The current study aimed to remediate membrane fouling by latex effluent by altering membrane surface charge or ionic strength of the effluent. Hydrophilic polysulfone and Ultrafilic flat membranes, with molecular weight cutoff (MWCO) of 60,000 and 100,000, respectively, and hydrophobic polyvinylidene difluoride membrane (PVDF; MWCO 100,000) were used under a constant flow rate and in cross-flow mode for ultrafiltration of latex effluent. The effect of linear alkyl benzene sulfonate (LAS) on the ionic strength of the effluent and the zeta potential of latex particles was investigated. LAS was also used to improve the anti-fouling properties of the membrane surface. The ionic strength of latex effluent was increased by raising its pH from 7 to 12, resulting in an increase of the zeta potential negativity of the latex particles from −26.61 to −42.66 mV. LAS was found to be an ineffective pretreatment for limiting the fouling propensity of latex effluent using hydrophilic membranes even at high concentration and long treatment times. It was concluded that LAS-treated membrane surface is much more favorable than pH changed feed pretreatment. The total mass of fouling decreased by 44.00 and 29.60%, when PVDF membrane surface was treated with LAS at a concentration of 1 × 10−4 g/L, and latex effluent at pH 11 was used, respectively.

Nomenclature

     
  • Cf

    Concentration of foulants in the feedwater (kg/m3)

  •  
  • mt

    Total mass of particles retained per unit membrane surface area (kg/m2)

  •  
  • mp

    Mass of particles attaching to membrane pores in a unit membrane surface area (kg/m2)

  •  
  • mc

    Total mass of particles in the cake layer per unit membrane surface area (kg/m2)

  •  
  • TMP

    Initial transmembrane pressure (psi)

  •  
  • Q

    Feed flow rate (L/min)

  •  
  • Vs

    Cumulative volume of the permeate normalized to membrane surface area (m3/m2)

  •  
  • αpm

    The attachment probabilities between a particle and the membrane (dimensionless)

  •  
  • αpp

    The attachment probabilities between two particles (dimensionless)

INTRODUCTION

The manufacture of paint products, reactor cleaning, and mixing basins generate a large quantity of wastewater. Paint effluents typically have high levels of biological oxygen demand (greater than 580 mg/L), chemical oxygen demand (greater than 5,500 mg/L), and high levels of suspended solids and turbidity (Dey et al. 2004). Therefore, the wastewater needs to be treated before it is discharged. The low-pressure membrane process is considered the most effective and sustainable method of addressing environmental problems in treating water and wastewater in order to meet or exceed stringent environmental standards (Dey et al. 2004). Nevertheless, membrane fouling is one of the primary operational concerns that hinder a more widespread application of ultrafiltration for a variety of contaminants. Examining the source and mechanisms of foulant attachment to the membrane's surface is critical when it comes to investigating membrane fouling and its practical effects. There exist two major forces contributing to foulant attachment, specifically: the dispersion interaction force and the polar interaction force (Israelachvili 1992). The Derjaguin, Landau, Verwey, and Overbeek theory quantified particle–surface interactions in aqueous environments by balancing the Van der Waals attraction force and electrostatic double layer forces between particles and the membrane's surface. These interactions elucidate the possible advantages of hydrophilizing the membrane's surface as an effective fouling remediation technique (Abdelrasoul et al. 2014a). Several methods have been used in order to increase the anti-fouling properties of hydrophobic membranes (Sui et al. 2012; Zuo & Wan 2013; Nikkola et al. 2014). Moreover, the ionic strength of the feed solution was found to significantly affect the fouling potential in wastewater effluent (Faibish et al. 1998; Jones & O'Melia 2000; Singh & Song 2005; Mika et al. 2006; Abdelrasoul et al. 2014b).

However, the experimental observations are not sufficient for a thorough understanding of the fouling potential of latex particles in solutions of different ionic strength. The relationship between the solution ionic strength and the particle-to-particle and particle-to-membrane attachment is critical for the elucidation of the underlying factors affecting membrane fouling. In our previous study, a mathematical model was developed using a homogeneous membrane with a uniform pore size for the ultrafiltration of latex effluent with a wide range of particle size distribution (Abdelrasoul et al. 2013a). This model accounts for the existing chemical attachments in membrane fouling, and incorporates the coupled effects of the chemical and physical parameters, allowing for a systematic understanding of the fouling phenomenon and its potential functions. The model has the potential to overcome the shortcomings of the various empirical and semi-empirical models (Viadero et al. 1999; Ho & Zydney 2002; Bruijn et al. 2005; Huanga et al. 2012; Abdelrasoul et al. 2013b), which are generally considered inadequate in fully assessing the effects of the membrane's physical properties on membrane fouling. As a consequence, these empirical models are insufficient in the cases where it is necessary to provide an accurate process generalization and scale-up. As we have demonstrated, the fouling attachment probabilities are dependent on the properties of foulants and membranes, operating conditions, and solution chemistry (Abdelrasoul et al. 2013c, 2014a). In addition, a mechanistic mathematical model that could be applied to heterogeneous membranes with non-uniform pore size was also developed (Abdelrasoul et al. 2014c).

Therefore, the aim of the present study is to remediate membrane fouling of latex effluent by altering the membrane surface charge or the ionic strength of the effluent either by a pH change or using anionic surfactants. The study investigated the influence of ionic strength, by means of varying the solution pH, and the effect of adding anionic surfactant on: (i) fouling attachments, (ii) the total mass of fouling, (iii) cumulative permeate volume per unit area, and (iv) specific power consumption. The impact of membrane surface treatment on improving the anti-fouling properties of the membrane and membrane fouling remediation is discussed below.

MATERIALS AND METHODS

Attachment mathematical model

A mechanistic model was developed in our previous studies for the deposition of the non-uniform latex particles on homogeneous and heterogeneous membranes with uniform and non-uniform pore size, respectively. In this model, the fouling was considered primarily with respect to the attachments among foulant entities (coagulation attachment), and with respect to the attachments between foulant and the membrane surface (depositional attachment; ) (Abdelrasoul et al. 2013a, 2014c). The model equations were used to calculate the depositional and coagulation attachments, presented in this study, using the experimentally measured values of the mass of fouling contributing to cake formation and pore blocking. The attachment probabilities have a value range from 0 to 1. Increasing the fouling attachment results in increasing the membrane fouling and vice versa.

Experimental set-up and procedure

A schematic diagram of the experimental set up is shown in Figure 1. Full details of the experimental set up, particulars of the procedure, and latex paint used can be found in our previous study (Abdelrasoul et al. 2013a). The membrane filtration unit is constructed of two acrylic plates held together by nuts and bolts. The gap between the two plates for the liquid flow was 2 mm. O-rings were used to seal the two plates of the unit. The membrane sheet area was 28 cm × 8 cm and was supported by a porous stainless steel disc, as shown in Figure 1. The procedure used to measure the total mass of fouling (mt), the mass of particles contributing to pore blocking (mp), and the mass of particles contributing to the cake layer (mc) was also reported previously (Abdelrasoul et al. 2013a). Each experimental run was repeated four times so as to ensure the accuracy of the results with 95% confidence intervals.
Figure 1

Schematic diagram of experimental set-up.

Figure 1

Schematic diagram of experimental set-up.

In the present study, polysulfone membrane with 60,000 molecular weight cutoff (MWCO) and chemical structure [OC6H4OC6H4SO2C6H4]n (GE Water & Process Technologies), Ultrafilic membrane with 100,000 MWCO and chemical structure (C3H3N)n (GE Water & Process Technologies), and polyvinylidene difluoride (PVDF) membrane with 100,000 MWCO and chemical structure (C2H2F2)n (Koch Membrane Systems) were used. The zeta potentials of the untreated membranes were −42.40, −41.50, and −2.50 mV for polysulfone, Ultrafilic, and PVDF, respectively. The solution temperature was maintained at room temperature (22–24°C). In order to analyze the influence of the surface charge and/or the anionic strength as the main process parameters, the ultrafiltration time for each experiment was kept constant (25 min).

For the membrane surface charge treatment with pH change, the membrane surface charge was adjusted using a solution in an MPT-2 autotitrator of the zeta potential analyzer (ZetasizerNano Series, Malvern Instruments Ltd, UK, ±0.01 MV). The zeta potential was adjusted by using 0.1N H2SO4 and 0.1N NaOH so as to increase the acidity or the alkalinity of the soaking solution. After this adjustment, the membrane sheets were immersed in alkaline or acid solutions for 2 h as this period was the optimal for the adsorption of the OH or H+ group on the membrane's surface. This time was estimated using preliminary experiments on the stability of the surface charge during 25 min of ultrafiltration. For the polysulfone membrane, zeta potentials of −10.00, −20.00, −30.00, −40.00, and −50.00 mV, respectively, were obtained at pH values of 2.4, 4.1, 5.7, 6.8, and 10.9. For the Ultrafilic membrane, zeta potentials of −15.00 and −30.00 mV, respectively, were obtained at pH values of 2.8 and 6. After each experimental run of the ultrafiltration process, the zeta potential of the back surface of the membrane sheet was measured in order to check the stability of the surface charge of the treated membranes. To increase accuracy of the zeta potential measurement, three identical types of measurements were performed. For each experimental run, a simulated latex effluent was used with a fixed pH value of 7. The zeta potential of latex particles at pH 7 was approximately −26.61 mV.

For the membrane surface charge treatments using anionic surfactant, the flat membranes were immersed in linear alkyl benzene sulfonate (LAS; chemical structure [(CH3)2(CH2)9CHC6H4-SO3Na]) at different concentrations and various treatment times. The critical micelle concentration (CMC) of LAS was 0.1 g/L. LAS was also added at different concentrations to the simulated latex effluent in order to investigate its effect on the ionic strength of the effluent and the zeta potential of latex particles. The latex effluent pH was adjusted using a pH transmitter (pH Transmitter 2100e, Mettler Toledo, Germany, <0.02 pH). The analytical methods were described in detail previously (Abdelrasoul et al. 2014a).

Specific power consumption calculations

During the filtration process, resistance to the permeate flow may increase due to the pore blockage and cake layer formation, resulting in membrane fouling. Hence, the permeate flux noticeably declines with filtration time. Higher permeate flux can be attained by augmenting the transmembrane pressure which in turn causes higher energy consumption. The specific power consumption per unit volume of filtrate is defined as: 
formula
1
where is the time-averaged transmembrane pressure throughout the filtration duration. [psia.min] can be calculated based on the area under the curve, as shown in Figure 2. Q [LPM] is the feed flow rate, and [m3] is the cumulative permeate volume.
Figure 2

Area under the curve represents AVG [psia.min].

Figure 2

Area under the curve represents AVG [psia.min].

RESULTS AND DISCUSSION

Simulated latex effluent treatment by pH change

The simulated latex effluent with a solid concentration of 1.30 kg/m3 has a pH value of 7. The zeta potential of latex particles at pH 7 is approximately −26.61 mV. As the pH increased from 7 to 11, the zeta potential negativity of latex particles increased significantly from −26.61 to −40.00 mV, as shown in Figure 3. The adsorption of OH groups on the surface of the latex particles at higher pH values, in turn, caused the negative charge on the surface to increase until it reached −42.66 mV at pH 12. In addition, the solution conductivity increased from 0.094 to 20.4 mS/cm when the solution pH increased from 7 to 11. Notably, the ionic strength is directly proportional to the solution conductivity. On the other hand, decreasing pH from 7 to 3 using sulfuric acid resulted in a substantial decrease in the zeta potential value from −26.61 to −11.20 mV. It is relevant to note that the zeta potential of the latex particles was −4.83 mV when pH was decreased to 3 using hydrochloric acid. This can be attributed to the more negative charge of sulfate ions (SO42–) in sulfuric acid in comparison to the chloride ion (Cl) of hydrochloric acid.
Figure 3

Effect of pH change of latex effluent on the zeta potential of latex particles and the membrane surface through the ultrafiltration process.

Figure 3

Effect of pH change of latex effluent on the zeta potential of latex particles and the membrane surface through the ultrafiltration process.

In addition, the zeta potential of each membrane surface was investigated at each pH value so as to simulate the effects of pH of the latex effluent through the ultrafiltration process. As the pH of the simulated latex effluent was increased from 3 to 11, the zeta potential of PVDF, Ultrafilic, and polysulfone membrane surfaces became increasingly negative: from −2.01 to −32.62 mV, −18.99 to −43.00 mV, and −5.67 to −41.98 mV, respectively, as shown in Figure 3. Notably, the hydrophilicity of the hydrophobic PVDF membrane surface was enhanced from −2.5 to −32.62 mV when the latex effluent pH was increased from 7 to 11, but this change resulted in an insignificant change in the surface of the Ultrafilic membrane from −41.5 to −43.00 mV. On the other hand, increasing the pH value from 7 to 11 resulted in a slight decrease in the membrane surface negativity of the polysulfone membrane from −42.40 to −41.98 mV. This may be attributed to the fact that 25 min of the ultrafiltration process were not sufficient enough to change the zeta potential of polysulfone membrane because of its unique chemical structure with the sulfone group.

The strongest influence of increasing the pH of simulated latex effluent through the ultrafiltration process occurred when the hydrophobic PVDF membrane was used (Figure 3). It is postulated that this strong influence is due to the high adsorption of OH groups on the hydrophobic membrane surface. At the feed flow rate of 4 LPM, feed concentration of 0.78 kg/m3, and transmembrane pressure of 15 psi, using a PVDF membrane and increasing the pH value from 7 to 11 resulted in an increase in the zeta potential negativity from −2.5 to −32.62 mV. In addition, the zeta potential negativity of latex particles increased significantly from −26.61 to −40.00 mV. As a consequence, the repulsion force between the latex particles and the membrane surface increased, causing the depositional attachment to decrease from 0.97 to 0.21. Hence, the total mass of fouling significantly decreased by 29.60 from 0.0125 to 0.0088 ± 0.0001 kg/m2, and the specific power consumption decrease substantially by 88.12% from 15.4 to 1.83 kWh/m3, while the cumulative permeate volume per unit area increase seven-fold from 0.01 to 0.07 ± 0.0002 m3/m2. Figure 4(a) shows a scanning electron microscopy (SEM) image of a PVDF membrane before the ultrafiltration process; Figure 4(b) and 4(c) are SEM images of PVDF membrane after ultrafiltration of effluent under conditions described above. At pH 7, the higher depositional attachment resulted in a greater number of latex particles deposited on the membrane surface, more blocked membrane pores, and a higher number of particles accumulated at the surface (Figure 4(b)). At pH 11 the higher repulsion force between the PVDF membrane of higher zeta potential negativity of –32.62 mV and the latex particles of higher negativity of –40.00 mV resulted in fewer particles deposited on the membrane surface and the presence of unrestricted membrane pores (Figure 4(c)).
Figure 4

SEM images of PVDF membrane surfaces (a) before ultrafiltration; (b) after ultrafiltration at [Q = 4 LPM], [Cf = 0.78 kg/m3], [15 psi] using latex effluent at pH of 7; (c) after ultrafiltration at [Q = 4 LPM], [Cf = 0.78 kg/m3], [15 psi] using latex effluent at pH of 11.

Figure 4

SEM images of PVDF membrane surfaces (a) before ultrafiltration; (b) after ultrafiltration at [Q = 4 LPM], [Cf = 0.78 kg/m3], [15 psi] using latex effluent at pH of 7; (c) after ultrafiltration at [Q = 4 LPM], [Cf = 0.78 kg/m3], [15 psi] using latex effluent at pH of 11.

The zeta potential negativity of polysulfone membrane decreased from −42.40 to −41.98 mV, while the zeta potential negativity of Ultrafilic membrane slightly increased from −41.50 to −43.00 mV. As a consequence, increasing the pH of latex effluent from 7 to 11 had an insignificant effect on the surface charges hydrophilicity improvement of polysulfone and Ultrafilic membranes. However, the increase of the solution pH had a significant effect on the latex particles’ surface charge negativity, as it increased from −26.61 to −40.00 mV. Thus, ultrafiltration using latex effluent at pH 11 and using polysulfone and Ultrafilic membranes was performed in order to investigate the effect of foulant surface charge on the membrane fouling, cumulative permeate flux, and the specific power consumption. At the feed flow rate of 4 LPM, feed concentration 1.30 kg/m3, and transmembrane pressure 25 psi, the total mass of fouling decreased by 11.11% from 0.0135 to 0.012 ± 0.0001 kg/m2, while the cumulative filtration volume increased by 10.43% from 0.115 to 0.127 ± 0.0014 m3/m2 using polysulfone membrane. Consequently, the specific power consumption was lowered by 20.56% from 2.14 to 1.7 kWh/m3.

Similarly, at the same operating conditions, the total mass of fouling decreased by 24% from 0.025 to 0.019 ± 0.0001 kg/m2, while the cumulative filtration volume was increased by 38.21% from 0.123 to 0.17 ± 0.0002 m3/m2. The decrease in total mass of fouling resulted in a decrease in the specific power consumption by 28.02% from 1.82 to 1.31 kWh/m3. As a result, the latex particles’ charge was found to have a significant effect on the repulsion/attraction forces between latex particles and membrane surfaces.

Simulated latex effluent treatment using anionic surfactant

As the anionic surfactant concentration increased from 0.0001 to 0.1 g/L, the ionic strength of the latex effluent increased, and the solution conductivity increased from 0.0944 to 6.5210 mS/cm. However, the zeta potential negativity decreased from −26.61 to −4.86 mV, as shown in Table 1. This can be attributed to the electrostatic repulsions between the highly charged latex surface and the anionic head groups. Therefore, the anionic surfactant stayed in the latex effluent, which resulted in the low electrical stability of colloids, and a significant decrease in the potential difference between the dispersion solution and the stationary layer of fluid attached to the dispersed latex particles. Consequently, the zeta potential negativity of the latex particles was reduced. Table 1 also illustrates the effects of LAS additions to the latex effluent on the zeta potential of the membrane surface through the ultrafiltration process. The LAS concentration of 0.0001 g/L was an optimum concentration for the enhancement of hydrophobic PVDF surface charge, while it had an insignificant effect on the Ultrafilic membrane at this concentration. However, the addition of LAS had the opposite effect on the polysulfone membrane (Table 1). This may be attributed to the unique chemical structure of the polysulfone membrane with the sulfone group. Accordingly, the repulsion between the functional group of anionic surfactant and the functional group of polysulfone can explain the non-changeable hydrophilicity of the membrane surface after LAS treatment.

Table 1

Zeta potential of latex particles and membrane surfaces at different LAS concentrations after ultrafiltration (15 min LAS addition)

Zeta potential of membrane surface after ultrafiltration process [mV]
LAS concentration [g/L]Zeta potential of latex particles [mV]Latex effluent conductivity [mS/cm]PolysulfoneUltrafilicPVDF
−26.61 0.0944 −42.40 −41.50 −2.50 
0.0001 −24.05 0.3601 −25.22 −41.97 −28.66 
0.001 −16.30 1.1020 −18.43 −27.51 −17.18 
0.01 −10.02 2.4030 −8.09 −18.77 −11.67 
0.1 −4.86 6.5210 0.96 3.58 −0.26 
Zeta potential of membrane surface after ultrafiltration process [mV]
LAS concentration [g/L]Zeta potential of latex particles [mV]Latex effluent conductivity [mS/cm]PolysulfoneUltrafilicPVDF
−26.61 0.0944 −42.40 −41.50 −2.50 
0.0001 −24.05 0.3601 −25.22 −41.97 −28.66 
0.001 −16.30 1.1020 −18.43 −27.51 −17.18 
0.01 −10.02 2.4030 −8.09 −18.77 −11.67 
0.1 −4.86 6.5210 0.96 3.58 −0.26 

The results in Table 1 also indicate that increasing the LAS concentration to more than 0.0001 g/L resulted in a decrease zeta potential negativity of PVDF and Ultrafilic membrane surfaces. The increase in the concentration of the anionic surfactants in the latex effluent caused a reduction in the potential difference between the effluent and the membrane surface. It should also be mentioned that the micellar ultrafiltration at the CMC of LAS of 0.1 g/L had the least favorable results for the zeta potential negativity of latex particles and featured a decrease from −26.61 to −4.86 mV. Furthermore, the zeta potential negativity of polysulfone, Ultrafilic, and PVDF changed from −42.40 to 0.96 mV, −41.50 to 3.58 mV, and −2.5 to −0.26 mV, respectively. The reason for this behavior stems from the fact that when the micelles were formed, as shown in Figure 5, they reduced the interfacial tension between the effluent and the membrane surface or the latex particle surface. In addition, the electrostatic repulsion between the highly charged membrane surfaces or the latex particles and the micelles increased, which resulted in the presence of the micelle in the solution. Hence, a substantial decrease in the potential difference between the dispersion solution and the surfaces can occur. It is notable that the maximum allowable concentration of LAS in wastewater discharged to the environment is 1 ppm, due to its high biodegradability based on its molecular weight. Hence, the toxicity of LAS at the concentration of 1 × 10−4 g/L, used in this study, was found to be safe for marine, freshwater, and estuarine environments (Morrow & Piwoni 1993).
Figure 5

Micelle formation of anionic surfactant.

Figure 5

Micelle formation of anionic surfactant.

The effect of LAS treatments for various times was investigated at a concentration of 0.0001 g/L and CMC concentration of 0.1 g/L. At zero concentration of LAS, the zeta potential of latex particles and the conductivity of simulated latex effluent are −26.61 mV and 0.0944 mS/cm, respectively. As shown in Table 2, after the simultaneous addition of LAS, the zeta potential of latex particles dropped to −0.94 and −2.13 mV, at concentration of 0.0001 and 0.1 g/L, respectively. Due to the simultaneous increase of anionic heads in the latex solutions, there was a significant change in charge of the dispersion medium with respect to the dispersed particle, which in turn resulted in a significant decrease in the potential difference between the dispersion solution and the dispersed latex particles. Consequently, the zeta potential negativity significantly decreased. The increase in latex effluent ionic strength through the addition of the anionic surfactant was due to the presence of the negative charge hydrophilic head of the anionic surfactant. Hence, the effluent conductivity was increased to 0.3070 and 6.4411 mS/cm, as LAS was simultaneously added at the concentration of 0.0001 and 0.1 g/L, respectively. The optimum time was determined to be 15 min for the addition of LAS as feed pretreatment. The zeta potential measurements of latex particles indicated that the treatment time was ineffective as a method of increasing the zeta potential negativity of latex particles, as shown in Table 2. Furthermore, the zeta potential of latex particles had very low negative values at the CMC of LAS (Table 2). Moreover, the treatment time had an insignificant effect on the solution ionic strength at both concentrations.

Table 2

Zeta potential of latex particles and latex effluent conductivity at various treatment times at LAS concentration of 0.0001 and 0.1 g/L

LAS concentration 0.0001 g/L
LAS concentration 0.1 g/L (CMC)
Time [min]Zeta potential of latex particles [mV]Latex solution conductivity [mS/cm]Zeta potential of latex particles [mV]Latex solution conductivity [mS/cm]
−0.94 0.3070 −2.13 6.4411 
15 −24.05 0.3601 −4.86 6.5210 
30 −22.30 0.3870 −1.84 6.5401 
60 −19.20 0.3870 −1.58 5.9806 
90 −19.76 0.3865 −1.64 6.0110 
LAS concentration 0.0001 g/L
LAS concentration 0.1 g/L (CMC)
Time [min]Zeta potential of latex particles [mV]Latex solution conductivity [mS/cm]Zeta potential of latex particles [mV]Latex solution conductivity [mS/cm]
−0.94 0.3070 −2.13 6.4411 
15 −24.05 0.3601 −4.86 6.5210 
30 −22.30 0.3870 −1.84 6.5401 
60 −19.20 0.3870 −1.58 5.9806 
90 −19.76 0.3865 −1.64 6.0110 

The influence of adding LAS at a concentration of 0.0001 g/L to the latex effluent 15 min before the ultrafiltration process using hydrophilic membranes was investigated. At the feed flow rate 4 LPM, feed concentration 1.30 kg/m3, and transmembrane pressure 25 psi, the total mass of fouling substantially increased from 0.0135 to 0.0931 ± 0.0001 kg/m2, while the cumulative filtration volume significantly decreased from 0.115 to 0.078 ± 0.0007 m3/m2 using the polysulfone membrane. Due to the decrease of the zeta potential negativity of latex particles from −26.61 to −24.05 mV, in addition to the substantial decrease of polysulfone surface charge negativity from −42.40 to −25.22 mV, there was a higher attraction force between foulants and the membrane surface. As a consequence, the depositional attachment increased from 0.76 to 0.91. The increase in the total mass of fouling caused an increase in the specific power consumption from 2.14 to 4.14 kWh/m3. At the same operating conditions, Ultrafilic membrane showed a slight increase in the total mass of fouling and a slight decrease in the cumulative permeate volume, from 0.025 to 0.0261 ± 0.0001 kg/m2 and from 0.123 to 0.119 ± 0.0007 m3/m2, respectively, and the specific power consumption slightly increased from 1.82 to 1.86 kWh/m3. This slight change was due more to the decrease in the zeta potential negativity of latex particles from −26.61 to −24.05 ± 0.01 mV than the slight enhancement of Ultrafilic surface hydrophilicity from −41.50 to −41.97 ± 0.01 mV. In this case, the deposition attachment was slightly increased from 0.7 to 0.72. Thus, the results indicate that the LAS addition was an ineffective pretreatment for limiting the fouling propensity of the latex effluent when using hydrophilic membranes.

However, the addition of LAS to the latex effluent had a definitively positive effect on the enhancement of the surface charge of the PVDF hydrophobic membrane from −2.5 to −28.66 mV. Thus, the repulsion force between the latex particles and the membrane surface increased even though the zeta potential negativity of the latex particles decreased from −26.61 to −24.05 mV. As a result, the depositional attachment decreased from 0.97 to 0.54, the total mass of fouling significantly decreased by 12.80% from 0.0125 to 0.0109 ± 0.0001 kg/m2, and the specific power consumption decreased by 34.15% from 15.4 to 10.14 kWh/m3. Nevertheless, the feed pretreatment, that includes the increasing of the solution alkalinity to pH 11, was found to enhance the PVDF surface charge more than the addition of 0.0001 g/L of LAS. The results obtained indicate that the total mass of fouling decreased by 29.60%, in case of the feed pretreatment when pH was increased to 11, and decreased to 12.80%, with the addition of 0.0001 g/L LAS. While the specific power consumption decreased by 88.12% in case of the feed pretreatment by pH change, and decreased by 34.15% where LAS was added.

Membrane surface treatment using anionic surfactant

As shown in Figure 6(a), the LAS treatment had a noticeable effect on hydrophobic PVDF and hydrophilic Ultrafilic membranes at low concentration of 0.0001 g/L. However, LAS treatment has an inverse effect on the polysulfone membrane. Notably, the results suggested that LAS treatment was ineffective for polysulfone membranes even in the instances where soaking was implemented for long periods of time. This may be attributed to the repulsion between the functional group of LAS and the functional group of polysulfone membrane. The original zeta potential value of each membrane is shown on the y-axis in Figure 6(a). The optimum enhancement of membrane surface hydrophilicity occurred during the LAS treatments at a concentration of 0.0001 g/L. The optimum time for LAS treatment was 15 and 20 min, for Ultrafilic and PVDF membranes, respectively. At the CMC, LAS treatment had an opposite effect on the zeta potential of membrane surfaces due to micelle formation, as shown in Figure 6(b). The surface charge negativity was turned into a positive charge from −42.40 to 1.62 mV, −41.50 to 4.07 mV, and −2.50 to 0.26 mV, for the polysulfone, Ultrafilic, and PVDF membranes, respectively. Figure 7(a) and 7(b) demonstrate the attraction between anionic surfactant hydrophilic heads and hydrophilic surfaces, as well as between anionic surfactant hydrophilic heads and hydrophobic membrane surfaces.
Figure 6

Zeta potential of membrane surfaces after LAS treatment at a concentration of (a) 1 × 10−4 g/L and (b) 0.1 g/L.

Figure 6

Zeta potential of membrane surfaces after LAS treatment at a concentration of (a) 1 × 10−4 g/L and (b) 0.1 g/L.

Figure 7

Schematic of adsorption LAS into the membrane surface: (a) hydrophilic membranes; (b) hydrophobic membranes.

Figure 7

Schematic of adsorption LAS into the membrane surface: (a) hydrophilic membranes; (b) hydrophobic membranes.

It should be mentioned that the LAS treatment for the membrane surface had a more noticeable effect than the addition of LAS to the latex effluent as a feed pretreatment before the ultrafiltration process. For instance, the zeta potential negativity of PVDF membrane was enhanced from −2.50 to −28.66 mV, and from −2.50 to −40.90 using LAS as feed pretreatment, and for membrane surface treatment, respectively. Furthermore, in the cases mentioned the negative effects of LAS addition to the effluent, as a feed pretreatment, on the zeta potential negativity of latex particles was avoided.

As shown in Figure 6, the results obtained indicated that the most significant effect of LAS, at a low concentration of 1 × 10−4 g/L, was on the hydrophobic PVDF membrane by increasing the membrane surface charge from −2.50 to −40.90 mV after 20 min of membrane treatment. This increase was due to the high electrostatic attraction between the anionic heads of LAS and the PVDF membrane surface, as shown in Figure 7(b). At the feed flow rate of 4 LPM, feed concentration of 0.78 kg/m3, and transmembrane pressure of 15 psi, the repulsion force between the latex particles of untreated latex effluent and the membrane surface increased. As a consequence, the depositional attachment (αpm) decreased from 0.97 to 0.1. Hence, the total mass of fouling experienced a substantial decrease of 44.00%, from 0.0125 ± 0.0001 to 0.007 ± 0.0001 kg/m2, the specific power consumption significantly decreased by 92.14%, from 15.4 to 1.21 kWh/m3, while the permeate flux perceptibly increased from 0.01 to 0.124 ± 0.0012 m3/m2. Figure 8(a) and 8(b) show SEM images of PVDF membrane after ultrafiltration at the operating conditions described using untreated and treated PVDF membrane surfaces, respectively. Figure 8(b) indicates less attachment and deposition of latex particles on the PVDF membrane surface.
Figure 8

SEM images of PVDF membrane surfaces after ultrafiltration at [Q = 4 LPM], [Cf = 0.78 kg/m3], [15 psi]: (a) at original zeta potential of −2.50 mV; (b) at zeta potential of −40.91 mV after 20 min treatment for the surface charge using LAS with concentration of 0.0001 g/L.

Figure 8

SEM images of PVDF membrane surfaces after ultrafiltration at [Q = 4 LPM], [Cf = 0.78 kg/m3], [15 psi]: (a) at original zeta potential of −2.50 mV; (b) at zeta potential of −40.91 mV after 20 min treatment for the surface charge using LAS with concentration of 0.0001 g/L.

It should be also mentioned that for treated PVDF membrane surface, using anionic surfactants at a low concentration of 0.0001 g/L had more favorable results than the best results obtained by feed pretreatment occurring due to the pH change. As such, the total mass of fouling decreased by 44.00 and 29.60%, respectively, for membrane treated with anionic surfactant, and for latex effluent feed at pH 11. Specific power consumption was substantially decreased by 92.14 and 88.12%, respectively, for membrane treated with anionic surfactant, and for latex effluent feed at pH 11. Although when the pH of the latex effluent was increased to 11 the membrane surface charge was changed from −2.50 to −32.62 mV due to the adsorption of the OH group on the membrane surface, in addition to the significant increase of the latex particle surface charge from −26.61 to −40.00 mV. As a consequence, the repulsion force between the membrane surface and the particles increased and the depositional attachment, αpm, decreased from 0.97 to 0.21. However, when the ultrafiltration process was performed with the PVDF-treated surface using anionic surfactants (zeta potential −40.90 mV) with untreated simulated latex effluent (particle zeta potential −26.61 mV), the depositional attachment was 0.1. This gives a strong indication of the major contribution of membrane surface charge to the attachment force between the membrane surface and the foulants, in comparison to the potential effects of the latex particle's charge.

In addition, the ultrafiltration process was performed using Ultrafilic membranes treated with 0.0001 g/L LAS for 15 min. At feed flow rate of 1 LPM, transmembrane pressure of 35 psi, and feed concentration of 1.3 kg/m3, the total mass of fouling decreased by 28.05% from 0.0278 ± 0.0001 to 0.02 ± 0.0016 kg/m2, due to the increase in surface charge negativity from −41.50 to −50.67 mV, which caused a significant decrease in the attraction force between latex particles and Ultrafilic membrane surface. Hence, the depositional attachment decreased from 0.98 to 0.84. Figure 9 shows SEM images of the Ultrafilic membranes after ultrafiltration using treated and untreated membranes. The latex particle deposition decreased under the same operating conditions (Figure 9(b)).
Figure 9

SEM images of Ultrafilic membrane surfaces after ultrafiltration at [Q = 1 LPM], [Cf = 1.3 kg/m3], [35 psi]: (a) at original zeta potential of −41.50 mV; (b) at zeta potential of −50.67 mV after 15 minutes treatment for the surface charge using LAS with concentration of 0.0001 g/L.

Figure 9

SEM images of Ultrafilic membrane surfaces after ultrafiltration at [Q = 1 LPM], [Cf = 1.3 kg/m3], [35 psi]: (a) at original zeta potential of −41.50 mV; (b) at zeta potential of −50.67 mV after 15 minutes treatment for the surface charge using LAS with concentration of 0.0001 g/L.

pH treatment for membrane surface charge

A change in the polysulfone surface charge from −10 to −50 mV was obtained with pH treatments of 2.4–10.9, respectively, using a 2-h soaking in the basic solution. Under operating conditions of 25 psi, 4.5 LPM, and 1.30 kg/m3, increasing the zeta potential of the membrane surface from −10.00 to −50.00 mV, resulted in a 65% reduction of depositional attachment (αpm) from 0.99 to 0.35, as shown in Figure 10(a). This could be attributed to the increased hydrophilicity of the membrane, which was created upon introducing more negative charges on the membrane surface. As a result, the electrostatic attraction force between the latex particles and the higher negatively charged membrane surface was significantly decreased. The depositional attachment was thus notably reduced. On the other hand, increasing the zeta potential negativity of the membrane surface caused an insignificant decrease in the coagulation attachment (αpp) by 5.3%, from 0.75 to 0.71, as shown in Figure 10(a). The decrease in the depositional attachment resulted in a significant increase in the cumulative filtration volume per unit area (Vs) from 0.015 to 0.123 ± 0.0018 m3/m2, an augmentation of about 10-fold, as shown in Figure 10(a). This could be attributed to the significant reduction in the depositional attachment that resulted in a lower frequency of particle attached to the membrane pores, i.e. less pore blockage for the filtrate passage through the membrane. As a consequence, the total mass of fouling diminished by 61%, from 0.018 to 0.007 ± 0.0001 kg/m2, as shown in Figure 10(b). Decreasing the mass of fouling resulted in a lower rate of the transmembrane pressure increase during the filtration process. Accordingly, the specific power consumption decreased dramatically by 92.5%, from 24.83 to 1.86 kWh/m3, as shown in Figure 10(b). These observations suggest that altering the particle-to-membrane attachment (αpm) by manipulating the zeta potential of the membrane surface could be essential in fouling remediation. Figure 11(a) is an SEM image of polysulfone membrane with the zeta potential of −50.00 mV after ultrafiltration at a transmembrane pressure of 25 psi, a feed flow rate of 4.5 LPM, and a feed concentration of 1.30 kg/m3. Figure 11(b) shows an SEM image of polysulfone membrane with the zeta potential of −10.00 mV after ultrafiltration at the same operating conditions. As shown in Figure 11(a), the lower depositional attachment caused a decrease in the mass of fouling, due to the reduced chances for particles to participate in particle-to-membrane attachment.
Figure 10

Effect of the zeta potential of polysulfone membrane surface at [25 psi], [Q = 4.5 LPM], [Cf = 1.3 kg/m3] on (a) fouling attachment probabilities (αpp, αpm) and cumulative filtration volume per unit area (Vs) [m3/m2]; (b) total mass of fouling (mt) [kg/m2] and the specific power consumption [kWh/m3].

Figure 10

Effect of the zeta potential of polysulfone membrane surface at [25 psi], [Q = 4.5 LPM], [Cf = 1.3 kg/m3] on (a) fouling attachment probabilities (αpp, αpm) and cumulative filtration volume per unit area (Vs) [m3/m2]; (b) total mass of fouling (mt) [kg/m2] and the specific power consumption [kWh/m3].

Figure 11

SEM images of polysulfone membranes after ultrafiltration at [25 psi], [Q = 4.5 LPM], [Cf=1.3 kg/m3] at the zeta potential of (a) −50.00 mV; (b) −10.00 mV.

Figure 11

SEM images of polysulfone membranes after ultrafiltration at [25 psi], [Q = 4.5 LPM], [Cf=1.3 kg/m3] at the zeta potential of (a) −50.00 mV; (b) −10.00 mV.

At the same operating conditions, the reduction of the negativity of the surface charge of Ultrafilic membrane, from −41.50 to −30.00 mV, caused a noticeable decrease in repulsion force between the membrane surface and latex particles. This in turn caused the depositional attachment (αpm) to increase from 0.70 to 0.85. A significantly higher αpm facilitated a larger number of particle attachments to the membrane surface, thus resulting in a greater pore blockage, a lowered cumulative filtration volume per unit area from 0.123 to 0.115 ± 0.0007 m3/m2, power consumption upsurge from 1.82 to 1.93 kWh/m3, and an overall increase in the mass of fouling from 0.025 to 0.03 ± 0.0003 kg/m2. A continued decrease in the surface negativity, from −30.00 to −15.00 mV, resulted in an additional increase in the depositional attachment (αpm), from 0.85 to 0.96. As a consequence, the cumulative filtration volume per unit area lowered from 0.115 to 0.03 ± 0.0012 m3/m2, and the mass of fouling was augmented from 0.03 to 0.05 ± 0.0002 kg/m2. The increase in the total mass of fouling resulted in the increase in the transmembrane pressure, which in turn caused a higher rate through the filtration process and power consumption increased from 1.93 to 7.5 kWh/m3. Notably, when using an Ultrafilic membrane with the zeta potentials of −41.50, −30.00, and −15.00 mV, the coagulation attachment (αpp) was 0.76, 0.76, and 0.77, respectively. This can be attributed to the fact that the particle-to-particle collisions and attachments are independent of the membrane surface charge. Figure 12 shows SEM images for Ultrafilic membrane after ultrafiltration at a transmembrane pressure of 25 psi, a feed flow rate of 4 LPM, and a feed concentration of 1.3 kg/m3, at the zeta potentials of −41.50, −30.00, and −15.00 mV. Figure 12(b) shows that decreasing the negativity of the surface charge of the Ultrafilic membrane caused a smaller number of clean pores, greater number of particle attachments to the membrane surface, and a higher chance of particle-to-particle collisions and attachments, when compared to Figure 12(a). Extensive particle deposition on the membrane surface was caused by a further decrease in the surface negativity, which ensured an even higher chance for the particles to contribute to the coagulation attachment and cake formation, as shown in Figure 12(c).
Figure 12

SEM images of Ultrafilic membranes after ultrafiltration at [25 psi], [Q = 4 LPM], [Cf=1.30 kg/m3] at zeta potential of (a) −41.50 mV; (b) −30.00 mV; (c) −15.00 mV.

Figure 12

SEM images of Ultrafilic membranes after ultrafiltration at [25 psi], [Q = 4 LPM], [Cf=1.30 kg/m3] at zeta potential of (a) −41.50 mV; (b) −30.00 mV; (c) −15.00 mV.

Stability of the treated membrane surface charge after the ultrafiltration process

It was necessary to check the zeta potential of the treated membrane after the ultrafiltration in order to ensure the stability of the surface charge throughout the filtration process. In the case of the stability of the pH treated membrane surface charge, polysulfone membrane zeta potentials of −10.00, −20.00, −30.00, −40.00, and −50.00 mV, respectively, were changed to the average zeta potentials of −13.20, −22.50, −32.40, −41.50, and −47.60 mV after the ultrafiltration process.

The simulated latex effluent had a pH of 7, which would affect the membrane surface charge during filtration. The zeta potential of polysulfone membrane was −42.40 mV at pH 7. It was observed that for the treated polysulfone membrane with a lower negativity, whenever the latex effluent was allowed to pass through the membrane for 25 min, the negativity of the zeta potential of membrane surface increased by 2.00–3.00 mV. On the other hand, for the treated polysulfone membranes with a negativity charge higher than the original zeta potential value, ultrafiltration of latex effluent caused a reduction of the negative charge on the membrane surface from −50 to −47.6 mV. This can be attributed to the influence of the ionic strength of the simulated effluent at pH 7 on the zeta potential of the treated membrane surfaces at higher pH values. On the other hand, the surface charge of the treated membranes using anionic surfactant was stable enough after the ultrafiltration of simulated effluent at pH 7. As such, the surface charge negativity of the treated PVDF membrane decreased only by 18.00% from −40.90 to −33.27 mV after 25 min of filtration time, when compared to the significant enhancement of the surface charge negativity of around 16-fold from −2.50 to −40.90 mV after 20 min of membrane treatment. Moreover, the surface charge negativity of treated Ultrafilic membrane decreased by 4.60% from −50.67 to −48.34 after 25 min of filtration time, if compared to the increased surface charge negativity by 22.10% from −41.50 to −50.67 mV after 15 min of membrane treatment. Table 3 summarizes the experiments conducted for each fouling remediation technique.

Table 3

The surface charges and depositional attachments for the experiments for three fouling remediation techniques: (a) simulated latex effluent treatment by pH change; (b) simulated latex effluent treatment using anionic surfactant; (c) membrane surface treatment using anionic surfactant; (d) pH treatment for membrane surface charge

 Operating conditions
Membrane usedQTMPCfZeta potential of untreated membrane surfaceZeta potential of untreated latex particlesαpmRemediation techniqueZeta potential of membrane surface after treatmentZeta potential of latex particles after treatmentαpm
PVDF 15 0.78 −2.50 −26.61 0.97 −32.62 −40.00 0.21 
Polysulfone 25 1.3 −42.40 −26.61 0.76 −25.22 −24.05 0.91 
Ultrafilic 25 1.3 −41.50 −26.61 0.70 −41.97 −24.05 0.72 
PVDF 25 1.3 −2.50 −26.61 0.97 −28.66 −24.05 0.54 
PVDF 15 0.78 −2.50 −26.61 0.97 −40.90 −26.61 0.10 
Ultrafilic 35 1.3 −41.50 −26.61 0.98 −50.67 −26.61 0.84 
Polysulfone 4.5 25 1.3 −10.00 −26.61 0.99 −50.00 −26.61 0.35 
Ultrafilic 25 1.3 −41.50 −26.61 0.70 −30.00 −26.61 0.85 
 Operating conditions
Membrane usedQTMPCfZeta potential of untreated membrane surfaceZeta potential of untreated latex particlesαpmRemediation techniqueZeta potential of membrane surface after treatmentZeta potential of latex particles after treatmentαpm
PVDF 15 0.78 −2.50 −26.61 0.97 −32.62 −40.00 0.21 
Polysulfone 25 1.3 −42.40 −26.61 0.76 −25.22 −24.05 0.91 
Ultrafilic 25 1.3 −41.50 −26.61 0.70 −41.97 −24.05 0.72 
PVDF 25 1.3 −2.50 −26.61 0.97 −28.66 −24.05 0.54 
PVDF 15 0.78 −2.50 −26.61 0.97 −40.90 −26.61 0.10 
Ultrafilic 35 1.3 −41.50 −26.61 0.98 −50.67 −26.61 0.84 
Polysulfone 4.5 25 1.3 −10.00 −26.61 0.99 −50.00 −26.61 0.35 
Ultrafilic 25 1.3 −41.50 −26.61 0.70 −30.00 −26.61 0.85 

CONCLUSIONS

The aim of the present study was to remediate membrane fouling of latex effluent by altering the membrane surface charge or the ionic strength of the latex effluent either through pH change or by using anionic surfactants. The results obtained in this study indicate that an increase in the ionic strength of the latex effluent was achieved by raising its pH, while LAS was found to be an ineffective pretreatment for limiting the fouling propensity of latex effluent using hydrophilic membranes even at high concentrations and long treatment times. On the other hand, LAS addition to the latex effluent had an overall positive effect on the enhancement of the surface charge of the PVDF hydrophobic membrane from −2.50 to −28.66 mV through the ultrafiltration process. The optimum time was determined as 15 min of LAS addition at the concentration of 0.0001 g/L as a feed pretreatment.

It was also concluded that for fouling remediation purposes the LAS-treated PVDF membrane surface is more favorable than pH-changed feed pretreatment. As such, the total mass of fouling decreased by 44.00 and 29.60%, for a membrane surface treated with anionic surfactant at a concentration of 0.0001 g/L, and latex effluent feed at pH 11, respectively. Surface charge of the membranes, treated by either a pH change or soaking in LAS as anionic surfactant, indicated good stability, sufficient for the ultrafiltration process. Increasing the zeta potential negativity of the latex particles or enhancing the membrane surface hydrophilicity caused a significant increase in the cumulative permeate flux, a substantial decrease in the total mass of fouling, and a noticeable decrease in the specific power consumption.

ACKNOWLEDGEMENTS

The authors are grateful for the financial support from the Natural Science and Engineering Research Council of Canada (NSERC). The assistance and facilities provided by the Department of Chemical Engineering, Ryerson University, have made this research possible and are also highly appreciated.

REFERENCES

REFERENCES
Abdelrasoul
A.
Doan
H.
Lohi
A.
2013a
A mechanistic model for ultrafiltration membrane fouling by latex
.
J. Membr. Sci.
433
,
88
99
.
Abdelrasoul
A.
Doan
H.
Lohi
A.
2013b
Fouling in membrane filtration and remediation methods
. In:
Mass Transfer – Advances in Sustainable Energy and Environment Oriented Numerical Modeling
(
Nakajima
Hironori
, ed.).
Intech Open Access Publisher
,
Croatia
, pp.
195
218
.
Abdelrasoul
A.
Doan
H.
Lohi
A.
2013c
Impact of operating conditions on fouling probability and cake height in ultrafiltration of latex solution
.
J. Membr. Sep. Technol.
2
,
134
147
.
Bruijn
J. P. F.
Salazar
F. N.
Borquez
R.
2005
Membrane blocking in ultrafiltration, a new approach to fouling
.
Food Bioprod. Process.
83
,
211
219
.
Dey
B. K.
Hashim
M. A.
Hasan
S.
Gupta
B. S.
2004
Microfiltration of water-based paint effluents
.
Adv. Environ. Res.
8
,
455
466
.
Huanga
J.
Guoa
S.
Zenga
G.
Xionga
Y.
Zhanga
D.
Tanga
X.
Xiea
G.
2012
Prediction of fouling resistance and permeate flux in cross-flowmicellar-enhanced ultrafiltration (MEUF)
.
Colloid Surf. A.
401
,
81
89
.
Israelachvili
J.
1992
Intermolecular & Surface Forces
, 2nd edn.
Academic Press Inc.
,
San Diego, CA
,
USA
.
Morrow
D.
Piwoni
M.
1993
LAS aquatic toxicity and biodegradation
.
Waste Management and Research Centre Reports
,
Champaign, IL, USA, TN93-029
.
Nikkola
J.
Sievänen
J.
Raulio
M.
Wei
J.
Vuorinen
J.
Tang
C.
2014
Surface modification of thin film composite polyamide membrane using atomic layer deposition method
.
J. Membr. Sci.
450
,
174
180
.
Sui
Y.
Wang
Z.
Gao
C.
Wan
Q.
Zhu
L.
2012
An investigation on the antifouling ability of PVDF membranes by polyDOPA coating
.
Desalin. Water Treat.
50
,
22
33
.
Viadero
R. C.
Vaughan
R. L.
Reed
B. E.
1999
Study of series resistances in high-shear rotary ultrafiltration
.
J. Membr. Sci.
162
,
199
211
.