Semiconductor photocatalysis, as a “green” technique for environmental contamination control, has attracted considerable attention since the pioneering work of Fujishima et al. ^[1]. Although TiO₂, as the earliest used photocatalyst, has been widely studied, its application suffers from the wide band gap that is only responsive to ultraviolet (UV) light, which constitutes only a small fraction of the solar spectrum. Currently, more attention has been devoted to the exploration and fabrication of visible-light-driven photocatalysts such as BiOI ^[2], Bi₂MO₆ (M = Mo or W) ^{[3, 4]}, MnO₂ ^[5], CdS ^[6], g-C₃N₄ ^[7]and AgX (X = Cl ^[8], Br ^[9], I ^[10]). Among them, Ag-containing nanomaterials are considered to be promising candidates (Ag₃VO₄ ^[11], AgSbO₃ ^[12], Ag₃PO₄ ^[13], Ag₂CO₃ ^[14] and Ag₆Si₂O₇ ^[15]) because of their narrow band gap and excellent photoconversion efficiency. Notably, the discovery of the Ag₃PO₄ photocatalyst has been almost universally considered as a major breakthrough in the field of visible-light-active photocatalysts with high photooxidative capability for water-splitting and organic dye decomposition. Recently, many studies have been performed to further enhance their photocatalytic activity. However, an insurmountable problem for nearly all the reported Ag-based semiconductor photocatalysts is their poor activity or stability during their photocatalytic process. Therefore, enhancing photocatalytic activity and stability for these Ag-based semiconductor photocatalysts is essential.

Formation of a heterojunction (or interface) over two semiconductors has already proven to be an effective way to improve photocatalytic activity or stability, according to previous reports on AgX/Ag₃PO₄ (X = Cl, Br, I) ^[16], TiO₂/Ag₃PO₄ ^[17], Ag₂O/Ag₂CO₃ ^[18], Ag₂CO₃/AgBr ^{[19, 20]}, Ag₃PO₄/BiOBr ^[21], g-C₃N₄/Ag₂CO₃ ^[22], GO-Ag₂CO₃^[23]and AgI/Ag₂CO₃ ^[24]. Theoretically, when a p-type semiconductor and n-type semiconductor form a p-n junction, a strong inner electric field will be formed near the junction, pointing from n toward p, owing to the arrangement of the high concentrations of negatively and positively charged ions. The difference of the electric potential in the electric field can enhance the separation of photogenerated electrons and holes, increasing the quantum efficiency of the photocatalytic reactions ^[29]. The p-n junction photocatalysts, p-CuBi₂O₄/n-TiO₂ ^[30], p-Ag₂O/n-Bi₂O₂CO₃ ^[31] and p-BiOI/n-TiO₂ ^[20], have been shown to exhibit a high photocatalytic efficiency compared with other heterojunctions.

Ag₂CO₃ has been confirmed to exhibit a high photocatalytic activity for organic pollutant degradation, because the incorporation of a nonmetallic p-block carbon element into Ag₂O can broaden the band gap, which can enhance the redox ability ^[32]. Therefore, Ag₂CO₃, as a p-type semiconductor, can form a p-n heterojunction with Ag₃PO₄, which is an n-type semiconductor. To the best of our knowledge, there have been no relevant reports concerning the Ag₃PO₄/Ag₂CO₃ heterojunction.

Inspired by the previous reports mentioned above, a series of novel Ag₃PO₄/Ag₂CO₃ p-n heterojunction composite photocatalysts were prepared through a simple coprecipitation process and then characterized by powder X-ray diffraction (XRD), transmission electron microscopy (TEM), energy-dispersive X-ray spectroscopy (EDS), ultraviolet-visible (UV-vis) diffuse reflectance spectroscopy and transient photovoltage (TPV) measurements. The photocatalytic activities of the samples were determined by measuring the degradation of rhodamine B (RhB) in an aqueous solution under UV-vis light irradiation. Additionally, the active species during photocatalytic reactivity were also investigated by the addition of free radical scavengers such as isopropanol (IPA), 1,4-benzoquinone (BQ) and sodium oxalate ^[33].

All chemicals (Sinopharm Chemical Reagent Co., Shanghai, China) were of analytical grade purity and used without further purification. Briefly, 252 mg NaHCO₃ and 312 mg NaH₂PO₄·2H₂O were dissolved in 80 mL deionized water under stirring. Then, 40 mL AgNO₃ aqueous solutions (approximately 0.3 mol/L) was added dropwise to the solution under stirring. The precipitate was collected by centrifugation, washed three times with deionized water after stirring for 12 h and dried in air at 70 °C for 12 h to yield the 40%-Ag₃PO₄/Ag₂CO₃ composite photocatalyst (40 mol% Ag₃PO₄ and 60 mol% Ag₂CO₃). For comparison, other samples, such as Ag₃PO₄, Ag₂CO₃, and Ag₃PO₄/Ag₂CO₃ with different Ag₃PO₄ concentrations (10%, 20%, 30%, 50%, 60%) were prepared in a similar procedure, but with different weight additives.

The crystal structure of the composite photocatalysts was characterized using a Bruker D8 Advance X-ray powder diffractometer under nickel-filtered Cu-K_α radiation operating at 40 kV and 40 mA. TEM images were recorded using a Tecnai G2 F20 transmission electron microscope at an accelerating voltage of 200 kV. UV-vis diffuse reflectance spectra (DRS) were measured using a Cary 5000 spectrophotometer (Agilent, Santa Clara, USA). The transient surface photovoltage data were recorded by a 500 MHz digital oscilloscope (TDS 3054C, Tektronix, Beaverton, OR, USA) in the procedure reported by Fan et al. ^[34]. We placed the powder sample on the ITO electrode and another ITO electrode was used to press it to obtain a film composed of the powder sample. A laser pulse (λ = 355 nm with a pulse width of 4 ns) using the third-harmonic from a Nd:YAG laser (Quantel Brilliant Eazy: BRILEZ/IR-10) was employed as the light source.

The photocatalytic performances of the prepared samples were evaluated by the degradation of RhB in a XPA-7 photochemical reactor (Xujiang electromechanical plant, Nanjing, China). A 500 W, high-pressure Xenon lamp with a 420-nm cutoff filter was used as the visible-light source. Photocatalyst (20 mg) was added to 50 mL of an aqueous solution of RhB with an original concentration of c₀ = 20 mg/L. The suspension was magnetically stirred for 30 min in the dark to obtain an adsorption-desorption equilibrium among the photocatalyst, organic substances and water prior to irradiation. During the whole process, the suspension was vigorously stirred in the photoreactor. Identical amounts (3 mL) of the suspension were removed and centrifuged at given intervals during illumination. The absorbance of the solution was measured using a UV-vis spectrophotometer (Agilent Cary 5000).

To detect the active species during photocatalytic reactivity, hydroxyl radicals (•OH), superoxide radicals (O₂^•−) and holes (h⁺) were investigated by adding 1.0 mmol/L IPA (a quencher of •OH), BQ (a quencher of O₂^•−) and sodium oxalate (a quencher of h⁺) into the water phase as a diagnostic tool. This method was similar to the former photocatalytic activity test.

Figure 1 shows the XRD patterns of the Ag₃PO₄/Ag₂CO₃ samples. The peaks at 18.5° (020), 20.6° (110), 32.6° (-101), 33.7° (-130), 37.1° (040) and 39.6° (031) are indexed to that of monoclinic Ag₂CO₃ (JCPDS 26-339). The peaks at 21.0° (110), 29.8° (200), 33.4° (210), 36.7° (211), 47.9° (310), 52.8° (222), 55.1° (320) and 57.4° (321) are indexed to that of cubic Ag₃PO₄ (JCPDS 70-702). No impurities were detected in Ag₃PO₄ and Ag₂CO₃. From Fig. 1, the diffraction peaks of both Ag₃PO₄ and Ag₂CO₃ can be observed in the Ag₃PO₄/Ag₂CO₃ composite samples when the Ag₃PO₄ concentration is below 40%. Along with the increase of the concentration of Ag₃PO₄, the intensities of the diffraction peaks of Ag₂CO₃ decrease and these of Ag₃PO₄ increase. When the concentration of Ag₃PO₄ was increased to 40%, there were no longer any diffraction peaks of Ag₂CO₃ and only that of Ag₃PO₄ was present, even with a further increase of the concentration of Ag₃PO₄.

Fig. 1. XRD patterns of the samples. (1) Ag2CO3; (2) Ag3PO4; (3) 10%-Ag3PO4/Ag2CO3; (4) 20%-Ag3PO4/Ag2CO3; (5) 30%-Ag3PO4/ Ag2CO3; (6) 40%(50%, 60%)-Ag3PO4/Ag2CO3.

Figure 2 shows the TEM images and the corresponding EDS spectra of the 40%-Ag₃PO₄/Ag₂CO₃ composite photocatalyst. Smaller Ag₃PO₄ particles were observed to be dispersed on the surface of larger Ag₂CO₃ particles. At these places, which are shown by arrows in Fig. 2(a) and (b), the interfaces were clearly connected by different materials, which are shown by the different light, shade and gray scale. The close proximity of the two semiconductors proved that a heterojunction was formed between Ag₃PO₄ and Ag₂CO₃ in the composites. To further distinguish the components of the composites, we employed EDS to confirm the compositions at selected areas. Only elemental P, not elemental C, was present at region D according to the quantification results of EDS. Elemental C, O, Ag and P, and the Cu signals resulting from the copper grid were present at region C. The atomic compositions of elemental C and P were 10.434% and 8.757%, respectively. These results indicate that Ag₂CO₃ and Ag₃PO₄ were present at region C, and Ag₂CO₃ was the dominant component. However, Ag₂PO₃ was the dominant component at region D. A core-shell Ag₂CO₃@Ag₃PO₄ structure was presumed in which the Ag₃PO₄ was coated on the Ag₂CO₃ surface.

Fig. 2. TEM images (a, b) and the corresponding EDS spectra (c, d) of 40%-Ag3PO4/Ag2CO3 composite photocatalyst.

Figure 3(a) shows UV-vis diffuse reflectance spectra of the as-prepared photocatalysts, which reflects the ability to harvest solar energy. The spectra displays that pure Ag₂CO₃ and Ag₃PO₄ have absorption edges at approximately 474 nm and 520 nm, respectively. When the Ag₃PO₄/Ag₂CO₃ heterostructure was formed, a mixed absorption character, owing to the spectral superposition of Ag₂CO₃ and Ag₃PO₄, was observed. In particular, when the content of Ag₃PO₄ was 30% and 40%, the absorption intensity of the heterostructure catalyst was remarkably enhanced in the visible region, compared with that observed for pure Ag₃PO₄.

The band gap energies (E_g) of Ag₂CO₃ and Ag₃PO₄ were estimated using Eq. (1) ^[34]:

where A, ν, E_g and K are the absorbance, light frequency, band gap energy and a constant, respectively. Therefore, the E_g of Ag₂CO₃ and Ag₃PO₄ were calculated to be 2.62 and 2.40 eV from the plots of (Ahν)² versus hν, as shown in Fig. 3(b). The E_g of the 40%-Ag₃PO₄/Ag₂CO₃ composite was 2.55 eV, which was between that of Ag₂CO₃ and Ag₃PO₄.

Fig. 3. UV-vis diffuse reflectance spectra of photocatalysts (a) and the band gap energies (Eg) of photocatalysts (b).

In addition, the valence band (VB) and conduction band (CB) edge potentials of Ag₂CO₃ and Ag₃PO₄ were also determined to study the energy band matching, using the following empirical formulas ^[34:

where E_VB is the VB edge potential, E_CB is the CB edge potential, X is the electronegativity of the semiconductor, E_e is the energy of free electrons on the hydrogen scale (about 4.5 eV) and E_g is the band gap energy of the semiconductor. Herein, the E_VB were estimated to be 2.83 and 2.86 eV for Ag₂CO₃ and Ag₃PO₄, respectively. Correspondingly, the E_CB of Ag₂CO₃ and Ag₃PO₄ were further calculated to be 0.21 and 0.43 eV, respectively. Therefore, the energy bands of Ag₂CO₃ and Ag₃PO₄ are well matched as an interactive structure.

The photocatalytic activities of the photocatalysts were evaluated by measuring the degradation of RhB in aqueous solution. Fig. 4 showed that the 40%-Ag₃PO₄/Ag₂CO₃ composite photocatalyst exhibits the highest degradation rate among all photocatalysts: RhB was degraded completely under Xe lamp irradiation for 15 min. When the pure Ag₃PO₄ and Ag₂CO₃ were used as the photocatalysts, the degradation rates of the RhB were only 40% and 10%, respectively. Including additional Ag₃PO₄ coupled into Ag₂CO₃ was good for enhancing the degradation rate when the composition of Ag₃PO₄ was below 40%. However, when the Ag₃PO₄ composition exceeded 40%, the degradation rate started to decrease.

Fig. 4. Dynamic curve of RhB degradation over different photocatalysts under visible light. Reaction conditions: 500 W Xe lamp, 420 nm cutoff filter, RhB solution 20 mg/L, catalyst 50 mg.

We also investigated the photocatalytic durability of the 40%-Ag₃PO₄/Ag₂CO₃ photocatalyst through five cycles for the photocatalytic decomposition of RhB solution, and the corresponding results were shown in Fig. 5(a). Fig. 5(b) showed that the photocatalytic degradation ability for the 40%-Ag₃PO₄/ Ag₂CO₃ photocatalyst continuously decreased as the number of cycling tests increased, which is proposed to arise from the decomposition of Ag₃PO₄ and Ag₂CO₃ to metallic silver (Ag⁰), as has been reported in many studies ^[34]. The decrease in photocatalytic efficiency is negligible after the third cycle, which can be attributed to the formed Ag⁰ acting as a good trap for the electrons and thus enabling activity stability in the usage process ^{[33, 37]}.

Fig. 5. (a) Cycling runs in photocatalytic degradation of RhB in the presence of 40%-Ag3PO4/Ag2CO3; (b) XRD patterns of used 40%-Ag3PO4/Ag2CO3 after five cycles.

Figure 6 shows the trapping experiments of the active species during the photocatalytic reaction under visible-light irradiation. IPA is a well-known •OH scavenger, which can react quickly with all the •OH produced in the catalyst/RhB system. The results showed that adding IPA as an •OH scavenger did not change the decolorization rate of the RhB over Ag₃PO₄, Ag₂CO₃, and Ag₃PO₄/Ag₂CO₃ catalysts. This implies that the •OH radicals do not play a dominant role in RhB degradation.

Fig. 6. Trapping experiments of the active species during the photocatalytic reaction.

However, the addition of sodium oxalate caused a noticeable decrease in the photocatalytic activity for the three photocatalysts. Sodium oxalate is an h⁺ scavenger and quickly reacts with h⁺ at the interface of the catalysts, reducing the number of active species available for RhB decolorization. Furthermore, the addition of BQ (a quencher of O₂^•−) led to an obvious decrease in the photocatalytic activity for the Ag₃PO₄ and Ag₃PO₄/Ag₂CO₃ catalysts. Combining the results obtained by the additions of IPA, BQ and sodium oxalate as scavengers, it can be concluded that h⁺ and O₂^•− are the principal active species for RhB degradation. Specifically, h⁺ is the predominant contributor for all the photocatalysts. The influence of sodium oxalate on the photocatalysts has been evaluated according to the equation: I = (D₁/D₀) × 100%, where D₀ is the degradation ratio of RhB without any additives and D₁ is the degradation ratio of RhB with the addition of sodium oxalate. The value of I for the 40%-Ag₃PO₄/ Ag₂CO₃ composite catalyst was 15.6%, which is much lower than that of Ag₃PO₄ (44.8%) and Ag₂CO₃ (57.1%). All results suggest that the 40%-Ag₃PO₄/Ag₂CO₃ composite photocatalyst is able to produce more free h⁺ active species than Ag₃PO₄ and Ag₂CO₃.

The dynamic information of photoinduced charge carriers in the system can be obtained by TPV, which contributes to a further understanding of the photocatalytic ability of the composite photocatalysts. Fig. 7 shows the TPV spectra of the pure Ag₃PO₄, Ag₂CO₃ and 40%-Ag₃PO₄/Ag₂CO₃ composite photocatalysts. According to previous reports by Wang’s group ^[yes], a positive TPV response implies that photoinduced electrons transfer towards the bulk and photoinduced holes transfer to the surface and accumulate there. While a negative TPV response implies that photoinduced holes transfer towards the bulk and photoinduced electrons transfer to the surface and accumulate there.

Fig. 7. TPV of the samples. The wavelength and intensity of the laser pulse are 355 nm and 200 μJ, respectively.

Figure 7 shows some interesting features. (1) TPV signals of Ag₃PO₄ and 40%-Ag₃PO₄/Ag₂CO₃, as n-type semiconductor and p-n heterojunction, respectively, are positive, indicating the accumulation of positive charge near the surface. However, TPV signals of Ag₂CO₃, as a p-type semiconductor, are negative, suggesting negative charge accumulation near the surface. (2) It is typical for the surface photovoltage in semiconductors, where charge carriers are separated, to drift in the electric field of the surface space-charge region when the time is longer than 1×10⁻⁷ s. (3) Furthermore, an obvious retardation is observed in the TPV of 40%-Ag₃PO₄/Ag₂CO₃ in comparison with that of bare Ag₃PO₄ at times longer than 1×10⁻⁷ s. Hence, it is possible that the presence of the interface between Ag₃PO₄ and Ag₂CO₃ can reduce the transfer rate and inhibit the recombination of photoinduced charge carriers. (4) The time of the maximums (t_max) of the TPV can provide information about the recombination time of the photoinduced charge carriers. The value of t_max for the 40%-Ag₃PO₄/Ag₂CO₃ composite catalyst (1×10⁻⁷- 1×10⁻³ s) is higher than that of the Ag₃PO₄ catalyst (5.74×10⁻⁸-1×10⁻⁴ s), which indicates that the recombination rate of charge carriers in the 40%-Ag₃PO₄/Ag₂CO₃ composite catalyst is lower than that in the Ag₃PO₄ catalyst.

On the basis of the above data, it is believed that a p-n heterojunction is successfully formed in the 40%-Ag₃PO₄/Ag₂CO₃ composite catalyst. This results in the effective separation of photoinduced electron-hole pairs and the low recombination rate of charge carriers, which correspond to its highest photocatalytic activity.

Based on the above calculations and experimental results, a rough schematic energy band model of the Ag₃PO₄/Ag₂CO₃ composite photocatalyst can be proposed, as depicted in Fig. 8. The CB position of Ag₂CO₃ is more negative than that of Ag₃PO₄ and the VB position of Ag₃PO₄ is more positive than that of Ag₂CO₃. Thus, under visible-light irradiation, the excited electrons in the CB of Ag₂CO₃ can rapidly transfer to the CB of Ag₃PO₄. At the same time, photoinduced holes in the VB of Ag₃PO₄ can transfer to the VB of Ag₂CO₃. However, the formation of the p-n heterojunction greatly increases the separation of photoexcited electrons and holes, suppressing electron-hole recombination, which results in the enhanced photocatalytic performance of the 40%-Ag₃PO₄/Ag₂CO₃ composite.

Fig. 8. Proposed photocatalytic mechanism for Ag3PO4/Ag2CO3 photocatalyst.

A series of novel Ag₃PO₄/Ag₂CO₃ composite photocatalysts were synthesized through a simple coprecipitation process and the 40%-Ag₃PO₄/Ag₂CO₃ composite photocatalyst showed impressive photocatalytic behavior compared with both Ag₃PO₄ and Ag₂CO₃ in the degradation of RhB under visible-light irradiation. The enhanced photocatalytic activity of 40%-Ag₃PO₄/ Ag₂CO₃ composite arose from the fact that a p-n heterojunction was successfully formed between n-type Ag₃PO₄ and p-type Ag₂CO₃, which led to the effective separation of photoinduced electron-hole pairs and the low recombination rate of charge carriers. Therefore, more free h⁺ active species could be produced to oxidize the organic pollutant.