Facile synthesis of TiO2/Ag3PO4 composites with co-exposed high-energy facets for efficient photodegradation of rhodamine B solution under visible light irradiation

In this study, TiO2/Ag3PO4 composites based on anatase TiO2 nanocrystals with co-exposed {101}, {010}/{100}, {001} and [111]-facets and Ag3PO4 microcrystals with irregular and cubic-like polyhedron morphologies were successfully synthesized by combining hydrothermal and ion-exchange methods. The anatase TiO2 nanocrystals with different high-energy facets were controllably prepared via hydrothermal treatment of the exfoliated [Ti4O9]2−/[Ti2O5]2− nanosheet solutions at desired pH values. The Ag3PO4 microcrystal with different morphologies was prepared via the ion-exchange method in the presence of AgNO3 and NH4H2PO4 at room temperature, which was used as a substrate to load the as-prepared anatase TiO2 nanocrystals on its surface and to form TiO2/Ag3PO4 heterostructures. The apparent rate constant of the pH 3.5-TiO2/Ag3PO4 composite was the highest at 12.0 × 10−3 min−1, which was approximately 1.1, 1.2, 1.4, 1.6, 13.3, and 24.0 fold higher than that of pH 0.5-TiO2/Ag3PO4 (10.5 × 10−3 min−1), pH 7.5-TiO2/Ag3PO4 (10.2 × 10−3 min−1), pH 11.5-TiO2 (8.8 × 10−3 min−1), Ag3PO4 (7.7 × 10−3 min−1), blank sample (0.9 × 10−3 min−1), and the commercial TiO2 (0.5 × 10−3 min−1), respectively. The pH 3.5-TiO2/Ag3PO4 composite exhibited the highest visible-light photocatalytic activity which can be attributed to the synergistic effects of its heterostructure, relatively small crystal size, large specific surface area, good crystallinity, and co-exposed high-energy {001} and [111]-facets. The as-prepared TiO2/Ag3PO4 composites still exhibited good photocatalytic activity after three successive experimental runs, indicating that they had remarkable stability. This study provides a new way for the preparation of TiO2/Ag3PO4 composite semiconductor photocatalysts with high energy crystal surfaces and high photocatalytic activity.


Introduction
With the rapid development of industrialization, energy and environmental crises have become the key factors restricting the sustainable development of human society. Therefore, it is very urgent to search for suitable semiconductor photocatalysts to make full use of solar energy to split water into hydrogen, convert carbon dioxide into fuels, store energy, and degrade the organic wastewater discharged from the textile industry. [1][2][3][4][5] In recent decades, different types of semiconductor photocatalysts, such as carbon-cloth functionalized transition metal based electrocatalysts, 6 quantum dot-based photocatalysts, 7-9 iron (Fe)-doped ZrO 2 , 10 metal-organic framework (MOF)-based heterostructured catalysts, 11 and ZnO/Bi 2 WO 6 nanohybrids 12 have been reported. Among the well-known oxide semiconductor photocatalysts, titanium dioxide (TiO 2 ) has been proven to be the best choice due not only to its excellent photooxidization ability and low cost but also its long-term photostability and chemical stability and innocuousness. 13,14 However, the photocatalytic efficiency of TiO 2 still needs to be further improved for its practical application. The photocatalytic efficiency of TiO 2 is mainly dependent on the phase structure, crystallization, crystal size, specic surface area, and surface energy. 15,16 However, based on the principle of surface energy minimization (0.44 J m À2 for {101} facets < 0.53 J m À2 for {010}/{100} facets < 0.90 J m À2 for {001} facets < 1.09 J m À2 for {110} facets < 1.61 J m À2 for {111} facets), the proportion of highenergy crystal surfaces in the natural and synthetic anatase TiO 2 crystals under equilibrium condition is very small, resulting in the dominant exposed {101} crystal facets (more than 94%) on its surface. 17,18 Since the pioneering work by Wen and coworkers on the synthesis of nanometer-sized anatase TiO 3 ] 2À nanosheets as the precursors in the presence and absence of capping agent, which displayed superior photocatalytic and photovoltaic performance. [23][24][25] Although the exposed high-energy crystal surface of anatase TiO 2 crystals will be conducive to improving the photocatalytic activity and dyesensitized solar energy performance, however, anatase TiO 2 crystals cannot suitable for applications under visible light irradiation due to its wide band gap (3.2 eV), resulting in the lower energy conversion efficiency in practical application. 26 In order to overcome above limitation, it is of great signicance to extend the light absorption range of the anatase TiO 2 crystals to the visible light region. [27][28][29] Silver orthophosphate (Ag 3 PO 4 ) is a semiconductor photocatalyst with a narrow band gap of 2.45 eV, and is oen to decompose organic contaminants and oxidize water to produce oxygen under visible light irradiation. 30 However, the narrow band gap energy and low valence band (VB) and conduction band (CB) position of Ag 3 PO 4 result in high recombination rate and weak redox capacity of photogenerated electrons and holes, which severely weaken the photocatalytic activity of Ag 3 PO 4 . 26,31 Therefore, it is an effective strategy to form a heterojunction by coupling anatase TiO 2 crystals with Ag 3 PO 4 photocatalyst for improving the photocatalytic activity under visible-light irradiation. 32, 33 Zhang et al. synthesized one-dimensional heterostructured Ag 3 PO 4 /TiO 2 photocatalyst with improved photocatalytic activity for degradation of 4-nitrophenol in simulant wastewater under visible light. 34 An et al. reported that the oating HGMs-TiO 2 /Ag 3 PO 4 composites exhibited superior photocatalytic performance than that of pure Ag 3 PO 4 and TiO 2 /Ag 3 PO 4 for degradation of methylene blue solution under visible light irradiation. 3 Xu et al. reported that the magnetic Ag 3 PO 4 /TiO 2 /Fe 3 O 4 heterostructured nanocomposite showed enhanced photocatalytic performance for the degradation of acid orange 7 under visible light irradiation. 35 Hamrouni et al. synthesized Ag doped TiO 2 -Ag 3 PO 4 (Ag@TiO 2 -Ag 3 PO 4 ) composites by coupling sol-gel and precipitation methods, which signicantly improved the photocatalytic activity than that of the TiO 2 -Ag 3 PO 4 and the benchmark TiO 2 Evonik P25 for degradation of 4-nitrophenol solution under solar light irradiation. 36 In this study, anatase TiO 2 nanocrystals with different high energy facets were successful synthesized by using the exfoliated two-dimensional [Ti 4 O 9 ] 2À /[Ti 2 O 5 ] 2À nanosheets, which were compounded with Ag 3 PO 4 microcrystals to form a series of heterostructured TiO 2 /Ag 3 PO 4 composites. To our knowledge, this is the rst time to study the TiO 2 /Ag 3 PO 4 photocatalysts formed by the combination of the anatase TiO 2 nanocrystals with high energy crystal surface and Ag 3 PO 4 with different morphologies. Various catalyst characterization of the synthesized TiO 2 /Ag 3 PO 4 composites conrmed that TiO 2 nanocrystals with co-exposed high-energy facets were successfully attached to the surface of Ag 3 PO 4 microcrystals. In comparison to the commercial TiO 2 and the pure Ag 3 PO 4 samples, the heterostructured TiO 2 /Ag 3 PO 4 composites exhibited good photocatalytic activity for the degradation of rhodamine B under visible light irradiation, which can be attributed to the separation of the e À (in Ag 3 PO 4 crystal) and h + (in TiO 2 nanocrystal) inhibits the charge recombination. For the as-prepared TiO 2 / Ag 3 PO 4 composites, the pH 3.5-TiO 2 /Ag 3 PO 4 exhibited the highest photocatalytic activity, which can be attributed to the synergistic effects of its relative small crystal size, large specic surface area, good crystallinity, and co-exposed high-energy {001} and [111]-facets. However, although the as-prepared TiO 2 /Ag 3 PO 4 composites exhibited good stability, the photocatalytic performance needs to be further improved for their practical application.

Synthesis of TiO 2 /Ag 3 PO 4 composites
The well dispersed TiO 2 colloidal suspensions were obtained by dispersed 1.2 g as-prepared TiO 2 in 200 mL deionized water under stirring for 2 h. Then, 0.3 g Ag 3 PO 4 precipitate was added to the above TiO 2 colloidal suspensions and kept under stirring for 2 h to generate TiO 2 /Ag 3 PO 4 composites (w(TiO 2 ) ¼ 80%, w(Ag 3 PO 4 ) ¼ 20%). Finally, the composites were collected by ltering, which were washed several times, and dried at room temperature.

Characterization
The crystal structure of obtained samples were characterized by powder X-ray diffractometer (XRD) on a XRD-6100 (Shimadzu, Kyoto, Japan) with monochromated Cu Ka radiation (l ¼ 1.5406 A). The data were collected for scatting angles (2q) from 5 to 80 with a scanning speed of 8 min À1 . The morphology of the samples were investigated by using cold eld emission scanning electron microscope (FESEM, JSM-7500F, Japan). The crystalline nanostructures were investigated using transmission electron microscopy (TEM), high-resolution transmission electron microscopy (HRTEM) (Tecnai G 2 F20 S-TWIN, FEI, America). The specic surface areas of the as-prepared samples were determined by using the Brunauer-Emmett-Teller (BET) method (Autosorb-IQ3, Quantachrome, America). UV-Vis-NIR spectra of the samples were obtained by using a Cary Series UV-Vis-NIR Spectrophotometer (Agilent Technologies, Cary 5000). The absorbance of rhodamine B solution was recorded within the wavelength range of 350-650 nm by using a TU-1901 UV-vis spectrophotometer (Beijing Purkinje General Instrument Co. Ltd).

Photocatalytic activity evaluations
The photocatalytic activities of the as-synthesized TiO 2 /Ag 3 PO 4 composites were evaluated by monitoring the degradation of rhodamine B (RhB). The irradiation source was provided by a 300 W xenon lamp equipped with a 400 nm cutoff light lter and the wavelength ranges from 400 nm to 600 nm. Typically, 75 mg TiO 2 /Ag 3 PO 4 composite was suspended in 150 mL RhB solution (10 ppm). Prior to illumination, the suspensions were magnetically stirred for 2 h in the dark to achieve adsorptiondesorption equilibrium. At intervals of 15 min, 5 mL of suspensions were taken out and centrifuged at 2500 rpm for 10 min to remove the TiO 2 /Ag 3 PO 4 composites. The changes of RhB concentration during xenon light irradiation were determined by using a TU-1901 ultraviolet-visible spectrophotometer at the maximum absorption wavelength of RhB (554 nm) with deionized water as the reference solution. For comparison, the commercial TiO 2 powder ($70.9% anatase and $29.1% rutile), and as-prepared Ag 3 PO 4 powder were also used as the photocatalytic references. The stability and recyclability of the TiO 2 / Ag 3 PO 4 composites were investigated by the degradation experiments of the 10 ppm RhB solution (150 mL).
Morphology of the TiO 2 /Ag 3 PO 4 composites and the pure Ag 3 PO 4 specimens was determined by FESEM. The FESEM  Fig. 3(c). Fig. 3(d) shows the representation FESEM image of the pH 5.5-TiO 2 / Ag 3 PO 4 composite prepared by mixed the pH 5.5-TiO 2 and Ag 3 PO 4 samples. As shown in Fig. 3(d), egg-like anatase nanocrystals with a size of 30-60 nm and 30-70 nm in length in high yield, some shuttle-like anatase nanocrystals with a size of about 20-50 nm in width and 60-150 nm in length, and cuboid-shaped anatase nanocrystals with a size of 30-70 nm in width and 55-90 nm in length are observed. The FESEM image in Fig. 3(e) shows that the pH 7.5-TiO 2 /Ag 3 PO 4 composite that has two main morphologies, shuttle-like anatase nanocrystals with 40-200 nm in length and 25-50 nm in width, and cuboid-shaped anatase nanocrystals with 40-95 nm in length and 30-55 nm in width. 20-95 nm) in diameter, and shuttle-like anatase nanocrystals with a length of about 30-210 (or 30-215 nm) nm and a width of about 20-65 (or 20-85 nm) nm for pH 9.5-TiO 2 /Ag 3 PO 4 (or pH 11.5-TiO 2 /Ag 3 PO 4 ) composites. FESEM images of Ag 3 PO 4 microcrystals are shown in Fig. 3(h) and (i), it can be seen that welldispersed irregular Ag 3 PO 4 polyhedrons with about 3-12 mm in length and 2.5-9.0 mm in width (or thickness), and cubic-like particles with the size about 1.5-7.0 mm were obtained. And the surface of the Ag 3 PO 4 crystals is rough, which is formed by the agglomeration of many nanoparticles with the size about 30-50 nm in diameter (Fig. 3(i)). Based on the above analysis, the Ag 3 PO 4 crystals were not observed in the TiO 2 /Ag 3 PO 4 composites, which can be ascribed to the fact that the sizes of Ag 3 PO 4 crystals were micrometer while the anatase TiO 2 crystals were nanometer, and TiO 2 nanocrystals were bound to the surface of Ag 3 PO 4 microcrystals.
The FESEM images and the corresponding elemental distribution maps of TiO 2 /Ag 3 PO 4 composites were achieved by energy dispersive spectrometer (EDS). As shown in Fig. 4, the appearance of Ag and P elements in EDS further demonstrated successful impregnation of Ag 3 PO 4 . The analysis of the results shows the atomic ratio of Ag to Ti is about 1 : 27.86, 1 : 71.55, 1 : 58.27, and 1 : 121 for pH 0.5-TiO 2 /Ag 3 PO 4 , pH 3.5-TiO 2 / Ag 3 PO 4 , pH 7.5-TiO 2 /Ag 3 PO 4 , and pH 11.5-TiO 2 /Ag 3 PO 4 composites, respectively.
The TEM and HRTEM images further reveal the detailed surface morphology of the obtained TiO 2 /Ag 3 PO 4 composites products, as shown in Fig. 5 and 6. For pH 0.5-TiO 2 /Ag 3 PO 4 , shuttle-like anatase nanocrystals with the length of about 30-85 nm and the width of about 15-25 nm, and square rod-shaped anatase nanocrystals with the length of about 25-140 nm and the width of about 15-50 nm are observed (Fig. 5(a)), which corresponds to the results of FESEM This journal is © The Royal Society of Chemistry 2020 RSC Adv., 2020, 10, 24555-24569 | 24561 ( Fig. 3(a)). The square rod-shaped nanocrystals with a lattice of 0.353 nm (or 0.359 nm) can be indexed to the (101) planes of the anatase, and the egg-like nanoparticle with a lattice fringes of 0.359 nm also can be indexed to the (101) planes of the anatase (Fig. 5(b)-(f)). The lateral planes of square rod-shaped nanocrystals are parallel to (101) planes, indicating that the exposed facets are {101} facets (Fig. 5(b) and (c)). The lattice fringe has d-spacing values of 0.236 and 0.246 nm, corresponding to (004) and (103) planes of anatase TiO 2 , respectively (Fig. 5(c)). The long axis of the shuttle-like anatase nanocrystals is perpendicular to (004) planes, indicating that the exposed facets are {001} facets of the top and bottom planes (Fig. 5(c)). In Fig. 5(e), the lattice fringes of the irregular crystals with lattice spacings of 0.235 and 0.353 nm can be assigned to the (004) and (101) planes of the anatase TiO 2 , respectively. And the angle between the (004) and (101) facets is 68 , implying that the irregular crystals expose {010} facets on its surface. The coexistence of various morphologies of the pH 3.5-TiO 2 /Ag 3 PO 4 composites was further investigated by TEM and HRTEM, as shown in Fig. 5(j-l). For the cuboid-shaped anatase nanocrystals, the TEM images depict the nanocrystals with 15-50 nm in length and 15-30 nm in width ( Fig. 5(g)), and the lattice fringe has d-spacing values of 0.353 (or 0.360) and 0.353 nm, corresponding to (101) and (011) planes of anatase TiO 2 , respectively ( Fig. 5(h) and (l)). The interior angle between (101) and (011) planes of 82 is in good agreement with the theoretical value, which indicates that the preferentially exposed crystal facets of the cuboid-shaped anatase is perpendicular to [111] crystal zone axis (expressed as [111]-facets). For the shuttle-like anatase nanocrystals, the TEM images depict the nanocrystals with 15-120 nm in length and 10-45 nm in width (Fig. 5(g)), and the lattice fringe has d-spacing values of 0.360 nm, corresponding to (101) planes of anatase TiO 2 (Fig. 5(k)). For the diamond-shaped anatase nanocrystals, the TEM images depict the nanocrystals with 35-85 nm in length and 15-35 nm in width ( Fig. 5(g)), and the lattice fringe has d-spacing values of 0.191 and 0.360 (or 0.364) nm, corresponding to (200) and (101) planes of anatase TiO 2 , respectively ( Fig. 5(h)-(j)). The lateral planes of the diamond-shaped anatase nanocrystals is parallel to (101) planes, indicating that the exposed facets are {101} facets of the lateral planes. For the square rodshaped anatase nanocrystals, the TEM images depict the nanocrystals with 40-135 nm in length and 20-30 nm in width ( Fig. 5(g)), and the lattice fringe has d-spacing values of 0.360 nm, corresponding to (101) planes of anatase TiO 2 (Fig. 5(l)). The top and bottom planes of square rod-shaped anatase nanocrystals are parallel to (101) planes, indicating that the exposed facets are {101} facets. Fig. 6(a)-(f) shows the TEM and HRTEM analysis results of the pH 7.5-TiO 2 /Ag 3 PO 4 composite. The size of cuboid-shaped anatase nanocrystals has a size about 30-60 nm in length and 20-35 nm in width, as shown in Fig. 6(a). The size of shuttle-like anatase nanocrystals is about 25-250 nm in length and 20-75 nm in width ( Fig. 6(a) and (d)), and the lattice fringe of 0.349, 0.476 and 0.360 nm corresponds to the distance between two adjacent (10-1), (002) and (101) planes of anatase TiO 2 , and the intersection angles between (10-1) and (002), (101) and (002), and (101) and (10-1) planes are 68.3 , 68.3 , and 43.4 , respectively, as shown in Fig. 6(b). The high crystallized shuttle-like TiO 2 surfaces with the clear lattice fringes of the anatase phase are also observed from Fig. 6(c) and (f). Two set of nonparallel lattice fringes with the d-spacing values of 0.349 and 0.238 nm, corresponding to (101) and (004) atomic planes of anatase phase (Fig. 6(c)). The lattice spacing of 0.352 and 0.476 nm of the truncated shuttle-like TiO 2 anatase TiO 2 , corresponding to the distance between two adjacent (101) or (002) planes, and the intersection angle between (101) and (002) planes is 68.3 , as shown in Fig. 6(e). Based on the above TEM and HRTEM analysis and the Wulff construction model, the shuttle-like anatase TiO 2 nanocrystals preferentially expose the {010} facets, {101} facets, and {001} facets on the four lateral planes, the eight isosceles trapezoid planes, and the two top/ bottom surfaces, respectively, and the directional grown direction is along the [001]-direction. The size of shuttle-like (or cuboid-shaped) anatase nanocrystals is about 50-180 nm (or 25-100 nm) in length and 25-50 nm (or 20-80 nm) in width, as shown in Fig. 6(g, j and m). {010} facets exposed TiO 2 exhibits a typical shuttle-like morphology with lattice fringes of 0.353 (or 0.360) and 0.174 (or 0.238, 0.482) nm attributed to (101)  The morphology and microstructure of the Ag 3 PO 4 crystals were further analyzed by TEM and HRTEM images, as shown in Fig. 7. As can be seen in Fig. 7(a), the obtained Ag 3 PO 4 crystals contains some irregular polyhedrons with the lengths of 1.0-3.7 mm and a cubic-like crystals with the lengths of about 1.75 mm and the widths of about 1.45 mm, respectively, which is in agreement with the results observed by the SEM images ( Fig. 3(h) and (i)). The lattice fringes of 0.269 (or 0.262) and 0.247 (or 0.239) nm match well with the (210) and (2-1-1) (or (211)) planes of irregular polyhedral Ag 3 PO 4 crystals, respectively ( Fig. 7(b)-(d)). And the angle between the (210) and (2-1) facets of 57 agrees well with the theoretical value 56.8 , according to calculated result from the lattice constants of Ag 3 PO 4 (cubic, space group P4 3n, JCPDS 06-0505, and a ¼ 6.013 A). Based on the above TEM and HRTEM analysis, the Ag 3 PO 4 specimens in the TiO 2 /Ag 3 PO 4 composites were not observed, which can be attributed to the deposition of nanoscale anatase TiO 2 crystals on the microsized Ag 3 PO 4 crystals via an in situ precipitation process.

Growth mechanism of the TiO 2 /Ag 3 PO 4 composites
According to the results of the XRD, SEM and HR-TEM observation, the possible growth mechanism for the formation of TiO 2 /Ag 3 PO 4 hybrids can be expressed as follows.
Acidic condition is benecial for reactions (3) and (4), neutral and basic conditions are favorable for reactions (5) and (6). In this process, the [Ti 4 O 9 ] 2À /[Ti 2 O 5 ] 2À nanosheets were transformed rstly to nanosheet-like anatase TiO 2 crystals by an in situ topotactic dehydration reaction. 37 Then the nanosheetlike anatase TiO 2 crystals were split into anatase TiO 2 nanocrystals with various morphologies and different exposed facets by dissolution-recrystallization process along their different planes.
The micro-sized Ag 3 PO 4 crystals were synthesized by using an ion-exchange method, using AgNO 3 and NH 4 H 2 PO 4 (3Ag + + H 2 PO 4 À ¼ Ag 3 PO 4 Y + 2H + ). The anatase TiO 2 nanocrystals with various morphologies and different exposed facets and Ag 3 PO 4 precipitate were well dispersed into deionized water under stirring to form suspension solution. The micro-sized Ag 3 PO 4 polyhedrons with larger particle surface, which could absorb more nano-sized anatase TiO 2 nanocrystals onto their surfaces via an in situ precipitation process to form the heterostructured TiO 2 /Ag 3 PO 4 composites.

UV-vis adsorption spectra of Ag 3 PO 4 , TiO 2 /Ag 3 PO 4 and TiO 2
The UV-visible absorption spectrum was applied to examine the optical properties of pure Ag 3 PO 4 , TiO 2 and TiO 2 /Ag 3 PO 4 composites. As observed in Fig. 8, the UV-Vis NIR spectrum of  pure TiO 2 sample only exhibits the fundamental absorption band edge (395 nm) in the UV light region, and the absorption band edge almost no more exists in the visible wavelength range. The pure Ag 3 PO 4 sample shows strong adsorption with absorption band edge at around 500 nm, which is equivalent to the band gap energy of 2.45 eV, in an good agreement with the results reported previously. 38 However, for the prepared TiO 2 / Ag 3 PO 4 composites at different values of pH, except for adsorption band edge (less than 408 nm) in the UV light region, a feature band edge (510 nm) of pure Ag 3 PO 4 appears in the visible light range based on the UV-Vis NIR spectrum. The absorption edges of TiO 2 /Ag 3 PO 4 composites are shied slightly toward higher wavelength relative to pure Ag 3 PO 4 , indicating TiO 2 in the composites is coupled to Ag 3 PO 4 . The above analysis show that the as-prepared TiO 2 /Ag 3 PO 4 composites can be used for visible light photocatalytic reactions.

Photocatalytic activities for the degradation of rhodamine B solutions
Recently, different types of photocatalysts, such as Mn-doped ZrO 2 , 39 carbon quantum dots, 40 MOFs, 11 BaTiO 3 , 41 were used to degrade the organic pollutants. In this study, the photocatalytic activities of the TiO 2 /Ag 3 PO 4 composites were evaluated by degradation of the carcinogenic textile dye rhodamine B (RhB, adsorption band: 554 nm). The degradation efficiency of all the specimens is expressed as (c 0 À c t )/c 0 Â 100%, where c 0 and c t represent the initial and residual concentration of the RhB, respectively. Prior to illumination, the suspensions were magnetically stirred in the dark for 2 h to make the RhB dyes reach achieve adsorption-desorption equilibrium on the surface of TiO 2 /Ag 3 PO 4 composites. 42 The adsorption values (mol(RhB) g(TiO 2 /Ag 3 PO 4 ) À1 ) of RhB on the surface of TiO 2 / Ag 3 PO 4 composites were 4.0 Â 10 À6 , 7.0 Â 10 À6 , 5.5 Â 10 À6 , and 4.5 Â 10 À6 mol g À1 for pH 0.5-TiO 2 /Ag 3 PO 4 , pH 3.5-TiO 2 / Ag 3 PO 4 , pH 7.5-TiO 2 /Ag 3 PO 4 , and pH 11.5-TiO 2 /Ag 3 PO 4 samples, respectively. These results indicated that the enhancement order of adsorption binding of the RhB to the TiO 2 /Ag 3 PO 4 was pH 0.5-TiO 2 /Ag 3 PO 4 < pH 11.5-TiO 2 /Ag 3 PO 4 < pH 7.5-TiO 2 /Ag 3 PO 4 < pH 3.5-TiO 2 /Ag 3 PO 4 , and that the strong anchoring of the RhB onto the surface of pH 3.5-TiO 2 /Ag 3 PO 4 could improve the photocatalytic activity. The commercial TiO 2 powder ($70.9% anatase and $29.1% rutile) and Ag 3 PO 4 powder were used as the photocatalytic references. Fig. 9(a) shows the variation of the absorption of rhodamine B (RhB) in the presence of pH 0.5-TiO 2 /Ag 3 PO 4 composite under the Xe light irradiation for 120 min. The peak position at 554 nm gradually moved towards the short-wavelength direction (i.e., hypsochromic shi) and the intensity gradually decreased, indicating the partial N-de-ethylation and the destruction of structure of the polycyclic aromatic hydrocarbon by the gradual decolorization of the RhB solution. 43 Aer exposure to visible light for 120 min, the degradation of RhB was as follows: pH 3.5-TiO 2 /Ag 3 PO 4 (75.6%) > pH 0.5-TiO 2 / Ag 3 PO 4 (72.2%) > pH 7.5-TiO 2 /Ag 3 PO 4 (65.8%) > pH 11.5-TiO 2 / Ag 3 PO 4 (61.3%) > Ag 3 PO 4 (56.0%) > blank (10.8%) > the commercial TiO 2 (5.8%), as shown in Fig. 9(b). Obviously, the as-prepared TiO 2 /Ag 3 PO 4 composites exhibit enhanced photocatalytic performance for the degradation of RhB compared to the commercial TiO 2 powder and Ag 3 PO 4 powder. The enhanced photocatalytic performance can be attributed to the TiO 2 /Ag 3 PO 4 heterostructures, which can absorb more visible light and inhibit the recombination of photoelectrons and holes. 34 Fig. 10 shows a possible photocatalytic mechanism for the photodegradation of RhB over the TiO 2 /Ag 3 PO 4 heterostructures under visible light irradiation. The valence band (VB) potential (+2.90 eV vs. NHE) and conduction band (CB) potential (+0.45 eV vs. NHE) of Ag 3 PO 4 are more positive than those of TiO 2 (VB potential: +2.70 eV, and CB potential: À0.30 eV), which imply that the photon generated electrons (e À ) of TiO 2 nanocrystal will be quickly transferred to the CB of Ag 3 PO 4 crystal, whereas the photon generated holes (h + ) of Ag 3 PO 4 crystal will be migrated to the VB of TiO 2 nanocrystal under visible light irradiation. 30,44 The separation of the e À (in Ag 3 PO 4 crystal) and h + (in TiO 2 nanocrystal) inhibits the charge recombination, which leads to the improvement of the photocatalytic activity of TiO 2 /Ag 3 PO 4 composites. 32 The h + and e À have oxidation and reduction, respectively. Under visible light irradiation, the h + in the VB of TiO 2 nanocrystal can directly oxidize the organic dye RhB and the water molecules adsorbed to the surface of TiO 2 photocatalyst to form RhB oxidation and cOH radicals, respectively. 3 At the same time, the e À in the CB of Ag 3 PO 4 crystal can directly reduce the oxygen molecules adsorbed to the surface of Ag 3 PO 4 photocatalyst to form strong oxidizing capacity of hydrogen peroxide (H 2 O 2 ) to oxidize and degradation RhB. Moreover, Ag 3 PO 4 is reduced to Ag by e À in the photocatalytic process. The 10 mg L À1 RhB solution (10 ppm) was not completely degraded due to the addition of more RhB solution (150 mL) and fewer catalysts (75 mg), and the liquid level of RhB solution was far away from the light source (25 cm). However, TiO 2 exhibited very low photocatalytic activity for the photodegradation of RhB, only 5.8% degradation efficiency, even lower than 10.8% for the blank without any photocatalysts under the Xe light irradiation for 120 min, implying that the TiO 2 actually had no any photocatalytic activity. Based on the discussion results of TiO 2 and the blank, it is reasonable that the presence of photocatalyst has a shielding effect on the degradation of RhB under the Xe light irradiation. 45 Since the process of photodegradation of RhB solution followed the pseudo-rst-order reaction kinetics model, the tted pseudo-rst-order reaction plots, the correlation coefficient and the corresponding apparent rate constant (k app ) are shown in Fig. 9(c) and (d), respectively. The correlation coefficients (R 2 ) are 0.968, 0.971, 0.943, 0.955, 0.997, 0.939, and 0.977 for the blank, TiO 2 , Ag 3 PO 4 , pH 0.5-TiO 2 /Ag 3 PO 4 , pH 3.5-TiO 2 /Ag 3 PO 4 , pH 7.5-TiO 2 /Ag 3 PO 4 , and pH 11.5-TiO 2 /Ag 3 PO 4 , respectively. The pH 3.5-TiO 2 /Ag 3 PO 4 composite exhibited the highest k app value (12.0 Â 10 À3 min À1 ), which was approximately 24.0, 13.3, 1.6, 1.4, 1.2, and 1.1 times larger than those of the commercial TiO 2 (0.5 Â 10 À3 min À1 ), blank (0.9 Â 10 À3 min À1 ), Ag 3 PO 4 (7.7 Â 10 À3 min À1 ), pH 11.5-TiO 2 /Ag 3 PO 4 (8.8 Â 10 À3 min À1 ), pH 7.5-TiO 2 /Ag 3 PO 4 (10.2 Â 10 À3 min À1 ), and pH 0.5-TiO 2 /Ag 3 PO 4 (10.5 Â 10 À3 min À1 ) samples, respectively. The pH 3.5-TiO 2 / Ag 3 PO 4 composite had the highest k app value, indicating that the pH 3.5-TiO 2 /Ag 3 PO 4 composite had the highest photocatalytic activity.
The stability and recyclability of photocatalyst is one of the important parameters for its practical applications. Herein, the stability and recyclability of the pure Ag 3 PO 4 and TiO 2 /Ag 3 PO 4 composites were evaluated by examining their recyclability in the photodegradation of RhB. Fig. 11 exhibited the repetitive photocatalytic degradation of RhB solution (10 mg L À1 , 150 mL)  This journal is © The Royal Society of Chemistry 2020 RSC Adv., 2020, 10, 24555-24569 | 24565 during three sequential runs under identical conditions. Aer each run, TiO 2 /Ag 3 PO 4 and Ag 3 PO 4 photocatalysts were collected by centrifugation and washed with deionized water for several times and the fresh RhB solutions with the same concentration (10 mg L À1 ) were used for next run. The photocatalytic efficiency of the TiO 2 /Ag 3 PO 4 and Ag 3 PO 4 photocatalysts remained almost unchanged aer three successive experimental runs, indicating that the synthesized photocatalysts had remarkable stability.
As is well-known, the photocatalytic activity is not only related to the heterostructure of TiO 2 nanocrystals, but also inuenced by other factors, such as crystalline phase, crystalline size, crystallinity, specic surface area, exposed facets, and so on. 46À48 For the synthesized TiO 2 /Ag 3 PO 4 composites, the crystalline form (anatase) and proportion (w(TiO 2 ) : w(Ag 3 PO 4 ) ¼ 4 : 1) of TiO 2 are the same, indicating that the inuence of TiO 2 crystal form and proportion on photocatalytic activity in TiO 2 /Ag 3 PO 4 composites is negligible. In the TiO 2 /Ag 3 PO 4 composites, the average crystalline sizes of pH 0.5-TiO 2 /Ag 3 PO 4 , pH 3.5-TiO 2 /Ag 3 PO 4 , pH 7.5-TiO 2 /Ag 3 PO 4 , and pH 11.5-TiO 2 / Ag 3 PO 4 were 58.5, 68.5, 87.5 and 80.0 nm, respectively, by measuring 200 particles in the FESEM images with Particle Size Distribution Calculation Soware (Fudan University, China). And the specic surface areas were 32.6, 27.8, 21.8, 24.4 m 2 g À1 for pH 0.5-TiO 2 /Ag 3 PO 4 , pH 3.5-TiO 2 /Ag 3 PO 4 , pH 7.5-TiO 2 / Ag 3 PO 4 , and pH 11.5-TiO 2 /Ag 3 PO 4 , respectively. It is known that smaller crystal size and larger specic surface (favorable for the RhB adsorption) contribute to the enhancement of photocatalytic activity in the photochemical reaction, which is attributed to its strong oxidation-reduction capability and more active sites. 49,50 However, the pH 3.5-TiO 2 /Ag 3 PO 4 composite displayed the highest photocatalytic activity, although the crystal size (68.5 nm) is much bigger than that (58.5 nm) of the pH 0.5-TiO 2 /Ag 3 PO 4 composite, and the specic surface area (27.8 m 2 g À1 ) slightly smaller than that (32.6 m 2 g À1 ) of the pH 0.5-TiO 2 /Ag 3 PO 4 composite, indicating that it is also very signicant to establish a balance between crystal size and specic surface area to improve the photocatalytic performance. On the other hand, the crystallinity of the pH 0.5-TiO 2 /Ag 3 PO 4 composite is better than that of pH 3.5-TiO 2 /Ag 3 PO 4 composite, which inhibits the recombination of photogenerated charge carriers (photogenerated electrons and holes), resulting in a relative good photoactivity. 15 The in the pH 0.5-TiO 2 /Ag 3 PO 4 , pH 3.5-TiO 2 /Ag 3 PO 4 , pH 7.5-TiO 2 / Ag 3 PO 4 , pH 11.5-TiO 2 /Ag 3 PO 4 composites, respectively. Hence, the improvement of photocatalytic activity of the pH 3.5-TiO 2 / Ag 3 PO 4 can also be attributed to the coexistence of high-energy {001} and [111]-facets. According to the discussion above, the pH 0.5-TiO 2 /Ag 3 PO 4 composite possesses a relative small crystal size, large specic surface area, good crystallinity, and coexposed high-energy {001} and [111]-facets, the synergistic effects resulting in the highest photocatalytic activity of the pH 3.5-TiO 2 /Ag 3 PO 4 composite.

Conclusions
In summary, TiO 2 /Ag 3 PO 4 composites composed of anatase TiO 2 nanocrystals with co-exposed {101}, {010}/{100}, {001} and [111]-facets and Ag 3 PO 4 microcrystals with irregular polyhedrons and cubic-like crystals were successfully synthesized by combining hydrothermal and ion-exchange methods. The Ag 3 PO 4 microcrystals were used as a substrate to load the anatase TiO 2 nanocrystals on their surface and to form TiO 2 / Ag 3 PO 4 heterostructures. To investigate the photocatalytic performance, the carcinogenic RhB solution was selected as model pollutant because it widely used in the textile industry. Compared with the commercial TiO 2 and the pure Ag 3 PO 4 microcrystals, the heterostructured TiO 2 /Ag 3 PO 4 composites exhibited excellent photocatalytic activity for the degradation of rhodamine B under visible light irradiation, which can be attributed to the separation of the e À (in Ag 3 PO 4 crystal) and h + (in TiO 2 nanocrystal) inhibits the charge recombination. For the as-prepared TiO 2 /Ag 3 PO 4 composites, the pH 3.5-TiO 2 /Ag 3 PO 4 composite exhibited the highest photocatalytic activity which mainly attributed to the synergistic effects of its relative small crystal size, large specic surface area, good crystallinity, and co-exposed high-energy {001} and [111]-facets. Moreover, this study provides new way for the preparation of TiO 2 /Ag 3 PO 4 composite semiconductor photocatalysts with high energy crystal surfaces. However, although the as-prepared TiO 2 / Ag 3 PO 4 composites exhibited good stability, the photocatalytic performance needs to be further improved for their practical application.

Conflicts of interest
The authors declare no conict of interest.