Durable, magnetic-responsive melamine sponge composite for high efficiency, in situ oil–water separation

Scheme 1 illustrates our relatively simple approach for preparing a superhydrophobic, magnetic-responsive MF sponge for oil–water separation, employing PDA layer, Fe₃O₄ NPs, and PTOS surface modification. Briefly, PDA layer was first introduced onto pristine MF sponge scaffold via self-polymerization of dopamine monomers, leading to a PDA/MF sponge. As shown in our previous study [41], the PDA layer not only increases the roughness of surface but also endows surface to be negatively charged for absorbing positive ions (such as Fe²⁺ and Fe³⁺) through favorable electrostatic interactions. When NH₃·H₂O, FeCl₂·4H₂O and FeCl₃·6H₂O were introduced into the reaction mixture containing MF sponge, nucleation and growth of Fe₃O₄ NPs were initiated on the surface of PDA/MF sponge through a co-precipitation method, which results in both magnetic-responsive property and enhanced surface roughness. Finally, superhydrophobic PTOS-Fe₃O₄@PDA/MF sponge was obtained through the hydrolysis of PTOS onto the sponge scaffold. Notably, our approach is highly scalable for practical applications, since it only requires low-cost reactive agents and each step is easy to conduct without requiring any strict reaction conditions, as opposed to most of the previously established methods that involve either high-cost materials (e.g. for graphene, MOF) or carefully conducted experimental conditions (e.g. inert atmosphere for surface-initiated polymerization) for the preparation of superhydrophobic materials [28, 42–45]. More importantly, compared with traditional methods such as immersion/dip-coating, in situ growth of inorganic NPs on the sponge surface usually leads to a significantly improved interfacial interaction for enhancing material durability. The as-prepared PTOS-Fe₃O₄@PDA/MF sponge exhibits excellent water repellence with a high WCA >160°, for both acid and alkali aqueous solutions. The sponge also demonstrates an excellent magnetic property for easy retrieval and outstanding oil uptake performance for both immiscible oil–water mixture and surfactant stabilized water/oil (W/O) emulsion, which will be discussed in details in the following content.

Zoom In Zoom Out Reset image size

Surface roughness is a key parameter for controlling surface hydrophobicity. According to Cassie–Baxter equations, the water contact angle of a surface relies on the volume fraction of the water/solid contact surface area, namely the ratio of the liquid/solid contact area to the total projected area of the drop basement. Therefore, producing roughness on the surface could effectively introduce 'air pockets' to achieve a superhydrophobic surface [46, 47]. As shown in figures 1, S1, and S2 (available online at stacks.iop.org/NANO/32/275705/mmedia), the surface morphology of sponges after each processing step was characterized by scanning electron microscopy (SEM). Compared to pristine MF (smooth surface from figures 1(a), (e)), the PDA/MF sponge shows a significantly increased roughness (figures 1(b), (f)) from the deposition of PDA layer. Introduction of Fe₃O₄ onto PDA/MF sponge further improves the surface roughness as illustrated in figures 1(c), (g), where NP aggregates are clearly visible. After the growth of low-surface-energy PTOS layer, PTOS-Fe₃O₄@PDA/MF sponge exhibits very limited changes in surface morphology (figures 1(d), (h), (i)). The elemental mapping images of PTOS-Fe₃O₄@PDA/MF sponge, from figures 1(j) to (p), confirm the homogeneous distributions of C, N, Fe, O, F, and Si elements onto the struts at a macro-scale, indicating the successful decoration of PDA layers, Fe₃O₄ NPs and PTOS, onto MF sponge surfaces. The heterogenous structures of different particles and layers at a nanoscale were confirmed by SEM images in figure S2.

Zoom In Zoom Out Reset image size

The surface wetting behaviors of the as-prepared sponges to water droplets (applied on their surface) were changed after each processing step/modification, as shown in figure 2. The inclusion of PDA into the sponge changes its color from white (pristine MF, figure 2(a)) to black-brownish color (figure 2(b)), which is retained after sequential modification steps, including adding Fe₃O₄ (figure 2(c)) and PTOS layer (figure 2(d)). For pristine MF sponge (figure 2(a)), the water droplets were completely wetted on the surface after being deposited onto the sponge surface, leading to a nearly 0° WCA (representing highly hydrophilic nature). However, this water-wetting behavior was significantly altered for PDA/MF, Fe₃O₄@PDA/MF, and PTOS- Fe₃O₄@PDA/MF sponges as shown in figures 2(b)–(d), with all results confirming that these sponges are highly hydrophobic (WCA all above 130°). After coating PDA on the MF sponge surface, the WCA increased to 134 ± 2.0° after being completely dried. This substantial increase in hydrophobicity cannot be simply explained by the physical natural of PDA layer since PDA is a hydrophilic polymer. Instead, this phenomenon is due to the synergistic effects of (1) selective exposure of benzene moieties of PDA instead of its hydroxyl groups to air, and (2) significantly improved surface roughness, which is also consistent with previous reports associated with 'unexpected' superhydrophobic behavior of PDA layers on other substrates [48]. The introduction of Fe₃O₄ NPs from in situ deposition onto the PDA layer leads to an increase of WCA to 154 ± 1.8°. After the modification of PTOS, the WCA of the PTOS-Fe₃O₄@PDA/MF sponge further increases to 165 ± 1.5°, showing excellent superhydrophobic property owing to the significantly increased surface roughness (attributed to PDA and Fe₃O₄) and reduced surface energy (attributed to PTOS), as previously discussed using Cassie–Baxter theory.

Zoom In Zoom Out Reset image size

The superhydrophobicity of the PTOS-Fe₃O₄@PDA/MF sponge is very stable even with highly acid and alkaline water (figure 2(e), pH value ranges from 1 to 14), which can fully meet the needs for some potentially harsh applications. When water was spilled onto the surface of PTOS-Fe₃O₄@PDA/MF sponge, it immediately bounced from the sponge surface with no residues left (figure 2(f)), which further confirms the superhydrophobicity of the PTOS-Fe₃O₄@PDA/MF sponge surface. The water-repelling property of pristine MF and PTOS-Fe₃O₄@PDA/MF sponge is shown in figure 2(g). The pristine MF sponge can completely sink into the water, while the PTOS-Fe₃O₄@PDA/MF sponge is only able to float on the water surface. Furthermore, when PTOS-Fe₃O₄@PDA/MF sponge is forced to be immersed into the water, air bubbles surrounding functionalized MF sponge surface were formed (figure 2(h)), and the sponge floats on the water surface again, immediately after the external force is removed, confirming the superhydrophobicity of MF sponge composite, attributed to the construction of micro/nanostructure (as observed from SEM, figure 1(d)) and the modification of low-surface-energy PTOS layers.

The surface elemental composition of the prepared sponges after each processing stage was investigated by XPS (figure 3(a)). The peaks at 286.9, 400.3, and 532.1 eV are attributed to the binding energy of C 1s, N 1s, and O 1s (from MF), respectively, which can be observed in the MF, PDA/MF, Fe₃O₄ @PDA/MF, and PTOS-Fe₃O₄ @PDA/MF sponges [49]. After Fe₃O₄ NPs were in situ co-precipitated onto the surface of PDA/MF sponges, additional peaks observed at 711.3 eV were ascribed to the binding energy of Fe 2p [50]. For PTOS-Fe₃O₄ @PDA/MF sponge, the new peaks of F 1s (689.0 eV) and Si 2p (102.3 eV) were observed [51, 52], corresponding to the introduction of PTOS layer. To further investigate the crystal structure of Fe₃O₄ within the PTOS-Fe₃O₄@PDA/MF sponge, X-ray diffraction (XRD, figure 3(b)) characterization was performed. Characteristics peaks of (220), (311), (400), (422), (511), and (440) were observed, corresponding to the cubic spinel structure of Fe₃O₄ (JCPDS card no. 19-0629) [41]. The thermal stability of the Fe₃O₄ @PDA/MF sponge and the loading amount of Fe₃O₄ NPs were determined by TGA, as shown in figure 3(c). For both PDA/MF and Fe₃O₄ @PDA/MF sponges, the weight loss from 375 to 400 °C is associated with the decomposition of methylene bridges from melamine sponge [28]. As a comparison, the total weight loss ratio of PDA-Fe₃O₄@MF decreases nearly by 28.1% at 800 °C, indicating a high loading content of Fe₃O₄ NPs by in situ co-precipitation method.

Zoom In Zoom Out Reset image size

The loading of Fe₃O₄ NPs not only improves hierarchical surface roughness of the PTOS-Fe₃O₄@PDA/MF sponge, but also enables the magnetic properties for easy retrieval. As shown in figures 4(a)–(c), the movement of the superhydrophobic PTOS-Fe₃O₄@PDA/MF sponge can be easily controlled by a magnet (0.2 T) to effectively absorb n-hexane (red) at desired locations in water solution (Video S1). The magnetic performance of Fe₃O₄ @PDA/MF and PTOS-Fe₃O₄ @PDA/MF sponges were both investigated by VSM (figure 4(d)), which showed them to be superparamagnetic at room temperature. The saturation magnetization (Ms) is 8.5 and 5.9 emu/g for the Fe₃O₄@PDA/MF and PTOS-Fe₃O₄@PDA/MF sponge, respectively. The slight decrease of the Ms of the PTOS-Fe₃O₄ @PDA/MF sponge was due to the introduction of PTOS, but it does not diminish the ability of retrieve these foams by a magnet as shown in figure 4(e).

Zoom In Zoom Out Reset image size

The PTOS-Fe₃O₄@PDA/MF sponge exhibits high potential for selectively absorbing organic contaminants in water solution, which is attributed to its superhydrophobicity and high porosity. Figure 5 and Video S(2–3) show that PTOS-Fe₃O₄@PDA/MF sponge can absorb both low-density n-hexane (figures 5(a)–(d)) and high-density chloroform (figures 5(e)–(h)) from aqueous solution. The saturated absorption capacity and recyclable performance of the PTOS-Fe₃O₄@PDA/MF sponge for various oils and organic solvents were also investigated, which are shown in figures 5(i), (j). The PTOS-Fe₃O₄@PDA/MF sponge can absorb nearly 59.3 g/g for n-hexane and 141.1 g/g for chloroform, which is higher than most of the previously reported MF sponge materials, as shown in table 1. For example, most of previous sorbents shown in the table have a capacity less than 100 g/g for organic solvents, while our MF sponge can absorb 141 g/g for chloroform. The WCA of this MF composite sponge is also among the highest in the table. Furthermore, compared to conventional dip-coating/blending methods, our in situ growth of NPs on the MF scaffold leads to a more robust sponge system. In order to confirm this advantage, a recycling experiment has been performed. As a result, only a very limited decease (from 141.1 to 136.7 g g⁻¹) in absorption capacity is observed after 30 cycles of absorption–desorption for chloroform, indicating the excellent durability of the PTOS-Fe₃O₄@PDA/MF sponge for recycled use.

Zoom In Zoom Out Reset image size

The prepared PTOS-Fe₃O₄@PDA/MF sponge (1 cm × 2 cm × 2 cm) was utilized as an efficient filter layer to separate the immiscible oil–water mixtures within a long neck funnel. As shown in figures 6(a)–(d), low-density n-hexane and high-density chloroform were red and water was blue. When the immiscible oil-water mixtures were introduced through the funnel containing PTOS-Fe₃O₄@PDA/MF sponge, oil was immediately absorbed in the PTOS-Fe₃O₄@PDA/MF sponge filter. As a result, water was effectively retained while oil was passed through the sponge (Video S4–5). Furthermore, other immiscible oil-water mixtures, including diesel oil, soybean oil, dodecane, n-hexadecane, mineral oil and toluene, can also be successfully separated by above method using this melamine sponge composite. According to figure 6(f), the E_s values for all immiscible oil–water mixtures using PTOS-Fe₃O₄@PDA/MF sponge are higher than 99.9%, demonstrating its excellent oil-water separation performance under the gravity filtration. In order to study the recyclability of the PTOS-Fe₃O₄@PDA/MF sponge for practical applications, separation efficiencies were measured repeatedly after each absorption-desorption cycle. As shown in figure 6(f), the PTOS-Fe₃O₄@PDA/MF sponge shows a very stable performance for oil–water separation performance, with only less than 2% performance deterioration after 15 cycles. This result confirms its robust mechanical durability and chemical stability.

Zoom In Zoom Out Reset image size

We also performed separation experiments with surfactant-stabilized water/oil (W/O) emulsions, which typically involve significantly more difficulty compared with immiscible oil–water mixtures. Herein, W/O type emulsions were prepared by the assistance of span 80 surfactant (1.0 mg mL⁻¹), in which water droplets are dispersed in the n-hexane phase (1/99, v/v). It can be observed from the optical images (figures 7(a)–(c)) that white cloudy span-80-stabilized emulsion became clear and transparent (indicative of no microphase separation) after passing through PTOS-Fe₃O₄@PDA/MF filtration layer under gravity-driven separation. Interestingly, even though the highly emulsified water droplets (ranging from 200 to 800 nm) are much smaller than the typical pore size of PTOS-Fe₃O₄@PDA/MF (>100 μm) from SEM images, successful oil–water separation was confirmed by both optical microscopy images and DLS measurement (figures 7(d), (e)). The excellent ability of this melamine sponge composite to remove water from oil matrix can be explained by the significantly decreased pore size upon squeezing into the funnel, which can effectively reject water molecules due to its superhydrophobic surface [61]. This result also indicates the broad applications of our sponge for separation of bulk oil–water emulsions.

Zoom In Zoom Out Reset image size

In order to address large-scale oil–water separation, in situ and continuous process is highly desired since other methods such as mechanical extrusion and gravity filtration are typically batch-process, time-consuming, and difficult to recover oils form water. Herein, we demonstrate the ability of our foam to in situ absorb large quantities of oils though a pump-driven process, as shown in figures 8(a)–(d) and Video S7. During this process, a PTOS-Fe₃O₄@PDA/MF sponge was embedded at the end of a water pipe (outlet) as a filter. The tubing was immersed into the oil–water mixture. Upon pumping, oil was continuously sucked into the pipe through the PTOS-Fe₃O₄@PDA/MF sponge and finally collected into the beaker. As shown in figure 8(c), oil can be completely removed and collected from the water/oil mixture, while the water remains in the solvent container. The recyclability was also investigated in figure 8(d), which shows an approximately 99.2% efficiency even after ten cycles. The mechanism for this oil–water separation can be explained by figure 8(e). When the PTOS-Fe₃O₄@PDA/MF sponge was introduced into oil/water mixture, oil was uptaken and collected rapidly by the peristaltic pump. When the oil on the water surface was absorbed, the sponge above the water was filled with air bubbles under the continuous suction of the pump [24, 62]. As suggested in figure S3, the introduction of air bubbles around the sponges allows the circulation of only the oil component, as the superhydrophobic surface of the PTOS-Fe₃O₄@PDA/MF sponge would reject water molecules under pressure. Furthermore, this high separation efficiency implies the stability of both structure and chemical composition of the as-prepared PTOS-Fe₃O₄@PDA/MF sponge, especially prepared through an effective and low-cost approach. We believe these results suggest the great potential of our highly durable, magnetic-responsive melamine sponge composite for high efficiency, in situ oil–water separation, especially for addressing large-scale oil spill applications.

Zoom In Zoom Out Reset image size