Surface modification of layered perovskite Sr2TiO4 for improved CO2 photoreduction with H2O to CH4

Layered perovskite Sr2TiO4 photocatalyst was synthesized by using sol-gel method with citric acid. In order to increase the surface area of layered perovskite Sr2TiO4, and thus to improve its photocatalytic activity for CO2 reduction, its surface was modified via hydrogen treatment or exfoliation. The physical and chemical properties of the prepared catalysts were characterized by X-ray diffraction, high-resolution transmission electron microscopy, elemental mapping analysis, energy-dispersive X-ray spectroscopy, N2 adsorption-desorption, UV-Vis spectroscopy, X-ray photoelectron spectroscopy, photoluminescence, and electrophoretic light scattering. CO2 photoreduction was performed in a closed reactor under 6 W/cm2 UV irradiation. The gaseous products were analyzed using a gas chromatograph equipped with flame ionization and thermal conductivity detectors. The exfoliated Sr2TiO4 catalyst (E-Sr2TiO4) exhibited a narrow band gap, a large surface area, and high dispersion. Owing to these advantageous properties, E-Sr2TiO4 photocatalyst showed an excellent catalytic performance for CO2 photoreduction reaction. The rate of CH4 production from the photoreduction of CO2 with H2O using E-Sr2TiO4 was about 3431.77 μmol/gcat after 8 h.

Recently, the development of TiO 2 photocatalysts with the perovskite structure ABO 3 , has attracted due to the unique perovskite structure, their composition can be easily changed at the A, and B sites and the metal introduced can be quantitatively substituted into the skeleton. Among the perovskite semiconductors, SrTiO 3 is widely used as a photocatalyst. Much like TiO 2 , SrTiO 3 has been combined with other species to form hybrid composites such as Mn/SrTiO 3 42 , Cu/SrTiO 3 43 , N-doped TiO 2 -SrTiO 3 44 , Fe 2 O 3 /SrTiO 3 45 , SrTiO 3 :Cr/Ta/F 46 , SrTiO 3 /HZSM-5 47 , SrTiO 3 /TiO 2 /H-titanate nanofiber 48 , SrTiO 3 :Rh/Sb 49 , La/Cr-doped SrTiO 3 50 , Pt/SrTiO 3 51 , Zn/SrTiO 3 52 , Ag 3 PO 4 /Cr-SrTiO 3 53 , and g-C 3 N 4 -SrTiO 3 :Rh 54 to improve its photocatalytic performance. Studies on other perovskite catalysts, including Ca x Ti y O 3 55 , and basalt fiber@PbTiO 3 56 , also have been recently reported. Another advantage of perovskite is that it forms a layered perovskite depending on the nature and contents of the A and B ions. Figure 1a shows the structure of a Ruddlesden-Popper A j+1 B j O 3j+1 perovskite. Here, when the j value increases, the structure tends towards an ABO 3 perovskite. In particular, A 2 BO 4 , for which j = 1, shows a layered structure with a large gap between each BO 6 octahedron.
In this study, we have attempted to improve the photocatalytic performance of Sr 2 TiO 4 layered perovskite by increasing its surface area (Fig. 1b). The surface of the synthesized catalyst was hydrogen treated or exfoliated to increase its interaction with the CO 2 feed material and thus amplify its CO 2 photoreduction activity. The characteristic properties of the synthesized catalysts were measured by using a variety of techniques such as using X-ray diffraction (XRD), high-resolution transmission electron microscopy (HR-TEM), N 2 adsorption-desorption isotherm analysis, UV-Vis spectroscopy, photoluminescence (PL), zeta potential analysis, and X-ray photoelectron spectroscopy (XPS). Furthermore, their photocatalytic activity for the reduction of CO 2 with H 2 O under UV light was studied.

Results
X-ray diffraction (XRD) patterns. The 59 . These structural changes are due to the fact that the Sr ions located between the Sr 2 TiO 4 layers are removed by HNO 3 treatment, and only the TiO 6 octahedra (corresponding to BO 6 ) remain. The crystal structure analysis reveals that the Ti forms a rutile structure in the Sr 2 TiO 4 and that the Sr is intercalated between the layers. The crystallite size of the catalysts was calculated using scherrer's equation. In the case of Sr 2 TiO 4 and H-Sr 2 TiO 4 , (013) plane was selected, and (110) plane was selected for E-Sr 2 TiO 4 . As a result, the crystallite size of Sr 2 TiO 4 , H-Sr 2 TiO 4 , and E-Sr 2 TiO 4 were found to be 11.568, 11.578, 1.165 Å, respectively.
High-resolution transmission electron microscopy (HR-TEM), element mapping and EDX analysis. The differences in the morphologies of Sr 2 TiO 4 and E-Sr 2 TiO 4 were investigated using HR-TEM and selected area electron diffraction (SAED) (Fig. 3). The larger particles are observed in Sr 2 TiO 4 , and whereas E-Sr 2 TiO 4 consists of the particles form a separate sheets or randomly folded sheets. The images show that the interplanar distance for E-Sr 2 TiO 4 is larger than that for Sr 2 TiO 4 . The d-spacings for the Sr 2 TiO 4 (013) plane and the E-Sr 2 TiO 4 (110) plane are 2.84 and 3.25 Å, respectively. These results are in accordance with the values derived from XRD patterns.
The compositions of Sr 2 TiO 4 and E-Sr 2 TiO 4 were analyzed by HR-TEM element mapping analysis and EDX, and the results are shown in Fig. 4 and Table 1. In Sr 2 TiO 4 , Sr and Ti ions are uniformly distributed throughout the particles and Sr ions are more abundant than Ti ions. The atomic percent values for Sr and Ti are 16 and 11%, respectively. However, the Sr ion content of E-Sr 2 TiO 4 is much lower than that of Sr 2 TiO 4 . Furthermore, EDX analysis showed that the Sr ion content, which is about 1.4-times that of Ti in Sr 2 TiO 4 , is reduced to just 0.07% that of Ti in E-Sr 2 TiO 4 . Thus, these results support the assertion that the layers are separated because Sr ions are removed from the interlayers by exfoliation. Figure 5 shows the N 2 adsorption-desorption isotherms at 77 K for P-25, which was used as a comparative sample, Sr 2 TiO 4 , H-Sr 2 TiO 4 , and E-Sr 2 TiO 4 . According to the IUPAC classification, the adsorption-desorption isotherm curves of all the catalysts belong to type III. Therefore, the synthesized catalysts are non-porous materials. However, the slight hysteresis in the curves is due to the bulk pores between the particles. The specific surface areas of Sr 2 TiO 4 and H-Sr 2 TiO 4 are 1.19 and 0.78 m 2 /g, respectively, which are very low. However, E-Sr 2 TiO 4 has a specific surface area of 358.54 m 2 /g, which is much larger than those of Sr 2 TiO 4 and H-Sr 2 TiO 4 . The increase in the specific surface are of E-Sr 2 TiO 4 is due not only to the removal of Sr ions from between the layers, but also to the separation of the layers, as shown in the HR-TEM images. The increase in catalyst surface area leads to an increase in the number of active sites for CO 2 and H 2 O to react, leading to an increase in reactivity. Therefore, E-Sr 2 TiO 4 was expected to exhibit improved catalytic activity compared to those of Sr 2 TiO 4 and H-Sr 2 TiO 4 . This specific surface area is considerably larger than that of the commercial catalyst P-25, which is 43.65 m 2 /g. Figures 6 and 7 show the UV-Vis spectra and Tauc's plots of P-25, Sr 2 TiO 4 , H-Sr 2 TiO 4 , and E-Sr 2 TiO 4 . The UV absorptions of Sr 2 TiO 4 and H-Sr 2 TiO 4 are blue-shifted compared to that of P-25 because of the influence of SrO, with its band gap of 5.71 eV 60 . Conversely, the absorbance of E-Sr 2 TiO 4 is shifted to a longer wavelength due to the removal of Sr ions. Most interestingly, it moved to longer wavelength than that of P-25. This is because E-Sr 2 TiO 4 has a rutile TiO 2 structure (band gap: 3.0 eV), as confirmed by the XRD analysis above. Therefore, the absorbance of E-Sr 2 TiO 4 is shifted to longer wavelength than that of P-25, which composed mainly of anatase TiO 2 (band gap: 3.2 eV). The band gap was calculated using the Tauc equation 61 :

Optical properties of photocatalysts.
where α, h, ν, A, and E bg represent the absorption coefficient, Plank's constant, light frequency, a constant, and band gap energy, respectively. In a plot of (αhν) 2 versus photon energy (hν), the intercept on the x axis gives the band gap. Using this method, the band gap of P-25, Sr 2 TiO 4 , H-Sr 2 TiO 4 and E-Sr 2 TiO 4 were calculated to be 3.16, 3.33, 3.34, and 3.03 eV, respectively. Therefore, E-Sr 2 TiO 4 has the narrowest band gap, making it most suitable as a photocatalyst. Photocatalysts with narrower band gaps have advantages in terms of photosensitization. However, in order to exhibit good performance in the current system, the band gap of a photocatalyst should also include the CO 2 / CH 4 and H 2 O/O 2 reduction potentials. Figure 8 shows the XPS valance band spectra of the catalysts. Based on the data obtained, the valance band value of the catalysts was confirmed, and the values for P-25, Sr 2 TiO 4 , H-Sr 2 TiO 4 , and E-Sr 2 TiO 4 are 2.46, 2.34, 1.60, and 2.06 eV, respectively. When the vacuum level of 4.5 eV is corrected to 0 V for a standard hydrogen electrode (SHE) and the work function of the XPS instrument is taken as 4.62 eV 62 ,   Figure 9 shows the energy diagrams obtained for the catalysts using the valance and conduction band values and the band gap. All catalysts contain the CO 2 /CH 4 and H 2 O/O 2 reduction potential. Therefore, the synthesized catalysts are suitable for the photoreduction of CO 2 with H 2 O to CH 4 .
In order to understand the recombination of excited electrons and holes, PL analysis was conducted, and the results are shown in Fig. 10. The PL spectra of the catalysts show a strong emission signal at 468.1 nm. The PL intensity of Sr 2 TiO 4 is smaller than that of P-25. This is due to a decrease in the number of excited electrons because of the wide band gap of Sr 2 TiO 4 . The intensity for H-Sr 2 TiO 4 is lower than that of Sr 2 TiO 4 . This is because the oxidation state of the exposed Ti on the surface is reduced to (4-δ) + , which is not +4, and the reduced Ti   suppresses the recombination of electrons and holes by trapping the excited electrons in the conduction band. The E-Sr 2 TiO 4 also exhibits a PL intensity lower than that of Sr 2 TiO 4 and much lower than that of P-25. Generally, excited electrons and holes move from the bulk of a particle to its surface where they react with reactants. The recombination of excited electrons and holes takes place in the bulk or on the surface of a particle during transport. When the particles are exfoliated, the internal area of the particles decreases and the distance to the surface for the electrons and holes decreases. Therefore, recombination inside the particles is also reduced. This is the reason that E-Sr 2 TiO 4 has a lower PL intensity than that of P-25. Thus, the above analysis indicates that H-Sr 2 TiO 4 and E-Sr 2 TiO 4 will be better photocatalysts than Sr 2 TiO 4 . X-ray photoelectron spectroscopy (XPS) analysis. The XPS spectra of the photocatalysts were obtained to confirm the oxidation state of the elements according to their chemical bonding, and the results are shown in Fig. 11. For Sr 2 TiO 4 , the peaks located at 133.38 and 134.88 eV are assigned to Sr-3d 5/2 and Sr-3d 3/2 core levels respectively. In H-Sr 2 TiO 4 , the Sr 3d peaks are shifted toward a slightly lower binding energy. In E-Sr 2 TiO 4 , the intensity of the 3d peaks is greatly reduced. This is due to the removal of Sr ions from the interlayer spaces, as described above. The Ti 2p 3/2 and 2p 1/2 peaks of Sr 2 TiO 4 are observed at 457.88 and 463.78 eV, respectively. The Ti 2p peaks for H-Sr 2 TiO 4 are shifted to a lower binding energy, similarly to the Sr 3d peaks. This is because some of the Sr 2+ and Ti 4+ ions are reduced by hydrogen to Sr (2-δ)+ and Ti (4-δ)+ , respectively. The Ti 2p peaks for E-Sr 2 TiO 4 have significantly different peak intensities to Sr 2 TiO 4 or H-Sr 2 TiO 4 , and its Ti 2p 3/2 and Ti 2p 1/2 peaks are observed at 458.18 and 464.08 eV, respectively. This binding energy is shifted slightly lower compared to that of P-25. Therefore, the Ti ions of E-Sr 2 TiO 4 are slightly more reduced ions than in P-25. It is believed that this can induce vacancies in the crystal framework and facilitate the movement of electrons and holes, which can be advantageous for photocatalytic activity.
There are two O 1 s peaks in Sr 2 TiO 4 and H-Sr 2 TiO 4 . The peak at ~529 eV is from oxygen bound to Ti, and the peak at ~531 eV corresponds to oxygen bound to Sr. For E-Sr 2 TiO 4 , only the peak corresponding to oxygen bonded to Ti is observed (at 529.48 eV), because the Sr is removed by exfoliation.
Zeta potential analysis of photocatalysts. Since the catalytic reaction takes place in H 2 O, it is important to study the dispersion of the catalyst particles in H 2 O. The zeta potentials were measured after dispersing the catalysts in distilled water or in bubbling-CO 2 solution (i.e., the reaction conditions), and the results are shown in Table 2. Generally, a larger absolute value for the measured zeta potential means that the particles are well dis-  Therefore, the degree of colloidal dispersion in the solution after CO 2 bubbling follows the order E-Sr 2 TiO 4 > P-2 5 > H-Sr 2 TiO 4 > Sr 2 TiO 4 . Thus, E-Sr 2 TiO 4 , P-25, and H-Sr 2 TiO 4 show good dispersion under the reaction conditions, which is considered advantageous for CO 2 photoreduction performance. However, in the case of Sr 2 TiO 4 , the zeta potential is low, so it is likely to exhibit poor catalytic performance owing to it being more agglomerated than the other catalysts.

Photocatalytic reduction of CO 2 with H 2 O, property after reaction, and mechanism.
The products obtained though CO 2 reduction using the catalysts synthesized in this study are CH 4 , H 2 , C 2 H 6 , C 2 H 4 , and CO. Figure 12 shows the accumulation of the products according to irradiation time. The main product is CH 4 and the product amounts follow the order CH 4 Figure 13 shows the product distribution on the catalysts. P-25 showed the highest CH 4 selectivity, which was 80 to 90%. Next, when E-Sr 2 TiO 4 was used, the CH 4 selectivity was high and its value was about 70%. Table 3 shows the quantum yield of the catalysts and the overall quantum yield was in the order of E-Sr 2 TiO 4 , P-25, H-Sr 2 TiO 4 , and Sr 2 TiO 4 , which were 2.83, 1.96, 2.93, and 6.20%, respectively. The quantum yield for CO 2 photoreduction to produce CH 4 of P-25, Sr 2 TiO 4 , H-Sr 2 TiO 4 and E-Sr 2 TiO 4 were 2.69, 1.28, 2.56, 5.21%, respectively. The quantum yield for other products was less than 1% for all catalysts. The XRD analysis of the catalysts after the reaction was carried out to confirm the structural stability and displayed in Fig. 14. From the XRD analysis results, it was observed that the catalysts structure was remained stable before and after the reaction. Therefore, the catalysts structure was stable during the reaction conditions.
The photoreaction of E-Sr 2 TiO 4 showing the best activity was repeated three times. The results for CH 4 production, the main product, are shown in Fig. 15. A slight amount of difference was observed in the results but   Table 3. The apparent quantum yield of catalysts.

Catalysts
SCientifiC REPORTS | 7: 16370 | DOI:10.1038/s41598-017-16605-w similar performance was maintained without deactivation in all three times. Therefore, it was confirmed that the E-Sr 2 TiO 4 was excellent in not only structural stability but also reusability during the reaction Based on these results, a plausible reaction pathway over E-Sr 2 TiO 4 , which has the best performance is proposed in Fig. 16. The excited electrons on the exfoliated improved catalyst surface react with CO 2 to produce •CO 2 radicals, and the holes react with H 2 O to produce OHand H + . Hydrogen radicals (•H) are formed by the reaction of H + with excited electrons, and then CO is produced by the reaction of •H and •CO 2 radicals. CH 4 , C 2 H 6 , and C 2 H 4 are produced finally as the CO and •H radicals continuously react.

Conclusion
In this study, nanosized layered perovskite Sr 2 TiO 4 photocatalyst was successfully synthesized by using sol-gel technique with the assistance of citric acid. The surface of layered perovskite Sr 2 TiO 4 photocatalyst was treated to improve the CO 2 photoreduction activity. The particles were treated with HNO 3 to remove the Sr ions present between the layers, and the layers were exfoliated by treatment with TPAOH. The catalyst, E-Sr 2 TiO 4 showed the rutile TiO 2 structure after exfoliation because the Sr 2 TiO 4 structure was collapsed. The shape of the exfoliated thin film was confirmed by TEM. In comparison to Sr 2 TiO 4 and H-Sr 2 TiO 4 photocatalysts, the exfoliated catalyst E-Sr 2 TiO 4 showed an excellent performance in CO 2 photoreduction to CH 4 , and after 8 h, 3431.77 μmol/g cat of CH 4 was generated. The reason for the excellent performance of E-Sr 2 TiO 4 can be explained by the following factors.
First, it has a narrow band gap compared to the other two catalysts, and exhibits reduced electron-hole recombination. Therefore, a relatively larger number of electrons and holes transferred to CO 2 and H 2 O. Next, a large amount of CO 2 and H 2 O can interact with the active sites on the surface because it has a large surface area and is well dispersed in the solution. Based on the excellent physical and photochemical properties of the exfoliated layered perovskite catalyst, it may be employed for different photocatalytic applications as well as the CO 2 photoreduction reactions.  Strontium nitrate (Sr(NO 3 ) 2 , 97.0%, Junsei Chemical, Japan) and titanium isopropoxide (Ti(OCH(CH 3 ) 2 ) 4 , TTIP, 98.0%, Junsei Chemical, Japan) were used as precursors. First, 0.1 mol of Sr(NO 3 ) 2 was dissolved in double distilled water (100 mL) with continuous stirring. Then, 10 mL of HNO 3 (60%, OCI company Ltd., Republic of Korea) was added with stirring to prevent hydrolysis. This solution was labeled A. In a separate vessel, 0.05 mol of TTIP was dissolved in EtOH (99.9%, OCI company Ltd., Republic of Korea), and then glacial acetic acid (CH 3 COOH, 99.0%, Ducsan, Republic of Korea) was added with stirring to prevent hydrolysis. This solution was labeled as B. The solutions A and B were then mixed with stirring, and citric acid monohydrate (C 6 H 8 O 7 ·H 2 O, 99.5%, Daejung Chemicals and Metals Co Ltd., Republic of Korea), which is a complexing agent for the gel, was added, and the mixture was stirred until it became homogeneous. Then, the solvent was removed without the temperature exceeding 323.15 K to obtain a sol gel, which was subsequently pretreated at 493.15 K. Finally, a white powder was obtained by thermal treatment at 1323.15 K for 6 h. In some cases, hydrogen treatment was then performed at 1123.15 K for 3 h in H 2 atmosphere. In other cases, exfoliation was achieved by a three-step process, as shown in Fig. 17. Ion exchange of Sr for H cations was carried out in 1 M HNO 3 for 5 days using ultrasonication. The powder obtained was treated in tetrapropylammonium hydroxide (TPAOH, 25.0%, in water, ACROS, Belgium) for 3 weeks using ultrasonication. The final white precipitate was washed several times with distilled water and ethanol and dried at 343 K for 24 h. The hydrogen-treated and exfoliated Sr 2 TiO 4 samples were labeled H-Sr 2 TiO 4 and E-Sr 2 TiO 4 , respectively.
Characterization of photocatalysts. The structures and crystallinities of the as-prepared Sr 2 TiO 4 , H-Sr 2 TiO 4 , and E-Sr 2 TiO 4 samples were confirmed with XRD (model MPD from PANalytical) using nickel-filtered CuKα radiation (40.0 kV, 30.0 mA). The morphologies were investigated using HR-TEM (Tecnal G2 F20 S-TWIN, FEI, Netherlands) operated at 200 kV. The presence of different elements was confirmed using the elemental mapping and energy dispersive X-ray spectroscopy (EDS) attached to the TEM setup. The specific surface areas (S BET ) were calculated according to the Brunauer-Emmett-Teller theory using a Belsorp II mini (BEL, Japan Inc.). The UV-Vis absorption spectra were obtained using a SCINCO Neosys-2000 spectrometer fitted with a reflectance sphere. PL profiles were obtained using a SCINCO FluoroMate FS-2 at room temperature using a He-Cd laser source at a wavelength of 325 nm. XPS measurements were performed on a K-alpha (Thermo Scientific, UK) using Al Kα X-rays as the excitation source. The zeta potential of the material was determined by electrophoretic mobility using an electrophoresis measurement apparatus (ELS 8000, Otsuka Electronics, Japan) with a plate sample cell. Electrophoretic light scattering (ELS) determination was performed in reference beam mode with a 670 nm laser light source at a modular frequency of 250 Hz and a scattering angle of 15°. The standard error of the zeta potential, converted from the experimentally determined electrophoretic mobility, was typically <1.5% with 5% error. To measure the zeta potentials, the samples were dispersed in deionized water or bubbling-CO 2 water at 0.1 wt%. The final zeta potentials were obtained by averaging 2 or 3 measurements.
Photocatalytic activity measurements. The photocatalytic tests for the reduction of CO 2 with H 2 O were performed in a photoreactor comprising a quartz chamber with a total volume of 150.0 cm 3 (Fig. 18). To photoreduce CO 2 , 0.01 g of the catalyst was placed in the reactor chamber with 50 mL of double distilled water, and the reactor was closed. A UV lamp (6 W/cm 2 , 20 cm length × 2.0 cm diameter, Shinan, Republic of Korea) emitting light at 365 nm was used to irradiate the reaction mixture. Supercritical-fluid-grade CO 2 with a certified maximum hydrocarbon content of <1 ppm was used as the reactant. Before the reaction was initiated by illumination, the reactor was purged with CO 2 gas for 5 min. The lamp was then switched on to start the experiment. The reaction temperature and pressure were maintained at 303 K and 1 atm, respectively. The gas products were analyzed using a gas chromatograph (iGC7200, DS Science, Republic of Korea) equipped with a thermal conductivity detector (TCD) and a flame ionization detector (FID). The product yield 33 and quantum yield 63 during reaction was calculated using following equation (2-3).