Removal of Atenolol and Isoproturon in Aqueous Solutions by

Article pubs.acs.org/IECR

Removal of Atenolol and Isoproturon in Aqueous Solutions by Adsorption in a Fixed-Bed Column José Luis Sotelo,* Gabriel Ovejero, Araceli Rodríguez, Silvia Á lvarez, and Juan García* 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 ABSTRACT: In this work, the removal of a β-blocker, atenolol, and a pesticide, isoproturon, from aqueous solutions by adsorption on granular activated carbon fixed-bed columns has been studied. The effect of important operation parameters on breakthrough curves as column length (Z = 1.0−3.0 cm), initial concentration of atenolol or isoproturon (C0 = 50.0−500.0 μg·L−1), and volumetric flow rate (Q = 1.5 mL·min−1) was studied. Breakthrough time of the bed was found to increase with an increase in the value of the column length, and with a decrease in the value of initial concentration or flow rate. Mathematical models, as Bohart−Adams, Thomas, Yoon−Nelson, and Clark were applied to the experimental data for the prediction of the breakthrough curves, and to determine the characteristic parameters of the bed.

1. INTRODUCTION There has been an increasing concern in recent years about the allocation and possible adverse effects on human health of the emerging contaminants in the aquatic environment. They are present in treated wastewater at trace levels (μg·L−1 to ng·L−1), and include personal care products, pharmaceuticals, pesticides, surfactants, flame retardants, etc. Some of the most widely and frequently used drug classes are employed in quantities similar to those of pesticides and in some countries some are even sold without the requirement of a prescription.1 In this sense, it is estimated that hundreds of tons of pharmaceutical compounds are produced and consumed in developed countries each year.2,3 The lack of validated analytical methods, nonuniform monitoring data, and the lack of definite information about the fate and effects of these compounds and/or their metabolites and transformation byproduct in the aquatic environment make accurate risk assessments problematic. The full extent and consequences of the presence of these compounds in the environment are therefore, still largely unknown.1 These compounds may occur in wastewater treatment plants effluents because they are truly persistent under the conditions of a conventional activated sludge process or because their microbial degradation was not fast enough to be completed within a low retention time. An incomplete degradation implies, however, that these compounds could be present after its discharge in the receiving water body.4 Therefore, these pharmaceuticals can persist in the environment and either via the food chain or via drinking water, can make their way back to humans. It is also accepted that some of these compounds are beginning to be associated with adverse developmental effects in aquatic organisms at environmentally relevant concentrations.1 One of these mentioned micropollutants are pesticides, which are widely used not only in agriculture but also in domestic and industrial activities. This fact resulted in the © 2012 American Chemical Society

presence of residues of these products and their metabolites in the environment.5 Some of the most commercially important herbicides belong to the urea family, containing over 20 related compounds, as chlorotoluron, diuron, or isoproturon, the latter being one of the target compounds of this work. Isoproturon is watersoluble, moderately hydrophobic, and weakly adsorbed by soils. It has been reported that concentrations in ground and surface water exceeded the limit levels.6 In this sense, many regulatory organisms, like the European Commission (EC), have adopted strict regulations trying to minimize the negative effects in the environment produced by these compounds. In the water policy field, the European Union has established different directives such as Water Framework Directive 2000/60/EC, the objective of which is to protect and prevent water quality.5,7 In 2008, the Directive 2008/105/EC was established, which include a list of 33 priority compounds in water to be controlled, where the third part of the list is pesticides, and one is isoproturon.5 On the other hand, among the several pharmaceutical compounds present in the environment, β-blockers constitute one therapeutic class of pharmaceuticals that is not effectively removed during wastewater treatment processes. These pharmaceuticals are used for various purposes, mainly for the treatment of cardiac malfunctions such as arrhythmia and hypertension, as well as post-treatment after a myocardial infarction.8 Its presence was reported in groundwater at concentrations between ng·L−1 and μg·L−1 concentrations. The adsorption technique is widely popular due to its simplicity, no use of chemical reactants, and the availability of a wide range of adsorbents. Therefore, it has been shown to be the most promising option for the removal of these nonbiodegradable organics from aqueous streams. Activated Received: Revised: Accepted: Published: 5045

February 7, 2012 March 5, 2012 March 10, 2012 March 10, 2012 dx.doi.org/10.1021/ie300334q | Ind. Eng. Chem. Res. 2012, 51, 5045−5055

Industrial & Engineering Chemistry Research

Article

Table 1. Operation Conditions for Atenolol and Isoproturon Adsorption in Fixed-Bed Column initial concentration (μg·L−1)

column length (cm) different initial concentration different volumetric flow rates different column lengths

volumetric flow rate (mL·min−1)

atenolol

isoproturon

atenolol

isoproturon

atenolol

isoproturon

2.0 2.0 2.0 2.0 1.0 2.0 3.0

2.0 2.0 2.0 2.0 1.0 2.0 3.0

300.0 500.0 100.0 100.0 100.0 100.0 100.0

50.0 150.0 100.0 100.0 200.0 200.0 200.0

1.5 1.5 2.0 3.0 1.5 1.5 1.5

1.5 1.5 2.0 3.0 1.5 1.5 1.5

batches. The eluate from the elution step was evaporated to dryness under a stream of nitrogen and the dry residue was reconstituted in 1.0 mL 20:80 (v/v) acetonitrile/phosphate buffer solution (pH 7) for atenolol and 60:40 (v/v) acetonitrile/ultrapure water for isoproturon. The extract was filtered through a 0.45 μm filter into a glass sample vial. The cartridge can be regenerated by washing thoroughly with acetonitrile and reused multiple times. 2.4. Adsorbent Characterization. Textural characterization of activated carbon was done by using N 2 adsorption−desorption at 77 K in a Micromeritics ASAP 2010 apparatus, and mercury intrusion porosimetry in a Thermo Finnigan Pascal 140−440. Thermogravimetric analysis (TGA) experiments were performed with a heating rate of 10 °C/min in inert atmosphere on a Seiko EXSTAR 6000 TGA Instrument, from 20 to 900 °C, with 30 mL min−1 helium flow rate. More details about the adsorbent characterization can be consulted elsewhere.11 2.5. Adsorption Experiments. Batch equilibrium experiments were conducted using conical flasks (250 mL) immersed in a water thermostatic bath at 30 ± 1 °C with both adsorbates. The suspensions containing different doses of activated carbon and the solutions of atenolol or isoproturon were shaken with a magnetic stirrer at constant temperature until equilibrium was reached. The equilibrium concentrations of each solution were determined. The amount of atenolol or isoproturon adsorbed on activated carbon was determined by the difference between the initial and remaining concentrations in solution at equilibrium time. The amount of adsorbed atenolol or isoproturon at equilibrium, qeq (mg·g−1) was calculated by the following expression:

carbon is the most common adsorbent for this type of process due to its effectiveness, versatility, and great adsorption capacity, which is mainly determined by its porous structure and its surface functional groups.9,10 Studies about micropollutants treatment by adsorption in fixed-bed column are few. To our knowledge, this is one of the first studies on the elimination of isoproturon and atenolol by activated carbon in fixed-bed column. In the present study, the effect of various operating variables as the inlet adsorbate concentration, flow rate, and bed length is studied. Several empirical models have been applied to predict the dynamic column adsorption, as Bohart−Adams, Thomas, Wolborska, and Yoon−Nelson.

2. MATERIALS AND METHODS 2.1. Adsorbent Material. Granular activated carbon (Filtrasorb 400) was used in this study (supplied by Calgon, France). Before use, the adsorbent was washed with water to remove surface impurities, followed by drying at 100 °C for 48 h. For the experiments, the size fraction between 0.5 to 0.589 mm was selected (ρP = 453.6 g·L−1, εP = 0.410). 2.2. Pharmaceutical Compounds. Atenolol and isoproturon (analytical grade) were purchased from Sigma-Aldrich (Steinheim, Germany), and used in the experiments directly without any further purification. Solutions of both compounds of the appropriate concentrations were prepared by diluting a stock solution. 2.3. Analytical Technique. All analyses were carried out by high-performance liquid chromatography technique, HPLC, using a chromatograph Varian ProStar 230 equipped with a UV−vis PDA detector under the following conditions: Mediterranea column C18 (4.6 mm i.d. × 250 mm., 5 μm particle size): 100 μL aliquots were injected into the chromatograph. For the atenolol analysis, the isocratic mobile phase was a 20:80 (v/v) acetonitrile/phosphate buffer solution (pH 7) at a flow rate of 1.0 mL·min−1. For the isoproturon analysis, the mobile phase used was a 60:40 (v/v) acetonitrile/ water at the same flow rate. Because of the low concentrations that these contaminants exhibit in natural water and wastewater (at μg·L−1 level), it was necessary to preconcentrate the samples by solid-phase extraction (SPE) on SEP Oasis C18 Waters cartridges (60 mg, 3 cm3) previous to their analysis. The cartridges were connected to an SPE manifold, which was connected to a vacuum pump. Prior to the extraction, the sorbent was conditioned by rinsing with 2.5 mL of methanol and 2.5 mL of ultrapure water. The sample solution (10.0 mL, adjusted to pH 8.0) was passed through the cartridge at a flow rate of 0.5−1.0 mL·min−1. The cartridge was then washed with 2.0 mL of ultrapure water to remove the coadsorbed matrix materials from the cartridge. The analytes retained on the sorbent were eluted with 4.0 mL of methanol, loaded in two

qeq =

(C0 − Ceq)V W

(1)

where Co and Ceq (mg·L−1) are the liquid-phase initial and equilibrium concentrations of atenolol or isoproturon, respectively. V is the volume of the solution (L), and W is the mass of adsorbent (g). The adsorption of atenolol and isoproturon on the activated carbon were also evaluated at constant temperature (30 °C) for the adsorption isotherms. Fixed-bed experiments were conducted using borosilicate glass columns of 6 mm i.d. and 30 cm length. The column was packed with the granular activated carbon and then filled with a layer of glass balls (1 mm in diameter) to compact the mass of adsorbent and to avoid dead volumes. The influent to the column was pumped using a Dinko multichannel peristaltic pump, model D25 V. Atenolol solutions with concentration in the range of 100− 500 μg·L−1 and volumetric flow rates in the range of 1.5−3.0 5046

dx.doi.org/10.1021/ie300334q | Ind. Eng. Chem. Res. 2012, 51, 5045−5055


Article

mL·min−1 were passed in the down-flow mode through the bed. In the case of isoproturon, the conditions were concentration in the range of 50−200 μg·L−1 and volumetric flow rate between 1.5 and 3.0 mL·min−1. The effluent was collected at time intervals, and its concentration was determined by HPLC. All experiments were performed at 25 ± 1 °C using ultrapure water at the pH value of the solution itself. Table 1 summarize the operating conditions for the atenolol and isoproturon experiments, respectively, which have been developed modifying the following parameters: initial influent concentration, flow rate, and weight of adsorbent or bed depth.

3. RESULTS AND DISCUSSION 3.1. Characterization of Granular Activated Carbon. The physicochemical properties of granular activated carbon determined in this study are shown in Table 2. This material

Figure 2. Adsorption isotherm of atenolol and isoproturon.

corresponds to type L3, indicative of a high affinity between the adsorbate and the adsorbent for low concentrations, which decreases as the concentration increases. These isotherms are characterized by a decrease in the slope of the curve with an increase in the concentration of adsorbate in the solution, due to a decrease of the available adsorption sites. BET model adsorption is used to describe this isotherm. This equation fit correctly the experimental isotherm data of the granular activated carbon. The estimated BET model parameters and the correlation coefficient are reported in Table 3.

Table 2. Physicochemical Properties of the Granular Activated Carbon F-400 parameters

value

surface area (m2·g−1) SBET (m2·g−1) micropore volume (cm3·g−1) pHPZC basicity (μeq·g−1) acidity (μeq·g−1)

384.0 997.0 0.26 7.6 462.0 802.0

exhibited a narrow pore size distribution and was essentially microporous. On the other hand, at least two different stages can be observed in the thermogravimetric analysis. The first major thermal effect, within the 475−500 °C range, can be attributed to the decomposition of surface oxygen groups such as ketones, ethers, and hydroxyls originally present in the structure of the granular activated carbon.11 Furthermore, the point of zero charge, pHPZC, of this material was obtained. The results indicated that the pHPZC was 7.6. 3.2. Batch Adsorption Experiments. For atenolol, the equilibrium state was considered reached after about 25 h, since the variations in equilibrium adsorption capacity did not change more than 5% (Figure 1). Therefore, the adsorption equilibrium capacity of the atenolol on the granular activated carbon versus the concentration of the aqueous solution at equilibrium is given in Figure 2. The isotherm, according to the Giles classification,

Table 3. BET and Freundlich Model Parameters Related to the Adsorption Isotherm of Atenolol and Isoproturon parameters

value BET Model Atenolol

qsat K C0 R2

80.4 16.7 115.0 0.9962 Freundlich Model Isoproturon

KF 1/n R2

282.2 7.3 0.9999

In the case of isoproturon, the equilibrium state was reached after about 300 h (Figure 1). This contaminant, as seen in Figure 2, presents a type of isotherm that can be fitted to Freundlich model (the R2 value for linear fit is 0.9999). This isotherm is L1-type according to the Giles classification, indicating a high affinity adsorbent−adsorbate and suggesting that isoproturon molecules are adsorbed in parallel to the carbon surface and that there is no major competition between adsorbate and water molecules for the active adsorption sites on the activated carbon. The Freundlich model parameters are shown in Table 3. 3.3. Fixed-Bed Adsorption Experiments. In the set of experiments reported in Table 1, the influence of the operation parameters, initial atenolol or isoproturon concentration, flow rate, and column length on the breakthrough curves were evaluated. 3.3.1. Effect of the Inlet Concentration. In the first case, the adsorption of atenolol by fixed bed was tested at various atenolol inlet concentrations. The breakthrough curves were obtained at initial atenolol concentrations from 300 to 500

Figure 1. Equilibrium time for atenolol and isoproturon. 5047



Article

Figure 3. Breakthrough curves of atenolol removal by granular activated carbon fixed-bed columns of (a) different initial atenolol concentrations (column length = 2.0 cm, flow rate = 1.5 mL·min−1); (b) different volumetric flow rates (column length = 2.0 cm, initial concn = 100 μg·L−1); (c) different column lengths (inlet atenolol concn = 100 μg·L−1, flow rate = 1.5 mL·min−1).

Figure 4. Breakthrough curves of isoproturon removal by granular activated carbon fixed-bed columns of (a) different initial isoproturon concentrations (column length = 2.0 cm, flow rate = 1.5 mL·min−1); (b) different volumetric flow rates (column length = 2.0 cm, initial concn = 100 μg·L−1); (c) different column lengths (inlet isoproturon concn = 200 μg·L−1, flow rate = 1.5 mL·min−1).

μg·L−1, flow rate of 1.5 mL·min−1, and a length of bed of 2.0 cm (Figure 3a). On the other hand, in the case of isoproturon, the breakthrough curves were obtained at initial isoproturon concentrations of 50.0 and 150.0 μg·L−1. The column length used was of 2.0 cm and the flow rate was 1.5 mL·min−1. The breakthrough curves obtained are shown in Figure 4a. In both cases, a decrease in the inlet concentration leads to an increase in the breakthrough time, as in this case the binding sites became more slowly saturated in the system. Besides, a decrease in the initial concentration gave a lower slope of the curve, which indicates a slower mass transport due to a decreased diffusion coefficient or decreased mass transfer coefficient. For atenolol, breakthrough (in all cases, considered at C/C0 = 0.15) occurred after 70.5 h at 300 μg·L−1 inlet atenolol concentration while breakthrough appeared after 23.7 h for an initial atenolol concentration of 500 μg·L−1. Therefore, in the case of isoproturon, breakthrough point occurred after 5.8 h at 150.0 μg·L−1 inlet isoproturon concentration and 309.3 h at 50.0 μg·L−1 inlet concentration. 3.3.2. Effect of the Flow Rate. In the atenolol removal studies in a fixed-bed column, the flow rate was between 2.0 and 3.0 mL·min−1, while the atenolol concentration and the bed depth were held constant at 100 μg·L−1 and 2.0 cm respectively. The breakthrough curves at different flow rates are given in Figure 3b.

For isoproturon, the effect of flow rate on isoproturon adsorption removal was tested at 2.0 mL·min−1 and 3.0 mL·min−1, using an initial isoproturon concentration of 100.0 μg·L−1 and a column length 2.0 cm. The breakthrough curves obtained are shown in Figure 4b. As expected, it can be seen in the figures that, at the higher flow rate, a decrease of the breakthrough time was observed. This can be due to the insufficient or limited residence time of the adsorbate in the column.12 In both cases, it can be observed that the breakthrough curve obtained at the higher flow rate, 3.0 mL·min−1, presents a higher slope, which indicates a decrease in the mass transfer resistance of the process. For atenolol, breakthrough (in both cases, considered at C/ C0 = 0.15) occurred after 171.6 h for 2.0 mL·min−1 flow rate, and after 123.8 h for 3.0 mL·min−1 flow rate. In the case of isoproturon, breakthrough occurred after 266.6 h at flow rate of 2.0 mL·min−1, while the breakthrough time appeared after 20.0 h in the case of a flow rate of 3.0 mL·min−1. 3.3.3. Effect of the Column Length. In the case of atenolol removal, the breakthrough curves were examined at different column lengths, from 1.0 to 3.0 cm, the initial concentration being held constant at 100 μg·L−1 and the flow rate at 1.5 mL·min−1. The results are shown in Figure 3c. For isoproturon, the breakthrough curves were obtained at the same column lengths, from 1.0 to 3.0 cm, using a constant flow rate of 1.5 mL·min−1 and an initial isoproturon concentration of 200 μg·L−1. The results can be seen in Figure 4c. 5048



Article

Table 4. Adsorption Capacities (qb, qs), MTZ, FBU, and Removal Percentage Values for Atenolol Adsorption initial concentration (μg·L−1)

column length (cm) parameter −1

qb (mg·g ) qs (mg·g−1) MTZ (cm) FBU removal percentage (%)

volumetric flow rate (mL·min‑1)

1.0

2.0

3.0

300.0

500.0

2.0

3.0

2.15 14.40 0.85 0.15 25.85

5.77 11.19 0.97 0.52 38.25

7.63 10.64 0.85 0.72 54.57

7.59 26.83 1.43 0.28 36.13

4.52 34.44 1.74 0.13 28.42

8.65 16.35 0.94 0.53 47.81

8.88 17.91 1.00 0.50 34.60

Table 5. Adsorption Capacities (qb, qs), MTZ, FBU, and Removal Percentage Values for Isoproturon Adsorption initial concentration (μg·L−1)

column length (cm)

volumetric flow rate (mL·min−1)

parameter

1.0

2.0

3.0

50.0

150.0

2.0

3.0

qb (mg·g‑1) qs (mg·g‑1) MTZ (cm) FBU removal percentage (%)

0.20 21.64 0.99 0.01 17.03

1.75 12.95 1.73 0.14 17.64

5.69 9.59 1.22 0.59 19.56

7.70 16.51 1.07 0.47 50.50

0.46 19.50 1.95 0.02 33.33

13.38 19.98 0.66 0.67 46.25

1.52 15.94 1.81 0.10 29.52

(g), Q is the volumetric flow rate (mL·min−1), C and C0 are the atenolol or isoproturon final and initial concentrations (mg·L−1), tb is the breakthrough time (h), and ts is the saturation time (h). The values of adsorption capacities at breakthrough and saturation time for atenolol and isoproturon are shown in Tables 4 and 5, respectively. Therefore, the MTZ may then be calculated by the ratio qb/ qs according to eq 4. This equation has a maximum value that corresponds to the bedś height (Z) and, as the mass transfer resistance decreases, this value leads to the ideal condition in which the MTZ value is zero and the breakthrough curve is a step-function.14

In both cases, it can be seen, as expected, that the higher bed length, 3.0 cm, leads to the higher breakthrough time, this parameter being lower when the column length is 1.0 and 2.0 cm, respectively. It can be observed that the slopes of the breakthrough curve are roughly similar, as a change of the column length, at the same concentration and flow rate, does not affect the mass transfer of the process. At this column length values, the difference between the slopes which is observed in Figures 3c and 4c is due to the front of the concentration in the bed being not fully developed.13 For atenolol, breakthrough (considered at C/C0 = 0.15) occurred after 24.3 h for 1.0 cm of bed length, 153.4 h for 2.0 cm of column length, and 316.0 h for a column length of 3.0 cm. For isoproturon, a breakthrough point (considered at C/C0 = 0.35) appeared after 1.3 h for 1.0 cm of bed length, 33.7 h for 2.0 cm of column length, and 188.9 h for 3.0 cm of column length. 3.4. Adsorption Parameters Estimation. For both adsorbates, atenolol, and isoproturon, it can be observed that the operational conditions of the process influence the mass transfer resistance and therefore change the adsorption parameters as adsorption capacities, mass transfer zone, and bed utilization values. Adsorption capacities at breakthrough time (qb) and at saturation time (qs), length of the mass transfer zone (MTZ), fractional bed utilization (FBU), and percentages of removal are important parameters, which directly affect the feasibility and economics of the sorption process. These values are reported in Table 4 for atenolol and Table 5 for isoproturon. Equations 2 and 3 were obtained through mass balance in the column by using the column’s saturation data, where the area below the curve at breakthrough time is proportional to qb and at bed saturation time is proportional to qs, according to the Geankoplis model:14 t ⎛

⎛ q ⎞ MTZ = Z ⎜⎜1 − b ⎟⎟ qs ⎠ ⎝

As can be observed in Table 4 and Table 5, a variation in the column length leads to a variation in the length of the mass transfer zone, MTZ. This is due to the column lengths tested being not high enough to have a fully developed profile, which is called the constant pattern behavior.13,15 Also, it can be seen that an increase of the initial atenolol or isoproturon concentration leads to a worse fractional bed utilization of the bed and to a higher adsorption capacity at saturation time, as qs, 34.4 mg·g−1 for 500.0 μg·L−1 of atenolol and 19.5 mg·g−1 for 150.0 μg·L−1 of isoproturon. The driving force for adsorption is the adsorbate concentration difference between the adsorbent surface and the solution. A high concentration difference provides a high driving force for the adsorption process, and this may explain why higher adsorption capacities were achieved in the column with a higher adsorbate concentration.12 3.5.1. Modeling of the Breakthrough Curves. The Bohart−Adams Model. Although this model16 was initially developed to describe the adsorption of chloride on charcoal, nowadays it is used for the quantitative description of other adsorption systems. This model was established based on the surface reaction theory, and it is assumed that equilibrium is not instantaneous. Therefore, the rate of adsorption was proportional to both the residual capacity of the activated carbon and the concentration of the adsorbates. Generally, the Bohart− Adams model is used for the description of the initial part of

⎞

qb =

C0Q m

∫0 b ⎜⎝1 − CC ⎟⎠ dt

qs =

C0Q m

∫0 s ⎜⎝1 − CC ⎟⎠ dt

0

t ⎛

(2)

⎞

0

(4)

(3)

where qb and qs are the breakthrough and saturation time capacities, respectively (mg·g−1), m is the mass of adsorbent 5049



Article

Figure 5. Time for breakthrough compared to column length for (a) atenolol and (b) isoproturon adsorption on GAC fixed-bed columns according to the Bohart−Adams model.

t = 93.8x − 112.9

the breakthrough curve, which relates C/C0 to time, t, for a continuous-flow adsorber column.12,17 Hutchins, linearized the Bohart−Adams equation and proposed a relationship between the time and the column length (Z), which is called bed-depth service time (BDST) model:18 t=

⎞ ⎛C N0 1 Z− ln⎜ 0 − 1⎟ C0U C0kAB ⎝ Cb ⎠

In general, a plot of t0.5 (the time at 50% breakthrough) versus Z (column length) should be a straight line passing through the origin. However, in both cases, the straight line does not pass through the origin. This behavior indicates that the adsorption of these compounds on activated carbon occurs through a complex mechanism and more than one rate-limiting step are involved in the adsorption process.19 So, a simplified model as Bohart−Adams is not suitable to define complex adsorption mechanisms, as organic compounds on activated carbon.19,20 The adsorption capacity (N0) as calculated from the slope of the 50% plot using eq 6 was 14.16 mg·g−1 for atenolol and 11.63 mg·g−1 for isoproturon. Therefore, N0 and k values as calculated from the slope and the intercept of the 15% plot are 10.24 mg·g−1 and 0.136 L·mg−1·h−1, for atenolol and 13.16 mg·g−1 and 0.027 L·mg−1·h−1 for isoproturon (at 35% breakthrough). 3.5.2. The Wolborska Model. The Wolborska model is used for the description of adsorption dynamics using mass transfer equations for diffusion mechanisms in the range of the lowconcentration breakthrough curve.21 The Wolborska model is given by the following equation:

(5)

where C0 is the initial concentration of adsorbate (mg·L−1), Cb is the desired concentration of adsorbate at breakthrough time (mg·L−1), kAB is the adsorption rate constant (L·mg−1·h−1), N0 is the adsorption capacity of the system (mg·L−1), Z is the column length (cm), U is the linear flow velocity (cm·h−1), and t is the service time of column under above conditions (h). Equation 2 can be defined by the next parameters:

a = slope =

N0 C0U

b = intercept =

⎞ ⎛C 1 ln⎜ 0 − 1⎟ C0kAB ⎝ Cb ⎠

(6)

(7)

When the concentration is half of the initial value, at 50% breakthrough, the values of C/C0 and t are 0.50 and t0.5, respectively. Therefore, the last term in eq 6 becomes 0, and the half-time t0.5 can be obtained as t0.5 =

N0 Z C0U

ln

β=

Figure 5 panels a and b show the Bohart−Adams plot (service time vs column length) for the removal of atenolol (Figure 5a) and isoproturon (Figure 5b) on a fixed-bed column at 15% and 50% breakthrough for atenolol, and at 35% and 50% breakthrough for isoproturon. The linear equations obtained in the case of atenolol, for 15% and 50% breakthrough, are next: (9)

t = 145.9x − 127.1

(10)

(13)

⎞ 4β0D U 2 ⎛⎜ ⎟⎟ 1 1 + − 2D ⎜⎝ U2 ⎠

(14)

where β is the kinetic coefficient of the external mass transfer (h−1), D is the axial diffusion coefficient (cm·h−1), and β0 is the external mass-transfer coefficient with a negligible axial dispersion coefficient, D, N0 is the adsorption capacity of the system (mg·L−1), U is the linear flow velocity (cm·h−1), and Z is the length of the fixed bed (cm). Wolborska observed that for short beds or high flow rates, the axial diffusion is negligible and β = β0, the external mass transfer coefficient.22 For atenolol and isoproturon cases, after applying eq 13 and the nonlinearized expression of eq 13, a good relationship between experimental and theoretical data (R2 < 0.6) could not be obtained, even in the region below to 50% saturation, for all

Therefore, for isoproturon, the equations obtained by Bohart− Adams plot are these (for 35% and 50% breakthrough): t = 82.9x + 66.2

βC0 βZ C = t− C0 N0 U

with

(8)

t = 201.7x − 128.1

(12)

(11) 5050



Article

Table 6. Predicted Parameters for Thomas Model and Model Deviations for Atenolol Adsorption Z (cm)

C0 (μg·L−1)

Q (mL·min−1)

kT (L·h−1·mg−1)

q0 exp (mg·g−1)

q0 cal (mg·g−1)

SE

MPSD

1.0 2.0 3.0 2.0 2.0 2.0 2.0

100.0 100.0 100.0 100.0 100.0 300.0 500.0

1.5 1.5 1.5 2.0 3.0 1.5 1.5

0.128 0.098 0.137 0.089 0.085 0.027 0.015

6.95 9.38 10.64 16.29 17.91 24.81 24.96

13.92 13.83 14.36 20.41 28.23 44.36 51.10

1.317 0.663 0.567 0.577 2.064 3.834 4.772

38.534 46.207 61.084 61.657 46.481 49.722 36.440

Table 7. Predicted Parameters for Thomas Model and Model Deviations for Isoproturon Adsorption Z (cm)

C0 (μg·L−1)

Q (mL·min−1)

kT (L·h−1·mg−1)

q0 exp (mg·g−1)

q0 cal (mg·g−1)

SE

MPSD

1.0 2.0 3.0 2.0 2.0 2.0 2.0

200.0 200.0 200.0 100.0 100.0 50.0 150.0

1.5 1.5 1.5 2.0 3.0 1.5 1.5

0.048 0.031 0.023 0.108 0.092 0.103 0.057

21.64 12.95 9.59 19.97 13.69 16.51 16.75

21.63 18.43 22.78 25.31 20.75 15.57 18.86

0.002 0.866 2.112 0.748 1.211 0.128 0.422

12.885 38.369 43.198 139.167 40.870 62.155 31.328

Figure 6. Experimental and predicted breakthrough curves of atenolol removal by granular activated carbon fixed-bed columns predicted by the Thomas model: (a) different initial atenolol concentrations (column length = 2.0 cm, flow rate = 1.5 mL·min−1); (b) different volumetric flow rates (column length = 2.0 cm; initial concn = 100 μg·L−1); (c) different column lengths (initial atenolol concn = 100 μg·L−1, flow rate = 1.5 mL·min−1).

Thomas model parameters, kT and q0, were determined by nonlinearized expression of eq 15 and are shown in Table 6 and Table 7 for atenolol and isoproturon, respectively. Therefore, adsorption capacity (q0) values calculated by this model were compared with experimental data using standard error of estimate (SE) method, with this expression:

breakthrough curves (varying inlet atenolol or isoproturon concentration, flow rate, and column length). 3.5.3. The Thomas Model. Developed by Thomas23 in 1944, this kinetic model is one of the most general and widely used. It assumes a Langmuir isotherm for equilibrium and a rate driving force obeying a second-order reversible reaction kinetics. This model is suitable for adsorption processes where the external and internal diffusion limitations are absent. Application of this model leads to some error in adsorption processes where firstorder reaction kinetics is followed.19,24 The adsorption capacity of an adsorbent is one of the parameters needed for the successful design of the adsorption process, and the Thomas model is used to accomplish this purpose.24 The linearized expression of the model has the following form: ⎛C ⎞ kTq0m ln⎜ − 1⎟ = − kTC0t Q ⎝ C0 ⎠

SE =

∑

(q0(expt) − q0(calcd))2 N

(16)

The estimation of error between the experimental and predicted values of C/C0 was done by using the modified form of the Marquardt’s percent standard deviation (MPSD) represented by eq 17 and the values in Tables 6 and 7. MPSD = 100

1 N−P

n ⎛ (C /C ) ⎞2 0 exp − (C /C0)calcd

∑ ⎜⎜

i=1 ⎝

(C /C0)expt

⎟ ⎟ ⎠i

(15)

(17)

where C0 is the initial concentration of atenolol or isoproturon (mg·L−1), q0 is the adsorption capacity of the system adsorbate−adsorbent (mg·L−1), kT is the Thomas rate constant (L·h−1·mg−1), m is the mass of the adsorbent in the column (g), and Q is the volumetric flow rate (L·h−1).

For atenolol, and as expected (Table 3), it can be observed from Table 6 that the values of q0 increased with an increase in the value of C0 and the volumetric flow rate. Therefore, for isoproturon, as it can be seen in Table 7, the values of q0 increased with an increase in the value of C0 and decreased with an increase in the value of the flow rate. 5051



Article

Figure 7. Experimental and predicted breakthrough curves of isoproturon removal by granular activated carbon fixed-bed columns predicted by the Thomas model: (a) different initial isoproturon concentrations (column length = 2.0 cm, flow rate = 1.5 mL·min−1); (b) different volumetric flow rates (column length = 2.0 cm, initial concn = 100 μg·L−1); (c) different column lengths (initial isoproturon concn = 200 μg·L−1, flow rate = 1.5 mL·min−1).

Table 8. Predicted Parameters for Yoon−Nelson Model and Model Deviations for Atenolol Adsorption Z (cm)

C0 (μg·L−1)

Q (mL·min−1)

kYN (h−1)

τ exp (h)

τ cal (h)

SE

MPSD

1.0 2.0 3.0 2.0 2.0 2.0 2.0

100.0 100.0 100.0 100.0 100.0 300.0 500.0

1.5 1.5 1.5 2.0 3.0 1.5 1.5

0.014 0.010 0.013 0.009 0.010 0.008 0.0077

75.00 305.75 478.45 332.06 233.50 262.65 123.00

137.72 307.32 479.12 347.82 276.30 316.90 227.09

11.853 0.234 0.102 2.207 8.560 10.639 19.004

37.885 46.207 66.805 56.794 45.195 57.416 36.441

Table 9. Predicted Parameters for Yoon−Nelson Model and Model Deviations for Isoproturon Adsorption Z (cm)

C0 (μg·L−1)

Q (mL·min−1)

kYN (h−1)

τ exp (h)

τ cal (h)

SE

MPSD

1.0 2.0 3.0 2.0 2.0 2.0 2.0

200.0 200.0 200.0 100.0 100.0 50.0 150.0

1.5 1.5 1.5 2.0 3.0 1.5 1.5

0.010 0.006 0.005 0.011 0.009 0.005 0.008

133.15 263.68 298.87 405.11 260.50 620.14 306.50

112.63 191.93 356.00 421.90 230.59 677.92 261.98

2.902 11.345 9.148 2.351 5.130 7.863 8.904

32.167 65.589 66.469 139.167 3.087 60.027 31.328

Figure 8. Experimental and predicted breakthrough curves of atenolol removal by granular activated carbon fixed-bed columns predicted by the Yoon−Nelson model: (a) different initial atenolol concentrations (column length = 2.0 cm, flow rate = 1.5 mL·min−1); (b) different volumetric flow rates (column length = 2.0 cm, initial concn = 100 μg·L−1); (c) different column lengths (initial atenolol concn = 100 μg·L−1, flow rate = 1.5 mL·min−1).

In Figures 6 (a−c) and 7 (a−c) the experimental and predicted breakthrough curves of atenolol and isoproturon removal by granular activated carbon packed columns using the Thomas model can be observed. 3.5.4. The Yoon−Nelson Model. The Yoon−Nelson model (1984) was originally focused on the adsorption of vapors or gases in activated coal. This is a relatively simple model which

assumes that the rate of decrease in the probability of adsorption for each adsorbate molecule was proportional to the probability of adsorbate adsorption and the probability of adsorbate breakthrough on the adsorbent.25,26 The Yoon− Nelson model is not only less complicated than other models, but also requires no detailed data concerning the characteristics 5052



Article

Figure 9. Experimental and predicted breakthrough curves of isoproturon removal by granular activated carbon fixed-bed columns predicted by the Yoon−Nelson model: (a) different initial isoproturon concentrations (column length = 2.0 cm, flow rate = 1.5 mL·min−1); (b) different volumetric flow rates (column length = 2.0 cm, initial concn = 100 μg·L−1); (c) different column lengths (initial isoproturon concn = 200 μg·L−1, flow rate = 1.5 mL·min−1).

For a particular adsorption process in fixed-bed column, the values of A and r can be determined from eq 19, thereby enabling the prediction of the breakthrough curve.28 Only for isoproturon, was the Freundlich constant, n = 7.3 (Table 3), used in the Clark model to calculate the model parameters A and r for the adsorption of this micropollutant on granular activated carbon on a fixed-bed column. These values and Marquardt’s percent standard deviation (MPSD) are shown in Table 10. Also, the experimental and theoretical

of the adsorbate and adsorbent and the physical properties of the adsorption bed.12,19,24 The Yoon−Nelson equation regarding a single-component system is expressed as ⎛ C ⎞ ln⎜ ⎟ = kYNt − τkYN ⎝ C0 − C ⎠

(18)

where C0 is the initial concentration of atenolol or isoproturon (mg·L−1), kYN is the Yoon−Nelson rate constant (h−1) and τ is the time required for reach 50% adsorbate breakthrough (h). The calculation of theoretical breakthrough curves requires the determination of the parameters kYN and τ for both compounds, which are reported in Table 8 and Table 9, for atenolol and isoproturon, respectively. The standard error of estimate (SE) method, used to evaluate the difference between predicted and experimental τ values, eq 16, and MPSD values are both reported in Tables 8 and 9. Figures 8 (a−c) and 9 (a−c) show the experimental and theoretical breakthrough curves, for atenolol and isoproturon, respectively, obtained at different inlet concentrations, flow rates, and column lengths using the Yoon−Nelson model. The figures and the data reported in Tables 8 and 9 indicated that the τ values predicted by the Yoon−Nelson model are, in general, relatively close to experimental results. 3.5.5. The Clark Model. Clark27 used the mass transfer coefficient concept in combination with the Freundlich isotherm to define a new relation for the breakthrough curve. The linearized expression of this model is ⎡⎛ C ⎞n − 1 ⎤ ln⎢⎜ 0 ⎟ − 1⎥ = −rt + ln A ⎢⎣⎝ C ⎠ ⎥⎦

Table 10. Predicted Parameters for Clark Model and Model Deviations for Isoproturon Adsorption

β νm(n − 1) U

Q (mL·min‑1)

r (h‑1)

A

MPSD

1.0 2.0 3.0 2.0 2.0 2.0 2.0

200.0 200.0 200.0 100.0 100.0 50.0 150.0

1.5 1.5 1.5 2.0 3.0 1.5 1.5

−0.052 0.055 0.044 0.039 0.031 0.018 0.030

543281.98 34657943.12 85239254.80 3376557626.00 219995.10 42138071.01 661511.85

22.106 35.967 47.002 181.024 43.663 69.734 28.757

4. CONCLUSIONS The present work is a study of the adsorption of atenolol and isoproturon, two contaminants that are representative of the group called emerging compounds, from aqueous solutions on a fixed-bed column of granular activated carbon. This technique has proven to be highly interesting and effective in both cases, even at μg·L−1 levels. The influence of the column length, initial adsorbate concentration, and volumetric flow rate on the shape of the breakthrough curves has been investigated. For both cases, the breakthrough time was found to decrease when the column length did, and when the initial adsorbate concentration and the volumetric flow rate increased. Therefore, it has been shown that a variation in the initial concentration or the volumetric flow rate changed the slope of the breakthrough curve, so the mass transfer resistance of the process is dependent on these parameters. Parameters as adsorption capacity at breakthrough time (qr) and saturation time (qs), length of the mass transfer zone (MTZ), fractional bed

(19)

(20)

and r=

C0 (μg·L‑1)

breakthrough curves obtained at different concentrations, flow rates, and column lengths using the Clark model are shown in Figure 10a−c.

with ⎛ C n−1 ⎞ A = ⎜⎜ 0n − 1 − 1⎟⎟ ert b ⎝ Cb ⎠

Z (cm)

(21)

where n is the Freundlich constant, Cb is the concentration of adsorbate at breakthrough time, tb (mg·L−1), and νm is the migration velocity of the concentration front in the bed (cm·h−1). 5053



Article

Figure 10. Experimental and predicted breakthrough curves of isoproturon removal by granular activated carbon fixed-bed columns predicted by Clark model: (a) different initial isoproturon concentrations (column length = 2.0 cm, flow rate = 1.5 mL·min−1); (b) different volumetric flow rates (column length = 2.0 cm, initial concn = 100 μg·L−1); (c) different column lengths (initial isoproturon concn = 200 μg·L−1, flow rate = 1.5 mL·min−1).

utilization (FBU), and percentages of adsorbate removal were obtained for the different operation conditions used in the adsorption experiments. The plot of 50% breakthrough versus column length, in the Bohart−Adams model, did not pass through the origin, indicating the adsorption of these micropollutants on activated carbon occurred through a complex mechanism, strongly dependent on diffusion inside the pore of the adsorbent. Therefore, the Thomas, Yoon−Nelson, and Clark models were applied to the experimental data for the prediction of the theoretical breakthrough curves and the parameters associated with these models, as adsorption capacity, adsorption rate constant, or time required for reach 50% adsorbate breakthrough. The Thomas and Yoon−Nelson models have been found more suitable for the mathematical description of atenolol and isoproturon removal in a fixed-bed column in the range of operation parameters studied.

■

■

AUTHOR INFORMATION

kAB = adsorption rate constant (L·mg−1·h−1) N0 = adsorption capacity of the system (mg·L−1) Z = column length (cm) U = linear flow velocity (cm·h−1) t = service time of column (h) β = kinetic coefficient of the external mass transfer (h−1) D = axial diffusion coefficient (cm·h−1) q0 = adsorption capacity (mg·L−1) kT = Thomas rate constant (L·h−1·mg−1) kYN = Yoon−Nelson rate constant (h−1) τ = time required for reach 50% adsorbate breakthrough (h) n = Freundlich constant Cb = concentration of adsorbate at breakthrough time (mg·L−1) νm = migration velocity of the concentration front in the bed (cm·h−1)

REFERENCES

(1) Fatta-Kassinos, D.; Meric, S.; Nikolaou, A. Pharmaceutical residues in environmental waters and wastewater: Current state of knowledge and future research. Anal. Bioanal. Chem. 2011, 399, 251− 275. (2) Kü mmerer, K. The presence of pharmaceuticals in the environment due to human use-present knowledge and future challenges. J. Environ. Manage. 2009, 90, 2354−2366. (3) Enick, O. V.; Moore, M. M. Assessing the assessments: Pharmaceuticals in the environment. Environ. Impact Assess. 2007, 27, 707−729. (4) Buttiglieri, G.; Knepper, T. P. Removal of emerging contaminants in wastewater treatment: Conventional activated sludge. Handb. Envon. Chem. 2008, 5, 1−35. (5) Pareja, L.; Martínez-Bueno, M. J.; Cesio, V.; Heinzen, H.; Fernández-Alba, A. R. Trace analysis of pesticides in paddy field water by direct injection using liquid chromatography-quadrupole-linear ion trap-mass spectrometry. J. Chromatogr. A 2011, 1218, 4790−4798. (6) Ding, Q.; Wu, H. L.; Xu, Y.; Guo, L. I.; Liu, K.; Gao, H. M.; Yang, H. Impact of low molecular weight organic acids and dissolved organic matter on sorption and mobility of isoproturon in two soils. J. Hazard. Mater. 2011, 190, 823−832. (7) Baskeyfield, D. E. H. A membrane-based inmunosensor for the analysis of the herbicide isoproturon. Anal. Chim. Acta 2011, 699, 223−231. (8) Ioannou, L. A.; Hapeshi, E.; Vasquez, M. I.; Mantzavinos, D.; Fatta-Kassinos, D. Solar/TiO(2) photocatalytic decomposition of beta-blockers atenolol and propranolol in water and wastewater. Sol. Energy 2011, 85, 1915−1926.

Corresponding Author

* Corresponding author. Tel.: +34-91-394-4117/5207. Fax: +34-91-394-4114. E-mail: [email protected]; juangcia@ quim.ucm.es. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors gratefully acknowledge the financial support from the Ministerio de Educación y Ciencia by CONSOLIDER Program through TRAGUA Network CSD2006-44. CTQ200802728, CTQ2011-27169 and Comunidad de Madrid through REMTAVARES Network S2009/AMB-1588.

■

NOMENCLATURE qb = breakthrough time capacity (mg·g−1) qs = saturation time capacity (mg·g−1) m = mass of adsorbent (g) Q = volumetric flow rate (mL·min−1) C0 = initial concentration (mg·L−1) C = final concentration (mg·L−1) Cb = desired concentration of adsorbate at breakthrough time (mg·L−1) tb = breakthrough time (h) ts = saturation time (h) 5054



Article

(9) Han, R.; Wang, Y.; Zhao, X.; Wang, Y.; Xie, F.; Cheng, J.; Tang, M. Adsorption of methylene blue by phoenix tree leaf powder in a fixed-bed column: Experiments and prediction of breakthrough curves. Desalination 2009, 245, 284−297. (10) Sze, M. F. F.; Lee, V. K. C.; McKay, F. Simplified fixed-bed column model for adsorption of organic pollutants using tapered activated carbon columns. Desalination 2008, 218, 323−333. (11) Rodriguez, A.; Garcia, J.; Ovejero, G.; Mestanza, M. Adsorption of anionic and cationic dyes on activated carbon from aqueous solutions: Equilibrium and kinetics. J. Hazard. Mater. 2009, 172, 1311−1320. (12) Aksu, Z.; Gönen, F. Biosorption of phenol by immobilized activated sludge in a continuous packed bed: Prediction of breakthrough curves. Process Biochem. 2004, 39, 599−613. (13) Sotelo, J. L.; Rodríguez, A.; Á lvarez, S.; García, J. Removal of caffeine and diclofenac on activated carbon in fixed-bed column. Chem. Eng. Res. Des. Published online 21-November-2011. (14) Vieira, M. G. A.; Gimenes, M. L.; da Silva, M. G. C. Modelling of the process of adsorption of nickel in bentonite clay. Chem. Eng. Trans. 2009, 17, 421−426. (15) Wankat, P. C. Rate-Controlled Separations; Kluwer: Amsterdam, The Netherlands, 1990. (16) Bohart, G. S.; Adams, E. Q. Some aspects of the behaviour of charcoal with respect to chlorine. J. Am. Chem. Soc. 1920, 42, 523− 544. (17) Baral, S. S.; Das, N.; Ramulu, T. S.; Sahou, S. K.; Das, S. N.; Chaudhury, G. R. Removal of Cr(VI) by thermally activated weed Salvinia cucullata in a fixed-bed column. J. Hazard. Mater. 2009, 161, 1427−1435. (18) Walker, G. M.; Weatherley, L. R. Adsorption of acid dyes onto granular activated carbon in fixed beds. Water Res. 1997, 31, 2093− 2101. (19) Singh, S.; Srivastava, V. C.; Mall, I. D. Fixed-bed study for adsorptive removal of furfural by activated carbon. Colloids Surf., A 2009, 332, 50−56. (20) Maji, S. K.; Pal, A.; Pal, T.; Adak, A. Modeling and fixed-bed column adsorption of As(III) on laterite soil. Sep. Purif. Technol. 2007, 56, 284−290. (21) Wolborska, A. Adsorption on activated carbon of paranitrophenol from aqueous-solution. Water Res. 1989, 23, 85−91. (22) Hamdaoui, O. Removal of copper (II) from aqueous phase by Purolite C100-MB cation exchange resin in fixed-bed columns: Modeling. J. Hazard. Mater. 2009, 161, 737−746. (23) Thomas, H. C. Heterogeneous ion exchange in a flowing system. J. Am. Chem. Soc. 1944, 66, 1664−1666. (24) Lopes, C. B.; Pereira, E.; Lin, Z.; Pato, P.; Otero, M.; Silva, C. M.; Duarte, A. C.; Rocha, J. Fixed-bed removal of Hg+2 from contaminated water by microporous titanosilicate ETS-4: Experimental and theoretical breakthrough curves. Microporous Mesoporous Mater. 2011, 145, 32−40. (25) Zhang, W.; Dong, L.; Yan, H.; Li, H.; Jiang, Z.; Kan, X.; Yang, H.; Li, A.; Cheng, R. Removal of methylene blue from aqueous solutions by straw based adsorbent in a fixed-bed column. Chem. Eng. J. 2011, 173, 429−436. (26) Yoon, Y.; Nelson, J. Application of gas adsorption kinetics. I. A theoretical model for respirator cartridge service time. Am. Ind. Hyg. Assoc. J. 1984, 45, 509−516. (27) Clark, R. M. Evaluating the cost and performance of field-scale granular activated carbon systems. Environ. Sci. Technol. 1987, 21, 573−580. (28) Srivastava, V. C. Prediction of breakthrough curves for sorptive removal of phenol by bagasse fly ash packed bed. Ind. Eng. Chem. Res. 2008, 47, 1603−1613.

5055


Removal of Atenolol and Isoproturon in Aqueous Solutions by

Recommend Documents