The various mechanisms of pollutant removal involve separation, adsorption, desorption, oxidation and degradation. 1 Typical adsorbents include gypsum, activated carbon, chitosan, fly ash, zeolite, clay minerals, peat and wood. 2 Application of nanomaterials, such as carbon nanotubes, microspheres and flowerlike ferric hydroxide 3 increases the surface area (higher adsorption capacity) of adsorbents. Moreover, additional catalytic materials can be included to degrade hazardous organics using released radical species (OH˙, HOO˙, or O 2 ˙-) to accomplish cleaning via an adsorption-oxidation-desorption mechanism. 4 For instance, the maximum adsorption capacity of 465.5 mg g -1 was reported for methylene blue using gelatin, iron oxide nanoparticles and carbon nanotube nanocomposites. 5 Using photocatalysts, such as TiO 2 , ZnO, Fe 2 O 3 , WO 3 , Ag 2 CO 3 , ZnS and BiTiO 3 , it is feasible to generate OH˙using solar light without additional chemicals. 6,7 Photocatalysts offer highly efficient and environmentally friendly routes to decompose undesired contaminants in water. Wide band gap semiconductors (e.g. TiO 2 , ZnO) utilize solar light to achieve oxidation and provide a promising green technology for water purification. 8,9 Upon illumination of semiconductors, electron-hole pairs are created. Holes react with water to form hydroxy radicals, while electrons react with oxygen to form superoxide anion. [10][11][12] However, bulk semiconductors suffer from low surface area and thus, a significantly lower number of catalytic sites are available. A high surface area of adsorbent/catalyst helps to capture a large amount of pollutants and reduce the amount of expensive adsorbents/ catalysts. 13 On the other hand, nanoparticles dispersed in rivers and lakes represent an environmental contamination themselves. Hence, it is highly desirable to encapsulate NPs in three-dimensional matrices, e.g. graphene supported catalysts for organic dyes degradation. 14 Hydrogels can absorb a large quantity of water and retain it without dissolution for applications including forward osmosis, desalination, wastewater clean-up, removal of microbial, chemical and biohazardous threats. 15 Hydrogels can swell or deswell in response to changes in pH, solvent composition, magnetic, electric fields and ultrasound irradiation. Permeability of hydrogel films can be controlled, for example, by tuning the composition ratio between polyethylene glycol diacrylate (PEGDA) and RNH 3 Cl monomers. 16 Water purifying materials can be physically entrapped in hydrogel network for adsorption, separation and removal of dyes and heavy metal ions. 17 Hybrid nanoparticle-hydrogel composites linked via physical adsorption or covalent bonds enable highly efficient adsorptive removal of aqueous pollutants. [18][19][20][21][22] Recently polymer nanocomposites were shown to have a high potential to avoid agglomeration of high surface area nanoparticles for water cleaning operations. 23 Other examples include, graphene oxide-based hydrogel particles are demonstrated for oil decontamination. 24 Hydrogels embedded with hydrated ferric oxide nanoparticles are used for arsenate removal. 25 A tripeptide-based hydrogel is shown for removal of toxic organic dyes and lead (Pb 2+ ) ions, 26 and laccase immobilized hydrogel are designed for removal of bisphenol A. 27 In addition, hydrated polymer networks and nanostructured hierarchical membranes can be prepared for water purification using solar-induced vapourization. 28,29 For instance, ionisable aromatic pollutants are removed from water using nano gamma-Fe 2 O 3 containing cationic hydrogel. 30 Moreover, water desalination using hydrogel membranes for obtaining solar-induced vapourization 31 and using forward osmosis 32 have recently attracted high attention. 33 Graphene-based hydrogels with agarose (AG) as a stabilizer, reducing agent and the polydopamine functionalized graphene hydrogels (PDA-GH) are shown to enhance degradation and removal of toxic dye pollutants, organic solvents, heavy metal ions (Cd 2+ , Pb 2+ , Cu 2+ ) and oils. 34 Certain challenging aspects of hydrogel filtration must be taken into account, such as embedded composite adsorbents/catalysts, adsorption kinetics, operations in wider pH ranges, time of filtration, regeneration, reusability and resources recovery. 35 Recently, the development of environmental micromotors powered by external magnetic fields, 36,37 as well as lightdriven (photocatalytic, plasmonic) [38][39][40][41][42] and chemical fuel powered catalytic/hybrid (i.e., H 2 O 2 ) micromachines has attracted attention. [43][44][45][46][47][48][49][50][51] Particularly, micromotors are capable to outswim diffusion, 52 generate reactive species, and, as a result, lead to more efficient decontamination of pollutants in shorter times. 53,54 65 For instance, methylene blue (MB) has been successfully removed using layered manganese oxide micromotors in H 2 O 2 solutions. 66 Herein, we report a water cleaning architecture employing encapsulated photocatalytic nanoparticles in poly(acid) hydrogel microcapsules. The microcapsules are fabricated using complex water-in-oil-in-water double emulsion drops produced from a glass capillary microfluidic devices. An aqueous dispersion of photocatalytic nanoparticles and methacrylic anhydride form the core and shell of the double emulsion drops, respectively. Upon photo-polymerization and hydrolysis, polyIJmethacrylic acid) hydrogel microcapsules containing photocatalytic nanoparticles in the aqueous core are obtained. Microcapsule shells exhibit reversible pHresponsive permeability, 68 and the dynamic capsules selfpropel in hydrogen peroxide solutions. 69 Thin shell hydrogel microcapsules with a photocatalytic aqueous core can potentially provide enhanced reaction, diffusion, and separation of molecular species in comparison to bulk hydrogels with embedded nanoparticles. 70 Dynamic properties of microcapsules (magnetic, self-propelled) and a prototype flow reactor device with a layer of microcapsules are demonstrated to reduce time required for MB removal.
A glass capillary microfluidic device is used to produce monodisperse double emulsion microdrops with the desired diameter, shell thickness and aqueous core. 67,68 Fig. 1a shows the experimental setup: two tapered cylindrical capillaries aligned inside a slightly larger square capillary. Methacrylic anhydride (MAAn), ethylene glycol dimethacrylate (EGDMA) are used as monomers and oil to fabricate the double emulsion drops with added radical photoinitiator 2-hydroxyl-2-methylpropiophenone (Darocur 1173). MAAn is water immiscible, leading to the formation of an acidprecursor-containing shell around a hydrophilic aqueous core. When hydrolysed to polyIJmethacrylic acid), the EGDMA cross-linker retains the structural integrity of the capsules. As shown in Fig. 1a-c, the outer phase (water + PVA), the inner phase (water + PVA + photocatalysts), and the middle phase (monomers) are used to assemble monodisperse double emulsion drops containing photocatalysts in the core. As illustrated in Fig. 1d, exposure to UV light leads to photopolymerization of the hydrophobic polyIJanhydride) networks. 68 After treating the microcapsules in pH 11 for 24 hours, the anhydride groups (red colour, Fig. 1d) of the polyIJanhydride) network are cleaved to carboxylic acid groups (blue colour), which results in a hydrophilic poly(acid) network.
Environmental Science: Nano Paper Fig. 2a-h illustrates optical microscopy and FESEM images of fabricated hydrogel capsules showing a visible aggregation of nanoparticles in the aqueous core. Fig. 2b-f shows the ZnO (Fig. 2b ande) and TiO 2 (Fig. 2c andf) nanoparticles in capsules. Fig. 2g andh shows the FESEM images of a single capsule and a collapsed capsule (diameter ∼200 μm, shell thickness 5 μm), respectively. SEM images of TiO 2 nanoparticles, ZnO nanoparticles, TiO 2 nanoparticles in microcapsule, ZnO nanoparticles in microcapsule, and the corresponding elemental mapping images of element C, O, Ti, and Zn are shown in Fig. S3. † The photocatalysts' particle sizes were measured by laser particle analyzer (Nano ZS90, Malvern, UK), the results are shown in Fig. S4. † Analysis of particle size distribution indicates that an average size of TiO 2 and ZnO nanoparticles are 259 nm and 456 nm, respectively. A Fourier transform infrared spectroscopy (FTIR) is used to confirm the hydrolysis of the capsules, shown in Fig. 2i. We see the presence of an absorption band around 3000-3500 cm -1 , corresponding to O-H stretching, which is typically observed only in the FT-IR spectrum of hydrolysed capsules. In addition, the contact angle is measured to study the wettability of the surfaces of the polyIJanhydride) network and the poly(acid) hydrogel, shown in the inset Fig. 2i (I) and (II). The measured contact angle of the polyIJanhydride) network is 83°, indicating that the polyIJanhydride) network is hydrophobic, while the contact angle of the poly(acid) hydrogel is close to 0°, proving that the polyIJanhydride) network is hydrolyzed to hydrophilic poly(acid) hydrogel. Fig. 2j-l shows the X-ray diffraction (XRD) and 2D wide angle X-ray scattering (WAXS) patterns (Fig. 2k andl) of the hydrophobic poly (anhydride) network and poly (methacrylic acid) hydrogel. In this measurement, both the polyIJanhydride) network and the hydrolyzed poly(acid) hydrogel are amorphous and the broad bands of poly(acid) hydrogel are at 18°and 32°indicating that after hydrolysis the poly(acid) hydrogel gets more ordered than the polyIJanhydride) network. Stability in salts was also measured by inserting capsules in 1 M NaCl and 1 M K 2 SO 4 aqueous solution for 7 days (Fig. S2 †).
2.1.1 Adsorption and photocatalytic degradation. MB is used as a model organic dye pollutant to estimate the adsorptive performance and photocatalytic oxidation using synthesized capsules. To demonstrate the adsorption and photocatalytic activity, samples are immersed in the MB aqueous solution. Fig. 3a-g shows the temporal evolution of MB concentrations in aqueous solution in the dark and under UV light. For comparing hydrogel capsules and particles, TiO 2 nanoparticles are mixed with homogeneous hydrogel precursor to form spherical microparticles (HHP) with diameters of ∼200 μm (optical and SEM images are shown in Fig. S1,ESI †). In the latter case, TiO 2 is mixed using the same procedure, followed by an exposure to UV light and The high adsorption efficiency of the samples is related to the porous structure of hydrogels. After 80 min soaking, UV light was turned on (Fig. 3). During exposure to UV light the photocatalytic degradation of MB is realized by photoinduced holes (h + ) and their secondary products ˙OH, HO 2 ˙. A new equilibrium is established due to consumption of MB in microcapsules, where more MB molecules are absorbed in the solution and the MB concentration decreases further, shown in Fig. 3h. A negligible change due to photodegradation is observed in the aqueous solution using MB (w/o microcapsules) and bare microcapsules (w/o photocatalyst), while ZnO and TiO 2 in capsules removed nearly 100% of the MB in aqueous solutions. PIC shows a photocatalytic removal of MB comparable to HHP (3.61% of the original MB concentration remained in HPP). Recently reported non-equilibrium uptake kinetics simulation results [70][71][72][73][74] indicate that hydrogels' polymer chains can restrict solute dynamics, and the mobility of co-solutes can be reduced due to specific adsorption. In agreement with this explanation, we detected that PIC have a slightly enhanced photocatalytic performance than TiO 2 in homogeneous particles. Both PIC and TiO 2 in HHP are decoloured upon UV exposure, meanwhile no significant change was detected in an aqueous solution with MB (w/o microcapsules) and bare capsules (w/o photocatalyst), shown in Fig. 5. PIC and HHP are decolored from the dark blue to the original transparent colour by exposing samples to UV light during 120 min. FTIR measurements were performed to confirm the adsorption and photocatalytic degradation of MB in PIC, as shown in Fig. 5. Bands at 1596 cm -1 and 1330 cm -1 correspond to CN aromatic and C-N aromatic bonds in MB, respectively, and band at 1247 cm -1 is attributed to the C-N aliphatic band in MB. After the adsorption process, the PIC were in blue colour, these bands appeared, indicating the presence of MB. And after UV exposure the PIC gets its original transparent colour, the bands disappeared, it suggests the cleavage of CN aromatic, C-N aromatic, and C-N aliphatic band because of photocatalytic degradation (Fig. 4).
2.1.2 A removal of MB using flow microreactor. Flow microreactors based on the hydrogel capsules with aqueous photocatalytic core and permeable shells allow the construction of a new prototype device. Flow microreactors enable excellent control over reaction rates, diffusion, dynamic stirring of photocatalytic performance towards the direct adsorption, degradation and separation of pollutants. In our method, a large amount of PIC is loaded into the flash chromatographic column (17.0 mm outer diameter, 13.4 mm inner diameter) with a fritted glass bed and a Teflon stopcock (Fig. 6a). When the valve is opened, the MB solution flows through the filter consisting of the hydrogel microcapsules, while MB molecules are adsorbed by the PIC in the column. The 20 ml of MB solution are cleaned within 97 seconds. The inset UV-vis spectrum in Fig. 6a indicates that nearly 100% MB is adsorbed by the PIC sample. In addition, the PIC in the flow reactor can be regenerated by a continuous exposure to UV light. Optical images are recorded every 30 min during the UV-regeneration process over 4.5 hours, shown in Fig. 6b. The colour of the PIC solution is initially intense blue and finally, it returns to their original colourless appearance (slightly brown from the optical images, which can be attributed to a tiny amount MB remained in PIC due to UV unexposed parts in flow reactor).
In summary, we demonstrate fabrication of poly(acid) hydrogel microcapsules containing photocatalytic nanoparticles in their aqueous core for the removal of MB pollutant. The synthesized PICs are characterized by FESEM, optical microscopy and FT-IR. Due to permeability of the hydrogel shell and its adsorptive performance, MB molecules can diffuse into the aqueous core of PIC, while ZnO and TiO 2 photocatalytic NPs stay trapped inside the functional core. A flow reactor device is designed to demonstrate the capabilities of a hydrogel capsule-filter, adsorbing nearly 100% of MB over short time. Afterwards, the PIC materials are regenerated using UV light. Hydrogel capsules can potentially trap high surface area photo-catalytic nanomaterials as a liquid dispersion within the dynamic absorbent materials that can be easily removed from the environment unlike free photocatalytic nanoparticles. In this study, hydrogel capsules containing liquid photocatalytic core remove MB at rates comparable to photocatalyst enclosed in homogeneous hydrogel particles. Time of molecules/particles mixing in a fluidic volume can be estimated by the following equation t mix = d c 2 /D, where d c is the diameter of the capsule and D is the diffusion coefficient. Using traffic time one can also estimate how often molecules/particles can meet in a fluidic volume t traffic = d c 3 /Dr, where r is the sum of molecule/ particle diameters. 69 As a result, a reduction of the diffusion constant in homogeneous hydrogels can increase time required for a pollutant adsorption/degradation. Consideration of typical mixing and traffic time in capsules and particles can provide insight into future design of more efficient chemical microreactors for adsorption/degradation of pollutants. For instance, it is of high interest to test efficiency of pollutants removal using significantly smaller capsules and photocatalytic nanoparticles.
Methacrylic anhydride (MAAn), ethylene glycol dimethacrylate (EGDMA), and radical photoinitiator 2-hydroxyl-2methylpropiophenone (Darocur 1173), were purchased from Aladdin Industrial Corporation (USA). Sodium dodecyl benzene sulfonate (SDBS), poly (vinyl alcohol) (PVA, Mw 13 000-23 000), hydrogen peroxide, zinc oxide, potassium hydroxide and titanium oxide were provided by Hushi Corporation (China). Capillaries were provided by Jinglai Corporation (China). Microscope slides were purchased from Sail Brand (China). Epoxy adhesive were purchased from Ailete (China).
(1) Preparation of the glass capillary microfluidic device. Two tapered cylindrical capillaries were obtained by fuse stretching (Capillary fuse stretcher (PC-10, NARISHIGE, Japan)) and their tips were polished using sand paper to the appropriate sizes. The injection capillary was treated with octadecyltrimethoxysilane to obtain hydrophobic property, and the collection capillary was treated with 2-(methoxyIJpolyethyleneoxy)propyl)trimethoxysilane to obtain hydrophilic property and prevent the double emulsion drops from wetting on the collection capillary wall. Capillaries were aligned inside of the square capillary (inner dimension: 1.05 mm × 1.05 mm) whose inner dimensions were larger than the outer diameter of the tapered capillaries. Capillaries were fixed with 5-minute epoxy adhesive.
(2) Preparation of solutions. Outer phase (water + PVA): 2 wt% PVA aqueous. Inner phase (water + PVA + photocatalysts): prepared by mixing 0.01 g ZnO or TiO 2 in 10 mL 2 wt% PVA aqueous, followed by the stirring the solutions for 6 hours.
Middle phase (monomers): the monomer mixture of EGDMA (5 mol% EGDMA) and MAAn (95 mol%) is degassed and 1 mol% Darocur 1173 is added before use.
(3) Microfluidic device assisted PIC preparation. As shown in Fig. 1a andb, the outer phase, the inner phase, and the middle phase are injected into the corresponding inlets, and the double emulsion drops with monodisperse diameters are formed between the tapered inlet capillary tip and the outlet capillary tip (shown in Fig. 1a andc). The drops are exposed to UV light (365 nm, 10 W) (AC 80-240 V, Zhongshan Zigu Lighting Appliance Factory, China). After that, the capsules are immersed in pH = 11 KOH solution for 1 day to be hydrolysed. The obtained samples are collected and washed with deionized water.
The morphologies of the microcapsules, PIC and Janus Pt/Fe/ PIC were characterized by an optical microscope (BX51, Olympus Corporation, Japan) and a field emission scanning electron microscope (FESEM) (Sigma, Zeiss, Germany) at an accelerating voltage of 3 kV. To confirm the hydrolysis of the samples, the particles and particles were analysed by a Fourier transform infrared spectrometer (FT-IR) (Nicolet 6700, ThermoFisher, USA) after drying in vacuum. WAXS profiles were obtained at the beamline BL14B1-XRD of Shanghai Synchronization Radiation Facility (SSRF). The wavelength was 0.1214 nm, the end station was equipped with a Huber5021 diffractometer, and the sample to detector the distance 240 mm was calibrated by lanthanum boride (LaB).
MB was employed as an organic dye contaminate to estimate the adsorptive performance and photocatalytic activity of the synthesized samples. For each experiment, 10 mL (5 mg L -1 ) MB solution was added into a quartz test tube and 0.05 g samples was mixed in the solution. The dark reaction was conducted at room temperature under a gentle stirring. The light reaction was conducted under a 10 W UV lamp (365 nm, 10 W) (AC 80-240 V, Zhongshan Zigu Lighting Appliance Factory, China) with a gentle stirring with a measured power density of 30 mW cm -2 . The MB concentration of every solution was measured by a UV-vis absorption spectrometer (UV-2550, Shimadzu Scientific Instruments, Japan) at intervals of 20 min. The following formula was used to calculate the removal efficiency (η) for CR:
where η is removal efficiency, C 0 and C are the initial MB concentration and the concentration of MB extracted within interval time, respectively. A flow reactor was designed by filling the PIC in a flash chromatographic column. 20 mL (5 mg L -1 ) MB solution was treated by the prepared flow reactor, the concentration before and after the treatment was measured by a UV-vis absorption spectrometer. After that, the flow reactor was exposed to a 10 W UV lamp to regenerate the PIC, and its picture was recorded by a camera (iPhone 8) every 0.5 h.
City. The work performed at Harvard University was supported by the NSF (DMR-1708729) and by the Harvard MRSEC (DMR-1420570).
Published on 02 January 2020. Downloaded by Harvard University on 11/23/2020 8:37:24 PM.View Article Online
This journal is © The Royal Society of Chemistry 2020