Exploring hierarchical porous silica-supported Ag₃PO₄ as high-efficient and environmental-friendly photocatalytic disinfectant

Tetraethyl orthosilicate (TEOS), triethanolamine (TEA), cetyltrimethylammonium p-toluenesulfonate (CTAT), 3-mercaptopropyltrimethoxysilane (MPTMS) and other chemicals were supplied by Sigma-Aldrich and used without any purification. De-ionized (DI) water was used though all experiments. E. coli was provided by the University of Adelaide.

h-SiO₂ NPs were synthesized through a soft-templating method with slight modification using TEOS as silica precursor, TEA as base catalyst, CTAT as structure-directing agent [20, 24]. Typically, 0.48 g CTAT and 0.087 g TEA were dissolved in 25 mL DI water (80 °C). 3.9 mL TEOS was then quickly added into the above mixture and refluxed for 2 h. White precipitates were collected by centrifugation and washed with DI water and then dried at 60 °C overnight. After that, surfactant templates were removed by calcination in air at 700 °C for 3 h at a ramp rate of 2 °C min⁻¹ to obtain the hierarchical silica.

h-SiO₂ surface modification with thiols (HS-h-SiO₂) was achieved by grafting MPTMS onto the above-prepared hierarchical silica [21]. 0.1 g h-SiO₂ was added into 2.5 ml n-hexane containing 0.2 g MPTMS, agitated for 18 h, centrifuged and then washed with absolute ethanol for three times.

Deposition of Ag₃PO₄ to the surface-functionalized h-SiO₂ was conducted using an in situ precipitation method. Briefly, HS-h-SiO₂ was dispersed in 10 mL AgNO₃ aqueous solution (1.8 mM). Then, the dispersion was agitated for 24 h at room temperature, allowing adsorption equilibrium. The silver ion saturated HS-h-SiO₂ was then filtered and re-suspended in 10 ml DI water. After that, 20 mL Na₂HPO₄ aqueous solution (0.6 mM) was added slowly, refluxed 3 h. Precipitates were collected, washed with DI water and ethanol and then dried overnight to obtain the Ag₃PO₄@h-SiO₂. Bulk Ag₃PO₄ NPs were synthesized using the same way without adding HS-h-SiO₂.

10 mg Ag₃PO₄@h-SiO₂ was dissolved in 5% HNO₃ for ICP-MS measurement, and actual Ag weight percentage was determined to be 11.07% (i.e. Ag₃PO₄ weight percentage of 15.82%) in Ag₃PO₄@h-SiO₂.

To better understand the prepared NPs, their morphology and structural analysis were studied by scanning electron microscopy (SEM), energy-dispersive X-ray (EDX), X-ray diffraction (XRD), and UV–vis. The sizes and specific surface areas were also investigated by dynamic light scattering (DLS) and Brunauer−Emmett–Teller (BET) methods.

To study the photocatalytic efficiencies of the prepared Ag₃PO₄@h-SiO₂, h-SiO₂ and Ag₃PO₄ NPs, methylene blue (MB) degradations were studied under visible-light irradiation with different nanoparticle dosages (0.010 g Ag₃PO₄@h-SiO₂, 0.0085 g h-SiO₂ and 0.0015 g Ag₃PO₄) dispersed in MB solution (20 mL, 15 μM) under vigorous agitation. All suspensions were firstly agitated in dark for 45 min to reach adsorption/desorption equilibrium before introducing light source. A 300 W Xe arc lamp (PerfectLight, PLS-SXE300C, Beijing) equipped with a UV cutoff filter (UVCUT420, λ > 420 nm) was used as the light source, which was located 20 cm from reaction solutions. To monitor the MB concentration change during the experiments, a specific amount of the suspension was collected at a set frequency, centrifuged and measured by using a UV–Vis spectrophotometer.

Escherichia coli was selected as a model pathogen to test disinfection efficiencies of the synthesized NPs. A chloride ion-free buffer solution was prepared and used through the whole disinfect experiments [18]. E.coli was first cultured to yield a suspension with viable cell density of about 10⁹ CFU mL⁻¹ [18]. Then, designed dosages of Ag₃PO₄@h-SiO₂ (2.0 mg), h-SiO₂ (1.7 mg) and Ag₃PO₄ (0.3 mg) were separately added into 20 mL of the above E coli suspension. During the experiment, the suspensions were agitated gently at room temperature and irradiated by visible light. At set time intervals, 0.1 ml aliquots were plated on agar plates. The viable E coli colonies were counted after those plates were incubated at 37 °C overnight. For comparison, dark control experiments without light irradiation and blank control in the absence of any NPs were performed. All experiments were carried out in triplicates. Kirby–Bauer method was also performed to compare the disinfection abilities of the synthesized NPs in a quantitative and visual way. The chosen pathogen was spread evenly on agar plates covered with small round filter disks carrying little volume (10 μL) of buffer solution and NP suspensions. Those agar plates were left in dark/ visible light for 10 min and then incubated overnight at 37 °C for observation.

The morphological change of E. coli before and after Ag₃PO₄@h-SiO₂ photo-disinfectant treatment was monitored by SEM. The bacteria (10⁹ CFU mL⁻¹) were inoculated in sterilized buffer solution containing 0.1 mg/mL Ag₃PO₄@h-SiO₂. The mixture was then irradiated under visible light for 50 min. After filtration, the critical point drying method was used to keep E. coli morphology [8, 18], and then observed using SEM technology.

The preparation of Ag₃PO₄@h-SiO₂ is schematically illustrated in Scheme 1. First, monodisperse hierarchical porous silica nanoparticles were synthesized through a soft-templating method [19]. h-SiO₂ nanoparticles were further surface functionalized with thiols by a post-synthesis grafting method. Silver ions were then loaded to thiols, and subsequently reacted with HPO₄²⁻ to in situ form Ag₃PO₄ nanocrystals inside the pores of h-SiO₂.

Morphologies of these NPs were examined by SEM and TEM. Figure 1a indicates that h-SiO₂ exhibits a brain-like spherical morphology with large pores and wrinkled surfaces. h-SiO₂ NPs have a uniform particle size of ~ 120 nm (Figure S1, SI) together with good dispersity in solution. TEM image (Fig. 1b) confirms the existence of hierarchical structures in h-SiO₂ NPs. Figure 1c shows that the Ag₃PO₄@h-SiO₂ well maintains the wrinkled surface, porous structure and good stability. The large specific surface area and hierarchical porous structure of Ag₃PO₄@h-SiO₂ provide promising microenvironment beneficial for molecular interfacial reaction, molecular adsorption capacity, photocatalytic kinetics and consequently photo-disinfection ability. The identical particle sizes of Ag₃PO₄@h-SiO₂ and h-SiO₂ (Figure S1, SI) indicate the presence of Ag₃PO₄ NPs inside the pores of h-SiO₂ rather than on surface. TEM image of Ag₃PO₄@h-SiO₂ (Fig. 1d) clearly depicts that Ag₃PO₄ are well scattered in the large pores of h-SiO₂ with an average size of ~ 10 nm, which is much smaller than bulk Ag₃PO₄ NPs being prepared in the absence of h-SiO₂ (Figure S2, SI). Moreover, due to the presence of thiols on h-SiO₂ surface, Ag₃PO₄ can be stably chelated in the pores of h-SiO₂. Elemental analysis was studied by EDX (Figure S3, SI), where oxygen and silicon peaks come from h-SiO₂. Besides, the observation of Ag and P peaks confirms the formation of Ag₃PO₄@h-SiO₂.

The crystallographic analysis of the h-SiO₂, Ag₃PO₄ and Ag₃PO₄@h-SiO₂ was examined by XRD (Fig. 2). All diffraction peaks of pristine Ag₃PO₄ can be indexed with the body-centered cubic (BCC) structure, which agree well with the JCPDS card no. 06–0505 [26]. In addition, it has been observed that the diffraction peaks for pristine Ag₃PO₄ and amorphous h-SiO₂ are well presented in the XRD pattern of Ag₃PO₄@h-SiO₂, confirming the successful preparation of h-SiO₂@Ag₃PO₄.

The UV–Vis study of the h-SiO₂, Ag₃PO₄ and Ag₃PO₄@h-SiO₂ has been presented in Figure S4 (SI). Bulk Ag₃PO₄ shows an apparent visible-light absorption band [26, 27]. h-SiO₂ exhibits no visible-light response because of its broad band gap of ~ 9 eV [28]. However, Ag₃PO₄@h-SiO₂ exhibits obvious visible-light absorbance, implying it is a potential visible-light inducing photocatalyst.

Nitrogen adsorption–desorption isotherms and pore size distributions (PSD) of h-SiO₂ and Ag₃PO₄@h-SiO₂ show type-IV isotherms with narrow hysteresis loops (Fig. 3) [29]. The BET surface areas of h-SiO₂ and Ag₃PO₄@h-SiO₂ are calculated to be 254 and 40 m²g⁻¹. The decrease in pore volume from 0.489 for h-SiO₂ to 0.135 cm³g⁻¹ for Ag₃PO₄@h-SiO₂ confirms that Ag₃PO₄ have been loaded to the pores of h-SiO₂. The size of loaded Ag₃PO₄ NPs of about 10 nm is much smaller than that being prepared in bulk (Figure S2, SI), and smaller Ag₃PO₄ NPs may provide larger specific surface areas, beneficial for enhancing its photocatalytic disinfection performance.

Photocatalytic efficiencies of the Ag₃PO₄@h-SiO₂ were preliminarily studied by degrading a model organic dye methylene blue (MB) under visible-light irradiation (Fig. 4). From the ICP-MS results, the silver content in Ag₃PO₄@h-SiO₂ is ~ 11 wt%, i.e. there is ~ 15 wt% of Ag₃PO₄ and ~ 85 wt% of h-SiO₂ in the Ag₃PO₄@h-SiO₂. In the systematic photocatalytic evaluation, comparable dosages of Ag₃PO₄@h-SiO₂ (0.5 mg/mL), h-SiO₂ (0.425 mg/mL, 85 wt% of Ag₃PO₄@h-SiO₂) and Ag₃PO₄ (0.075 mg/mL, 15 wt% of Ag₃PO₄@h-SiO₂) were used for MB degradation. Blank control experiments were conducted as a reference. All suspensions were agitated in dark for 45 min to achieve adsorption/desorption equilibrium before visible light was introduced (Fig. 4, inset). h-SiO₂ achieves the highest MB adsorption (~ 69.2%) compared to Ag₃PO₄@h-SiO₂ (~ 52.2%) and bulk Ag₃PO₄ (~ 23.7%), which can be attributed to their different specific surface areas. Beyond the adsorption/desorption equilibrium, photocatalytic experiments of MB degradation were initiated by introducing visible light. After 35 min of visible-light irradiation, there was no change for MB alone, showing MB cannot be remarkably degraded by visible light. Similarly, MB was not degraded in the presence of h-SiO₂, which can be ascribed to its low responsiveness to visible light. Using bulk Ag₃PO₄, MB can be degraded completely within 35 min. It is notable that only 15 min is required to achieve complete MB degradation using the Ag₃PO₄@h-SiO₂ under same experimental condition. These enhanced photocatalytic MB degradation of Ag₃PO₄@h-SiO₂ is attributed by its superior structure advantage, where the hierarchical structures of Ag₃PO₄@h-SiO₂ facilitate MB physical adsorption and interfacial photocatalytic degradation associated with its high porosities and large surface areas.

The disinfection performance of Ag₃PO₄@h-SiO₂ as a photocatalyst was investigated using E. coli as a representative pathogen. To avoid protein adsorption and AgCl precipitation, all disinfection evaluations were conducted in chlorine-free buffer media [4, 18]. Prior to experiments, E. coli were thoroughly washed and re-suspended in chlorine-free buffer solution.

Like previous MB photocatalytic experiment, comparable dosages of Ag₃PO₄@h-SiO₂ (100 μg/mL), h-SiO₂ (85 μg/mL, ~ 85% of Ag₃PO₄@h-SiO₂) and Ag₃PO₄ (15 μg/mL, 15% of Ag₃PO₄@h-SiO₂) were used in disinfection tests at room temperature. Figure 5 shows the E. coli disinfection profiles for the above NPs under visible-light irradiation. As references, dark controls were performed under the same NP dosages without illumination, and blank control was carried out under visible-light irradiation but without adding any NPs. Like MB degradation, visible-light irradiation does not display any disinfection activity without NPs. Due to its poor response to visible light, h-SiO₂ shows identical 15.2% disinfection over 50 min under both illumination and dark conditions. Slight SiO₂ toxicity to both E. coli and B. subtilis under both light and dark conditions has been reported by Laura [30], and it also conforms with its low photocatalytic ability on MB degradation. Experimental data show that 47.2% E. coli are killed by bulk Ag₃PO₄ under 50 min visible-light irradiation, while 25.3% E. coli death in the dark control. Ag₃PO₄ has a BCC structure and low solubility in water [8, 13, 26], and the released silver can kill bacteria. [31, 32] Thus, the dark killing of bulk Ag₃PO₄ is mainly due to released silver [8, 12], while its enhanced disinfection activity under visible-light illumination is ascribed to its additional visible-light-driven photocatalytic ability [9, 33], 34. Therefore, both photo-responsiveness and sliver release dominate the overall disinfection of bulk Ag₃NO₄.

Figure 5 reveals the remarkable disinfection performance of Ag₃PO₄@h-SiO₂, where ~ 85.3% disinfectant efficiency can be achieved in 50 min visible-light irradiation by adding 0.1 mg/mL of Ag₃PO₄@h-SiO₂. At a comparable dosage of NPs, Ag₃PO₄@h-SiO₂ shows approximately double disinfection efficiency over bulk Ag₃PO₄ under 30 min visible-light irradiation. Interestingly, the Ag₃PO₄@h-SiO₂ demonstrates much lower disinfection efficiency in dark than bulk Ag₃PO₄. These results indicate that Ag₃PO₄@h-SiO₂ are more physically and chemically stable than bulk Ag₃PO₄ as evident from the low silver release from Ag₃PO₄@h-SiO₂. Hence, Ag₃PO₄@h-SiO₂ presents obvious visible-light induced photocatalytic disinfection activity, largely surpassing h-SiO₂ and bulk Ag₃PO₄. These disinfection results match well with the superior photocatalytic MB degradation ability of Ag₃PO₄@h-SiO₂. It has been documented that the disinfectant activity of Ag-based NPs is size-dependent, and smaller particles always present higher disinfection efficiency when an equivalent amount of silver mass is used as small particles size gives a large surface area to volume ratios [12, 35]. Our results reveal that Ag₃PO₄@h-SiO₂ exhibits higher stability with much lower silver release than bulk Ag₃PO₄. That result can be ascribed to the presence of functional thiols on h-SiO₂ surface, which makes Ag₃PO₄ NPs more stable [21, 36]. The high chemical and physical stability without obvious silver leaching is an environmental-friendly and economic outcome of using Ag₃PO₄@h-SiO₂ as a potential photocatalytic disinfectant for water and wastewater treatment. Most importantly, minimization of silver release does not sacrifice the overall disinfection performance of Ag₃PO₄@h-SiO₂ due to the high surface areas of incorporated Ag₃PO₄. Therefore, the disinfection performance of hierarchical Ag₃PO₄@h-SiO₂ under visible-light irradiation is dominated by its enhanced photo-disinfection activity.

The above experimental findings were further confirmed by another qualitative disinfection tests using the Kirby–Bauer method [18]. Similarly, comparable dosages of NPs (4.25 mg/ml h-SiO₂, 0.75 mg/ml Ag₃PO₄ and 5.00 mg/ml Ag₃PO₄@h-SiO₂) were used based on their contents in Ag₃PO₄@h-SiO₂. Figure 6 indicates that tests with buffer solution only and h-SiO₂ do not show obvious inhibition zones under both dark and visible-light irradiation. Those results agree well with those disinfection activities in liquid medium. Bulk Ag₃PO₄ exhibits a more obvious inhibition zone with diameter of 8 mm than that of Ag₃PO₄@h-SiO₂ in dark, confirming its higher silver release from bulk Ag₃PO₄. Superior photocatalytic disinfection performance of Ag₃PO₄@h-SiO₂ under visible-light irradiation can be confirmed by a noticeable inhibition zone (13 mm), which is much larger than that of bulk Ag₃PO₄.

To further investigate the role of Ag₃PO₄@h-SiO₂ during the photo-disinfection process, morphological change in E. coli was studied by SEM before (Fig. 6c-d) and after (Fig. 6e-f) exposing the pathogen strain to Ag₃PO₄@h-SiO₂ suspensions under visible-light radiation. The pristine E. coli strain exhibits a well-defined IB structure and an evenly distributed interior content, indicating the structural integrity of bacteria [35]. However, after exposure to Ag₃PO₄@h-SiO₂ and 50 min visible-light irradiation, E. coli are severely damaged and show a typical shape with collapsed outer membrane and indented surface, confirming the superior photo-disinfection activity of Ag₃PO₄@h-SiO₂.

From the above evaluation, the excellent photocatalytic disinfection ability of Ag₃PO₄@h-SiO₂ is associated with its chemical and structure advantages. Based on energy levels, a plausible mechanism diagram is depicted in Fig. 7a. Under visible-light irradiation, the photo-excited electrons in nanostructured Ag₃PO₄ jump from its valence band (VB) to conduction band (CB) (Eq. 1) [37‐39], whereas the left holes can be trapped by OH⁻ or H₂O to generate reactive oxidant species such as OH· and H₂O₂ (Eqs. 2‐4) [27],40. In addition, the accumulated photo-excited electrons in the CB involve a multiple-electron reduction reaction with O₂ to yield H₂O₂ (Eq. 5) [18, 41, 42].

For the visible-light irradiated Ag₃PO₄@h-SiO₂ aqueous system, H₂O₂, OH· and photo-generated holes are potential reactive species for pathogen disinfection [43]. To further understand detailed Ag₃PO₄@h-SiO₂ disinfection mechanism, various quenching experiments were conducted by introducing different scavengers of 2-propanol, Cr (VI), EDTA and Fe (II). 2-propanol is an OH· quencher to suppress OH·-mediated reactions [43]. The formation of OH· radical can be further confirmed by photoluminescence technique (Figure S5). EDTA is a scavenger for photo-generated holes [39]. Cr (VI) can capture photo-excited electrons to eliminate H₂O₂ generation (Eqs. 4 and 5) [44]. Fe (II) enhances OH· generation through the Fenton reaction [45].

Figure 7b compares the time-dependent photo-disinfection of E. coli exposure to Ag₃PO₄@h-SiO₂ and visible-light irradiation with and without scavengers. The disinfection efficiency decreases slightly after adding 2-propanol as an OH· scavenger, implying OH· is involved in the photo-disinfection but not dominative [46]. More drop of photo-disinfection efficiency after introducing Cr (VI) as an e⁻ scavenger suggests the H₂O₂ generated via the multiple-electron reduction reaction with O₂ near the CB of Ag₃PO₄ (Eq. 5) also contribute the disinfection process. However, the addition of Fe (II) as a H₂O₂ scavenger leads to obvious decrease on photo-disinfection efficiency, indicating H₂O₂ plays a large part in the disinfection process. Through the Fenton reaction with H₂O₂, Fe (II) introduction promotes OH· generation [43] but the obvious reduction on photo-disinfection after adding of Fe (II) in this study confirms the important role of H₂O₂ rather than OH·. In addition, the disinfection efficiency greatly drops when introducing EDTA as a quencher of photo-generated holes, indicating holes play a major role in this photo-disinfection process because of their intrinsic oxidative ability. Therefore, h⁺ and H₂O₂ are dominative reactive species for photocatalytic disinfection in the presence of Ag₃PO₄@h-SiO₂ [47].