Micellar-enhanced ultrafiltration (MEUF) was applied to the separation of phenolic compounds p-nitrophenol (PNP), p-chlorophenol (PCP), p-cresol (PC) and phenol (P) from effluents using a hydrophilic polyethersulfone ultrafiltration membrane. Cationic cetylpyridinium chloride (CPC), nonionic TX-100 and anionic sodium dodecyl benzene sulfonate (SDBS) were chosen as the surfactants. Several important parameters, i.e. the separation efficiency, the distribution coefficient of phenolic compounds and the removal ratio of surfactants, were investigated. It was shown that the separation efficiency and the distribution coefficient of phenolic compounds ascended with the increasing surfactant concentration and could be arranged as the following order: PNP > PCP > PC > P. Moreover, in the case of phenolic compound separation, CPC achieved the highest treatment efficiency, and the separation efficiency of SDBS was a little lower than that of TX-100. The removal ratios of the same surfactant when treating different phenolic effluents were nearly similar. However, when treating the same phenolic compound, the sequence of the surfactant rejection was in the following order: TX-100 > CPC > SDBS. These results indicate that CPC has a distinct superiority in the treatment of phenolic effluents via the MEUF process, and PNP easily solubilizes in the surface of the micelles.
Phenolic compounds are known as important environmental pollutants owing to their harmful influences on organisms at very low concentration and persistence in the environment (Zidi et al. 2010). Moreover, some of these have been considered as the environmental priority pollutants by the EPA and EU (Zidi et al. 2010). The primary sources of these compounds in the ecological environment are the effluents, which are discharged from coking plants, petroleum refining, gashouses and the manufacturing process of insulating materials, drugs, paper, and so forth (Praveen & Loh 2013). Many approaches have been described in the literature for the separation of these compounds from effluents, for example, photocatalytic degradation (Teh & Mohamed 2011), precipitation (Nicell et al. 1995), biocatalytic oxidation (Pérez-Prior et al. 2012), and electrochemical oxidation (Sripriya et al. 2007). However, these approaches had their own inherent flaws like high cost, high-energy consumption, inconvenient operation or their failure to treat effluents containing micromolecular phenolic compounds effectively (Nicell et al. 1995; Sripriya et al. 2007; Teh & Mohamed 2011; Pérez-Prior et al. 2012). Meanwhile, all the treatment methods listed above were aimed at the destruction of the phenolic compounds present in effluents despite the phenolic natural antioxidant and medicinal properties. Therefore, developing a low-energy consumption and high-efficiency separation technology for phenolic effluent treatment and the concentration and recycle of phenolic compounds is of a great interest.
During the past 50 years, membrane separation technology (MST) has been widely applied (Praveen & Loh 2013). According to the membrane pore size, the MST, which is generally based on the use of polymer membranes and ceramic membrane, can be classified as reverse osmosis (RO), nanofiltration (NF), ultrafiltration (UF) and microfiltration (MF). RO, NF and UF are regarded as the best available methods for the separation of some of the organic micromolecular contaminants in the aquatic environment (Zhang et al. 2012). However, compared with UF, the permeate fluxes of RO (Baudequin et al. 2014) and NF (Dražević et al. 2014) are very low, and the pressures needed are comparatively high. Therefore, the advent of micellar-enhanced ultrafiltration (MEUF), a separation technology based on surfactants, has had a significant influence on the separation of micromolecular contaminants (Mehling et al. 2012). When the quantity of surfactant added into the aqueous solution exceeds or equals its critical micellar concentration (CMC), surfactant molecules in the aqueous solution will aggregate to form large amphiphilic and transparent micelles (Zeng et al. 2008). Zhang et al. (2012) performed some experiments with phenol and showed that the retention of phenol ascends with the increase of feed surfactant concentration, but diminishes with the increase of feed phenol concentration. As the concentration of phenol increases, the amount of phenol in the micelles increases until the solubility of phenol is saturated, thus leading to the decrease of phenol retention. As a result, some of the phenol molecules are adsorbed in or on the micelles because of solubilization or electrostatic interaction (Zhang et al. 2012). Then, the micelles containing solubilized solutes are intercepted by the UF membrane with a proper molecular weight cut-off pore size.
In recent years, the separation of phenolic compounds with MEUF has been studied by numerous researchers (Lee et al. 2005; El-Abbassi et al. 2011; Huang et al. 2012a, b). The treatment of olive mill wastewater containing polyphenols has shown a high-efficiency via MEUF (El-Abbassi et al. 2011), and the MEUF was applied in the separation of organic and inorganic contaminants using a mixed surfactant (Lee et al. 2005). So far, wastewater containing phenol has been treated by MEUF with various surfactants (Huang et al. 2012a, b). However, the mechanism of phenolic compound rejection via surfactant was ignored.
The aim of this research is to elucidate the governing mechanisms between surfactants and phenolic compounds when treating phenolic effluents and concentrating phenolic compounds via MEUF. The application of MEUF in the phenolic effluent treatment is promising due to its high separation efficiency and low cost (Zeng et al. 2008). In this study, phenol, p-nitrophenol (PNP), p-chlorophenol (PCP) and p-cresol (PC) were chosen as organic pollutants in the effluents. Three representative surfactants, positively charged cationic cetylpyridinium chloride (CPC), uncharged Triton X-100 (TX-100) and negatively charged sodium dodecyl benzene sulfonate (SDBS), were used in MEUF experiments. Several parameters, including the concentrations of surfactants and phenolic compounds in the permeate and retentate, were measured to illustrate the different mechanisms between micelles and phenolic compounds.
CHEMICALS AND METHODS
|Name .||Chemical formula .||Molecular weight .||Type .|
|Name .||Chemical formula .||Molecular weight .||Type .|
|Name .||MWa (g/mol) .||Type .||CMC (mM) .||Aggregation numberb .||MW of micellec (g/mol) .|
|Name .||MWa (g/mol) .||Type .||CMC (mM) .||Aggregation numberb .||MW of micellec (g/mol) .|
aMean surfactant molecular mass.
bMean aggregation number of surfactant micelle.
cMean molecular mass of surfactant micelle.
UF experiments were performed in a cross-flow membrane module modified from a Millipore Labscale™ TFF System. The pore size of the polyethersulfone (PES) membrane (Pellicon XL Device, Millipore Corporation, USA) is 10,000 Da and the membrane effective filtration area is 50 cm2. The pH and temperature limit of the membrane is 1–14 and 4–45 °C, respectively. In addition, the maximum operating pressure of the module is 5.6 bar.
The calculated quantities of phenolic compounds and surfactants were weighed and dissolved in 1,000 ml ultra-pure water as the feed solution. The feed solution was stirred adequately and settled for 30 min to ensure that the micelles were formed and the solutes were solubilized in them. All the UF experiments were conducted at room temperature of 25 °C and a stabilized operating pressure of 1.5 bar. The permeate was retained in the permeate tank and the retentate was returned back to the feed tank. The schematic diagram of the experimental apparatus is shown in Figure 2. When the volume of the solution in the feed tank was 200 ml, the experiment was ended. During the MEUF experiments, the phenolic compound concentrations added into the feed solution were kept constant at 1 mM, while the surfactant concentrations varied from 0.5 to 4 mM. After each experiment, the concentrations of the permeate and retentate were measured.
After each run, the UF membrane was washed with the pure water without pressure for 1 h, and then the ultra-pure water was used to rinse out most of the phenolic compounds and surfactants remaining within the membrane at a low pressure for 30 min. Finally and the most importantly, the ultra-pure water was filtered through the membrane to ensure that the permeate flux recovered up to 95% of the initial water flux and filled in the membrane cell to protect the membrane.
In the presence of CPC, the concentrations of CPC and phenolic compounds were determined using a high-performance liquid chromatograph provided with a 25 × 0.46 cm Spherisorb ODS2 column (particle size 5 μm) and measured spectrophotometrically at 220 nm. The operating conditions were as follows: mobile phase acetonitrile/water (50/50 by volume), injection volume 20 μl, flow rate 1 ml/min, temperature 30 °C.
In the presence of TX-100, phenolic compound concentrations in the permeate and retentate and TX-100 concentration in the permeate were determined by the extinction coefficients method using a UV spectrophotometer (Shimadzu UV-2550) at wavelengths of 270 nm and 274 nm.
In the presence of SDBS, phenolic compound concentrations in the permeate and retentate were measured at a wavelength of 270 nm. The concentration of SDBS was measured by colorimetric method with methylene blue (ISO-7875-1-1996, as described in Fang et al. (2008)).
RESULTS AND DISCUSSION
The separation efficiency and distribution coefficient of phenolic compounds
Variations of the separation efficiency and distribution coefficient of phenolic compounds with feed surfactant concentration
The variation of the separation efficiency of phenolic compounds with feed surfactant concentration is shown in Figure 3(a)–3(c) for CPC, TX-100 and SDBS, respectively, and the variation of the distribution coefficient of phenolic compounds with feed surfactant concentration is shown in Figure 4(a)–4(c) for CPC, TX-100 and SDBS, respectively.
As shown in Figures 3 and 4, the separation efficiency and distribution coefficient of phenolic compounds ascend with the increasing feed surfactant concentration. This clearly indicates that the amount of micelles increases with the feed surfactant concentration, leading to increased phenolic compounds solubilizing in the micelles (Purkait et al. 2005a, b), which are subsequently retained by the UF membrane. Considering surfactant separation, it may be observed from these figures that the rejection of different pollutants is different. The separation efficiency and distribution coefficient of phenolic compounds are different (higher for PNP and lower for P). The pollutants used here are hydrophilic (P and PNP) and hydrophobic (PC and PCP) in nature (Purkait et al. 2005a, b). So the hydrophilic P and PNP readily get solubilized on the micelle-water surface (Bhat et al. 1987; Xu et al. 2010). The sequence of the solubilization is PNP < P, which is the same as the sequence of the polarity of P and PNP. The reason is that highly polar solutes tend to solubilize in the vicinity of the micellar surface and the polar groups interact strongly with the ionic and/or polar groups of the surfactant molecules (Bhat et al. 1987). In contrast, PC and PCP possess a hydrophobic nature. Hence, they can be easily solubilized in the core of micelles (Xu et al. 2010). In previous studies about the solubilization of phenolic compounds in surfactant micelles (Xu et al. 2010; Bhat et al. 1987), the authors found out that when the concentration of phenolic compounds was very low (less than or equal to 1 mM), the influence of phenolic compounds on the micellar form can be neglected, and the location of phenolic compound solubilization on the micelles did not change. Hence, phenolic compounds have little influence on the MEUF performance in this concentration.
As we know, the micelles can be hardly formed when the surfactant concentration is below its CMC value; however, as presented in Figures 3(a)–3(c) and 4(a)–4(c), the phenolic compounds obtain rejections as well in such conditions, for the following three causes. Firstly, during the MEUF experiment, the surfactant molecules gradually accumulate on the UF membrane surface, leading to an accumulation layer on it, in which the surfactant concentration may surpass its CMC value and form micelles to bind some phenolic compound molecules (Zeng et al. 2008; Li et al. 2011). Secondly, the phenolic compound molecules can absorb on the UF membrane surface and in the pores of the UF membrane during the MEUF experiment (Purkait et al. 2005a, b). Finally, when the concentration of the feed surfactant is close to and lower than its CMC value, there are premicelles formed in these surfactant aqueous solutions, and they also can solubilize some phenolic compound molecules (Lu et al. 2009). Consequently, they describe a rise in the separation efficiency and distribution coefficient of phenolic compounds with the increasing feed SDBS concentration ranging from 0.5 to 2.0 mM in Figures 3(c) and 4(c).
Comparisons of the separation efficiency and distribution coefficient of different phenolic compounds with different surfactants
Comparisons of the separation efficiency and distribution coefficient of the same phenolic compounds using different surfactants are presented in Figures 5 and 6, respectively. The feed concentrations of surfactants and phenolic compounds were fixed at 4 mM and 1 mM, respectively. It may be observed from these figures that the separation efficiency and distribution coefficient of the four chosen phenolic compounds are the best with CPC, and the minimum is observed with TX-100. This clearly shows that the cationic surfactant has better efficiency on the separation of phenolic compounds during the MEUF experiment. This is because the phenolic compounds can be ionized into an anion phenoxide group and H+ ions in the feed solution, and the CPC has a cationic hydrophilic pyridinium head group (Purkait et al. 2005a, b), which leads to a better efficiency on the phenolic compound rejection for CPC due to the similarity–intermiscibility and electrostatic interaction (Bhat et al. 1987; Purkait et al. 2005a, b; Zhang et al. 2012). Because of the electrostatic attraction, separation efficiency and distribution coefficient of the phenolic compounds obtained by cationic surfactant CPC are better than those obtained by anionic surfactant SDBS (Huang et al. 2012a, b; Zhang et al. 2012), which is similar to methylene blue solubilized in the SDBS micelles (Zaghbani et al. 2007). In contrast, it is clear that the ionic surfactant has a much better effect on the solubilization of phenolic compounds than has the nonionic surfactant, for the nonionic surfactant cannot ionize to solubilize phenolic compound molecules, and the interaction force that exists between them is only a hydrophobic interaction, which is extremely weak (Zaghbani et al. 2007). Figures 5 and 6 indicate that the electrostatic interaction is more powerful than the hydrophobic interaction, whereupon the nonionic surfactant has less influence on the solubilization of phenolic compounds than ionic surfactant.
MEUF of CPC, TX-100 and SDBS
Variations of the removal ratio of surfactants with feed surfactant concentration
The variations of surfactant rejection with feed surfactant concentration (0.5, 1, 2, 3 and 4 mM) are presented in Figure 7. Figure 7 depicts a rapid increase in the removal ratio of three surfactants during MEUF of phenolic compounds at low surfactant concentration (≤2 mM), and then the growth trend becomes gradually slow when the surfactant concentration is higher than 2 mM. With the further increasing of feed surfactant concentrations, more and more micelles in the aqueous solution are formed, leading to the retaining of more surfactant molecules by the UF membrane. Meanwhile, with the increasing of the feed surfactant concentration, the amount of micelles which accumulate on the surface of the UF membrane has gradually increased, and then the concentration polarization occurs. This results in an increase in the convective transport of the solutes to the permeate side, thereby increasing the permeate concentration and subsequently decreasing the observed retention with time. But the decrease in observed retention is marginal in the presence of surfactant, as the free solute over the membrane surface is much less compared to the without-surfactant case. (Purkait et al. 2005a, b). As a consequence, the permeate concentration of surfactant increases slightly, leading to decreased growth rate of the surfactant rejection. When treating different phenolic compounds with the same surfactant, the surfactant rejection is almost the same. The reason is that the relationship between the phenolic compounds in aqueous solution and the surfactant rejection can be ignored (Zhang et al. 2012).
Theoretically, there are no micelles formed when the surfactant concentration is below its CMC value. Hence, there is no surfactant rejection. However, the surfactant removal ratio is extremely high, which reached 86–94%, as can be seem from Figure 7(a) and 7(c). Owing to the concentration polarization effect, a gel layer was formed on the membrane surface (Huang et al. 2012a, b). In the gel layer, the surfactant concentration may exceed its CMC value, and this leads to the formation of the micelle (Huang et al. 2012a, b). Then, the micelles are retained by the UF membrane.
Comparisons of the removal ratio of different surfactants for the different phenolic compounds
Comparisons of the removal ratio of different surfactants at the same phenolic are described in Figure 8. As shown in Figure 8, the removal ratio of the three chosen surfactants can be arranged as the following sequence: TX-100 > CPC > SDBS. For instance, considering PNP alone, the rejection of surfactant is 96.89%, 95.05% and 94.51% for TX-100, CPC and SDBS, respectively. The great TX-100 rejection is probably on account of its small CMC value and large micelle size, which are favorable to form micelles and beneficial to be intercepted by the UF membrane (Huang et al. 2012a, b; Zhang et al. 2012).
MEUF has been regarded as a significant technology for the separation of phenolic compounds from effluents. In this study, it is found that the separation efficiency and distribution coefficient of phenolic compounds are all in the order of PNP > PCP > PC > P, which indicates that PNP possesses greater polarity and easily solubilizes in the vicinity of the surfactant micellar surface. Meanwhile, in terms of each phenolic compound, the separation efficiency and distribution coefficient of phenolic compounds can be arranged in the following surfactant order: CPC > SDBS > TX-100. The study clearly shows that the CPC surfactant has higher solubilizing capabilities for phenolic compounds and can achieve better efficiency in the separation of phenolic compounds during MEUF. The reason is that the CPC head group is similar to that of phenolic compounds, but its charge is opposite. Moreover, the removal ratios of the same surfactant are almost identical when treating different phenolic compounds. Also, considering phenolic compound separation, the removal ratios of surfactants are in the order of TX-100 > CPC > SDBS, which is decided by the CMC value and size of the micelles.
From the experimental results, the bind-force between CPC and phenolic compound is the highest than that of TX-100 and SDBS. So CPC is considered the most efficient surfactant for the separation of phenolic compounds from effluents via MEUF on PES membrane, and the phenolic compound with greater polarity can be easily bound with the micelles.
This study was ﬁnancially supported by the National Natural Science Foundation of China (51178172, 51039001, 51308076, 51378190), the Project of Chinese Ministry of Education (113049A) and the Research Fund for the Program for Changjiang Scholars and Innovative Research Team in University (IRT-13R17).