Durable Ag/AgCl nanowires assembled in a sponge for continuous water purification under sunlight

Abstract

A Ag/AgCl nanowire (NW) sponge and a confined flow design were put forward to improve the photodegradation efficiency of immobilized Ag/AgCl NW catalysts. In consideration of the high cost of Ag/AgCl NWs, an in situ recovery strategy was also provided to extend the lifetime of the Ag/AgCl NW sponge.

---

Conceptual insights

We for the first time fabricated robust 3D Ag/AgCl nanowire (NW) networks on a polymer sponge with interconnected macropores through a low temperature welding method. This binder free method avoids contaminating the surface of Ag/AgCl NWs. Compared to 2D films, these 3D Ag/AgCl nanowire networks expose more catalyst surface and their unique open macropore systems also benefit the fast diffusion of reactants to the catalyst surface. We put forward a confined flow design for the photoreactor based on the above 3D Ag/AgCl NW networks to shorten the diffusion distance of organic pollutants to the catalyst surface and at the same time enhance the mass transfer speed. This design improves the photodegradation efficiency more than two times compared with conventional lamellar flow design. This photoreactor can be easily scaled up to one square meter size. To the best of our knowledge, this has been the first huge photoreactor immobilized with Ag/AgCl nanostructures so far. Furthermore, our proposed in situ chemical recovery strategy, which can extend the lifetime of the immobilized Ag/AgCl NW and thus reduce the material consumption, makes the Ag/AgCl NW based photoreactor more cost efficient and more competitive.

Photocatalytic water purification by using sunlight as the energy source has been widely considered as a low-cost, environment friendly and sustainable technology for the decomposition of toxic organic pollutants into innocuous products (e.g., CO₂, H₂O).^1–3 One obstacle to its practical application is the inefficient use of visible-light which accounts for 43% of total sunlight. Silver halide is unstable under sunlight and has been used as a photosensitive material in the field of photographic films for many years. Recently, Wang et al.⁴ found that after reducing some Ag⁺ ions on the surface region of AgCl particles to Ag⁰ species, the AgCl particles with silver nanoparticles formed on the surface exhibited ultra-high photocatalytic activity under visible-light. The degradation rate of methyl orange (MO) over Ag/AgCl hybrid particles under visible-light irradiation was faster than that over N-doped TiO₂ by a factor of eight. This high performance is due to the surface plasmon resonance of silver nanoparticles which improves the absorption of visible light, and the enhanced separation of photogenerated electrons and holes at the metal-semiconductor interfaces.^5,6 Besides the high photodegradation performance, the physicochemical structure of Ag/AgCl hybrid particles also exhibited high stability under UV and visible light, their photocatalytic activity experienced little change even after ten times of repeated usage. After this study, various Ag/AgCl based photocatalysts were fabricated and also showed high photodegradation efficiency to organic species under visible light.^7–17 Ag/AgCl based photocatalysts have been considered as promising materials for high speed water purification by using sunlight as the energy source due to their extremely high photodegradation efficiency to organic compounds.

To promote the practical application of Ag/AgCl based photocatalysts in the industrial process, the complex and costly post-separation of these nanoscale particles from water should be solved.¹ The immobilization of Ag/AgCl nanostructures on the substrates of photoreactors is the most common approach to realize the instant separation of Ag/AgCl photocatalysts and water, and purify waste water in a continuous mode. However, directly mixing the Ag/AgCl nanostructures with binders and coating this mixture on the planner substrates of previously reported photoreactors^18–20 would result in 2D closely packed Ag/AgCl films with reduced active sites due to the contamination of the catalyst surface by binders and the low surface area of 2D planner substrates, which greatly decreases the photodegradation speed per unit area of the photoreactor. In addition, the large diffusion distance of organic molecules to the surface of Ag/AgCl films in the big reaction chambers of the photoreactors also reduces the photocatalytic water purification efficiency. Therefore, developing binder-free immobilization methods and novel photoreactors to improve the photocatalytic efficiency of immobilized Ag/AgCl nanostructures while having high throughput is highly in demand.

Herein, as shown in Scheme 1, we use a polymer sponge as the substrate of the photoreactor to immobilize Ag/AgCl nanowire (NW) networks (the resulting product is denoted as PS@Ag/AgCl NW), forming robust three dimensional (3D) Ag/AgCl NW networks. To avoid losing the intrinsic high performance of Ag/AgCl NWs, we developed a low temperature thermal welding method to immobilize intertwined Ag/AgCl NW networks on the backbones of the sponge. Compared to 2D Ag/AgCl hybrid films,²¹ 3D Ag/AgCl NW network structures expose more catalyst surface, and their unique open macropore systems facilitate the passage of water and photon transfer to the catalyst surface. Furthermore, we fabricated a new photoreactor, in which the void space is completely occupied by the 3D Ag/AgCl NW networks, to confine the contaminated water flowing only through the irregular interconnected macropores. The confined flow in the photoreactor not only significantly decreased the diffusion distance of pollutants to Ag/AgCl NW surfaces from centimeters to micrometers, but also increased the probability of contact between pollutants and Ag/AgCl NW surfaces due to the enhanced fluid disturbance. The high photodegradation efficiency of 3D Ag/AgCl NW networks in the confined flow design was tested through photodegradation of methyl orange (MO), and was 4 times higher than that of 2D Ag/AgCl films with a laminar gap. More importantly, the throughput of the photoreactor based on 3D Ag/AgCl NW networks with confined flow could reach up to 9600 L h⁻¹ m⁻², which is two orders higher than that of the microfluidic photoreactor (ca. 40 L h⁻¹ m⁻²).²² We further developed a facile in situ chemical recovery method to extend the lifespan of Ag/AgCl NWs on backbones of a sponge and thus reduce the material consumption. We anticipate that the 3D structuring of Ag/AgCl NWs, the photoreactor with confined water flow design and the in situ recovery strategy will promote the practical application of Ag/AgCl nanostructures for high speed, low cost and solar-powered water purification.

Scheme 1 Schematic illustration of a photoreactor based on the welded Ag/AgCl hybrid nanowire (Ag/AgCl NW) coated sponge. Welded Ag/AgCl NW are tightly fixed on the backbone of a polymer sponge. The shell of the photoreactor wraps the Ag/AgCl NW coated sponge (PS@Ag/AgCl NW) with no gap left. During the usage of PS@Ag/AgCl NW, its color will change from purple to grey due to the photoreduction of Ag⁺ to Ag⁰. And its photocatalytic activity will gradually decrease. But through an in situ chemical recovery treatment to reduce Ag⁰ back to Ag⁺, the degraded performance of PS@Ag/AgCl NW after prolonged usage can be recovered.

The fabrication procedures of 3D Ag/AgCl nanowire networks with high mechanical stability are illustrated in Fig. 1a. Briefly, a polymer sponge (PS, ESI,† Fig. S1) was firstly coated with silver nanowires (AgNWs) through dipping and drying processes as previously reported methods.^23–25 Then the AgNW coated polymer sponge was heated at 180 °C for 30 min to obtain 3D AgNW networks with welded joints (denoted as PS@AgNW). By dipping the PS@AgNW into FeCl₃ solution, AgNW on the sponge was oxidized to AgCl nanowires. Finally, polymer sponge supported welded 3D Ag/AgCl NW networks (denoted as PS@Ag/AgCl NW, Fig. 1b) were obtained by a photo irradiation treatment to reduce partial AgCl to metallic Ag on the surface of AgCl nanowires. X-ray powder diffraction (XRD, ESI,† Fig. S2) of the PS@Ag/AgCl NW product clearly shows that the cubic phase of AgCl (JCPDF file: 31-1238) coexists with the cubic phase of Ag (JCPDF file: 65-2871). The comparison of microstructures of PS, PS@AgNW, and PS@Ag/AgCl NW (Fig. 1c, d, and e respectively) indicates that 3D Ag/AgCl NW networks were formed on the microfiber surface of the sponge. The magnified SEM (Fig. 1f) shows that the AgNW on the backbone of the sponge are welded together at the contact point after thermal treatment. The magnified SEM (Fig. 1g) also shows that the weld Ag/AgCl NWs were intertwined together rather than the weak physical contact between the nanowires by direct dip-coating of Ag/AgCl NWs on the polymer sponge (ESI,† Fig. S3).

Fig. 1 (a) Schematic illustration of the fabrication processes of polymer sponge supported 3D Ag/AgCl NW (PS@Ag/AgCl NW) networks. (b) A photo of a typical sample of PS@Ag/AgCl NW. (c–e) SEM images of the original polymer sponge, PS@AgNW and PS@Ag/AgCl NW, respectively. (f and g) High magnified SEM image of the welded AgNW coating on the polymer sponge and the welded Ag/AgCl NW coating on the polymer sponge, respectively. (h) The weight retention of 3D Ag/AgCl NW networks with welded joints and that with non-welded joints (direct coating method, ESI,† Fig. S3) after 12 h of water rush test at different water flow rates (volume of water per square sponge piece per hour).

The anti-peeling of Ag/AgCl NWs on the polymer sponge under water rush was examined by a homemade apparatus (ESI,† Fig. S4). The results (Fig. 1h) show that the weight of welded Ag/AgCl NWs on the sponge substrate experienced little change after 12 hours of water flow wash even at a high speed of 180 m³ m⁻² h⁻¹, while non-welded Ag/AgCl NWs on the sponge substrate only retained 63% of the original weight under the same conditions, indicating that the weld treatment played an important role in the formation of mechanically stable 3D Ag/AgCl NW networks.

Compared with laminar flow in the conventional photoreactors fixed with 2D thin films,^{18,19,26–28} the water flow through the whole irregular 3D Ag/AgCl NW networks would greatly enhance the contact between reactants and the surfaces of Ag/AgCl NWs. To this end, we made up a prototype photoreactor to confine the water flowing through the inner space of the 3D Ag/AgCl NW networks. As is illustrated in Fig. 2a, PS@Ag/AgCl NW was embedded into the middle cavity of the patterned silica rubber mat, and then sandwiched between two polymethyl methacrylate (PMMA) sheets (for detailed methods see the ESI,† Fig. S5). The advantage of this design is that the PS@Ag/AgCl NW closely comes into contact the silicon rubber wall and organic glasses, leaving no gap except the macropores of PS@Ag/AgCl NWs for water to flow through (ESI,† Fig. S5c). Therefore, the micro sized pores (∼300 μm) serve as reaction chambers, which will efficiently shorten the diffusion distance of organic pollutants from water to the surfaces of Ag/AgCl NWs. In addition, the interconnected tortuous macropores of the 3D Ag/AgCl NW networks act as static mixers and cause enhanced fluid disturbance, which greatly increase the probability of contact between organic pollutants and catalyst surfaces.

Fig. 2 (a) Fabrication of a 3D Ag/AgCl NW nanowire networks based photoreactor with confined water flow. (b) Optical image of a photoreactor in which a typical PS@Ag/AgCl NW with 3 mm of thickness was fixed. (c) Schematic of the MO degradation test on three different photoreactors: (i) Ag/AgCl NW films on a filter paper substrate (2 cm × 5 cm), the gap between top glass and the film is 2 mm; (ii) the gap between top glass and the PS@Ag/AgCl NW piece (2 cm × 5 cm × 3 mm) is 2 mm; (iii) the same as (ii) expect that no gap is left. (d) MO degradation by the three different photoreactors versus time under the same light irradiation (300 W Xe arc lamp, 420 nm–700 nm, ca. 100 mW cm⁻²), water flow rate (87.3 mL min⁻¹) and catalyst loading density (1.26 mg cm⁻²). (e) Optical image of a polite modal photoreactor with one square meter size. (f and g) Optical images of original MO solution (8 mg L⁻¹) and the degraded MO solution using the pilot-scale photoreactor under sunlight (42 mW cm⁻²) at a flow rate of 1 L min⁻¹. (h) UV-Vis spectra of the original MO solution (black curve), 2 L of the MO solution after flowing through the pilot-scale photoreactor in the dark at a flow rate of 1 L min⁻¹ (red curve), and the degraded MO solution mentioned in Fig. 3g (green curve).

The advantages of a 3D Ag/AgCl NW network structure and confined flow design were demonstrated by photo bleaching methyl orange (MO) in a continuous-recirculation mode. As shown in Fig. 2c, a self-priming pump was used to feed the MO solution (30 mL, 8 mg mL⁻¹) to the photoreactor which was under visible light irradiation (300 W Xe arc lamp equipped with an UV-cutoff filter (≥420 nm), ca. 100 mW cm⁻²) and circulate the outlet solution back to the MO reservoir. Meanwhile, the performance of 2D Ag/AgCl NW films (Fig. 2c(i), ESI,† Fig. S6) and the photoreactor based on the same 3D Ag/AgCl NW networks with a laminar gap (Fig. 2c(ii)) was also tested under the same light irradiation, water flow rate and mass loading of catalysts, respectively. Fig. 2d clearly shows that the degradation efficiency of the PS@Ag/AgCl NW based photoreactor with a laminar gap (Fig. 2c(ii)) is more than 2 times higher than that of Ag/AgCl NW films, demonstrating the advantage of 3D structured Ag/AgCl NWs. However, the degradation efficiency of PS@Ag/AgCl NW with a laminar gap (Fig. 2c(ii)) is only half of that with confined flow design (Fig. 2c(iii)). This supports our design principle that confined flow induces close contact of MO molecules and Ag/AgCl NWs, leading to a higher photocatalytic efficiency. The mass transfer efficiency of 3D Ag/AgCl NW networks with confined flow can be further improved by increasing the flow rate which has great influence on the intensity of disturbance (ESI,† Fig. S7a). It should be noted that, at the flow rate of 160.9 mL min⁻¹, the throughput of the photoreactor could reach up to 9600 L h⁻¹ m⁻². The utilization of solar energy per unit area for the 3D Ag/AgCl NW networks can be maximized by increasing the mass loading of Ag/AgCl NWs on per volume of the polymer sponge and the thickness of PS@Ag/AgCl NWs (ESI,† Fig. S7b and c).

To demonstrate the easily scalable fabrication and high performance of the 3D Ag/AgCl NW networks with confined flow under sunlight for practical application, a pilot-scale photoreactor with one square meter size was fabricated (Fig. 2e). The performance of this photoreactor was evaluated by bleaching MO in the real sunlight irradiation. Fig. 2f and g show that 2000 mL of yellow MO solution (8 mg L⁻¹) was completely bleached by the photoreactor in 2 min (ESI,† Movie S1). It should be noted that the absorption of MO molecules on the PS@Ag/AgCl NW monolith contributes partially to the concentration decrease of the MO solution. From the absorption spectra of the MO solution (2 L) that flowed through the photoreactor in the dark at a flow rate of 1 L min⁻¹ (Fig. 2h), we can see that 14.6% of the MO molecules was absorbed by the PS@Ag/AgCl NW monolith.

In addition to the high efficiency and high throughput of the photoreactor based on 3D Ag/AgCl NW networks with confined flow under visible light, the service time of the Ag/AgCl NW catalysts in the water treatment is also a key parameter. To evaluate the stability of 3D Ag/AgCl NW networks, a long-time bleaching experiment for MO was performed. As illustrated in Fig. 3a, the MO solution was injected into the photoreactor using a syringe pump and degraded under white Light Emitting Diode (LED) lamp light irradiation (25 mW cm⁻²). The residential time of MO in the photoreactor chamber was set to 30 seconds. The MO concentration of the outlet solution was measured using a UV-Vis spectrophotometer. Fig. 3b shows that the degradation efficiency of the PS@Ag/AgCl NW to MO gradually decreased to 42% of its beginning value after 300 min of continuous operation. The XRD patterns of original and used PS@Ag/AgCl NWs show that the relative intensity of AgCl compared to Ag decreased after the long-time operation (Fig. 3c), which indicates that the performance deterioration of PS@Ag/AgCl NWs was caused by the gradual reduction of AgCl to elemental Ag during the photocatalytic process. This was further confirmed by X-ray photoelectron spectroscopy (XPS) of the original and used PS@Ag/AgCl NWs. From the XPS spectrum (ESI,† Fig. S8), the calculated surface mole ratio of the metallic Ag⁰ to Ag⁺ of original PS@Ag/AgCl NWs was 1 : 8.4. After 2.5 h of usage, the ratio of Ag⁰ to Ag⁺ increased to 1 : 4.1.

Fig. 3 (a) Schematic illustration of the stability evaluation system for the PS@Ag/AgCl NW monolith. Commercial Light Emitting Diode (LED) light was chosen as the light source, for the LED light can continue working for more than 100 hours. The thickness of the PS@Ag/AgCl NW monolith is 1 mm, and the Ag/AgCl NW mass loading is 2.2 mg cm⁻³. (b) Change in MO degradation versus time for the original PS@Ag/AgCl NW monolith. (c) XRD patterns of the original, used and in situ recovered PS@Ag/AgCl NW monolith. (d) Degradation – in situ chemical recovery cycle test for the PS@Ag/AgCl NW monolith. Each degradation cycle lasted for 2.5 h. (e) Schematic illustration of the in situ chemical recovery strategy used to extend the lifetime of PS@Ag/AgCl NW.

After understanding the inactivation mechanism, FeCl₃ solution was continuously pumped through the photoreactor for 15 min to in situ oxide Ag back to AgCl. XRD (Fig. 3c) and XPS (ESI,† Fig. S8) show that AgCl was regenerated in the used PS@Ag/AgCl NW. The data in Fig. 3d demonstrate that the performance of the used PS@Ag/AgCl NW was recovered by the in situ chemical recovery processes. After repeating the degradation and recovery processes nine times, the curve of degradation efficiency versus usage time almost remained the same as that of the original PS@Ag/AgCl NW. Therefore, even though the degradation efficiency of the PS@Ag/AgCl NW over MO reduced gradually during the photocatalytic processes due to the partial conversion of Ag⁺ to Ag⁰, the efficiency of PS@Ag/AgCl NW could be repeatedly recovered by reducing Ag⁰ back to Ag⁺ by in situ injecting FeCl₃ solution through the 3D Ag/AgCl NW networks for 15 min (Fig. 3e). And as a result, the PS@Ag/AgCl NWs can be used for a long time while retaining high photodegradation efficiency.

Conclusions

In summary, we developed a welding method to fabricate robust 3D Ag/AgCl nanowire (Ag/AgCl NW) networks which exhibit higher photodegradation efficiency than 2D Ag/AgCl NW films, and proposed a new photoreactor with confined flow to further enhance the mass transfer efficiency of the 3D Ag/AgCl nanowire networks. This photoreactor with high throughput can be easily scaled up to one square meter size and exhibits high performance under sunlight. Furthermore, we demonstrated that the lifetime of the 3D Ag/AgCl nanowire networks can be greatly prolonged by a facile in situ chemical recovery method. The 3D structuring of Ag/AgCl nanowire networks, combined with the confined flow design and in situ chemical recovery strategy, will pave the way for the practical application of Ag/AgCl nanostructures in solar-powered water purification with high efficiency and throughput in the future.

Acknowledgements

S.H.Y. acknowledges the funding support from the National Basic Research Program of China (Grants 2014CB931800, 2013CB933900), the National Natural Science Foundation of China (Grants 21431006, 91227103, 21061160492, J1030412), and the Chinese Academy of Sciences (Grant KJZD-EW-M01-1).

Notes and references

1. M. N. Chong, B. Jin, C. W. K. Chow and C. Saint, Water Res., 2010, 44, 2997.
2. C. Chen, W. Ma and J. Zhao, Chem. Soc. Rev., 2010, 39, 4206.
3. A. Kubacka, M. Fernandez-Garcia and G. Colon, Chem. Rev., 2011, 112, 1555.
4. P. Wang, B. Huang, X. Qin, X. Zhang, Y. Dai, J. Wei and M. H. Whangbo, Angew. Chem., Int. Ed., 2008, 47, 7931.
5. Z. Z. Lou, Z. Y. Wang, B. B. Huang and Y. Dai, ChemCatChem, 2014, 6, 2456.
6. P. Wang, B. B. Huang, Y. Dai and M. H. Whangbo, Phys. Chem. Chem. Phys., 2012, 14, 9813.
7. C. H. An, S. N. Peng and Y. G. Sun, Adv. Mater., 2010, 22, 2570.
8. M. Choi, K. H. Shin and J. Jang, J. Colloid Interface Sci., 2010, 341, 83.
9. Y. Y. Li and Y. Ding, J. Phys. Chem. C, 2010, 114, 3175.
10. Y. G. Sun, J. Phys. Chem. C, 2010, 114, 2127.
11. L. Han, P. Wang, C. Z. Zhu, Y. M. Zhai and S. J. Dong, Nanoscale, 2011, 3, 2931.
12. J. Jiang and L. Z. Zhang, Chem. – Eur. J., 2011, 17, 3710.
13. H. Xu, H. M. Li, J. X. Xia, S. Yin, Z. J. Luo, L. Liu and L. Xu, ACS Appl. Mater. Interfaces, 2011, 3, 22.
14. H. Zhang, X. F. Fan, X. Quan, S. Chen and H. T. Yu, Environ. Sci. Technol., 2011, 45, 5731.
15. M. S. Zhu, P. L. Chen and M. H. Liu, ACS Nano, 2011, 5, 4529.
16. J. H. Kou and R. S. Varma, ChemSusChem, 2012, 5, 2435.
17. H. X. Shi, J. Y. Chen, G. Y. Li, X. Nie, H. J. Zhao, P. K. Wong and T. C. An, ACS Appl. Mater. Interfaces, 2013, 5, 6959.
18. J. Senthilnathan and L. Philip, Sol. Energy, 2012, 86, 2735.
19. G. Zayani, L. Bousselmi, F. Mhenni and A. Ghrabi, Desalination, 2009, 246, 344.
20. N. M. Mahmoodi, M. Arami and N. Y. Limaee, J. Hazard. Mater., 2006, 133, 113.
21. L. Han, Z. K. Xu, P. Wang and S. J. Dong, Chem. Commun., 2013, 49, 4953.
22. L. Lei, N. Wang, X. M. Zhang, Q. D. Tai, D. P. Tsai and H. L. W. Chan, Biomicrofluidics, 2010, 4, 043004.
23. J. Ge, Y.-D. Ye, H.-B. Yao, X. Zhu, X. Wang, L. Wu, J.-L. Wang, H. Ding, N. Yong, L.-H. He and S.-H. Yu, Angew. Chem., Int. Ed., 2014, 53, 3612.
24. J. Ge, H.-B. Yao, X. Wang, Y.-D. Ye, J.-L. Wang, Z.-Y. Wu, J.-W. Liu, F.-J. Fan, H.-L. Gao, C.-L. Zhang and S.-H. Yu, Angew. Chem., Int. Ed., 2013, 52, 1654.
25. H.-B. Yao, J. Ge, C.-F. Wang, X. Wang, W. Hu, Z.-J. Zheng, Y. Ni and S.-H. Yu, Adv. Mater., 2013, 25, 6692.
26. Y.-L. Kuo, T.-L. Su, K.-J. Chuang, H.-W. Chen and F.-C. Kung, Environ. Technol., 2011, 32, 1799.
27. X. Wang, X. Tan and T. Yu, Ind. Eng. Chem. Res., 2014, 53, 7902.
28. P.-J. Lu, C.-W. Chien, T.-S. Chen and J.-M. Chern, Chem. Eng. J., 2010, 163, 28.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5mh00069f

‡ These authors contributed equally to this work.

---