High Photocatalytic Activity of g-C₃N₄/La-N-TiO₂ Composite with Nanoscale Heterojunctions for Degradation of Ciprofloxacin

Abstract

Ciprofloxacin (CIP) in natural waters has been taken as a serious pollutant because of its hazardous biological and ecotoxicological effects. Here, a 3D nanocomposite photocatalyst g-C₃N₄/La-N-TiO₂ (CN/La-N-TiO₂) was successfully synthesized by a simple and reproducible in-situ synthetic method. The obtained composite was characterized by XRD, SEM, BET, TEM, mapping, IR, and UV-vis spectra. The photocatalytic degradation of ciprofloxacin was investigated by using CN/La-N-TiO₂ nanocomposite. The main influential factors such as pH of the solution, initial CIP concentration, catalyst dosage, and coexisting ions were investigated in detail. The fastest degradation of CIP occurred at a pH of about 6.5, and CIP (5 mg/L starting concentration) was completely degraded in about 60 min after exposure to the simulated solar light. The removal rates were rarely affected by Na⁺ (10 mg·L⁻¹), Ca²⁺ (10 mg·L⁻¹), Mg²⁺ (10 mg·L⁻¹), and urea (5 mg·L⁻¹), but decreased in the presence of NO₃⁻ (10 mg·L⁻¹). The findings indicate that CN/La-N-TiO₂ nanocomposite is a green and promising photocatalyst for large-scale applications and would be a candidate for the removal of the emerging antibiotics present in the water environment.

1. Introduction

Antibiotic pollution poses a serious threat to human health through drinking water and/or the food-chain, especially in developing countries, and how to efficiently remove it has gained great attention over the past [1,2,3,4,5,6]. Fluoroquinolones (FQs) are one of the most consumed antibiotics and have a total consumption of about 25 million kilograms every year in China [7,8]. Ciprofloxacin (CIP), as the second-generation fluoroquinolone, is extensively used in humans and animals to sterilize and fight infections due to their good oral intake and large activity spectrum [9]. Ultimately, it enters the WWTPs (wastewater treatment plants) as a result of being poorly adsorbed in the body and excreted unchanged in feces and urine [10,11]. As the quinolone ring is degraded difficultly, CIP is inefficiently removed by most wastewater and sewage treatment plants via conventional wastewater treatment technologies or is directly poured into water bodies in countries lacking effluent treatments [12]. CIP has been detected frequently in various groundwater [13], surface water [14], and seawater [15] throughout the world [12,13,14,15,16,17,18,19,20]. Therefore, an effective treatment technology is necessary to transform CIP into noN-Toxic, pharmaceutically less active, or more biodegradable species in order to face CIP pollution problems in water.

Semiconductor photocatalysis, which has attracted wide attention by virtue of low- cost, energy-savings, non-toxicity, and high catalytic degradation activity of organic pollutants in purifying water under solar-light irradiation, has been employed to address the problem of antibiotic wastewater and it has made great progress over the past decade [19,20,21,22,23]. A series of previous studies was focused on the photocatalytic degradation of CIP and the data are listed in

Table 1. Photocatalytic activity of semiconductor materials for CIP degradation under visible light irradiation.

Photocatalysts	C0 (mg/L)	pH	Degradation Rate (%)	Light Irradiation Wavelength (λ)	Incident Light Power	Ref
Commercial anatase TiO2	1500 ppm	5.8	Kapp = 0.02 min−1	Simulated visible light	Commercial visible metal halide lamp	[24]
CeO2–Ag/AgBr	10	-	93.05%/120 min	λ > 420 nm	A 300 W Xe lamp	[25]
FeS/Fe2S3/zeolite	2	-	6%/90 min	λ > 380 nm	Tungsten lamp	[26]
Mesoporous TiO2 nanocrystallite aggregates	160	-	96.05%/6 h	200 nm ~1000 nm	A 500 W Xe lamp	[27]
ZnO	4	7/10	48%/60 min	365 nm	A xenon lamp	[28]
3D γ-Fe2O3@ZnO core-shell photocatalyst	10	5.8	92.5%/60 min	Simulated sunlight	A 300 W Xenon lamp	[29]
Bi2O3/(BiO)2CO3	10	4.0~8.3	93.4%/30 min	Simulated solar light	A 300 W Xenon lamp	[30]
Ag@PCNS/BiVO4	10	-	92.6%/120 min 17.5%/120 min	λ > 420 nm λ > 760 nm	-	[31]
ZnAl-LDH/g-C3N4	20	-	84.1%/150 min	-	A 35 W Xenon lamp	[35]
Ag2CrO4/Ag/BiFeO3@RGO	10	6	96%/60 min	Visible light	450 mW cm−2	[33]
TiO2 nanorod/g-C3N4 nanosheet	4.97	6.3	93.4%/60 min	Simulated sunlight	A 500WXenon lamp	[34]
TiO2@g-C3N4 hollow	-	-	74%/120 min	Simulated sunlight	-	[32]
CN/La-N-TiO2	10	6.0–7.0	96.8%/60 min	λ > 420 nm	A 300 W Xe lamp	our work

C₀: the initial concentration. PMS: peroxymonosulfate.

. Gad-Allah et al. employed commercial anatase TiO₂ to degrade CIP under simulated sunlight, and the best reaction rate was obtained in natural ciprofloxacin pH (5.8) and 1000 mg/L TiO₂ [24]. Wen et al. synthesized a novel Z-scheme CeO₂–Ag/AgBr photocatalyst for photocatalytic degradation of ciprofloxacin, the degradation rate could reach 93.05% in 120 min under visible light irradiation [25]. Rahmani-Aliabadi et al. evaluated the photocatalytic activities of FeS/Fe₂S₃/zeolite by photodegradation of ciprofloxacin, and the best catalysis conditions are pH 3.7, catalyst dose of 2.8 g·L⁻¹, 3.8 mg·L⁻¹ of CIP for 102 min irradiation of the suspension [26]. Gan et al. employed different post−treatment methods to synthesize three kinds of mesoporous TiO₂ nanocrystallite aggregates and comparatively studied the removal capability of CIP by adsorption in the dark and photocatalysis under light irradiation [27]. El-Kemary et al. carried out systematic research on the photocatalytic activity of the prepared ZnO nanoparticles for the degradation of ciprofloxacin drug under UV light irradiation in aqueous solutions of different pH values [28]. Li et al. successfully prepared the magnetic 3D γ-Fe₂O₃@ZnO core-shell photocatalyst, which exhibits excellent degradation efficiency of CIP up to 92.5% under simulated sunlight irradiation [29]. Chen et al. studied the CIP decay over Bi₂O₃/(BiO)₂CO₃ heterojunctions under simulated solar light irradiation, where the removal efficiency achieved 93.4% in 30 min [30], and so on [16,20,29,31,32,33,34,35,36,37,38,39,40,41].

From Table 1, it is can be found that the existing research mainly focuses on simulating wastewater degradation in the laboratory and little research on the catalytic degradation of CIP by the semiconductor photocatalytic systems in different water matrices (river water, synthetic wastewater similar by composition to wastewater of pharmaceutical industry). Therefore, further optimization and innovation of the semiconductor photocatalysts for catalytic degradation of CIP in actual wastewater are still in great demand. Hence, in this work, the semiconductor photocatalysis (CN/La-N-TiO₂) will be applied to remove ciprofloxacin in an actual water environment.

TiO₂-assisted photocatalysis has attracted widespread attention to purify polluted water by the removal of organic pollutants, owing to its good stability, low-budget, non-toxicity, and excellent photocatalytic performance [42,43]. It has been demonstrated that TiO₂ doped with elements such as N, P, S, B, C, Co, and La, is an effective way to introduce impurity energy levels, change the intrinsic band structure, reduce the bandgap and improve the photocatalytic capability under visible light [39,41,44]. Lanthanide is a good candidate because its low-valent doping results in the formation of oxygen vacancies on the exterior or surface lattice and hence improves the catalytic activity [41,45,46]. In our previous research work, TiO₂ co-doped with nitrogen and lanthanum/gadolinium ([Xe]5d¹6s²/[Xe]4f⁷5d¹6s² electronic configuration) was successfully synthesized using urea and lanthanum nitrate as a precursor, and the result is that La or Gd-N codoping could significantly improve the photocatalytic activity and thermal stability of TiO₂ [42,47]. In addition to ion doping, the highly optically active heterojunction structures also could effectively reduce recombination of electron-hole pairs, broaden the spectral response range and enhance photocatalytic activity. g-C₃N₄ nanosheets with hybrid engineering architectures exhibit superior chemical stability and photocatalytic decomposition activity of organic pollutants under visible light due to distinct electronics and morphological behavior [22,23,34,41,48,49]. The addition of g-C₃N₄ nanosheet with La-N-TiO₂, as the appropriate band position of La-N-TiO₂ can match well with that of g-C₃N₄ to effectively hinder the recombination of photo-generated carriers, which improves the intrinsic electrical conductivity, results in a good electrochemical performance of g-C₃N₄/La-N-TiO₂ composites.

Hence, in this work, the composite materials, made of graphitic carbon nitride/lanthanum and nitrogen co-doped titanium dioxide (g-C₃N₄/La-N-TiO₂), have been successfully fabricated for the efficient photocatalytic degradation of ciprofloxacin (CIP). A 96.8% degradation of CIP was achieved in 60 min under the conditions of 0.75 g·L⁻¹ CN/La-N-TiO₂, 10 mg·L⁻¹ CIP, and pH of 6.5. It was applied to remove CIP in the collected Danjiangkou Water sample, and the degradation efficiency was close to that of laboratory simulated wastewater. The intermediates and the suggested mineralization pathway of CIP were analyzed and discussed. The findings indicate that CN/La-N-TiO₂ nanocomposite is a green and promising photocatalyst for large-scale applications and would be a candidate for the removal of the emerging antibiotics present in the water environment.

2. Materials and Methods

2.1. Materials

We purchased urea (CH₄N₂O), La(NO₃)₃, dicyandiamideand (99%), and TiCl₄ from Sinopharm Chemical Reagent Co. (Shanghai, China) Ti(OCH(CH₃)₂)₄ and P123 (PEO-PPO-PEO) were purchased by Sigma-Aldrich. The water used throughout all of the experiments was purified using a Millipore system.

2.2. Synthesis of g-C₃N₄/La-N-TiO₂ (CN/La-N-TiO₂) Composite Nanomaterials

g-C₃N₄ was prepared according to a reported procedure [50]. Given amounts of P123, La(NO₃)₃, and urea were mixed in 20 mL anhydrous ethanol, 0.44 mL TiCl₄, and 2.1 mL (CH₃CH₃CHO)₄Ti were added after vigorous stirring for 2 h at room temperature, then it continued to stir for 3 h. The resulting mixture was sealed in a crucible with a cover to aging for 48 h at room temperature. Before drying at 60 °C for 8 h, a certain amount of dicyandiamide (20, 30, 50, 70, or 100 mg) was added. Then, the dried samples were subjected to calcination at the temperature of 580 °C for 5 h with a heating rate of 2.0 °C·min⁻¹. Finally, g-C₃N₄/La-N-TiO₂(CN/La-N-TiO₂) powder products were obtained and ground thoroughly. In control experiments, the samples of different elements doped were synthesized under identical conditions, marked as La-N-TiO₂, CN/N-TiO₂, and CN/La-TiO₂.

2.3. Spectroscopic and Electron Microscopic Characterization

Physicochemical properties of the prepared catalyst were thoroughly characterized as follows: X-ray diffraction (XRD) patterns were recorded on XD-3 X-ray diffraction, fitted with a Cu Kα radiation (λ = 1.54 Å). Fourier transform infrared (-FTIR) spectroscopy was measured on a Bruker VERTEX 70v FTIR Spectrometer for samples embedded in KBr pellets. Surface morphologies and structure features were observed with a field emission scanning electron microscope (FE-SEM, Hitachi s-4800, Tokyo, Japan) at an accelerating voltage of 10 kV, transmission electron microscopic (TEM, FEI-Talos F200S, Hong Kong, China) at an accelerating voltage of 200 kV, and STEM mapping. UV-vis diffuse reflectance spectra (DRS) of the samples were collected on a Perkin Elmer Lambda 950 UV-vis spectrophotometer equipped with an integrating sphere attachment (BaSO₄ as the reference) within the range of 200–1000 nm. The elemental composition and the chemical state of the samples were confirmed by X-ray photoelectron spectroscopy (Thermo Scientific Escalab 250Xi, Waltham, MA, USA) with a Mg Kα source.

2.4. Photocatalytic Tests

The visible light photocatalytic performances of CN/La-N-TiO₂, CN/N-TiO₂, CN/La-TiO₂, and La-N-TiO₂ were evaluated by the degradation of ciprofloxacin in an aqueous solution containing a concentration of (5–20) mg/L ciprofloxacin and 0.75 g/L catalyst in 200 mL glass vessels. The suspension was stirred vigorously in dark for 30 min to achieve adsorption-desorption equilibrium between the catalysts and the ciprofloxacin, and then was irradiated under a 350 W Xe lamp light source (PLS-SXE300, Shanghai Bilang Plant, Shanghai, China) with an optical filter, λ > 420 nm. Sample was taken out periodically of the reactor, then centrifuged at 9000 rpm and analyzed by a UV-vis spectrophotometer (JASCO V-750, Tokyo, Japan). The rate of photocatalytic ciprofloxacin degradations was obtained by recording the changes in absorbance of the reaction solution. HR-MS (Thermo Scientific Q Exactive, Waltham, MA, USA) was used to analyze the byproducts formed during the photocatalytic degradation of CIP.

3. Results and Discussion

3.1. Photocatalysts Characterization

3.1.1. XRD and -FTIR Spectra

The crystal structure of the as-prepared g-C₃N₄, La-N-TiO₂, and CN/La-N-TiO₂ was characterized by XRD, and the results are shown in Figure 1a. The diffraction peaks at 2θ = 25.29°, 37.87°, 48.09°, 53.83°, 55.13°, and 62.81° in the XRD pattern of La-N-TiO₂/CN/La-N-TiO₂ correspond to (101), (004), (200), (105), (211), and (204) diffraction planes of anatase TiO₂ (JCPDS No. 98-4921), correspondingly [34]. The diffraction peak at 2θ = 27.0° (indexed as the (002) peak for graphitic materials) of g-C₃N₄ is attributed to the interplanar stacking of aromatic systems, which slightly shifted to 27.61° following the hybridization of La-N-TiO₂ and g-C₃N₄, revealing the reduction of layer spacing and the formation of denser g-C₃N₄ structures [23,51]. It is obvious that La-N-TiO₂ and CN/La-N-TiO₂ exhibit similar diffraction patterns, except for a characteristic peak at 2θ = 27.0° for CN/La-N-TiO₂, so it can be concluded that g-C₃N₄ was successfully decorated in the composite. The amount of dicyandiamide on the crystal structure of CN/La-N-TiO₂ composite was investigated (Figure S1a, Supporting Information). In Figure S1b, when the amount of dicyanide added is 70 mg, CN/La-N-TiO₂ composite exhibits the strongest photocatalytic activity. In this study, the optimal amount of dicyanide is 70 mg.

FTIR was used to further study the surface functional groups and chemical structures of g-C₃N₄ and CN/La-N-TiO₂ (Figure 1b). The vibration peaks in the range of 1200–1700 cm⁻¹ are attributed to the typical CN heterocycles rings stretching [52]. The characteristic absorption band at 811 cm⁻¹ could be observed, which is commonly ascribed to an s-triazine ring system [53]. The broad band adsorption peak in the range of 3150–3340 cm⁻¹ corresponds to the stretching vibration modes of the terminal NH_x group [54,55]. In the case of CN/La-N-TiO₂, in addition to the above peaks of g-C₃N₄, there are other strong absorption peaks at 500–800 cm⁻¹, which are related with the Ti-O and Ti-O-Ti bands [34].

3.1.2. Morphology Analysis

The surface morphology and structure of representative samples were examined by SEM and TEM (Figure 2). It can be found from Figure 2a that the appearance of these incompletely separated La-N-TiO₂ nanoparticles is spherical-like -nanoparticles with an average particle size of 10–20 nm. Figure 2b shows that the 2D-g-C₃N₄ nanosheets own curved nanosheets with lamellar morphology, and their surface is smooth and soft. As Figure 2c shows, the surface of the g-C₃N₄ nanosheet is covered by the La-N-TiO₂ particles to form a highly uniform composite heterojunction structure of the CN/La-N-TiO₂ composite. The decorated and robustly 2D-g-C₃N₄ attached with the structure of the La-N-TiO₂ nanoparticle to form a 3D nanocomposite photocatalyst g-C₃N₄/La-N-TiO₂(CN/La-N-TiO₂), which can provide a larger surface area, enhance the surface permeability and porous property. The TEM image provides a more evident observation of the two components in Figure 2d. The black and dark gray irregular dots located on the g-C₃N₄ nanosheet were from La-N-TiO₂ nanoparticles. The darker part with a spherical shape should be La-N-TiO₂ and the lighter part is g-C₃N₄. The size of La-N-TiO₂ in the CN/La-N-TiO₂ composite is smaller than that of the pristine La-N-TiO₂, which is due to the confinement of CN against the self-assembly of small component TiO₂ nanoparticles to a certain extent [56]. Two clearly crystal lattice fringes of 0.388 and 0.365 nm can be observed in Figure 2d. The resolved d spacing matched well with the (101) planes of TiO₂.

The TEM elemental mapping images of CN/La-N-TiO₂ are shown in Figure 3a–g, which reveal the uniform and homogenous presence of Ti, O, N, and La elements and further demonstrate that La-N-TiO₂ nanoparticles were evenly dispersed over the g-C₃N₄ nanosheet. It may be beneficial for the interfacial electron transfer between g-C₃N₄/La-N-TiO₂.

The composition and element valence of g-C₃N₄, La-N-TiO₂, and CN/La-N-TiO₂ materials were investigated by XPS measurement (Figure 4). The full survey XPS spectrum (Figure 4a) shows obvious signals of Ti 2p, O 1s, C 1s, N 1s, and La 3d, clearly indicating the existence of C, O, Ti, N, and La in CN/La-N-TiO₂. This result is consistent with the chemical composition, as proven by the XRD and TEM mapping.

3.1.3. XPS Analysis

As can be seen from Figure 4b, binding energy peaks at 458.7 and 464.4 eV in the high-resolution Ti 2p spectrum are assigned to Ti 2p_3/2 and Ti 2p_1/2, respectively, revealing the characteristic value of Ti(IV) oxidation state in TiO₂ [42]. Compared with La-N-TiO₂, there is little change in Ti 2p peaks for the CN/La-N-TiO₂.

The high-resolution O 1s XPS spectrum (Figure 4c) of La-N-TiO₂ is fitted with two peaks centered at 529.89 and 531.93 eV, which correspond to Ti-O-Ti and O-H species, respectively. Three fitted peaks located around 529.88, 530.53, and 531.93 eV in the high-resolution O 1s spectrum of CN/La-N-TiO₂ correspond to Ti-O-Ti, O-H species, and adsorbed molecular oxygen groups, respectively [39]. There is a positive shift of O 1s peaks for the CN/La-N-TiO₂ (from 529.88 to 529.89 eV). This shift is a consequence of the interaction between La-N-TiO₂ and g-C₃N₄ in heterojunctions to increase the effective positive charge on the O species [27].

There are mainly two carbon species presented in the C1s spectra (Figure 4d), which are ascribed to the C-C and/or adventitious carbon at 284.83 eV and sp²-bonded carbon atom in the N-C=N group at 288.33 eV, respectively [23]. The sp²-bonded carbon atom is considered the major carbon in g-C₃N₄.

The N 1s spectra of g-C₃N₄ could be fitted into three peaks (sp²-hybridized N involved in triazine rings (C-N=C) at 396.70 eV, tertiary N group (N-(C)₃) at 398.70 eV and uncondensed amino group (-NH_x) in the aromatic skeleton at 400.30 eV), indicating the presence of sp²-bonded graphitic carbon nitride (g-C₃N₄). The weak peak signal at 403.4 eV is attributed to the charge effect or positive charge localization of the heterocycles. The basic substructure units of g-C₃N₄ polymers [55]. There are two peaks in the XPS spectrum of N1s at 399.70 eV and 400.60 eV for La-N-TiO₂, which suggest two coexisting chemical states of the N in doped TiO₂ substitution doping (N-Ti–O) and N existed in the TiO₂ lattice (Ti–O–N–O) [43]. The N 1s spectra of CN/La-N-TiO₂ could be deconvoluted into four peaks at 398.5 eV, 398.8 eV, 400.1 eV, and 401.3 eV, respectively, corresponding to sp²-hybridized N (C-N=C), tertiary N group (N-(C)₃), uncondensed amino group (-NH_x), and substitution doping nitrogen (N-Ti–O). Compared with g-C₃N₄ and La-N-TiO₂, there is an obvious binding energy shift of the N 1speaks in CN/La-N-TiO₂, indicating that strong interaction exists between g-C₃N₄ and La-N-TiO₂ [42,54].

According to the XPS investigation (as shown in Figure 4f), there are four peaks located at (834.8 eV, 838.40 eV) and (852.35 eV, 855.54 eV), which are attributed to La 3d_5/2 and La 3d_3/2, respectively, confirming that La³⁺ has been successfully doped into CN/La-N-TiO₂ samples. As the radius of La³⁺(1.03 Å) is larger than the radius of Ti⁴⁺ (0.745 Å), La³⁺ can only form the La-O-Ti bond on the surface, and the La-O-La bond on the interstitial site of TiO₂ lattice, instead of entering into the bulk lattice of TiO₂ [41,57].

The atomic ratios of the relative elements, calculated from the XPS spectra, are summarized in

Table 2. Atom concentrations, specific surface area (S_BET), pore size, and band gap of the synthesized samples.

Catalysts	C1/at.%	C2/at.%	Ti/at.%	O/at.%	N/at.%	La/at.%	Surface Area (m2/g)	Av Pore Size(nm)	Pore Volume (cm3/g)
La-N-TiO2	15.71	–	19.51	51.62	1.55	0.6	83 [47]	5.10	0.17
g-C3N4	–	31.298	–	–	66.789	–	50 [42]	–	–
CN/La-N-TiO2	12.64	6.94	19.15	48.69	12.52	0.06	122	8.76	0.28

. The C/N ratios for all cases of heating dicyandiamide are lower than 0.75 for the ideal crystal g-C₃N₄ [23,51,52,53,55]. From Table 2, the g-C₃N₄ nanosheet in this work is the ideal crystal. Combining the experimental results of XRD, SEM, TEM, and XPS, it can be concluded that the chemical combination existed on the interface between La-N-TiO₂ and g-C₃N₄ nanosheet.

3.1.4. BET Surface Area Analysis

In Figure 5a, the N₂ adsorption/desorption isotherm of CN/La-N-TiO₂ shows a typical type IV curve with an H1-type hysteresis loop according to IUPAC classification, confirming the presence of mesopores [42]. Two peaks are observed from the pore distribution curve (Figure 5b), one at 4 nm is attributable to the confined self-assembled La-N-TiO₂ component nanoparticles, and the other broad peak centering at 50 nm is probably formed from gaps between CN nanosheets and La-N-TiO₂, caused by the steric confinement of crumpled and entangled nanosheet structure of seeded CN against La-N-TiO₂ growth [56]. The data of S_BET, total pore volume, and average pore diameter of La-N-TiO₂, g-C₃N₄, and CN/La-N-TiO₂ are summarized in Table 2. Compared with g-C₃N₄ (50 m²·g⁻¹) and La-N-TiO₂ (83 m²·g⁻¹), CN/La-N-TiO₂ possesses the highest surface area of ∼122 m²·g⁻¹ and larger pore size of ∼8.76 nm and the total pore volume of 0.28 cm³·g⁻¹. The higher specific surface areas of the as−prepared hybrids benefit the penetration and adsorption of pollutants on the surface of photocatalysts, which may result in a quick transport of reaction species onto the active sites.

3.1.5. Spectral Properties

The characteristic light absorption spectra of the catalysts are presented in Figure 6a. The absorption edges of La-N-TiO₂ and g-C₃N₄ are 400 nm and 475 nm, respectively. The absorption edges of CN/La-TiO₂ and CN/La-N-TiO₂ redshift to 525 nm and 550 nm. In addition, their light absorption intensities further increased compared with g-C₃N₄ and La-N-TiO₂, which indicates the broadened light absorption range and the enhanced light-harvesting capacity of CN/La-TiO₂ and CN/La-N-TiO₂. Compared with g-C₃N₄, CN/N-TiO₂ shows a weakened light absorption intensity and a narrowed light absorption range, which may be due to the additional amount of g-C₃N₄. The bandgap energy (E_g) is the important parameter for predicting photophysical and photochemical properties of semiconductors [42,58]. The E_g is determined by using the relationship between (αhυ)^1/2 and photon energy and obtained through the linear extrapolation of the intercepts on the x axis (Figure 6b). The E_g of g-C₃N₄, La-N-TiO₂, CN/N-TiO₂, CN/La-TiO₂, and CN/La-N-TiO₂ are thereby estimated to be 2.70 eV, 2.98 eV, 2.77 eV, 2.62 eV, and 2.46 eV, respectively, which is consistent with the redshift of their absorption edges.

3.2. Photocatalytic Performance

Many studies have demonstrated that catalyst dosage, solution pH, and coexisting cations have a considerable impact on the removal efficiency of the pollutant. From an economic perspective and for designing large-scale treatment processes, the determination of optimum parameters for the removal of CIP by CN/La-N-TiO₂ is important. The main influential factors such as pH of the solution, initial CIP concentration, catalyst dosage, and -coexisting ions were investigated in detail and shown in Figure 7a–f.

3.2.1. Effect of Catalyst Dosage

The effect of catalyst dosage was explored and shown in Figure 7a. When the CN/La-N-TiO₂ dosage increased from 0.1 g/L to 1 g/L, the removal efficiency of CIP increased from 51.5% to 96.8%. This is because the number of the active sites on the catalyst surface for radical formation increased by increasing the catalyst dosage, leading to a higher photocatalysis activity. However, when the CN/La-N-TiO₂ dosage was further increased above 0.75 g/L, the removal efficiency of CIP did not significantly vary. For the above-mentioned facts, 150 mg was considered the optimum dosage for the treatment of 200 mL CIP solution with a concentration of 10 mg/L.

3.2.2. Effect of pH

In Figure 8, the zeta potential values measured for CN/La-N-TiO₂ as a function of pH are plotted. It is observed that at pH values around 3 the zeta potential is about +24 mV, but at pH values between 4 and 5, there is a rapid decrease of the zeta potential, the IEP occurring at pH 4.6 [59]. Hence, the catalyst surface is negatively charged at above a value of pH 4.6, while it is positively charged at below a value of pH 4.6. CIP is an amphoteric drug with a pK_a value of 6.1 (for –COOH) and 8.9 (for -NH- and -N- on the piperazinyl ring), where the isoelectric point lies at pH = 7.4 [28,60]. Therefore, CIP zwitterion will present three different ionic forms, mainly depending on the pH value of the solution. As CIP and CN/La-N-TiO₂ are pH sensitive, pH will influence the characteristics of both CIP and CN/La-N-TiO₂ as well as the interactions between them which change as pH changes.

The plots of the effect of pH (2–11.8) on CIP photodegradation with an initial concentration of 10 mg L⁻¹ were studied in presence of CN/La-N-TiO₂ under visible light and shown in Figure 7b. The 0.05 mol·L⁻¹ HCl or 0.05 mol·L⁻¹ NaOH solution was used to adjust the pH value of the simulated CIP wastewater solution. It can be seen from Figure 7b, the degradation is strongly inhibited at highly acidic and basic pH and the degradation efficiency reduces from 96.8% (pH = 6.5) to 35.2% at pH = 2 and 30.5% at pH = 11.8. It is because CIP and CN/La-N-TiO₂ are negatively charged leading to decreased interactions at higher pH, while the generation of ·OH radicals is inhibited in an alkaline medium. It can be concluded that the optimum pH value of the CIP wastewater solution is in the pH range 6–7.

3.2.3. Effect of Initial CIP Concentrations

It can be found in Figure 7c that CIP with an initial concentration of 5 mg/L has been completely removed after 60 min of irradiation, while the relatively stable degradation efficiency is observed when the concentration increases to 10 mg/L. However, a further increase in initial CIP concentration from 10 to 20 mg/L resulted in a decrease in degradation efficiency from 96.8% to 31.2%. This is probably due to the number of surface-active sites on the catalyst provided by 150 mg CN/La-N-TiO₂ are not enough for a higher initial CIP concentration. Additionally, the higher the concentration, the more intermediates are generated. The generated intermediates would compete with the CIP molecules to contact the active sites on the catalyst surface, so as to inhibit the degradation of CIP.

3.2.4. Different Photocatalysts

Under the optimal conditions, where the pH value is 6.5, initial CIP concentration is 10 mg·L⁻¹, and the catalyst dosage is 0.75 g/L, the degradation experiment of CIP was carried out. As shown in Figure 7d, on a blank experiment, the self-photolysis removal of CIP was negligible. Under simulated solar- light irradiation, 15.2 %, 30.2%, 65.4%, 85.8%, and 96.8% of the CIP were degraded in 60 min by g-C₃N₄, La-N-TiO₂, CN/N-TiO₂, CN/La-TiO₂, and CN/La-N-TiO₂, respectively. CN/La-N-TiO₂ shows the highest CIP degradation of 96.8% in 60 min of irradiation followed by CN/La-TiO₂ (85.8%) and CN/N-TiO₂ (65.4%), indicating that the combination among La-N-co-doped TiO₂ and g-C₃N₄ hybrid photocatalyst can significantly enhance the photocatalytic properties.

3.2.5. Effect of Coexisting Cations

Natural water body contains various ions and organics, which would affect the photodegradation of CIP. In this work, removal of CIP by CN/La-N-TiO₂ was tested in presence of Na⁺ (10 mg·L⁻¹), Ca²⁺ (10 mg·L⁻¹), Mg²⁺ (10 mg·L⁻¹), NO₃⁻ (10 mg·L⁻¹), and urea (5 mg·L⁻¹) (Figure 7e). The concentrations of ions and urea used are representative of Danjiangkou’s natural water body. It can be found in Figure 7e that Na⁺, Ca²⁺, and Mg²⁺ have little effect on photocatalytic activity, where NO₃⁻ and urea obviously hindered the photocatalytic degradation of CIP.

3.2.6. Degradation in Water Sample of Dangjiangkou

In order to study the effect of the actual water application of CN/La-N-TiO₂. The degradation experiment of CIP in the water sample of Dangjiangkou was studied. As shown in Figure 7f, there is little reduction in degradation rate compared to laboratory simulated CIP wastewater. Therefore, CN/La-N-TiO₂ has a good practical application prospect.

Furthermore, the stability and reusability of CN/La-N-TiO₂ were evaluated via four successive cycles of CIP photocatalytic degradation experiments, where the removal rate of CIP remained almost unchanged in the cycles (Figure 9a). Meanwhile, to further explore the photocatalytic stability, the XRD patterns of CN/La-N-TiO₂ before and after the cycling experiment were examined (Figure 9b). It can be found that the crystallinity was almost unchanged throughout the 4 cycles. Thus, the synthesized CN/La-N-TiO₂ exhibits good stability and reusability, a property that is very in demand for practical applications.

3.2.7. Proposed Pathways of CIP Degradation

In order to estimate the toxicity of intermediates and better elucidate the mineralization pathway of CIP, the byproducts formed during the 10 min and 30 min -photocatalytic experiment were analyzed by HR-MS. According to the byproducts and a literature study [9,10,12,17,19,61], the probable degradation pathway is conceived and proposed in Figure 10. A high-intensity peak (m/z = 332) is corresponding with the original CIP molecule and the intensity of the CIP peak decreased to 20% after 10 min of degradation reaction in presence of CN/La-N-TiO₂ under visible light. Beyond CIP, six other -byproducts are detected including BP-1 (m/z = 362), BP-5 (m/z = 346), BP-2 (m/z = 306), BP-3 (m/z = 263), BP-4 (m/z = 246), and BP-6 (m/z = 200) (In Figure S2). It can be observed that ··O₂⁻and h⁺ direct attack as the main attacking species to attack the aromatic tertiary amine group, carboxyl group, and the secondary aliphatic amine group in the piperazine ring, which are three principal reactive sites in CIP in this photocatalytic system [62]. Generally, the intermediates formed showed less toxicity than the original chemical, which proved that CN/La-N-TiO₂ is a promising photocatalyst in the CIP wastewater treatment [10,12].

3.2.8. Radical Trapping Experiment and Possible Photocatalytic Mechanism

In order to identify the active oxidative species in the photocatalytic process, trapping experiments are carried out in the CN/La-N-TiO₂ hybrid photocatalytic system. Scheme 1a shows the photocatalytic degradation rate of CN/La-N-TiO₂ in the presence of different scavengers. The photocatalytic efficiency decreases a little by the addition of t-BuOH (TBA, ·OH scavenger, decreases to 90.5%), while it is remarkably suppressed when EDTA-2Na (h⁺ scavenger decreases to 62.6%) and p-Benzoquinone (p-BQ, ·O₂⁻ decreases to 70.8%) were added to the system. Therefore, it can be inferred that the active species affect the photocatalytic activity in the following order: h⁺ > ·O₂⁻ > ·OH [32].

Based on the above results, here a possible photocatalytic mechanism of g-C₃N₄/La-N-TiO₂ composite was discussed and shown in Scheme 1b. The following empirical equation: $E_{VB}=\chi -E_c+0.5E_g$ were also employed to calculate the VB and CB potentials, where χ is the electronegativity of the semiconductor. For g-C₃N₄ and TiO₂, χ is 4.73 and 5.81, respectively [42,45]. E_c is the energy of free electrons based on the hydrogen scale (4.50 eV). The E_VB and E_CB edge potentials of La-N-TiO₂ were calculated to be +2.80 eV and −0.18 eV, respectively, and those of g-C₃N₄ were +1.55 eV and −1.25 eV, respectively.

Under simulated solar light illumination, both g-C₃N₄ and La-N-TiO₂ could be stimulated and generate electron-hole (e⁻-h⁺) pairs [42,45]. Due to the CB potential of g-C₃N₄ (−1.25 eV) being lower than that of La-N-TiO₂ (−0.18 eV), the photo-induced electrons would shift from the CB of g-C₃N₄ to the CB of g-C₃N₄ through the heterointerface. The e⁻ generated in g-C₃N₄ goes to TiO₂, where La³⁺ with 4f orbit can easily capture O₂ adsorbed on the surface of TiO₂ to form superoxide anion (·O₂⁻) [23,41,45,56].

Part of the holes (h⁺) in the VB of TiO₂ can easily be transferred to the VB of g-C₃N₄. Due to the lower reduction potential of the VB of g-C₃N₄ (1.55 eV vs. NHE), h⁺ in the VB of g-C₃N₄ can be used to degrade CIP directly. The remaining h⁺ at the VB potential of La-N-TiO₂ (2.80 V vs. NHE) can oxidize water into hydroxyl radicals (·OH), owing to their potential being higher than the H₂O/·OH potential (2.68 eV vs. NHE) [42]. Thus, during the photocatalytic reaction, h + and ·O₂⁻ are the most important active species.

$$\mathrm{g}-{\mathrm{C}}_3{\mathrm{N}}_4+\mathrm{hv}\to \mathrm{g}-{\mathrm{C}}_3{\mathrm{N}}_4\left({\mathrm{e}}^{-}+{\mathrm{h}}^{+}\right)$$

(1)

$$\mathrm{La}-\mathrm{N}-{\mathrm{TiO}}_2+\mathrm{hv}\to \mathrm{La}-\mathrm{N}-{\mathrm{TiO}}_2\left({\mathrm{e}}^{-}+{\mathrm{h}}^{+}\right)$$

(2)

$${\mathrm{h}}^{+}+{\mathrm{H}}_2\mathrm{O}\to ·\mathrm{OH}$$

(3)

$${\mathrm{La}}^{3+}+{\mathrm{e}}^{-}+{\mathrm{O}}_2\to ·{\mathrm{O}}_2{}^{-}$$

(4)

$${\mathrm{h}}^{+}\left(\mathrm{main}\right)+·{\mathrm{O}}_2{}^{-}\left(\mathrm{main}\right)+·\mathrm{OH}+\mathrm{CIP}\to \mathrm{degradation}\mathrm{products}$$

(5)

4. Conclusions

A 3D nanocomposite photocatalyst g-C₃N₄/La-N-TiO₂ was successfully synthesized by a simple and reproducible in-situ synthetic method. g-C₃N₄/La-N-TiO₂ exhibits high photocatalytic activity for the degradation of ciprofloxacin. A 96.8% degradation of CIP was achieved in 60 min under the conditions of 0.75 g·L⁻¹ CN/La-N-TiO₂, 10 mg·L⁻¹ CIP, and pH of 6.5. The excellent photocatalytic performance of CN/La-N-TiO₂ on degrading CIF is attributed to the formation of a junction that efficiently improves charge separation and wide spectrum response. The removal rates were rarely affected by Na⁺ (10 mg·L⁻¹), Ca²⁺ (10 mg·L⁻¹), Mg²⁺ (10 mg·L⁻¹), and urea (5 mg·L⁻¹), where the concentrations of ions and urea used are representative of Danjiangkou natural water body. In addition, the intermediates formed showed less toxicity than the original chemical, which proved that CN/La-N-TiO₂ nanocomposite is a green and promising photocatalyst for large-scale applications and would be a candidate for the removal of the emerging antibiotics present in the water environment.