Chemical-activated carbons from peach stones for the adsorption of emerging contaminants in aqueous solutions

Chemical Engineering Journal 279 (2015) 788–798

Contents lists available at ScienceDirect

Chemical Engineering Journal journal homepage: www.elsevier.com/locate/cej

Chemical-activated carbons from peach stones for the adsorption of emerging contaminants in aqueous solutions Silvia Álvarez Torrellas a,b, Rafael García Lovera b, Néstor Escalona b, Catherine Sepúlveda b, José Luis Sotelo a, Juan García a,⇑ a Grupo de Catálisis y Procesos de Separación (CyPS), Departamento de Ingeniería Química, Facultad de Ciencias Químicas, Universidad Complutense de Madrid, Avda. Complutense s/n, 28040 Madrid, Spain b Grupo de Materiales Microporosos en Catálisis y Adsorción, Departamento de Físico-Química, Facultad de Ciencias Químicas, Universidad de Concepción, Edmundo Larenas 129, Concepción, Región del Biobío, Chile

h i g h l i g h t s

g r a p h i c a l a b s t r a c t

 Chemical activation and

a r t i c l e

i n f o

Article history: Received 8 April 2015 Received in revised form 26 May 2015 Accepted 27 May 2015 Available online 1 June 2015 Keywords: Adsorption Chemical activation Fixed-bed column Mesoporous carbons Emerging contaminants

360 Adsorption capacity (mg/g)

characterization of raw peach stones based-activated carbons.  Chemical and thermal treatments in order to modify the chemical surface carbons.  Batch and dynamic adsorption tests of several emerging contaminants were performed.  Competitive effect occurring between adsorbate and water for the active sites.  TPD studies played an important role in the interpretation of chemical surface groups.

320

PC He-PC

280 240

O-PC

200 160 120 0.5

1.0 1.5 2.0 2.5 3.0 Acid sites (mmol/g carbon)

3.5

a b s t r a c t Activated carbons from peach stones by chemical activation with H3PO4(s) have been prepared, leading to high surface area and well-developed mesoporous materials. Further oxidation and gas phase treatments were applied in order to analyze the influence of the chemical surface groups on the adsorption behavior. The adsorption of three emerging compounds: a stimulant (caffeine), an anti-inflammatory drug (diclofenac) and a psychiatric drug (carbamazepine) from ultrapure water through batch and dynamic tests was investigated. The characterization of the adsorbents was conducted by N2 adsorption–desorption isotherms, scanning electron microscopy (SEM), infrared spectrometry (FTIR), thermal programmed decomposition (TPD), zero charge and isoelectric points (pHZPC, pHIEP) determination and the evaluation of the total acidity by potentiometric titration with n-butylamine. TPD profiles and FTIR spectra revealed the presence of carboxylic, phenolic, carbonyl and quinonic functionalities in the carbonaceous surfaces. S-type adsorption isotherms, as Giles classification, were obtained, indicating a competitive effect between aqueous solution and the target adsorbates. Carbamazepine adsorption capacity was higher than caffeine and diclofenac, reaching 335 mg/g, attributed to its hydrophobic character and water solubility properties. The oxidation of the activated carbon greatly enhanced the hydrophilic character of the material, decreasing the adsorption capacity and highly affecting on the breakthrough times and adsorption capacity values in the fixed-bed adsorption process. Ó 2015 Elsevier B.V. All rights reserved.

⇑ Corresponding author. Tel.: +56 220 4324, +34 91 394 5207. E-mail addresses: [email protected], [email protected] (J. García). http://dx.doi.org/10.1016/j.cej.2015.05.104 1385-8947/Ó 2015 Elsevier B.V. All rights reserved.

Caffeine Diclofenac Carbamazepine

S.Á. Torrellas et al. / Chemical Engineering Journal 279 (2015) 788–798

1. Introduction Activated carbon is a carbon-based material which is not truly amorphous but has a microcrystalline structure. These materials have been used as excellent and versatile adsorbents in several gas and liquid phase applications due to their highly developed porosity and extended apparent surface area. The volume of the pores in activated carbons is generally defined as being higher than 0.2 cm3/g, and the internal surface area is larger than 400 m2/g as measured by Brunauer–Emmett–Teller (BET) equation. The physicochemical properties of the material and subsequent adsorbent behavior highly depend on the nature of the raw material used, the activating agent and the conditions of the carbonization and activation processes [1]. Most organic materials rich in carbon, mainly coals, wood [2], coconut shell, peat [3], agricultural by-products such as fruit stones [4], seeds hulls [5], straw and stalks [6], lignite, coal, petroleum, coke, etc. can be used as precursor materials for activated carbon preparation. The selection of the material is based mainly on (i) low in inorganic matter; (ii) availability and cost; (iii) low degradation upon storage and (iv) ease of activation. The most common processes of activation are physical and chemical. The physical activation involves two steps: (i) carbonization and (ii) the control gasification of a carbonaceous precursor using a gas stream (O2, CO2, steam water, etc.) at high temperatures. In the chemical activation process, the raw material is impregnated using a chemical agent (H3PO4, ZnCl2, H2SO4, H3BO3, KOH, NaOH, etc. solutions) and the resulting solid is activated at lower temperatures than those used in the physical methodology [7,8]. Some previous works report several conditions using different precursors and chemical activating agents [9,10]. Lignocellulosic materials constitute the more commonly used precursors (around 45 wt% of the total raw materials used for the manufacture of activated carbons). Low contents of inorganic materials are important in the preparation of activated carbons with low ash content, but relatively high volatile content is also needed for the control of the carbon porosity. Both characteristics are common to most of lignocellulosic materials used for the production of activated carbons. These cellulose-type materials such as wood, sawdust, nutshells and fruit stones are mainly used in the chemical activation methodology [7]. In the chemical activation the porosity is generated through dehydration reactions in the carbonaceous structure [7]. Earlier studies have shown that chemical activation using phosphoric acid solutions at moderate temperatures generates, in some lignocellulosic materials, a high surface area and a balanced degree of micro and mesoporosity percentages, useful for a wide range of applications as adsorbents. The development of surface area appears to be greatly dependent on the subsequent heat treatment temperature [11]. The authors have selected peach stones as precursor material due to their availability and desirable physical characteristics as activated carbon precursor. Carbon-oxygen surface groups are the most important surface groups that influence the surface characteristics such as the wettability, polarity, and acidity, and the physico-chemical properties such as catalytic, electrical, and chemical reactivity of the activated carbons. In chemical activation the carbon material develops oxygenated functional groups (more reactive), which has an important role on the adsorption capacity of water and other polar compounds [12]. Therefore, these surface groups could be modified by chemical and/or thermal treatments in order to improve the adsorption properties. Emerging contaminants of concern are new substances detected in waste waters at low concentration levels, ranging from lg/L to ng/L. The results of advanced water analysis have revealed

789

the presence of pharmaceuticals, personal care products (PPCPs), endocrine disrupting compounds, polybrominated flame retardants (PBDEs), perfluorinated compounds (PFCs), etc. even in drinking waters. Some representative compounds of these categories are the well-known stimulant substance, caffeine, an alkaloid occurring in more than 60 plant species and whose consumption is very regular over the world, about being the global data 70 mg per person per day. Caffeine is a compound detected in wastewater, surface water, and groundwater worldwide [13]. Therefore, there is evidence that pharmaceutical compounds can reach detectable concentrations in wastewaters due to they are not completely degraded after consumption. Diclofenac, a popular analgesic belonging to the group of the non-steroidal anti-inflammatory drugs, has been frequently identified in effluents from domestic wastewater treatment plants and in rivers [14]. Carbamazepine, an antiepileptic drug, is one of the most frequently detected pharmaceutical residues in water bodies. This drug, together with caffeine, has been proposed as an anthropogenic marker in water streams in numerous studies [15]. Activated carbon adsorption has revealed worldwide as effective in the removal of organic compounds at the concentration range in they are present in waste waters, no generating secondary by-products which can be more harmful than the original compounds. In addition, it has no high energy costs associated and it is easy of operate [16,17]. Numerous studies have evaluated the efficiency of adsorption technique in the removal of emerging contaminants, such as pharmaceuticals and endocrine disruptors, onto activated carbons in ultrapure water and in the presence of natural organic matter [18]. One of our goals has been to develop a granular activated carbon using peach stones as a starting material by acid phosphoric solution-activation, obtaining a high surface area and mesoporosity activated carbon for the removal of several polar emerging compounds, such as caffeine, diclofenac and carbamazepine. Another goal has also been to study the overall changes on the carbon surface after an oxidation treatment with HNO3 and a thermal treatment under inert atmosphere, studying the influence of these surface modifications on the aqueous phase adsorption properties. Therefore, from our knowledge, this is the first work testing the role of the external media, i.e. atmosphere on the regeneration of oxygen-surface functionalities in the modified activated carbons. 2. Materials and methods 2.1. Materials The target compounds, caffeine, diclofenac and carbamazepine, were purchased from Sigma to Aldrich (Steinheim, Germany), with a purity higher than 98% and used as received in the experiments. Ortho-phosphoric acid (85 wt%) was purchased from Panreac and nitric acid (69.5 wt%) was obtained from Carlo Erba. The working solutions were obtained by diluting a stock solution previously prepared using ultrapure water. The physicochemical properties of the compounds are shown in Table 1. The molecular structures of the substances are depicted in Fig. S1 (Supplementary Material). In the dynamic experiments, before column packing, the activated carbons were sieved at 0.5–0.589 mm and washed with boiling water to remove the impurities into the pores and finally they were dried in oven at 110 °C for 24 h. 2.2. Preparation of the activated carbon The activated carbon was prepared by chemical activation using phosphoric acid as activating agent. The precursor material, peach

790

S.Á. Torrellas et al. / Chemical Engineering Journal 279 (2015) 788–798

Table 1 Physico-chemical properties of the studied emerging compounds. Compound

Molecular weight (g/mol)

Log Kow

pKa

Water solubility (mg/L)

Molecular size (nm)

Caffeine

194.2

0.98  0.87  0.56

318.1 236.3

10.4 0.6 4.15 13.9

21600

Diclofenac Carbamazepine

0.07 (exp) 0.16 (calc) 0.70 2.45

2430 17.7

0.97  0.96 1.2  0.92  0.58

stones, was crushed and sieved to get a particle size ranging from 0.883 to 1.0 mm. The first step, chemical impregnation, was carried out in a round-bottom flask reactor, where 60 g of precursor reacted with a H3PO4 solution (12 mol/L) at 85 °C for 6 h. After impregnation, the solid were filtered under vacuum to remove the excess of phosphoric acid and calcined under an air flow (50 cm3/min) at 435 °C during 4 h, defining a heating ramp of 5 °C/min. The resulting carbon was thoroughly washed with ultrapure water in order to remove the remaining phosphoric acid until reach pH 6.5. Finally, the solid was dried in oven at 110 °C for 24 h. The resulting material was named as PC activated carbon. 2.3. Modification of the activated carbons In order to enhance the quantity of oxygenated surface functional groups, the original PC material was modified by a liquid phase oxidation treatment reported in other previous works [19]: the solid was kept in contact to a HNO3 solution (6 mol/L) for 1 h at 80 °C in a 250 mL round-bottom flask, using a ratio of 10 mL of acid solution/1.0 g of carbon. The recovered solids was washed with deionized water until neutrality and further dried in oven for 24 h at 110 °C, resulting in O-PC material. With the objective of decreasing the quantity of oxygenated surface functional groups, a gas phase thermal treatment was carried out; the original solid was treated under a helium flow (50 cm3/min) at 875 °C during 5 min, resulting in the He-PC material.

Acid site concentrations and acid strength measurements of the activated carbons were determined using a potentiometric method [20], where the dried solid (0.15 g) was suspended in 90 mL acetonitrile and agitated for 3 h. Further, the suspension was titrated with a 0.1 M n-butylamine solution in acetonitrile at 0.05 mL/2 min. The variation in electric potential was registered on a Denver Instrument UltraBasic pH-mV meter. According to this procedure the initial electric potential is related to the strength of the acid sites as well as the number of acid sites is determined by the amount of base added until obtain the plateau in the titration curves. Isoelectric point (pHIEP) of the materials was determined through electrophoretic migration measurements using a Zeta-MeterPlus 3.0 equipment. For this characterization a suspension of 0.025 g of sample in 200 mL ultrapure water was used. The pH solution was adjusted with either 0.1 M HCl or KOH solutions. The isoelectric point was calculated by the graphical representation of zeta potential versus pH solution. The point of zero charge (pHPZC) was calculated using the mass titration method proposed by Noh and Schwarz [21]. This method is based on the fact that pH approaches the limiting value (pH1) by addition of solid powder to an aqueous medium. This limiting pH1 value, at a high solid content, is then equal to pHPZC. Experimentally various suspensions of dried carbon (with weights ranging from 0.1 to 1.0 g) in 40 mL distilled hot water were prepared in a flask. The boiling was controlled by a thermometer to ensure that no false boiling, due to trapped gases, occurred. After boiling, the solid was immediately filtered and pH was measured at 50 ± 5 °C.

2.4. Characterization of the activated carbons

2.5. Adsorption experiments

The textural characterization of the activated carbons was carried out by N2 adsorption–desorption isotherms at 196 °C using a TRISTAR Micromeritics apparatus (Micromeritics, Norcross, Atlanta, USA). 0.1 g of sample, previously degassed at 250 °C for 3 h under nitrogen flow, was used. The micropore volume (V0) was obtained using Dubinin–Radushkevich equation. The mesopore volume (Vm) was calculated by subtracting the total pore volume (Vpore), obtained at 0.95 relative pressure value, and the micropore volume. The morphological studies were carried out by scanning electronic microscopy (SEM) technique. The studies were developed using a JEOL JSM 6400 microscope equipped with a thermoionic cathode (tungsten filament) and 25 kV detector. Surface chemical characterization was deeply analyzed by several procedures in order to identify the different nature of oxygenated surface functional groups. The Fourier transformed infrared spectrums in the range 4000–400 cm1 were recorded using a Thermo Nicolet FT-IR spectrophotometer. KBr-activated carbon pellets were prepared using 0.1 wt% of activated carbon. The pellets were dried under an infrared lamp before recording the spectra. The thermal programmed decomposition (TPD) essays were carried out in a quartz reactor heated by an electric furnace under helium flow (50 cm3/min) as inert carrier gas, establishing a heating range from 25 °C to 1050 °C (with a heating rate of 10.25 °C/min). The gases evolved during the thermal treatment were detected by a thermal conductivity detector.

Batch adsorption experiments were performed on a LabMate orbital shaker. In order to obtain the adsorption equilibrium time, the evolution of the adsorbate concentration was studied by adding 0.12 g of activated carbon to adsorbate solutions (C0 = 100 mg/L) in 50 mL-vessels. The experiments were carried out at controlled shaking (250 rpm) and temperature (30 °C) until reach the equilibrium. Therefore, in order to determine the adsorption isotherms, the target adsorbate solutions (C0 = 100 mg/L) were putted in contact to different known weights of activated carbon, in 50 mL-vessels, at control shaking (250 rpm) and constant temperature (30 °C). When the adsorption equilibrium was reached, the adsorbent was removed from the solution by filtration and the residual adsorbate concentration was analyzed by a Shimadzu UV-2401PC UV-Vis spectrophotometer. The amount of adsorbed compound at equilibrium time, which represents the adsorption capacity, qe (mg/g), can be determined by the next expression:

qe ¼

ðC 0  C e Þ  V W

ð1Þ

where, C0 and Ce (mg/L) are the initial and equilibrium adsorbate concentrations, respectively; V is the volume of solution (L) and W is the mass of adsorbent (g). In order to evaluate the dynamic performance of the adsorbent, fixed-bed adsorption experiments were conducted. Column tests

S.Á. Torrellas et al. / Chemical Engineering Journal 279 (2015) 788–798

were carried out in glass columns with an inner diameter of 0.6 cm and 30 cm of length. The columns were filled with glass balls in order to avoid dead volumes and preferential channels in the adsorption operation. The influent solutions of known adsorbate concentration (C0 = 15.0 mg/L) was pumped into the column by a Dinko D-25V multichannel peristaltic pump at up-flow mode ensuring a constant volumetric flow rate (Q = 3.0 mL/min). Experiments were conducted at different weights of adsorbent, ranging from 0.6 to 1.0 g. Effluent samples were collected at different time intervals and analyzed by UV–Vis spectrometry. All experiments were carried out at natural pH of the adsorbate solution and room temperature.

3. Results and discussion 3.1. Characterization of the activated carbons 3.1.1. Textural properties of activated carbons The textural properties of all the synthesized activated carbons were characterized by N2 adsorption–desorption at 196 °C. Fig. S2 (in the Supplementary Material) shows the N2 adsorption–desorption isotherms at 196 °C for the activated carbon obtained from peach stones and the modified materials. The adsorption–desorption isotherms can be classified as Type I isotherm and Langmuir monolayer adsorption, as it is defined in Brunauer, Deming, Deming and Teller classification (BBDT) [22], indicating that micropore filling can be described as the Polanyi– Dubinin theory. The activated carbons possess a microporous character, although, as a comparison, the N2 adsorbed volume at P/P0 values >0.10 is higher than that obtained for commercial microporous carbons such as Calgon F-400 [23], being the mesoporosity structure more developed in these synthesized materials. Hysteresis cycles of the three N2 adsorption–desorption isotherms can be classified as type H4 [24], showing a broader hysteresis loop than the obtained for the activated carbonF-400. This is due to the mesoporous character of the prepared adsorbents. This is in accordance to the determined textural parameters, surface area (SBET), total pore volume (Vpore), Dubinin– Radushkevich micropore volume (V0) and mesopore volume (Vm), which are reported in Table 2. The mesoporosity ratio for PC activated carbon was found to be 0.31, exhibiting an acceptable micropore volume value, 0.56 cm3/g. The pore size distributions (PSD) of the synthesized materials obtained by DFT method can be shown in Fig. S3 (Supplementary Material). An unimodal distribution and centered in the mesoporous range, with a medium pore size of 2.7 nm, can be seen. The different modification procedures did not seem to have much influence on the pore size, since O-PC and He-PC samples showed average pore sizes very close to one another. Nevertheless, for the modified carbons the differential pore volume decreased, which is in accordance to the results reported in Table 2. From Table 2 and the N2 adsorption–desorption isotherms can be concluded that the chemical/thermal treatments have not dramatically altered the textural properties of the modified carbons. Therefore, thus from our results and those found in the literature [25], the BET specific surface area, micropore volume and the total

Table 2 Textural parameters of the synthesized activated carbons. Adsorbent

SBET (m2/g)

Vpore (cm3/g)

V0 (cm3/g)

Vm (cm3/g)

PC O-PC He-PC

1216 959 1064

0.81 0.57 0.65

0.56 0.42 0.46

0.25 0.14 0.18

791

pore volume decreased upon nitric acid oxidation of the original activated carbon. An effect that may explain this result is the formation of macropores in the oxidation process, due to the destruction of the pore walls; this phenomenon is usually associated to the use of high concentration oxidizing agents. Taking in account that in this work, carried out at mild conditions, the pore size distribution of the oxidized carbon is centered in the same medium pore size value than original carbon, this is indicating that additional porosity has not been developed in the carbon structure [26]. So, the results can be explained by the formation of a higher quantity of oxygenated groups at the entrance or on the micropore walls, blocking the micropores and decreasing the specific surface and the micropore volume values [27]. The slight decreasing of the specific surface area of He-PC could be related to the partial gasification of the carbonaceous structure, due to the removal of the oxygenated surface groups during the heat treatment. It has been studied in the literature that the decreasing in the micropore volume is higher when the treatment temperature increases [26]. 3.1.2. Morphological studies of activated carbons The morphology of the prepared activated carbons was determined by scanning electron microscopy. SEM micrographs show a well-developed and homogeneous structure, with high porosity. The external surface of the material, with well-defined walls can be seen in Fig 1a and b shows a smooth surface, probably due to a fracture of the material, with scattered pores of different sizes. Modified materials (Fig. 1c–f) preserve similar structures to original carbon. 3.1.3. Chemical surface composition of activated carbons The FT-IR spectra of the prepared activated carbons are shown in Fig. 2. The spectra show vibration bands characteristic of carbonaceous materials. It was observed a high-intensity band at 3400 cm1 which is assigned to hydroxylic groups. The bands at 2850–2920 cm1 identify the stretching vibrations of aliphatic groups –CH2–. The band at 1629 cm1 may be assigned to a combination of C@C stretching vibration of the aromatic ring structures or attributed to systems such as diketone, ketoester and quinone [28]. The bands in the range from 1558 to 1461 cm1 are assigned to a combination of the carboxyl C@O stretching of non-aromatic carboxylic acid and lactone structures. Finally, the vibration band at 1118 cm1 probably can be assigned to the stretching C–O vibrations of carboxylate and ether structures, and the bending O–H modes of phenol structures [29–31]. Also the intensity of the vibration bands is dramatically high in the nitric acid oxidized carbon, particularly at 1118 cm1. Nevertheless, in this material can be found a vibration band at 1729 cm1, specifically assigned to C@O stretching vibrations of carboxylic groups [32]. The spectra did not reveal the presence of nitrogenated-groups into the oxidized carbon surface. Vibration bands at 1667, 2256 and/or 1525 cm1 reported by Akhter et al. [33], which are usually assigned to –O@N@O, –N@C@O and –NO2 groups, respectively, have not been found in our work. This result could be attributed to the use of a treatment temperature higher than those reported by Akhter, 80 °C versus 22–55 °C, or to the presence of aqueous phase in contact to the activated carbon. FT-IR spectra results are in accordance to the temperature-programmed desorption studies. The TPD profiles, recorded from 25 to 1050 °C of the synthesized materials are depicted in Fig. 3a. PC activated carbon showed a high quantity of oxygenated surface groups, mainly at two ranges of temperature. It decomposes from 200 to 400 °C (low-temperature), bands which are usually assigned to carboxylic groups [34], and it can be supposed that

792

S.Á. Torrellas et al. / Chemical Engineering Journal 279 (2015) 788–798

(a)

(b)

(c)

(d)

(e)

(f)

3400 cm

-1

2850 cm -1 2920 cm

-1

1729 cm

2360 cm

1461 cm -1 1558 -1cm 1629 cm -1

-1

Absorbance (u.a.)

1118 cm

-1

PC O-PC He-PC

-1

Fig. 1. SEM micrographs of (a and b) Original activated carbon; (c and d) Oxidized activated carbon; (e and f) High temperature treated activated carbon.

500

1000

1500 2000 2500 3000 -1 Wavelength (cm )

3500

4000

Fig. 2. FT-IR spectra of the synthesized activated carbons.

they have been generated when the sample was exposed to air at room temperature, due to these groups are not stable at the synthesis temperature, 435 °C [26]. The other groups decompose from 500 to 900 °C (high-temperature) are assigned to phenols (500– 750 °C), high-temperature anhydrides (800–900 °C) and carbonyl and quinone groups (650–950 °C) [35,36].

Important surface changes were introduced by chemical modifications. As expected, after oxidation, a significant increase in the concentration of the oxygenated surface groups was observed [37]. The nitric acid oxidation generates both the increase of low-temperature, hydroxyl and carboxylic groups, and the high-temperature, phenolic, carbonyl, anhydride and quinonic surface groups. In contrast, the high-temperature treatment under inert atmosphere generates an important decreasing in the amount of functional groups, mainly those decomposing in the low-temperature range. Only high-temperature, carbonyl and quinonic groups, are observed in the TPD profile of He-PC material. These groups could be removed at treatment temperatures higher than 875 °C [38]. Two 875 °C and 1050 °C ‘‘in situ’’ treatments, developed at the same volumetric flow rate of He used in the modification resulting in He-PC material, were carried out. These treatments allowed verify that the further generation of low-temperature functional groups in He-PC material can be attributed to the handling of the material afterwards, concluding that the gas-phase treatment in the reactor was successfully developed. The results obtained from these TPD studies are shown in Fig. 3b. From ‘‘in situ’’ treatments profiles can be evidenced that the oxidation in the air atmosphere generates the recovery of the surface groups which were initially removed in the thermal treatment. In addition, from Fig. 3b could

793

S.Á. Torrellas et al. / Chemical Engineering Journal 279 (2015) 788–798

(a)

(b) PC O-PC He-PC

-1

-1

a.u. (g )

a.u. (g )

875 C treatment in reactor 875 C "in situ" treatment 1050 C "in situ" treatment

0

200

400

600 T (ºC)

800

0

1000

200

400

600 T (ºC)

800

1000

Fig. 3. (a) TPD profiles of the synthesized activated carbons; (b) TPD profiles of ‘‘in situ’’ high-temperature treatments.

be observed that the quantity of high-temperature oxygenated functional groups decreased when the temperature increased from 875 to 1050 °C. The acidity or polar nature of the synthesized carbons is supported by the measurements of isoelectric point (pHIEP) and point of zero charge (pHPZC) of the samples (Fig. 4a and b). The isoelectric point value is representative of the external surface charges of the carbon, whereas the point of zero charge varies in response to the net total (external and internal) surface charge of the particle [39]. PC and O-PC materials showed pHIEP values of 3.2 and 2.5 ± 0.25, respectively. So, the sample oxidized with HNO3 (O-PC) presented a pHIEP value lower than that observed for the original sample, attributed to an increasing in the acidic oxygenated surface groups. Higher hydrophobicity of He-PC carbon hindered the neutralization of the surface sites. So, as can be seen in Fig. 4a, it was not possible to obtain the whole curve and the pHIEP value was estimated by extrapolation, close to 3.4 ± 0.25. pHPZC values are in agreement to the acidic/basic nature of the materials (Fig. 4b). As a consequence of the higher degree of oxidation of the O-PC sample, the estimated value of pHPZC, 2.1, is lower than that obtained for PC carbon, 3.1. Despite of the oxidation in the environment of He-PC sample, this sample showed a higher pHPZC value, 4.3, indicating a more basic character. This supports that the high-temperature groups, mainly remained in He-PC carbon, are weaker acidic sites than lower temperature surface groups. This statement is in accordance to the determination of the total acid sites by n-butylamine titration.

The total acidity and the relative strength of the acid sites were measured by potentiometric titration using n-butylamine in a non-aqueous media, in order to mitigate the role of the water molecules through the generation of new oxygenated surface groups on the surface carbon. Some authors [40,41] have reported a scale of acid strength measurement, indicating: the electrode potential variation (Ei) > 100 mV, for very strong sites; 0 < Ei < 100 mV, for strong acid sites; 100 < Ei < 0 mV, for weak sites and Ei < 100 mV, for very weak acid sites. The initial electrode potential (Ei) indicates the maximum acid strength of the sites and the value of mmol n-butylamine/g solid, where the plateau is reached in the neutralization profiles, indicates the total number of acid sites [40]. From the experimental results can be concluded that He-PC sample possess weaker acid sites, with a maximum initial potential (Ei) of 56.9 mV and a total acid sites equivalent to 0.5 mmol n-butylamine/g carbon. The oxidation of the original carbon increased the acidic strength (Ei = 1.6 mV) and generated more acidic sites, 3.5 mmol n-butylamine/g carbon. These results should be compared to those obtained for the original activated carbon, with a maximum potential of 28.2 mV and a total number of acid sites of 2.0 mmol/g solid. 3.2. Adsorption of emerging contaminants The adsorption of three emerging contaminants, a stimulant, caffeine, an anti-inflammatory, diclofenac and a psychiatric drug,

(a)

(b) PC O-PC He-PC

30

10 0

pH

Zeta potential (mV)

20

PC O-PC He-PC

6

4

-10 -20

2

-30 1

2

3

4 pH

5

6

7

0

2

4 6 8 Carbon fraction (wt%)

10

Fig. 4. Determination of the (a) isoelectric point (pHIEP) and the point of zero charge (pHPZC) of the synthesized carbons.

12

S.Á. Torrellas et al. / Chemical Engineering Journal 279 (2015) 788–798

carbamazepine in aqueous solution was carried out. The high content of oxygenated-surface groups in the carbon structure enhanced the adsorption of polar compounds, since in aqueous phase adsorption a larger affinity towards water molecules was established. Fig. 5 shows the rate profiles of the contaminants adsorption onto the original carbon (PC activated carbon). For the three compounds, e.g., caffeine, diclofenac and carbamazepine, the adsorption equilibrium was achieved in the first 2–3 h. It was considered that the equilibrium was attained when the adsorbate concentration measured in the aqueous phase was constant. The obtained equilibrium times express extremely high adsorption rates if they are compared to adsorption equilibrium of days to weeks exhibited by microporous commercial activated carbons such as Calgon F-400 in previous works of our group [23,42]. The results of this work are in agreement to higher mesoporosity showed by the laboratory-made activated carbons. The adsorption isotherms of caffeine on the three carbon samples at 30 °C are depicted in Fig. 6. The differences observed in the three materials are mainly a consequence of the control of the chemical surfaces adsorbents of the adsorption process. According to Giles classification [43], adsorption isotherms could be classified as S-3 type, indicating a competence between the caffeine molecule and water molecules for the active sites. The highly acidic water pH, 4.8, favoured this competitive effect. Fig. 6 shows that for caffeine equilibrium concentrations of up to 20 mg/L, the activated carbons presented relatively large caffeine adsorption capacities; however, for caffeine concentrations higher than 20 mg/L, a sigmoid trend occurred due to a synergistic effect, which favours the adsorption of caffeine molecules in aqueous solution by adsorbate–adsorbate interactions to the adsorbed molecules. This sudden increasing in the adsorption capacity can be attributed to a more favorable reorientation of the molecules, creating more availability for adsorption in the active sites [44]. Other authors explain this effect through the capillary action in mesopores [45]. The oxidized carbon showed a much higher hydrophilic character, driving a decreasing in the caffeine adsorption capacity from 260 mg/g to 126 mg/g. Müller and Gubbins [46] reported that the oxygenated sites on the surface of carbon pores can drastically change the adsorption characteristics. When such sites are present, strong bonds are formed between these active sites and water molecules, and these adsorbed molecules become nucleation sites for other water molecules to adhere to, forming three-dimensional clusters on the entrance of the pores which may block some of the

300 PC O-PC He-PC

250 200 qe (mg/g)

794

150 100 50 0 0

10

20

30 40 Ce (mg/L)

50

60

70

Fig. 6. Adsorption isotherms of caffeine onto the synthesized activated carbons (C0 = 100 mg/L, V = 50 mL, T = 30 °C, 250 rpm).

adsorption sites. This phenomenon, named in the literature as water adsorption mechanism, has been studied by several authors [47–49]. The adsorption isotherm obtained for He-PC carbon showed similar trend to PC carbon curve. These results lead us to propose that carboxylic groups play an important role in the caffeine adsorption process, and affirm that groups at 400–700 °C temperature range – which show large differences in TPD profiles of both carbons – are not influencing on the adsorption. The adsorption isotherms of diclofenac by the three synthesized activated carbons are shown in Fig. 7. In this case, the adsorption capacity become significant only at diclofenac equilibrium concentrations higher than 20 mg/L, being indicative that at diclofenac concentrations lower than 20 mg/L weaker interactions adsorbate–adsorbent were occurring. The further progressive filling of the pores could be attributed to the cooperation between adsorbate and solute, increasing the diclofenac adsorption capacity [44]. Diclofenac adsorption capacity values are lower than the obtained for caffeine on the same adsorbents and as it was expected, the oxidized carbon showed lower adsorption capacity values. Related to the study of the adsorption of hydrophobic organic contaminants, such as 2,4,6-trichlorophenol and diuron, onto carbonaceous adsorbents, can be found similar adsorption isotherms in the literature [50,51]. Carbamazepine, with a more hydrophobic character than the other two contaminants, presented a steeper adsorption isotherm

45 40

80

35

70

25

q (mg/g)

30

60

15 10

150

5 0 0,0

0,5

1,0

40

1,5

2,0 2,5 Time (h)

3,0

3,5

4,0

qe (mg/g)

50 q (mg/g)

PC O-PC He-PC

200

20

100

30 20

50

Caffeine Diclofenac Carbamazepine

10

0

0 0

10

20

30 Time (h)

40

50

60

Fig. 5. Adsorption kinetic curves of caffeine, diclofenac and carbamazepine on PC activated carbon (W = 0.12 g, C0 = 100 mg/L, V = 50 mL, T = 30 °C, 250 rpm).

0

20

40 Ce (mg/L)

60

80

Fig. 7. Adsorption isotherms of diclofenac onto the synthesized activated carbons (C0 = 100 mg/L, V = 50 mL, T = 30 °C, 250 rpm).

795

S.Á. Torrellas et al. / Chemical Engineering Journal 279 (2015) 788–798

360

PC 320 Adsorption capacity (mg/g)

curve when the solute aqueous concentration was higher than 30 mg/L. Therefore, the adsorption capacity values were larger compared to caffeine and diclofenac, reaching to 335 mg/g for the original carbon sample. As it could be expected, the oxidized carbon led to a significant decreasing in the carbamazepine removal, attributed to the competitive effect with water molecules for the available active sites. The carbamazepine adsorption isotherms are depicted in Fig. 8. Chemical-activated carbons seem not to be good adsorbents for carbamazepine removal. Similar results were found in the work of Aktas and Çeçen [52], studying the adsorption of phenol by chemical activated-carbons. We have worked with concentrations around of 100 mg/L in batch mode and 15 mg/L in fixed-bed experiments. These concentrations are not in the realistic field, since the treated micro-pollutants have been found in real aqueous matrixes in concentration levels of lg/L–ng/L. Therefore, the concentration gradient is extremely important in the adsorption experiments development, in order to obtain an acceptable saturation of the monolayers in the adsorbent surface. At higher concentrations, the adsorption tests will be deal more readily, enhancing the removal rate of the process. Additionally, it is necessary the use of a more advanced analytical methodology to determine from lg/L to ng/L concentrations, this is LC/MS chromatography. From the experimental results, when competition with the aqueous medium is occurring, only the adsorption of caffeine should be efficient at real concentration conditions. Adsorption capacity versus total number of acid sites of the synthesized carbons is depicted in Fig. 9. The equilibrium adsorption capacity values were obtained from the adsorption isotherms for the target compounds onto the three tested adsorbents (Figs. 6–8). In the Figure can be seen a general tendency of decreasing in the adsorption capacity with the increasing of the hydrophilic nature of the carbon, this is the number of acid sites, in mmol/g, excepting for the adsorption of carbamazepine onto PC-carbon. This decreasing would be attributed to the increasing of the amount of oxygenated groups in the surface activated carbon and to the diminishing of p–p interactions responsible of chemisorption bonds. The increasing in the quantity of acidic sites in the activated carbons deal with a competition of the target compounds with the aqueous medium for the active sites, resulting in a drawback of the micro-pollutants adsorption capacity. Consequently, at acidic pH of 4.8, the more polar the surface of the carbon, the lower is the adsorption capacity of the target pollutants, being increased the water adsorption phenomenon, extensively studied by Müller and Gubbins [46], and blocking some of the carbon active sites

He-PC

280 240

O-PC 200 160 120 0.5

1.0

1.5 2.0 2.5 3.0 Acid sites (mmol/g carbon)

PC O-PC He-PC

300

qe (mg/g)

250 200 150 100 50 0 0

10

20

30 Ce (mg/L)

40

50

60

Fig. 8. Adsorption isotherms of carbamazepine onto the synthesized activated carbons (C0 = 100 mg/L, V = 50 mL, T = 30 °C, 250 rpm).

3.5

Fig. 9. Influence of the chemical surface of carbons on the adsorption capacity.

adsorbing the micro-pollutants. This effect is described in the literature by other authors as solvent effect [55]. The experimental data of caffeine, diclofenac and carbamazepine adsorption have been adequately described by Sips adsorption isotherm equation. Since adsorption isotherms for all the target compounds show a sigmoidal profile, the application of usual models, such as Langmuir or Freundlich equations, result in an inadequate agreement, due to the commonly used adsorption models do not have consider the competitive effect occurring in the adsorption process. Sips model, expressed by Eq. (2), is able to have in account the competition between adsorbates and aqueous medium, dealing with a good correlation to the experimental data. This has been reported previously in the literature [53]. 1=n

qe ¼

qsat  ðb  C e Þ 1 þ ðb  C e Þ

1=n

ð2Þ

where, qsat, is the Sips maximum adsorption capacity, b, the Sips equilibrium constant and n is the Sips model exponent, which can be defined as a measurement of the heterogeneity of the adsorption system; as higher is n value, higher is the system heterogeneity. The adsorption parameters of the model are reported in Table S1 (in the Supplementary Material). The accuracy of the prediction of the experimental data by Sips model was evaluated by using the standard error of estimate (SE), Eq. (3); the fitting to the experimental data was accomplished by using the non-linear expression of Sips model and the software Solver of Microsoft Excel:

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Pn 2 i¼1 ðq0ðexpÞ  q0ðcalÞ Þ SE ¼ N 350

Caffeine Diclofenac Carbamazepine

ð3Þ

where, q0(exp), is the experimental adsorption capacity values, q0(cal), is the predicted adsorption capacity values by Sips model and N is the number of experimental points. General high accuracy to the experimental data was obtained by Sips adsorption model. Higher system heterogeneity was observed for caffeine onto the synthesized activated carbons, followed by diclofenac and finally carbamazepine system. Fixed-bed adsorption experiments were assessed for the three emerging compounds onto the synthesized adsorbent materials. The resulting breakthrough curves from the adsorption experiments are depicted in Fig. 10a–c. 0.8 g of activated carbon was used for caffeine adsorption experiments, while 0.6 g of activated carbon was used for the other two adsorbates. This is due to experimental restrictions in the preparation of the activated carbons, since it was necessary to optimize the weights of adsorbent used in the several adsorption experiments, e.g., batch and column tests.

S.Á. Torrellas et al. / Chemical Engineering Journal 279 (2015) 788–798

(a)

(b)

1.0

1.0

0.8

0.8

0.6

0.6 C/C0

C/C0

796

0.4 PC O-PC He-Pc

0.2

0.4 PC O-PC He-PC

0.2

0.0

0.0 0

20

40

60

80 t (h)

100

120

140

0

20

40

60

80 t (h)

100

120

140

160

(c) 1.0 0.8

C/C0

0.6 0.4 PC O-PC He-PC

0.2 0.0 0

20

40

60

80 100 t (h)

120

140

160

180

Fig. 10. Breakthrough curves for (a) caffeine adsorption (adsorbent weight: 0.8 g); (b) diclofenac adsorption (adsorbent weight: 0.6 g); (c) carbamazepine adsorption (adsorbent weight: 0.6 g) onto the synthesized activated carbons at C0 = 15 mg/L, Q = 3.0 mL/min, T = 20 °C, pH solution = 6.3.

Therefore, as it was confirmed in previous works [42,54], the difference of 0.2 g of activated carbon used in the fixed-bed experiments did not deal to significance changes in the adsorption efficiency of the column. From the experimental data, adsorption capacities at breakthrough (qb) and saturation (qs) times, mass transfer zone lengths (MTZ), fractional bed utilizations (FBU) and adsorbate removal percentages were determined. Breakthrough times were obtained for C/C0 values of 0.15 for caffeine adsorption, 0.27 and 0.40 for diclofenac and 0.30 and 0.52 for carbamazepine adsorption. These values are exceptionally high due to the low interaction of the activated carbons towards the pollutants, specially the more hydrophobic, diclofenac and carbamazepine. Saturation times were determined for C/C0 = 1.0. Therefore, MTZ parameter is calculated by the ratio qb/qs, according to Eq. (4).

  q MTZ ¼ Z  1  b qs

molecules in the aqueous medium. The parameters mentioned above are shown in Table S2 (Supplementary Material). So the general behaviour of the studied compounds is an overall disfavoured adsorption performance, due to a lower affinity of the adsorbate to the carbon active centres. This phenomenon is more evident in the case of the oxidized activated carbon, obtaining an abrupt reduction of the breakthrough time and a significant decreasing in the adsorption capacity at saturation time and in the removal efficiencies (Table S2). Therefore, large similarities between adsorption parameters of PC and He-PC activated carbons were found, as it can be evidenced in Table S2. So from the adsorption experiments it could be concluded that carboxylic acid groups (located at the low-temperature range of the temperature programmed profile) are the surface functionalities directly involved in the adsorption mechanism.

ð4Þ

where, qb is the adsorption capacity at breakthrough time, qs is the adsorption capacity at saturation time and Z is the length of the adsorbent bed. This equation provides a maximum value corresponding to the total bed height and when the mass transfer resistance decreases, MTZ value leads to the ideal condition, being MTZ equal to zero and the breakthrough curve, a step-function [56]. Furthermore, the unsatisfactory dynamic adsorption observed in this work could be attributed to the type and quantity of oxygen-containing surface groups on the activated carbons, shown by TPD studies; and the inefficient attraction of the carbon towards the organic pollutants due to the competition exerted by the water

4. Conclusions An important degree of transport pores – mesoporoses – in activated carbons from peach stones by H3PO4-chemical activation was obtained. So the textural characterization results showed that the use of H3PO4 as an activating agent is highly efficient in the production of activated carbons with high specific surface area and a good development of porosity. Highly favorable textural parameters were obtained for the original carbon, e.g., a SBET of 1216 m2/g and a pore volume of 0.81 cm3/g. Chemical and thermal treatments were carried out in order to study the influence of the chemical properties of the materials on the adsorption

S.Á. Torrellas et al. / Chemical Engineering Journal 279 (2015) 788–798

experiments. These modifications did not generate significant changes in the textural and morphological properties of the original adsorbent. Therefore, it was observed that the high acidic properties, this is the hydrophilic character, exhibited by the prepared activated carbons, increased with the chemical oxidation of the solid using HNO3, leading to a drawback in the adsorption affinity of the carbon chemical surface towards the studied emerging compounds. S-type adsorption isotherms and a decrease in the breakthrough time and the adsorption capacity at saturation time were obtained, due to the competitive effect occurring between the adsorbate and the water molecules for the active sites in the carbonaceous structure. This statement is based on the Monte Carlo molecular simulation model, proposed by Müller and Gubbins, reporting the more preferential adsorption of water molecules on the oxygenated surface groups, forming three-dimensional clusters, which may block the internal pores. Therefore, the TPD studies of the thermally modified-sample evidenced the recovery of a large quantity of oxygenated surface groups by the contact to the air atmosphere. The highest adsorption capacity value, as great as 335 mg/g, was obtained for carbamazepine removal, a compound that exhibits higher hydrophobicity and much lower water solubility value than caffeine and diclofenac. The experimental adsorption results suggest that the textural properties do not have influence on the adsorption removal of the organic compounds. So it seems that the low-range temperature functionalities control the adsorption mechanism of caffeine, diclofenac and carbamazepine from aqueous solutions. It could be concluded that the knowledge/determination of the physicochemical properties of the organic compounds, e.g., hydrophobicity, water solubility and molecular size, and of the adsorbent: surface area, pore volume, nature of the chemical surface groups, charge properties of the particles (pHZPC, pHIEP), and number of total acid sites and their acid strength provides a better understanding of the removal of the adsorbates in the aqueous phase. In this work the characterization of the adsorbents through thermal programmed decomposition (TPD) studies played a very important role in the interpretation of the chemical surface groups in the adsorbents surfaces. Acknowledgements The authors gratefully acknowledge the financial support from Ministerio de Economía y Competitividad, TRAGUANET Network CTM2014-53485-REDC, from Comunidad de Madrid through REMTAVARES Network S2013/MAE-2716 and Santander-Universidades Institution. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.cej.2015.05.104. References [1] R.C. Bansal, M. Goyal, Activated carbon and its surface structure, in: Activated Carbon Adsorption, CRC Press, Boca Raton, USA, 2005, pp. 1–5. [2] M. Morita, M. Higuchi, I. Sakata, Binding of heavy metal ions by chemically modified wood, J. Appl. Polym. Sci. 34 (1987) 1013–1023. [3] T. Viraraghavan, A. Kapoor, Adsorption of mercury from wastewater by peat, J. Environ. Sci. Health A 30 (1995) 553–566. [4] F. Rodríguez Reinoso, M. Molina-Sabio, Activated carbons from lignocellulosic materials by chemical and/or physical activation: an overview, Carbon 30 (1992) 1111–1118. [5] S. Omar, B. Girgis, F. Taha, Carbonaceous materials from seed hulls for bleaching of vegetable oils, Food Res. Int. 36 (2003) 11–17. [6] O. Ioannidou, A. Zabaniotou, Agricultural residues as precursors for activated carbon production –a review, Renew. Sustain. Energy Rev. 11 (2007) 1966– 2005.

797

[7] H. Marsh, F. Rodríguez Reinoso, Production and reference material, in: Activated Carbon, Elsevier Science, 2006, pp. 454–457. [8] V.O. Njoku, B.H. Hameed, Preparation and characterization of activated carbon from corncob by chemical activation with H3PO4 for 2,4dichlorophenoxyacetic acid adsorption, Chem. Eng. J. 173 (2011) 391–399. [9] Y. Muñoz, R. Arriagada, G. Soto-Garrido, R. García, Phosphoric and boric acid activation of pine sawdust, J. Chem. Technol. Biotechnol. 78 (2003) 1252–1258. [10] Y. Muñoz-González, R. Arriagada-Acuña, G. Soto-Garrido, R. García-Lovera, Activated carbons from peach stones and pine sawdust by phosphoric acid activation used in clarification and decolorization processes, J. Chem. Technol. Biotechnol. 84 (2009) 39–47. [11] C.A. Toles, W.E. Marshall, M.M. Johns, Granular activated carbons from nutshells for the uptake of metals and organic compounds, Carbon 35 (1997) 1407–1414. [12] R.B. Rios, F.W.M. Silva, A.E.B. Torres, D.C.S. Azevedo Jr., C.L. Cavalcante, Adsorption of methane in activated carbons obtained from coconut shells using H3PO4 chemical activation, Adsorption 15 (2009) 271–277. [13] I.J. Buerge, T. Poiger, M.D. Müller, H.-R. Buser, Caffeine, as anthropogenic marker for wastewater contamination of surface waters, Environ. Sci. Technol. 37 (2003) 691–700. [14] H.-R. Buser, T. Poiger, M.D. Müller, Occurrence and fate of the pharmaceutical drug diclofenac in surface waters: rapid photodegradation in a lake, Environ. Sci. Technol. 32 (1998) 3449–3456. [15] M. Clara, B. Strenn, N. Kreuzinger, Carbamazepine as a possible anthropogenic marker in the aquatic environment: investigations on the behavior of carbamazepine in wastewater treatment and during groundwater infiltration, Water Res. 38 (2004) 947–954. [16] R.P.S. Suri, J.C. Crittenden, D.W. Hand, Removal and destruction of organic compounds in water using adsorption, steam regeneration, and photocatalytic oxidation processes, J. Environ. Eng. ASCE 125 (1999) 897–905. [17] J.L. Sotelo, G. Ovejero, A. Rodríguez, S. Álvarez, J. Galán, J. García, Competitive adsorption studies of caffeine and diclofenac aqueous solutions by activated carbon, Chem. Eng. J. 240 (2014) 443–453. [18] A. Rossner, S.A. Snyder, D.R.U. Knappe, Removal of emerging contaminants of concern by alternative adsorbents, Water Res. 43 (2009) 3787–3796. [19] C. Aguilar, R. García, G. Soto-Garrido, R. Arriagada, Catalytic oxidation of aqueous methyl and dimethylamines by activated carbon, Top. Catal. 33 (2005) 201–206. [20] K.N. Rao, K.M. Reddy, N. Lingaiah, I. Suryanarayana, P.S.S. Prasad, Structure and reactivity of zirconium oxide-supported ammonium salt of 12molybdophosphoric acid catalysts, Appl. Catal. A 300 (2006) 139–146. [21] J.S. Noh, J.A. Schwarz, Estimation of the point of zero charge of simple oxides by mass titration, J. Colloid Interface Sci. 130 (1989) 157–164. [22] S. Brunauer, L.S. Deming, W.E. Deming, E. Teller, On a theory of the van der Waals adsorption of gases, J. Am. Chem. Soc. 62 (1940) 1723–1732. [23] J.L. Sotelo, A.R. Rodríguez, M.M. Mateos, S.D. Hernández, S.A. Torrellas, J.G. Rodríguez, Adsorption of pharmaceutical compounds and an endocrine disruptor from aqueous solutions by carbon materials, J. Environ. Sci. Health B 47 (2012) 640–652. [24] J.M. Martín Martínez, Adsorción física de gases y vapores por carbones, Universidad de Alicante. Secretariado de Publicaciones, Alicante, 1990. [25] A. Valente, A.M. Botelho do Rego, M.J. Reis, I.F. Silva, A.M. Ramos, J. Vital, Oxidation of pinnae using transition metal acetylacetonate complexes immobilized on modified activated carbon, Appl. Catal. A 207 (2001) 221–228. [26] C. Moreno-Castilla, M.A. Ferro-García, J.P. Joly, I. Bautista-Toledo, F. CarrascoMarín, J. Rivera-Utrilla, Activated carbon surface modifications by nitric acid, hydrogen peroxide, and ammonium peroxy disulfate treatments, Langmuir 11 (1995) 4386–4392. [27] A.N.A. El-Hendawy, Influence of HNO3 oxidation on the structure and adsorptive properties of corncob-based activated carbon, Carbon 41 (2003) 713–722. [28] P.E. Fanning, M.A.A. Vannice, DRIFTS study of the formation of surface groups on carbon by oxidation, Carbon 31 (1993) 721–730. [29] S. Biniak, G. Szymanski, J. Siedlewski, A. Swiatkowski, The characterization of activated carbons with oxygen and nitrogen surface groups, Carbon 35 (1997) 1799–1810. [30] J.J. Venter, M.A. Vannice, Applicability of ‘‘drifts’’ for the characterization of carbon-supported metal catalysts and carbon surfaces, Carbon 26 (1988) 889– 902. [31] C. Ishizaki, I. Martí, Surface oxides structures on a commercial activated carbon, Carbon 19 (1981) 409–412. [32] A. Macías-García, M.A. Díaz-Díez, E.M. Cuerda-Correa, M. Olivares-Marín, J. Gañán-Gómez, Study of the pore size distribution and fractal dimension of HNO3-treated activated carbons, Appl. Surf. Sci. 252 (2006) 5972–5975. [33] M.S. Akhter, A.R. Chughtai, D.M. Smith, Reaction of hexane soot with nitrogen dioxide/nitrogen oxide (N2O4), J. Phys. Chem. 88 (1984) 5334–5342. [34] Y. Otake, R.G. Jenkins, Characterization of oxygen-containing surface complexes created on a microporous carbon by air and nitric acid treatment, Carbon 31 (1993) 109–121. [35] K.J. Zielke, K.J. Hüttinger, W.P. Hoffman, Surface-oxidized carbon fibers: I. Surface structure and chemistry, Carbon 34 (1996) 983–998. [36] B.K. Pradhan, N.K. Sandle, Effect of different oxidizing agent treatments on the surface properties of activated carbons, Carbon 37 (1999) 1323–1332. [37] W. Cheah, S. Hosseini, M.A. Khan, T.G. Chuah, T.S.Y. Choong, Acid modified carbon coated monolith for methyl orange adsorption, Chem. Eng. J. 215–216 (2013) 747–754.

798

S.Á. Torrellas et al. / Chemical Engineering Journal 279 (2015) 788–798

[38] J.L. Figueiredo, Functionalization of porous carbons for catalytic applications, J. Mater. Chem. A 1 (2013) 9351–9364. [39] G. Newcombe, R. Hayes, M. Drikas, Granular activated carbon: importance of surface properties in the adsorption of naturally occurring organics, Colloids Surf. A 78 (1993) 65–71. [40] R. Cid, G. Pecchi, Potentiometric method for determining the number and relative strength of acid sites in colored catalysts, Appl. Catal. A 14 (1985) 15– 21. [41] M. Abdollahi-Alibeik, I. Mohammadpoor-Baltork, Z. Zaghaghi, B.H. Yousefi, Efficient synthesis of 1,5-benzodiazepines catalyzed by silica supported 12tungstophosphoric acid, Catal. Commun. 9 (2008) 2496–2502. [42] J.L. Sotelo, G. Ovejero, A. Rodríguez, S. Álvarez, J. García, Adsorption of carbamazepine in fixed bed columns: experimental and modelling studies, Sep. Sci. Technol. 48 (2013) 2626–2637. [43] C.H. Giles, T.H. MacEwan, S.N. Nakhwa, D. Smith, Studies in adsorption. Part XI: A system of classification of solution adsorption isotherms, and its use in diagnosis of adsorption mechanisms and in measurement of specific surface areas of solids, J. Chem. Soc. (1960) 3973–3993. [44] V. Gómez, M.S. Larrechi, M.P. Callao, Kinetic and adsorption study of acid dye removal using activated carbon, Chemosphere 69 (2007) 1151–1158. [45] C.-T. Hsieh, W.-S. Fan, W.-Y. Chen, Impact of mesoporous pore distribution on adsorption of methylene blue onto titania nanotubes in aqueous solution, Micropor. Mesopor. Mater. 116 (2008) 677–683. [46] E.A. Müller, K.E. Gubbins, Molecular simulation study of hydrophilic and hydrophobic behaviour of activated carbon surfaces, Carbon 36 (1998) 1433– 1438.

[47] A.S. Mestre, J. Pires, J.M.F. Nogueira, A.P. Carvalho, Activated carbons for the adsorption of ibuprofen, Carbon 45 (2007) 1979–1988. [48] O. Mahajan, C. Moreno-Castilla, P. Walker, Surface-treated activated carbon for removal of phenol from water, Sep. Sci. Technol. 15 (1980) 1733–1752. [49] P. Pendleton, S.H. Wong, R. Schumann, G. Levay, R. Denoyel, J. Rouquerol, Properties of activated carbon controlling 2-methylisoborneol adsorption, Carbon 35 (1997) 1141–1149. [50] D. Krishnaiah, S.M. Anisuzzaman, A. Bono, R. Sarbatly, Adsorption of 2,4,6trichlorophenol (TCP) onto activated carbon, J. King Saud Univ. Sci. 25 (2013) 251–255. [51] M. Al Bahri, L. Calvo, J. Lemus, M.A. Gilarranz, J. Palomar, J.J. Rodríguez, Mechanistic understanding of the behaviour of diuron in the adsorption from water onto activated carbon, Chem. Eng. J. 198–199 (2012) 346–354. [52] O. Aktas, F. Cecen, Effect of type of carbon activation on adsorption and its reversibility, J. Chem. Technol. Biotechnol. 81 (2006) 94–101. [53] S.W. Nahm, W.G. Shim, Y.K. Park, S.C. Kim, Thermal and chemical regeneration of spent activated carbon and its adsorption property for toluene, Chem. Eng. J. 210 (2012) 500–509. [54] J.L. Sotelo, A. Rodríguez, S. Álvarez, J. García, Removal of caffeine and diclofenac on activated carbon in fixed bed column, Chem. Eng. Res. Des. 90 (2012) 967– 974. [55] A.P. Terzyk, Further insights into the role of carbon surface functionalities in the mechanism of phenol adsorption, J. Colloid Interf. Sci. 268 (2003) 301–329. [56] M.G.A. Vieira, M.L. Gimenes, M.G.C. da Silva, Modelling of the process of adsorption of nickel in bentonite clay, Chem. Eng. Trans. 17 (2009) 421–426.