Microfluidic Fabrication of Monodisperse and Recyclable TiO2-Poly(ethylene glycol) Diacrylate Hybrid Microgels for Removal of Methylene Blue from Aqueous Medium

Nearly monodisperse titanium oxide–polyethylene glycol diacrylate [TiO2–P(EGDA)] hybrid microbeads containing 0.5 wt % TiO2 nanoparticles entrapped within a P(EGDA) cross-linked polymeric network were synthesized using a modular Lego-inspired glass capillary microfluidic device. TiO2–P(EGDA) hybrid microgels were characterized by optical microscopy, scanning electron microscopy, X-ray diffraction, energy dispersive X-ray spectroscopy, and thermogravimetric analysis. The fabricated TiO2–P(EGDA) hybrid microgel system showed 100% removal efficiency of methylene blue (MB) from its 1–3 ppm aqueous solutions after 4 h of UV light irradiation at 0.2 mW/cm2 at the loading of 25 g/L photocatalyst beads in the reaction mixture, corresponding to the loading of naked TiO2 of just 0.025 g/L. No decrease in photocatalytic efficiency was observed in 10 repeated runs with recycled photocatalyst using a fresh 1 ppm MB solution in each cycle. The rate of photocatalytic degradation was controlled by the UV light irradiance, catalyst loading, and the initial dye concentration. Physical adsorption of MB onto the surface of composite microgel was also observed. The adsorption data was best fitted with the Langmuir adsorption isotherm and the Elovich kinetic model. TiO2–P(EGDA) microgel beads are biocompatible, can be prepared with a tunable size in the microfluidic device, and can easily be separated from the reaction mixture by gravity settling. The TiO2–P(EGDA) system can be used for the removal of other toxic dyes and micropollutants from industrial wastewater.


■ INTRODUCTION
Methylene blue (MB) is a widely used organic dye in industrial processes and can be found in large quantities in textile wastewater.It is highly poisonous and has toxic effects on humans and aquatic life. 1−3 Therefore, various strategies have been developed to remove MB from aqueous solutions. 4,5owever, adsorption and photocatalysis have gained the most attention as sustainable and eco-friendly technologies.Photocatalytic degradation of methylene blue has become a benchmark reaction to test the photocatalytic activity of organic−inorganic hybrid materials because the degradation can be easily monitored by a simple UV−visible spectrophotometric method.Polymeric microparticles and TiO 2 nanoparticles (NPs) have been widely used as adsorbents and photocatalysts, respectively, for the removal of MB from wastewater. 6,7However, adsorption of MB onto polymeric particles is a slow process, and the adsorption capacity is often relatively small. 8On the other hand, the use of naked TiO 2 NPs for photocatalytic degradation of MB is associated with some issues, such as the high aggregation (agglomeration) tendency of TiO 2 NPs in aqueous solutions which results in reduced light absorbance and photocatalytic activity, and challenging separation from the reaction mixture after photocatalysis.−20 For example, Sun et al. 19 developed an aminofunctionalized TiO 2 surface molecularly imprinted polymeric system using sodium lignosulfonate as a functional monomer for highly selective removal of MB via adsorption and photocatalysis methodologies.Idris et al. 21reported a TiO 2 / poly(vinyl alcohol)/cork hybrid material for scavenging of MB using adsorption and photocatalysis strategies.Wei et al. 22 designed TiO 2 −P(EGDA) composite film using photopolymerization of P(EGDA) in the presence of TiO 2 NPs and used the resulting film for adsorptive and photocatalytic removal of Congo red dye from aqueous medium.The loading of TiO 2 NPs into polyethylene glycol diacrylate [P(EGDA)] hydrogel beads is advantageous because P(EGDA) is a hydrophilic, transparent, and biocompatible polymer that can entrap and immobilize TiO 2 NPs while allowing UV light and reactants to pass through.In previously reported studies, TiO 2 NPs have been incorporated only within bulk P(EGDA) hydrogel systems 23 or P(EGDA) films. 22These studies encouraged us to load TiO 2 NPs into P(EGDA) microbeads for adsorption and photocatalytic applications because micro hybrid beads are more efficient adsorbents and photocatalysts than bulk TiO 2 −P(EGDA) hybrid systems.To the best of our knowledge, fabrication of monodisperse TiO 2 -P(EGDA) hybrid microgels using Lego-inspired microfluidic device for adsorptive and photocatalytic removal of MB from water has not yet been reported in the literature.P(EGDA) microgel particles are highly monodisperse and big enough to be easily separated from water by gravity, with a separation efficiency of 100%.So, they can easily be recycled, and their mesh size is large enough for MB to diffuse inside but small enough to prevent loss of TiO 2 NPs.
In the present work, we have used a novel Lego-inspired microfluidic device developed by our group to incorporate TiO 2 NPs into P(EGDA) microgel for the first time.TiO 2 NPs were dispersed in a prepolymer mixture, and the dispersion was emulsified by 3D microfluidic flow focusing, followed by off-chip photopolymerization of generated droplets.TiO 2 -P(EGDA) composite microgel particles have several advantages for adsorption and photocatalysis over bulk composite materials, including higher surface to volume ratio, high UV light transparency, and easy handling.Moreover, microgel particles can be efficiently brought into contact with the reaction mixture and used in various reactor designs, including fluidized bed reactors, packed bed reactors, and stirred tank reactors.
■ EXPERIMENTAL SECTION Materials.P(EGDA) (M w = 700 g/mol), titanium(IV) oxide (mixture of rutile and anatase, Degussa P25, <100 nm particle size (BET), purity 99.5%), and 2-hydroxy-40-(2-hydroxyethoxy)-2-methyl-propiophenone (Irgacure 2959) were purchased from Sigma-Aldrich, UK.MB was obtained from Fisher Bio-Reagents (Fisher Scientific, UK).XIAMETER PMX-200 Silicone Fluid, 100 cSt, and XIAMETER RSN-0749 Resin (50/50 mixture of trimethylsiloxysilicate and cyclomethicone) were received from Dow Chemical, USA.All chemicals were used as such without any further purification.Aqueous solutions were prepared using ultrapure water from a Millipore Milli-Q Plus 185 water purification system.Synthesis of TiO 2 −P(EGDA) Hybrid Microgels.Water−oil (W/O) emulsion droplets containing TiO 2 NPs, P(EGDA), and Irgacure 2959 were generated by 3D counter-current flow focusing in a two-phase Lego-inspired glass capillary device, as shown in Figure 1a. 24The schematic illustration for the photo-cross-linking of P(EGDA) in the presence of TiO 2 NPs is given in Figure 1b.A full description of the microfluidic process and the experimental setup was provided in our previous work. 25Briefly, the dispersed phase was injected through an outer glass capillary into a tapered tip of the coaxially aligned inner capillary.
The dispersed phase jet was broken up into droplets due to jet instability induced by high flow velocity of the continuous phase at the tip of the inner capillary.The continuous phase was supplied countercurrently through the annular gap between the two capillaries.The device was placed on the stage of an inverted biological microscope, which served to control the microfluidic process.
The continuous phase was silicone oil (XIAMETER PMX-200) containing 3.5 wt % RSN-0749 surfactant and the dispersed phase was an aqueous solution containing 98.5 wt % P(EGDA) prepolymer, 1 wt % Igracure 2959, and 0.5 wt % TiO 2 .Pure P(EGDA) beads were prepared using the dispersed phase containing 99 wt % P(EGDA) prepolymer and 1 wt % Igracure 2959.The flow rate of the continuous phase and the dispersed phase was adjusted to 0.8 and 0.16 mL/h, respectively, using Harvard Apparatus model 11 Elite syringe pumps (Biochrom Ltd., UK).Both liquids were delivered from SGE gastight glass syringes (10 mL, Sigma-Aldrich, UK) connected to the device via polyethylene tubing (I.D. 0.86 mm, O.D. 1.52 mm, Smiths Medicals, Luton, UK).The tubing for the delivery of the dispersed phase was covered by aluminum foil to avoid premature polymerization of P(EGDA) by ambient light.The produced droplets were collected in a Petri dish prefilled with the continuous phase and exposed to UVA light (365−395 nm) for 6.5 min from a UVAHAND 250GS lamp (Dr.Honle AG, Gilching, Germany) to cure the droplets.The UV light irradiance was 75 mW/cm 2 , as measured by a radiometer (365 nm, VLX-3W).The beads were washed 5−7 times with acetone and then several times with deionized water to fully remove the oil phase and unreacted species.
Characterization of TiO 2 −P(EGDA) Hybrid Beads.Bright-field microscopy images of W/O emulsion droplets and TiO 2 −P(EGDA) hybrid beads were captured by using a CCD camera (Retiga 6000, Canada) interfaced to a computer running Q-capture software.The surface morphology of TiO 2 −P(EGDA) hybrid beads was examined by a field emission scanning electron microscope (JEOL, JSM 7800F, Tokyo, Japan) with backscatter (BED-C) detector mode operating at an acceleration voltage of 5.0 kV.X-ray diffraction (XRD) analysis was performed by using a D2 PHASER XRD diffractometer (Bruker UK Limited, UK).The TGA 550 instrument (TA Instruments, USA) was used to monitor the weight loss of dried particles of pure P(EGDA) microgel and TiO 2 −P(EGDA) hybrid microgel (12 mg each) under a nitrogen atmosphere in the temperature range of 20− 600 °C at a rate of 10 °C/min.
Photocatalytic Degradation of MB.In photocatalytic experiments, 0.13 g/L hybrid beads were stirred in 20 mL of MB solution (1 or 3 ppm), while the suspension was exposed to UV light at 2, 4, or 8 W/m 2 .In the experiments performed using different catalyst loadings, 100−500 mg of hybrid beads were added into 10 mL of 10 ppm MB solution, and the mixture was irradiated by UV light at 8 W/m 2 .In all experiments, the solution pH was kept at 7, and the residual MB concentration in the liquid phase was measured using a UV−visible spectrophotometer (NanoDrop One/OneC, Thermo Fisher Scientific, USA).The absorption spectra were acquired in the wavelength range 350−850 nm.In the adsorption experiments, 55 mg of hybrid beads were added into 30 mL of the dye solution containing between 0.25 and 7.5 ppm of MB at pH 7, and the mixture was stirred in the dark for a set amount of time.Adsorption isotherms were determined by stirring the suspension for 48 h and measuring the equilibrium MB concentration in the liquid phase.The adsorption kinetics was investigated by monitoring a decrease in the MB concentration with time in the first 18 min of the adsorption process.The control experiments were carried out under similar conditions.The size of the TiO 2 −P(EGDA) hybrid beads was slightly less than that of the P(EGDA) beads.TiO 2 NPs embedded in the polymer matrix act as a physical cross-linker to cause shrinkage of the polymeric network.The average diameter of TiO 2 −P(EGDA) hybrid beads and pure P(EGDA) beads was 238 and 243 μm, respectively.Small dots uniformly distributed within the hybrid beads in Figure 2b  shown in Figure 3b have a spherical shape with occasional wrinkles on the surface.TiO 2 NPs are visible in Figure 3b in the form of dot-like protrusions scattered on the particle surface.Surface wrinkles on the microgel surface occur due to nonuniform droplet polymerization as a result of inhibition of free radicals at the droplet surface by oxygen dissolved in silicone oil.Nonpolymerized PEGDA molecules at the particle surface cause swelling of the cross-linked polymer in the interior of the particle, resulting in stresses that lead to surface buckling and wrinkles. 26oreover, SEM, in combination with energy dispersive Xray spectroscopy (EDX), was used to investigate the distribution of TiO 2 particles within hybrid beads.For this purpose, a single TiO 2 −P(EGDA) bead was cut into two pieces, and its SEM analysis with EDX mapping was  performed.The SEM image of the cut bead is shown in Figure 4a.Bright dots at the particle surface and in the cross section confirm the presence of TiO 2 NPs both in the interior of the beads and on the surface.TiO 2 NPs in the interior of the beads should be accessible to the dye molecules since MB molecules have a rectangular cross section with a surface area of 0.76 × 0.33 nm, 27 and the mesh size of P(EGDA) polymer network in aqueous medium at pH 7 is about 1.1 nm, as calculated based on the end-to-end distance of PEGDA prepolymer chains and the swollen polymer volume fraction.
The EDX elemental mapping in Figure 4b shows that titanium is distributed in all regions on the surface and in the interior of the beads.Green dots of different sizes and shapes suggest that TiO 2 NPs are partially aggregated, which is expected since no dispersant was used to stabilize TiO 2 NPs.The agglomeration of TiO 2 NPs can be reduced by adding a water-soluble surfactant 28 or by coating TiO 2 NPs with polymers such as polyethylene glycol (PEG), polyacrylamide (PAM), and poly(acrylic acid) (PAA) for steric stabilization of TiO 2 sols. 29he EDX spectrum of the TiO 2 −P(EGDA) particle shown in Figure 4c reveals the presence of titanium (Ti) on the bead surface in addition to carbon (C) from the polymer backbone, oxygen (O) from TiO 2 NPs and oxygen-containing groups of P(EGDA) (carbonyl and ether), gold (Au), and palladium (Pd) from the sputter coating process, and silicon (Si) from the traces of silicon-based oil and surfactant remaining on the particle surface after washing.Other elements present in small quantities, such as sulfur and sodium, are impurities from the raw materials.
TGA analysis of P(EGDA) and TiO 2 −P(EGDA) beads was performed to determine their thermal stability and operating temperature ranges for photocatalysis and regeneration.The plots of the normalized sample weight against temperature for both P(EGDA) and TiO 2 −P(EGDA) beads are shown in Figure 5.The weight changes of P(EGDA) as a function of temperature are represented by the black line in Figure 5.A small weight loss in the region from 20 to 150 °C was due to water evaporation from the beads.When temperature was further increased, more significant weight loss was recorded due to chemical degradation of the sample.The maximum rate of weight loss was observed in the temperature range of 320− 380 °C which can be attributed to the degradation of functionalities present in the polymeric network of P(EGDA) and degradation of the main backbone of polymeric chains.The weight changes of TiO 2 −P(EGDA) beads are shown by the red line in Figure 5.A slight weight loss in the range 20− 200 °C was due to removal of water from the beads.Interestingly, TiO 2 −P(EGDA) hybrid beads started to degrade at a higher temperature than pure P(EGDA) beads, and the maximum rate of weight loss due to degradation of polymeric network was also observed at higher temperatures (350−400 °C).Therefore, TiO 2 −P(EGDA) hybrid beads exhibit higher thermal stability than pure P(EGDA) beads, probably due to the higher thermal stability of TiO 2 NPs compared to P(EGDA).The residual weight was nearly the same for both samples due to the small loading of NPs in the polymer matrix.The improved thermal stability of hybrid beads compared to pure polymer beads is in agreement with previously reported TGA results for TiO 2 -polymer nanocomposites 30,31 and could be a useful feature in the case of thermal regeneration of the beads.The presence of TiO 2 NPs in P(EGDA) beads not only  improves their thermal stability but also imparts photocatalytic activity.
The XRD analysis was performed to confirm the entrapment of TiO 2 NPs within P(EGDA) beads (Figure 6).Two broad peaks at 2θ values of 20 and 42.5°were related to P(EGDA).These broad peaks are common in the XRD pattern of both pure and hybrid beads, revealing that P(EGDA) has an amorphous structure in both samples.In the XRD pattern of TiO 2 −P(EGDA) beads, the small peaks at 2θ values of 25.3, 37.7, 49.2, 55.4, 62.9, 69.4,70.6, and 75.6°correspond to the (101), ( 200), ( 211), ( 204), ( 116), (220), and (215) planes of TiO 2 , which is a clear indication of successful loading of TiO 2 NPs into P(EGDA). 32This result was further confirmed by recording an XRD pattern of TiO 2 NPs (gray line in Figure 6).Both samples (TiO 2 NPs and TiO 2 −P(EGDA) beads) have peaks at the same 2θ values.These sharp peaks were missing in the XRD pattern of the pure beads (red line).
Adsorption of MB on TiO 2 −P(EGDA) Hybrid Beads.The adsorption of MB onto TiO 2 −P(EGDA) beads occurs under the influence of van der Waals forces and hydrogen bonding.The adsorption equilibrium was studied by adding 55 mg of the beads into 30 mL of MB solution at 21 °C and pH 7. The initial dye concentration was 0.25, 0.5, 1, 2.5, 5, and 7.5 ppm and the suspension was stirred in the dark until the adsorption equilibrium had been established.The equilibrium MB concentration in the liquid phase was measured based on the height of the absorbance peak at 664 nm.
Four different adsorption isotherm models, Freundlich (FR), Temkin (TM), Dubinin−Radushkevich (DR), and Langmuir (LM), were used to correlate the mass of MB adsorbed per unit mass of beads (q e ) and the equilibrium  concentration of MB in the liquid phase (C e ). 33,34A linear form of the FR isotherm is given by eq 1 where K f and n are Freundlich constants, which can be determined by plotting ln q e vs ln C e , as shown in Figure 7a.
The values of K f and n were calculated from the intercept and slope of the line of best fit, respectively (Table 1).However, since the value of the regression coefficient (R 2 = 0.705) is very low, the adsorption of MB on TiO 2 −P(EGDA) microgel does not follow the FR model.A linear form of the TM adsorption model is given by eq 2 where B T and K T are TM constants, which can be determined by plotting q e vs ln C e , as shown in Figure 7b.The values of B T and K L found from the intercept and slope of the best-fit line are shown in Table 1.However, since R 2 is only 0.508, the TM adsorption model is not applicable here.
A linear form of the DR adsorption isotherm is given by eq 3 where q DR (mg/g) is the theoretical saturation capacity of MB on the adsorbent surface, while ε and β (mol 2 /kJ 2 ) are DR constants.The value of ε was calculated by using the expression, ε = RT ln[1 + 1/C e ], where R is the universal gas constant and T is the absolute temperature.The values of q DR and β were determined from the slope and intercept of the ln q e vs ε 2 graph, as shown in Figure 7c, while the mean free energy per molecule was calculated from the equation, E = [1/ (2β) 1/2 ].The obtained values of q DR , β, and E are given in Table 1.Since R 2 is 0.6458, the adsorption of MB on TiO 2 − P(EGDA) does not follow DR adsorption isotherm modeling, which means that the adsorption of MB onto the hybrid microgel particles is based on layer-by-layer surface coverage rather than pore filling.The LM adsorption model considers monolayer adsorption with no further adsorption after formation of the monolayer because an equilibrium is established.A linear form of the LM model is given by eq 4 where q m is the maximum adsorption capacity of hybrid beads.The values of q m and b were determined from the slope and intercept of the C e /q e vs C e plot shown in 7d and shown in Table 1.A high R 2 value (0.9634) suggests that the adsorption of MB on a hybrid microgel is following the LM adsorption mechanism.Furthermore, the separation factor (R L ) was calculated from the expression, R L = 1/(1 + bC 0 ), where C 0 is the initial dye concentration in the solution.The R L value is used to check the favorability of the adsorption process. 35,36The favorable adsorption is characterized by the value of R L in the range of 0−1. 37,38The R L value for the adsorption of MB on TiO 2 −P(EGDA) beads was found to be 0.084, which confirms a favorable adsorption process in this case.It can be concluded that LM is the best fitted model to describe the adsorption of MB on the TiO 2 −P(EGDA) hybrid beads.−41 Kinetics Models for Adsorption of MB on TiO 2 − P(EGDA) Hybrid Beads.The kinetics of adsorption of MB on TiO 2 −P(EGDA) microgel has been studied using various kinetic models, such as pseudo-first-order, pseudo-secondorder, and Elovich models. 42,43A linear form of the pseudofirst-order and pseudo-second-order kinetic models is given by eqs 5 and 6, respectively ) ln e t 1 e (5) where q e and q t (mg/g) are the mass of MB adsorbed per unit mass of microgel beads at equilibrium and at any time, respectively, k 1 (min −1 ) is the pseudo-first-order rate constant, and k 2 (g•mg −1 •min −1 ) is the second-order rate constant.Equation 5 was tested in Figure 8a by plotting ln(q e − q t ) as a function of time, t.The values of k 1 and q e determined from the slope and intercept of the graph are given in Table 2.
In Figure 8b, t/q t was plotted against time, t using eq 6.The values of k 2 and q e determined from the slope and intercept of the line are given in Table 2.The R 2 values for fitting the pseudo-first and pseudo-second order rate model are 0.912 and 0.682, respectively, confirming that the pseudo-first-order kinetic model is more accurate for the adsorption of MB on the hybrid beads.This finding is in agreement with the behavior observed in earlier studies for the adsorption of MB on polyaniline/graphene oxide or polyaniline/reduced graphene oxide composites. 44oreover, the Elovich chemisorption model was also employed where α and β are the initial adsorption rate constant and the desorption constant, respectively.The fitting of the experimental data to the Elovich kinetic model is shown in Figure 8c.The values of α and β determined from the intercept and slope of the q t vs ln(t) given in Figure 8c  disposed of in large quantities into water bodies from various industries like textiles, paint, leather, food, and pharmaceuticals. 45,46They can be removed from water by chemical oxidation, adsorption, and photocatalysis.Photocatalysis holds several advantages over competing technologies, including complete and fast dye removal without causing secondary pollution, no need for additional chemical agents, and low operating costs since the process can be driven by solar energy. 47The added benefits of using hybrid TiO 2 -filled hydrogel beads for photocatalysis are their tunable size, biocompatibility, and excellent recyclability, due to the large size of microbeads compared to the size of TiO 2 NPs.In this study, the photocatalytic degradation of MB was chosen as a model reaction to demonstrate the photocatalytic activity of the fabricated beads.For this purpose, 20 mL of a 1 ppm MB solution containing 0.13 g/L of TiO 2 −P(EGDA) (CB) was exposed to UV light at 2, 4, and 8 W/m 2 irradiances, and the progress of dye degradation was monitored by spectrophotometric analysis, as shown in Figure 9.During photocatalytic degradation, the central heterocyclic ring of MB is broken and sulfhydryl (−S + �) group is oxidized by hydroxyl radicals ( • OH) to produce a sulfonyl group. 48A steady decrease in the height of the absorption peak at 664 nm was observed at all light intensities, since the absorption wavelength of the sulfonyl group is less than 180 nm.This result shows that TiO 2 − P(EGDA) (CB) hybrid beads are photocatalytically active, even at very low UV light irradiance of 2 W/m 2 .As can be seen from Figure 9, the UV light irradiance had a large impact on the rate of degradation of MB.When the solution was irradiated with 8 W/m 2 , the peak height was reduced from 0.22 to 0.02 in 1.5 h, as shown in Figure 9a.At 4 W/m 2 , the peak height was reduced to the same level in 2 h, as reflected in Figure 9b.When the UV light irradiance was 2 W/ m 2 , it took 5 h for the peak height to be reduced to 0.02, Figure 9c.A strong correlation between the reaction time and the UV irradiance confirms that the decrease in absorbance was due to a photocatalytic reaction and not due to physical adsorption or catalysis.The degradation efficiency of MB is generally smaller when less UV radiation is emitted from the sample since less reactive oxygen species are produced.At the low UV irradiances used in this study, the rate of photocatalytic degradation is usually proportional to the irradiance level. 49At high UV irradiances (>25 W/m 2 ), the dye degradation rate is expected to be independent of the intensity of UV light since a dynamic equilibrium is reached between the reactions leading to electron−hole pair formation and electron−hole recombination.
Three control experiments were performed to confirm that both TiO 2 −P(EGDA) (CB) beads and UV light are needed for the rapid degradation of MB (Figure 10).In the first experiment, 1 ppm of MB solution was irradiated with 8 W/m 2 UV light without adding TiO 2 −P(EGDA) beads.No decrease in absorption peak was observed within 6 h, confirming that UV light cannot degrade MB without hybrid beads, as given in Figure 10a.In the next experiment, hybrid beads were added to 1 ppm MB solution, and the reaction mixture was stirred in the dark.No prominent decrease of absorbance peak was observed, as shown in Figure 10b, and a small decline in absorbance from 0.24 to 0.20 after 6 h can be attributed to the Table 2. Kinetic Parameters Determined from the Pseudo-First-Order, Pseudo-Second-Order, and Elovich Models for the Adsorption of MB onto TiO 2 −P(EGDA) Hybrid Beads a Pseudo-first-order kinetic parameters k 1 (min −1 ) q e (mg•g −1 ) R 2 (�) 0.132 2.98 0.912 Pseudo-second-order kinetic parameters k 2 (g•mg −1 •min −1 ) q e (mg•g −1 ) R 2 (/) 3.5 0.0158 0.682 Elovich model parameters α (mg•g −1 •min −1 ) β (g•mg −1 ) R 2 (�) 0.0881 11.5 0.955 a The coefficients of determination, R 2 , are also provided to estimate the goodness of fit in each case.
dye adsorption on the beads surface.When the same MB solution was exposed to UV light at the same irradiance (8 W/ m 2 ) and catalyst loading (25 g/L), the absorbance decreased from 0.22 to 0.02 in 1.5 h, Figure 9a.Therefore, both hybrid beads and UV light are needed for the photocatalytic degradation of MB.
In the third experiment, P(EGDA) beads were added to 1 ppm of MB solution, and the dye removal was monitored over 6 h under the UV light irradiation of 8 W/m 2 .As shown in Figure 10c, again a very small decrease in the absorbance peak occurred, which can be attributed to the physical adsorption of MB to P(EGDA) beads.Interestingly, the absorbance decreased by a smaller amount in the presence of pure polymer beads than TiO 2 −P(EGDA) (CB) beads, meaning that the dye adsorption was enhanced by entrapped TiO 2 NPs.This experiment confirmed that P(EGDA) microgel particles cannot serve as photocatalysts for the degradation of MB, unless TiO 2 NPs are entrapped into the P(EGDA) matrix.
The reaction progress was also monitored by plotting C t /C 0 against time for the experiments carried out at different light intensities, initial concentrations of MB, and catalyst loadings, Figure 11a,b.As a comparison, the results of three control experiments are also plotted in Figure 11a.
The C t /C 0 vs time plots for the degradation of MB from 1 and 3 ppm solutions at different UV light irradiances at the constant loading of catalytic beads (25 g/L) are shown in Figure 11a.The results obtained with pure TiO 2 NPs are also included in this figure.In this case, 0.13 g/L of pure TiO 2 was added to 1 ppm MB solution and the sample was exposed to UV light at 8 W/m 2 .It should be noted that 0.13 g/L pure TiO 2 is equivalent to 25 g/L catalytic beads since the content of TiO 2 in the catalytic beads was 0.5 wt %.After 0.5 h, the catalytic beads were more efficient than pure TiO 2 and degraded 61% of MB, while pure TiO 2 NPs removed only 40% of the dye.However, pure TiO 2 NPs were more efficient over longer times and degraded 85% of MB after 1 h and 100% of MB after 1.5 h.On the other hand, catalytic beads degraded 78% of MB after 1 h, 93% after 1.5 h, and 100% after 2 h.A relatively small difference in performance between pure TiO 2 and TiO 2 -loaded beads shows that the catalytic beads exhibit small internal diffusional resistances and high transmittance of UV light.
Except for irradiation at 2 W/m 2 , the MB degradation efficiency was almost 100% after 4 h of irradiation.Rauf et al. 50ave achieved degradation efficiency of MB of 70% from 1.6 ppm solution after 4 h of irradiation using 0.64 g/L photocatalyst (Cr−Ti binary oxide with 10 mol % Cr 3+ ).Although the TiO 2 loading in this study was much smaller (0.025 g/L), the dye degradation efficiency was significantly higher.
By increasing the UV light irradiance, the degradation rate of MB increases since more electron−hole pairs are generated in TiO 2 NPs.The C t /C 0 vs time graph for photocatalytic degradation of a 10 ppm MB solution using different catalyst loadings is given in Figure 11b.The rate of photocatalytic degradation increases by increasing the amount of TiO 2 − P(EGDA) beads added to the reaction mixture, and the reaction can be completed faster.The Langmuir−Hinshelwood (LH) kinetic model is widely accepted to describe the kinetics of photocatalytic degradation of MB. 1 The LH mechanism involves the coadsorption of MB and H 2 O molecules on the catalyst surface, followed by a surface reaction between the adsorbed water molecules and positive holes in the valence band, leading to the generation of hydroxyl radicals.The dissolved oxygen can also be adsorbed on the catalyst surface and react with electrons in the conduction band to form a superoxide anion (O 2 •− ).The formation of O 2 •− triggers a series of reactions leading to the formation of various other reactive oxygen species, such as hydrogen peroxide (H 2 O 2 ), hydroperoxyl radical (HO 2 •− ), and hydroxyl radical ( • OH). 51inally, the released reactive oxygen species react with the adsorbed MB molecules, causing their breakdown into smaller molecules. 48According to the LH model, the greater the catalyst surface area, the more adsorption sites there are and the faster the reaction, as shown in Figure 11b.
Initially, the reaction rate strongly depends on the catalyst loading since the rate-limiting factor at low catalyst loadings is the availability of active sites on the catalyst surface.As the catalyst loading increases, this effect weakens and the ratelimiting factor becomes the rate of surface reaction, as can be seen by the small difference in the reaction rate when 20 and Langmuir 25 g/L beads were added.Further increase in catalyst loading might have a negative impact on the reaction rate due to light scattering by the beads, leading to poor light utilization. 52ithout TiO 2 NPs in the polymer matrix or without UV light, the removal efficiency of MB was only 13 and 16%, respectively, as compared to ∼100% in the presence of photocatalytic beads and UV light, Figure 11b.Also, when no particles were added to the reaction mixture, the removal efficiency was zero, irrespective of the irradiation time and UV light irradiance.
The reaction kinetics was investigated using the pseudo-firstorder model based on eq 8 i k j j j j j y A linear portion of the ln(C t /C 0 ) vs time plots shown in Figure 12a was used to calculate the apparent rate constant (k app ) for photocatalytic degradation of MB under different conditions, and the obtained k app values are listed in Table 3.
As shown in Figure 12b, the apparent rate constant increased linearly with increasing catalyst loading, reflecting the fact that the number of active adsorption sites on the catalyst surface is a linear function of the catalyst loading.Also, at constant UV light irradiance (8 W/m 2 ) and the beads loading in the solution (25 g/L), k app decreased from 2.52 to 0.819 h −1 when the initial MB concentration increased from 1 to 10 ppm (Table 3).Due to increased adsorption of MB molecules on the catalyst surface at the higher dye concentration, a smaller fraction of active sites was available for the adsorption of water molecules, and a smaller number of • OH radicals were produced. 52Also, as the dye concentration increases, more and more UV light is absorbed by the dye molecules dissolved in the liquid phase and adsorbed onto the catalyst surface, which can prevent photons from reaching the surface of TiO 2 NPs. 53omparison of Photocatalytic Activity of TiO 2 − P(EGDA) with Previous Studies.Usually, TiO 2 NPs are coated on the surface of a support (polystyrene, 54 Si@Fe, 55 reduced graphene oxide) 56 for photocatalytic degradation of MB due to which the photocatalytic system does not remain stable for their reuse in large number of cycles.Moreover, high light irradiance is needed for the complete degradation of MB.A comparison of photocatalytic activity of TiO 2 −P(EGDA) with work in terms of photocatalytic activity and recyclability (removal efficiency at different cycles) reported in the literature is presented in Table 4.
A catalyst recyclability study was also performed by repeating degradation experiments with 1 ppm of MB solution 10 times under 8, 4, and 2 W/m 2 UV irradiance using the same photocatalyst.After each cycle, the photocatalyst was separated from the reaction mixture by gravity settling, which took only 10 s.After separation, the photocatalytic beads were regenerated by washing 5−7 times with pure acetone, followed by 3−5 washes with deionized water.The highest removal efficiency of MB (nearly 100%) was achieved at the highest irradiance level (8 W/m 2 ).It should be noted that the removal efficiency at smaller UV irradiances would eventually reach 100% if the time of UV exposure in each cycle was longer.Moreover, no loss in photocatalytic efficiency of TiO 2 − P(EGDA) (CB) beads was observed after 10 cycles, indicating that TiO 2 NPs were neither released from the polymer matrix nor deactivated.Furthermore, a slight increase in photocatalytic activity occurred in the first several cycles, probably due to complete removal of surfactant molecules from the catalyst surface.As shown in Figure 4, the silicon surfactant was not fully removed from the particle surfaces during washing.Therefore, some active sites on the catalyst surface were covered by surfactant molecules and unavailable to the reactant molecules.During UV irradiation, the surfactant molecules were degraded by the photocatalyst just like the dye and, consequently, the removal efficiency was higher in the subsequent cycles.At the smaller light intensities, more cycles were needed to fully remove the surfactant molecules from the interface.A steady-state removal efficiency of MB under the experimental conditions shown in Figure 13 varied from 72% at 2 W/m 2 to 92% at 8 W/m 2 .

■ CONCLUSIONS
Nearly monodisperse P(EGDA) microgel beads and TiO 2 − P(EGDA) hybrid beads were successfully prepared by 3D   hydrodynamic flow focusing using a Lego-inspired glass capillary microfluidic device.The synthesized hybrid microgel beads were perfectly spherical in shape, with a smooth surface, and dot-like protrusions from embedded TiO 2 NPs.The hybrid microgel beads showed excellent photocatalytic activity for the removal of a trace amount of methylene blue (1−10 ppm) from aqueous solutions.The adsorption data were found to be best fitted with the Langmuir adsorption isotherm.P(EGDA) not only stabilizes TiO 2 NPs but also enhances their recyclability and photocatalytic ability.The mesh size of the P(EGDA) polymer network was large enough to allow diffusion of MB molecules to the surface of embedded TiO 2 NPs, which allowed for the dye removal efficiency of nearly 100% from 1 and 3 ppm solutions.Photocatalytic activity of the hybrid system could be controlled by varying the reaction conditions, such as UV light irradiance and the amount of TiO 2 −P(EGDA) beads in the reaction mixture.The photocatalytic microgel beads could be recycled and used in multiple cycles without any loss of their photocatalytic efficiency.The developed hybrid beads can be used in the future for photocatalytic degradation of other dyes and other persistent organic pollutants.The microfluidic device can be cheaply fabricated by computer numerical control (CNC) milling and is easy to assemble, operate, and dismantle for cleaning.The size of the microgel beads can be precisely controlled by hydrodynamic conditions and flow geometry in the microfluidic device, which allows for precise tuning of the photocatalytic performance of the beads and their behavior in the reactor, such as settling time and flow properties.

Figure 1 .
Figure 1.(a) Microfluidic fabrication of TiO 2 −P(EGDA) hybrid microbeads for the photocatalytic removal of methylene blue from aqueous solution.(b) Chemistry of photo-cross-linking of P(EGDA) in the presence of TiO 2 NPs.

■
RESULTS AND DISCUSSION Fabrication and Characterization of TiO 2 −P(EGDA) Hybrid Beads.Microscopic images of P(EGDA) beads and TiO 2 −P(EGDA) hybrid beads are shown in Figure 2a,b respectively.Both P(EGDA) beads and TiO 2 −P(EGDA) hybrid beads are nearly monodisperse, as can be seen by their hexagonal arrangement on a microscope slide, and have a spherical shape without any fused particles.
indicate the presence of embedded TiO 2 NPs.The morphological investigations of TiO 2 −P(EGDA) hybrid beads were carried out by SEM.The SEM images of TiO 2 −P(EGDA) hybrid beads at two different magnifications are shown in Figure 3a,b.TiO 2 −P(EGDA) hybrid beads

Figure 2 .
Figure 2. Optical microscopic images of (a) pure P(EGDA) beads and (b) TiO 2 −P(EGDA) hybrid beads.The same scale bar applies to both images.

Figure 3 .
Figure 3. SEM images of TiO 2 −P(EGDA) beads at different magnifications, indicating their size, shape, and surface morphology.The scale bar is 100 μm in (a) and 10 μm in (b).

Figure 4 .
Figure 4. (a) SEM image of a single TiO 2 −P(EGDA) bead cut with a knife to expose TiO 2 NPs inside the bead; (b) EDX mapping of titanium distribution within the bead; (c) EDX spectrum of a single hybrid bead.

Figure 8 .
Figure 8. Fitting various kinetic models to the experimental data for the adsorption of MB upon the surface of TiO 2 −P(EGDA) beads.(a) Pseudo-first-order model; (b) Pseudo-second-order model; (c) Elovich model.In each case, 55 mg of the beads were added into 30 mL of a stirred MB solution (1 ppm).

Figure 9 .
Figure 9. UV−visible absorption spectra of MB solution as a function of time in the presence of TiO 2 −P(EGDA) (CB) hybrid beads at (a) 8; (b) 4; and (c) 2 W/m 2 UV light irradiance.In each experiment, 0.13 g/L TiO 2 −P(EGDA) beads were added into 20 mL of well-stirred MB solution (1 ppm).

Figure 10 .
Figure 10.UV−visible absorption spectra of 1.0 ppm of MB solution as a function of time (a) under UV light (8 W/m 2 ) but without photocatalyst; (b) in the presence of TiO 2 −P(EGDA) catalytic beads (25 g/L) (CB) in the dark; and (c) in the presence of P(EGDA) beads (25 g/L) under UV light (8 W/m 2 ).

Figure 11 .
Figure 11.(a) C t /C 0 vs time curves for photocatalytic degradation of 1 and 3 ppm MB solution under different reaction conditions; (b) C t /C 0 vs time curves for photocatalytic degradation of 10 ppm MB solution under 8 W/m 2 UV irradiation at different loadings of TiO 2 −P(EGDA) (CB) hybrid beads.

Figure 12 .
Figure 12.(a) Plots of ln(C t /C 0 ) vs time for the photocatalytic degradation of MB solution of different initial concentrations under different UV light intensities; (b) apparent rate constant (k app ) for photocatalytic degradation of 10 ppm MB solution under 8 W/m 2 UV irradiation in the presence of different amounts of TiO 2 −P(EGDA) catalytic beads.

Table 1 .
Adsorption Parameters for MB Adsorption on TiO 2 −P(EGDA) Hybrid Beads Were Calculated from Langmuir, Freundlich, Dubinin−Radushkevich, and Temkin Models are shown in Table2.Based on the R 2 values, the Elovich model can be considered as the most accurate model for the prediction of the adsorption kinetics of MB on the TiO 2 −P(EGDA) microgel.Photocatalytic Activity of TiO 2 −P(EGDA) Hybrid Beads.Dyes are important industrial chemicals, which are

Table 3 .
Apparent Rate Constant (k app ) for the Photocatalytic Degradation of MB at Different Loadings of Catalytic Beads in the Dye Solution, UV Light Irradiances, and Initial Dye Concentrations