Tetracycline, an artificial synthetic drug, has been widely used in livestock farming and aquaculture for disease treatment. At the same time, pharmaceutical industrial wastewater discharged to nature has caused environmental problems. As one of the most environmentally friendly, low cost, and promising technologies, photocatalysis has been applied to the treatment of such wastewater.
Stimuli-responsive photocatalysts, namely "intelligent" photocatalysts, have received increasing interest in recent years [1]. Using thermo-responsive polymers to load semiconductors or combine them with noble metal particles to prepare "intelligent" photocatalysts can allow their photocatalytic activity to be switched on and off by adjusting the temperature of the surrounding environment below or above the lower critical solution temperature [2, 3].
Poly N-isopropylacrylamide (PNIPAM) is one of the most common thermo-sensitive polymers [4]. PNIPAM molecular chains show hydrophilicity and become highly swollen at temperatures lower than the lower critical solution temperature (LCST, 32 ℃), which arises from the hydrogen bonding interactions between the amide groups of PNIPAM and water molecules. Conversely, with increasing temperature, the PNIPAM molecular chains shrink rapidly as these hydrogen bonds are destroyed, resulting in hydrophobicity [5, 6]. With the rapid development of temperature sensitive materials, many studies have been carried out to exploit the thermo-responsivity of PNIPAM, such as the investigation of Au-Pt/thermo-responsive copolymer microgels at the oil/water interface [7, 8] and thermo-sensitive core/shell microgels with carbon cores [9], which have proved that PNIPAM exhibits effective intelligent control performance.
However, the application of the special properties of PNIPAM to the field of photocatalysis has rarely been reported [10-12]. Numerous semiconductor photocatalysts have been widely investigated for the removal of organic contaminants. Among these semiconductors, g-C3N4 has been systematically and deeply investigated owing to its stable chemical structure, nontoxicity, and low cost [13-15]. However, there are still many issues which have greatly limited the further development and application of g-C3N4 in the field of photocatalysis [16]. For example, g-C3N4 is not controllable, is difficult to recover from the solution for reuse [17], and has low visible-light efficiency [18]. To overcome the above-mentioned disadvantages, in this work we introduced PNIPAM and Fe3O4 to the surface of g-C3N4 using a simple and facile method.
Currently, Fe3O4 has been paid much attention owing to its unique physical chemical properties, and its use has spread into many areas, including drug delivery systems, biological medicine [19, 20], sewage treatment [21], and even photocatalysis [22]. Particularly, great progress has been made in its application to photocatalysis. In previous studies, many researchers have coupled Fe3O4 with other semiconductors, producing WO3/Fe3O4[23], TiO2/Fe3O4[24], and even graphene-Fe3O4[25]. These composite photocatalytic materials have shown ideal performance. For this reason, we also introduced magnetic Fe3O4 to the surface of g-C3N4. On one hand, Fe3O4 can be considered to be a better conductor because its conductivity is about 1.9 × 106S/m[26]. Therefore, Fe3O4 can rapidly transmit electrons and decrease the recombination rate of photogenerated electrons and holes, which significantly enhances photocatalytic activity. On the other hand, the favorable magnetic characteristics of Fe3O4 makes it easy to separate from a reaction system using an external magnet [27], which improves the recycling efficiency [28]. Therefore, as a conductive magnetic substance, Fe3O4 will not only enhance the recycling rate, but also the photocatalytic activity of the intelligent material.
With the above in mind, we report the design and fabrication of an intelligent recyclable photocatalyst in this paper. We introduced thermo-sensitive PNIPAM to the surface of Fe3O4/g-C3N4, the photocatalytic activity of which can be externally modulated as shown in Fig. 1. At temperatures lower than 32 ℃, the thermo-sensitive polymer is swollen. At temperatures higher than 32 ℃, the thermo-sensitive polymer dehydrates to the collapsed state owing to the breakdown of the delicate hydrophilic/hydrophobic balance in the spherical structure distances [29]. The findings reported here for the novel PNIPAM/Fe3O4/g-C3N4composite are critical and necessary inputs in the development of photocatalytic degradation/purification processes from both academic and industrial points of view. We also expect that the modification of photocatalytic materials with thermo-sensitive materials will enable the intelligent removal of organic pollutants in waste water.
The fabricated intelligent composite PNIPAM/Fe3O4/g-C3N4 photocatalyst was characterized by a series of methods including X-ray diffraction (XRD), Fourier-transformed infrared spectroscopy (FT-IR), transmission electron microscopy (TEM), Raman spectroscopy, UV-visible spectroscopy (UV-vis), elemental analysis, thermo-gravimetric analysis (TGA), and vibrating sample magnetometry (VSM). Photoluminescence (PL) spectra and the transient photocurrent responses were also measured. Finally, the degradation mechanism and intermediate products of tetracycline were also determined by mass spectrometry (MS).
Melamine powder (99.0%), CH3COOHNa·3H2O (98.0%), polyethylene glycol 1500 (PEG, 1500), and ethylene glycol (98.0%) were supplied by Aladdin Chemistry Co. Ltd. (Shanghai, China) Azobisisobutyronitrile (AIBN, CP), N- isopropylacrylamide (NIPAM, 98.0%), N, N'-methylenebis (acrylamide) (MBAA, 97.0%) were purchased from Sinopharm Chemical Reagent Co., Ltd. (China). Fe (NO3)3·9H2O (98.0%), ethanol (95.0%), tetracycline, and all other chemicals used in the experiments were purchased from Shanghai Chemical Reagent Co., Ltd. (China). Distilled water was used throughout.
Pure g-C3N4was synthesized by annealing melamine in a muffle furnace. Briefly, 2.0 g melamine was placed in an open crucible, heated to 500 ℃ from room temperature and held for 2 h, then heated to 550 ℃ by a ramping rate of 2.3 ℃/min and held for another 2 h. After the alumina crucible cooled to room temperature, the as-prepared g-C3N4 was ground into powder for further use.
Fe3O4/g-C3N4was synthesized as follows: Firstly, 0.5 g g-C3N4 powder was added to 30 mL ethanediol containing 0.18 g Fe (NO3)3·9H2O, 0.25 g C2H3O2Na·3H2O, 0.05 g PEG, and 0.005 g PVP, and ultrasonically dispersed for 1 h. The suspension was then transferred into a 50 mL Teflon-lined stainless steel autoclave and heated to 200 ℃ for 15 h. Finally, the Fe3O4/g-C3N4 product was obtained
PNIPAM/Fe3O4/g-C3N4 was synthesized as follows: Appropriate amounts of Fe3O4/g-C3N4and NIPAM were added to 100 mL of a distilled water/methanol mixture (25/75, V/V). The mixture was sonicated for 10 min, and 0.01 g MBAA and 0.005 g AIBN were then quickly added. The polymerization was conducted at 65 ℃ under the nitrogen protection for 5 h. After completion of the reaction, the suspension was separated with a magnet and then dried.
XRD patterns of the samples were recorded at room temperature using a powder X-ray diffractometer (Rigaku, Japan) with Ni-filtrated Cu Kα radiation (40 kV, 200 mA) at 10°-80° with a scanning step of 0.02°/0.2 s. FT-IR spectra were obtained on a Nicolet Magna-IR 550 within the wavelength range of 500-4000 cm-1. KBr was used as the reference sample. TEM images were obtained with a JEM-2100 transmission electron microscopy (JEOL, Japan). The magnetic properties of the samples were measured with a VSM (HH-15, Jiangsu University, China). TGA was carried out using a thermal analyzer (NETZSCH-Gerȁtebau GmbH, Germany) in air, from room temperature to 800 ℃ with heating rate of 5 ℃/min. Raman experiments were performed using the 532 nm laser of a DXR spectrometer at room temperature. The elemental content of each sample was analyzed using an elemental analyzer (EA-1112A, Thermos, Italy). The PL spectra of the as-prepared samples were investigated using a Cary Eclipse Spectrophotometer (VARIAN, USA) equipped with xenon lamp at an excitation wavelength of 343 nm. UV-vis diffuse reflectance spectra (DRS) were obtained at room temperature for dry-pressed disk samples using a 2450 spectrometer (Shimadzu Co., Japan) equipped with an integrated sphere accessory, and using BaSO4 as the background. The degradation intermediate products of aqueous tetracycline solution were determined using a Thermo LXQ ion trap mass spectrometer (MS).
To investigate the adsorption capability of the photocatalysts, adsorption experiments were carried out in a GHX-2 photochemical reactor (Yangzhou University Technology Ltd, Yangzhou, China). The reactor consisted of a quartz reactor and an illumination source. The quartz reactor was connected to a DFY-5 low temperature cooling liquid reaction bath (Gongyi City Yuhua Instrument Co., Ltd. Gongyi, China) to keep the solution at a certain temperature. 0.1 g photocatalyst was added to 100 mL 20 mg/Ltetracycline solution in the photochemical reactor and stirred for 70 min in the dark at 20 ℃ or 45 ℃, during which the absorbance of the tetracycline was monitored every 10 min.
The photocatalytic activity of the samples was evaluated using the decomposition of tetracycline under visible light irradiation (300 W Xe lamp). The photo-degradation was performed as follows: the binding experiments were carried out at 20 ℃ and 45 ℃. The tetracycline solution (20 mg/L, 100 mL) containing 0.1 g photocatalyst was ultrasonicated for 5 min and kept in the dark for 1 h under continuous magnetic stirring to achieve adsorption-desorption equilibrium between the tetracycline and photocatalyst. At 20 min intervals, 6 mL of the reaction suspension was sampled and magnetic separated, and the absorbance of the tetracycline was measured. Additionally, radical capture experiments were carried out by adding 0.1 mmol triethanolamine (TEOA, a quencher of h+), 0.1 mmol isopropanol (IPA, a quencher of •OH), or 0.1 mmol benzoquinone (BQ, a quencher of •O2-).
The concentration of tetracycline in reaction solutions were determined by UV-vis spectroscopy at a wavelength of 357 nm. The photocatalytic efficiency of tetracycline degradation was calculated using the following formula: η=(C0-Ct)/C0, where η is photocatalytic efficiency, C0 is the concentration of reactant before illumination, and Ctis the reactant concentration after illumination time t.
Fig. 2 shows the XRD patterns of the as-prepared g-C3N4, Fe3O4/g-C3N4, and PNIPAM/Fe3O4/g-C3N4. Each sample exhibited an intense diffraction peak at 2θ=27.4°, corresponding to the (002) interplanar stacking structure of crystalline graphite [30]. The XRD pattern of Fe3O4/g-C3N4 exhibited the characteristic peaks of Fe3O4 at 2θ=30.2°, 35.5°, 43.2°, 53.4°, 57.3° and 62.6°[28]. The diffraction peak at 27.4° was slightly lower in intensity than that of the pure g-C3N4, which resulted from the Fe3O4 restraining the stacking of g-C3N4 perpendicular to the (002) direction. The peak intensity of Fe3O4 (35.5°) was remarkably weakened in the diffraction pattern of PNIPAM/Fe3O4/g-C3N4. This result implies that PNIPAM was formed on the surface of the Fe3O4/g-C3N4, and further indicated the successful formation of PNIPAM/Fe3O4/g-C3N4 ternary composites. The presence of PNIPAM was subsequently confirmed by FT-IR, EDS, TEM, and elemental analysis.
The functional groups of pure g-C3N4, Fe3O4/g-C3N4, and PNIPAM/Fe3O4/g-C3N4 were analyzed by FT-IR spectroscopy. As shown in Fig. 3, all the materials demonstrated a series of multiple bands. Those located in the 1200-1600 cm-1 region (1250, 1329, 1419, 1577, and 1635 cm-1) corresponded to the typical stretching modes of CN heterocycles [31], such as C-N and C=N stretching [32]. The sharp band at 808 cm-1 was assigned to the characteristic typical bending vibration of tri-s-triazine units [33]. The broad peaks from 3400 to 2800 cm-1 arose from the stretching vibrational modes of primary amines. In the FT-IR spectrum of Fe3O4/g-C3N4, the absorption peaks of g-C3N4 become slightly weaker, which indicated that the hydrothermal treatment had no effect on the chemical structure of g-C3N4. The absorption peaks of PNIPAM/Fe3O4/g-C3N4 indicated that the g-C3N4 retained the same chemical structure as pure g-C3N4 after coating with PNIPAM. Peaks at 2969, 1657, and 1390 cm-1 from C-H, O=C-NH2, and-CH (CH3)2, respectively, could be seen clearly, and were caused by introducing PNIPAM to the Fe3O4/g-C3N4 surface [34]. In brief, it could be inferred from the FT-IR results, that the incorporation and existence of Fe3O4 and PNIPAM in the composite would not affect the chemical functionality of g-C3N4.
In the Raman spectra shown in Fig. 4(a), peaks located at 468 cm-1, 974 cm-1, 1248 cm-1, and 1485 cm-1 that stemmed from the vibration modes of CN were clearly seen for bothFe3O4/g-C3N4(1) and PNIPAM/Fe3O4/g-C3N4(2). The prominent peak at 684 cm-1 represents the symmetric stretch of oxygen atoms along the Fe-O bond, which indicates that the Fe3O4 was successfully loaded on the surface of the g-C3N4 [22]. A series of modes could be detected in the amplified Raman spectrum of PNIPAM/Fe3O4/g-C3N4 in Fig. 4(b), including CH3 twisting (838 cm-1), CH3 rocking (tert-butyl, 918 cm-1), C-C skeletal in-phase stretching (1106 cm-1), N-C bond stretching (1146 cm-1), CH3 bending (scissor deformation, 1449 cm-1) amide 2 N-H deformation (1544 cm-1), and amide 1 C=O stretching (1646 cm-1) [35]. These results confirmed the existence of g-C3N4, Fe3O4, and PNIPAM in the composite.
The morphology and microstructure of the prepared photocatalysts were characterized by TEM and EDS. As shown in Fig. 5(a), g-C3N4 exhibited a stacked pimple-like structure and consisted of aggregates of several micrometers in size resulting from the thermal polycondensation of melamine. Fig. 5(b) shows that in Fe3O4/g-C3N4, the g-C3N4 sheets served as a support and surfactant to bind with the Fe3O4, which was mainly present as 3D microspheres with an average diameter of 0.05-0.1 μm. After hybridization with the pure g-C3N4, the Fe3O4 microspheres were well interwoven with the g-C3N4, forming a 3D hybridized structure expected to facilitate the transfer of photogenerated carriers and improve the photocatalytic performance. In the image of PNIPAM/Fe3O4/g-C3N4 displayed in Fig. 5(c), a thin layer appeared to be assembled on the surface of the Fe3O4/g-C3N4, and the Fe3O4 3D microspheres could not be seen clearly. This change from the morphology seen in Fig. 5(b) must have been caused by the covering of the material with PNIPAM. Therefore, the TEM image further confirmed the existence of PNIPAM in the composite. Accordingly, the EDS results obtained for the different samples were also different. As shown in Fig. 5(a1), the EDS spectrum of the pure g-C3N4 samples revealed the existence of C and N elements. For Fe3O4/g-C3N4 (Fig. 5(b1)), besides N and C elements, Fe and O elements were also detected, indicating that the samples consisted of Fe3O4 and g-C3N4. Compared with those of Fe3O4/g-C3N4, the C and N intensities of the PNIPAM/Fe3O4/g-C3N4 were enhanced, while the Fe intensity was decreased. These results further indicated that thermo-sensitive PNIPAM was successfully introduced to the composite.
The elemental contents of the different samples are shown in Table 1. The C/N molar ratio in g-C3N4 was nearly 3/4, close to the theoretical value. This further proved that g-C3N4 was successfully prepared [36]. Furthermore, the contents of N, C, and H in Fe3O4/g-C3N4 were decreased compared with those in pure g-C3N4. The content of Fe3O4 was nearly 20%, which was consistent with the initial dosage, meaning that Fe3O4 was successfully attached to the surface of the g-C3N4. Moreover, the contents of N, C, and H in PNIPAM/Fe3O4/g-C3N4 were significantly higher than those in Fe3O4/g-C3N4, indicating that the surface of Fe3O4/g-C3N4 was successfully modified with PNIPAM.
Fig. 6 displays the TG and DSC curves of PNIPAM/Fe3O4/g-C3N4. The TG curve of PNIPAM/Fe3O4/g-C3N4 exhibited a slight mass loss of about 4.31% from room temperature up to 200 ℃. This mass loss may have been caused by the evaporation of water [37]. As the temperature was increased to 300 ℃, a further mass loss of 10.09% was obtained, which may have been caused by the loss of PNIPAM from the surface of the Fe3O4/g-C3N4. Finally, the mass of the sample decreased rapidly in the temperature range from 450 ℃ to 600 ℃, indicating that 71.02% of the sample was g-C3N4, which underwent combustion in the air [38, 39]. The remaining mass (15.0%) was attributed to Fe3O4, because the melting point of Fe3O4 is much higher than the experimental temperature. From the DSC curves, it could be seen the thermal decomposition of PNIPAM/Fe3O4/g-C3N4occurred at 300 and 550 ℃. Hence, this analysis demonstrated that PNIPAM/Fe3O4/g-C3N4 had good thermal stability and that its composition was consistent with the original materials.
The magnetic properties of PNIPAM/Fe3O4/g-C3N4 were measured by sweeping an applied field over the sample from 10 to 10000 kOe at room temperature. Fig. 7(a) shows the hysteresis loop of the as-prepared sample, the magnetization saturation value of which reached 13.80 emu/g. Good superparamagnetism is essential for the collection and reuse of the photocatalyst in multiple cycles. The inset of Fig. 7(a) shows that the photocatalyst was strongly attracted to a permanent magnet. Therefore, the PNIPAM/Fe3O4/g-C3N4 photocatalyst could be rapidly separated from suspension under an applied magnetic field. To further examine the stability of the as-prepared intelligent PNIPAM/Fe3O4/g-C3N4, its reusability was tested in five successive tetracycline degradation experiments. As shown in Fig. 7(b), the as-prepared photocatalyst retained over 80% of its original photocatalytic activity after five successive experimental runs, which is very important for practical application. Therefore, the PNIPAM/Fe3O4/g-C3N4 can be used efficiently in environmental protection.
PL measurements and photocurrent-time tests have been widely used to evaluate the photocatalytic quantum efficiency of photocatalysts [40, 41]. As shown in Fig. 8(a), pure g-C3N4 exhibited luminescence in a broad range (400-600 nm) centered at 450 nm at room temperature with an excitation wavelength of 330 nm. This finding suggested that photoinduced e--h+ pairs were generated and recombined within the pure g-C3N4. The PL intensity of PNIPAM/Fe3O4/g-C3N4 was obviously decreased compared with that of the pure g-C3N4, suggesting a decreased recombination rate of photogenerated e--h+pairs in PNIPAM/Fe3O4/g-C3N4. The above results indicated that efficient separation, transport, and capture of the photo-generated carriers were realized after the formation of Fe3O4 and introduction of PNIPAM to the composite. The photocurrent-time (I-t) curves are displayed in Fig. 8(b). The generated photocurrents were reproducible and stable during three intermittent on-off irradiation cycles. The prompt increase in the photocurrent response from the light-off to the light-on state was ascribed to the quick separation and transport of photo-generated electrons. PNIPAM/Fe3O4/g-C3N4 exhibited a higher photocurrent response than pure g-C3N4. The results of the photo-electrochemical experiments were consistent with the PL measurements, further indicating that PNIPAM/Fe3O4/g-C3N4 effectively slowed the combination rate of photo-generated e--h+ pairs.
The UV-vis diffuse reflectance spectra of Fe3O4, pure g-C3N4, and PNIPAM/Fe3O4/g-C3N4 are shown in Fig. 8(c). The pure g-C3N4 exhibited photo-absorption from the UV light range to visible light shorter than 460 nm. Notably, PNIPAM/Fe3O4/g-C3N4 exhibited a red shift in its absorbance edge up to 800 nm, indicating a smaller band gap, and its absorption intensity in the visible light region was higher than that of pure g-C3N4. The prominent red shift of the composite photocatalyst compared with pure g-C3N4 can be attributed to the presence of PNIPAM and Fe3O4 on the g-C3N4surface, which improved its optical absorption [42-44]. Firstly, the good conductivity of Fe3O4 effectively promoted the separation of electron-hole pairs, which accounts for the band gap transition of the photo-generated electrons, and then enhanced the absorption of the composite in the visible light region. Secondly, the change in the color of the sample also cause an increase in absorbance.
Next, the active species generated by visible-light irradiation were investigated by adding different scavengers to the photocatalytic system. The results are shown in Fig. 8(d). The photocatalytic performance of PNIPAM/Fe3O4/g-C3N4 changed slightly when IPA was added as scavenger of •OH [45], indicating that •OH was not the major oxidative species in the PNIPAM/Fe3O4/g-C3N4 photocatalytic reaction system. However, the photocatalytic performance decreased upon the addition of BQ as a scavenger for photogenerated superoxide radical species in the solution [46, 47]. The tetracycline photo-degradation rate was significantly decreased to only 49%, which indicated that superoxide radicals were the main active species. When TEOA, which is a scavenger for h+ [48], was added to the reaction system, the photocatalytic reaction was significantly inhibited. These results indicated that superoxide radicals and h+ were the predominant contributors to the photocatalytic reaction.
Fig. 9(a) shows the absorption curves of tetracycline molecules by PNIPAM/Fe3O4/g-C3N4 at different temperatures. The solution systems almost reached adsorption equilibrium within 60 min. PNIPAM/Fe3O4/g-C3N4 exhibited the highest adsorption capacity of 9.0 mg/g at 20 ℃. Interestingly, the adsorption capacity of PNIPAM/Fe3O4/g-C3N4 was reduced at 45 ℃, which may have been because PNIPAM is hydrophilic and swells in water at lower temperatures (T < 32 ℃), so the tetracycline molecules could reach the surface of PNIPAM easily. In contrast, at high temperatures (T > 32 ℃), PNIPAM shrinks and becomes hydrophobic [49-52], preventing the tetracycline molecules from being adsorbed on the surface of PNIPAM/Fe3O4/g-C3N4. Therefore, the absorption capacity of PNIPAM/Fe3O4/g-C3N4 was increased by the change in polarity of the PNIPAM induced by its volume transition. Moreover, a superior adsorption capacity is beneficial for photocatalytic activity because of the increased contact between target molecules and photocatalyst. PNIPAM/Fe3O4/g-C3N4 had a higher adsorption capacity at 20 ℃ than at 45 ℃ and exhibited a higher degradation rate of about 80%. As shown in Fig. 9(b), the degradation rate decreased to 54% when the temperature was increased to 45 ℃. This results provides direct evidence that the thermo-sensitive composite had a switchable organic pollutant degradation ability based on temperature control.
The variation in the absorbance of tetracycline on PNIPAM/Fe3O4/g-C3N4 was also measured. Fig. 10(a) illustrates that the absorbance of tetracycline decreased obviously with increasing light irradiation time at 20 ℃. No absorbance peak was observed after visible-light irradiation of 120 min, which indirectly confirmed that the tetracycline molecules were destroyed or decomposed into smaller molecules [53, 54]. These results further indicated that the swelling of the polymer at low temperature made it more hydrophilic and more lacunose, which allowed the tetracycline molecules to reach the surface of the PNIPAM easily for photocatalytic degradation by active species. A similar experiment was conducted at 45 ℃ (Fig. 10(b)), and the results showed that the absorbance was slightly changed compared with that at 20 ℃. This result can be attributed to the lower binding ability of PNIPAM at higher temperature meaning that the tetracycline molecules could not be adsorbed on the PNIPAM/Fe3O4/g-C3N4 surface, as discussed above. Therefore, PNIPAM/Fe3O4/g-C3N4, showed temperature responsive intelligent binding properties and the temperature was a crucially important factor for its efficient degradation of organic pollutants. Therefore, from the experimental results, catalytic activity of the intelligent PNIPAM/Fe3O4/g-C3N4 photocatalyst was controllable by adjusting the temperature.
Based on the experimental results and theoretical analysis, a possible mechanism for the photo-degradation of tetracycline by PNIPAM/Fe3O4/g-C3N4 is illustrated in Fig. 11. As previously reported, based on density functional theory (DFT), g-C3N4 has appropriate conduction band (CB) and valence band (VB) edge potentials at-1.13 and 1.57 eV, respectively [55, 56]. Under visible light irradiation, electrons are excited from the VB to the CB of g-C3N4, and then migrate to the surface of the Fe3O4, which effectively inhibits the direct recombination of electrons and holes and thus enhances the photocatalytic activity [20]. The photo-generated holes (h+) in g-C3N4 can react with tetracycline, causing it to degrade into small inorganic molecules such as CO2 and H2O. Meanwhile, the photo-generated electrons in g-C3N4 are transferred to Fe3O4. In the Fe3O4 crystals, the electrons are good reductants that can capture adsorbed oxygen molecules, well-known electron accepters, onto the composite catalyst surface and react with them, generating superoxide radicals •O2-[57]. The generated •O2- may also react with H+ to further produce •OH. Both superoxide radicals and •OH have high oxidation activity, and can finally decompose tetracycline. Therefore, the efficient photocatalytic degradation of tetracycline proceeds smoothly. Based on the above analysis, the charge carrier transfers and degradation reactions are proposed as follows:
MS is widely used to probe the possible mechanism of photodegradation and identify the intermediates. In the mass spectrum obtained for tetracycline degradation, many extra peaks appeared alongside that of tetracycline (including Mr/z=478, 401, 373, 359, 348, 292, 246, 164). As time went on, the characteristic peak of tetracycline (Mr/z=445) decreased in intensity. This result meant that the tetracycline molecules were decomposed into other intermediate degradation products, which were gradually decomposed to CO2, H2O, and other small molecules. The mechanism speculated according to the MS results is illustrated in Fig. 12.
Under visible light excitation, B was formed by the loss of the-CONH2 group from A (tetracycline). The process of "A to C" occurred through the loss of-CONH2 and-(CH3)2 from the tetracycline. The formation of "D" must have been caused by •OH substitution for H in-N (CH3)2. "B, C, and D to E" resulted from the removal of the-CONH2, -CH3, and-N (CH3)2 groups from tetracycline, respectively. The degradation process of "E to F" occurred through the breakage of a C=C double bond and loss of •OH. Likewise, "F to G" was caused by the breakage of a C=C double bond, following which •OH and-NH2 were easily lost and a rearrangement reaction occurred. Along with this reaction, the formation of "H" resulted from the breakage of the σ bond and rearrangement reaction. "I" was formed from "H" by the opening of C=C bonds, and the loss of molecular C6H11O. Finally, under visible-light irradiation, these products were further decomposed into small gaseous components.
In summary, an intelligent PNIPAM/Fe3O4/g-C3N4 composite photocatalyst was successfully fabricated by a hydrothermal method and thermal photoinitiation technology. The photocatalyst exhibited reversible thermo-induced swelling/shrinking transition, and its catalytic activity could be modulated through the corresponding temperature-dependent binding behaviors. The rate of tetracycline degradation over the catalyst was more significant at 20 ℃ than at 45 ℃. Moreover, PNIPAM/Fe3O4/g-C3N4 could be easily separated from the reaction system and exhibited little deactivation after five repeated use. Importantly, the tetracycline was decomposed into harmless small molecules, as confirmed by MS. The results of this study prove the feasible and promising use of thermo-sensitive composites in the photo-degradation of organic pollutants and intelligent catalysis.
2017,
Vol. 38 Issue (3): 447-457 
PDF
2017,
Vol. 38 Issue (3): 447-457 PDF