Equilibrium and kinetic studies of caffeine adsorption from aqueous solutions on thermally modified Verde-lodo bentonite

Equilibrium and kinetic studies of caffeine adsorption from aqueous solutions on thermally modified Verde-lodo bentonite

Applied Clay Science 168 (2019) 366–373 Contents lists available at ScienceDirect Applied Clay Science journal homepage: www.elsevier.com/locate/cla...

2MB Sizes 0 Downloads 20 Views

Applied Clay Science 168 (2019) 366–373

Contents lists available at ScienceDirect

Applied Clay Science journal homepage: www.elsevier.com/locate/clay

Research paper

Equilibrium and kinetic studies of caffeine adsorption from aqueous solutions on thermally modified Verde-lodo bentonite Maria Fernanda Oliveira, Meuris G.C. da Silva, Melissa G.A. Vieira


Department of Processes and Products Design, School of Chemical Engineering – University of Campinas, Albert Einstein Avenue, 500, 13083-852 Campinas, São Paulo, Brazil



Keywords: Adsorption Caffeine Bentonite

The removal of pharmaceutical compounds from water and wastewater is a subject of interest to the scientific community, since these substances have been related to several environmental and health problems. Usual techniques are not effective to remove pharmaceuticals and adsorption is an alternative technology with high potential to treat contaminated water. This work aimed to investigate the use of thermally modified bentonite Verde-lodo for batch adsorption of caffeine from aqueous solution. A kinetic study occurred at atmospheric pressure and 25 °C. The adsorption equilibrium time was 40 h, and the models adjusted to experimental data were pseudo-first order, pseudo-second order, Boyd and intraparticle diffusion and the pseudo-second order model showed the best fit. The resistance to external film was the limiting step on mass transfer. The equilibrium study was performed at temperatures of 15, 25, 40 and 60 °C. The Langmuir model was adjusted to data and the highest adsorption capacity was obtained at 60 °C (0.73 mmol/g). The Freundlich and Dubinin-Radushkevich models were also fitted to the experimental curves. The first one showed the best fit for temperatures of 40 and 60 °C and the second was the best for temperatures of 15 and 25 °C. The characterization techniques of scanning electron microscopy, mercury porosimetry, helium pycnometry and nitrogen physisorption indicated changes on bentonite's surface, such as increase on the bulk density, reduction on skeletal density and reduction in the volume of micropores and mesopores.

1. Introduction In recent decades, there was an increase in the interest of the scientific community in investigating the presence of pharmaceutical compounds in water and wastewater (de Andrade et al., 2018). Among these emerging contaminants, caffeine is a pharmaceutical active compound largely consumed, recalcitrant and persistent to usual water treatments, and continuously released in the environment (Chen et al., 2002; Seiler et al., 1999; Zarrelli et al., 2014). Even when present in water at low concentrations, caffeine can negatively affect the metabolism of fish, amphibians, and reptiles living in these environments (Brausch et al., 2012; Fraker and Smith, 2004; Santos-Silva et al., 2018). Therefore, there is a need to find specific treatments for efficient caffeine removal from treated water before disposing it in water bodies. Among different alternative technologies for pharmaceutical removal in water treatment, adsorption stands out as a promising technique due to its high efficiency, and lower costs and higher simplicity if compared to other methods like membrane filtration or oxidative processes (Ahmaruzzaman, 2008). Although activated carbon appears to

be one of the most commonly used adsorbents for pharmaceutical removal, as well as one of the most efficient, its use is sometimes restricted due to its high price and the difficulty in its regeneration and disposal (Bolong et al., 2009; De Gisi et al., 2016). These facts contribute to increasing the interest in investigating the use of a large variety of non-conventional low-cost adsorbents for pharmaceutical removal in wastewater. (de Andrade et al., 2018). Some studies involving pharmaceutical adsorption by these materials were already published (Antunes et al., 2012; Bekçi et al., 2006; Zhang et al., 2014; Zheng et al., 2013), but there is still a lack of information regarding caffeine adsorption. To the best of the authors' knowledge, there are only a few studies about wastewater treatment aiming caffeine removal by activated carbons (Couto et al., 2015; Sotelo et al., 2012a,b; Sotelo et al., 2014) or alternative adsorbents (Álvarez et al., 2013; Cabrera-Lafaurie et al., 2015; Cabrera-Lafaurie et al., 2012; Maia et al., 2017; Okada et al., 2015; Portinho et al., 2017; Sotelo et al., 2013; Yamamoto et al., 2018; Yamamoto et al., 2016). The use of calcined Verde-lodo (CVL) bentonite as an adsorbent for caffeine have shown favorable results in a previous study made by the authors

Corresponding author. E-mail address: [email protected] (M.G.A. Vieira).

https://doi.org/10.1016/j.clay.2018.12.011 Received 31 July 2018; Received in revised form 7 December 2018; Accepted 9 December 2018 0169-1317/ © 2018 Elsevier B.V. All rights reserved.

Applied Clay Science 168 (2019) 366–373

M.F. Oliveira et al.

(Maia et al., 2017), which led to the interest in evaluating the applicability of the adsorbent in batch systems. The CVL bentonite is an abundant low-cost adsorbent which has been successfully employed in works about metal ions removal (Almeida Neto et al., 2012; Cantuaria et al., 2016, Cantuaria et al., 2014; de Freitas et al., 2018; Freitas et al., 2017), and the material's adsorption capacity was justified by the authors due to its high specific area, great mechanical and chemical stability, and ion exchange capacity. The objective of this research was to evaluate the kinetic and equilibrium aspects of caffeine adsorption onto CVL bentonite and analyze the structural changes in the material caused by the adsorption process. 2. Material and methods 2.1. Caffeine solutions All caffeine aqueous solutions used in this study were prepared by dissolution of a standard stock solution of caffeine with purity content above 99.9% in deionized water. Absorbance measurements were performed with a Shimadzu UV–vis spectrophotometer (UVmini-1240) at λmax = 273 nm. 2.2. Adsorbent The Verde-lodo bentonite, from Boa Vista-PB (Brazil), was obtained in its raw form. The material was ground and sieved at particles with average size of 0.855 mm. The particles were calcined in a muffle oven (Quimis) at 500 °C for 24 h, in order to increase material stability (Cantuaria et al., 2014). 2.3. Adsorption kinetics The kinetic study was carried out mixing 0.16 g of CVL bentonite with samples of 100 mL of solutions containing caffeine at different concentrations (0.184, 0.460 and 0.736 mmol/L). The solutions were stirred at 200 rpm on a shaker Jeio Tech (SI-600R) and, in predetermined times (0 to 48 h), each sample was collected, filtrated through 0.45 μm PTFE syringes, and the concentration of caffeine was determined using a spectrophotometer. All experiments were conducted at constant temperature of 25 ± 1 °C. At a determined time t, the relationship between the amount of caffeine adsorbed per mass of CVL, q(t) (mmol/g), and the caffeine concentration, C(t) (mmol/L), is given by Eq. (1).

q (t ) = (V / m)(C0 − C (t ))


where t is the time (h), V is the solution volume (L), m is the adsorbent mass (g), and C0 is the initial caffeine concentration (mmol/L). The highest deviation obtained for each calculated value of caffeine concentration in duplicate was 3.24%. The models of pseudo-first order (Eq. (2)) (Lagergren, 1898), pseudo-second order (Eq. (3)) (Ho and McKay, 1998), Boyd (Eq. (4)) (Boyd et al., 1947), and intraparticle diffusion (Eq. (5)) (Weber and Morris, 1963) were adjusted to the experimental data using non-linear regression.

q (t ) = qe (1 − e−K1 t ) q (t ) =

Fig. 1. Pseudo-first order and pseudo-second order adjustments for caffeine adsorption at initial concentration of (a) 0.184 (b) 0.460 (c) 0.736 mmol/L.

equilibrium (mmol/g), t is the time (h), K1 is the rate constant of pseudo-first order model (h−1), K2is the rate constant of pseudo-second order model (g/(mmol.h)), F(t) is fractional attainment of equilibrium at a time t calculated by F = q(t)/qe, Bt is a constant calculated by Eqs. (6) and (7) (Reichenberg, 1953), Ki is the intraparticle diffusion rate constant (mmol/(g.h0.5)), and I is the constant related to the boundary layer thickness (mmol/g). If F > 0.85:


K2 qe 2t 1 + K2 qe t

F (t ) = 1 −

6 π2

(3) ∞


q (t ) = Ki t 1/2 + I

1 exp(−n2Bt ) n2

π2 Bt = −ln ⎛ (1 − F ) ⎞ ⎠ ⎝6



And if F ≤ 0.85:

where q(t) is the quantity of caffeine adsorbed onto CVL at a time t (mmol/g), qe is the amount of caffeine adsorbed onto CVL at 367


Applied Clay Science 168 (2019) 366–373

M.F. Oliveira et al.

radius of the adsorbent particle, considering it is spherical (m).

Table 1 Calculated parameters for pseudo-first order, pseudo-second order Boyd and intraparticle diffusion models. Model


Experimental Pseudo-first order

Pseudo-second order

Boyd Intraparticle diffusion

qe (mmol/g) qe (mmol/g) K1 (h−1) R2 AICC qe (mmol/g) K2 (g/ (mmol.h)) R2 AICC Def×109 (m2/h) R2 Ki (mmol/(g.h)) I (mmol/g) R2

2.4. Adsorption isotherms

Solution initial concentration (mmol/L) 0.184



0.08 0.07 0.12 0.98 −353 0.09 1.55

0.14 0.14 0.09 0.99 −334 0.18 0.48

0.16 0.16 0.12 0.97 −302 0.19 0.70

0.99 −384 1.1 0.98 0.01 0.03 0.95

1.00 −337 1.7 0.86 0.02 0.03 0.93

0.98 −316 1.4 0.93 0.02 0.03 0.93

Equilibrium experiments were performed keeping the ratio of 0.16 g of clay per 100 mL of solution used on the kinetic study. Initially, 50 mL of solutions of different concentrations were prepared and samples of 5 mL were collected from all the solutions to measure the initial concentration (from 0.06 to 6 mmol/L). The clay was added to each sample of the 45 mL of the remaining solution, the mixtures were under constant 200 rpm agitation for 48 h. After that, the solutions were filtrated and the final concentration of caffeine was determined using a spectrophotometer. The experiments were conducted at the temperatures of 15, 25, 40, and 60 °C. At equilibrium, the amount of caffeine adsorbed per mass of CVL is given by Eq. (9).

qe = (V / m)(C0 − Ce )


where V is the solution volume (L), m is the adsorbent mass (g), C0 is the initial caffeine concentration (mmol/L), and Ce is the concentration of caffeine at equilibrium (mmol/L). The deviation calculated for the value of caffeine concentration in duplicate oscillated from 0.67% up to 3.24%. The models of Langmuir (Eq. (10)) (Langmuir, 1918), Freundlich (Eq. (11)) (Freundlich and Hatfield, 1926), and Dubinin-Radushkevich (D-R) (Eq. (12)) (Dubinin and Radushkevich, 1947) were fitted to the obtained data using non-linear regression.

qe =

qmax KL Ce (1 + KL Ce )


qe = KF Ce1/ n


qe = Xm exp(−KDR ε 2)


where qe is the amount of caffeine adsorbed per mass of CVL at equilibrium (mmol/g), qmax is the Langmuir maximum adsorption capacity (mmol/g), KL is the Langmuir isotherm constant (L/mmol), Ce is the concentration of caffeine at equilibrium (mmol/L), KF is the Freundlich isotherm constant [(mmol/g).(L/mmol)1/n], n is a dimensionless empirical constant, Xm is the adsorption capacity of D-R isotherm (mmol/ g), KDR is the constant associated to the mean free energy of sorption by the relation E = 1/ 2KDR (J/mol) (Hobson, 1969), E is the mean free energy of sorption (J/mol),ε is Polanyi potential calculated by ε = − RT ln (Ce/CS) (Yang and Xing, 2010), in which R is the universal gas constant (J/(mol.K)), T is the temperature (K), and Cs is the caffeine solubility in water for a given temperature (mmol/L), which were calculated according to the literature (Shalmashi and Golmohammad, 2010; Yalkowsky et al., 2016). 2.5. Clay characterization Scanning electron microscopy (SEM) was used to evaluate the surface morphology of the material. Clay particles were covered by a thin layer of gold and placed into a support. The micrographs were obtained with a LEO 440i instrument, with 150 times and 3000 times zoom. Mercury porosimetry determined the bulk densities of the clay samples and helium pycnometry the skeletal densities. Mercury porosimetry analysis was performed on a Micromeritics (AutoPore IV), with a pressure range of 0.5 to 52,000 psia (3.5 to 359,000 kPa). The analysis of helium gas pycnometry was conducted on a Micromeritics AccuPyc II 1340 instrument. The porosity was calculated using the relation between bulk and skeletal densities (Rouquerol et al., 2013). The specific surface area was obtained using N2 physisorption (BET method). The samples were treated in vacuum at 300 °C for 3 h and analyzed in a Quantachrome (NOVA 1200e) instrument. The volumes of mesopores and micropores were calculated according to (Gañán-

Fig. 2. Curves for (a) Boyd and (b) intraparticle diffusion model curves obtained for initial concentration of 0.460 mmol/L.

Bt = 2π −

π 2F πF − 2π 1 − 3 3


It is possible to calculate the value of the effective diffusion coefficient of caffeine inside the clay particle using Eq. (8) (Reichenberg, 1953).


Deff π 2 (8)

r2 2

where Deff is the effective diffusion coefficient (m /h), and r is the 368

Applied Clay Science 168 (2019) 366–373

M.F. Oliveira et al.

Fig. 3. Langmuir, Freundlich and Dubinin-Radushkevich curves obtained for caffeine adsorption onto CVL at (a) 15, (b) 25, (c) 40, and (d) 60 °C.

0.16 mmol/g for an initial concentration of 0.736 mmol/L. Both pseudo-first order and pseudo-second order models seem to fit satisfactorily the experimental data. The parameters for the adjusted models are shown in Table 1. The good adjustment of pseudo-first order model, according to (Lagergren, 1898), indicates the beginning of adsorption process is relatively fast and, when close to its final step, the kinetic curve approaches asymptotically the equilibrium. The same behavior, according to (Simonin, 2016), can also be described by the pseudo-second order kinetic, which indicates that the final step is slow, possibly due to the diffusion into smaller pores. Calculated R2 for both models are high (close to 1) and very similar in all cases, but pseudo-second order model shows lower values of corrected Akaike information criteria (AICc), which indicates that this model is the one that best described experimental data (Bonate, 2011). Even so, according to Fig. 1 and Table 1, the pseudo-first order model was the one that estimated the values of qe closer to the experimental ones. Fig. 2 shows the models of Boyd and intraparticle diffusion fitted to the obtained data for the initial concentration of 0.460 mmol/L. The parameters for the adjustment of Boyd and intraparticle diffusion models, for all the three studied initial concentrations, are shown in Table 1. Fig. 2 (a) shows that the curve from Boyd Model is not linear, suggesting that the pore-diffusion is not the rate controlling step and indicating the existence of a step of resistance to mass transfer in the external film. Fig. 2 (b) shows a graphic with three distinct linear regions. These regions were determined by the linear fit of the data and the calculated R2 values were up to 0.99. The second linear region indicates that there is a step of intraparticle diffusion even though this is not the limiting step of the adsorption process. The observed multilinearity shows that adsorption process happens in different steps (ElKhaiary and Malash, 2011; Tran et al., 2017). Besides pointing to the existence of a resistance to external film mass transfer, Boyd model also indicates that the effective mass transfer coefficient of caffeine onto

Table 2 Calculated parameters for the equilibrium study. Model

Experimental Langmuir




qmax (mmol/g) qmax (mmol/g) KL (L/mmol) R2 AICc KF [(mmol/g).(L/ mmol)1/n] n R2 AICc Xm (mmol/g) KDR×109 (mol2/J2) E (kJ/mol) Water solubility at 1 atm (mmol/L) R2 AICc

Temperature (°C) 15




0.20 0.21 5.77 0.95 −163 0.18

0.21 0.21 7.86 0.94 −151 0.18

0.27 0.28 4.21 0.89 −160 0.21

0.39 0.73 0.44 0.95 −207 0.23

3.57 0.98 −178 0.28 5.0 9.94 54

4.50 0.93 −149 0.28 3.3 12.23 104

3.48 0.97 −190 0.44 3.3 12.24 279

1.98 0.98 −229 1.60 5.2 9.80 1044

0.98 −186

0.96 −160

0.97 −185

0.97 −220

Gómez et al., 2006).

3. Results and discussion 3.1. Adsorption kinetics Fig. 1 shows the models of pseudo-first order and pseudo-second order fitted to the obtained data. It is possible to observe that the equilibrium time was 40 h for all the three cases and in 24 h approximately 90% of the total adsorbed caffeine was already removed. The maximum value of caffeine adsorbed during the kinetic studies was 369

Applied Clay Science 168 (2019) 366–373

M.F. Oliveira et al.

Table 3 Maximum Langmuir adsorption capacities for several adsorbent materials used for caffeine removal. Adsorbent

qmax (mmol/g)


Bentonite CVL

0.21 0.21 0.28 0.73 0.41 0.26 0.95 0.27 1.15 1.39 0.96 1.09 0.82 0.80 0.35 0.50 1.89 0.35 2.33 1.40 1.35 1.19 0.15 0.13 0.13 0.21 0.21 0.21 0.25

This work

Carbon xerogel Carbon xerogel treated with nitric acid Carbon xerogel treated with nitrogen Carbon xerogel treated with sulfuric acid Comercial Norit1 GAC 1240 plus Comercial Norit1 GAC 1240 plus functionalized in inert atmosphere Babassu coco derived AC Babassu coco derived AC functionalized in inert atmosphere Dende coco derived AC Dende coco derived AC functionalized in inert atmosphere Raw grape stalk Grape stalk modified by phosphoric acid Grape stalk derived AC Commercial Mizulite montmorillonite Commercial AC FP-3 Commercial AC Calgon

Multiwalled carbon nanotubes

Carbon nanofibers

Sepiolite Minclear SG36

(15 °C) (25 °C) (40 °C) (60 °C) (30 °C) (30 °C) (30 °C) (30 °C) (23 °C) (23 °C) (23 °C) (23 °C) (23 °C) (23 °C) (room) (room) (room) (25 °C) (25 °C) (30 °C) (40 °C) (65 °C) (30 °C) (40 °C) (65 °C) (30 °C) (40 °C) (65 °C) (25 °C)

(Álvarez et al., 2015) (Álvarez et al., 2015) (Álvarez et al., 2015) (Álvarez et al., 2015) (Couto et al., 2015) (Couto et al., 2015) (Couto et al., 2015) (Couto et al., 2015) (Couto et al., 2015) (Couto et al., 2015) (Portinho et al., 2017) (Portinho et al., 2017) (Portinho et al., 2017) (Shiono et al., 2017) (Shiono et al., 2017) (Sotelo et al., 2012b)

(Sotelo et al., 2012b)

(Sotelo et al., 2012b)

(Sotelo et al., 2012b)

Fig. 4. Micrographs of CVL before contamination with zooms of (a) 150× and (b) 3000× and after contamination with zooms of (c) 150× and (d) 3000×.


Applied Clay Science 168 (2019) 366–373

M.F. Oliveira et al.

et al., 2013). The values of E, calculated using the D-R model, oscillated between 9.80 and 12.24 kJ/mol, suggesting the occurrence of ion exchange (Mobasherpour et al., 2014; Tran et al., 2017). This ion exchange occurs due to the ability of caffeine in exchange ions through the pyrimidine and imidazole rings with siloxanes from the clay (Okada et al., 2015). Table 2 also shows that for the temperatures of 15 and 25 °C, the model Dubinin-Radushkevich exhibited the highest values of R2 and lowest for AICc, and for the temperatures of 40 and 60 °C, the Freundlich model which best described experimental data. The excellent adjustment of Freundlich model, along with the tendency of caffeine adsorption without saturation (Fig. 3), suggest a heterogeneous adsorption with multilayers, which characterizes physical adsorption (Rouquerol et al., 2013; Tran et al., 2017; Yu et al., 2001). Since CVL contains Ca2+, Mg2+, and Na+, it can be classified as a polycationic bentonite (Almeida Neto et al., 2012). The presence of ions Na+ and Ca2+ is related to the high adsorption capacity of montmorillonites. The material containing these ions can adsorb water molecules, which increases the interlayer spacing and exposes the surface of the material to a greater amount of caffeine solution (Yamamoto et al., 2018). The calcination process removes the water from the montmorillonite structure, increase the distance between the layers of the material and, consequently, its adsorption capacity (Yamamoto et al., 2016). A nonionic organic compound, such as caffeine, can intercalate between the layers of the calcined montmorillonite material, which explains the occurrence of the studied adsorption phenomenon. The combination of the results obtained during the equilibrium study indicates that adsorption of caffeine onto CVL clay, for the studied range of temperatures, is a combination of weak chemisorption (ion-exchange) and physisorption. Another evidence that justifies the combination of physisorption and ion-exchange are the values of qmax showed in Table 2. The Adjustment of Langmuir model shows that for the temperature of 15 °C, the maximum Langmuir adsorption capacity, qmax, was slightly higher than the capacity at 25 °C and, for higher temperatures, the values of qmaxtended to increase. The values of qmax at 15 and 25 °C can be justified because it is possible that adsorption presents a temperature range in which the process is exothermic and other in which it is endothermic. At lower temperatures, it is common that the dominant phenomenon is the physisorption and the quantity of solute removed by adsorption decreases while temperature increases. At higher temperatures, chemisorption takes place as the dominant phenomenon, and the quantity of adsorbed material starts to increase along with temperature (Guo et al., 2005; Hills Jr., 1977; Tran et al., 2016). Besides indicating the nature of adsorption, the values of qmax also indicate that CVL has a maximum adsorption capacity comparable to other adsorbents used to remove caffeine from water. Table 3 summarizes the results found in literature for Langmuir maximum capacities of different adsorbents. Although the values of qmax for CVL on caffeine removal are still smaller than the capacities of activated carbons, the bentonite applied in this work can be considered a promising material to be used for caffeine adsorption. While activated carbons showed maximum capacity values from two to ten times greater than the carbon materials, these adsorbents can reach prices 20 times higher than bentonite clays (de Andrade et al., 2018).

Table 4 Density and porosity data of CVL without and with adsorbed caffeine.


Bulk density (g/cm ) Skeletal density (g/cm3) Porosity



1.91 2.67 28%

2.06 2.57 20%

Fig. 5. Nitrogen adsorption and desorption isotherms for VLC and VLC + CAF. Table 5 BET analysis results.


Surface area (m /g) Volume of micropores (cm3/g) Volume of mesopores (cm3/g)



74.9 20.0 24.8

25.1 6.1 14.1

CVL are of the order of 10−9 m2/h. The values of the rate constant for internal diffusion were larger for larger initial concentrations, but they did not vary meaningfully. All four models adjustments showed values of R2 > 0.8, which indicates that they can be considered satisfactory (Simonin, 2016). Pseudo-first and Pseudo-second order models, as well as the intraparticle diffusion, suggested that that the adsorption happens in different steps and the model of Boyd pointed the external mass transfer resistance as the limiting step of the adsorption process. 3.2. Adsorption equilibrium Fig. 3 shows Langmuir, Freundlich and Dubinin-Radushkevich models adjusted to obtained experimental data. Despite the isotherms obtained for 15 and 25 °C are very similar, the adsorption is favored at higher temperatures, which suggests that the process is a chemical phenomenon (Hills Jr., 1977) or a combination of chemisorption and physisorption (Tran et al., 2016). The experimental curves show that equilibrium adsorption isotherm is favorable (McCabe et al., 2004). Table 2 summarizes the calculated parameters for the three adjusted models. Although the values of R2 in Table 2 were high enough to indicate that the isotherms are Langmuir type, Fig. 3 indicates that the values of experimental qe obtained at temperatures of 40 and 60 °C show a tendency to increase their value along with the increase of Ce, without saturation, suggesting that the adsorption does not occur in monolayer (Sotelo et al., 2013; Tran et al., 2017). The same behavior was verified in a previous study about Caffeine adsorption by ion-exchange montmorillonites (Yamamoto et al., 2018) and by natural sepiolite (Sotelo

3.3. Clay characterization Fig. 4 shows the micrographs of CVL clay. The SEM analysis showed that CVL before caffeine adsorption (Fig. 4 (a) and 4 (b)) has a more rough and deformed structure. The presence of caffeine on the clay surface after adsorption (Fig. 4 (c) and 4 (d)) reduced the appearance of the material's lamellar structure. Table 4 summarizes the data obtained during the analysis of mercury porosimetry and helium gas pycnometry. After the caffeine adsorption, there was an increase in the bulk density and a decrease in skeletal density, as shown in Table 4. The first observation can be 371

Applied Clay Science 168 (2019) 366–373

M.F. Oliveira et al.


justified because adsorption occurs mostly in mesopores and micropores, which increases the material mass but does not change the volume detected by the porosimeter. The decrease in skeletal density happens because the adsorbed compound has a lower density than the adsorbent, and when adsorption occurs, the increase in the material's mass and volume do not follow the same proportion. The complete isotherm of adsorption and desorption of N2 was obtained by BET technique. The relative pressure ranged from 0.05 to 0.99. Fig. 5 shows the curves obtained for the CVL before and after the adsorption of caffeine. According to IUPAC classification, presented by Thommes et al. (2015), the isotherms shown in Fig. 5 are type IV, what is a characteristic of mesoporous materials. The hysteresis phenomenon was observed for the two curves, which indicates occupation and evacuation of the mesopores by capillary condensation (Thommes et al., 2015). BET was also used to obtain the specific surface areas of CVL and CVL + CAF and the volumes of micropores and mesopores. The obtained data are shown in Table 5. CVL has a specific surface area of 74.9 m2/g before adsorption of caffeine and 25.1 m2/g after adsorption. The volumes of micropores and mesopores for the clays before and after adsorption are, respectively, 20.0 and 24.8 cm3/g, and 6.1 and 14.1 cm3/g. The decrease in both surface specific area and pore volume after adsorption happens because the caffeine molecule occupies the previously available sites. This occupation also justifies the increase of the particle volume of the clay responsible for decreasing the skeletal density, a result evidenced by helium pycnometry. The occurrence of adsorption on the micropores and mesopores can be associated with the existence of a steric effect between Caffeine and CVL (Cabrera-Lafaurie et al., 2015). This fact justifies the results obtained in the kinetic study, which indicated the existence of a step of intraparticle diffusion during the adsorption and also the absence of a monolayer (Sotelo et al., 2013).

Ahmaruzzaman, M., 2008. Adsorption of phenolic compounds on low-cost adsorbents: a review. Adv. Colloid Interf. Sci. 143, 48–67. https://doi.org/10.1016/j.cis.2008.07. 002. Álvarez, S., Sotelo, J.L., Ovejero, G., Rodríguez, A., García, J., 2013. Low-cost adsorbent for emerging contaminant removal in fixed-bed columns. Chem. Eng. Trans. 32, 61–66. https://doi.org/10.3303/CET1332011. Álvarez, S., Ribeiro, R.S., Gomes, H.T., Sotelo, J.L., García, J., 2015. Synthesis of carbon xerogels and their application in adsorption studies of caffeine and diclofenac as emerging contaminants. Chem. Eng. Res. Des. 95, 229–238. https://doi.org/10. 1016/j.cherd.2014.11.001. Antunes, M., Esteves, V.I., Guégan, R., Crespo, J.S., Fernandes, A.N., Giovanela, M., 2012. Removal of diclofenac sodium from aqueous solution by Isabel grape bagasse. Chem. Eng. J. 192, 114–121. https://doi.org/10.1016/j.cej.2012.03.062. Bekçi, Z., Seki, Y., Yurdakoç, M.K., 2006. Equilibrium studies for trimethoprim adsorption on montmorillonite KSF. J. Hazard. Mater. 133, 233–242. https://doi.org/10.1016/j. jhazmat.2005.10.029. Bolong, N., Ismail, A.F., Salim, M.R., Matsuura, T., 2009. A review of the effects of emerging contaminants in wastewater and options for their removal. Desalination 238, 229–246. https://doi.org/10.1016/j.desal.2008.03.020. Bonate, P.L., 2011. Pharmacokinetic-Pharmacodynamic Modeling and Simulation. Springer Science & Business Media. Boyd, G.E., Adamson, A.E., Myers Jr., L.S., 1947. The exchange adsorption of ions from aqueous solutions by organic zeolites II. Kinet. J. Am. Chem. Soc. 69, 2836–2848. https://doi.org/10.1021/ja01203a066. Brausch, J.M., Connors, K.a., Brooks, B.W., Rand, G.M., 2012. Human pharmaceuticals in the aquatic environment: a review of recent toxicological studies and considerations for toxicity testing. Rev. Environ. Contam. Toxicol. https://doi.org/10.1007/978-14614-3137-4. Cabrera-Lafaurie, W.A., Román, F.R., Hernández-Maldonado, A.J., 2012. Transition metal modified and partially calcined inorganic–organic pillared clays for the adsorption of salicylic acid, clofibric acid, carbamazepine, and caffeine from water. J. Colloid Interface Sci. 386, 381–391. https://doi.org/10.1016/j.jhazmat.2014.03.009. Cabrera-Lafaurie, W.A., Román, F.R., Hernández-Maldonado, A.J., 2015. Single and multi-component adsorption of salicylic acid, clofibric acid, carbamazepine and caffeine from water onto transition metal modified and partially calcined inorganicorganic pillared clay fixed beds. J. Hazard. Mater. 282, 174–182. https://doi.org/10. 1016/j.jhazmat.2014.03.009. Cantuaria, M.L., Almeida Neto, A.F., Nascimento, E.S., Andreo Dos Santos, O.A., Vieira, M.G.A., 2014. Analysis of silver’s adsorption on calcined verde-lodo clay. In: XX Congresso Brasileiro de Engenharia Química. Florianópolis - SC, pp. 1–8. Cantuaria, M.L., Almeida Neto, A.F., Nascimento, E., Vieira, M.G.A., 2016. Adsorption of silver from aqueous solution onto pre-treated bentonite clay: complete batch system evaluation. J. Clean. Prod. 112, 1112–1121. https://doi.org/10.1016/j.jclepro.2015. 07.021. Chen, Z., Pavelic, P., Dillon, P., Naidu, R., 2002. Determination of caffeine as a tracer of sewage effluent in natural waters by on-line solid-phase extraction and liquid chromatography with diode-array detection. Water Res. 36, 4830–4838. https://doi.org/ 10.1016/S0043-1354(02)00221-X. Couto, O.M., Matos, I., da Fonseca, I.M., Arroyo, P.A., da Silva, E.A., de Barros, M.A.S.D., 2015. Effect of solution pH and influence of water hardness on caffeine adsorption onto activated carbons. Can. J. Chem. Eng. 93, 68–77. https://doi.org/10.1002/cjce. 22104. de Andrade, J.R., Oliveira, M.F., da Silva, M.G.C., Vieira, M.G.A., 2018. Adsorption of Pharmaceuticals from Water and Wastewater using Nonconventional Low-cost Materials: a Review. Ind. Eng. Chem. Res. 57, 3103–3127. https://doi.org/10.1021/ acs.iecr.7b05137. de Freitas, E.D., de Almeida, H.J., de Almeida Neto, A.F., Vieira, M.G.A., 2018. Continuous adsorption of silver and copper by Verde-lodo bentonite in a fixed bed flow-through column. J. Clean. Prod. 171, 613–621. https://doi.org/10.1016/j. jclepro.2017.10.036. De Gisi, S., Lofrano, G., Grassi, M., Notarnicola, M., 2016. Characteristics and adsorption capacities of low-cost sorbents for wastewater treatment: a review. Sustain. Mater. Technol. 9, 10–40. https://doi.org/10.1016/j.susmat.2016.06.002. Dubinin, M.M., Radushkevich, L.V., 1947. Equation of the characteristic curve of activated charcoal. In: Proc. Acad. Sci. vol. 331–333. pp. 875–890. El-Khaiary, M.I., Malash, G.F., 2011. Common data analysis errors in batch adsorption studies. Hydrometallurgy 105, 314–320. https://doi.org/10.1016/j.hydromet.2010. 11.005. Fraker, S.L., Smith, G.R., 2004. Direct and interactive effects of ecologically relevant concentrations of organic wastewater contaminants on Rana pipiens tadpoles. Environ. Toxicol. 19, 250–256. https://doi.org/10.1002/tox.20017. Freitas, E.D., Carmo, A.C.R., Almeida Neto, A.F., Vieira, M.G.A., 2017. Binary adsorption of silver and copper on Verde-lodo bentonite: Kinetic and equilibrium study. Appl. Clay Sci. 137, 69–76. https://doi.org/10.1016/j.clay.2016.12.016. Freundlich, H., Hatfield, H., 1926. Colloid [and] Capillary Chemistry. Methuen, London. Gañán-Gómez, J., Macías-García, A., Díaz-Díez, M.A., González-García, C., Sabio-Rey, E., 2006. Preparation and characterization of activated carbons from impregnation pitch by ZnCl2. Appl. Surf. Sci. 252, 5976–5979. https://doi.org/10.1016/j.apsusc.2005. 11.011. Guo, J., Xu, W.S., Chen, Y.L., Lua, A.C., 2005. Adsorption of NH3onto activated carbon prepared from palm shells impregnated with H2SO4. J. Colloid Interface Sci. 281, 285–290. https://doi.org/10.1016/j.jcis.2004.08.101. Hills Jr., C.G., 1977. An Introduction to Chemical Engineering Kinetics & Reactor Design.

4. Conclusions This work investigated the adsorption of caffeine by bentonite Verde-lodo thermally treated in batch systems. The obtained results support studies regarding the adsorption of caffeine by VLC in fixedbed. From the kinetic study, both pseudo-first and pseudo-second order models described satisfactorily the adsorption of caffeine adsorption onto CVL bentonite, but pseudo-second order showed slightly higher values of R2 and lower values of AICc. The kinetic data suggested that external film resistance is the limiting step on mass transfer during adsorption. The equilibrium study resulted in a Langmuir maximum adsorption capacity of 0.73 mmol/g (60 °C), what indicates that the material is promising to be used in future studies about caffeine adsorption. The results also indicated that the adsorption of caffeine by CVL is a combination of physisorption and ion-exchange. Characterization studies indicated that caffeine modified the clay surface, making it smoother, reducing its pore volume and increasing bulk density. The fact that calcined Verde-lodo clay had shown itself to be an excellent adsorbent for caffeine suggest that the material can be evaluated as adsorbent for other pharmaceutical compounds and can lead to future environmental benefits if used in large scale. Acknowledgements This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES) Finance Code 001. The authors also would like to thank CNPq, and FAPESP [Proc. 2016/05007-1] for the financial support and Dolomil Industrial Ltda. and MagisPharma for the donation of the clay and the caffeine, respectively, used in this work. 372

Applied Clay Science 168 (2019) 366–373

M.F. Oliveira et al.

Sotelo, J.L., Rodríguez, A., Álvarez, S., García, J., 2012a. Removal of caffeine and diclofenac on activated carbon in fixed bed column. Chem. Eng. Res. Des. 90, 967–974. https://doi.org/10.1016/j.cherd.2011.10.012. Sotelo, J.L., Rodriguez, A.R., Mateos, M.M., Hernandez, S.D., Torrellas, S.A., Rodriguez, J.G., 2012b. Adsorption of pharmaceutical compounds and an endocrine disruptor from aqueous solutions by carbon materials. J Env. Sci Heal. B 47, 640–652. https:// doi.org/10.1080/03601234.2012.668462. Sotelo, J.L., Ovejero, G., Rodríguez, A., Álvarez, S., García, J., 2013. Study of natural clay adsorbent sepiolite for the removal of caffeine from aqueous solutions: batch and fixed-bed column operation. Water Air Soil Pollut. 224 (15). https://doi.org/10. 1007/s11270-013-1466-8. Sotelo, J.L., Ovejero, G., Rodríguez, A., Álvarez, S., Galán, J., García, J., 2014. Competitive adsorption studies of caffeine and diclofenac aqueous solutions by activated carbon. Chem. Eng. J. 240, 443–453. https://doi.org/10.1016/j.cej.2013.11. 094. Thommes, M., Kaneko, K., Neimark, A.V., Olivier, J.P., Rodriguez-Reinoso, F., Rouquerol, J., Sing, K.S.W., 2015. Physisorption of gases, with special reference to the evaluation of surface area and pore size distribution (IUPAC Technical Report). Pure Appl. Chem. https://doi.org/10.1515/pac-2014-1117. Tran, H.N., You, S.J., Chao, H.P., 2016. Thermodynamic parameters of cadmium adsorption onto orange peel calculated from various methods: a comparison study. J. Environ. Chem. Eng. 4, 2671–2682. https://doi.org/10.1016/j.jece.2016.05.009. Tran, H.N., You, S.J., Hosseini-Bandegharaei, A., Chao, H.P., 2017. Mistakes and inconsistencies regarding adsorption of contaminants from aqueous solutions: a critical review. Water Res. 120, 88–116. https://doi.org/10.1016/j.watres.2017.04.014. Weber, W.J., Morris, J.C., 1963. Kinetics of adsorption carbon from solutions. J. Sanit. Engeering Div. Proc. 89, 31–60. Yalkowsky, S.H., He, Y., Jain, P., 2016. Handbook of Aqueous Solubility Data. CRC press. Yamamoto, K., Shiono, T., Matsui, Y., Yoneda, M., 2016. Changes the structure and caffeine adsorption property of calcined montmorillonite. Int. J. Geomate 11, 2301–2306. Yamamoto, K., Shiono, T., Yoshimura, R., Matsui, Y., Yoneda, M., 2018. Influence of hydrophilicity on adsorption of caffeine onto montmorillonite. Adsorpt. Sci. Technol. 36, 967–981. https://doi.org/10.1177/0263617417735480. Yang, K., Xing, B., 2010. Adsorption of organic compounds by carbon nanomaterials in aqueous phase: Polanyi theory and its application. Chem. Rev. 110, 5989–6008. https://doi.org/10.1021/cr100059s. Yu, Y., Zhuang, Y.-Y., Wang, Z.-H., 2001. Adsorption of Water - Soluble Dye onto Functionalized Resin. J. Colloid Interface Sci. 242, 288–293. https://doi.org/10. 1006/jcis.2001.7780. Zarrelli, A., Dellagreca, M., Iesce, M.R., Lavorgna, M., Temussi, F., Schiavone, L., Criscuolo, E., Parrella, A., Previtera, L., Isidori, M., 2014. Ecotoxicological evaluation of caffeine and its derivatives from a simulated chlorination step. Sci. Total Environ. 470–471, 453–458. https://doi.org/10.1016/j.scitotenv.2013.10.005. Zhang, Y., Shen, Z., Dai, C., Zhou, X., 2014. Removal of selected pharmaceuticals from aqueous solution using magnetic chitosan: sorption behavior and mechanism. Environ. Sci. Pollut. Res. 21, 12780–12789. https://doi.org/10.1007/s11356-0143212-1. Zheng, H., Wang, Z., Zhao, J., Herbert, S., Xing, B., 2013. Sorption of antibiotic sulfamethoxazole varies with biochars produced at different temperatures. Environ. Pollut. 181, 60–67. https://doi.org/10.1016/j.envpol.2013.05.056.

John Wiley & Sons, Inc., Madison, Wisconsin. Ho, Y.S., McKay, G., 1998. A comparison of chemisorption kinetic models applied to pollutant removal on various sorbents. Process. Saf. Environ. Prot. 76, 332–340. https://doi.org/10.1205/095758298529696. Hobson, J., 1969. Physical adsorption isotherms extending from ultrahigh vacuum to vapor pressure. J. Phys. Chem. 309, 2720–2727. https://doi.org/10.1021/ j100842a045. Lagergren, S., 1898. Zur theorie der sogenannten adsorption gelöster stoffe. Kungliga Svenska Vetenskapsakademiens. Handlingar 24, 1–39. Langmuir, I., 1918. The adsorption of gases on plane surfaces of glass, mica and platinum. J. Am. Chem. Soc. 40, 1361–1403. https://doi.org/10.1021/ja02242a004. Maia, G.S., Andrade, J.R., Oliveira, M.F., Vieira, M.G.A., Silva, M.G.C., 2017. Affinity studies between drugs and clays as adsorbent material. Chem. Eng. Trans. 57, 583–588. https://doi.org/10.3303/CET1757098. McCabe, W.L., Smith, J.C., Harriot, P., 2004. Unit Operations of Chemical Engineering. McGraw-Hill College. Mobasherpour, I., Salahi, E., Ebrahimi, M., 2014. Thermodynamics and kinetics of adsorption of Cu(II) from aqueous solutions onto multi-walled carbon nanotubes. J. Saudi Chem. Soc. 18, 792–801. https://doi.org/10.1016/j.jscs.2011.09.006. Neto, Almeida, De, A.F., Vieira, M.G.A., da Silva, M.G.C., 2012. Cu(II) adsorption on modified bentonitic clays: different isotherm behaviors in static and dynamic systems. Mater. Res. 15, 114–124. https://doi.org/10.1590/S151614392011005000089. Okada, T., Oguchi, J., Yamamoto, K.I., Shiono, T., Fujita, M., Iiyama, T., 2015. Organoclays in water cause expansion that facilitates caffeine adsorption. Langmuir 31, 180–187. https://doi.org/10.1021/la503708t. Portinho, R., Zanella, O., Féris, L.A., 2017. Grape stalk application for caffeine removal through adsorption. J. Environ. Manag. 202, 178–187. https://doi.org/10.1016/j. jenvman.2017.07.033. Reichenberg, D., 1953. Properties of ion-exchange resins in relation to their structure. III. Kinetics of exchange. J. Am. Chem. Soc. 75, 589–597. https://doi.org/10.1021/ ja01099a022. Rouquerol, J., Rouquerol, F., Llewellyn, P., Maurin, G., Sing, K.S.W., 2013. Adsorption by Powders and Porous Solids: Principles, Methodology and Applications. Academic Press, San Diego. https://doi.org/10.1002/vipr.19990110317. Santos-Silva, T.G., Montagner, C.C., Martinez, C.B.R., 2018. Evaluation of caffeine effects on biochemical and genotoxic biomarkers in the neotropical freshwater teleost Prochilodus lineatus. Environ. Toxicol. Pharmacol. 58, 237–242. https://doi.org/10. 1016/j.etap.2018.02.002. Seiler, R.L., Zaugg, S.D., Thomas, J.M., Howcroft, D.L., 1999. Caffeine and Pharmaceuticals as Indicators of Waste Water Contamination in Wells. Ground Water 37, 405–410. Shalmashi, A., Golmohammad, F., 2010. Solubility of caffeine in water, ethyl acetate, ethanol, carbon tetrachloride, methanol, chloroform, dichloromethane, and acetone between 298 and 323 K. Lat. Am. Appl. Res. 40, 283–285. Shiono, T., Yamamoto, K., Yotsumoto, Y., Kawai, J., Imada, N., Hioki, J., Naganuma, H., Eguchi, T., Kurihara, M., Yoshida, A., Deuchi, K., 2017. Selective decaffeination of tea extracts by montmorillonite. J. Food Eng. 200, 13–21. https://doi.org/10.1016/j. jfoodeng.2016.12.015. Simonin, J.P., 2016. On the comparison of pseudo-first order and pseudo-second order rate laws in the modeling of adsorption kinetics. Chem. Eng. J. 300, 254–263. https://doi.org/10.1016/j.cej.2016.04.079.