Study on the removal of organic micropollutants from aqueous and ethanol solutions by HAP membranes with tunable hydrophilicity and hydrophobicity
a b s t r a c t
A biocompatible and uniquely defined hydroxyapatite (HAP) adsorption membrane with a sandwich structure was developed for the removal of organic micropollutants for the first time. Both the adsorption and membrane technique were used for the removal of organic micropollutants. The hy- drophilicity and hydrophobicity of the HAP adsorbent and membrane were tunable by controlling the surface structure of HAP. The adsorption of organic micropollutants on the HAP adsorbent was studied in batch experiments. The adsorption process was fit with the Freundlich model, while the adsorption kinetics followed the pseudo-second-order model. The HAP membrane could remove organic micro- pollutants effectively by dynamic adsorption in both aqueous and ethanol solutions. The removal effi- ciencies of organic micropollutants depended on the solution composition, membrane thickness and hydrophilicity, flow rate, and the initial concentration of organic micropollutants. The adsorption ca- pacities of the HAP membrane with a sandwich structure (membrane thickness was 0.3 mm) were 6700, 6510, 6310, 5960, 5490, 5230, 4980 and 4360 L m—2 for 1-naphthyl amine, 2-naphthol, bisphenol S, propranolol hydrochloride, metolachlor, ethinyl oestradiol, 2,4-dichlorophenol and bisphenol A, respectively, when the initial concentration was 3.0 mg L—1. The biocompatible HAP adsorption mem- brane can be easily regenerated by methanol and was thus demonstrated to be a novel concept for the removal of organic micropollutants from both aqueous and organic solutions.
1.Introduction
The pervasive pollution of organic micropollutants, such as phenol, aniline and nitrobenzene, in water resources has led to serious effects on aquatic ecosystems and human health (Barbosa et al., 2016; Zgheib et al., 2012). These aromatic organic micro- pollutants have strong adverse effects on organisms and will react with free chlorine to produce chlorinated by-products, which are more stable and poisonous (Chan et al., 2003). So far, numerous studies have investigated the removal and eradication of organic micropollutants from aqueous solutions (Allabashi et al., 2007; Cui et al., 2016; Zhou et al., 2015). There are three kinds of primary methods for the removal of organic micropollutants from waste- water: physical, such as adsorption (Chen et al., 2015; Yu et al., 2016; Zou et al., 2016b), membrane separation (Martinez et al., 2013; Pastrana-Martinez et al., 2015), steam stripping (Braass et al., 2003), ion exchange (Bauerlein et al., 2012), and solvent extraction (Lopez-Montilla et al., 2005); chemical, such as oxida- tion (Zhang et al., 2014) and phase transfer catalysis (Wu et al., 2016); and biological (Gasperi et al., 2010; Rosal et al., 2010) treatments. Despite the availability of the above-mentioned pro- cesses for the removal of organic micropollutants, adsorption and membrane separation still remain two of the best methods (He et al., 2015).Commonly, in the removal of organic micropollutants, the twomethods of adsorption and membrane separation are used sepa- rately. The adsorption process is essentially the immobilization of target molecules through biological recognition, electrostatic in- teractions, chemical reactions and hydrophobic interactions (He et al., 2016, 2017). This method can remove soluble and insoluble organic micropollutants effectively without generation of hazard- ous by-products (Kanno et al., 2014). However, the adsorbents are sometimes difficult to separate from water in practical applications (Budyanto et al., 2015; Nie et al., 2012). Membrane separation is defined as the rejection of target molecules by size exclusion (Jin et al., 2012).
In traditional membrane separation processes, size exclusion is difficult to achieve, but the membrane can be easilytemplate. Second, NaOH (10.0 mL, 0.50 g), CaCl2 (10.0 mL, 0.11 g) and NaH2PO4 (10.0 mL, 0.12 g) solutions were added dropwise intothe mixture of ethanol and oleic acid. Then, the resulting system was heated at 180 ◦C for 24 h and then cooled to room temperature.Finally, the HAP adsorbent was collected by centrifugation and washed with ethanol and deionized water before being dried at60 ◦C.The HAP adsorption membrane was fabricated through a simple process of suction filtration. The powdered HAP nanowires were dispersed in ethanol and then poured onto filter paper in a Buchner funnel. With the help of a vacuum pump, the white HAP membrane was fabricated on filter paper and separated with a tweezer.To evaluate the crystallinity and functional groups on the adsorbent, the HAP adsorbent was characterized by X-ray diffrac- tion (X’Pert ProMPD, Cu-Ka radiation wavelength of 1.5418 Å) and Fourier transform infrared (FT-IR) spectroscopy (Nexus-870 spec- trophotometer). The textural properties of the adsorbent (specific area, pore volume and average pore diameter) were tested on a Coulter Omnisorp 100 CX Brunauer-Emmett-Teller (BET) analyserusing nitrogen adsorption with a degassing temperature of 80 ◦C.Field-emission scanning electron microscopy (FE-SEM) images were taken by a Sirion 200 field-emission scanning electron mi- croscope. Transmission electron microscopy (TEM) images were obtained from a JEOL JEM-2010 instrument operated at 100 kV. The contact angle of the HAP membrane was measured by an OCA20 system equipped with a CCD camera and the SCA 20 software (Dataphysics GmbH, Germany).
The zeta potential was determined on a Delsa Nano C/Z. Ultraviolet-visible (UVevis) spectra were obtained on a Cary 5000 Varian UVevis spectrometer and recorded in a range of 200e800 nm, corrected against the background spectrum, and normalized to zero absorbance at 800 nm.The thickness of the HAP membrane was calculated from the following equation:separated from water (Ajji and Ali, 2010; Cassano et al., 2013). Thus, the combination of adsorption and membrane separation would remove organic micropollutants more effectively without second-d = mrs(1)ary pollution (He et al., 2015). It is highly important to find mate- rials that can be used as both adsorbents and membranes.In this work, a biocompatible and uniquely defined adsorption membrane for the rapid removal of organic micropollutants from aqueous and ethanol solutions was prepared. Hydroxyapatite (HAP), Ca5(PO4)3$(OH), was chosen as the membrane material for its high biocompatibility and adsorption capacity (Hammari et al., 2004; Jimenez-Reyes and Solache-Rios, 2010; Prabhu and Mee- nakshi, 2014). Because of the special structure of ultralong HAP nanowires, HAP nanowires were fabricated into an adsorption membrane for organic micropollutants removal. The adsorption and desorption behaviour of organic micropollutants on the HAP adsorbent and membrane were investigated. The effects of solution composition, membrane thickness and hydrophilicity, flow rate, and the initial concentration of organic micropollutants were studied. The adsorption capacities of HAP membranes for organic micropollutants were also investigated.
2.Experimental
In our research, both a HAP adsorbent and a HAP adsorption membrane were synthesized. First, 6.0 mL oleic acid was dissolved in 10.0 mL ethanol under magnetic stirring, which acted as a softwhere d is the thickness of the HAP membrane, m is the weight ofthe HAP membrane, r is the density of the HAP membrane, and s is the area of the HAP membrane.The removal of organic micropollutants from aqueous solution was investigated in batch adsorption experiments, and the adsorption kinetics and isotherms are depicted in S2 (see Supplementary Material).The organic micropollutants used in the adsorption experiments included the following: 1-naphthyl amine (1-NA), an azo dye pre- cursor and known carcinogen (Refat et al., 2014); 2-naphthol (2- NO), a model for various naphthol pollutants (Zhou et al., 2016); 2,4-dichlorophenol (2,4-DCP), an intermediate in herbicide pro- duction and degradation product of the antibacterial agent triclo- san (Latch et al., 2005); bisphenol A (BPA) and bisphenol S (BPS), endocrine disruptors with greater environmental persistence that exist in many polycarbonates (Ike et al., 2006); metolachlor, one of the most common herbicides present in groundwater (Benner et al., 2013); ethinyl oestradiol, an oestrogen mimic used in oral contra- ceptives that has caused the collapse of fish populations (Kidd et al., 2007); and propranolol hydrochloride, a beta-blocker used to treat hypertension, which has been detected in effluent streams at concentrations similar to blood serum levels of its users (Kostichet al., 2014). The structures and relevance of each tested emerging organic micropollutant are shown in Fig. 1, Tables S1 and S2 (see Supplementary Material).
The concentrations of the organic micropollutants in solution at high levels were measured by UVevis spectroscopy and based on calibration, were determinedfor BPA (lmax = 276 nm), BPS (lmax = 259 nm), 2-NO (lmax = 273 nm), 1-NA (lmax = 305 nm), 2,4-DCP, (lmax = 284 nm), metolachlor (lmax = 265 nm), ethinyl oestradiol (lmax = 305 nm) and propranolol hydrochloride (lmax = 290 nm).The UV calibrations of the eight organic micropollutants aredepicted in Fig. S1 (see Supplementary Material). The concentra- tions of organic micropollutants at low levels were analysed using high-performance liquid chromatography-tandem mass spec- trometry (HPLC-MS, Agilent Technologies 1200 Series for HPLC and 6410 Triple Quad. for MS).In the membrane experiments, the removal efficiencies of organic micropollutants from both aqueous and ethanol solution by the HAP membrane were investigated by a filtration system we assembled as shown in Fig. S2 (see Supplementary Material). The HAP membranes with different hydrophilicities were placed in front of the strainer (diameter: 4.2 cm), and the strainer was then fixed at the bottom of the column. With the help of a diaphragm boost pump and fluid flowmeter, water or ethanol permeated through the HAP membranes at a controlled flow rate, while the organic micropollutants were adsorbed. The effects of flow rate, solution composition, thickness and hydrophilicity of the HAP membrane, and the initial concentration of organic micropollutants on the removal efficiencies were investigated. The amount of water filtered and the adsorption capacity of the HAP membrane were also evaluated.To evaluate the economic applicability of the HAP adsorbent and membrane, regeneration should be the most important aspect to be investigated. To estimate the desorption and regeneration of theHAP adsorbent and HAP membrane for organic micropollutants, the HAP adsorbent and HAP membrane were regenerated by soaking in 100 mL methanol for 2 h and were then washed by water. Six adsorption/desorption cycles were conducted. After every cycle, the residual concentration of organic micropollutants was measured.
3.Results and discussion
SEM images of the HAP adsorbent are shown in Fig. 2a, and ultralong HAP nanowires 40e60 mm in length and 40e100 nm in diameter were obtained. The TEM images are depicted in Fig. 2b, showing that the nanowires were not singular but instead were composed of numerous thin wires 10 nm in diameter. The single- crystalline structure of the HAP nanowires was demonstrated by the selected-area electron diffraction (SAED) pattern, which is shown in the inset of Fig. 2b.Optical and SEM images of the HAP membrane are depicted in Fig. 2c and d. A smooth and dense membrane surface was obtained. To evaluate the hydrophilic/hydrophobic features of the HAP membrane, the static contact angles of the HAP membrane are shown in the inset of Fig. 2c. Both hydrophilic and hydrophobicmembranes were obtained with contact angles of 0◦, 31.4◦ and146.4◦, which are numbered 1#, 2# and 3#, respectively. The transformation between hydrophilic and hydrophobic could be due to the long hydrophobic hydrocarbon chains of oleic acid molecules or oleate groups adsorbed on the surface of the HAP nanowires. This can be accomplished by adjusting the amount of NaH2PO4 and heating time. The synthetic conditions to obtain HAP membraneswith contact angles of 0◦, 31.4◦ and 146.4◦ are shown in Table S3(see Supplementary Material). The tunable hydrophilicity and hy- drophobicity enables HAP to be an excellent adsorbent in both aqueous and ethanol solutions. To further utilize this property, two new-style HAP membranes with a sandwich structure, numbered 4# and 5#, were prepared. The 4# membrane had three layers; the upper and lower layers were hydrophilic, while the middle layer was hydrophobic. The 5# membrane had the opposite structure;the upper and lower layers were hydrophobic, while the middle layer was hydrophilic.
The thicknesses of the three layers were controlled to be consistent. A cross-sectional SEM image of the 4# membrane is shown in Fig. S3 (see Supplementary Material), from which we can see that the membrane had three layers with well- marked interfaces. The thickness of each layer was 100 mm. The membrane had a good density, which ensured the stability of the membrane. These results further demonstrate the sandwich structure of the membrane.XRD patterns of the HAP adsorbent and membrane are shown in Fig. 3a. Both the adsorbent and membrane had a single phase (Lu et al., 2014). The diffraction peaks were successfully indexed to hydroxyapatite, Ca5(PO4)3OH (JCPDS 09-0432) (Akram et al., 2015; Jimenez-Reyes and Solache-Rios, 2010; Long et al., 2008). The N2 adsorption-desorption isotherm and BJH adsorption pore size dis- tribution of the HAP adsorbent and membrane are depicted inFig. 3b. The BET surface area of the HAP adsorbent and membrane were 77.05 and 70.29 m2 g—1, which are higher than most HAP materials (6.00 m2 g—1 (Utara and Klinkaewnarong, 2015),10.87 m2 g—1 (Sahu and Mohanty, 2015), 23.17 m2 g—1 (Alshemary et al., 2015), 36.00 m2 g—1 (Mohammad et al., 2015), 47.23 m2 g—1(Shi et al., 2015), 48.70 m2 g—1 (Sudhakar et al., 2016), 69.68 m2 g—1(Akram et al., 2015)). A wide pore size distribution of less than 5 nm was observed for both the HAP adsorbent and membrane, which is considered to be the gaps between the nanowires, as shown in the TEM results. The pore volume of the HAP membrane was decreased compared with the HAP adsorbent of the same pore diameter, which resulted in a decrease in the BET surface area.The FT-IR spectrum of the HAP adsorbent is depicted in Fig. 3c. The peaks at 1095 and 1030 cm—1 were attributed to asymmetric stretching vibrations of the phosphate groups, while the peaks at632, 605 and 558 cm—1 correspond to the bending mode of the phosphate groups (Hammari et al., 2004; Wang et al., 2011). The weak band at approximately 3560 cm—1, which was overlapped with the broad band of adsorbed water at approximately 3400 cm—1, was attributed to the hydroxyl groups (Mohammad et al., 2015).
The UVevis spectrum of the HAP adsorbent is shown in Fig. 3d, and no response in the ultraviolet region was observed, which means that the HAP adsorbent will not affect the detection of organic micropollutants. All these results indicate the formation of the ultralong HAP adsorbent and membrane with large surface area and its potential in the removal of organic micropollutants.The adsorption of organic micropollutants by HAP was investi- gated over a range of adsorbent doses. Various doses of hydrophilic HAP adsorbent in the range of 0e0.2 g L—1 were investigated for theremoval of organic micropollutants. Flasks containing the organicmicropollutants and HAP adsorbent were shaken at 150 rpm at 25 ◦C for 12 h. After that, the residual concentration of organicmicropollutants in the supernatant liquid was measured. The removal efficiencies of organic micropollutants with various adsorbent doses are shown in Fig. S4 (see Supplementary Material), and it was observed that the removal efficiencies of the eight organic micropollutants increased with adsorbent dose, with theequilibrium dose at 0.1 g L—1. Taking economic applicability into account, the adsorbent dose was selected to be 0.05 g L—1 in thefollowing experiments, as most of the adsorption was complete at this dose.Fig. S5 (see Supplementary Material) depicts the adsorption kinetics of the eight organic micropollutants on the HAP adsorbent at a dose of 0.05 g L—1 at pH 7.0. The adsorption of organic micro- pollutants was a fast process, as it reached equilibrium in 30 min. To further investigate the adsorption of organic micropollutants overtime, the adsorption data were simulated by the pseudo-second- order kinetics model (Lin et al., 2015). Linear plots of t/qt versus time t in the pseudo-second-order kinetics model of various organic micropollutants were achieved and are shown in the inset of Fig. S5 (see Supplementary Material).
Parameters in the pseudo- second-order models for the different organic micropollutants are listed in Table S4 (see Supplementary Material). Based on the values of the correlation coefficient (R2), the adsorption kinetics of the eight organic micropollutants can be well described by the pseudo- second-order kinetics model (Jia et al., 2015; Zhu et al., 2015), which demonstrates that the rate depended on the adsorption capacity but not the concentration of the adsorbate (Ho and McKay, 2000; Ho and Ofomaja, 2006).Adsorption isotherms were measured to study the adsorption performance and mechanism of the HAP adsorbent. Two adsorp- tion models, the Langmuir model and the Freundlich model, were used to simulate the experimental data (Zou et al., 2016a, 2016c). Relative parameters in the Langmuir and Freundlich models were calculated and are listed in Table 1. The Freundlich isotherm model,used for fitting the adsorption of the eight organic micropollutants, is shown in Fig. 4aeb. The Freundlich model gave a better fit to the experimental data than the Langmuir model, indicating that the adsorption process took place on a heterogeneous surface that varies in surface coverage (Jimenez-Reyes and Solache-Rios, 2010; Medellin-Castillo et al., 2014). This was caused by the assembly of the HAP nanowires into one nanowire, as the TEM results showed. The maximum of the adsorption capacity (qm) for BPA, BPS, 2-NO, 1- NA, 2,4-DCP, metolachlor, ethinyl oestradiol and propranolol were 104.20, 143.07, 148.54, 160.44, 116.15, 132.09, 127.27 and139.29 mg g—1, respectively, when the initial concentration was 10 mg L—1. To further investigate the differences in the adsorptioncapacities for these eight organic micropollutants, the adsorption capacity and zeta potential of each organic micropollutant were investigated, and the results are depicted in Fig. 4c. The zeta po- tential of the HAP adsorbent at pH = 7 was 3.4 mV, which meant that the HAP adsorbent had a tendency to adsorb negativelycharged organic micropollutants. It followed that the adsorption capacities of the eight organic micropollutants decreased with their zeta potentials.
Thus, we can conclude that the variation in adsorption capacity among BPA, BPS, 2-NO, 1-NA, 2,4-DCP, meto- lachlor, ethinyl oestradiol and propranolol were caused by elec- trostatic interactions between HAP and the organic micropollutants.To evaluate the adsorption capacity, the qm values of the organic micropollutants on various materials are listed in Table 2 for comparison. Obviously, the adsorption capacities of the HAPadsorbent in this work were much higher than those of other materials (Ahmed et al., 2014; Dehghani et al., 2016; Gundogdu et al., 2012; Imyim and Prapalimrungsi, 2010; Jin et al., 2015; Otero et al., 2014; Rovani et al., 2014; Song et al., 2011; Zhou et al., 2016).To investigate the effect of membrane hydrophilicity andsolution composition on the removal efficiencies of organic micropollutants, working solutions of organic micropollutants were prepared with deionized water and ethanol. The HAP mem- branes numbered 1#, 2#, 3#, 4# and 5# were utilized. The removal efficiencies of BPS by the HAP membranes with different hydro- philicities in aqueous solutions containing different ratios of ethanol are shown in Fig. 4d. The ratio of ethanol in solution rangedfrom 0 to 1. In aqueous solution (ratio of ethanol = 0), the removalefficiencies increased with membrane hydrophilicity; however, the 4# membrane presented the best removal efficiency. This could beexplained by the outer layer of the 4# membrane being highly hydrophilic, and the hydrophilic groups on the surface made it easy for water to pass through. Meanwhile, the middle layer was highly hydrophobic, and organic micropollutants were strongly adsorbed and immobilized on this layer as water penetrated the membrane. To further demonstrate this phenomenon, the adsorption kinetics and isotherms of the hydrophobic HAP adsorbent in aqueous so- lution was investigated, and the results are shown in Fig. S6 (see Supplementary Material).
Due to the hydrophobicity of HAP, the adsorption capacity and adsorption rate both decreased compared with the hydrophilic HAP adsorbent. In ethanol solution (ratio ofethanol = 1), the removal efficiencies decreased with hydrophi-licity, and the best removal efficiency was found for the 3# mem- brane. This is because the 3# membrane was highly hydrophobic, which made it easy to interact with ethanol and the organic micropollutants. In solution containing both water and ethanol, more ethanol led to lower removal efficiencies by the hydrophilic membrane, and a higher removal efficiency was obtained for the hydrophobic membrane. Moreover, the sandwich structure of the 4# and 5# membranes showed better removal efficiencies than the single-layer structure membranes. In our following membrane experiment, the 4# membrane, with a sandwich structure, was chosen for the removal study.The removal efficiencies by HAP membranes with different thicknesses were studied and is shown in Fig. 5a, where the removal efficiencies of BPA and BPS increased with membrane thickness. When the thickness of the HAP membrane increasedfrom 50 to 350 mm, at a flow rate of 30 mL min—1 and an initial concentration of 3.0 mg L—1, the removal efficiencies of BPA and BPSincreased from 69.54% to 95.25% and 71.54%e97.25%. This was because the increase in membrane thickness increased the avail- able interaction sites for the adsorption of organic micropollutants.The removal efficiencies by the HAP membranes at different flow rates was studied. Filtration experiments were conducted at flow rates of 10e80 mL min—1, where the thickness of the HAPmembrane was 300 mm and the initial concentration was3.0 mg L—1. The results are expressed in Fig. 5b, when the flow rate increased from 10 to 80 mL min—1, the removal efficiencies of BPAand BPS decreased from 88.21% to 70.31% and 93.21%e74.31%. This can be explained by the increase in flow rate decreasing the contact time between organic micropollutants and the membrane surface.The removal efficiencies by the HAP membranes in different concentrations of organic micropollutants were studied. BPA and BPS in different initial concentrations were used in the filtration experiments at a flow rate of 30 mL min—1. The results are depictedin Fig. 5c.
When the initial concentration of BPA and BPS increased from 0.5 to 5.0 mg L—1, the removal efficiencies decreased from 90.23% to 72.81% and 91.36%e80.75%.The adsorption capacities of the HAP membrane were investi- gated and are depicted in Fig. 5d. The HAP membrane was inca- pable of removing organic micropollutants as the value of Ce/C0 increased to 1.0. When the initial concentration of organic micro-pollutants was 3.0 mg L—1, with a membrane thickness of 300 mm and a flow rate of 30 mL min—1, the maximum amount of water that the HAP membrane (with an area of 13.85 cm2) can address was9280, 9020, 8740, 8250, 7610, 7240, 6900 and 6040 mL for 1-NA, 2-NO, BPS, propranolol, metolachlor, ethinyl oestradiol, 2,4-DCP and BPA. Thus, the adsorption capacities of the HAP membrane were 6700, 6510,6310, 5960, 5490, 5230, 4980 and 4360 L m—2 for 1-NA,2-NO, BPS, propranolol, metolachlor, ethinyl oestradiol, 2,4-DCPand BPA when the concentration was 3.0 mg L—1. The membrane performances over time are depicted in Fig. S7 (see Supplementary Material). At an initial fluoride concentration of 3.0 mg L—1, the treatment time by the HAP membrane was 3095, 3008, 2908, 2749,2539, 2408, 2317 and 2000 min for 1-NA, 2-NO, BPS, propranolol, metolachlor, ethinyl oestradiol, 2,4-DCP and BPA. As we can see, the performance of the HAP membrane was maintained at a high level even after a day of operation.The removal efficiencies of BPA and BPS by the regenerated HAP adsorbent and membrane are depicted in Fig. S8 (see Supplementary Material). After six cycles with an initial concen- tration of 3.0 mg L—1, the removal efficiencies of the HAP adsorbent and membrane for BPA and BPS remained almost unchanged. Theconcentration of BPA and BPS in the methanol regeneration solu- tion is shown in Table S6 (see Supplementary Material). Obviously, the adsorbed BPA and BPS were suitable for desorption and regeneration by methanol. What was significant here was that we added 200 mg of HAP adsorbent into 200 mL of BPA and BPS so- lution in the first cycle, but only 191 mg of adsorbent was recycled after the first regeneration, while 184, 179, 166, and 157 mg of adsorbent was obtained over the next four cycles. This could be caused by incomplete collection during centrifugation.
Moreover, the concentration of Ca and P for the HAP adsorbent and membrane in the adsorption and regeneration solutions was measured and is given in Table S7 (see Supplementary Material), which demon- strates that both the HAP adsorbent and membrane were stable enough during application.The concentration of organic micropollutants used in the abovestudy (0.1e10 mg L—1) was higher than realistic environmental concentrations, which are at low levels on the order of mg L—1(Kostich et al., 2014). In addition, the substances in realistic water environments are far more complex than in the regular water that was tested. To investigate the practical implications of our research, the HAP membrane was used for the treatment of underground water in Xingwang Village, Inner Mongolia of China, in which organic micropollutants were added artificially at a concentrationof 50 mg L—1. The characteristics of the underground water samplesare shown in Table S5 (see Supplementary Material), which demonstrate the complexity of the substances. The removal effi- ciencies of organic micropollutants by the HAP adsorbent and membrane from realistic water samples are shown in Fig. S9 (see Supplementary Material). At low levels of mg L—1 concentrations,the removal efficiencies of both the HAP adsorbent and membranewere at high levels above 90% and were slightly affected by the existing substances. This application implies that the HAP adsor- bent and membrane are able to be used in practical drinking water treatments.
4.Conclusions
A novel HAP adsorbent and membrane were prepared for the removal of organic micropollutants for the first time. The hydro- philicity and hydrophobicity of the HAP materials were tunable. Both the adsorption and membrane technique were used for the removal of organic micropollutants. The adsorption data could be well described by the Freundlich model, while the adsorption kinetics followed the pseudo-second-order model. Electrostatic interactions were involved in the adsorption of organic micro- pollutants. The HAP membrane could remove organic micro- pollutants effectively by dynamic adsorption in both aqueous and ethanol solutions. The removal efficiencies of organic micro- pollutants depended on the solution composition, membrane hy- drophilicity, membrane thickness, flow rate and the initial concentration of organic micropollutants. The adsorption capacities of the HAP membranes with a sandwich structure were 6700, 6510, 6310, 5960, 5490, 5230, 4980 and 4360 L m—2 for 1-naphthyl amine, 2-naphthol, S(-)-Propranolol bisphenol S, propranolol hydrochloride, metolachlor, ethinyl oestradiol, 2,4-dichlorophenol and bisphenol A, respectively, when the concentration was 3.0 mg L—1, with a membrane thickness of 300 mm and a flow rate of 30 mL min—1. The as- synthesized HAP adsorption membrane was thus demonstrated to be a novel concept for the removal of organic micropollutants from both aqueous and organic solutions.