Superhydrophilic and Underwater Superoleophobic Cotton Fabric for Oil-Water Separation and Removal of Heavy-Metal Ion.

Superhydrophilic and underwater superoleophobic cotton fabric (named CS-CF-PDA, or m-CF) was prepared by modifying the cotton fabric (CF) with dopamine (DA) and chitosan (CS). The oil-water separation and heavy-metal ion (e.g., Cu(II)) adsorption performances of m-CF were investigated, and m-CF was characterized by field emission scanning electron microscopy (FE-SEM), energy-dispersive spectroscopy (EDS), Fourier transform infrared (FT-IR), and thermogravimetric analysis (TGA). The results showed that the underwater oil contact angle (UWOCA) of m-CF was more than 156°. The m-CF was used to treat artificial oily wastewater containing Cu(II) under room temperature and atmospheric pressure and gravity, by which the separation efficiency, water flux, and Cu(II) removal rate could reach 99%, 17 400 L·m-2·h-1, and 89%, respectively. Additionally, in the process of continuous treatment of oily wastewater, the water flux slightly decreased; on the contrary, the Cu(II) removal rate decreased significantly to 67% within 120 s. Cu(II) was one of the reasons for the decrease of water flux. The m-CF of adsorbed Cu(II) could be leached with HCl (0.1 mol·L-1) solution, and the Cu(II) desorption rate could reach over 95% within 120 s. After strong acid, strong alkali, high salt, and abrasion treatment, the UWOCAs of m-CF were still higher than 150°. In a word, in terms of oil-water separation, m-CF exhibited good acid, alkali, salt, and abrasion resistances. Also, it is an underwater superoleophobic material involving simple preparation, low cost, and environmental friendliness, which could remove the floating oil and heavy-metal ions from wastewater and has good industrial application prospects.

Introduction

With the rapid development of the petroleum and chemical industries, the discharge of oily wastewater and the risk of oil spill accidents increased;[1−3] it was the goal of people’s continuous efforts to develop efficient and environmentally friendly methods for treating oily wastewater. So far, many technologies have been reported for the disposal of oily wastewater, including flotation, centrifuges, oil skimmers, membrane separation, etc.[4] The membrane separation method based on a superwetting surface has been attracting extensive attention in the field of oily wastewater treatment because of its simple operation and high separation efficiency.[5−8] Notably, oily wastewater contains other pollutants, such as organic dyes and heavy-metal ions, among which heavy-metal ions are one of the most common and harmful pollutants in water pollution.[9,10] Adsorption is one of the most commonly used methods to remove heavy metals because of its simple operation, easy recycling, and low cost.[9,11−13] A superhydrophilic membrane with functional groups for heavy-metal ion adsorption could simultaneously remove oil and heavy-metal ions from wastewater, which simplifies the purification process of oily wastewater containing heavy-metal ions and reduces the cost of wastewater treatment.[14−20] Wang et al.[14] prepared an organic-inorganic composite membrane by the electrospinning technique modified with 3-[(trimethoxysilyl) propyl]-diethylenetriamine to obtain a superhydrophilic and underwater superhydrophobic membrane. The membrane was used to treat artificial emulsified oily wastewater with the desired concentrations of heavy-metal ions Pb(II), Cr(III), and Ni(II); the water flux could reach 1517 ± 53 L·m–2·h–1, and the removal efficiencies of both oil and heavy-metal ions were larger than 99%.

Cotton fabric (CF), which is abundant in nature, easily biodegradable, and environmentally friendly, is widely used as a substrate for the preparation of oil–water separation membranes.[3,8,21] Chitosan (CS) is a natural polymeric alkaline polysaccharide with abundant sources, degradable, and eco-friendly; it contains a large number of functional groups (such as −NH2 and −OH, etc.) that can interact with heavy-metal ions,[9,22,23] resulting in its being widely used to treat heavy-metal wastewater by adsorption.[8,11,24] Therefore, environmentally friendly membrane materials for oil–water separation and heavy-metal ion removal can be obtained by using the CS-modified CF.[25] Krishnamoorthi et al.[25] prepared a superhydrophilic CA-CS-CF membrane by using laccase to initiate the oxidation of catechol groups in caffeic acid (CA) to generate reactive quinoids, which were subsequently cross-linked with the amino groups of CS to coat CS on CF. They found that the membrane after multilayer compaction was used to treat artificial emulsified oily wastewater containing heavy-metal ions (Cu(II), Pb(II), and Hg(II)), which yielded good water flux, oil–water separation efficiency, and heavy-metal removal rate.

In general, there are a few reports on superhydrophilic and underwater superhydrophobic cotton fabric materials for simultaneous removal of heavy-metal ions and oils from wastewater.[25] Dopamine is an established mussel-inspired molecule that forms a hydrophilic surface when self-polymerized to form polydopamine (PDA) on various substrates.[25,26] CS can be coated on the substrate by cross-linking the amino group on it with the quinone structure in the polydopamine molecule.[25,27−29] Therefore, we propose to prepare a modified CF by coating the CS on CF with PDA to treat oily wastewater to study its performances of oil–water separation and removal of heavy-metal ions, which can provide technical support for the industrial application of environmentally friendly and low-cost cotton fabric superwetting membranes.

Experimental Section

Materials

Chitosan (CS, deacetylation degree ≥ 95%) was purchased from Aladdin industrial Corporation. Dopamine hydrochloride (DA, 98%) was purchased from Sinopharm Chemical Reagent Co., Ltd. Tri-(hydroxymethyl) aminomethane hydrochloride (Tris-HCl, AR) was purchased from Chengdu Aikeda Chemical Reagent Co., Ltd. All of the reagents (hexane, toluene, glacial acetic acid, absolute ethyl alcohol, HCl, NaOH, NaCl, Sudan red III, CuCl2·2H2O (AR)) were used without further purification. Deionized water was self-made through a laboratory water purifier. The fabric used was commercially available cotton fabric (CF, thickness 0.4 mm).

Preparation of m-CF

Figure shows the preparation process of the modified CF (named m-CF). The CF was ultrasonically cleaned with ethanol for 20 min and deionized water three times, and then dried in the oven at 60 °C before use. Preparation of modified CF was done according to the literature:[30] 4 mg·mL–1 dopamine hydrochloride (DA) was dissolved in 100 mL of Tris-HCl buffered aqueous solution (pH 8.5); the cleaned CF was put into the prepared DA buffer solution. Oxidative polymerization of DA generated polydopamine (PDA) after 24 h of reaction with constant shaking and at room temperature, which was deposited on the surface of CF. The obtained hydrophilic modified CF (named CF-PDA) was rinsed several times with deionized water, then dried in vacuum at 60 °C for 6 h. 2 wt % of chitosan (CS) solution was prepared by dissolving CS in 1% glacial acetic acid solution under magnetic stirring condition. The CF-PDA was immersed in CS solution for 10 h, then washed with 0.1% NaOH and a large amount of deionized water, and then dried in vacuum at 60 °C for 6 h to prepare the m-CF.

Figure 1

Schematic Representation of the Preparation of m-CF.

Characterization Analysis

The surface morphologies and element mappings of the samples were observed by field emission scanning electron microscopy (FE-SEM, Tescan Mira 3XH) at an accelerating voltage of 3 kV and by energy-dispersive spectroscopy (EDS, Aztec X-MaxN80), respectively. Thermogravimetric analysis (TGA) was performed under air atmosphere with 10 °C·min–1 heating rate from 25 to 600 °C, and the weight of the samples was about 8–10 mg. Fourier transform infrared (FT-IR) spectrum was recorded to investigate the modified effect and chemical compatibility between pristine CF and modified CF, and the testing wavenumber range was 4000–400 cm–1. Determination of oil in water was done with an infrared spectrophotometry oil measuring instrument (JLBG-126). The wettability of the samples was characterized by a contact angle measuring instrument. The infiltration in air and underwater oil contact angle were measured with 10 μL of water and n-hexane, respectively. The time required for the water droplet (10 μL) from just touching the sample surface to spread (i.e., the water contact angle was 0°) was named the water drop infiltration time (WIT).

Simultaneous Removal of Oil and Cu(II) by m-CF

The separation experiment of 100 mL of artificial oily wastewater containing oil and Cu(II) was carried out using m-CF (effective area was 9.6 cm2) under gravity and pre-wetted by water prior to use. Toluene (dyed by Sudan Red III) was used as the oil phase (30 mL); deionized water and CuCl2 were used to prepare the water solution, wherein the concentration of Cu(II) was 5 mg·L–1 (70 mL, pH 5.6). For clearly, the artificial oily wastewater was denoted as “oil–water”, and the Cu(II) containing artificial oily wastewater was denoted as “Cu(II)–oil–water”. The concentration of Cu(II) was determined via an inductively coupled plasma atomic emission spectrophotometer (ICP-AES). The oil–water separation efficiency and water flux are calculated using eqs (1) and (2), respectively. The Cu(II) removal rate and adsorption capacity are calculated using eqs (3) and (4), respectively.where E and F are the oil–water separation efficiency (%) and water flux (L·m–2·h–1), respectively. R and Q are the removal rate (%) and adsorption capacity of Cu(II) (mg/g), respectively. M0 and M1 represent the weight of water or water solution in the original mixture and the collected water after separation (g), respectively. S is the effective area of m-CF (m2), V1 is the volume of water (L) passing through the m-CF, and t is the permeate time (h). C0 and C are the initial and final concentrations of copper ions in water (mg/L), respectively; V is the volume (L) and M is the weight of the m-CF (g).

Chemical Stability of m-CF

To evaluate the chemical stability of m-CF,[30,31] it was immersed in strong acid (HCl solution, 0.5 mol·L–1), strong alkali (NaOH solution, 0.5 mol·L–1), and high salt solution (NaCl solution, with a concentration of 3.5, 10, and 36%, respectively) for 20 days, washed with deionized water, and dried at 60 °C. The obtained samples were named HCl-m-CF, NaOH-m-CF, NaCl(3.5%)-m-CF, NaCl(10%)-m-CF, and NaCl(36%)-m-CF, respectively. Then, their wettability and the adsorption capacity were investigated.

Abrasion Treatment of m-CF

The mechanical stability and adsorption capacity of m-CF were evaluated by rubbing on a sandpaper,[32] in which 500# sandpaper served as an abrasion surface, m-CF was placed and moved on the sandpaper under a 200 g weight, and m-CF was dragged back and forth (40 cm as a cycle) in one direction with a speed of 4 cm·s–1, and then the wetting behavior was tested after 300 cycles of sandpaper abrasion. The sample was named abrasion-m-CF after abrasion treatment.

Results and Discussion

Characterizations

The surface morphologies of the samples were investigated by SEM and EDS. According to Figure A,B, the surface of CF was smooth.[30] PDA particles with different sizes and irregular shapes were distributed on the surface of CF-PDA, which contained 4.9 wt % of N element. Compared with CF-PDA, the particle size on the surface of m-CF was more unifrom, and the content of N element was slightly reduced (4.2 wt %). DA was oxidized and polymerized on the surface of CF to form PDA particles, and long-chain CS were “wound” around the PDA particles when grafted on the CF-PDA surface, which made the depth of the “gully” formed between PDA particles shallower.[30] The chemical formulas of DA and CS were C8H11NO2 and (C6H11NO4), respectively. The content of N in CS was 8.7% by weight, which was slightly less than that of DA (9.2% by weight). Therefore, the N content of CS grafted on CF-PDA was slightly low.

Figure 2

Characterization and analysis of samples by SEM (A), EDS (B), EDS mapping images (C), and DTG (D): (a) CF, (b) CF-PDA, and (c) m-CF.

The DTG curves (differential form of TGA) are shown in Figure D. The weight loss peak of CF (Figure D(a)) at 346 °C can be attributed to the decomposition of the cellulose chain of the cotton fabric.[24,33,34] The weight loss peak of CF-PDA (Figure D(b)) appeared at 325 and 442 °C, respectively. Besides, the weight loss was lower than that of CF. The weight loss peak at 442 °C could be attributed to PDA, which may have degraded some substances on the fiber under alkaline environment (pH 8.5), leading to the easy oxidation and decomposition of the fiber and decrease of the weight loss at low temperature. The weight loss peaks of m-CF (Figure D(c)) appeared at 325 and 465 °C, respectively. The peak value of high-temperature weight loss increased and its shape widened compared with CF-PDA. The high-temperature peak could be attributed to the decomposition of CS and PDA, and the grafting of CS may also have changed the thermal decomposition performance of PDA, resulting in the increase of the width and peak value of the high-temperature weight loss peak of m-CF.

The above analyses showed that both PDA and CS were successfully coated on CF and the distribution of C, O, and N elements on m-CF was consistent (Figure C).

Wettability of the Sample

For the porous material that is hydrophilic and rich in capillaries, the capillary pressure is one of the main driving forces of water penetration into it. According to the Young–Laplace equation (eq (), the smaller the contact angle of water in the inner wall of the capillary (i.e., the stronger the hydrophilicity), the greater the capillary pressure.where Δp is the pressure difference between the upper and lower liquid surfaces of the capillary, θWA is the water contact angle of the water on the capillary wall, R is the capillary radius, and γWA is the surface tension of water.

The water drop infiltration time (WIT) on porous materials can be calculated by Washburn equation (as shown in eq (). Equation ( shows that the longer the WIT on the porous material, the larger the contact angle of water on the capillary wall, and the weaker the hydrophilicity of the material.where t is the infiltration time of the water droplet on porous materials, l is the capillary length, and η is the viscosity of water.

Obviously, the longer the WIT on the porous material, the lesser the capillary force that pushes the water through the membrane.

According to the Wenzel equation (as shown in eq (), the surface roughness reduces the water contact angle and enhances the hydrophilicity of the material.where θWA′ is the water contact angle of the rough surface, and r is the surface roughness factor (i.e., the roughness factor is the ratio of actual surface area to the apparent surface area; r is always larger than 1).

Figure shows the experimental results of wettability of the samples. According to Figure , the WIT values of CF, CF-PDA, and m-CF were 855, 354, and 117 s, respectively, and the underwater oil contact angles (UWOCA) were 139.5, 147.3, and 156.1°, respectively. CF was mainly composed of cellulose, which was rich in hydrophilic groups (−OH). The surface or capillary wall of CF was hydrophilic, and the water contact angle was less than 90° (θWA < 90°). According to the Young–Laplace equation (eq (), the capillary force generated on the CF can absorb water into the capillary. DA had polar groups (such as −OH and −NH2) as well as oxidative polymerization on the surface of CF to form PDA particles, making the surface of the material rough (Figure B(b)). Compared with CF, therefore, the hydrophilicity of the CF-PDA surface or capillary inner wall increased and the WIT of the material decreased; at the same time, the UWOCA increased. We knew that the content of non-hydrocarbon elements (O and N) in CS ((C6H11NO4)) was higher than DA (C8H11NO2), which enhanced the hydrophilicity of m-CF, while UWOCA increased and showed the superoleophobic state (θOWS > 150°).

Figure 3

Surface wettability of the samples toward water and oil on the surface of (a) CF, (b) CF-PDA, and (c) m-CF (①–④optical snapshots for the dynamic contact processes of the water droplets; ⑤UWOCA).

Performance of m-CF Experiments

Figure A shows the experimental phenomenon of oil–water separation. It can be seen from the figure that both oil (red) and water can pass through CF; in contrast, for CF-PDA and m-CF, water can completely permeate through, while oil cannot.[30] Among them, it took a shorter time (only 15 s) for water to pass thoroughly through m-CF. Figure B shows the photographs of the critical heights of the oil column of CF, CF-PDA, and m-CF, which were 2.2, 9.0, and 11.5 cm, respectively. Besides, the loading of PDA and CS on CF (i.e., m-CF), compared with unmodified CF, brought about a decrease in R and an increase in θOWS, resulting in the increase in the HOC (see eq ). Therefore, m-CF had a higher oil repellent ability than CF-PDA.

Figure 4

Experimental phenomena of the treatment of Cu(II)–oil–water with the samples at room temperature (A); the height of the oil column of the samples (B): (a) CF, (b) CF-PDA, and (c) m-CF.

When using a hydrophilic-oleophobic material for oil–water separation, the combined action of three forces determined whether the oil droplets could permeate through the filter membrane (as shown in Figure ): (1) The porous material surface is superoleophobic under water; the oil droplets on the surface of the modified CF are subjected to an upward capillary force. (2) If the density of oil is lighter than that of water, it will be subjected to upward net buoyancy. (3) Neglecting the net buoyancy, the oil droplets will permeate through the filter membrane when the capillary force is less than the static pressure of the liquid. According to Young–Laplace equation and hydrostatics equation, we can know the critical liquid height when the oil droplet can pass through the filter membrane (as shown in eqs (8) and (9)); when the actual liquid height is larger than the critical liquid height, the oil droplet will penetrate the filter membrane.where HOC is the critical height of oil, HWC is the critical height of water, ρO and ρW are the densities of oil and water, respectively, γow is the oil–water interfacial tension, and θOWS is the contact angle of the underwater oil on the solid surface (i.e., UWOCA).

Figure 5

Schematic diagram of forces acting on the oil droplets on the “water-removing” type porous materials (Note: the oil droplets are very small and the static pressure of the water column at 2R height is negligible). (a) Oil above the filter membrane, (b) water above the filter membrane, and (c) forces analysis of the oil droplet.

Commonly if ρO < ρW, then HOC > HWC. Therefore, when the total heights of liquid and oil in the oil–water separation equipment are less than HWC, it can ensure that the oil droplets do not pass through the filter membrane and successfully separate oil from water. Obviously, the larger the θOWS, the higher the HWC, or the smaller the particle size of the oil droplets that can just pass through the filter membrane under certain conditions, the higher the oil–water separation efficiency.

The loading of PDA and CS on CF (i.e., m-CF), compared with unmodified CF, led to a decrease in R and an increase in θOWS, resulting in the increase in HWC. Therefore, when the oil–water separation conditions are the same, the m-CF material can carry out oil–water separation successfully while CF cannot. These results are consistent with those in Figure .

Figure shows the continuous treatment of the solution containing oil and Cu(II) wastewater (i.e., Cu(II)–oil–water), which was divided ten times to investigate the performances of m-CF. As shown in Figure A,B, respectively, the E value of m-CF was larger than 99%, which was almost unchanged with the increasing time of the experiments, and the F values of the first time and the tenth time were about 17 400 and 15 400 L·m–2·h–1, respectively, which was decreased by 11%. The R value is shown in Figure C, which was 89% in the first experiment and decreased by 57% in the tenth time. Meanwhile, in a similar study by Krishnamoorthi et al.,[25] the water flux, oil–water separation efficiency, and heavy-metal ion removal rate were 50 050 L·m–2·h–1·bar–1, 99, and 99%, respectively. Compared with our study, the reason for the higher water flux and heavy-metal ion removal rate of the membrane may be that it was allowed under pressure and was treated by multilayer compaction. Figure D shows that the Cu(II) cumulative adsorption capacity on the membrane after ten times of continuous treatment of wastewater by m-CF (named used-I(10)-m-CF) was about 11.0 mg·g–1. The above results indicated that (1) the m-CF had good oil–water separation efficiency and water flux, as well as removed trace amounts of Cu(II) from water solution. (2) With the increasing times of the experiments, the cumulative adsorption capacity of Cu(II) on m-CF increased, while the strength of the adsorption site or non-adsorption site decreased, resulting in the reduction in the removal rate of Cu(II).[35,36] (3) With the increasing times of the experiments, the water flux decreased gradually, while the oil–water separation efficiency remained almost unchanged. What was the reason for the decrease of water flux of the m-CF? Therefore, the used-I(10)-m-CF was analyzed by SEM, EDS, FT-IR, and DTG, as shown in Figure .

Figure 6

Simultaneous removal of oil and Cu(II) by m-CF at room temperature; oil–water separation efficiency (A), flux (B), Cu(II) removal rate (C), and cumulative adsorption capacity of Cu(II) (D).

Figure 7

Wettability and characterization analysis of used-I(10)-m-CF. Wettability (A) (①–④optical snapshots for the dynamic contact processes of the water droplets,⑤UWOCA), SEM (B), EDS (C), DTG (D), and FT-IR (E): (a) m-CF and (b) used-I(10)-m-CF.

As shown in Figure A,C, the WIT of used-I(10)-m-CF was 185 s and UWOCA was 155.3°. Compared with m-CF, the surface roughness of used-I(10)-m-CF decreased slightly and the Cu element appeared with a content of 1.5 wt %, which was close to the result in Figure D. The oil was slightly soluble in water (under the conditions of room temperature and atmospheric pressure, the solubility of toluene in water is about 0.5 g·L–1). It may be that when the aqueous phase containing trace amounts of oil permeated m-CF, the oil was absorbed and filled in the “gully” on the surface of the membrane, resulting in the slightly decreased roughness of used-I(10)-m-CF.

Figure D shows that the two thermal degradation peaks of used-I(10)-m-CF were higher than those of m-CF, which were 334 and 491 °C, respectively. It was possible that the adsorption of Cu(II) increased the difficulty of pyrolysis of used-I(10)-m-CF.

As shown in Figure E, the absorption peak of used-I(10)-m-CF at 2995–2800 cm–1 (methyl and methylene) was slightly stronger than that of m-CF. The methyl or methylene content per unit mass of toluene was higher than that of cellulose, DA, or CS. It may be that used-I(10)-m-CF adsorbed trace amounts of toluene and slightly enhanced its adsorption peak at 2995–2800 cm–1.

CF and modified CF are porous materials rich in capillaries,[37] where the inner surface area is much greater than the outer surface area. Pollutants of toluene and Cu(II) were mainly adsorbed on the inner surface (i.e., the wall of the capillary) and affected the WIT of modified CF, while they hardly affected the UWOCA. Therefore, WIT increased with the time of the experiments, the capillary force of driving water through m-CF decreased, and the F value also decreased. Even so, the values of E and UWOCA were almost unchanged.

Adsorption–Desorption Cu(II) Performances of m-CF

The results of the studies mentioned in Section showed that m-CF had a certain ability of simultaneous removal of Cu(II) and oil from oily wastewater. On continuous feeding of the oily wastewater, within 120 s (that is, the total time of seven consecutive feeds of oily wastewater) the Cu(II) removal rate decreased from 89 to 67%, and less than 38% in the tenth second. Thus, the following questions were mentioned: (1) Can the removal rate of Cu(II) be increased by changing the operating conditions? (2) Can m-CF regenerate quickly and recover the ability of Cu(II) adsorption? For these reasons, the Cu(II) static adsorption and dynamic desorption experiments of m-CF were carried out, as shown in Figure .

Figure 8

Cu(II) static adsorption experiments (A–C), Cu(II) dynamic desorption experiment of the used-I(10)-m-CF (D). (a–c) Treatment of Cu(II)–water by m-CF, CF-PDA, and CF, respectively, (d) treatment of Cu(II)–water–dissolved oil by m-CF.

The static adsorption experiment of m-CF was carried out with artificial wastewater containing 50 mg·L–1 Cu(II) (i.e., Cu(II)–water). Fifty milliliters of wastewater and about 0.40 ± 0.08 g of the m-CF sample were added to a 250 mL conical flask at a constant shaking for 6 h, and then the liquid phase was taken for further analysis. The artificial wastewater containing Cu(II) was prepared with CuCl2 and deionized water, and the initial pH of the wastewater was adjusted using 0.1 mol·L–1 NaOH or HCl. The adsorption capacity of Cu(II) is calculated by eq (. used-I(10)-m-CF was fixed on the sand-core apparatus, and the desorption experiment was leached with the HCl solution of 0.1 mol·L–1 at 20 mL·min–1. The desorption rate (named D) of Cu(II) is calculated according to eq (.where D is the desorption rate (%) of Cu(II), C1 and V2 are the concentration of Cu(II) in the desorption solution (mg·L–1) and the volume of desorption solution (L), and MCu(II) is the initial Cu(II) content on used-I(10)-m-CF (mg).

It can be seen from Figure A that the Cu(II) equilibrium adsorption capacities of CF, CF-PDA, and m-CF were 2.59, 3.60, and 4.90 mg·g–1, respectively. Compared with CF, the Cu(II) adsorption capacity of CF-PDA and m-CF increased by 28.1 and 47.1%, respectively. The above results showed that DA and CS loaded with heavy-metal adsorption functional groups (such as −NH2 and −OH, etc.) could improve the modified CF adsorption capacity.[9,38] As shown in Figure A(a), the adsorption capacity of Cu(II) could reach 50% of the equilibrium adsorption capacity of m-CF within 1 min, indicating that m-CF can remove Cu(II) from wastewater in a short time. Figure B,C shows that the effects of pH and temperature on the adsorption of Cu(II) by m-CF were consistent with the results of other studies with CS as the main heavy-metal adsorption composition.[12,39] As the pH increased, the adsorption capacity of Cu(II) decreased slowly,[12,40] and the adsorption capacity decreased with increasing temperature.[41,42]Figure A,C shows that the equilibrium adsorption capacity of m-CF was about 4.90 mg·g–1 under the experimental conditions of pH 5.6 and Cu(II) concentration of 50 mg·L–1. The literature showed that −OH and −NH2 have a strong binding ability to heavy-metal ions, and −NH2 is the main heavy-metal adsorption site under the condition of weak acid.[9,12,39] Sahebjamee et al.[42] prepared the CS/PVA/PEI membrane for Cu(II) removal, for which the equilibrium adsorption capacity of Cu(II) was 125 mg·g–1 under the conditions of 25 °C, pH 6, and Cu(II) concentration of 30 mg·L–1. It was possible that there was either a small amount of PDA and CS on the modified CF or the density of the exposing adsorption groups of the heavy metal was low. The Cu(II) adsorption capacity of m-CF was less than that of other adsorbents based on CS modification, resulting in the low removal rate of the simultaneous removal of Cu(II) from wastewater separation.

Figure D shows that when the m-CF containing 11.0 mg·g–1 Cu(II) (i.e., used-I(10)-m-CF, see Figure D) was leached with the HCl solution of 0.1 mol·L–1 at 20 mL·min–1, the desorption rate of Cu(II) was more than 95% within 120 s. According to Figures D and 8D, we therefore considered that when using multiple oil–water separation devices in parallel, one of the devices simultaneously carried out oil–water separation and adsorption of Cu(II), while another performed the regeneration process, which could be contributing to the simultaneous removal of oil and Cu(II) by m-CF. However, in order to make m-CF suitable for practical application in industry, it is necessary to improve the membrane preparation and desorption processes to increase the adsorption capacity and desorption rate.

Effect of Cu(II) on Oil–Water Separation

In order to investigate the effect of Cu(II) on the oil–water separation performance, experiments similar to the ones described above in Section , were carried out with oily wastewater (i.e., oil–water system, 30 mL of toluene, and 70 mL of deionized water) and Cu(II)-containing oily wastewater (i.e., Cu(II)–oil–water system) using m-CF. To make it clear that the m-CF adsorption of Cu(II) (which contained 4.90 mg·g–1 Cu(II) and was named Cu(4.90 mg/g)-m-CF) was carried out on an oil–water system, see Figure .

Figure 9

Effect of Cu(II) on the oil–water separation of m-CF at room temperature: (a) treatment of oil–water and (b) Cu(II)–oil–water by m-CF, and (c) treatment of oil–water by Cu(4.90 mg/g)-m-CF.

Figure shows the m-CF-treated Cu(II)–oil–water and oil–water system, respectively. The F values of the first experiment were both 17 400 L·m–2·h–1. But the F value of Cu(II)–oil–water was 14 500 L·m–2·h–1 at the 20th treatment, which decreased by 16.7%, and the F value of oil–water was 16 400 L·m–2·h–1, which only decreased by 5.7%. Obviously, the Cu(II) in wastewater had negatively affected the F value of m-CF.

Figure shows that the F value of Cu(4.90 mg/g)-m-CF was 16 400 L·m–2·h–1 for the first time, which decreased by 6.1% at the 20th treatment. However, the decrease rate of the F value of Cu(4.90 mg/g)-m-CF was much smaller than that of m-CF-treated Cu(II)–oil–water (16.7%). In addition, the Cu(II) contents of the samples treated with oil–water by Cu(4.90 mg/g)-m-CF and Cu(II)–oil–water treated by m-CF for 20 times were analyzed, which were 4.20 and 13.8 mg·g–1, respectively. The above phenomena showed that the increase of the content of Cu(II) in m-CF will lead to a significant decrease in F value.

It can be seen from Figure that Cu(II) had no influence on the E value, which was more than 98%.

Effect of Dissolved Oil on the Adsorption of Cu(II)

In this study, the effect of trace amounts of dissolved oil in water on Cu(II) removal was studied. Taking 200 mL of deionized water and 30 mL of toluene, the mixture was stirred magnetically at room temperature for 24 h and subjected to stratification. Then, the water phase was taken to further prepare the artificial wastewater with Cu(II), which contained 252 mg·L–1 dissolved toluene (determined with an infrared spectrophotometry oil measuring instrument) and 50 mg·L–1 Cu(II) (i.e., Cu(II)–water-dissolved oil). Subsequently, an adsorption experiment was carried out. The results are shown in Figure A(d); the equilibrium adsorption capacity of m-CF was 4.50 mg·g–1, which was 8.2% lower than that for the treatment with “Cu(II)-water” wastewater (Figure A(a)). The large π bond in the molecule of toluene endowed it with a certain polarity, and it could interact with polar groups (such as −NH2 and −OH) on m-CF, which slightly reduced the adsorption capacity of Cu(II).

Stability of m-CF

In this work, we investigated the wetting behavior and static adsorption of Cu(II) by m-CF after acid, alkali, and salt immersion treatments and after sandpaper abrasion, as shown in Figure .

Figure 10

Surface wettability and the Cu(II) adsorption capacity of m-CF after treatment in an artificial environment of acid, alkali, salt, or abrasion (A) and (B), respectively. (a) m-CF, (b) HCl-m-CF, (c) NaOH-m-CF, (d) NaCl(3.5%)-m-CF, (e) NaCl(10%)-m-CF, (f) NaCl(36%)-m-CF, (g) abrasion-m-CF (①–④optical snapshots for dynamic contact processes of the water droplets, ⑤UWOCA).

As shown in Figure A, the WITs of m-CF, HCl-m-CF, NaOH-m-CF, NaCl(3.5%)-m-CF, NaCl(10%)-m-CF, NaCl(36%)-m-CF, and abrasion-m-CF were 117, 215, 268, 50, 60, 45, and 180 s, respectively, and UWOCAs were 156.1, 154.9, 154.6, 157.6, 156.1, 154.1, and 152.6°, respectively. There results showed that (1) the WIT of m-CF increased after immersing in acid or alkaline solution for 20 days. On the contrary, (2) the WIT of NaCl-m-CF decreased after immersing for 20 days, and (3) after abrasion, the WIT of m-CF increased. Fortunately, (4) the UWOCAs of all samples were still larger than 150°.

From Figure B, compared with m-CF, the equilibrium adsorption of Cu(II) by HCl-m-CF, NaCl(36%)-m-CF, NaOH-m-CF, and abrasion-m-CF were all decreased to 4.04, 4.64, 3.18, and 4.36 mg·g–1, respectively, i.e., the equilibrium adsorption capacity of Cu(II) decreased by 17.6, 5.3, 35.1, and 11.0%, respectively. What was the reason for the decrease of the equilibrium adsorption of Cu(II) on the m-CF? In order to illustrate the effects of m-CF after treatment in harsh environment on the wettability and Cu(II) adsorption, SEM and EDS analyses were performed for HCl-m-CF, NaOH-m-CF, NaCl(36%)-m-CF, and abrasion-m-CF, respectively, as shown in Figure .

Figure 11

SEM and EDS of m-CF after treatment in an artificial environment of acid, alkali, salt, or abrasion. (a) HCl-m-CF, (b) NaOH-m-CF, (c) NaCl(36%)-m-CF, and (d) abrasion-m-CF.

It can be seen from Figure a that there were etching holes on the surface of HCl-m-CF, and the content of N element (4.1 wt %) was close to that of m-CF. The H+ of HCl entered the internal structure of CF, and, adhering to the β-1,4 glycosidic bond of CF, could hydrolyze cellulose, which led to the partial breakage of the cellulose macromolecules and decrease of its polymerization.[43] CS was similar to the cellulose structure (β-1,4 glycosidic bonds), which could be hydrolyzed in acidic aqueous solution, resulting in partial hydrolysis shedding or molecular chain breaking of CS, and forming etching holes. It may also be that the shedding amounts of CS were not much, so that the content of N element did not change obviously. However, the hydrophilicity and the content of polar functional groups (−NH3, −OH) chelating with Cu(II) were decreased, resulting in decrease of the equilibrium adsorption of Cu(II).

As shown in Figure b, compared with m-CF, the surface roughness of NaOH-m-CF and the content of N element (3.3 wt %) decreased obviously. The loss of PDA deposited on m-CF and the CS adhered on PDA were attributed to the dissolution of PDA by NaOH;[44] thus, the roughness, the content of N element, and the hydrophilicity of the NaOH-m-CF decreased. What’s worse, the equilibrium adsorption capacity of Cu(II) was greatly reduced.

From Figure c, it can be seen that the surface of NaCl(36%)-m-CF was still rough, the content of N was 4.3 wt %, and there were Na and Cl elements, which may be because NaCl did not dissolve or decompose PDA and CS on m-CF, so that the morphology of NaCl-m-CF was similar to that of m-CF. Additionally, it may be that Na+ and Cl– adsorbed on the surface of m-CF and increased the hydrophilicity of the capillary wall,[45] so the WIT of NaCl-m-CF decreased and UWOCA was almost unchanged. However, the residual Na(I) on m-CF competed with Cu(II) for the active site,[40,46] resulting in the decrease of the equilibrium adsorption of Cu(II).

It can be seen from Figure d that abrasion-m-CF was partially worn, resulting in the decrease of N content (3.3 wt %), and indicating that parts of PDA and CS were lost, compared with m-CF, causing the WIT to be increased. It was possible that m-CF immersed in acid, alkali, or salt solution, which had a uniform effect on the inner and outer surfaces of the porous materials. In contrast, the abrasion treatment had a great influence on the outer surface of the porous m-CF and the pores changed. Immersing in acid, alkali, or salt solution changed the WIT of m-CF but UWOCA was almost unchanged, while abrasion treatment changed both WIT and UWOCA (UWOCA decreased, but was still larger than 150°).

In a word, although the F value of the oil–water separation of m-CF decreased to some extent under the condition of strong acid or strong alkali or abrasion, as well as the removal ability of Cu(II), its E value remained almost unchanged. It showed that m-CF still had good oil–water separation ability under harsh environments (strong acid, strong alkali, high salt and scouring, etc.).

Conclusions

Superhydrophilic and underwater superoleophobic cotton fabric (i.e., m-CF) was prepared using DA and CS, whereby m-CF simultaneously removed the oil and Cu(II) from oily wastewater under gravity, and the oil–water separation efficiency, the water flux, and removal rate of Cu(II) were more than 99%, 17 400 L·m–2·h–1, and 89%, respectively. The oil–water separation efficiency was not affected during the continuous treatment of oily wastewater containing Cu(II), while the water flux of m-CF decreased slowly. Besides, in about 120 s, the Cu(II) removal rate decreased to 67%. The reason we proposed for the decrease of water flux was that both the dissolved oil and Cu(II) on the m-CF and, in turn, trace amounts of dissolved toluene in the water had a slight effect on the removal of Cu(II).

After immersing in strong acid, strong alkali, and high salt solution of m-CF for 20 days at room temperature, and abrasion with 500# sandpaper, the equilibrium adsorption capacity of Cu(II) decreased by 17.6, 35.1, 5.3, and 11.0%, respectively. Although the water droplet infiltration time was prolonged by different degrees (except for the case of high salt), the UWOCAs were still in superoleophobic state. In terms of oil–water separation, m-CF had good acid resistance, alkali resistance, salt resistance, and wear resistance.

The m-CF on which Cu(II) was adsorbed was leached with 0.1 mol·L–1 HCl solution at 20 mL·min–1, and the desorption rate within 120 s was more than 95%. The decline rate of the Cu(II) adsorption capacity of m-CF was approximately equal to its desorption rate. Thus, the use of multiple oil–water separation devices in parallel would be conducive to the industrial applications of m-CF.