Photocatalytic Oxidative Degradation of Carbamazepine by TiO 2 Irradiated by UV Light Emitting Diode

: Here, ultraviolet light-emitting diodes (UV-LED) combined with TiO 2 was used to investigate the feasibility of carbamazepine (CBZ) degradation. The e ﬀ ects of various factors, like crystal form of the catalyst (anatase, rutile, and mixed phase), mass concentration of TiO 2 , wavelength and irradiation intensity of the UV-LED light source, pH of the reaction system, and coexisting anions and cations, on the photocatalytic degradation of CBZ were studied. The mixed-phase (2.8 g / L) showed the best degradation e ﬃ ciency at 365 nm among three kinds of TiO 2 , wherein CBZ (21.1 µ M) was completely oxidized within 1 h. The results of batch experiments showed that: (i) CBZ degradation e ﬃ ciency under UV-LED light at 365 nm was higher than 275 nm, due to stronger penetrability of 365 nm light in solution. (ii) The degradation e ﬃ ciency increased with increase in irradiation intensity and pH, whereas it decreased with increase in initial CBZ concentration. (iii) The optimal amount of mixed-phase TiO 2 catalyst was 2.8 g / L and excessive catalyst decreased the rate. (iv) The co-existence of CO 32 − , HCO 3 − , and Fe 3 + ions in water signiﬁcantly accelerated the degradation rate of photocatalytic CBZ, whereas Cu 2 + ions strongly inhibited the degradation process of CBZ. · OH was found to be the main active species in the UV-LED photocatalytic degradation of CBZ. UV-LED is more environmentally friendly, energy e ﬃ cient, and safer, whereas commercial TiO 2 is economical and readily available. Therefore, this study provides a practically viable reference method for the degradation of pharmaceuticals and personal care products (PPCPs).


Introduction
In recent years, pharmaceuticals and personal care products (PPCPs) have gained more attention, due to the manufacturing and application of pesticides and medicines for treating humans and livestock [1].PPCPs and its metabolites can be found in water bodies during regular inspections in aquatic environments conducted world-wide.Since, most PPCPs have very stable and complex molecular structures they cannot be easily captured and degraded by organisms.This increases the environmental pollution and is potentially harmful to human health [2,3].At the same time, the presence of trace amounts of chemical pollutants in the aquatic environment may lead to the Table 1.Molecular structure and solubility of carbamazepine (CBZ).

Chemical Formula
Molecular Weight/(g•mol −1 ) Solubility (20 entire ecosystem [4][5][6].Therefore, PPCP must be completely eliminated from the natural water bodies to reduce its threat to all biomes.Carbamazepine (CBZ, C15H12N2O), a dibenzazepine derivative, has molecular structure similar to that of tricyclic antidepressants.At present, it is widely used to treat trigeminal neuralgia, epilepsy, and mental illness [7].Table 1 enlists the molecular structure and solubility of CBZ.Due to its stable structure, the removal efficiency of CBZ is generally less than 10% by traditional sewage treatment methods like coagulation/flocculation, chlorination, and microbial degradation [8].The rest of it (major portion) is discharged into different aqueous environments either as a raw drug or its degradation intermediates.It is now commonly found in sewage, surface water, groundwater, and even drinking water.Its residues can trigger homologous estrogen activity and cause reproductive toxicity [9,10].Hence, there is an urgent need to develop an environmentally friendly, college-level method to degrade or remove CBZ from the water bodies.Various advanced oxidation process (AOP) methods are remediation techniques, effective in the removal of PPCPs.They include treatment methods using UV radiations in combination with H2O2, ozone, or titanium dioxide [11][12][13][14].Most AOP methods can effectively remove CBZ residues (93-100%) from water [15][16][17][18].One kind of AOP, semiconductor based photocatalytic oxidation technology, has the characteristics of environmental protection and shows great application prospects in the treatment of organic pollutants.Hydroxyl radicals have extremely strong oxidizing ability and, similar to most of the macromolecular organic pollutants having complex structures, can be oxidized and decomposed into smaller molecular weight and non-toxic organic compounds, or even completely decomposed into inorganic substances.Nano-semiconductor TiO2 particles are endowed with characteristics like green environmental protection, high chemical stability, large specific surface area, and low cost.They have become increasingly popular as photocatalytic semiconductor materials, because of their ability to photocatalytically decompose water [19].Many previous studies by scholars have shown that TiO2 photocatalysis by ultraviolet irradiation is an efficient and environment-friendly technique for removing organic antibiotic compounds from the aquatic environment.However, the main limitation is the source of UV light in the practical applications of photocatalysis.The most commonly used source, a low and medium pressure mercury lamp, has many short-comings, such as fragility, bulkiness, serious heat radiations, high energy consumption, short life (500-2000 h), and so on [20].In addition, mercury ultraviolet lamps emit light of only a fixed wavelength in a very narrow range and also have a risk of pollution due to mercury leakage [21].
In recent years, UV-LED technology has been widely explored for its applications, due to its unique advantages such as environmental protection, long service life, and short warm-up time.LED lights require low energy input and each LED lamp bead is very small, making it flexible and easy to install and transport [22,23].In addition, the emission wavelength of UV-LEDs can be easily adjusted according to the material composition of the lamp beads.Today, UV-LED is a more powerful narrow-band light emitting device with shorter relaxation times [24].Compared with traditional ultraviolet light, ultraviolet LED technology is more suitable for wastewater treatment, s as it is easy to regulate.For example, Cai et al. used UV-A type LED light to irradiate TiO2 nanoparticles to photo-catalyze antibiotics in wastewater, and their research showed that continuous UV-LED/TiO2 photocatalytic technology could be applied to 100 ppb concentration of sulfamethoxazole/methoxybenzyl above 320 nm.The degradation rate of pyridine was >90% [25].
Various advanced oxidation process (AOP) methods are remediation techniques, effective in the removal of PPCPs.They include treatment methods using UV radiations in combination with H 2 O 2 , ozone, or titanium dioxide [11][12][13][14].Most AOP methods can effectively remove CBZ residues (93-100%) from water [15][16][17][18].One kind of AOP, semiconductor based photocatalytic oxidation technology, has the characteristics of environmental protection and shows great application prospects in the treatment of organic pollutants.Hydroxyl radicals have extremely strong oxidizing ability and, similar to most of the macromolecular organic pollutants having complex structures, can be oxidized and decomposed into smaller molecular weight and non-toxic organic compounds, or even completely decomposed into inorganic substances.Nano-semiconductor TiO 2 particles are endowed with characteristics like green environmental protection, high chemical stability, large specific surface area, and low cost.They have become increasingly popular as photocatalytic semiconductor materials, because of their ability to photocatalytically decompose water [19].Many previous studies by scholars have shown that TiO 2 photocatalysis by ultraviolet irradiation is an efficient and environment-friendly technique for removing organic antibiotic compounds from the aquatic environment.However, the main limitation is the source of UV light in the practical applications of photocatalysis.The most commonly used source, a low and medium pressure mercury lamp, has many short-comings, such as fragility, bulkiness, serious heat radiations, high energy consumption, short life (500-2000 h), and so on [20].In addition, mercury ultraviolet lamps emit light of only a fixed wavelength in a very narrow range and also have a risk of pollution due to mercury leakage [21].
In recent years, UV-LED technology has been widely explored for its applications, due to its unique advantages such as environmental protection, long service life, and short warm-up time.LED lights require low energy input and each LED lamp bead is very small, making it flexible and easy to install and transport [22,23].In addition, the emission wavelength of UV-LEDs can be easily adjusted according to the material composition of the lamp beads.Today, UV-LED is a more powerful narrow-band light emitting device with shorter relaxation times [24].Compared with traditional ultraviolet light, ultraviolet LED technology is more suitable for wastewater treatment, s as it is easy to regulate.For example, Cai et al. used UV-A type LED light to irradiate TiO 2 nanoparticles to photo-catalyze antibiotics in wastewater, and their research showed that continuous UV-LED/TiO 2 photocatalytic technology could be applied to 100 ppb concentration of sulfamethoxazole/methoxybenzyl above Catalysts 2020, 10, 540 3 of 13 320 nm.The degradation rate of pyridine was >90% [25].Liang et al. found that an UV-LED/TiO 2 photocatalytic reaction system, using periodic lighting control, was extremely effective for the degradation and removal of PPCP.At the same time, the use of porous titanium dioxide substrate significantly reduced the energy requirements of the reaction system [26].It was also found that under certain experimental conditions, the UV-LED/TiO 2 process could effectively decompose methylene blue (MB) in water.This indicated that UV-LED is a radiation source that effectively stimulates the photocatalytic action of TiO 2 , and this technology has a high potential for photodegradation [27].In all, the UV-LED/TiO 2 system is a valuable technology, especially for wastewater treatment containing PPCP.
At present, there is only limited research on photocatalytic degradation of CBZ, using a more energy-saving and environmentally safe UV-LED source and an economical and readily available commercial TiO 2 , which inspired us to explore the feasibility of this study.Herein, single-factor experiments were conducted to study the effects of crystal forms of nano-TiO 2 , the amount of TiO 2 , the wavelength of UV-LED and its irradiation intensity, CBZ concentration in water, pH of aqueous solution, and coexisting anions and cations common in water, on the degradation efficiency of CBZ.Our research demonstrated the feasibility of this UV LED combined with TiO 2 photodegradation technology and explored the best operating conditions in actual water treatment project.

Characterization of Three Different Crystal Forms of TiO 2 Particles
Scanning electron microscopy (SEM) was employed to characterize the morphologies of the three different phases of TiO 2 .Figure 1a-c showed the particle sizes of TiO 2 nanoparticles, viz.anatase, rutile, and mixed phases, to be 20-100, 50-100, and 20-100 nm, respectively.The UV-Vis diffuse reflectance spectra (UV-Vis DRS, Figure 1d) showed that all the three phases of TiO2 nanoparticles could absorb light below 420 nm.The N2 adsorption-desorption tests were used to calculate the surface areas of the three crystal forms of TiO 2 .According to Table 2 and Figure 1e, the surface areas of the anatase, rutile, and mixed phases were 78.7, 32.2, and 102.6 m 2 /g, respectively, by Brunner−Emmet−Teller (BET).X-ray diffraction patterns of three forms of TiO 2 nanoparticles can be seen in Figure 1f.Anatase and rutile matched with their respective standard cards (JCPDS 89-4921) and (JCPDS 21-1276).The equation, X = 1/(1 + 0.8I A /I R ), was used to calculate the rutile content in the mixed phase, where X was the mass fraction of the rutile phase.I A and I R in the equation were the 2θ values of the characteristic peaks of anatase and rutile forms, which were 25.3 • and 27.4 • , respectively.Herein, the ratio of anatase to rutile forms in the mixed crystal TiO 2 form was about 83:17, which was similar to the common commercial P25 (80:20).

Control Experiments of CBZ Degradation
The control experiments for CBZ degradation were also conducted.Figure 2a shows that the CBZ degradation failed to proceed in the absence of any of the following: illumination or the mixed-phase TiO 2 photocatalyst.Owing to the adsorption of CBZ on the surface of TiO 2 , its concentration decreased by 5%, after reaching adsorption equilibrium.The degradation rate (η t ) of CBZ was calculated according to Equation (1) (Where η t is the conversion rate (percentage) at a given time t, C 0 is the initial concentration of the reagent, and C t is the residual concentration at a given time t).The concentration of CBZ decreased sharply under irradiation of UV-LED light for 60 min in the presence of mixed-phase TiO 2 , wherein it was almost completely degraded (CBZ degradation rate reached 98.5%).Since, the valence band potential (3.1 eV vs. NHE) was more positive than H 2 O/•OH (2.74 eV vs. NHE), the TiO 2 photocatalyst excited by UV-LED, could produce highly oxidized photogenerated holes and hydroxyl radicals that could oxidize the CBZ molecules into H 2 O, CO 2 , and other non-toxic small molecules.It is speculated that the CBZ photocatalytic degradation step may be a common photocatalytic reaction, as shown in reaction formulae 1-4.In addition, the CBZ degradation process was consistent with the pseudo-first-order kinetic models with a degree of fit R 2 > 0.99 (Figure 2b).These results indicated that the photocatalytic method of UV-LED combined with TiO 2 photocatalyst is a technology that can be implemented to effectively remove CBZ molecules from an aqueous environment.

Control Experiments of CBZ Degradation
The control experiments for CBZ degradation were also conducted.Figure 2a shows that the CBZ degradation failed to proceed in the absence of any of the following: illumination or the mixed-phase TiO2 photocatalyst.Owing to the adsorption of CBZ on the surface of TiO2, its concentration decreased by 5%, after reaching adsorption equilibrium.The degradation rate (ηt) of CBZ was calculated according to Equation (1) (Where ηt is the conversion rate (percentage) at a given time t, C0 is the initial concentration of the reagent, and Ct is the residual concentration at a given time t).The concentration of CBZ decreased sharply under irradiation of UV-LED light for 60 min in the presence of mixed-phase TiO2, wherein it was almost completely degraded (CBZ degradation rate reached 98.5%).Since, the valence band potential (3.1 eV vs. NHE) was more positive than H2O/•OH (2.74 eV vs. NHE), the TiO2 photocatalyst excited by UV-LED, could produce highly oxidized photogenerated holes and hydroxyl radicals that could oxidize the CBZ molecules into H2O, CO2, and other non-toxic small molecules.It is speculated that the CBZ photocatalytic degradation step may be a common photocatalytic reaction, as shown in reaction formulae 1-4.In addition, the CBZ degradation process was consistent with the pseudo-first-order kinetic models with a degree of fit R 2 > 0.99 (Figure 2b).These results indicated that the photocatalytic method of UV-LED combined with TiO2 photocatalyst is a technology that can be implemented to effectively remove CBZ molecules from an aqueous environment.nm; I0 = 774 µW/cm 2 ; and operating pH = 7.6.

Effect of UV-LED Wavelength and its Intensity
Figure 3 shows the effects of two different UV-LED emission wavelengths on CBZ degradation.The CBZ photocatalytic degradation rate (98.5%) using 365 nm irradiation was significantly higher than that of 275 nm (36.78%).The degradation rate constant at 365 nm irradiation was 0.067 min −1 , whereas at 275 nm it was only 0.007 min −1 .Generally, the energy of the shorter-wavelength ultraviolet rays is higher than longer-wavelength, however, its penetrating power is weaker [27,28]  Catalysts 2020, 10, 540 5 of 13

Effect of UV-LED Wavelength and Its Intensity
Figure 3 shows the effects of two different UV-LED emission wavelengths on CBZ degradation.The CBZ photocatalytic degradation rate (98.5%) using 365 nm irradiation was significantly higher than that of 275 nm (36.78%).The degradation rate constant at 365 nm irradiation was 0.067 min −1 , whereas at 275 nm it was only 0.007 min −1 .Generally, the energy of the shorter-wavelength ultraviolet rays is higher than longer-wavelength, however, its penetrating power is weaker [27,28] Therefore, we speculate that the difference in degradation effects could be due to the differences in penetrating power.Compared with 275 nm irradiations, the use of 365 nm UV-LED source is a more cost-effective, energy-saving, and efficient method for photocatalytic oxidation of CBZ.

Effect of UV-LED Wavelength and its Intensity
Figure 3 shows the effects of two different UV-LED emission wavelengths on CBZ degradation.The CBZ photocatalytic degradation rate (98.5%) using 365 nm irradiation was significantly higher than that of 275 nm (36.78%).The degradation rate constant at 365 nm irradiation was 0.067 min −1 , whereas at 275 nm it was only 0.007 min −1 .Generally, the energy of the shorter-wavelength ultraviolet rays is higher than longer-wavelength, however, its penetrating power is weaker [27,28] Therefore, we speculate that the difference in degradation effects could be due to the differences in penetrating power.Compared with 275 nm irradiations, the use of 365 nm UV-LED source is a more cost-effective, energy-saving, and efficient method for photocatalytic oxidation of CBZ.The radiation intensity of UV-LED is one of the important factors affecting the photocatalytic reaction, and it directly determines the number of photo-generated electrons and holes in the photocatalytic system.The effects of UV radiations of different intensities (220, 332, 437, 551, 663, and 774 µW/cm 2 ) on the degradation rate of CBZ at a fixed wavelength of 365 nm, were studied.Obviously, as shown in Figure 4a, the degradation rate increased with increase in light intensity.In particular, CBZ could be completely degraded at an irradiation intensity of 774 µW/cm 2 within 60 min.However, the degradation rate of CBZ at a light intensity of 220 µW/cm 2 within 60 min was The radiation intensity of UV-LED is one of the important factors affecting the photocatalytic reaction, and it directly determines the number of photo-generated electrons and holes in the photocatalytic system.The effects of UV radiations of different intensities (220, 332, 437, 551, 663, and 774 µW/cm 2 ) on the degradation rate of CBZ at a fixed wavelength of 365 nm, were studied.Obviously, as shown in Figure 4a, the degradation rate increased with increase in light intensity.In particular, CBZ could be completely degraded at an irradiation intensity of 774 µW/cm 2 within 60 min.However, the degradation rate of CBZ at a light intensity of 220 µW/cm 2 within 60 min was only 57.23%.When the light intensity was changed from 220 µW/cm 2 to 774 µW/cm 2 , the reaction rate constant for photocatalytic degradation showed a 6.7-fold increase (Figure 4b).

Effect of Crystalline Form of Nano-TiO2
The photocatalytic performances of the three crystalline forms of TiO2 on CBZ degradation were studied.The degradation curves and pseudo first-order kinetic models are shown in Figure 5.The degradation rate of anatase form (87.78%) was significantly higher than the rutile form (80.29%), within 60 min.The mixed-phase TiO2 showed the highest rate and could degrade 98.5% of CBZ The photocatalytic performances of the three crystalline forms of TiO 2 on CBZ degradation were studied.The degradation curves and pseudo first-order kinetic models are shown in Figure 5.The degradation rate of anatase form (87.78%) was significantly higher than the rutile form (80.29%), within 60 min.The mixed-phase TiO 2 showed the highest rate and could degrade 98.5% of CBZ within 60 min.At the same time, the reaction rate constant of mixed-phase was as high as 0.067 min −1 , which was nearly 2 and 3 times higher than those of anatase and rutile forms, respectively.Previous studies had demonstrated that the photogenerated carriers could easily transfer at the interface between the anatase and rutile forms, which promoted charge separation efficiency and photocatalytic activity [29].

Effect of Crystalline Form of Nano-TiO2
The photocatalytic performances of the three crystalline forms of TiO2 on CBZ degradation were studied.The degradation curves and pseudo first-order kinetic models are shown in Figure 5.The degradation rate of anatase form (87.78%) was significantly higher than the rutile form (80.29%), within 60 min.The mixed-phase TiO2 showed the highest rate and could degrade 98.5% of CBZ within 60 min.At the same time, the reaction rate constant of mixed-phase was as high as 0.067 min −1 , which was nearly 2 and 3 times higher than those of anatase and rutile forms, respectively.Previous studies had demonstrated that the photogenerated carriers could easily transfer at the interface between the anatase and rutile forms, which promoted charge separation efficiency and photocatalytic activity [29].

Effect of Initial CBZ Concentration
CBZ solutions of four different concentrations (10.6, 21.2, 31.7, and 42.3 µM) were used to evaluate the effect of initial concentration of CBZ on degradation.As shown in Figure 6, when the CBZ concentration in the original aqueous solution increased from 10.6 to 42.3 µM, the degradation rate decreased slightly.However, the rate constant decreased sharply from 0.084 to 0.040 min −1 .The reason for this phenomenon was that when the initial CBZ concentration was high, a higher number of CBZ molecules competed for the photogenerated holes and hydroxyl radicals in the system.The reduction in the main active material for photocatalytic reaction reduced the degradation rate of CBZ.CBZ concentration in the original aqueous solution increased from 10.6 to 42.3 µM, the degradation rate decreased slightly.However, the rate constant decreased sharply from 0.084 to 0.040 min −1 .The reason for this phenomenon was that when the initial CBZ concentration was high, a higher number of CBZ molecules competed for the photogenerated holes and hydroxyl radicals in the system.The reduction in the main active material for photocatalytic reaction reduced the degradation rate of CBZ.

Effect of Nano-TiO2 Dosage
The effect of amount of photocatalyst on CBZ degradation was investigated by changing the mass concentration of initial TiO2 (mixed crystal form) in the range of 0.8-3.6 g/L. Figure 7 shows the degradation curves of CBZ with different dosages of TiO2.When the amount of TiO2 was 0.8 g/L, the degradation rate reached 87.78% within 60 min.The degradation rate gradually reached its

Effect of Nano-TiO 2 Dosage
The effect of amount of photocatalyst on CBZ degradation was investigated by changing the mass concentration of initial TiO 2 (mixed crystal form) in the range of 0.8-3.6 g/L. Figure 7 shows the degradation curves of CBZ with different dosages of TiO 2 .When the amount of TiO 2 was 0.8 g/L, the degradation rate reached 87.78% within 60 min.The degradation rate gradually reached its maximum value (98.5%) when 2.8 g/L of TiO 2 was used (Figure 7a).Further, as the TiO 2 dosage was further increased from 2.8 to 3.6 g/L, the degradation efficiency decreased.High concentration of catalyst increases light scattering and reduces light penetration, which decreases the utilization of light.These factors were responsible for the decrease in degradation activity.CBZ underwent fastest degradation, with a rate constant of 0.058 min −1 , when the TiO 2 mass concentration was 2.4 g/L (Figure 7b).

Effect of Nano-TiO2 Dosage
The effect of amount of photocatalyst on CBZ degradation was investigated by changing the mass concentration of initial TiO2 (mixed crystal form) in the range of 0.8-3.6 g/L. Figure 7 shows the degradation curves of CBZ with different dosages of TiO2.When the amount of TiO2 was 0.8 g/L, the degradation rate reached 87.78% within 60 min.The degradation rate gradually reached its maximum value (98.5%) when 2.8 g/L of TiO2 was used (Figure 7a).Further, as the TiO2 dosage was further increased from 2.8 to 3.6 g/L, the degradation efficiency decreased.High concentration of catalyst increases light scattering and reduces light penetration, which decreases the utilization of light.These factors were responsible for the decrease in degradation activity.CBZ underwent fastest degradation, with a rate constant of 0.058 min −1 , when the TiO2 mass concentration was 2.4 g/L (Figure 7b).

Effect of pH
During the entire process of photocatalytic oxidation, the pH of reaction system also greatly influences the degradation efficiency of the target compounds, which in turn is mainly affected by the generation of hydroxyl radicals [30].Figure 8 shows the degradation of CBZ under different pH conditions (3.0, 5.0, 7.0, 9.0, and 11.0).As the pH increased from 3.0 to 11.0, the degradation rate increased gradually.The reaction rate was the fastest when the pH was 11 and its rate constant reached 0.110 min −1 , which was about 4 times than that at pH = 3.0.The •OH species was mainly formed from OH − , and the production efficiency of OH − was much higher than that of H 2 O to produce •OH.Therefore, an aqueous alkaline environment was more conducive for the degradation of CBZ.

Influence of Co-Existing Ions in the Aquatic Environment
Coexistence of different anions and cations in water has different effects on the photocatalytic degradation of CBZ [31].The concentration of a fixed coexisting anion and cation in the CBZ degradation reaction system was 1 mmol/L, keeping other conditions unchanged (C(mixed-phase TiO 2 ) = 2.0 g/L; C(CBZ) 0 = 18.8 µM; operating temperature = 20 • C; wavelength of UV-LED = 365 nm; I 0 = 774 µW/cm 2 ; and operating pH = 7.6).The effects of common anions and cations in water on the degradation of CBZ were investigated.As shown in Figure 9, the presence of Cl − in the system had little effect on the photodegradation process of CBZ, whereas NO 3 − had a slight inhibitory effect.
The SO 4 2− promoted the degradation to certain extent, while the presence of CO 3 2− and HCO 3 − significantly promoted the photocatalytic degradation of CBZ.When a large number of CO 3 2− and HCO 3 − coexisted in the solution, the solution was alkaline.An increase in the number of OH − ions was useful for the formation of •OH, which improved the photocatalytic activity of the system.Hu et al. [32] reported that CO 3 2− and HCO 3 − promoted the photocatalytic degradation of sulfamethoxazole, which was consistent with the results of this study.

Effect of pH
During the entire process of photocatalytic oxidation, the pH of reaction system also greatly influences the degradation efficiency of the target compounds, which in turn is mainly affected by the generation of hydroxyl radicals [30].Figure 8 shows the degradation of CBZ under different pH conditions (3.0, 5.0, 7.0, 9.0, and 11.0).As the pH increased from 3.0 to 11.0, the degradation rate increased gradually.The reaction rate was the fastest when the pH was 11 and its rate constant reached 0.110 min −1 , which was about 4 times than that at pH = 3.0.The •OH species was mainly formed from OH − , and the production efficiency of OH − was much higher than that of H2O to produce •OH.Therefore, an aqueous alkaline environment was more conducive for the degradation of CBZ.

Influence of Co-Existing Ions in the Aquatic Environment
Coexistence of different anions and cations in water has different effects on the photocatalytic degradation of CBZ [31].The concentration of a fixed coexisting anion and cation in the CBZ degradation reaction system was 1 mmol/L, keeping other conditions unchanged (C(mixed-phase TiO2) = 2.0 g / L; C(CBZ)0 = 18.8 µM; operating temperature = 20 °C; wavelength of UV-LED = 365 nm; I0 = 774 µW/cm 2 ; and operating pH = 7.6).The effects of common anions and cations in water on the degradation of CBZ were investigated.As shown in Figure 9, the presence of Cl -in the system had little effect on the photodegradation process of CBZ, whereas NO3 -had a slight inhibitory effect.The SO4 2− promoted the degradation to certain extent, while the presence of CO3 2− and HCO3 − significantly promoted the photocatalytic degradation of CBZ.When a large number of CO3 2− and HCO3 − coexisted in the solution, the solution was alkaline.An increase in the number of OH -ions was useful for the formation of •OH, which improved the photocatalytic activity of the system.Hu et al. [32] reported that CO3 2− and HCO3 − promoted the photocatalytic degradation of sulfamethoxazole, which was consistent with the results of this study.The effects of four commonly coexisting cations on the degradation of CBZ in natural water are shown in Figure 10.The presence of Ca 2+ , Mg 2+ , and Cu 2+ ions inhibited the photocatalytic degradation to a certain extent and the order of their inhibition effects was: Cu 2+ > Ca 2+ > Mg 2+ .The Fe 3+ ions had a positive effect, and 100% CBZ degradation was observed in less than 20 min.The photocatalytic degradation of CBZ promoted by Fe 3+ ions could be attributed to the fact that Fe 3+ traps photo-generated electrons generated by TiO 2 , thereby effectively reducing the chance of electron-hole recombination.In addition, the addition of Fe 3+ ions could form a Fenton system, thereby increasing the number of free active•OH species in the reaction system, which significantly increased the degradation capacity and degradation rate of CBZ.promoted by Fe 3+ ions may be attributed to the fact that Fe 3+ traps photo-generated electrons generated by 245 TiO2, thereby effectively reducing the chance of hole and electron recombination.In addition, the addition of 246 Fe 3+ ions may form a Fenton system in the system, increase the amount of free active •OH in the reaction 247 system, and significantly increase the degradation capacity and degradation rate of CBZ.248 249 degradation to a certain extent and the order of their inhibition effects was: Cu > Ca > Mg .The Fe 3+ ions had a positive effect, and 100% CBZ degradation was observed in less than 20 min.The photocatalytic degradation of CBZ promoted by Fe 3+ ions could be attributed to the fact that Fe 3+ traps photo-generated electrons generated by TiO2, thereby effectively reducing the chance of electron-hole recombination.In addition, the addition of Fe 3+ ions could form a Fenton system, thereby increasing the number of free active•OH species in the reaction system, which significantly increased the degradation capacity and degradation rate of CBZ.

Influence of Free Radical Scavenging
The •OH radical is considered to be the main active species in TiO2 photocatalytic degradation reaction [33].By adding different concentrations of t-butanol (TBA) as the OH scavenger to the reaction system, the effect of number of active •OH radicals on the photocatalytic degradation was studied.As shown in Figure 11, the presence of TBA significantly inhibited the photocatalytic

Influence of Free Radical Scavenging
The •OH radical is considered to be the main active species in TiO 2 photocatalytic degradation reaction [33].By adding different concentrations of t-butanol (TBA) as the OH scavenger to the reaction system, the effect of number of active •OH radicals on the photocatalytic degradation was studied.As shown in Figure 11, the presence of TBA significantly inhibited the photocatalytic degradation of CBZ.With increase in TBA concentration in the reaction system, the inhibitory effect was enhanced.When the TBA concentration in the system was 10 mM, the degradation efficiency was 19.91% after 60 min of UV-LED irradiation.As the TBA concentration increased to 50 mM, the degradation efficiency was only 6.58%.These experiments also show that OH species are indeed the main active species in CBZ degradation.
Catalysts 2020, 10, x FOR PEER REVIEW 10 of 13 was enhanced.When the TBA concentration in the system was 10 mM, the degradation efficiency was 19.91% after 60 min of UV-LED irradiation.As the TBA concentration increased to 50 mM, the degradation efficiency was only 6.58%.These experiments also show that OH species are indeed the main active species in CBZ degradation.

Materials
Carbamazepine (≥98%, Shandong Xiya Chemical Co.Ltd, Linyi,Shandong,China ) A stock mother liquor of CBZ, with a concentration of 211 µmol/L, was prepared and then serially diluted with ultrapure water, according to the concentration required for each experiment.Acetonitrile (HPLC grade purity, Merck KGaA, Darmstadt, Germany).HCl (0.1 mol/L), NaOH (0.01 mol/L), and phosphate buffer (0.01 mol/L) solutions were used to adjust the pH of the reaction system.The water used was ultra-pure and was prepared using Millipore Milli-Q ultra-pure water system, with a resistivity of 18.2 MΩ•cm.Three commercial nano TiO2 catalysts having different crystal forms: pure anatase, mixed-phase, and rutile forms (>98%, Shandong Xiya Chemical Co.Ltd, Linyi,Shandong,China ).All other reagents were from Damao Chemical Reagent Factory (Analytical grade, Tianjin, China).

Materials
Carbamazepine (≥98%, Shandong Xiya Chemical Co. Ltd., Linyi, Shandong, China) A stock mother liquor of CBZ, with a concentration of 211 µMol/L, was prepared and then serially diluted with ultrapure water, according to the concentration required for each experiment.Acetonitrile (HPLC grade purity, Merck KGaA, Darmstadt, Germany).HCl (0.1 mol/L), NaOH (0.01 mol/L), and phosphate buffer (0.01 mol/L) solutions were used to adjust the pH of the reaction system.The water used was ultra-pure and was prepared using Millipore Milli-Q ultra-pure water system, with a resistivity of 18.2 MΩ•cm.Three commercial nano TiO 2 catalysts having different crystal forms: pure anatase, mixed-phase, and rutile forms (>98%, Shandong Xiya Chemical Co. Ltd., Linyi, Shandong, China).All other reagents were from Damao Chemical Reagent Factory (Analytical grade, Tianjin, China).

Test Device
The photocatalytic degradations were performed in a 500 mL open plexiglass container.In order to prevent light energy, UV-LED light, and heat energy, a layer of aluminum foil was tightly wrapped around the outer wall of the reactor.The UV-LED tube was placed in a cylindrical radiation window having a diameter of about 5 cm, which was made of quartz.The light source was composed of two self-made UV-LED units, and the working wavelengths of the two units were 275 and 365 nm.Each unit contained 40 LED chips (manufactured by Shenzhen Julan Technology Co. Ltd., China).The output of 275 nm was about 2.8 mW with a current of 40 mA and the output of 365 nm was about 700 mW with a current of 500 mA.The fluence rate was measured in all experiments using a potassium ferrioxalate actinometer [34].A low-temperature constant temperature circulating water tank was used to circulate cooling water on the outer wall of the reactor in order to control the temperature during the operation at 20 ± 0.5 • C. The intensity of the radiations irradiated into the reactor was measured using an IL-2400 irradiation photometer (International Light, USA).The diagram of UV-LED/TiO 2 photocatalytic device used in this study is shown in Figure 12.

Analytical Method
In the UV-LED/TiO 2 photocatalytic experiment, 2 mL of the sample suspension was withdrawn at regular time intervals, and the suspension was filtered through a 0.22 µM syringe filter (Tianjin Jinteng Experimental Equipment Co. Ltd., Tianjin, China).The concentration of CBZ in the test sample was measured by HPLC (Waters e2695) using a Symmetry C18 column [35].The reaction conditions were as follows: the mobile phase consisted of 60% acetonitrile and 40% pure water.The injection volume was 100 µL, the flow rate was 1 mL/min, the column temperature was maintained at 35 • C, and the UV-LED wavelength of CBZ was below 286 nm.

Analytical Method
In the UV-LED/TiO2 photocatalytic experiment, 2 mL of the sample suspension was withdrawn at regular time intervals, and the suspension was filtered through a 0.22 µm syringe filter (Tianjin Jinteng Experimental Equipment Co. Ltd., Tianjin, China).The concentration of CBZ in the test sample was measured by HPLC (Waters e2695) using a Symmetry C18 column [35].The reaction conditions were as follows: the mobile phase consisted of 60% acetonitrile and 40% pure water.The injection volume was 100 µL, the flow rate was 1 mL/min, the column temperature was maintained at 35 °C, and the UV-LED wavelength of CBZ was below 286 nm.

Conclusions
The main conclusions of this study were as follows:

Figure 6 .
Figure 6.(a) Degradation of CBZ by UV-LED/TiO2 for different initial concentrations of CBZ (b) The pseudo-first-order kinetic models of CBZ degradation by UV-LED/TiO2 for different initial concentrations of CBZ.Operating temperature = 20 °C; wavelength of UV-LED = 365 nm; operating pH = 7.6; and C (mixed-phase TiO2) = 2.0 g/L.

Figure 6 .
Figure 6.(a) Degradation of CBZ by UV-LED/TiO 2 for different initial concentrations of CBZ (b) The pseudo-first-order kinetic models of CBZ degradation by UV-LED/TiO 2 for different initial concentrations of CBZ.Operating temperature = 20 • C; wavelength of UV-LED = 365 nm; operating pH = 7.6; and C (mixed-phase TiO 2 ) = 2.0 g/L.

Figure 6 .
Figure 6.(a) Degradation of CBZ by UV-LED/TiO2 for different initial concentrations of CBZ (b) The pseudo-first-order kinetic models of CBZ degradation by UV-LED/TiO2 for different initial concentrations of CBZ.Operating temperature = 20 °C; wavelength of UV-LED = 365 nm; operating pH = 7.6; and C (mixed-phase TiO2) = 2.0 g/L.

Figure 9 .
Figure 9. (a) Degradation of CBZ by UV-LED/TiO 2 in presence of different co-existing anions (b) The pseudo-first-order kinetic models of CBZ degradation by UV-LED/TiO 2 in presence of different co-existing anions.

Figure 10 .
Figure 10.(a) Degradation of CBZ by UV-LED/TiO2 with different co-existing cations.(b) The pseudo-first-order kinetic models of CBZ degradation by UV-LED/TiO2 with different co-existing cations.

Figure 10 .
Figure 10.(a) Degradation of CBZ by UV-LED/TiO 2 with different co-existing cations.(b) The pseudo-first-order kinetic models of CBZ degradation by UV-LED/TiO 2 with different co-existing cations.

Figure 11 .
Figure 11.(a) Degradation of CBZ by UV-LED/TiO2 using different concentrations of TBA (b) The pseudo-first-order kinetic models of CBZ degradation by UV-LED/TiO2 using different concentrations of TBA.

Figure 11 .
Figure 11.(a) Degradation of CBZ by UV-LED/TiO 2 using different concentrations of TBA (b) The pseudo-first-order kinetic models of CBZ degradation by UV-LED/TiO 2 using different concentrations of TBA.