Removal of CrO4 anions from waters using surfactant ...

Separation and Purification Technology 48 (2006) 270–280

Removal of CrO4 anions from waters using surfactant enhanced hybrid PAC/MF process Canan Akmil Basar a , Coskun Aydiner b,∗ , Serdar Kara b , Bulent Keskinler b b

a Department of Chemical Engineering, Inonu University, Malatya 44069, Turkey Department of Environmental Engineering, Gebze Institute of Technology, Gebze 41400, Kocaeli, Turkey

Received 12 May 2005; received in revised form 19 July 2005; accepted 20 July 2005

Abstract The removal of CrO4 2− anions from aqueous solution using surfactant added hybrid powdered activated carbon (PAC) and crossflow microfiltration (CFMF) process was performed. Experimental runs were studied at various amounts of cetyl trimethyl ammonium bromide (CTAB) as cationic surfactant, PAC as adsorbent and CrO4 2− ions in feed solution. CFMF unit was operated at constant values of the transmembrane pressure (150 kPa), crossflow velocity (1.18 m/s) and temperature (30 ◦ C). Hybrid PAC/MF process performance was investigated in terms of CTAB and CrO4 2− rejections and, the flux decline. It was shown that the flux declined with time and the rejections varied for various properties of feed solution. CTAB and CrO4 2− rejections could be achieved as 91 and 97.2% at the conditions of 0.5 g PAC/L, 5 mM CTAB, 0.2 mM CrO4 2− and 120-min process time. It was found that over the critic micelle concentration (CMC), both CTAB and CrO4 2− rejections were in the higher values according to the values under the CMC. In addition, membrane blocking mechanisms were investigated to clarify the cake formation mechanism due to PAC. For all experimental conditions, the best results of blocking models obtained for cake filtration, intermediate blocking and standard pore blocking, respectively. It was understood that these blocking mechanisms occurred simultaneously on membrane surface in terms of high r2 values. A cake layer formed on membrane surface (cake filtration). Pore entrance blocked partially due to particles bridging over pore opening (intermediate blocking). Free CTAB aggregates with or without CrO4 2− anions, entered into the membrane pores, caused to a layer formation by adsorption and entrapment on the membrane’s pore walls (standard pore blocking). Besides, cake resistance (Rc ), specific cake resistance (α), modified fouling index (MFI) and total dried solid mass of cake per unit membrane area (ω) were also determined to comprehend the reasons of clogging within the membrane and on the surface due to both CTAB aggregates and PAC, respectively. Membrane fouling (MFI) increased with the increasing of PAC, CTAB and CrO4 2− amounts in feed solution. As a result, flux, rejections and fouling in membrane were shown as a dynamic function of PAC, CTAB and CrO4 2− as the properties of the feed solution. © 2005 Elsevier B.V. All rights reserved. Keywords: Hybrid PAC/MF process; CrO4 removal; Surfactant; Crossflow microfiltration; Membrane fouling

1. Introduction Heavy metals in water and wastewater are hazardous to environment and human body due to their accumulation in the food chain and their persistence in nature. Therefore, the removal of heavy metal contaminants from aqueous waste streams is one of the most important environmental problems related to serious health hazardous and threatens water supplies [1,2]. In order to prevent the heavy metal pollu∗ Corresponding author. Tel.: +90 262 7542360x2316; fax: +90 262 7542382. E-mail address: [email protected] (C. Aydiner).

1383-5866/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.seppur.2005.07.033

tion in water environment, chemical precipitation, electrodeposition, solvent extraction, ion exchange, adsorption and biological treatment methods are generally used as well as membrane process. Despite the fact that there are many different treatment alternatives in application, the adsorption process is to be the most effective technique in terms of avoiding low efficiency in diminishing of metal ions to trace levels and high costs [3]. But this technique is a slow process and the performance of the process is limited by equilibrium [4]. Membrane filtration processes are widely used in various water and wastewater treatment applications. In the removal of heavy metals from aqueous solutions by membrane process, reverse osmosis (RO) [5], nanofiltration (NF)

C.A. Basar et al. / Separation and Purification Technology 48 (2006) 270–280

271

Nomenclature membrane filtration area (m2 ) Brunauer–Emmett–Teller algorithm crossflow microfiltration critic micelle concentration metal concentration in feed solution (mM) metal concentration in permeate (mM) surfactant concentration in feed solution (mM) surfactant concentration in permeate (mM) the suspended solid concentration in the feed solution (g/L) CTAB cetyl trimethyl ammonium bromide dp particle diameter (m) ED electrodialysis J permeate flux (m3 /(m2 h)) initial permeate flux (m3 /(m2 h)) Jo k mass transfer coefficient (s1−n /mn−2 ) kb mass transfer coefficient for complete pore blocking (1/h) kc mass transfer coefficient for cake filtration (h/m2 ) ki mass transfer coefficient for intermediate blocking (m) kS mass transfer√coefficient for standard pore blocking (1/( mh)) MEUF micellar-enhanced ultrafiltration MF microfiltration MFI modified fouling index (s/m2 ) total dried mass of cake (kg) MP MT total mass in feed (kg) n filtration constant which indicates a value for each fouling mechanism in Hermia’s equation (0, 1.0, 1.5 and 2.0) NF nanofiltration p porosity of cake layer (%) PAC powdered activated carbon Rc cake resistance (m-1 ) Rm hydrodynamic resistance of membrane (m−1 ) RMe metal rejection total resistance in membrane (m−1 ) RT RO reverse osmosis RS surfactant rejection (%) t filtration time (h) TOC total organic carbon UF ultrafiltration UV ultraviolet V the permeate volume per unit filtration area (m3 /m2 ) Vp total volume of permeate (m3 )

Am BET CFMF CMC CMeF CMeP CSF CSP CSS

Greek letters α specific cake resistance (m/kg) P transmembrane pressure (kPa)

µ ω

dynamic viscosity of permeate (Pa s) total dried mass of cake per unit membrane area (kg/m2 )

[6,7], electrodialysis (ED) [8,9] and micellar-enhanced ultrafiltration (MEUF) [10–13] are practically used. On the other hand, among the membrane processes, microfiltration (MF) and UF are especially preferred by reason of low pressure requirement and flexible operation in water and wastewater treatment. Although the high efficiency is achieved for solid–liquid separation by MF and UF membranes, these membranes cannot remove the soluble organics and inorganic impurities such as heavy metals [14]. Nevertheless, in many studies in the last couple of years, it was shown that both MF and UF can be effectively used to remove soluble materials from wastewaters as a hybrid process combined with conventional treatment methods as well as membrane bioreactors [15–30]. Hybrid or integrated membrane filtration processes compared with traditional wastewater treatment processes provide some advantages such as high quantity of treated wastewater, high removal efficiency, fouling control, low energy consumption and lower back-washing time [31–33]. Crossflow microfiltration (CFMF) has been shown effective in a great number of processes, including the removal of colloidal organic and inorganic solids and various anions and cations from aqueous streams with the aid of surfactants and macromolecules [34–36]. Surfactants are adsorbed onto solids with a high adsorptive capacity as a fast process [37]. They can be used as an additive material in surfactantbased environmental technologies to increase the removal of contaminants from water and wastewater by adsorption [38,39]. In recent years, the use of surfactants in water and wastewater treatment can be especially preferred for the preconcentration and the separation of metal ions and other toxic substances [40,41]. At the process, the metallic ion species which can bind electrostatically to the polar head of the surfactant are trapped to be attracted by the micelle surface [40,42,43]. Because of the adsorption is the best practical technique to remove metals from waste aqueous streams, PAC as common adsorbent material is preferred in the hybrid membrane applications. In the literature, there are many researches on the removal of soluble organics or inorganic materials by adsorbent/membrane system from waters and wastewaters [19,22,35,44,45]. While MEUF method has been extensively investigated for the removal of heavy metals and organic pollutants [4,10,46–48], surfactant added PAC/ MF hybrid process has not been studied in the removal of heavy metals. By this hybrid process, the adsorption capacity of PAC with surfactant could be higher than that of without surfactant. PAC, which is used as an effective adsorption material in water and wastewater treatment, could be easily removed from treated water by MF process. High metal and surfactant removal efficiencies like in MEUF could be

272


achieved. Besides, the hybrid process could be operated at less pressure ranges according to MEUF. Thus, this technique could be used as a promising technology to remove heavy metals from water and wastewater. In this study, the removal of CrO4 2− anions from aqueous solution was carried out by using surfactant added hybrid PAC/CFMF process. Experimental runs were studied at various characteristics of feed solution. The parameters investigated were the amounts of surfactant, PAC and metal contents. CFMF unit was kept constant for the transmembrane pressure (150 kPa), the crossflow velocity (1.18 m/s) and the temperature (30 ◦ C). Hybrid system performance for various conditions in feed solution was explored in terms of the surfactant and metal rejections and, for the flux decline. In addition, the blocking mechanisms of the membrane were investigated for the formation of the cake layer on the membrane surface due to PAC. Besides, cake resistance, specific cake resistance, modified fouling index and total dried solid mass of cake per unit membrane area were also determined to clarify the clogging mechanism within and on the membrane due to both surfactant and PAC, respectively.

2. Materials and methods 2.1. Materials PAC was obtained from Merck and used in the experiments as adsorbent. The BET surface area of PAC was 465 m2 g−1 . The average grain size was determined to be 30 ␮m using a particle size analyzer (Malvern). The particle size distribution of the PAC was approximately found between 1 and 100 ␮m [45]. Cetyl trimethyl ammonium bromide, CTAB, (C19 H42 NBr) was selected as cationic surfactant. CTAB was obtained from Fluka at 98% purity and was used as received. The CMC of CTAB was reported as 0.92 mM at 25 ◦ C [49]. Microfiltration membrane supplied by Schleicher & Schuell was antistrophic cellulose acetate with a pore diameter size of 0.2 ␮m. Analytical grade Na2 CrO4 ·4H2 O, was obtained from Merck at > 99% purity.

Fig. 1. Experimental set-up of continuous CFMF unit.

angular filtration channel of length 70 mm, width 40 mm and thickness 2 mm. In operation, feed tank was first filled with surfactant and metal ion solution at the concentrations of 0.2–5 mM and 0.2–0.8 mM, respectively. Then, PAC ranging between 0.5 and 2 g/L was added to the tank. Experiments were carried out at various amounts of CTAB, PAC and CrO4 2− , respectively. Membrane filter was used in order to separate PAC particles from the solution. Two manually controlled valves were placed to the inlet and the outlet of filtration cell to create a pressure difference across the membrane. The pressure values were measured in the inlet and the outlet of the filtration unit, and underneath the membrane using pressure gauges. The filtrate passing through the filter was re-circulated to the feeding tank to maintain similar conditions throughout the experiment. However, discrete samples were taken and recorded on the computer for the analyses of surfactant and metal concentrations. During the filtration, permeate flow rate, feed flow rate, temperature and transmembrane pressure values were also monitored and recorded on the computer. Permeate flow rate was measured with a balance (Precisa 300 s) at different time intervals throughout the filtration and recorded on the computer through all experiments.

2.2. Methods 2.2.1. Experiments The experimental set-up of CFMF was presented in Fig. 1. The set-up was equipped with a flow circuit in which 20 L of feed solution containing surfactant, metal ion and PAC. The feed solution was pumped continuously through a crossflow filtration cell. The crossflow velocity and transmembrane pressure were kept constant throughout experiments at 1.18 m/s and 150 kPa, respectively. The temperature of the feed suspension was hold up at 30 ◦ C using a plate heat exchange placed in the feed tank, which had its own cooling circuit. The filtration cell was constructed from stainless steel. Flat sheet membrane with effective surface area of 28 cm2 was placed on the flat cell surface to form a one-sided rect-

2.2.2. Analysis 2.2.2.1. Surfactant and metal concentrations. CTAB was analyzed by UV-absorption method using UV (UV-160A SHIMADZU). UV absorption for aqueous samples was measured at 375 nm of wavelength. The results were also checked by measurement of TOC (Becham 915A) in these samples. Reproducibility was confirmed as ± 0.9%. The analyses of Cr(VI) in aqueous samples were determined using a spectrophotometer (UV-160A SHIMADZU) at 540 nm after complexation with 1,5-diphenylcarbazide according to the standard methods [50]. Reproducibility was confirmed as ± 1%. In addition, the viscosity values of the permeating solutions were measured by Brookfield LVDV-E Viscometer. They were found same as that of water (10−3 Pa s).


273

Table 1 The main and linearized model equations of membrane pore blocking Model

Main equation

Linearized equation

e−Kb t

ln(J −1 ) = ln(Jo−1 ) + kb t −1/2 J −1/2 = Jo + ks t J −1 = Jo−1 + ki t J −2 = Jo−2 + kc t

2.2.2.2. Flux decline and membrane blocking. In the literature, there are too many models to predict flux decline during MF. The oldest of these models is the resistance model based on the cake filtration theory. In this theory, a sieving mechanism is dominant and a cake layer of rejected particles forms on the membrane surface, when a suspension contains particles which are too large to enter the membrane pores. The cake layer called secondary membrane formed on membrane surface provides an additional resistance to filtration. It was assumed that the cake layer and membrane may be considered as main resistances in series that cause to decline of the permeate flux with time. The permeate flux is described by Darcy’s law [51–53].

where t is the filtration time (s), V the permeate volume per unit filtration area, µ the dynamic viscosity of permeate, Rm the hydrodynamic resistance of membrane, P the applied transmembrane pressure, α the specific cake resistance and CSS is the suspended solid concentration in the feed solution. MFI is defined as the gradient of the linear region found in the linear relationship between t/V and V:

Complete pore blocking Standard pore blocking Intermediate blocking Cake filtration

J≡

1 dVp P = Am dt µRT

J J J J

= Jo −2 = Jo (1 + (1/2)Ks (AJo )1/2 t) −1 = Jo (1 + Ki AJo t) −1/2 = Jo (1 + 2Kc (AJo )2 t)

(1)

where J is permeate flux, Am membrane filtration area, Vp total volume of permeate, t filtration time, P transmembrane pressure, µ dynamic viscosity of permeate and RT is total resistance in membrane. The mechanism of membrane blocking is analyzed according to the blocking laws. These laws that describe the transition mechanism from pore blocking to cake formation are generally expressed with a physical model developed by Hermia as follows [33,52,53]: n d2 t dt = k 2 dV dV

(2)

where k is mass transfer coefficient and n is filtration constant. The values of n of 0, 1, 1.5 and 2 define cake, intermediate, standard and complete blocking filtration, respectively. Eq. (2) is frequently used to identify the filtration mechanism using experimental data of membrane filtration. The main and linearized equations of each blocking model are achieved with substituting n values in Eq. (2) and rearranging this equation. The both equations of the membrane blocking models were given in Table 1 [53–55].

MFI =

µαCss 2P

(4)

The specific cake resistance, α representing the hydrodynamic resistance to the flow due to secondary membrane, is an important parameter. This parameter can be used for better comprehending the reasons of the permeate flux decline with fouling in membrane. Therefore, it is valuable to investigate the variation of α by system parameters such as characteristics of feed solution, membrane type and pore size, applied P and crossflow velocity, etc. α is calculated from the value of MFI using Eq. (4) [35]. 2.2.3. Calculations 2.2.3.1. Surfactant and metal rejections. The efficiency of MF process was evaluated using surfactant and metal rejections, which were determined using the following relationship, respectively: RS = 1 −

CSP CSF

RMe = 1 −

CMeP CMeF

(5) (6)

where RS and RMe are the rejections, CSP and CMeP , the concentrations in permeate (mM) and CSF and CMeF are the concentrations in feed solution (mM) for surfactant and metal, respectively. 2.2.3.2. Cake properties. The characteristics of the cake layer deposited on the membrane surface were evaluated. For this, total dried mass of cake (MP ) was gravimetrically measured according to the standard methods for each experimental run [50]. The total dried mass of cake per unit membrane area (ω) was determined using Eq. (7):

2.2.2.3. Cake filtration. Modified fouling index (MFI), based on the cake filtration theory, is widely used for the prediction of the fouling potential of membrane processes. MFI is analyzed from the plot of t/V versus V using general cake filtration equation in constant pressure [35,56,57],

ω=

t µRm µαCss V = + V P 2P

The transition of both PAC particles and surfactant as monomers and micelles to the membrane was investigated

(3)

MP Am

(7)

274


Fig. 2. The values of the (a) rejections, (b) flux and (c) total resistance obtained for various CTAB amounts (PAC = 0.5 g/L and CrO4 2− = 0.2 mM, t = 120 min).

in respect of the deposition of mass within and on the membrane from the feed. The transition is determined with MP /MT ratio, where MP and MT are the total dried mass of cake and the total mass in feed, respectively. On the other hand, the cake resistance (Rc ) is calculated by means of using the Eq. (8). Rc = αω

(8)

The portioning of Rc to the total resistance was determined from Rc /RT ratio.

3. Results and discussion 3.1. The effect of the variation of CTAB amount The effect of CTAB amount on the surfactant and CrO4 2− rejections was investigated for various concentrations (0.2, 1.0 and 5.0 mM) of the under and over the CMC at 120 min process time. The relationship for the flux, J and the total resistance in membrane, RT versus time was determined. The results were presented in Fig. 2. As seen from Fig. 2(a–c), the rejections and RT increased slightly opposed to J values when the CTAB concentration increased from 0.2 to 5 mM. Under the CMC, both CTAB and CrO4 2− rejections were in the lower values than that of the values above the CMC. The increasing of the rejections with CTAB amount above the CMC can be explained as the formation of surfactant aggregates (bilayers, hemi-micelles and

micelle like structures) in the feed solution. Main removal mechanisms seem to be interaction of CTAB monomers and aggregates with PAC for under and above CMC, respectively. The micelles could not pass through the membrane. They caused to a secondary layer formation both on membrane surface and within membrane pores. [45]. Therefore, the result of the increased rejections of CTAB and CrO4 2− above the CMC was most likely the secondary membrane formation by the surfactant micelles. This phenomenon also increased the total resistance and decreased the flux values. On the other hand, it was obviously seen from Fig. 2 that the CrO4 2− rejection was higher than the CTAB rejection, especially under the CMC. This phenomenon occurs due to trapped of CrO4 2− ions by amphiphilic surfactant aggregates adsorbed on PAC, which is described as adsolubilization in the literature [58]. Consequently, it can be explained that the “adsorption/adsolubilization” of CrO4 2− ions by means of PAC combined with surfactant aggregates is more effective to achieve high CTAB and CrO4 2− rejections, than that of with CTAB monomers. Several properties for membrane cake layer were examined to better understand the membrane fouling and the mechanisms related to the formation of cake layer on and within the membrane material and to better evaluate the effectiveness of hybrid system performance at various CTAB amounts. The characteristics parameters for cake layer were presented in Table 2. As seen from Table 2, α increased with the increasing of CTAB amount due to the hydrodynamic resistance to the

Table 2 The characteristics parameters related to cake layer formed on and within the membrane at various CTAB amounts (PAC = 0.5 g/L and CrO4 2− = 0.2 mM, t = 120 min) CTAB (mM)

α (1011 ) (m/kg)

Rc (1011 ) (m−1 )

RT (1011 ) (m−1 )

r2a

MFI (s/m2 )

ω (kg/m2 )

Rc /RT (%)

MP /MT (%)

0.2 1.0 5.0

2.108 18.759 56.578

1.355 2.680 2.021

5.274 16.829 19.228

0.9959 0.9769 0.9834

351.4 3126.5 9429.7

0.643 0.143 0.036

28.35 21.39 10.51

17.91 3.90 0.86

a

r2 values obtained from the t/V vs. V plot.


flow of secondary membrane while Rc first increased and then decreased. Membrane fouling (MFI) increased at same values with the increase levels of α. On the other hand, ω, Rc /RT and MP /MT values decreased with increasing of CTAB amount opposed to α and MFI values. Specific cake resistance, α is defined as the resistance per unit thickness of cake layer and expressed by Carmen–Kozeny equation as follows [52]: α = 180

(1 − p)2 dp2 p3

(9)

where p and dp are the porosity of cake layer and particle diameter, respectively. According to Eq. (9), α decrease with the increasing of particle diameter causing the increase of porosity. Accordingly, it can be said that particle diameter and porosity in cake layer decreased depending on the increasing of α. This result indicated that a cake layer consisting less porous and smaller particles formed on membrane with increasing of CTAB amount. The value of MP /MT decreased with increasing of CTAB. While fouling in membrane (MFI) increased with increasing of CTAB, the decreasing of MP /MT shown that much CTAB participation on and into membrane occurred. CTAB aggregates with or without CrO4 2− superseded the PAC particles in cake layer formed on membrane surface as well as free CTAB aggregates deposited within the membrane pores provided higher CrO4 2− rejection. This phenomenon caused to decrease of Rc against α which increased distinctly at 5 mM. Besides, the decreasing of Rc /RT was shown that the total of adsorption and pore blocking resistances increased due to the transition of CTAB aggregates into the membrane pores according to resistance-in-series model approach [59]. 3.2. The effect of the variation of PAC amount The effect of PAC amount on the CTAB and CrO4 2− rejections was studied at various concentrations ranging between 0.5 and 2 g/L for the process time of 60 min. In addition, the flux and total resistance variations with time were also determined. The results were shown in Fig. 3.

275

The results clearly demonstrated that CTAB and CrO4 2− rejections decreased partly when the PAC concentration increased from 0.5 to 2 g/L. Although the increasing of PAC amount anticipates increasing the removal of surfactant and metal ion, one by one, this expectation was not observed in this stage of the study due to coexistence of surfactant and metal ion with PAC in feed solution. CTAB adsorption on PAC occurs dominantly by hydrophobic bonding mechanism as well as ion exchange and ion pairing mechanisms [45,60,61]. It can be said that the direct adsorption of free CrO4 2− anions which need to long time according to that of adsorption of micellar-enhanced CrO4 2− , further occurred due to increase of PAC amount in feed solution. This situation caused to less CTAB adsorption by hydrophobic bonding on PAC surface. In this case, in spite of the increase of adsorption of free CrO4 2− anions, much micellar-enhanced CrO4 2− formed depending on much CTAB aggregates remained in feed solution. Therefore, the competitive adsorption of CTAB and CrO4 2− by PAC affected CTAB and CrO4 2− rejections due to complex interaction mechanisms. At the end of the process time, contrary to expectation, both CTAB and CrO4 2− rejections decreased with increasing of PAC amount in feed solution. When the PAC concentration increased from 0.5 to 2 g/L, CTAB and CrO4 2− rejections decreased from 88.6 and 94.8% to 78.5 and 90.2%, respectively. The decreasing levels were found approximately 11.4 and 4.9% for CTAB and CrO4 2− rejections, respectively. It can be shown that the direct adsorption of free CrO4 2− anions on PAC surface was more effective than that of micellar-enhanced. On the other hand, RT increased clearly opposed to J values with process time. The increasing of RT suggested that further PAC particles participated into the secondary membrane. However, it is not clear that how the transition of free CTAB aggregates into the membrane pores affect the total resistance in membrane. Flux decline with time occurred especially at 2 g/L compare with 0.5 and 1 g/L values. The data on the variation of characteristics properties of cake layer with the variation of PAC amount were given in Table 3.

Fig. 3. The values of the (a) rejections, (b) flux and (c) total resistance obtained for various PAC amounts (CrO4 2− = 0.2 mM and CTAB = 5 mM, t = 60 min).

276


Table 3 The characteristics parameters related to cake layer formed on and within the membrane at various PAC amounts (CrO4 2− = 0.2 mM and CTAB = 5 mM, t = 60 min) PAC (g/L)

α (1011 ) (m/kg)

Rc (1011 ) (m−1 )

RT (1011 ) (m−1 )

r2a

MFI (s/m2 )

ω (kg/m2 )

Rc /RT (%)

MP /MT (%)

0.5 1.0 2.0

67.716 37.83 30.1065

1.557 1.666 2.796

16.829 17.550 22.751

0.9979 0.9980 0.9995

11286 12610 20071

0.023 0.044 0.093

9.26 9.50 12.27

0.55 0.57 0.63

a

See Table 2.

It was seen from Table 3, α decreased with increasing of PAC opposed to Rc variation while membrane fouling (MFI), ω, Rc /RT and MP /MT values increased. These variations exhibit that there are some differences in the formation mechanisms of cake layers occurred at various amounts of PAC and CTAB. According to Eq. (9), particle diameter and porosity increased depending on the decreasing of α value with increasing of PAC. Cake layer consisting bigger particles as to be more porous formed on membrane surface. This result indicated that the participation of PAC particles to the membrane increased partially with increasing of PAC. Therefore, both Rc and Rc /RT values like that of MFI and ω values increased. The value of MP /MT increased very few with increasing of PAC. It was provided the more contact of free CrO4 2− anions with PAC as well as CTAB aggregates by means of retaining of more PAC particles in feed solution. 3.3. The effect of the variation of CrO4 2− amount The effect of CrO4 2− amount on the CTAB and CrO4 2− rejections was carried out at various concentrations ranging between 0.2 and 0.8 mM for the process time of 60 min. The variations of the flux and the total resistance in membrane versus time were determined and the results were given in Fig. 4. CTAB rejection increased distinctly while CrO4 2− rejection decreased partly with increasing of CrO4 2− amount from 0.2 to 0.8 mM. It was shown in our previous study [60] that PAC has negative zeta-potential value. CTAB adsorption on PAC causes to raise the positive values of zeta-potential of PAC. On the other hand, the CrO4 2− addition to feed solu-

tion decreases the zeta-potential value of CTAB adsorbed PAC due to the adsorption and ion-pairing. At this point, it might be said that the increasing of CrO4 2− concentration in feed solution induced to decrease the zeta-potential of PAC and then the CTAB rejection increased. However, it was not observed the same variation trend for CrO4 2− rejection. Due to the adsorption of much CTAB aggregates by PAC particles, less micellar-enhanced CrO4 2− remained in feed solution. In addition, less CTAB aggregates were deposited in membrane pore and the effectiveness of enhancement of free CrO4 2− anions by free CTAB aggregates in pore walls decreased. RT increased clearly opposed to J values with process time. The variation of RT and J values was approximately found at the same trends of variation with time. This result gives an expected situation that for all CrO4 2− concentrations in feed the same removal mechanism forms stably. The increasing of RT or the decreasing of J indicated that further fouling occurred within or on membrane. In our previous study [62], CrO4 2− removal by CTABenhanced CFMF process was studied. In this study, the effect of the addition of PAC to CTAB-enhanced CFMF process on the removal of CrO4 2− from water was investigated. The CrO4 2− and CTAB rejections for both studies were presented in Table 4. It was shown from Table 4 that while the removal of CrO4 2− by CTAB-enhanced CFMF process was approximately in the range of 18–78%, it could be increased in the range of approximately 88–90% with the addition of PAC to the CTAB-enhanced CFMF process. Besides, CTAB rejection that was in the range of approximately 10–54% for CTAB-enhanced CFMF, was increased to 78.5–92.5 by

Fig. 4. The values of the (a) rejections, (b) flux and (c) total resistance obtained for various CrO4 2− amounts (CTAB = 5 mM and PAC = 2 g/L, t = 60 min).


277

Table 4 CrO4 2− and CTAB rejections in CTAB-enhanced CFMF process and CTAB-enhanced PAC/CFMF hybrid process CTAB-enhanced CFMFa

CTAB/CrO4 2−

25.00 12.50 6.25 a

CTAB-enhanced PAC/CFMF

RMe (%)

RS (%)

RMe (%)

RS (%)

77.9 39.2 18.1

53.6 23.1 10.4

90.2 89.9 88.0

78.5 86.0 92.5

Experimental conditions: CTAB = 4.5 mM, P = 150 kPa, temperature = 30 ◦ C, crossflow velocity = 6 m/s.

CTAB-enhanced PAC/CFMF. These results presented that the addition of PAC to surfactant-enhanced MF process increased both CrO4 2− and CTAB rejections to the level about 90%. The data on the variation of characteristics properties of cake layer with the variation of CrO4 2− amount were given in Table 5. As seen from Table 5, all parameters related to the resistances in membrane and total dried mass deposited within or on membrane increased with increasing of CrO4 2− amount. According to Eq. (9), particle diameter and porosity decreased depending on the increasing of α value with increasing of CrO4 2− amount. Cake with smaller particle diameter formed a less porous layer on membrane surface. The variation of ␻ and MP /MT values indicated that the total mass of the cake increased. These results exhibited that much free CTAB aggregates participated to the cake layer with increasing of CrO4 2− amount. On the other hand, Rc values increased higher levels according to α and MFI values which varied at same levels with increasing of CrO4 2− amount. This result showed that free CTAB aggregates were further deposited in membrane in addition to the participation of them to the cake layer. Besides, the increasing of Rc and Rc /RT values demonstrated the increasing of cake resistance because of the participation of free CTAB aggregates to both cake layer and membrane’s inner. Therefore, surfactant rejection increased with increasing of CrO4 2− amount. However, the increasing levels of all parameters related to the characteristics of cake layer were not higher than that of CrO4 2− amount. This put forward that although CTAB rejection increased by means of the PAC, the participation and the deposition occurred to the cake layer and within the membrane pores, respectively. However, CrO4 2− rejections partially decreased because of less micellar-enhanced CrO4 2− remained in feed solution. As a result, less effectiveness for the enhancement of free CrO4 2− anions by free CTAB aggregates in pore walls

was relatively achieved according to increasing of CrO4 2− amount. 3.4. The analysis of blocking mechanisms in membrane Membrane pore blocking models were used to comprehend the blocking within and on membrane surface by PAC particles and CTAB aggregates. The results for various properties of feed solution were presented in Table 6. It was clearly seen that for all experimental conditions, the best results of blocking models obtained for cake filtration, intermediate blocking and standard pore blocking, respectively. It was seen that cake formation was determined as dominant mechanism for membrane blocking because of n values changed from 0.222 to 0.673 in all experimental runs. On the other hand, it can be said that these three blocking mechanisms occurred simultaneously on membrane surface due to high r2 values. Among these three models, cake filtration and intermediate blocking models presented better agreements with results. This explained that both cake layer formed on membrane surface (cake filtration) and pore entrance was partially blocked due to particles bridging over pore opening (intermediate blocking). In spite of PAC having bigger particle size than the pore size of membrane, standard pore blocking mechanism was observed on the membrane. Although, PAC particles do not pass into the membrane pores, free CTAB aggregates enter into the membrane pores. The membrane wall is gradually covered by a layer of CTAB aggregates. As a result of this, the adsorption and the entrapment of free CTAB aggregates on the pore wall or in the membrane support can enhance the CrO4 2− removal. On the other hand, the results obtained for complete pore blocking reached partly to acceptable levels with increasing of both PAC and CrO4 2− amounts. This demonstrated that only a small part of PAC particles in feed solution, having similar

Table 5 The characteristics parameters related to cake layer formed on and within the membrane at various CrO4 2− amounts (CTAB = 5 mM and PAC = 2 g/L, t = 60 min) CrO4 2− (mM)

α (1011 ) (m/kg)

Rc (1011 ) (m−1 )

RT (1011 ) (m−1 )

MFI (s/m2 )

r2a

ω (kg/m2 )

Rc /RT (%)

MP /MT (%)

0.2 0.4 0.8

30.1065 35.3025 51.621

2.796 3.425 5.531

22.751 23.626 27.923

20071 23535 34414

0.9995 0.9989 0.9975

0.093 0.097 0.107

12.27 14.52 19.77

0.63 0.65 0.72

a

See Table 2.

0.552 0.127 0.572 0.773 0.847 0.816 0.794 0.917 14.835 −1.1502 0.6470 1.5480 0.6273 0.8170 0.6832 0.3225

sizes to membrane pore size, caused moderately the membrane clogging.

4. Conclusions

The value of process time is 120 min for these experimental runs while for others it is 60 min. These values could not be calculated because of the intersection of the model plot has negative value. a

0.5 0.5 0.5 0.5 1 2 2 2 2a 3a 4 5 6 7 8

b

0.4795 4.9938 5.7535 8.9420 7.3620 16.738 17.976 21.501 0.2 0.2 0.2 0.2 0.2 0.2 0.4 0.8

9.54 5.86 9.48 2.55 5.49 2.43 4.45 8.92

0.528 0.673 0.518 0.444 0.222 0.488 0.306 0.432

–b 0.580 0.825 0.552 0.662 0.564 0.374

–b

0.989 0.993 0.909 0.969 0.976 0.976 0.958 0.956

4.831 1.659 0.557 0.707 0.524 0.547 0.492 0.348

0.4090 1.3376 1.0797 1.4147 1.3914 2.6549 2.6878 2.5639

0.975 0.976 0.848 0.942 0.964 0.950 0.928 0.951

4.320 1.410 0.552 0.681 0.515 0.525 0.476 0.340

0.2837 0.5143 0.3364 0.6255 0.4301 0.7580 0.7444 0.6300

0.918 0.938 0.810 0.924 0.953 0.933 0.908 0.946

526.2 0.737 5.734 12.406 4.099 3.967 3.463 2.428

r2 Jo r2 kS Jo r2 ki kc Jo

1a

0.2 1 5 5 5 5 5 5

n k (104 ) CrO4 (mM) CTAB (mM) PAC (g/L) Run

Table 6 The results of membrane pore blocking models for all experimental runs

Cake filtration

r2

Jo

Standard pore blocking Intermediate blocking

kb


Complete Pore Blocking

278

The removal of CrO4 2− anions from aqueous solution using surfactant added hybrid PAC/CFMF process was investigated. It was found that RT increased slightly opposed to J values with increasing of PAC, CTAB and CrO4 2− amounts in feed solution. The rejections were determined to be a dynamic function of both PAC and CTAB aggregates depending on the characteristics of feed solution. Membrane fouling (MFI) increased with increasing of PAC, CTAB and CrO4 2− amounts in feed solution. Fouling in membrane was shown as a dynamic function of PAC, CTAB aggregates and CrO4 2− like in the rejections. The membrane pore blocking models were analyzed for all experimental conditions. The best results of blocking models obtained for cake filtration, intermediate blocking and standard pore blocking, respectively. These blocking mechanisms occurred simultaneously on membrane surface because of high r2 values. This situation demonstrated that a cake layer formed on membrane surface (cake filtration), pore entrance blocked partially due to particles bridging over pore opening (intermediate blocking). On the other hand, free CTAB aggregates with or without CrO4 2− anions, entered into the membrane pores, caused to an additional layer formation by adsorption and entrapment on the membrane’s pore walls (standard pore blocking). Consequently, surfactant added hybrid PAC/CFMF process seems to be very effective to remove CrO4 2− anions from water. The process can be used in the treatment of waters and wastewaters as a novel metal removal process. It would be more economical with the use of low cost adsorbent materials.

Acknowledgements The authors would like to thank to Research Council of Ataturk University due to the funding support of this study.

References [1] Y.K. Bayhan, B. Keskinler, A. Cakici, M. Levent, G. Akay, Removal of divalent heavy metal mixtures from water by Saccharomyces cerevisiae using crossflow microfiltration, Water Res. 35 (9) (2001) 2191–2200. [2] A. Bingol, H. Ucun, Y.K. Bayhan, A. Karagunduz, A. Cakici, B. Keskinler, Removal of chromate anions from aqueous stream by a cationic surfactant-modified yeast, Bioresour. Technol. 94 (2004) 245–249. [3] S.S. Banerjee, R.V. Jayaram, M.V. Joshi, Removal of nickel(II) and zinc(II) from wastewater using fly ash and impregnated fly ash, Sep. Sci. Technol. 38 (5) (2003) 1015–1032. [4] M.K. Purkait, S. DasGupta, S. De, Removal of dye from wastewater using micellar-enhanced ultrafiltration and recovery of surfactant, Sep. Purif. Technol. 37 (2004) 81–92.

C.A. Basar et al. / Separation and Purification Technology 48 (2006) 270–280 [5] J.J. Qin, M.H. Oo, M.N. Wai, C.M. Ang, F.S. Wong, H. Lee, A dual membrane UF/RO process for reclamation of spent rinses from a nickel-plating operation—a case study, Water Res. 37 (2003) 3269–3278. [6] A.W. Mohamma, R. Othaman, N. Hilal, Potential use of nanofiltration membranes in treatment of industrial wastewater from Ni–P electroless plating, Desalination 168 (2004) 241–252. [7] Y. Ku, S.W. Chen, W.Y. Wang, Effect of solution composition on the removal of copper ions by nanofiltration, Sep. Purif. Technol. 43 (2005) 135–142. [8] T. Mohammadi, A. Moheb, M. Sadrzadeh, A. Razmi, Modeling of metal ion removal from wastewater by electrodialysis, Sep. Purif. Technol. 41 (2005) 73–82. [9] I. Frenzel, H. Holdik, D.F. Stamatialis, G. Pourcelly, M. Wessling, Chromic acid recovery by electro-electrodialysis. I. Evaluation of anion-exchange membrane, J. Membr. Sci., in press. [10] C.C. Tung, Y.M. Yang, C.H. Chang, J.R. Maa, Removal of copper ions and dissolved phenol from water using micellar-enhanced ultrafiltration with mixed surfactants, Waste Manage. 22 (2002) 695– 701. [11] M. Aoudia, N. Allal, A. Djennet, L. Toumi, Dynamic micellar enhanced ultrafiltration: use of anionic (SDS)–nonionic(NPE) system to remove Cr3+ at low surfactant concentration, J. Membr. Sci. 217 (2003) 181–192. [12] K. Baek, J.W. Yang, Micellar-enhanced ultrafiltration of chromate and nitrate: binding competition between chromate and nitrate, Desalination 167 (2004) 111–118. [13] L. Gzara, M. Dhahbi, Removal of chromate anions by micellarenhanced ultrafiltration using cationic surfactants, Desalination 137 (2001) 241–250. [14] Y. Watanabe, K. Kimura, T. Suzuki, Membrane application to water purification process in Japan—Development of hybrid membrane system, in: International Specialized Conference on Membrane Technology in Environmental Management, Tokyo, Japan, November 1–4, 1999. [15] C. Campos, I. Baudin, J.M. Lainé, Adsorption performance of powdered activated carbon-ultrafiltration systems, Water Supply 1 (5–6) (2001) 13–19. [16] Y. Matsui, A. Yuasa, K. Ariga, Removal of synthetic organic chemical by PAC-UF systems—I. Theory and modeling, Water Res. 35 (2) (2001) 455–463. [17] Y. Matsui, F. Colas, A. Yuasa, Removal of synthetic organic chemical by PAC-UF systems—II. Model application, Water Res. 35 (2) (2001) 464–470. [18] D. Abdessemed, G. Nezzal, Treatment of primary effluent by coagulation–adsorption–ultrafiltration for reuse, Desalination 152 (2002) 367–373. [19] G. Klomfas, K. Konieczny, Fouling phenomena in unit and hybrid processes for potable water treatment, Desalination 163 (2004) 311–322. [20] H. Park, K.H. Choo, C.H. Lee, Flux Enhancement with powdered activated carbon addition in the membrane anaerobic bioreactor, Sep. Sci. Technol. 34 (14) (1999) 2781–2792. [21] M. Shioyama, T. Kawanishi, S. Yokoyama, M. Nuno, T. Yamamoto, Development of advanced ceramic membrane filtration system combined with ozonation and powdered activated carbon treatment, Water Supply 1 (5–6) (2001) 91–96. [22] N. Cicek, A review of membrane bioreactors and their potential application in the treatment of agricultural wastewater, Can. Biosyst. Eng. 45 (2003) 637–649. [23] G. Gehlert, M. Abdulkadir, J. Fuhrmann, J. Hapke, Dynamic modeling of an ultrafiltration module for use in a membrane bioreactor, J. Membr. Sci. 248 (2005) 63–71. [24] W.S. Guo, W.G. Shim, S. Vigneswaran, H.H. Ngo, Effect of operating parameters in a submerged membrane adsorption hybrid system: experiments and mathematical modeling, J. Membr. Sci. 247 (2005) 65–74.

279

[25] P. Schoeberl, M. Brik, M. Bertoni, R. Braun, W. Fuchs, Optimization of operational parameters for a submerged membrane bioreactor treating dyehouse wastewater, Sep. Purif. Technol. 44 (1) (2005) 61–68. [26] K.A. Matis, N.K. Lazaridis, A.I. Zouboulis, G.P. Gallios, V. Mavrov, A hybrid flotation–microfiltration process for metal ions recovery, J. Membr. Sci. 247 (2005) 29–35. [27] B. Schlichter, V. Mavrov, H. Chmiela, Study of a hybrid process combining ozonation and membrane filtration–filtration of model solutions, Desalination 156 (2003) 257–265. [28] B. Schlichter, V. Mavrov, H. Chmiel, Study of a hybrid process combining ozonation and microfiltration/ultrafiltration for drinking water production from surface water, Desalination 168 (2004) 307– 317. [29] P. Drogui, S. Elmaleh, M. Rumeau, C. Bernard, A. Rambaud, Hybride process, microfiltration–electroperoxidation for water treatment, J. Membr. Sci. 186 (2001) 123–132. [30] M.H. Al-Malack, A.A. Bukhari, N.S. Abuzaid, Crossflow microfiltration of electrocoagulated kaolin suspension: fouling mechanism, J. Membr. Sci. 243 (2004) 143–153. [31] H. Zhou, D.W. Smith, Advanced technologies in water and wastewater treatment, J. Environ. Eng. Sci. 1 (4) (2002) 247–264. [32] Saarland University, Department of Process Technology, State of the art in the removal of toxic metals and the Metassep Project: Selective separation of toxic metals from specific industrial wastewater streams for water and metals reuse—overview, The Project Report for Fifth Framework Programme of The European Commission, Saarbrücken, Germany, 2004. [33] N.K. Lazaridis, C. Blöcher, J. Dorda, K.A. Matis, A hybrid MF process based on flotation, J. Membr. Sci. 228 (2004) 83–88. [34] G. Akay, B. Keskinler, A. Cakici, U. Danis, Phosphate removal from water by red mud using crossflow microfiltration, Water Res. 32 (1998) 717–726. [35] E. Yildiz, Phosphate removal from water by fly ash using crossflow microfiltration, Sep. Purif. Technol. 35 (2004) 241–252. [36] C. Aydiner, I. Demir, E. Yildiz, Modeling of flux decline in crossflow microfiltration using neural networks: the case of phosphate removal, J. Membr. Sci. 248 (2005) 53–62. [37] C.M. Gonzalez-Garcıa, M.L. Gonzalez-Martın, A.M. GallardoMoreno, V. Gomez-Serrano, L. Labajos-Broncano, J.M. Bruque, Removal of an ionic surfactant from wastewater by carbon blacks adsorption, Sep. Sci. Technol. 37 (12) (2002) 2823–2837. [38] D. Talbot, A. Bee, C. Treiner, Adsolubilization of 4-nitrophenol at a kaolinite/water interface as a function of pH and surfactant surface coverage, J. Colloid Interface Sci. 258 (2003) 20–26. [39] K. Esumi, K. Yoshida, K. Torigoe, Y. Koide, Sorption of 2-naphthol and copper ions by cationic surfactant-adsorbed laponite, Colloid Surf. A 160 (1999) 247–250. [40] T. Cserhati, E. Forgacs, G. Oros, Biological activity and environmental impact of anionic surfactants, Environ. Int. 28 (2002) 337–348. [41] C.D. Stalikas, Micelle-mediated extraction as a tool for separation and preconcentration in metal analysis, Trends Anal. Chem. 21 (5) (2002) 343–355. [42] C.C. Tung, Y.M. Yang, C.H. Chang, J.R. Maa, Removal of copper ions and dissolved phenol from water using micellar-enhanced ultrafiltration with mixed surfactants, Waste Manage. 22 (2002) 695– 701. [43] L. Yurlova, A. Kryvoruchko, B. Kornilovich, Removal of Ni(II) ions from wastewater by micellar-enhanced ultrafiltration, Desalination 144 (1–3) (2002) 255–260. [44] R.S. Juang, W.C. Lee, C.L. Chen, Removal of sodium dodecyl benzene sulfonate and phenol from water by a combined PAC adsorption and cross-flow microfiltration process, J. Chem. Technol. Biotechnol. 79 (2004) 240–246. [45] C.A. Basar, A. Karagunduz, A. Cakici, B. Keskinler, Removal of surfactants by powdered activated carbon and microfiltration, Water Res. 38 (2004) 2117–2124.

280


[46] B. Keskinler, U. Danis, A. Cakici, G. Akay, Chromate removal from water using surfactant enhanced crossflow filtration, Sep. Sci. Technol. 32 (1997) 1899–1920. [47] J.F. Scamehorn, S.D. Christian, D.A. El-Sayed, H. Uchiyama, Removal of divalent metal cations and their mixtures from aqueous streams using micellar-enhanced ultrafiltration, Sep. Sci. Technol. 29 (7) (1994) 809–830. [48] C.K. Liu, C.W. Li, Combined electrolysis and micellar enhanced ultrafiltration (MEUF) process for metal removal, Sep. Purif. Technol. 43 (2005) 25–31. [49] M.J. Rosen, Surfactants and Interfacial Phenomena, second ed., John Wiley and Sons, New York, 1989. [50] American Public Health Association (AWWA, APHA), Standard Methods for the Examination of Water and Wastewater, 20th ed., Washington, DC, 1998. [51] M. Rahimi, S.S. Madaeni, K. Abbasi, CFD modeling of permeate flux in cross-flow microfiltration membrane, J. Membr. Sci. 255 (1–2) (2005) 23–31. [52] Y. Lee, M.M. Clark, Modeling of flux decline during crossflow ultrafiltration of colloidal suspensions, J. Membr. Sci. 149 (1998) 181–202. [53] E. Yildiz, A. Nuhoglu, B. Keskinler, G. Akay, B. Farizoglu, Water softening in a crossflow membrane reactor, Desalination 159 (2003) 139–152. [54] M.K. Purkait, P.K. Bhattacharya, S. Dea, Membrane filtration of leather plant effluent: flux decline mechanism, J. Membr. Sci. 258 (1–2) (2005) 85–96.

[55] A.L. Lim, R. Bai, Membrane fouling and cleaning in microfiltration of activated sludge wastewater, J. Membr. Sci. 216 (2003) 279– 290. [56] J.H. Roorda, J.H.J.M. van der Graaf, New parameter for monitoring fouling during ultrafiltration of WWTP effluent, Water Sci. Technol. 43 (10) (2001) 241–248. [57] S.F.E. Boerlage, M.D. Kennedy, M.R. Dickson, D.E.Y. El-Hodali, J.C. Schippers, The modified fouling index using ultrafiltration membranes (MFI-UF): characterization, filtration mechanisms and proposed reference membrane, J. Membr. Sci. 197 (2002) 1–21. [58] D. Talbot, A. Bee, C. Treiner, Adsolubilization of 4-nitrophenol at a kaolinite/water interface as a function of pH and surfactant surface coverage, J. Colloid Interface Sci. 258 (2003) 20–26. [59] I.S. Chang, S.N. Kim, Wastewater treatment using membrane filtration-effect of biosolids concentration on cake resistance, Process Biochem. 40 (2005) 1307–1314. [60] C.A. Basar, A. Karagunduz, B. Keskinler, A. Cakici, Effect of presence of ions on surface characteristics of surfactant modified powdered activated carbon (PAC), Appl. Surf. Sci. 218 (2003) 169– 174. [61] A. Gurses, M. Yalcin, M. Sozbilir, C. Dogar, The investigation of adsorption thermodynamics and mechanism of a cationic surfactant, CTAB, onto powdered active carbon, Fuel Process. Technol. 81 (2003) 57–66. [62] B. Keskinler, U. Danis, A. Cakici, G. Akay, Chromate removal from water using surfactant-enhanced crossflow filtration, Sep. Sci. Technol. 32 (11) (1997) 1899–1920.