Thermodynamic analysis of Cr(VI) extraction using TOPO impregnated membranes

Journal of Hazardous Materials 314 (2016) 204–210

Contents lists available at ScienceDirect

Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat

Thermodynamic analysis of Cr(VI) extraction using TOPO impregnated membranes Prashant Praveen, Kai-Chee Loh ∗ Department of Chemical and Biomolecular Engineering, National University of Singapore, 4 Engineering Drive 4, 117585, Singapore

h i g h l i g h t s • • • • •

Cr(VI) extraction by extractant impregnated membranes (EIM) was investigated. EIM exhibited high extraction efficiency, mass transfer rate and stability. Mass transfer mechanism was proposed based on kinetics and equilibrium data. Uptake of Cr(VI) by EIMs was endothermic and spontaneous. Cr(VI) extraction by EIMs was dominated by physical interactions.

a r t i c l e

i n f o

Article history: Received 20 January 2016 Received in revised form 20 April 2016 Accepted 21 April 2016 Available online 22 April 2016 Keywords: Adsorption Extraction Hollow fiber membrane Chromium Trioctylphosphine oxide

a b s t r a c t Solid/liquid extraction of Cr(VI) was accomplished using trioctylphosphine oxide impregnated polypropylene hollow fiber membranes. Extraction of 100–500 mg/L Cr(VI) by the extractant impregnated membranes (EIM) was characterized by high uptake rate and capacity, and equilibrium was attained within 45 min of contact. Extraction equilibrium was pH-dependent (at an optimal pH 2), whereas stripping using 0.2 M sodium hydroxide yielded the highest recovery of 98% within 60 min. The distribution coefficient was independent of initial Cr(VI) concentration, and the linear distribution equilibrium isotherm could be modeled using Freundlich isotherm. The mass transfer kinetics of Cr(VI) was examined using pseudo-second-order and intraparticle diffusion models and a mass transfer mechanism was deduced. The distribution coefficient increased with temperature, which indicated endothermic nature of the reaction. Enthalpy and entropy change during Cr(VI) extraction were positive and varied in the range of 37–49 kJ/mol and 114–155 J/mol, respectively. The free energy change was negative, confirming the feasibility and spontaneity of the mass transfer process. Results obtained suggest that EIMs are efficient and sustainable for extraction of Cr(VI) from wastewater. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Rapid industrialization has led to the rise in the presence of heavy metals in water resources [1]. One of the most widely distributed heavy metal in the environment is chromium, which is released from various industrial processes such as electroplating, tanning, wood preservation and textile synthesis [2]. Among the two valence forms of chromium: Cr(VI) is considered more hazardous than Cr(III) due to its high toxicity, carcinogenicity and recalcitrance. The concentration of Cr(VI) in industrial effluents can reach 500 mg/L, although the regulatory discharge limit is <1 mg/L [3].

∗ Corresponding author. E-mail address: [email protected] (K.-C. Loh). http://dx.doi.org/10.1016/j.jhazmat.2016.04.054 0304-3894/© 2016 Elsevier B.V. All rights reserved.

Conventional methods of Cr(VI) treatment via chemical reduction is costly and also generates secondary pollutants [4]. On the other hand, solvent extraction [5] or adsorption [6] can selectively recover and recycle the heavy metal from wastewater, and these are often preferred in heavy metal removal from wastewater. Solvent extraction using organic extractants dispersed in the wastewater, such as trioctylphosphine oxide (TOPO) [7] and tributyl phosphate [8], is widely used in the recovery of Cr(VI) from wastewater. TOPO, in particular, is considered a very effective extractant for Cr(VI) extraction due to its high selectivity, excellent chemical stability and low water solubility [5]. However, the disadvantage of solvent extraction is phase dispersion, which results in emulsification and high energy costs. Moreover, the large amount of organic solvent required to perform solvent extraction renders the process expensive and inefficient, particularly at low metal ion concentrations [3].

P. Praveen, K.-C. Loh / Journal of Hazardous Materials 314 (2016) 204–210

Phase-dispersion associated challenges can be mitigated by physically separating the aqueous and organic phases using semipermeable membranes or by using supported liquid membranes [9,10]. While these configurations offer advantages of low operating costs, low solvent requirements, complete solvent recycle and reuse, the high mass transfer resistance and lack of longterm stability of membrane-based extractors are major let downs [11]. Dispersion-free solvent extraction can also be achieved using solvent-encapsulated polymeric microcapsules [9,12]. But the preparation of these microcapsules is tedious and a significant amount of the extractant may be lost in the process [13]. Recently, extractant impregnated membranes (EIM) have been used in the recovery of organic pollutants from wastewater [13,14]. The EIMs are based on impregnation of a solid organic extractant in hydrophobic polymeric membranes. Unlike solvent-encapsulated microcapsules, EIM preparation does not involve any polymerization. The semi-permeable membranes are simply wetted with the extractant dissolved in a volatile carrier solvent; the carrier solvent is subsequently evaporated and the extractant is immobilized in the membrane walls. The EIMs have the advantages of high solute retention capacity, high solute uptake rate, large interfacial area and non-dispersive mass transfer [14]. Furthermore, extraction using the EIMs is considered ‘solventless’, as only a small amount of the carrier solvent/extractant is required for impregnation. More importantly, due to stable TOPO impregnation, the EIMs can be used repeatedly without performance deterioration, which enhances the sustainability of the separation. So far, extraction using the EIMs has been focused on the equilibrium and kinetics studies, and thermodynamics aspect of the process has not been investigated. Moreover, the use of EIMs in wastewater treatment has been limited to the recovery of aromatic compounds from wastewater, and their effectiveness in the removal of heavy metals has not been demonstrated. In this research, TOPO-based EIMs were used in extractive recovery of Cr(VI) from wastewater. Experiments have been performed under various operating conditions of pH, temperatures and Cr(VI) concentrations. Thermodynamic analysis of the experimental results has been conducted based on adsorption kinetic models, adsorption isotherms, enthalpy, entropy and free energy of the extraction process.

205

2.2. EIM preparation The EIMs were prepared by impregnating Accurel PP 50/280 polypropylene hollow fiber membranes (Membrana GmbH, Germany) with TOPO. Membrane specifications are described in the authors’ previous work [11]. The hollow fiber membranes were cut into small pieces of 6 cm length each, and bundles of 20 membrane pieces were prepared by applying epoxy resins at both ends of the membranes (Araldite, England). For immobilization, the bundles were added into DCM solution containing 400 g/L TOPO, and stirred on a shaking water bath at 30 ◦ C for 1 h at 150 rpm to wet the membranes properly. The wetted membranes were then removed and rinsed twice with ultrapure water to remove excess solution present on the outer surfaces of the membranes. Finally, the membranes were air dried for 24 h to evaporate DCM, leaving TOPO inside the fibers. The resulting EIMs were washed thoroughly to remove loosely held TOPO from the membranes. A detailed description of the impregnation process is given in Fig. S1 in the online version at DOI: 10.1016/j.jhazmat.2016.04.054 (Supplementary material). TOPO impregnation in the polypropylene membranes was estimated by measuring the changes in the weight of the membrane bundles before and after impregnation. The membranes were also weighed prior to the bundling process to prevent interference from variations in epoxy weight.

2.3. Experimental procedure The extraction/stripping experiments were carried out in batch mode in a 50 mL Falcon tube with 40 mL of aqueous solution at pH 1–3 on a thermomixer (MKR11, HLC Biotech, Germany) operating at 25–35 ◦ C and 200 rpm. A total of 15 bundles (with 20 membranes each) of EIMs were used to investigate the extraction of Cr(VI) in every batch. The total mass of the EIMs used for extraction was 0.62 g, which was equivalent to 15.6 g-EIMs/L-reactor volume. Samples were collected from the tubes periodically to measure aqueous concentrations of Cr(VI), whereas the amount of Cr(VI) extracted by the EIMs at any time was calculated by material balance. Cr(VI) was stripped using 40 mL of 0.2 M sodium hydroxide solution, after quickly rinsing them in water to wash away any wastewater from the EIM surfaces. Each experiment was repeated three times to ascertain reproducibility. In all the experiments, the standard deviation between triplicates was mostly less than 5%.

3. Theoretical 2. Material and methods 3.1. Adsorption kinetics modeling 2.1. Chemicals and analytical methods All chemicals used in this research were of analytical grade and were used as received from the supplier. Potassium dichromate was dissolved in hydrochloric acid (pH 2) to make a stock solution of 1 g/L. The EIMs were characterized using a scanning electron microscope (SEM) (JEOL JSM-5600LV) after sputtering with platinum. Cr(VI) concentration in the aqueous phase was determined by measuring the absorbance at 540 nm using an UV–vis spectrophotometer (UV-1240, Shimazdu, Japan) following coloration of the samples using diphenylcarbazide method [4]. Cr(VI) concentration in the solid phase (EIMs) was calculated by mass balance. The equilibrium extraction capacity calculations were made based on the mass of the EIMs which was 0.62 g. TOPO concentration in dichloromethane (DCM) was analyzed by gas chromatography equipped with a flame ionization detector (Clarus 600, Perkin Elmer, USA).

To investigate the mass transfer kinetics of Cr(VI) from the aqueous phase to the EIMs, pseudo-first-order, pseudo-second-order and intraparticle diffusion models were fitted to the experimental data [15]. Pseudo-first-order (PFO) rate expression can be written as [13]: dQt = k1 (Qe − Qt ) dt

(1)

where Qt and Qe are the amounts of Cr(VI) present in the EIMs at time t and at equilibrium, respectively, and k1 is the PFO rate constant. Eq. (1) can be integrated to obtain: loge (Qe − Qt ) = loge Qe − k1 t

(2)

With experimental data of Qt and t, Qe and k1 are obtained as adjustable parameters from curve fitting.

206

P. Praveen, K.-C. Loh / Journal of Hazardous Materials 314 (2016) 204–210

200

dQt = k2 (Qe − Qt )2 dt

160

(3)

Eq. (3) can be integrated to obtain: 1 t t = + Qt Qe k2 Qe2

(4)

The parameters Qe and k2 can be calculated from the slope and the intercept, respectively, of the straight line when experimental t/Qt is plotted against t. In adsorption systems in which intraparticle diffusion is rate limiting, the adsorption kinetics is commonly studied using the intraparticle diffusion model described by: Qt = ki t 0.5 + C

Cr(VI) conc. (mg/L)

The mass transfer rate for pseudo-second-order (PSO) kinetics is given by the expression:

Extraction

Stripping

120 80 40 0 0

15

(5)

30

45

60

t (min)

where ki and C are intraparticle diffusion constants calculated from the slope and intercept, respectively, of the straight-line portions of Qt vs. t0.5 [13].

Fig. 1. Extraction and stripping of Cr(VI) by EIMs. Extraction was carried out at pH 2 and 30 ◦ C. Stripping was performed using 0.2 M NaOH.

3.2. Adsorption isotherm

4. Results and discussion

Adsorption isotherms are often used to gain insights into the mass transfer mechanism, surface properties and affinity of the sorbent for the solute. Langmuir isotherm is expressed as [16]:

4.1. EIM characterization

Qe = Qmax

 kC  L e 1 + kL Ce

(6)

where Ce is the equilibrium aqueous Cr(VI) concentration (mg/L), kL is the Langmuir adsorption constant (L/mg) and Qmax is the maximum adsorption capacity (mg/g). The parameters kL and Qmax can be calculated from the slope and intercept, respectively, of the straight line for 1/Qe plotted against 1/Ce . Freundlich isotherm assumes adsorption on heterogeneous sites with non-uniform distribution of energy levels and is given by [16]: 1 n

Qe = kF Ce

(7)

(mg/g-(L/mg)1/n ) and n are constants and can be estimated

where kF from the slope and intercept of the straight line obtained by plotting loge Qe against loge Ce , respectively. 3.3. Thermodynamic properties Thermodynamic properties under a given set of conditions can identify the spontaneity and feasibility of a process or reaction. The Enthalpy Change (H) (kJ/mol) in a process can be calculated using the Clapeyron-Clausius equation [17]: ln Ce,cal =

H − ln Ko RT

(8)

where Ce,cal is Cr(VI) concentration in the EIMs (mg/L) as calculated by Freundlich isotherm at given Qe values, T is the reaction temperature (K), Ko is the Clapeyron-Clausius constant (mg/g) and R is the gas constant at 8.314 J/mol-K. Gibbs Free Energy Change (G) (kJ/mol) is the fundamental criterion to predict the spontaneity of a reaction. It is related to the Freundlich constant, n, by the following equation [17]: G = −nRT

(9)

Entropy Change (S) (kJ/mol-K) is the measure of the level of disorder in a thermodynamic system. S can be obtained through the Gibbs-Helmholtz equation [17]: S =

H − G T

(10)

The EIMs were characterized through weight measurement and SEM. Imaging. Upon TOPO impregnation, the weight of the supporting membranes increased by about 55%, and TOPO density was calculated as 1.21 g-TOPO/g-polypropylene membranes (excluding weight of epoxy), which was equivalent to 8.5 g-TOPO/L-reactor volume. The SEM images (Fig. S1 in the online version at DOI: 10.1016/j.jhazmat.2016.04.054) showed non-uniform distribution of TOPO within the membrane walls, with a higher TOPO density along an outer membrane surface. These results are consistent with previous studies on the synthesis of EIMs [13,14]. 4.2. Extraction/stripping Fig. 1 shows the extraction of 200 mg/L Cr(VI) by the EIMs at pH 2 and T = 30 ◦ C, and subsequent stripping of the extracted Cr(VI) from the EIMs by 0.2 M NaOH. The mass transfer of Cr(VI) from aqueous phase to the EIMs was quick, and equilibrium was achieved within 45 min of contact. About 70% of the initial Cr(VI) could be extracted into the solid phase, resulting in a Ce of 61.3 mg/L. Based on mass balance and total weight of EIMs, Qe obtained was 9.10 mg/g. During stripping, more than 98% of the extracted Cr(VI) could be recovered within 45 min of equilibration, which indicated that the stripping was effective, and the EIMs could be reused for further extraction. At the end of each extraction experiment, samples were collected from the aqueous phase to determine whether any TOPO had leaked from the membranes. No TOPO was detected in the aqueous phase in any of the experiments, which indicated the high stability of TOPO impregnation into the membranes. In order to further demonstrate the stability of TOPO impregnation in the EIMs, 10 cycles of Cr(VI) extraction and stripping were performed consecutively with the same sets of the EIMs. In these experiments, the extraction performance and stripping efficiency (>97%) remained unchanged, and average Qe was obtained as 9.21 ± 0.33 mg/g (Fig. S2 in the online version at DOI: 10.1016/ j.jhazmat.2016.04.054). These results indicated that Cr(VI) extraction by the EIMs was stable and reproducible in the long run. 4.2.1. Effects of pH on extraction/stripping Fig. 2 shows the effects of pH of the feed solution on extraction equilibrium of Cr(VI). At pH 2 and 3, the time required to reach equilibrium was about the same at 45 min, although a higher amount

P. Praveen, K.-C. Loh / Journal of Hazardous Materials 314 (2016) 204–210

14 pH 1

12

pH 2

100

pH 3

90 % Recovery

10 Qe (mg/g)

207

8 6 4

80 70 60

2 0 0

15

30

45

50

60

0.2

0.05

t (min) Fig. 2. Effects of pH on the extraction kinetics of Cr(VI) by EIMs.

0.8

Fig. 3. Effects of NaOH concentration on the recovery of Cr(VI) from EIMs.

of Cr(VI) could be extracted into the EIMs at pH 2. The Qe values at pH 2 and 3 were 9.1 and 3.2 mg/g, respectively. The equilibrium trend observed at pH 1 was quite different from those observed at pH 2 or 3. At pH 1, Qt reached a high value of 11.6 mg/g only after 5 min of equilibration. However, this distribution was not stable, and Cr(VI) concentration in the EIMs decreased steadily after reaching the maxima. After 60 min of equilibration, only 7.0 mg/g Cr(VI) remained in the EIMs, which did not appear to be the equilibrium concentration, as Cr(VI) distribution between the two phases was not showing signs of stabilization even at the end of the experiment. The changes in the extraction equilibrium of Cr(VI) with pH could be a result of the various ionic forms of the chromate ion (HCrO4 − , CrO4 2− , HCr2 O7 − , Cr2 O7 2− ) in aqueous solution. One of these ions dominate depending on the pH, as well as the total amount of Cr(VI) present in the solution [3,5]. Typically, Cr2 O7 2− anions dominate in acidic solution, but these readily convert into HCrO4 − when the total Cr(VI) concentration in the aqueous (or nonaqueous) phase decreases to lower values. The extraction equilibria of Cr(VI) with TOPO at acidic pH can be described by the chemical reactions between Cr(VI) anions and TOPO molecules. These reactions may involve one or more of TOPO molecules, and water molecules may also participate in the chemical reactions [7]. These reactions can be represented by the following [18]: Cr2 O7 2− + 2H+ + 3TOPO ↔ H2 Cr2 O7 · (TOPO)3

(11)

HCrO4 − + H+ + TOPO ↔ H2 CrO4 · (TOPO)

(12)

As more Cr(VI) reacted with TOPO and is transferred from aqueous phase to the EIMs, its concentration in the aqueous phase is reduced resulting in an increase in the concentration of HCrO4 − in the aqueous phase. The decrease in the extraction capacity of Cr(VI) at pH 1 can be attributed to the presence of HCrO4 − in the aqueous phase at low pH. H2 CrO4 ↔ HCrO4 − + H+

0.5

NaOH conc. (M)

(13)

It appears that the rise in H+ concentration at pH 1 was able to shift Eq. (13) to the left, which resulted in deionization of the chromate ion [19]. However, the primary factor for the reversal in Cr(VI) binding with TOPO at pH 1 could be attributed to irreversible conversion of Cr(VI) to Cr(III) under low pH [6]. Since Cr(III) had lower affinity for TOPO, their rising concentration would have reversed the mass transfer process of Cr(VI) from the EIMs to the aqueous phase to maintain equilibrium. These observations are consistent with that reported in the literature, wherein the extraction efficiency of Cr(VI) ions by TOPO increased with decreasing pH, until a critical pH is reached, below which extraction performance decreased steadily [3].

Fig. 3 shows the results of stripping of Cr(VI) from the EIMs using various concentrations of NaOH, ranging from 0.05 to 0.8 M. The recovery of Cr(VI) from the EIMs increased from 90 to 98% when NaOH concentration was increased from 0.05 to 0.2 M. However, when NaOH concentration was increased further, instead of an improvement in Cr(VI) recovery, the amount of Cr(VI) stripped from the EIMs decreased. The optimum NaOH concentration was found to be 0.2 M at which Cr(VI) recovery was 98%. These stripping reactions of Cr(VI) solvation complexes with NaOH can be represented by the following equations: H2 Cr2 O7 · (TOPO)3 + 3OH− ↔ Cr2 O7 2− + 3TOPO + 3H2 O −

−

H2 CrO4 · TOPO + OH ↔ HCrO4 + TOPO + H2 O

(14) (15)

During stripping, H2 Cr2 O7 ·(TOPO)3 and H2 CrO4 ·TOPO reacted with NaOH and Cr(VI) was recovered in the aqueous phase. Since the predominant form of Cr(VI) in aqueous solution at pH > 7 is CrO4 2− , Cr2 O7 2− and HCrO4 − readily converted into CrO4 2− at the solid/liquid interface, when contacted with NaOH [18]. The effective stripping reaction can thus be written as: H2 CrO4 · TOPO + 2OH− ↔ CrO4 2− + TOPO + 2H2 O

(16)

It is possible that at high NaOH concentration above 0.2 M, rapid stripping of Cr(VI) from the EIMs, and its conversion into CrO4 2− , could have resulted in the generation of large amounts of CrO4 2− in the aqueous phase. This might have shifted the reaction in Eq. (16) to the left, resulting in a lower recovery of Cr(VI) at higher NaOH concentration. Such variations in the stripping efficiency with NaOH concentration have been attributed to a shift in the reversible reaction [20]. All further Cr(VI) extraction experiments were carried out at the optimal wastewater pH of 2, whereas all the stripping experiments were performed using 0.2 M NaOH. 4.2.2. Effects of Cr(VI) concentration Fig. 4 shows the extraction equilibrium between the aqueous phase and the EIMs during Cr(VI) extraction. The Qe values at equilibrium were recorded as 5.2, 9.0, 13.2, 17.1 and 21.9 mg/g, at initial Cr(VI) concentrations of 100, 200, 300, 400 and 500 mg/L, respectively, at pH 2 and 30 ◦ C. These results indicate a linear relationship between equilibrium extraction capacity and initial Cr(VI)concentration within 100–500 mg/L concentration range. The distribution coefficient was independent of initial Cr(VI) concentration and it was estimated as 0.141 L/g at 30 ◦ C. Similar trends were observed at the operating temperatures of 25 and 35 ◦ C, yielding constant distribution coefficients of 0.108 and

208

P. Praveen, K.-C. Loh / Journal of Hazardous Materials 314 (2016) 204–210

(a)

30

200 mg/L

30 °C 20

300 mg/L

400 mg/L

20

25 °C

15

Qe (mg/g)

Qe (mg/g)

25

35 °C

25

10 5

15 10 5

0 0

30

60

90

120

150

180

0 0

15

30

Ce (mg/L)

(b)

25 200 mg/L

4.2.3. Temperature Apart from the effects of initial Cr(VI) concentration, Fig. 4 also shows the effects of operating temperature on Cr(VI) removal by the EIMs. While the distribution equilibrium trends remained Table 1 Langmuir and Freundlich isotherm parameters at different temperatures.

Langmuir

Qmax (mg/g) kL (␮g/L) R2

Freundlich

kF ((mg/g)(L/mg)1/n ) n R2

Values T = 25 ◦ C

T = 30 ◦ C

T = 35 ◦ C

74.6 1.76 0.99

172.4 1.28 0.99

58.1 3.26 0.99

0.17 1.10 0.99

0.19 1.07 0.99

0.21 1.02 0.99

300 mg/L

400 mg/L

20

Qe (mg/g)

0.194 L/g, respectively. Since the distribution coefficient (Qe /Ce ) did not change at constant temperature and pH, it was an indication that initial Cr(VI) concentration, in the range of 100–500 mg/L, did not influence the distribution equilibrium between the solid/liquid extraction system. During liquid/liquid extraction using organophosphorus extractants such as TOPO, the distribution coefficients of metals and organics is related to the initial concentrations of the solute [21]. Typically, a rising solute concentration favors higher distribution in the aqueous phase and it has been observed phenol extraction by the EIMs [13]. This is attributed to the reactive extraction process, which is based not on the solubility of the solute in the organic phase, but on the chemical reaction between the solute and the extractant. Since solventless extraction by the EIMs was a solid/liquid mass transfer process, which resembled sorption, the equilibrium distribution of Cr(VI) on the EIMs could be analyzed using adsorption isotherms [13]. The equilibrium data presented in Fig. 4 were analyzed using the Langmuir and the Freundlich isotherms. The model parameters obtained are listed in Table 1. Based on the computed values of R2 (>99%), it was observed that both of these isotherms could fit the experimental data with reasonable accuracy. However, the model parameters calculated for the Langmuir isotherm were inconsistent. On the other hand, Freundlich isotherm parameters were consistent and appeared more reliable. Therefore, Freundlich isotherm was used to predict the distribution equilibrium in this case. The dotted lines in Fig. 4 show the fit of the experimental data to the Freundlich isotherm.

Parameters

60

t (min)

Fig. 4. Distribution equilibrium of Cr(VI) between aqueous phase and the EIMs at different temperatures and Cr(VI) concentrations. The markers indicate the experimental data and the dotted lines indicate Freundlich isotherm.

Isotherm

45

15 10 5 0

1

2

3

4

5

6

7

8

t 0.5 (min0.5) Fig. 5. Extraction kinetics of Cr(VI) on the EIMs fitted to: (a) pseudo-second-order model, and; (b) intraparticle diffusion model. The markers indicate the experimental data and the dotted lines indicate model predictions.

unchanged with temperature, the amount of Cr(VI) extracted into the EIMs at given Ce increased with increasing temperature. Since Cr(VI) extraction by the EIMs was favored at higher temperature, it can be inferred that the solvation reactions between Cr(VI) and TOPO could be endothermic [17]. The equilibrium studies could not be carried out at temperatures higher than 35 ◦ C due to the relatively low melting point of TOPO. Although the melting point was close to 60 ◦ C, heat generated at temperatures above 40 ◦ C could induce phase transformation of the extractant, turning the crystalline solid into a gel-like substance, which could slowly leak out of the membranes. 4.3. Mass transfer kinetics Fig. 5a shows the representative concentration profiles of Cr(VI) extracted into the EIMs, at initial Cr(VI) concentrations of 200–400 mg/L, at pH 2 and 30 ◦ C. The equilibrium between the two phases was quick, and more than 85% Cr(VI) was extracted into the solid phase within 15 min of equilibration. Equilibrium was attained within 40–60 min in each case. In order to calculate Cr(VI) uptake rate by the EIMs, the experimental data were fitted to PFO and PSO kinetic models. The calculated values of the model parameters are listed in Table 2. The experimental data were in good agreement with the PFO model based on the high R2 values. However, the PFO modeled Qe at various Cr(VI) concentrations were significantly lower than those obtained experimentally. For example, the experimental Qe at 30 ◦ C

P. Praveen, K.-C. Loh / Journal of Hazardous Materials 314 (2016) 204–210

6

Table 2 Pseudo-first-order, pseudo-second-order and intraparticle diffusion model parameters for extraction of Cr(VI) by EIMs. Parameters

−1

200

300

400

4

Pseudo-first-order

k1 (min ) Qe (mg/g) R2

0.101 7.1 0.99

0.132 11.3 0.97

0.101 14.9 0.97

Pseudo-second-order

k2 (g/mg min) Qe (mg/g) R2

0.027 9.7 0.99

0.015 15.3 0.99

0.012 21.6 0.99

ki (mg/g min0.5 ) R2

1.85 0.99

3.64 0.99

4.93 0.98

Intraparticle diffusion

5

Cr(VI) conc. (mg/L)

ln Ce

Kinetic model

209

3 2 1 0 3.24

and initial concentrations of 200, 300 and 400 mg/L Cr(VI) were 9.1, 14.0 and 20.2 mg/g, respectively (the difference in Qe between Figs. 4 and 5 at 400 mg/L Cr(VI) is due to measured initial Cr(VI) concentration varying between 380 and 430 mg/L), whereas those predicted by the PFO model were 7.1, 11.3 and 14.9 mg/g, respectively. On the other hand, PSO kinetic model resulted in an excellent fit with the experimental data and the theoretical Qe values, computed using the model, corroborated the experimental results. Therefore, PSO model was better suited to describe the uptake rate of Cr(VI) on the EIMs. The dotted lines in Fig. 5a show the extraction kinetics of Cr(VI) at different concentrations, as predicted by the PSO model. As can be seen, there was an excellent correlation between the experimental and the modeled mass transfer kinetics. The rate constant k2 decreased with increasing Cr(VI) concentration, which could be attributed to the increase in the diffusion resistance as a higher amount of Cr(VI) diffused inside the EIMs, when TOPO molecules impregnated nearer the outer surface of the EIMs reached equilibrium with the metal ions. These results are consistent those obtained in the extraction of phenol by the EIMs [13]. To determine whether intraparticle diffusion was a controlling factor for Cr(VI) uptake by the EIMs, the amount of Cr(VI) extracted by the EIMs was plotted against the square root of elapsed time. The representative concentration profiles at initial Cr(VI) concentrations of 200–400 mg/L are indicated in Fig. 5b. The experimental data could be fitted with a straight line in the first 15 min, while data after 15 min could be described by another distinct straight line. The calculated rate constants along with the correlation coefficients are also listed in Table 2. These results indicated that intraparticle diffusion was indeed involved in the mass transfer of Cr(VI) to the EIMs. However, intraparticle diffusion was not rate-limiting because the straight lines did not pass through the origin, suggesting the presence of a diffusion boundary layer. The two straight lines in Fig. 5b also suggested that the extraction of Cr(VI) by the EIMs occurred in two stages. The first sharper region was attributed to Cr(VI) uptake nearer the external surface of the EIMs where TOPO density was quite high. The intraparticle diffusion was not a controlling factor at this stage and it was likely that Cr(VI) uptake rate was controlled by the diffusion of Cr(VI) through the aqueous boundary layer. The second stage commenced after a

4 mg/g

9 mg/g

19 mg/g

24 mg/g

3.27

13 mg/g

3.30

3.33

3.36

1/T x 1000 (K-1) Fig. 6. Enthalpy changes during the extraction of Cr(VI) by EIMs.

significant amount of Cr(VI) (about 85%) had been adsorbed into the EIMs. During this stage, Cr(VI) diffused deeper into the membranes to reach TOPO impregnated sites. This intraparticle diffusion occurred until equilibrium was reached. Based on the results from kinetic modeling and the analysis of the reactive extraction using Eqs.(11)–(16), the mass transfer mechanism of extraction/stripping of Cr(VI) using EIMs can be elucidated. The mass transfer steps are as follow: (1) Cr(VI) in aqueous solution was ionized into HCrO4 − and Cr2 O7 2− , which diffused through the aqueous boundary layer to the liquid/solid interface; (2) most of the HCrO4 − and Cr2 O7 2− were extracted at the solid/liquid interface by TOPO impregnated nearer the outer surfaces of the EIMs to form H2 Cr2 O7 . (TOPO)3 and H2 CrO4 ·TOPO solvation complexes; (3) some of the HCrO4 − and Cr2 O7 2− anions diffused into the membrane pores to locate TOPO impregnated sites and they formed reaction complexes with TOPO within the membrane pores; (4) when pH of the aqueous solution was rendered alkaline for Cr(VI) stripping, OH− ions diffused to the membrane surface and into the membrane pores to generate HCrO4 − and Cr2 O7 2− from the solvation complexes; (5) when HCrO4 − and Cr2 O7 2− diffused out of the EIMs into the aqueous solution, they were converted into CrO4 2− at the solid/liquid interface, and (6) CrO4 2− diffused through the aqueous boundary layer into the bulk solution. 4.4. Extraction thermodynamics In order to investigate the thermodynamics of the solventless extraction of Cr(VI) by the EIMs, thermodynamic properties such as H, G and S were estimated using Eqs. (8)–(10). The thermodynamic properties obtained under the various operating conditions investigated are summarized in Table 3. Fig. 6 shows the Clapeyron−Clausius equation plot between ln Ce, cal and T at various Qe values for computation of H. The H varied in the range of 37–49 kJ/mol at given Qe values ranging from 4 to 24 mg/g. Since H > 0 under the experimental conditions, it

Table 3 Enthalpy, entropy and free energy change at different temperatures and Cr(VI) concentration for extraction of Cr(VI) by EIMs. Qe (mg/g)

H (kJ/mol)

G (kJ/mol) ◦

4 9 13 19 24

37.8 42.5 45.0 47.3 48.8

S (J/mol-K) ◦

T = 25 C

T = 30 C

−2.72

−2.70

◦

T = 25 ◦ C

T = 30 ◦ C

T = 35 ◦ C

−2.61

117.7 133.5 141.9 149.7 154.5

115.8 131.4 139.6 147.3 152.0

114.2 129.5 137.6 145.2 149.8

T = 35 C

210

P. Praveen, K.-C. Loh / Journal of Hazardous Materials 314 (2016) 204–210

was concluded that the extraction of Cr(VI) by the EIMs was an endothermic process, and the increasing temperature favored the mass transfer process [22]. These results are consistent with the finding reported earlier in Section 4.2.3. Generally, uptake of a solute from an aqueous solution to a solid sorbent can be classified in two groups: physisorption and chemisorption. This classification is based on H values obtained during the mass transfer process. If the interaction between the solute and the sorbent is based on chemical bonding, H is greater than 42 kJ/mol [22], whereas H varies between 2 and 40 kJ/mol when the interactions are based on physical forces [17]. However, H during Cr(VI) extraction by the EIMs varied in the range of 38–49 kJ/mol, which made it difficult to predict the nature of interaction between Cr(VI) and TOPO. It was possible that both physical and chemical interactions were present, although one of them dominated. Further evidence of the nature of interactions between Cr(VI) and TOPO was inferred from computing G values at different operating temperatures. When temperatures was varied in the range of 25–35 ◦ C, G varied from −2.5 to −3.0 kJ/mol. G for physical sorption is usually between 0 to −20 kJ/mol, whereas that for chemical sorption is between −80 to −400 kJ/mol [17]. Hence, it can be inferred that the solventless extraction of Cr(VI) by the EIMs was dominated by the physical interactions rather than chemical bonding. Moreover, G obtained under all the experimental conditions was negative, which showed that the extraction was a spontaneous process. S obtained during solventless extraction at different Cr(VI) concentrations and temperatures were positive, and varied in the range of 117–155 J/mol K. The positive values of S suggested that Cr(VI) extraction by the EIMs was not an enthalpy driven process, but it was an entropy-driven process [22]. These S values also indicated an increase in the randomness at the solid/liquid interface, during the fixation of Cr(VI) ions on the EIM surface [23]. 5. Conclusions TOPO-based EIMs have been prepared for solventless extraction of Cr(VI) from aqueous solution. The effects of pH on the extraction and stripping have been investigated. Extraction optimum was achieved at pH 2, whereas stripping with 0.2 M NaOH resulted in the highest recovery. The mass transfer kinetics have been modeled using pseudo-second-order and intraparticle diffusion models, whereas the distribution equilibrium has been examined using Freundlich isotherm, and the distribution coefficients were found to be independent of initial Cr(VI) concentrations. The thermodynamic parameters of enthalpy, entropy and free energy have been computed, which indicated that TOPO uptake by the EIMs was endothermic and spontaneous. References [1] S.E. Bailey, T.J. Olin, R.M. Bricka, D.D. Adrian, A review of potentially low-cost sorbents for heavy metals, Water Res. 33 (1999) 2469–2479. [2] V.K. Gupta, A. Rastogi, A. Nayak, Adsorption studies on the removal of hexavalent chromium from aqueous solution using a low cost fertilizer industry waste material, J. Colloid Interface Sci. 342 (2010) 135–141.

[3] R.A. Kumbasar, Extraction of chromium(VI) from multicomponent acidic solutions by emulsion liquid membranes using TOPO as extractant, J. Hazard. Mater. 167 (2009) 1141–1147. [4] H. Deveci, Y. Kar, Adsorption of hexavalent chromium from aqueous solutions by bio-chars obtained during biomass pyrolysis, J. Ind. Eng. Chem. 19 (2013) 190–196. [5] T.C. Huang, C.C. Huang, D.H. Chen, Extraction equilibria of chromium(VI) from sulfuric acid solutions with tri-n-octylphosphine oxide, J. Chem. Technol. Biotechnol. 73 (1998) 43–53. [6] W. Zheng, J. Hu, Z. Han, E. Diesel, Z. Wang, Z. Zheng, C. Ba, J. Langer, J. Economy, Interactions of Cr(VI) with hybrid anion exchange/porous carbon fibers in aqueous solution at natural pH, Chem. Eng. J. 287 (2016) 54–61. [7] F.J. Alguacil, A. Lopez-Delgado, M. Alonso, A.M. Sastre, The phosphine oxides Cyanex 921 and Cyanex 923 as carriers for facilitated transport of chromium(VI)-chloride aqueous solutions, Chemosphere 57 (2004) 813–819. [8] M. Frites, M.H. Samar, N. Boutefnouchet, A. Abbaci, Extraction of chromium(VI) from aqueus solution by liquid surfactant membrane, Asian J. Chem. 17 (2005) 1397–1405. [9] X. Ma, Y. Li, X. Li, L. Yang, X. Wang, Preparation of novel polysulfone capsules containing zirconium phosphate and their properties for Pb2+ removal from aqueous solution, J. Hazard. Mater. 188 (2011) 296–303. [10] C.O. Illanes, N.A. Ochoa, J. Marchese, Kinetic sorption of Cr(VI) into solvent impregnated porous microspheres, Chem. Eng. J. 136 (2008) 92–98. [11] P. Praveen, K.-C. Loh, Simultaneous extraction and biodegradation of phenol in a hollow fiber supported liquid membrane bioreactor, J. Membr. Sci. 430 (2013) 242–251. [12] M. Whelehan, U. von Stockar, I.W. Marison, Removal of pharmaceuticals from water: using liquid-core microcapsules as a novel approach, Water Res. 44 (2010) 2314–2324. [13] P. Praveen, K.-C. Loh, Solventless extraction/stripping of phenol using trioctylphosphine oxide impregnated hollow fiber membranes—experimental & modeling analysis, Chem. Eng. J. 255 (2014) 641–649. [14] P. Praveen, K.-C. Loh, Trioctylphosphine oxide-impregnated hollow fiber membranes for removal of phenol from wastewater, J. Membr. Sci. 437 (2013) 1–6. [15] Y.S. Ho, G. McKay, Pseudo-second order model for sorption processes, Process Biochem. 34 (1999) 451–465. [16] K.Y. Foo, B.H. Hameed, Insights into the modeling of adsorption isotherm systems, Chem. Eng. J. 156 (2010) 2–10. [17] T. Qiu, Y. Zeng, C. Ye, H. Tian, Adsorption thermodynamics and kinetics of p-xylene on activated carbon, J. Chem. Eng. Data 57 (2012) 1551–1556. [18] T.C. Huang, C.C. Huang, D.H. Chen, Transport of chromium(VI) through a supported liquid membrane containing tri-n-octylphosphine oxide, Sep. Sci. Technol. 33 (1998) 1919–1935. [19] Z.K. Wang, C.L. Ye, X.Y. Wang, J. Li, Adsorption and desorption characteristics of imidazole-modified silica for chromium(VI), Appl. Surf. Sci. 287 (2013) 232–241. [20] P. Usapein, A. Lothongkum, P. Ramakul, U. Pancharoen, Efficient transport and selective extraction of Cr(VI) from waste pickling solution of the stainless steel-cold rolled plate process using Aliquat 336 via HFSLM, Korean J. Chem. Eng. 26 (2009) 791–798. [21] S. Lux, M. Siebenhofer, Investigation of liquidliquid phase equilibria for reactive extraction of lactic acid with organophosphorus solvents, J. Chem. Technol. Biotechnol. 88 (2013) 462–467. [22] J.B. Yang, M.Q. Yu, T. Qiu, Adsorption thermodynamics and kinetics of Cr(VI) on KIP210 resin, J. Ind. Eng. Chem. 20 (2014) 480–486. [23] H. Javadian, M. Ghaemy, M. Taghavi, Adsorption kinetics, isotherm, and thermodynamics of Hg2+ to polyaniline/hexagonal mesoporous silica nanocomposite in water/wastewater, J. Mater. Sci. 49 (2014) 232–242.