CPA (100, 500, and 750 mg/L) was measured at pH 5, 7, 9 and 11.It is worth mentioning that the mass of CPA adsorbed evaluated for MM-CPA was augmented by raising the initial concentration.The results indicated that when the CPA initial concentration increased, the zeta potential became less negative and even positive because the CPA adsorbed molecules balanced the negatively charged surface of MM.This tendency was observed at pH 5, 7 and 9 but not at pH 11, at which the zeta potentials of MM-CPA became slightly more negative than that of MM.At pH 11, the interactions between the π electrons of the CPA molecule and π electrons of MM basal planes were fortified because many acidic sites were neutralized by the oxygen-containing functional groups that can extricate π electrons from the MM basal planes, causing a decrease in π-π stacking interactions (Salam and Burk 2008;Salame and Bandosz 2003).
The change in the zeta potential, ΔZP, was computed with the equation given below to analyze the variations in the zeta potential regarding the mass of CPA adsorbed: where (ZP) MM-CPA is the zeta potential of MM-CPA (mV), and (ZP) MM is the zeta potential of pristine MM (mV).The ΔZP can be positive or negative when the ZP is reduced or increased by the adsorption of CPA, respectively.The ΔZP vs. mass of CPA adsorbed graph is depicted in Fig. 8.At pH 5 and 7, ΔZP increased drastically and almost linearly while raising the mass of CPA adsorbed so that the adsorption of CPA 1+ balanced the negative charge of MM.At pH = 9, ΔZP increased less pronouncedly than at pH 5-7, but not linearly.Although ΔZP started to decline at pH 9, it still had a considerable value that may indicate possible electrostatic interactions.At pH 11, ΔZP was negative, showing a reduction with growing q, opposite to the trend observed in the pH range of 5-9.Therefore, electrostatic interactions have no contribution at pH = 11, and π-π stacking interactions are the only mechanism involved in CPA adsorption.

Conclusions
CPA adsorption onto the commercial ACs MM, M10 and GB with acidic and basic character surfaces was investigated.First, the textural properties of MM, M10 and GB were obtained.All three adsorbents were micro-mesoporous ACs with large surface areas (1107, 766, and 812 m2 g −1 ).The pH PCZ was 2.9, 9.5 and 10.2 for MM, M10 and GB, correspondingly.Quantification of active sites indicated that the surface character of MM is acidic, and those of M10 and GB are basic, agreeing with their pH PZC values.SEM micrographs revealed the irregular morphology of all three ACs.
The adsorption capacities of the three ACS were compared at pH = 7 and MM presented the highest CPA adsorption capacity.This high capacity was attributed to the large (10) ΔZP =(ZP) MM-CPA − (ZP) MM number of acidic sites of MM and the pH that favor CPA adsorption according to the CPA speciation diagram.The influence of pH on CPA adsorption by MM with the highest adsorption capacity was then analyzed.The results revealed that pH had a major influence on the adsorption capacity, with pH 11 highly favoring CPA adsorption (564.4 mg/g).The temperature affected lightly the adsorption capacity.However, the isosteric heat of adsorption indicated that CPA adsorption on MM was endothermic and occurred through physical interactions.Finally, zeta potential measurements demonstrated that π-π stacking was the main mechanism of CPA adsorption on MM at pH = 11.

Fig. 1
Fig. 1 Chlorphenamine chemical structure (a) and speciation diagram (b) C e micropore volume was 82% (Figure S.2(b)).GB was mainly composed of micropores and some narrow mesopores, with a micropore volume percentage of 79%, as shown in Figure S.2 (c).
(a), (b), and (c).The three ACs exhibited similar morphology with irregular edges.MM comprised small granules, whereas M10 and GB were composed of large granules and some cavities.The FTIR spectra of CPA, MM, MM-CPA, M10, M10-CPA, GB, and GB-CPA are shown in Figure S.4.Regarding the FTIR spectra of activated carbons, the broad band centered at 3375 cm −1 represents the O-H stretching vibration(Heidari et al. 2014;Zheng et al. 2016), and the band at 2354 cm −1 indicates the O = C = O asymmetric vibration of the CO 2 molecule adsorbed(Zheng et al. 2016).The peak

Fig. 3
Fig. 3 Comparison of the adsorption capacity of M10 and GB towards CPA at pH = 7, T 25 °C, I = 0.01 N (a) and pH = 11, T = 25 °C, I = 0.01 N (b).The lines represent the Radke-Prausnitz isotherm Fig. 4 Effect of solution pH on the CPA adsorption capacity of MM at T = 25 °C and I = 0.01 N. The lines represent the Radke-Prausnitz isotherm

Fig. 6
Fig. 6 Removal percentages at different initial concentrations at T = 25 °C.Effect of pH at m/V = 1.25 g/L (a).Effect of AC dosage at pH = 9 (b)

Fig. 7 Fig.
Fig. 7 Zeta potential of pristine MM in the pH range of 2-11 and MM-CPA regarding the CPA initial concentration at pH of 3, 5, 7, 9, and 11

Table 1
Physicochemical properties of CPA

Table 2
Stoeckli et al. 2001)) the activated carbons a Total pore volume determined at P/P 0 ≈ 0.95 b Mean pore diameter evaluated by Eq. (1) c Micropore volume calculated by the Dubinin-Radushkevich method(Lowell et al. 2004)d Mean micropore width estimated with the Stoeckli equation(Stoeckli et al. 2001)e Micropore surface area calculated by the Stoeckli equationStoeckli et al. 2001)

Table 4
Parameters of the adsorption isotherms and their corresponding average percentage deviation.Ionic strength = 0.01 N

Table 5
Percentage of surface area occupied by CPA adsorbed