Experimental and Theoretical Approach to Multicomponent Adsorption

Dec 31, 2013 - Carla B. Vidal , Diego Q. Melo , Giselle S.C. Raulino , Adriana D. da Luz , Cleuzir da Luz , Ronaldo F. Nascimento. Desalination and Wa...
1 downloads 0 Views 492KB Size
Article pubs.acs.org/jced

Experimental and Theoretical Approach to Multicomponent Adsorption of Selected Aromatics on Hydrophobically Modified Zeolite Carla Bastos Vidal,*,† Giselle Santiago Cabral Raulino,† Adriana Dervanoski da Luz,‡ Cleuzir da Luz,‡ Ronaldo Ferreira do Nascimento,§ and Denis De Keukeleire∥ †

Department of Hydraulic and Environmental Engineering, Federal University of Ceará. Rua do Contorno, S/N Campus do Pici, Bl. 713, 60451-970 Fortaleza, Brazil ‡ Food Engineering Department, Santa Catarina State University, BR 282, Km 573, 89870-000 Pinhalzinho, Santa Catarina, Brazil § Department of Analytical Chemistry and Physical Chemistry, Federal University of Ceará. Rua do Contorno, S/N Campus do Pici, Bl. 940, 60451-970 Fortaleza, Brazil ∥ Faculty of Pharmaceutical Sciences, Ghent University, Harelbekestraat 72, B-9000 Ghent, Belgium ABSTRACT: The adsorption of aqueous solutions of BTEX (benzene, B; toluene, T; ethyl benzene, E; and xylenes, X) on hydrophobically modified zeolite was investigated. Multicomponent kinetics and equilibrium studies were carried out using a batch system. Furthermore, a mathematical model was studied that considers the mass transfer kinetics in a fixed-bed adsorption system. The influences of external mass transfer as well as the constant adsorption equilibrium and intraparticle diffusion resistance on breakthrough curves were evaluated. The adsorption kinetics was adjusted to the homogeneous diffusion model. The breakthrough times of the BTEX compounds increased with an increase in the bed height of the adsorbent and decreased with an increase in the flow. The mathematical model and numerical methodology that were applied represented the data of the present adsorption process with good accuracy.

1. INTRODUCTION Benzene, toluene, ethylbenzene, and xylene (BTEX compounds) occur in wastewaters from chemical and petrochemical industries at levels that are hazardous when discharged in the environment. These compounds have high toxicity, and because of this, severe regulations have been imposed on the concentrations of BTEX compounds in discharge wastewaters. Various conventional and advanced technologies, including adsorption using activated carbon or synthetic zeolites as adsorbent, are constantly used for the treatment of contaminations with BTEX compounds, but cost reduction remains an important issue.1,2 Zeolites are natural alumino-silicate minerals that are characterized by cagelike structures, high surface areas, and efficient cation-exchange capacities. Voluminous cationic surfactant molecules, such as hexadecyltrimethyl ammonium bromide, have a strong affinity for the zeolite surface and positively replace charged inorganic counterions. They also neutralize the negative surface charge of the zeolite. The surfactant molecules impart hydrophobic properties to the zeolite surface, allowing the adsorbent to retain organic compounds including BTEX compounds.3 Conceptual adsorption parameters are used to predict the quality of the effluent under different operating conditions. From the batch studies, it was possible to obtain of the isotherm constants and mass transfer coefficients. The homogeneous solid-phase diffusion model was used to obtain the parameters that are responsible for mass transfer operation, © 2013 American Chemical Society

such as the external mass transfer coefficient and the intraparticle diffusivity or surface diffusion coefficient.4 In this context, the aim of this research was to study the adsorption equilibrium and kinetics of the BTEX compounds in aqueous solution, batch reactor at 28 ± 2 °C, using modified zeolite as the adsorbent and then use the parameters obtained for simulating the breakthrough curves of BTEX compounds in multicomponent column adsorption. The mathematical model of pore diffusion, which considers the resistances to mass transfer to the adsorbent particles in addition to the competitive nonlinear adsorption isotherm, was used in the present work. The finite volume method was used in the discretization of the equations using the weight upstream differencing scheme interpolation function along the column and the central difference scheme along the particle, and the algorithm was implemented in the FORTRAN programming language.

2. MATERIALS AND METHODS 2.1. Chemicals and Solutions. The functionalizing agent used was hexadecyltrimethyl ammonium bromide (SigmaAldrich, St. Louis, MO, USA) with Faujasite type NaY zeolite (Policlay Nanotech, Fortaleza, Ceará, Brazil) and deionized water. BTEX in methanol (2 g·L−1 of each component: Received: June 3, 2013 Accepted: December 18, 2013 Published: December 31, 2013 282

dx.doi.org/10.1021/je400780f | J. Chem. Eng. Data 2014, 59, 282−288

Journal of Chemical & Engineering Data

Article

where r is the position in the radius from the center of the particle considered to be spherical, t is the time, and Def is the effective diffusivity coefficient. 2.4. Determination of the BTEX Compounds by Gas Chromatography (GC) Coupled to Mass Spectrometry (MS). Quantitative analyses of the compounds were performed by GC/MS model QP2010 Plus from Shimadzu. The carrier gas was Helium,which was used at a flow rate of 1 mL·min−1 and a pressure of 53.5 kPa. One microliter aliquots were injected using the splitless mode and separated on a DB-5 column (Agilent, Santa Clara, CA, USA) (length, 30 m; i.d., 0.25 mm; and film thickness, 0.25 mm). The temperature program started at 40 °C for 1 min, increased at a rate of 5 °C min−1 to 70 °C, and then increased at a rate of 20 °C min−1 to 200 °C. The compounds p-xylene and m-xylene were not resolved with this method and were treated as a single compound. 2.6. Numerical Methodology. The numerical methodology used in the present study is based on the model established by Chatzopoulos and Varma,9 which describes the adsorption process of removing toluene in a fixed-bed column using activated carbon. The mathematical model is a grouped model of diffusion in the pores, which considers the mass transfer resistance inside and outside the particle adsorbent. Mathematical modeling considers the equations of chemicalspecies conservation in the solid and liquid phases, which describe the variation of the solute concentration inside the fixed-bed column and the particle with respect to time and position. The surface diffusion coefficient (Ds) was assumed to increase exponentially with surface coverage according to eq 4.

benzene, toluene, ethylbenzene, and xylene) was obtained from Sigma, and methanol (HPLC grade) and distilled water were used for the adsorption experiments. The solutions that served for this purpose were prepared by dilution of appropriate stock solutions. 2.2. Hydrophobically Modified Zeolite Characterization. The hydrophobically modified zeolite was subjected to the pore-size distribution analysis, which was calculated using the BJH method,5 and the specific surface area was determined using the BET (Brunauer, Emmett, and Teller) method6 at 77 K (−196 °C) with an accelerated surface area and porosimetry system (model ASAP 2020, Micromeritics, Norcross, GA, USA). Zeta potential measurement of hydrophobically modified zeolite was obtained using a Zetasizer Nano (model ZEN 3500, Malvern, Worcestershire, UK). 2.3. Adsorption Isotherms. Equilibrium adsorption tests were carried out to determine the maximum capacity of the adsorbent for the multicomponent adsorption of BTEX. During the isotherm tests, 300 mg of the adsorbent was added to 10 mL of a BTEX solution with pH 7 in glass flasks that were shaken for 11 h at 300 rpm and 28 ± 2 °C. The initial concentrations of each studied compound varied from 2 mg· L−1 to 60 mg·L−1 (in duplicate tests). The amount of the BTEX compounds adsorbed was calculated by the following mass-balance eq 1.

qe =

V (C in − Ce) M

(1)

where V is the volume of the initial solution, Cin is the initial concentration of the BTEX compounds, M is the mass of adsorbent, and Ce is its concentration at equilibrium. The multicomponent isotherm model used in this work was the Langmuir isotherm, expressed by eq 2, to fit to the experimental multicomponent data.4 qei =

⎡ ⎛ ⎤ q ⎞⎥ ⎟⎟ Ds(q) = Def exp⎢k ⎜⎜ ⎢⎣ ⎝ qsat ⎠⎥⎦

where Def is the diffusion coefficient at q = 0, k is a parameter of eq 3, q is the concentration of the solute in the solid phase, and qsat is the saturation concentration of the solute at the surface. Equation 3 was incorporated into the adsorption model to take into account the variation in the surface diffusion coefficient, Ds, as a function of time and position within the particles.9 Setting k = 0 in eq 4 also allowed for the use of a constant (concentration-independent) Ds. In this case, the concentration of the solute in the liquid phase, C, varies with the axial position, z, and the time, t, whereas the concentration in the solid phase, q, is a function of the radial position, r, inside the particle. Assuming spherical adsorbent particles, an isothermal process, and fast adsorption kinetics, the solute mass balance in the solid phase is given by eq 5.

qmax KL1Ce1 n

1 + ∑i = 1 KLiCei

(2)

where qe1 is the adsorbed quantity of each component per gram of adsorbent at equilibrium, Ce1 is the concentration of each component at equilibrium, KL1 is the monocomponent (noncompetitive) Langmuir adsorption constant of the single component, KLi is the individual Langmuir adsorption constant, and ni is the multicomponent (competitive) Langmuir adsorption constant of each component 2.4. Adsorption Kinetics. Three-hundred milligrams of hydrophobically modified zeolite was added to 10 mL of the BTEX solution (10 mg·L−1 of each compound) in glass flasks (50 mL) that were shaken at 300 rpm and pH 7.7 Supernatant aliquots were collected at different times, varying from 30 min to 2880 min. Experiments were performed in duplicate. The experimental data of the adsorption kinetics of the multicomponent BTEX compounds were adjusted by the homogeneous diffusion model of a particle according to eq 3.8 ∂q ∂q ⎫ 1 ∂⎧ = 2 ⎨r 2Def ⎬ ∂t ∂r ⎭ r ∂r ⎩

(4)

⎡ q ⎤ ∂q ⎫ ⎪ ⎪ ∂q D ∂⎧ 2 ⎢k ⎥ ⎬ = ef2 ⎨ r exp ⎪ ∂t ⎢⎣ qsat ⎥⎦ ∂r ⎪ r ∂r ⎩ ⎭

(5)

with the following initial (IC) and boundary conditions (BC) (3)

IC: t = 0,

0 ≤ r ≤ R,

0 ≤ z ≤ L,

with the following initial (IC) and boundary conditions (BC) q(r , 0) = q0 ,

q(r , t ) = q0 ,

∂q ∂r

BC1: t > 0,

=0 r=0

(3a) 283

0 ≤ z ≤ L,

∂q ∂r

r=0

qi = 0

(5a)

(5b)

dx.doi.org/10.1021/je400780f | J. Chem. Eng. Data 2014, 59, 282−288

Journal of Chemical & Engineering Data

Article

Table 1. Parameters Obtained from the Multicomponent Langmuir Isotherm2 Qmax KL R2 E

Langmuir

benzene

toluene

ethylbenzene

m,p-xylene

o-xylene

150.420 0.0000 0.918 0.828

152.413 0.015 0.829 0.588

162.219 0.392 0.771 1.923

175.332 0.105 0.764 0.922

164.579 0.145 0.717 1.598

BC2: t > 0, 0 ≤ z ≤ L , ⎡ ⎛ ⎤ q ⎞⎥ ∂q ⎟⎟ = k f (C − Ce) Def ρs exp⎢k ⎜⎜ ⎢⎣ ⎝ qsat ⎠⎥⎦ ∂r r = R

I for unmodified zeolite. This indicates that the sample had uniform microporous diameters of 7.4 Å, as calculated by the BJH method. The BET surface area was 650.3 m2·g−1. 3.1.2. Zeta Potential. The zeolite adsorbent has a negative charge on its surface. Surface modifications of zeolite with the surfactant at different concentrations resulted in the change of the zeta potential from negative to positive and a transition from a monolayer to a double layer on the modified zeolite.2,10 Zeolites modified with surfactant whose concentrations are below the CEC exhibit a negative charge, indicating that exchangeable active sites on the external surface of the zeolite are not completely covered and a submonolayer is formed. The negative surface of the zeolite progressively decreases with increasing surfactant load, becoming more hydrophobic as it approaches the isoelectric point (zeta potential = 0). In the modification with 100% of the CEC, the surface cations of the zeolite were replaced by cationic surfactant molecules, leaving it uncharged and more hydrophobic. The isoelectric point at which the concentration of surfactant adsorbed on the zeolite corresponds to the CEC (2.9 mmol·g−1), indicating a complete monolayer surface of surfactant, which is a good indicator for exhibiting a maximum hydrophobically modified surface.2,10 For values above the isoelectric point, surfactant molecules form a second layer or an incomplete second layer, reversing the surface charge and becoming hydrophilic once again. A stable surfactant monolayer formed on the zeolite surface is capable of sorbing nonpolar organic molecules because of exposed hydrophobic tail groups. Bowman et al.11 suggested that the sorption of benzene, toluene, and p-xylene by surfactant-modified zeolite was controlled by a partitioning mechanism, and a strong correlation between the compound’s sorption coefficient (Kd) and its octanol−water partition coefficient (Kow) was noted.12 Some authors have studied the variation of the pH during adsorption of BTEX compounds on carbon nanotubes oxidized with NaOCl, and they observed no significant change in adsorption capacity related to pH.2,10,13,14 In this work, the zeolite was initially introduced in its sodium form, having pH values that revealed near neutrality of the materials. 3.2. Adsorption Isotherms. The adsorption isotherms of the selected aromatics on the hydrophobically modified zeolite were obtained by constructing curves correlating the concentration of the compounds in the solid phase versus the concentration of the compounds in the liquid phase at room temperature (28 ± 2 °C). The experimental data were applied to the multicomponent Langmuir model. The parameters determined from the model are shown in Table 1. The Langmuir model proposes a homogeneous adsorption mechanism, taking the surface uniformity of the adsorbent and the adsorption sites as energetically identical. According to the literature, the experimental data for low values of Ce better fit the Langmuir isotherm.1 3.2. Adsorption Kinetics. The study of the adsorption kinetics of the BTEX compounds in solution by the modified

(5c)

where ρs is the solid density, kf is the external mass transfer coefficient, and Ce is the solute concentration in the liquid phase at the solid−liquid interface. The initial condition used to describe the model for the solid phase requires the concentration in the solid phase to be equal to zero for adsorption along the bed at time zero. The boundary conditions employed in the model are symmetry and equality of flows. In the absence of axial solute dispersion in the bed, the solute mass balance in the fluid phase along with the relevant IC and BC is expressed by eq 6, and eqs 6a and 6b are the initial and boundary conditions, respectively, for this equation. v ∂C ∂C 3 (1 − ε) =− s − k f (C − Ce) ∂t ε ∂z R ε

IC: t = 0,

0 ≤ z ≤ L,

BC: t > 0,

z = 0,

(6)

C=0

(6a)

C = C in(t )

(6b)

where νs is the superficial velocity of the liquid in the bed, ε is the bed porosity, and R is the radius of the adsorbent particle. The time-varying BC at the bed inlet in eq 6b was employed primarily to lend greater flexibility to the model so that it can handle experiments under variable influent-concentration conditions. However, this feature of the model also allowed for the successful description of discontinuities in the experimental breakthrough curves caused by small variations of the influent concentration, Cin, with time resulting from the batchwise production of feed solution. Because of the assumption of fast intrinsic adsorption kinetics, the solidphase and liquid-phase solute concentrations at the solid− liquid interface can be related through the equilibrium isotherm expressed by eq 2. The finite volume method was used to discretize the conservation equations. The choice is due to the fact that it ensures the conservation of the quantities involved, both at the elementary level and globally. In this work, we used the explicit formulation and one-dimensional structured mesh for storing discrete points. In the computational mesh, we used a colocated arrangement of variables, where all variables are stored in the center of the control volumes. For the evaluation of the variables and their derivatives on the faces of the control volumes, the weight upstream differencing scheme (WUDS) along the column and the central difference scheme (CDS) throughout the particle were applied.

3. RESULTS AND DISCUSSION 3.1. Zeolite Characterization. 3.1.1. Nitrogen Adsorption−Desorption Isotherms. The adsorption−desorption isotherms of nitrogen followed the IUPAC classification type 284

dx.doi.org/10.1021/je400780f | J. Chem. Eng. Data 2014, 59, 282−288

Journal of Chemical & Engineering Data

Article

zeolite was conducted to determine the time required for the system to reach the adsorption equilibrium. The results showed that in 6 h the equilibrium time was reached for all compounds (Figure 1). The adsorption kinetics is initially fast because of

Table 2. Parameters Used to Obtain the Multicomponent Breakthrough Curves of BTX Compounds (Benzene, Toluene, and o-Xylene) on Thermally Activated Charcoal from Coconut Shellsa

a

Figure 1. Multicomponent adsorption kinetics of BTEX compounds and adjustment of effective diffusivity (Def). Conditions: 10 mg·L−1, M = 300 mg, V = 10 mL, and T = 28 °C ± 2 °C.

the adsorption process occurring at the external surface and is followed by a slow adsorption step on the inner surface of the adsorbent. If the adsorbent has a low microporosity not accessible to molecules of the solute, then the adsorption kinetics are faster when compared to an adsorbent with a high volume of micropores. The volume of the micropores in zeolites is considered large, thus explaining the slower adsorption kinetics. Figure 1 shows the adsorption kinetics of multicomponent BTEX compounds. The concentration of each contaminant studied was 10 mg·L−1, the temperature was kept constant at 28 ± 2 °C, and the initial pH was 7. The experimental data of the adsorption kinetics of the multicomponent BTEX compounds are shown in Figure 1. 3.3. Validation and Results. The equations that describe the adsorption of the BTX compounds (benzene, toluene, and o-xylene) in the mixture using a fixed-bed column were solved, with the aim of validating the proposed mathematical model and the numerical methodology developed. The results obtained by the simulation are compared with the experimental data of Luz et al.15 A column (7.0 cm × 1.2 cm i.d.) was used to obtain the experimental data. The column was filled with thermally activated charcoal from coconut shells. A detailed description of the experiment can be found in Luz et al.15 The input parameters of the model to determine the concentration profiles of the BTX compounds in the mixture are shown in Table 2. Figure 2 shows the breakthrough curve of the BTX compounds (benzene, toluene, and o-xylene) (C/Cin) using the parameters shown in Table 2 by applying the mathematical model. The numerical results in Figure 2 presented a greater deviation in relation to the experimental data for the multicomponent breakthrough curves, mainly for benzene and toluene, which presented lower affinity for the solid phase. According to Luz,15 this discrepancy can be due to the short bed height and high flow rate used, resulting in a desorption of compounds weakly adsorbed.

parameters

benzene

toluene

o-xylene

Cin (mg·L−1) εL (adim) Def (cm2·s−1) Dm (cm2·s−1) ρs (g·L−1) dp (cm) Dc(cm) kf (cm·s−1) Q (mL·min−1) T (°C) L (cm) KL (L·g−1) qmax (mg·g−1)

50 0.41 9.30·10−11 9.8·10−6 1850 0.085 1.20 6.3107·10−3 40 23 7.0 0.0490 124.77

50 0.41 9.70·10−9 8.6·10−6 1850 0.085 1.20 6.3107·10−3 40 23 7.0 0.0497 150.42

50 0.41 9.50·10−7 8.4·10−6 1850 0.085 1.20 6.3107·10−3 40 23 7.0 0.0405 165.07

According to Daifullah and Girgis.16

Figure 2. Experimental and simulated breakthrough curves of BTX compounds (benzene, toluene, and o-xylene) obtained for different initial concentrations.16

According to Sulaymon and Ahamed,4 this deviation is related to the Biot number, that is, an adimensional number, expressed by eq 7.

BiMi =

k f Lc Def

(7)

The Biot number is a relation between the resistance to internal mass transfer by diffusion and the resistance to external mass transfer by convection. In the present work the following Biot numbers were found for the BTX compounds: BiMb= 4.75·108, BiMt = 4.55·106, and BiMx = 4.65·104. The competitive adsorption rate will decrease as the average Biot number increases for each adsorbate, leading to a lower breakpoint because of the low intraparticle resistance and the reduction in the contact time. The increase in the bed height leads to an increase in the competitive adsorption rate, and the displacement of the weak components will be greater.4 A higher Biot number was observed for benzene and toluene in the BTX mixture, where a lower intraparticle diffusivity of these compounds occurs compared with o-xylene. This corroborates the mathematical model and the numerical methodology used, demonstrating that they represent the real adsorption process with good accuracy and allowing other situations to be simulated. 285

dx.doi.org/10.1021/je400780f | J. Chem. Eng. Data 2014, 59, 282−288

Journal of Chemical & Engineering Data

Article

Table 3. Parameters Used to Obtain the Multicomponent Breakthrough Curves of the BTEX (Benzene, Toluene, Ethylbenzene, and Xylene) Compounds on Modified Zeolite parameters

benzene

toluene

ethylbenzene

m,p-xylene

o-xylene

Cin (mg·L−1) εL (adim) Dm (cm2·s−1) Def (cm2·h−1) ρs (g·L−1) dp (cm) Dc (cm) Q (mL·min−1) T (°C) L (cm) KL (L·g−1) qmax (mg·g−1)

10 0.2631 9.8·10−6 1.39·10−9 472.8 0.0045 1.2 40 28 7.0 0.000 150.420

10 0.2631 8.6·10−6 1.11·10−10 472.8 0.0045 1.2 40 28 7.0 0.015 152.413

10 0.2631 7.8·10−6 1.11·10−9 472.8 0.0045 1.2 40 28 7.0 0.392 162.219

10 0.2631 8.4·10−6 5.56·10−10 472.8 0.0045 1.2 40 28 7.0 0.105 175.322

10 0.2631 8.4·10−6 9.72·10−10 472.8 0.0045 1.2 40 28 7.0 0.145 164.579

3.3.1. Simulation of the Kinetics of Adsorption of the BTEX Compounds in a Fixed-Bed Column. Table 3 provides the conditions and parameters necessary for determining the breakthrough curves of the BTEX compounds in the mixture using the equilibrium and kinetic parameters experimentally determined by Vidal et al.2 The external mass transfer coefficients in the packed bed model for each solute were evaluated by the correlation of Wilson and Geankoplis4 (eq 8). Shi =

1.09 1/3 1/3 Sci Re εb

(8) Figure 3. Numerical results of the multicomponent breakthrough curves of the BTEX compounds (benzene, toluene, ethylbenzene, and xylene) compounds: Cin = 10 mg·L−1, Q = 40 mL·min−1, L = 7.0 cm, and εL = 0.26.

For a Re number between 0.0015 and 55, we have an Shi value expressed by eq 9, Sci expressed by eq 10, and Re expressed by eq 11. Shi =

Sci =

Re =

kf dp Dm, i

(9)

observed as shown in Table 1, according to the parameters obtained. This compound was followed, in terms of adsorption capacity, by m,p-xylene, o-xylene, ethylbenzene, toluene, and benzene. According to Su et al.,14 Luz et al.,15 and Daifullah and Girgis,16 the decrease in solubility (B, 790 mg·L−1 >T, 530 mg· L−1 > E, 152 mg·L−1 > o,m,p-X 150.5 mg·L−1) and increase in the molar mass (B, 78 g < T, 92 g < E, o,m,p-X, 106 g) resulted in better adsorption of BTEX compounds. In the case of multicomponent systems, the active sites are available in a large amount at the initial stage, and the compounds have free access to them. With the advancement of time, the strongly adsorbed compounds tend to adsorb in the active sites occupied first by weakly adsorbed compounds, thereby displacing these sites. As a result, the concentration of the weakly adsorbed compounds is higher in the fluid phase of the fixed-bed adsorbent.4 Consequently, the weakly adsorbed compounds elute from the column, resulting in a final concentration higher than the initial one. This can be observed in the curves of the compounds o-xylene, ethylbenzene, toluene, and benzene, which, upon reaching saturation (C/Cin = 1), start to elute, giving room for the m,p-xylene compound that is more strongly adsorbed and whose curve shows the profile of a common rupture curve. These results are consistent with the observations of Sulaymon and Ahmed,4 who studied the competitive adsorption of furfural and phenolic compounds in a fixed bed of activated carbon, and Luz et al.,15 who studied the competitive adsorption of BTX compounds on coconut shell activated carbon.

μw ρw Dm, i

(10)

ρw vd p μw

(11)

where Sci is the Schmidt number, Shi is the Sherwood number, Re is the Reynolds number, dp is the particle diameter, Bi is the Biot number, μw is the viscosity of water, ρw is the density of water, εb is the bed porosity, ν is the interstitial velocity, and Dm,i is the molecular diffusion coefficients of benzene, toluene, ethylbenzene, m,p-xylene, and o-xylene in aqueous solution, which are listed in Table 3. Figure 3 shows the numerical results of the multicomponent breakthrough curves of the BTEX compounds using an initial concentration, Cin, of 10 mg·L−1for each contaminant in the mixture, a volumetric flow, Q, of 40 mL·min−1, a bed porosity, ε, of 0.26, and a bed height, L, of 7.0 cm. The order of adsorption of the compounds in the active sites of the adsorbent was m,p-xylene, o-xylene, ethylbenzene, toluene, and benzene. The same order was found by Su et al.,14 Daifullah and Girgis,16 and Yu et al.17 The preferred adsorption of the compounds in this order can be explained by the decrease in solubility and the increase in molecular weight.16 A higher adsorption for m,p-xylene, which had a larger structure, greater molar mass, and lower water solubility, was 286

dx.doi.org/10.1021/je400780f | J. Chem. Eng. Data 2014, 59, 282−288

Journal of Chemical & Engineering Data

Article

With the objective of predicting the operating conditions of an adsorption process, we performed a sensitivity parametric analysis of the operational conditions. The parameters studied include volumetric flow, bed height, and bed porosity. A numerical experiment was performed for three cases specified in Table 4 for the parameters shown in Table 3. Table 4. Operational Conditions Used in the Numerical Simulations case

simulations

concentration mg·L

1 2 3

1 2 5 6 8 9 10

10 10 10 10 10 10 10

−1

flow

bed height −1

mL·min 30 40 40 40 40 40 40

porosity

cm

εL

7.0 7.0 7.0 10.0 7.0 7.0 7.0

0.26 0.26 0.26 0.26 0.26 0.36 0.41

Figure 5. Numerical results of the multicomponent breakthrough curves of the BTEX (benzene, toluene, ethylbenzene, and xylene) compounds for different bed heights (L): Cin = 10 mg·L−1, Q = 40 mL· min−1, and εL = 0.26.

The breakthrough times of the BTEX compounds increased because of the increasing of the height of the adsorbent bed. The literature reports that the higher the bed height of the adsorbent is the longer the service time of the column will be, taking into account that the surface area of the material increases as the number of active sites available for interaction adsorbate−adsorbent increases. In addition, the adsorption capacity of the material also increases with the increase of the adsorbent bed height.18,20,21 The HRT varies from 11.8 s for the 7.0 cm bed height to 16.9 s for the 10 cm bed height, thereby increasing the service life of the column. Figure 6 displays the numerical results of the multicomponent breakthrough curves of the BTEX compounds

Figure 4 shows the numerical results of the multicomponent breakthrough curves of the BTEX compounds using an initial

Figure 4. Numerical results of the multicomponent breakthrough curves of the BTEX (benzene, toluene, ethylbenzene, and xylene) compounds for different flows (Q): Cin = 10 mg·L−1, L = 7.0 cm, and εL = 0.26.

concentration, Cin, of 10 mg·L−1 for each compound, a bed porosity, εL, of 0.26, and a bed height, L, of 7.0 cm for two different flows of the system (Q = 40 mL·min−1 and 30 mL· min−1). As the flow increases, the breakthrough times of the BTEX compounds decrease as well the adsorption capacity of the zeolite for each compound. The increase in flow rates implies a reduction in the hydraulic retention time (HRT) of the compounds inside the column. According to Cooney,18 the HRT is a typical parameter of design and operation for the usage of columns. Long residence times can lead to a decrease in contaminant removal, whereas shorter times do not allow an effective contact for the interaction between the adsorbent and the adsorbate.17,19 The HRT is given by the ratio between the volume of the reactor (in this case, the column) and the flow. The HRT decreased from 15.6 s to 11.8 s when the flow rate increased from 30 mL min−1 to 40 mL min−1, resulting in a decrease of the adsorption capacity. Figure 5 shows the numerical results of the multicomponent breakthrough curves of the BTEX compounds using a initial concentration, Cin, of 10 mg·L−1 for each compound, a bed porosity, εL, of 0.26, and a volumetric flow rate, Q, of 40 mL· min−1 for two different bed heights (L = 7.0 cm and 10 cm).

Figure 6. Numerical results of the multicomponent breakthrough curves of the BTEX (benzene, toluene, ethylbenzene, and xylene) compounds for different porosities (εL): Cin = 10 mg·L−1, Q = 40 mL· min−1, and L = 7.0 cm.

using a concentration, Cin, of 10 mg·L−1 for each compound, a bed height, L, of 7 cm, and a volumetric flow rate, Q, of 40 mL· min−1 for three different values of the bed porosity (εL = 0.26, 0.36, and 0.41). As the porosity of the bed increases, the void space also increases, thus decreasing the number of active sites inside the column. As a result, the adsorption capacity of the bed is smaller than the breakthrough times. This effect was most pronounced for the compounds whose rupture time was increased. For toluene, the breakthrough curves for the three porosities practically overlap, whereas for m,p-xylene, the breakthrough time increased from approximately 8 h to 10 h when the porosity decreased from 0.41 to 0.26. Thus, for all cases, it can be observed in different simulations that the concentration of a solute that has a low affinity for the 287

dx.doi.org/10.1021/je400780f | J. Chem. Eng. Data 2014, 59, 282−288

Journal of Chemical & Engineering Data

Article

adsorbent increases in the fluid phase, overcoming the dimensionless concentration. This can be explained because at the beginning of the adsorption process there is a large amount of active sites available where all compounds can adsorb easily. Over time, the bed becomes saturated by contaminants present in the mixture, reducing the amount of active sites in the bed. Thus, the compounds that showed high affinity for the active sites began competing with compounds that had low affinity, thus desorbing them. As the compounds with low affinity are desorbed, an increase in concentration of weak compounds in the outlet of the bed is apparent, thereby increasing their concentrations in the fluid phase (sum of the concentration of entry and the amount desorbed). Therefore, the breakthrough curves of compounds that adsorb weakly become flat, and their interruptions occur rapidly. This can also be explained by the fact that the driving force for mass transfer increases with the increasing of the solute concentration, resulting in reduction of competitive adsorption. Similar results were found by Sulaymon and Ahmed4 and Su et al.14

hydroxide cetylsulfonyl acetate in solution. Mater. Lett. 2008, 62, 219− 221. (6) Brunauer, S.; Emmett, P. H.; Teller, E. Adsorption of gases in multimolecular layers. J. Am. Chem. Soc. 1938, 60, 309−319. (7) Chandrasekhar, S.; Pramada, P. N. Investigation on the synthesis of zeolite NaX from Kerala kaolin. J. Porous Mater. 1999, 6, 283−297. (8) Ruthven, D. M. Principles of Adsorption and Adsorption Processes; John Wiley & Sons: New York, 1984. (9) Chatzopoulos, D.; Varma, A. Aqueous-phase adsorption and desorption of toluene in activated carbon fixed beds: Experiments and model. Chem. Eng. Sci. 1995, 50, 127−141. (10) Li, Z.; Bowman, R. S. Regeneration of surfactant-modified zeolite after saturation with chromate and perchloroethylene. Water Res. 2001, 35, 322−326. (11) Bowman, R. S.; Haggerty, G. M.; Huddleston, R. G.; Neel, D.; Flynn, M. M. Sorption of nonpolar organic compounds, inorganic cations, and inorganic anions by surfactant-modified zeolites. In Surfactant-Enhanced Subsurface Remediation; Sabatini, D. A., Knox, R. C., Harwell, J. H., Eds.; American Chemical Society: Washington, DC, 1995. (12) Chiou, C. T. Theoretical considerations of the partition uptake of nonionic organic compounds by soil organic matter. In Reactions and Movement of Organic Compounds in Soils; Sawhney, B. L., Brown, K., Eds.; Soil Science Society of America: Madison, WI, 1989. (13) Haggerty, G. M.; Bowman, R. S. Sorption of chromate and other inorganic ions by organo-zeolite. Environ. Sci. Technol. 1994, 28, 452− 458. (14) Su, F.; Lu, C.; Hu, S. Adsorption of benzene, toluene, ethylbenzene and p-xylene by NaOCl-oxidized carbon nanotubes. Colloids Surf., A 2010, 353, 83−91. (15) Luz, A. D.; de Souza, S. M. G. U.; Luz, C.; de Paula, R. V. R.; de Souza, A. A. U. Multicomponent adsorption and desorption of BTX compounds in coconut shell activated carbon: Experiments, mathematical modeling, and numerical simulation. Ind. Eng. Chem. Res. 2013, 52, 7896−7911. (16) Daifullah, A. A. M.; Girgis, B. S. Impact of surface characteristics of activated carbon on adsorption of BTEX. Colloids Surf. 2003, 214, 181−193. (17) Yu, F.; Ma, J.; Wu, Y. Adsorption of toluene, ethylbenzene and m-xylene on multi-walled carbon nanotubes with different oxygen contents from aqueous solutions. J. Hazard. Mater. 2011, 192, 1370− 1379. (18) Cooney, D. O. Adsorption Design for Wastewater Treatment; CRC Press: Boca Raton, FL, 1999. (19) Ayoob, S.; Gupta, A. K.; Bhakat, P. B. Analysis of breakthrough developments and modeling of fixed bed adsorption system for As(V) removal from water by modified calcined bauxite (MCB). Sep. Purif. Technol. 2007, 52, 430−438. (20) Han, R.; Zou, W.; Li, H.; Li, Y.; Shi, J. Copper(II) and lead(II) removal from aqueous solution in fixed-bed columns by manganese oxide coated zeolite. J. Hazard. Mater. 2006, 137, 934−942. (21) Malkoc, E.; Nuhoglu, Y. Fixed bed studies for the sorption of chromium (VI) onto tea factory waste. Chem. Eng. Sci. 2006, 61, 4363−4372.

4. CONCLUSIONS An experimental and numerical study of adsorption of BTEX compounds (benzene, toluene, ethylbenzene, and xylene) on modified zeolite in aqueous solution was carried out. The multicomponent Langmuir model was used to describe the equilibrium in the multicomponent system. The kinetics in a batch reactor showed that the multicomponent adsorption equilibrium was reached after 6 h. The numerical results presented a greater deviation in relation to experimental data for the multicomponent breakthrough curves, mainly for benzene and toluene, which presented lower affinity for the solid phase. With an increase in the bed height, the breakthrough times of the BTEX compounds increased. As the flow increased, the breakthrough times of the BTEX compounds underwent a decrease. The adsorption process was represented with good accuracy by the mathematical model and the numerical methodology used, allowing other situations to be simulated.



AUTHOR INFORMATION

Corresponding Author

*Tel.: +55 85 96778999. E-mail: [email protected]. Notes

The authors declare no competing financial interest.



REFERENCES

(1) Vidal, C. B.; Barros, A. L.; Moura, C. P.; De Lima, A. C. A.; Dias, F. S.; Vasconcellos, L. C. G.; Fechine, P. B. A.; Nascimento, R. F. Adsorption of polycyclic aromatic hydrocarbons from aqueous solutions by modified periodic mesoporousorganosilica. J. Colloid Interface Sci. 2011, 357, 466−473. (2) Vidal, C. B.; Raulino, G. S. C.; Barros, A. L.; Lima, A. C. A.; Ribeiro, J. P.; Pires, M. J. R.; Nascimento, R. F. BTEX removal from aqueous solutions by HDTMA-modified Y zeolite. J. Environ. Manage. 2012, 112, 178−185. (3) Ranck, J. M.; Bowman, R. S.; Weeber, J. L.; Katz, L. E.; Sullivan, E. J. BTEX removal from produced water using surfactant-modified zeolite. J. Environ. Eng. 2005, 131, 434−442. (4) Sulaymon, A. H.; Ahmed, K. W. Competitive adsorption of furfural and phenolic compounds onto activated carbon in fixed bed column. Environ. Sci. Technol. 2008, 42, 392−397. (5) Zhang, G.; Shen, K.; Zhao, D.; Yuan, Y.; Wang, G. Preparation of uncoated iron oxide nanoparticles by thermal decarboxylation of iron 288

dx.doi.org/10.1021/je400780f | J. Chem. Eng. Data 2014, 59, 282−288