Nb/N Co-Doped Layered Perovskite Sr2TiO4: Preparation and Enhanced Photocatalytic Degradation Tetracycline under Visible Light

Sr2TiO4 is a promising photocatalyst for antibiotic degradation in wastewater. The photocatalytic performance of pristine Sr2TiO4 is limited to its wide bandgap, especially under visible light. Doping is an effective strategy to enhance photocatalytic performance. In this work, Nb/N co-doped layered perovskite Sr2TiO4 (Sr2TiO4:N,Nb) with varying percentages (0–5 at%) of Nb were synthesized by sol-gel and calcination. Nb/N co-doping slightly expanded the unit cell of Sr2TiO4. Their photocatalytic performance towards antibiotic (tetracycline) was studied under visible light (λ > 420 nm). When Nb/(Nb + Ti) was 2 at%, Sr2TiO4:N,Nb(2%) shows optimal photocatalytic performance with the 99% degradation after 60 min visible light irradiation, which is higher than pristine Sr2TiO4 (40%). The enhancement in photocatalytic performance is attributed to improving light absorption, and photo-generated charges separation derived from Nb/N co-doping. Sr2TiO4:N,Nb(2%) shows good stability after five cycles photocatalytic degradation reaction. The capture experiments confirm that superoxide radical is the leading active species during the photocatalytic degradation process. Therefore, the Nb/N co-doping in this work could be used as an efficient strategy for perovskite-type semiconductor to realize visible light driving for wastewater treatment.


Introduction
Tetracycline (TC) is a widely used broad-spectrum antibiotic in medical treatment, animal husbandry, and aquaculture [1]. The accumulation of antibiotics in the environment leads to drug-resistant bacteria, which could bring serious threats to the ecological environment and human health [2][3][4][5]. Tetracycline is not only hard to self-degrade in the natural environment [3], but also difficult to be eliminated by conventional techniques.
Photocatalytic technology [6][7][8][9] utilizes a photocatalyst to generate light-generated holes and electrons under irradiation. Holes or electrons react with H 2 O or O 2 to generate superoxide radicals and hydroxyl radicals with strong oxidation capacity, thereby thoroughly oxidizing and degrading organic pollutants [10] in wastewater. The core of photocatalytic oxidation technology are photocatalysts. Among many photocatalytic materials, strontium titanate has attracted wide attention because of its advantages, such as being inexpensive, pollution-free, and light corrosion resistant [11]. However, strontium titanate also has the disadvantages of having a large bandgap and high photo-generated carrier recombination rate. Layered perovskite Sr 2 TiO 4 exhibits higher photocatalytic performance than perovskite SrTiO 3 , due to the particularly layered crystal structure and typical 2D charge transportation properties [12,13]. The wide bandgap of Sr 2 TiO 4 exhibits photocatalytic activity only under ultraviolet light, which gravely affects further enhancement of photocatalytic performance. Ion doping can reduce the bandgap [14] of the perovskite material [11], including single metal doping, nonmetal doping, and co-doping. For instance, by introducing Cr [15], Ag [16], F [17], chalcogens [18], La/N [13], Cr/F [19], La/Rh [20], and La/Fe [12], the light response range of Sr 2 TiO 4 is extended. Properties of Sr 2 TiO 4 doping with different ions was shown in Table 1. Among them, co-doping has gained more interest, because co-doping is expected to maintain the charge-balance without forming oxygen vacancies.  [12] Among various doping elements for oxide semiconductors, doping the nitrogen at the O site can narrow the band gap through the hybridization N 2p with the O 2p [21]. Meanwhile, N-doped oxide semiconductors are often accompanied by oxygen vacancies [22] due to charge compensation. Too high of a concentration of oxygen vacancies will act as a charge carrier recombination center, which is detrimental to photocatalytic activity [16,23,24]. The preparation methods of Sr 2 TiO 4 mainly includes conventional solid-state reactions, the molten-salt method, and sol-gel method. Among them, the sol-gel synthesis process can achieve molecular-level doping.
This study chooses N-doped Sr 2 TiO 4 , where N 3− replaces O 2− to regulate the band structure. In order to control the concentration of oxygen vacancies, we replace Ti 4+ (r = 0.061 nm) ions with Nb 5+ (r = 0.064 nm) to balance the negative charge introduced by the unequal substitution of N 3− and O 2− Nb/N co-doped Sr 2 TiO 4 , which perhaps obtains charge-balanced Sr 2 TiO 4 :N,Nb and avoids too high of a concentration of oxygen vacancies. The amount of the dopant should also be optimized. Hence, we perform an investigation on Nb/N co-doped layered perovskite Sr 2 TiO 4 for photocatalytic degradation tetracycline by varying Nb 5+ doping content. To the authors' knowledge, although there have been several studies on the doping modification of Sr 2 TiO 4 [12,13,[15][16][17]19,20] for photocatalytic hydrogen production, there is no literature report on the preparation and photocatalytic degradation performance of Nb/N co-doped Sr 2 TiO 4 . In this work, Sr 2 TiO 4 :N,Nb(2%) shows the best photocatalytic degradation performance towards tetracycline under visible light. The superior photocatalytic performance can be attributed to N-doping, narrowing the bandgap and Nb-doping compensating charge imbalance. This work affords a new insight to realizing visible-light-driven perovskite-type semiconductors.

Materials Characterization
As shown in Figure 1, the crystal structure of Sr 2 TiO 4 , Sr 2 TiO 4 :N and Sr 2 TiO 4 :N,Nb with different Nb-doping content were investigated by X-ray diffraction (XRD). There was only one phase of Sr 2 TiO 4 (JCPDS NO:39-1471) without preferred growth orientation in the XRD patterns. Characteristic peaks of 2θ at 23. cording to the XRD patterns, it can be concluded that both Nb 5+ and N 3− were doped into the lattice of Sr2TiO4 successfully. The average crystallite sizes were calculated by Scherrer's equation [14] of the Sr2TiO4 (103) XRD reflection and were shown in Table 2.  The morphological characterization of Sr2TiO4 and Sr2TiO4:N,Nb(2%) were shown in Figure 2. After being calcined at 1000 °C for 4 h, both pristine Sr2TiO4 and Sr2TiO4:N,Nb(2%) were composed of irregular particles ranging from tens of nanometers to 1 micron.  After N-doping, the peak of (110) shown by Sr 2 TiO 4 :N in XRD shifted to a lower angle, meaning larger d-spacing than that of Sr 2 TiO 4 , and this larger spacing proved that N-doped Sr 2 TiO 4 successfully. This is due to the N 3− radius (0.146 nm) being slightly larger than the O 2− radius (0.140 nm); when N 3− enters the Sr 2 TiO 4 crystal lattice, the crystal lattice expands. According to Bragg's equation 2dsinθ = λ, with the increase of d-spacing, the diffraction angle 2θ shifts to a lower angle. Similar to Sr 2 TiO 4 :N, 2θ of Sr 2 TiO 4 :N,N shifted to a lower direction gradually with the increase of Nb-doping content, as shown in Figure 1. Because the Nb 5+ radius (0.064 nm) and N 3− radius (0.146 nm) are slightly larger than the Ti 4+ radius (0.061 nm) and O 2− radius (0.140 nm), respectively, the more the Nb-doping content, the more the lattice expansion [25,26], and the more diffraction angle offset. According to the XRD patterns, it can be concluded that both Nb 5+ and N 3− were doped into the lattice of Sr 2 TiO 4 successfully. The average crystallite sizes were calculated by Scherrer's equation [14] of the Sr 2 TiO 4 (103) XRD reflection and were shown in Table 2. The morphological characterization of Sr 2 TiO 4 and Sr 2 TiO 4 :N,Nb(2%) were shown in Figure 2. After being calcined at 1000 • C for 4 h, both pristine Sr 2 TiO 4 and Sr 2 TiO 4 :N,Nb(2%) were composed of irregular particles ranging from tens of nanometers to 1 micron. After N-doping, the peak of (110) shown by Sr2TiO4:N in XRD shifted to a lower angle, meaning larger d-spacing than that of Sr2TiO4, and this larger spacing proved that Ndoped Sr2TiO4 successfully. This is due to the N 3− radius (0.146 nm) being slightly larger than the O 2− radius (0.140 nm); when N 3− enters the Sr2TiO4 crystal lattice, the crystal lattice expands. According to Bragg's equation 2dsinθ = λ, with the increase of d-spacing, the diffraction angle 2θ shifts to a lower angle. Similar to Sr2TiO4:N, 2θ of Sr2TiO4:N,N shifted to a lower direction gradually with the increase of Nb-doping content, as shown in Figure  1. Because the Nb 5+ radius (0.064 nm) and N 3− radius (0.146 nm) are slightly larger than the Ti 4+ radius (0.061 nm) and O 2− radius (0.140 nm), respectively, the more the Nb-doping content, the more the lattice expansion [25,26], and the more diffraction angle offset. According to the XRD patterns, it can be concluded that both Nb 5+ and N 3− were doped into the lattice of Sr2TiO4 successfully. The average crystallite sizes were calculated by Scherrer's equation [14] of the Sr2TiO4 (103) XRD reflection and were shown in Table 2.  The morphological characterization of Sr2TiO4 and Sr2TiO4:N,Nb(2%) were shown in Figure 2. After being calcined at 1000 °C for 4 h, both pristine Sr2TiO4 and Sr2TiO4:N,Nb(2%) were composed of irregular particles ranging from tens of nanometers to 1 micron.    Figure 4a shows the absorbance of Sr2TiO4, Sr2TiO4:N, and Sr2TiO4:N,Nb. The UV-Vis absorption spectra of the pristine and doped Sr2TiO4 showed a significant difference. The absorption edge of pristine Sr2TiO4 is located in the ultraviolet region. However, the absorption edge of N doping and Nb/N co-doping Sr2TiO4 were significantly red and shifted into the visible light region, and Sr2TiO4:N,Nb(2%) showed the strongest absorption capacity of visible light. According to the previous literature, N doping can reduce the band gap through hybridization of N 2p with the O 2p, raising the valence band maximum. The energy level of Nb 5+ is similar to titanium 3d orbital energy, so Nb can replace Ti and mix Nb 4d with Ti 3d orbitals without lowering the conduction band minimum [27]. The same observation has been seen in other metal oxides doped with nitrogen, such as N-doped SrTiO3 [28], N-doped NaLaTiO4 [29], N-doped Sr2TiO4 [29], Nb/N co-doped TiO2 [27], and so on. The bandgap Eg deduced from the Tauc plots in Figure 4b were 3.48 eV, 3.02 eV, and 2.95 eV for pristine Sr2TiO4, Sr2TiO4:N, and Sr2TiO4:N,Nb(2%), respectively. Pristine Sr2TiO4 had a wide bandgap (3.48 eV), which was consistent with the previous literature [13]. N doping and Nb/N co-doping can reduce the bandgap of Sr2TiO4. Similar phenomena have also been observed in other materials involving N doping [13,24,30]. The chemical state and chemical composition on the surfaces of the pristine and doped Sr2TiO4 were investigated by XPS. The complete measurement scan in Figure 5a showed the presence of Sr, Ti, O, N (except pristine Sr2TiO4), and Nb (only for Sr2TiO4:N,Nb(2%)). Figure 5b shows the fine spectra of Ti. The presence of Ti 4+ was validated by peaks at 458 and 464 eV, which were attributable to Ti 2p3/2 and Ti 2p1/2 [15,20].  The energy level of Nb 5+ is similar to titanium 3d orbital energy, so Nb can replace Ti and mix Nb 4d with Ti 3d orbitals without lowering the conduction band minimum [27]. The same observation has been seen in other metal oxides doped with nitrogen, such as N-doped SrTiO 3 [28], N-doped NaLaTiO 4 [29], N-doped Sr 2 TiO 4 [29], Nb/N co-doped TiO 2 [27], and so on. The bandgap E g deduced from the Tauc plots in Figure 4b were 3.48 eV, 3.02 eV, and 2.95 eV for pristine Sr 2 TiO 4 , Sr 2 TiO 4 :N, and Sr 2 TiO 4 :N,Nb(2%), respectively. Pristine Sr 2 TiO 4 had a wide bandgap (3.48 eV), which was consistent with the previous literature [13]. N doping and Nb/N co-doping can reduce the bandgap of Sr 2 TiO 4 . Similar phenomena have also been observed in other materials involving N doping [13,24,30].   Figure 4a shows the absorbance of Sr2TiO4, Sr2TiO4:N, and Sr2TiO4:N,Nb. The UV-Vis absorption spectra of the pristine and doped Sr2TiO4 showed a significant difference. The absorption edge of pristine Sr2TiO4 is located in the ultraviolet region. However, the absorption edge of N doping and Nb/N co-doping Sr2TiO4 were significantly red and shifted into the visible light region, and Sr2TiO4:N,Nb(2%) showed the strongest absorption capacity of visible light. According to the previous literature, N doping can reduce the band gap through hybridization of N 2p with the O 2p, raising the valence band maximum. The energy level of Nb 5+ is similar to titanium 3d orbital energy, so Nb can replace Ti and mix Nb 4d with Ti 3d orbitals without lowering the conduction band minimum [27]. The same observation has been seen in other metal oxides doped with nitrogen, such as N-doped SrTiO3 [28], N-doped NaLaTiO4 [29], N-doped Sr2TiO4 [29], Nb/N co-doped TiO2 [27], and so on. The bandgap Eg deduced from the Tauc plots in Figure 4b were 3.48 eV, 3.02 eV, and 2.95 eV for pristine Sr2TiO4, Sr2TiO4:N, and Sr2TiO4:N,Nb(2%), respectively. Pristine Sr2TiO4 had a wide bandgap (3.48 eV), which was consistent with the previous literature [13]. N doping and Nb/N co-doping can reduce the bandgap of Sr2TiO4. Similar phenomena have also been observed in other materials involving N doping [13,24,30]. The chemical state and chemical composition on the surfaces of the pristine and doped Sr2TiO4 were investigated by XPS. The complete measurement scan in Figure 5a showed the presence of Sr, Ti, O, N (except pristine Sr2TiO4), and Nb (only for Sr2TiO4:N,Nb(2%)). Figure 5b shows the fine spectra of Ti. The presence of Ti 4+ was validated by peaks at 458 and 464 eV, which were attributable to Ti 2p3/2 and Ti 2p1/2 [15,20]. The chemical state and chemical composition on the surfaces of the pristine and doped Sr 2 TiO 4 were investigated by XPS. The complete measurement scan in Figure 5a showed the presence of Sr, Ti, O, N (except pristine Sr 2 TiO 4 ), and Nb (only for Sr 2 TiO 4 :N,Nb(2%)). Figure 5b shows the fine spectra of Ti. The presence of Ti 4+ was validated by peaks at 458 and 464 eV, which were attributable to Ti 2p 3/2 and Ti 2p 1/2 [15,20]. The Ti 2p signal was gradually weakened by simultaneous Nb/N doping, confirming the substitution of Ti with Nb. A similar phenomenon was also observed in La/Fe co-doped Sr 2 TiO 4 [12]. The peaks located around 529 eV and 531 eV belonged to the lattice oxygen and surface OH − groups [31], respectively ( Figure 5c). The signals of lattice oxygen gradually diminished along with N doping, meaning the substitution of lattice oxygen with N 3− for Sr 2 TiO 4 :N and Sr 2 TiO 4 :N,Nb(2%), which was consistent with XRD results (Figure 1). Due to the strong signal of surface OH − , it can be inferred that all samples are hydrophilic [13]. The peak around 398 eV was assigned to lattice N 3− as shown in Figure 5d. Along with Nb/N co-doping, this signal of N 1s was enhanced, which demonstrated the introduction of Nb promoting N doping. The 3d 5/2 and 3d 5/2 of Nb 3d electrons were detected at binding energies of 207 eV and 209 eV [32], indicating the existence of Nb 5+ . It was worth noticing that binding-energy shifted towards lower binding energy after Nb/N co-doping, which suggests Nb 5+ doping based on N 3− doping was hole-doping. A similar shift was observed in SrCuO 2 [33]. The Ti 2p signal was gradually weakened by simultaneous Nb/N doping, confirming the substitution of Ti with Nb. A similar phenomenon was also observed in La/Fe co-doped Sr2TiO4 [12]. The peaks located around 529 eV and 531 eV belonged to the lattice oxygen and surface OH − groups [31], respectively ( Figure 5c). The signals of lattice oxygen gradually diminished along with N doping, meaning the substitution of lattice oxygen with N 3− for Sr2TiO4:N and Sr2TiO4:N,Nb(2%), which was consistent with XRD results( Figure  1). Due to the strong signal of surface OH − , it can be inferred that all samples are hydrophilic [13]. The peak around 398 eV was assigned to lattice N 3− as shown in Figure 5d. Along with Nb/N co-doping, this signal of N 1s was enhanced, which demonstrated the introduction of Nb promoting N doping. The 3d5/2 and 3d5/2 of Nb 3d electrons were detected at binding energies of 207 eV and 209 eV [32], indicating the existence of Nb 5+ . It was worth noticing that binding-energy shifted towards lower binding energy after Nb/N co-doping, which suggests Nb 5+ doping based on N 3− doping was hole-doping. A similar shift was observed in SrCuO2 [33]. The charge carrier behaviors of the pristine Sr2TiO4 and Sr2TiO4:N,Nb with different Nb-doping content were examined by PL with an excitation wavelength of 325 nm at room temperature. As shown in Figure 6, the emission intensity of Nb/N co-doped Sr2TiO4 with different Nb doping content were lower than that of pristine Sr2TiO4, implying that the charge separation capability was enhanced significantly, especially for Sr2TiO4:N,Nb(2%). The negative charge (O vacancy) introduced by the unequal replacement of N 3− and O 2− can be balanced by Nb-doping. Therefore, Nb/N co-doping Sr2TiO4 can reduce the photoelectron-hole recombination probability, which was beneficial to the photocatalytic reaction. Nb 5+ , as a charge balancing agent, can compensate for a charge imbalance caused by N substitution of O,  The charge carrier behaviors of the pristine Sr 2 TiO 4 and Sr 2 TiO 4 :N,Nb with different Nb-doping content were examined by PL with an excitation wavelength of 325 nm at room temperature. As shown in Figure 6, the emission intensity of Nb/N co-doped Sr 2 TiO 4 with different Nb doping content were lower than that of pristine Sr 2 TiO 4 , implying that the charge separation capability was enhanced significantly, especially for Sr 2 TiO 4 :N,Nb(2%). The negative charge (O vacancy) introduced by the unequal replacement of N 3− and O 2− can be balanced by Nb-doping. Therefore, Nb/N co-doping Sr 2 TiO 4 can reduce the photoelectron-hole recombination probability, which was beneficial to the photocatalytic reaction. Nb 5+ , as a charge balancing agent, can compensate for a charge imbalance caused by N substitution of O, Nb • Ti + N O , just as La played a charge balancing role in La/N co-doped Sr 2 TiO 4 [13], La • Sr + N O .

Photoelectrochemical Properties of Sr2TiO4:N,Nb
To further confirm the effect of Nb doping on photo-carriers separation and migration of Sr2TiO4:N,Nb(2%), photochemical measurements were performed. As shown in Figure Figure 7b shows the electrochemical impedance spectroscopy (EIS) of Sr2TiO4, Sr2TiO4:N, and Sr2TiO4:N,Nb(2%) and the equivalent circuit, which can further verify the charge transport and transfer. It is well known that the smaller resistance radius indicates the higher separation and migration of electrons and holes. Compared with Sr2TiO4, Sr2TiO4:N, and Sr2TiO4:N,Nb(2%) had a smaller resistance radius, indicating more efficient separation of photocarriers. The result was consistent with the PL spectra and transient photocurrent response. The enhancement of the photocurrent density was attributed to the sufficient light absorption and effective photocarriers separation by introducing Nb 5+ based on N 3− doping.
The specific surface area was tested by an N2 adsorption-desorption isotherm at 77 K. As shown in Figure 8, both Sr2TiO4 and Sr2TiO4:N,Nb(2%) had typical IV isotherms and H3 hysteresis loops. H3 type hysteresis loops can be observed virtually on adsorbents

Photoelectrochemical Properties of Sr2TiO4:N,Nb
To further confirm the effect of Nb doping on photo-carriers separation and migration of Sr2TiO4:N,Nb(2%), photochemical measurements were performed. As shown in Figure 7a Figure 7b shows the electrochemical impedance spectroscopy (EIS) of Sr2TiO4, Sr2TiO4:N, and Sr2TiO4:N,Nb(2%) and the equivalent circuit, which can further verify the charge transport and transfer. It is well known that the smaller resistance radius indicates the higher separation and migration of electrons and holes. Compared with Sr2TiO4, Sr2TiO4:N, and Sr2TiO4:N,Nb(2%) had a smaller resistance radius, indicating more efficient separation of photocarriers. The result was consistent with the PL spectra and transient photocurrent response. The enhancement of the photocurrent density was attributed to the sufficient light absorption and effective photocarriers separation by introducing Nb 5+ based on N 3− doping.
The specific surface area was tested by an N2 adsorption-desorption isotherm at 77 K. As shown in Figure 8, both Sr2TiO4 and Sr2TiO4:N,Nb(2%) had typical IV isotherms and H3 hysteresis loops. H3 type hysteresis loops can be observed virtually on adsorbents  Figure 7b shows the electrochemical impedance spectroscopy (EIS) of Sr 2 TiO 4 , Sr 2 TiO 4 :N, and Sr 2 TiO 4 :N,Nb(2%) and the equivalent circuit, which can further verify the charge transport and transfer. It is well known that the smaller resistance radius indicates the higher separation and migration of electrons and holes. Compared with Sr 2 TiO 4 , Sr 2 TiO 4 :N, and Sr 2 TiO 4 :N,Nb(2%) had a smaller resistance radius, indicating more efficient separation of photocarriers. The result was consistent with the PL spectra and transient photocurrent response. The enhancement of the photocurrent density was attributed to the sufficient light absorption and effective photocarriers separation by introducing Nb 5+ based on N 3− doping.
The specific surface area was tested by an N 2 adsorption-desorption isotherm at 77 K. As shown in Figure 8, both Sr 2 TiO 4 and Sr 2 TiO 4 :N,Nb(2%) had typical IV isotherms and H3 hysteresis loops. H3 type hysteresis loops can be observed virtually on adsorbents with a lamellar structure. At high specific pressure, the curves showed an obvious hysteresis loop, which indicated that there was a mesoporous structure in the samples. The average pore diameter was determined using the Barrett-Joyner-Halenda (BJH) methods. BJT desorption average pore diameter (4 V/A) of Sr 2 TiO 4 and Sr 2 TiO 4 :N,Nb(2%) were 3824 nm, and 23.22 nm, respectively. The specific surface areas of Sr 2 TiO 4 and Sr 2 TiO 4 :N,Nb(2%) were 1.213 m 2 /g and 6.718 m 2 /g, respectively, which were very low. Specific surface area was not the main reason for the improvement of the photocatalytic degradation performance of Sr 2 TiO 4 :N,Nb(2%). with a lamellar structure. At high specific pressure, the curves showed an obvious hysteresis loop, which indicated that there was a mesoporous structure in the samples. The average pore diameter was determined using the Barrett-Joyner-Halenda (BJH) methods. BJT desorption average pore diameter (4 V/A) of Sr2TiO4 and Sr2TiO4:N,Nb(2%) were 38.24 nm and 23.22 nm, respectively. The specific surface areas of Sr2TiO4 and Sr2TiO4:N,Nb(2%) were 1.213 m 2 /g and 6.718 m 2 /g, respectively, which were very low. Specific surface area was not the main reason for the improvement of the photocatalytic degradation performance of Sr2TiO4:N,Nb(2%).

Photocatalytic Degradation of TC
As shown in Figure 9a, the adsorption capacity of photocatalysts towards TC after stirring for 30 min in the dark shows that Sr2TiO4:N,Nb(2%) exhibited the best adsorption property among them. However, the adsorption capacity was low, which was consistent with BET results. The photocatalytic properties of Sr2TiO4, Sr2TiO4:N, and Sr2TiO4:N,Nb(2%) were investigated by photocatalytic degradation of TC with a 1 g·L −1 photocatalyst under visible light. A blank experiment without a photocatalyst was also carried out to exclude the impact of photolysis. The self-degradation rate was 3.6% after 60 min due to negligible photooxygenation reactions in the presence of dissolved oxygen (Figure 9b) [34,35]. As shown in Figure 9b, the photocatalytic performance of all Nb/N codoping Sr2TiO4 exhibited a higher degradation rate than prime Sr2TiO4 under the same conditions. The photocatalytic degradation rate of Sr2TiO4:N,Nb increased as Nb-doping content increased from 0% to 2%. When the content of Nb exceeded 2%, the degradation rate of Sr2TiO4:N,Nb decreased, which may be due to the introduction of new defects caused by excessive Nb, ig.

' Ti Ti
Ti Nb + • . Sr2TiO4:N,Nb(2%) exhibited the best photocatalytic degradation rate (99%), which was higher than that of prime Sr2TiO4 (40%) and Sr2TiO4:N (94%). On the one hand, Nb/N co-doping improved the light absorption in the visiblelight region. On the other hand, the recombination probability of Sr2TiO4:N,Nb decreased, especially Sr2TiO4:N,Nb(2%). For comparison, the degradation rates of Sr2TiO4 towards different organic pollutants are shown in Table 3. To the authors' knowledge, the photocatalytic degradation performance of N-doped Sr2TiO4 and Nb/N co-doped Sr2TiO4 has not been reported.

Photocatalytic Degradation of TC
As shown in Figure 9a, the adsorption capacity of photocatalysts towards TC after stirring for 30 min in the dark shows that Sr 2 TiO 4 :N,Nb(2%) exhibited the best adsorption property among them. However, the adsorption capacity was low, which was consistent with BET results. The photocatalytic properties of Sr 2 TiO 4 , Sr 2 TiO 4 :N, and Sr 2 TiO 4 :N,Nb(2%) were investigated by photocatalytic degradation of TC with a 1 g·L −1 photocatalyst under visible light. A blank experiment without a photocatalyst was also carried out to exclude the impact of photolysis. The self-degradation rate was 3.6% after 60 min due to negligible photooxygenation reactions in the presence of dissolved oxygen (Figure 9b) [34,35]. As shown in Figure 9b, the photocatalytic performance of all Nb/N co-doping Sr 2 TiO 4 exhibited a higher degradation rate than prime Sr 2 TiO 4 under the same conditions. The photocatalytic degradation rate of Sr 2 TiO 4 :N,Nb increased as Nb-doping content increased from 0% to 2%. When the content of Nb exceeded 2%, the degradation rate of Sr 2 TiO 4 :N,Nb decreased, which may be due to the introduction of new defects caused by excessive Nb, ig. Nb • Ti + Ti Ti . Sr 2 TiO 4 :N,Nb(2%) exhibited the best photocatalytic degradation rate (99%), which was higher than that of prime Sr 2 TiO 4 (40%) and Sr 2 TiO 4 :N (94%). On the one hand, Nb/N co-doping improved the light absorption in the visible-light region. On the other hand, the recombination probability of Sr 2 TiO 4 :N,Nb decreased, especially Sr 2 TiO 4 :N,Nb(2%). For comparison, the degradation rates of Sr 2 TiO 4 towards different organic pollutants are shown in Table 3. To the authors' knowledge, the photocatalytic degradation performance of N-doped Sr 2 TiO 4 and Nb/N co-doped Sr 2 TiO 4 has not been reported.
The photocatalytic degradation and fitting results of the TC kinetics followed pseudofirst-order kinetics, as shown in Figure 9c. Sr 2 TiO 4 :N,Nb(2%) showed a maximum apparent reaction rate constant of 0.0655 min −1 which was nearly 24.9 times that of Sr 2 TiO 4 (0.00263 min −1 ) and just 1.25 times that of Sr 2 TiO 4 :N (0.0522 min −1 ). This indicated that N doping had the main effect whereas adding Nb yielded only a slight increase to photocatalytic degradation rates. The maximum apparent reaction rate constant obtained by Sr 2 TiO 4 :N,Nb(2%) was attributed to an appropriate Nb/N co-doping concentration. The photocatalytic degradation and fitting results of the TC kinetics followed pseudo-first-order kinetics, as shown in Figure 9c. Sr2TiO4:N,Nb(2%) showed a maximum apparent reaction rate constant of 0.0655 min −1 which was nearly 24.9 times that of Sr2TiO4 (0.00263 min −1 ) and just 1.25 times that of Sr2TiO4:N (0.0522 min −1 ). This indicated that N doping had the main effect whereas adding Nb yielded only a slight increase to photocatalytic degradation rates. The maximum apparent reaction rate constant obtained by Sr2TiO4:N,Nb(2%) was attributed to an appropriate Nb/N co-doping concentration.
To examine the stability and reusability of Sr2TiO4:N,Nb(2%), five cycles of photodegradation experiments were carried out. After each cycle, the photocatalyst was centrifuged (at 9000 rpm) and washed several times with deionized water for several times and re-placed in a deionized water solution containing fresh TC. As shown in Figure 9d, the degradation rate of Sr2TiO4:N,Nb(2%) from 99% to 90% after five repeated cycles without an obvious decrease, indicating that Sr2TiO4:N,Nb(2%) exhibited relatively good stability and repeatability. The decrease of degradation performance could be ascribed to the mass loss in the centrifugation and washing process. Sr2TiO4:N,Nb(2%) shows good potential in photocatalytic degradation of antibiotic pollutants.
To further understand the key active species in the photocatalytic process of TC degradation, active species trapping experiments were performed, as shown in Figure 10  To examine the stability and reusability of Sr 2 TiO 4 :N,Nb(2%), five cycles of photodegradation experiments were carried out. After each cycle, the photocatalyst was centrifuged (at 9000 rpm) and washed several times with deionized water for several times and re-placed in a deionized water solution containing fresh TC. As shown in Figure 9d, the degradation rate of Sr 2 TiO 4 :N,Nb(2%) from 99% to 90% after five repeated cycles without an obvious decrease, indicating that Sr 2 TiO 4 :N,Nb(2%) exhibited relatively good stability and repeatability. The decrease of degradation performance could be ascribed to the mass loss in the centrifugation and washing process. Sr 2 TiO 4 :N,Nb(2%) shows good potential in photocatalytic degradation of antibiotic pollutants.
To further understand the key active species in the photocatalytic process of TC degradation, active species trapping experiments were performed, as shown in Figure 10. Isopropyl alcohol (IPA, 0.01 M), triethanolamine (TEOA, 0.01 M), p-benzoquinone (BQ, 0.005 M), and AgNO 3 (0.01M) were added as scavengers for hydroxyl radical (·OH), hole (h + ), superoxide radical (·O − 2 ), and photogenerated electrons (e − ) to TC solution at the presence of Sr 2 TiO 4 :N,Nb(2%), respectively. When the IPA, TEOA, and AgNO 3 was added, the photodegradation rate of TC only slightly decreased. However, when BQ was added, the degradation rate of TC reduced greatly, indicating that superoxide radical (·O − 2 ) was the main active species during the photocatalytic degradation of TC.

Improvement Mechanism of Photocatalytic Performance
The effect of N doping and Nb/N co-doping on the valence bands of Sr 2 TiO 4 was evaluated by XPS valence band (VB) spectra. As shown in Figure 11 [16,17,20,41], and the conduction band minimum of N-doped Sr 2 TiO 4 and Nb/N co-doped Sr 2 TiO 4 have not been reported.

Improvement Mechanism of Photocatalytic Performance
The effect of N doping and Nb/N co-doping on the valence bands of Sr2TiO4 was evaluated by XPS valence band (VB) spectra. As shown in Figure 11 [16,17,20,41], and the conduction band minimum of N-doped Sr2TiO4 and Nb/N co-doped Sr2TiO4 have not been reported.  According to our experimental results of XPS valence band spectra ( Figure 11) and UV-Vis spectra (Figure 4b), schematic band structures of Sr 2 TiO 4 , Sr 2 TiO 4 :N, and Sr 2 TiO 4 :N,Nb(2%) were illustrated in Figure 12. The pure Sr 2 TiO 4 had a wide bandgap. In contrast, Sr 2 TiO 4 :N displayed visible light response derived from N orbital. The state of the oxygen vacancies (accompanied by N-doping) was located below the conduction band minimum, which was consistent with previous literature about SrTiO 3 co-doped with N and La [21]. Regarding Sr 2 TiO 4 :N,Nb(2%), Nb/N co-doping uplifted the valence band furthermore. As we know, the more negative the valance band, the higher reduction ability of photogenerated electrons to generate superoxide radicals (O − 2 ). The conduction band potential of Sr 2 TiO 4 :N,Nb(2%) was more negative than Sr 2 TiO 4 , Sr 2 TiO 4 :N, and O 2 /O − 2 (−0.28 eV), so electrons can more easily transfer to the oxygen molecules, producing O − 2 . The uplift of conduction band after cationic doping had also been observed in Ag doping Sr 2 TiO 4 [16]. According to our experimental results of XPS valence band spectra ( Figure 11) and UV-Vis spectra (Figure 4b), schematic band structures of Sr2TiO4, Sr2TiO4:N, and Sr2TiO4:N,Nb(2%) were illustrated in Figure 12. The pure Sr2TiO4 had a wide bandgap. In contrast, Sr2TiO4:N displayed visible light response derived from N orbital. The state of the oxygen vacancies (accompanied by N-doping) was located below the conduction band minimum, which was consistent with previous literature about SrTiO3 co-doped with N and La [21]. Regarding Sr2TiO4:N,Nb(2%), Nb/N co-doping uplifted the valence band furthermore. As we know, the more negative the valance band, the higher reduction ability of photo-generated electrons to generate superoxide radicals (▪O ). The conduction band potential of Sr2TiO4:N,Nb(2%) was more negative than Sr2TiO4, Sr2TiO4:N, and O 2 /▪O (−0.28 eV), so electrons can more easily transfer to the oxygen molecules, producing ▪O . The uplift of conduction band after cationic doping had also been observed in Ag doping Sr2TiO4 [16]. Hydroxyl radical ( OH ) may be difficult to produce by the photogenerated holes, directly, because the valence band position (+2.12 eV) in Sr2TiO4:N,Nb(2%) was higher than that of OH/H 2 O (+2.27 eV) and slightly more positive than that of OH/OH (1.99 eV). Therefore, ▪O is the predominant active species in the degradation of TC, which coincided with the previous result ( Figure 10). A similar situation was also discovered in Bi2WO6 [43,44].

Chemicals
The chemicals in this work were obtained from Shanghai Maclin Biochemical Technology Co., Ltd. (Shanghai, China) and Tianjin Yongda Chemical Reagent Co., Ltd. (Tianjin, China), without any further purification.

Synthesis of Sr2TiO4
Firstly, 0.01 mol tetrabutyl titanate was dissolved in 50 mL of absolute ethanol, and 0.07 mol citric acid was injected as a complex agent (solution A). Secondly, 0.02 mol Sr(NO3)2 was dissolved in deionized water (solution B). Solution B was added dropwise to solution A to gain a transparent mixed solution. The mixed solution was heated and stirred in a water bath at 60 °C to form a homogeneous transparent sol. After aging for 24 h, the sol was dried at 120 °C for several hours to form a dry gel. Finally, the obtained gel was ground and calcined at 1000 °C for 4 h (heating rate 10 °C·min −1 ) to obtain Sr2TiO4 powder. Hydroxyl radical (OH) may be difficult to produce by the photogenerated holes, directly, because the valence band position (+2.12 eV) in Sr 2 TiO 4 :N,Nb(2%) was higher than that of OH/H 2 O (+2.27 eV) and slightly more positive than that of OH/OH − (1.99 eV). Therefore, O − 2 is the predominant active species in the degradation of TC, which coincided with the previous result ( Figure 10). A similar situation was also discovered in Bi 2 WO 6 [43,44].

Chemicals
The chemicals in this work were obtained from Shanghai Maclin Biochemical Technology Co., Ltd. (Shanghai, China) and Tianjin Yongda Chemical Reagent Co., Ltd. (Tianjin, China), without any further purification.

Synthesis of Sr 2 TiO 4
Firstly, 0.01 mol tetrabutyl titanate was dissolved in 50 mL of absolute ethanol, and 0.07 mol citric acid was injected as a complex agent (solution A). Secondly, 0.02 mol Sr(NO 3 ) 2 was dissolved in deionized water (solution B). Solution B was added dropwise to solution A to gain a transparent mixed solution. The mixed solution was heated and stirred in a water bath at 60 • C to form a homogeneous transparent sol. After aging for 24 h, the sol was dried at 120 • C for several hours to form a dry gel. Finally, the obtained gel was ground and calcined at 1000 • C for 4 h (heating rate 10 • C·min −1 ) to obtain Sr 2 TiO 4 powder.

Characterization
The crystal structure analysis was carried out by a Japanese D/MAX2500PC X-ray diffractometer employing Cu Kα radiation; the surface morphology was investigated by a Japan JEM-2010 transmission electron microscope (TEM); a specific surface and pore size analyzer (3H-2000PM1, BeiShiDe Instrument Technology Co., Ltd., Beijing, China) was used to calculate the specific surface area (BET) and pore size of the sample; the optical properties were investigated by an Ultraviolet-Visible spectrophotometer (Lambda750, PerkinElmer, Norwalk, CT, USA); the photoluminescence (PL, SR830) with the excitation at 325 nm was characterized using a fluorescence spectrometer; the chemical composition and valence band spectrum were detected by X-ray photoelectron spectroscopy (Thermo Scientific, Waltham, MA, USA, ESCALAB 250XI).

Photoelectrochemical Measurements
Photocurrent density-time (I-t) and electrochemical impedance spectroscopy (EIS) were characterized on an electrochemical workstation (CHI660D, Shanghai Chenhua Instrument, Shanghai, China) through a three-electrode system. Three-electrode system was set up using 0.5 M Na 2 SO 4 solution, Ag/AgCl as reference electrode, Pt foil as counter electrode, and photoelectrodes fabricated by these sample powders on FTO glass as working electrodes. During the measurements of photocurrent density, simulated sunlight was irradiated with an intensity of 100 mW·cm −2 (AM1.5G filter) under chopped light at a bias of 1.23 V vs. RHE. The EIS was measured with a frequency from 100 kHz to 0.01 Hz with an AC amplitude of 5 mV.

Photocatalytic Degradation of TC
The photocatalytic performances of the samples were assessed by photocatalytic degradation towards TC. In total, 50 mg Sr 2 TiO 4 or 50 mg Sr 2 TiO 4 :N,Nb was dispersed in 50 mL of TC solution at a concentration of 20 mg·L −1 . Before irradiation, the mixtures were continuously stirred in the dark for 30 min to reach adsorption-desorption equilibrium. Then, a photocatalytic reaction was performed under visible light using a 5 W LED lamp (white light, λ > 420 nm, 250 mW/cm 2 ). Extract 3 mL of TC solution every 10 min and centrifuge at 9000 rpm for 5 min. The TC degradation rate was calculated as follows: where C 0 (A 0 ) and C t (A t ) represent the TC's concentration (absorbance at 357 nm) of initial and after min irradiation, respectively. In order to ascertain the active species during the photocatalytic degradation process, the active species capture experiment was carried out, similar to the photocatalytic degradation of TC experiments, except that TC solution was replaced by TC and scavenger.

Conclusions
In summary, Nb/N co-doped layered perovskite Sr 2 TiO 4 (Sr 2 TiO 4 :N,Nb) with varying percentages (0-5 at%) of Nb were successfully synthesized by sol-gel and calcination. Nb/N could co-dope into Sr 2 TiO 4 with a slight unit cell expansion, which was confirmed by X-ray diffraction. Nb/N co-doped Sr 2 TiO 4 showed better photocatalytic performance for tetracycline photocatalytic degradation than pristine Sr 2 TiO 4 under visible light, especially Sr 2 TiO 4 :N,Nb(2%). Sr 2 TiO 4 :N,Nb(2%) showed optimal photocatalytic performance with the 99% degradation after 60 min visible light irradiation, which was higher than pristine Sr 2 TiO 4 (40%). Nb/N co-doping enhanced the photocatalytic activity by broadening light response to the visible region and reducing the photogenerated carrier recombination. Nb/N co-doping had a slight effect on morphology and the average grain size of Sr 2 TiO 4 . Sr 2 TiO 4 :N,Nb(2%) showed good stability and recyclability after five cycles photocatalytic degradation reaction. The superoxide radical (·O − 2 ) was the leading contributor to tetracycline degradation. Nb/N co-doping strategy in this work affords insight to realizing visible-light-driven perovskite-type semiconductors for wastewater treatment.
Author Contributions: Conceptualization, J.W. and Y.Z.; methodology, Y.Z. and X.Z.; validation, P.L.; data curation, J.W. and P.L. writing-original draft preparation, J.W. and P.L. All authors have read and agreed to the published version of the manuscript.