A general salt-resistant hydrophilic/hydrophobic nanoporous double layer design for efficient and stable solar water evaporation distillation

Abstract

A novel and elegant hydrophilic/hydrophobic nanoporous double layer structure was designed and developed for efficient long-term water desalination. It contained a hydrophobic salt-resistant hierarchical layer of well-defined Cu₂SnSe₃ (or Cu₂ZnSnSe₄) nanosphere arrays for broad solar harvesting and water vapor evaporation, and a hydrophilic filter membrane for continuous water supply and vapor generation. The as-fabricated self-floatable devices achieve remarkable solar water evaporation performances (average evaporation rate: 1.657 kg m⁻² h⁻¹ and solar thermal conversion efficiency: 86.6% under one sun) with super stability for water distillation from seawater and wastewater containing organic dyes, heavy metals and bacteria.

---

Conceptual insights

Water pollution and shortage have become some of the most serious issues in modern society. Solar energy driven clean water production from seawater and wastewater has been regarded as an economic and sustainable solution. Recently, examples of all-hydrophilic structures based on carbon and plasmonic photothermal conversion materials have been fabricated and are widely accepted to be favourable for efficient solar pure water evaporation. However, in the actual solar desalination process through vapor evaporation, salt accumulation on the vapor generator, resulting from hydrophilic features, blocks the device. In this work, we have developed a general salt-resistant hydrophilic/hydrophobic nanoporous double layer structure, containing superior light absorption materials Cu₂SnSe₃ and Cu₂ZnSnSe₄, which can achieve both highly efficient solar evaporation performance and long-term continuously stable distillation. Such rational design offers new insight and a significant step forward in the development of advanced materials and structures for practical clean energy utilization, water purification and environment remediation.

With the tremendous development of modern industry, ever-increasing energy and resource consumption has caused serious water pollution and shortage problems. Currently, one-third of the world's population is living with the problem of clean water shortages.¹ Seawater is a seemingly unlimited resource for clean water since the oceans cover 70.8% of the Earth’s surface area and make up 97% of the total water reservoir in the world.² Seawater desalination has been acknowledged as the most applicable route for providing a supply of large amounts of fresh water for drinking as well as irrigation. The state-of-the-art seawater desalination technologies can be generally divided into two categories. One is reverse osmosis (RO), and the other is thermal desalination.³ RO technology is the most energy-economic desalination technology at present.⁴ However, the requirement of a high-pressure supply and RO membrane fouling make this technology inapplicable in direct seawater treatment.⁵ Thermal desalination was the first technology for drinking water treatment, where the thermal energy comes from the residual steam of a power plant, and the desalination installation as well as the cost are heavily reliant on the heat source distribution.³

Clean water production by employing solar energy has been considered as an economic and sustainable solution.⁶ The costs of installation and maintenance of solar desalination facilities are significantly lower than those of RO desalination and traditional thermal evaporation, and are thus suitable for medium-sized clean water supplies.⁷ Solar desalination can be realized by a thermal collector with a circulation water system and a floating-type solar evaporation unit.⁸ The floating-type unit is currently a research hotspot for seawater desalination and wastewater remediation.^9–15 The keys to this type of solar evaporation unit are the designs of light-to-thermal conversion materials and water–vapor transport structures. Recently, several examples of carbon and wood-based and plasmonic photothermal conversion materials have been demonstrated for highly efficient solar evaporation performances.^16–48 These research results draw a clear picture of how to achieve highly efficiency desalination. One strategy involves using highly efficient solar absorption materials such as carbon materials, sulfides, and nano-sized gold particles.^15,27 The other is to ensure enough thermal insulation between the evaporation material and the bulk water while keeping a water supply by the capillary effect.^8,32

However, for practical applications, the desalination of seawater and wastewater with undesirably high salinity is still a big challenge in terms of making devices that are capable of long-term steady work. During water evaporation, salt continuously accumulates on the surface of the light-absorbing materials.^{8,15,19,27,43} The aggregated crystals reduce the capillary effect for water pumping, decrease the vapor pressure, and hinder the subsequent water evaporation, and are highly likely to destroy the device, especially under a high evaporation rate.^8,38 This unsolved problem is retarding the development of efficient devices with long-term performances suitable for practical applications. Until now, this problem has been rarely emphasized and an effective solution to this issue is still lacking. Very recently, plasmonic wood with microchannels,⁴³ washable graphene foam²⁷ and carbon black nanoparticles coated polymethylmethacrylate/polyacrylonitrile Janus absorber by electrospinning⁴⁸ have exhibited the potential for stable solar desalination under intermittent working conditions.

Hydrophobic surfaces with natural non-wetting features offer an effective approach for preventing salt water infiltration. However, such all-hydrophobic structures have obviously lower evaporation rates compared to hydrophilic ones due to most of the water being blocked.^13,21,30 We proposed a novel hydrophilic/hydrophobic nanoporous double layer (HHNDL) structure, which combines the advantages of both a hydrophilic material that can guarantee water storage and supply for efficient vapor generation and a porous hydrophobic layer that can resist salt deposition and provide vapor evaporation pathways, enabling long-term stability of the device under continuous working conditions. As a proof of concept, we selected well-defined superior light absorption materials Cu₂SnSe₃ (CTSe) and Cu₂ZnSnSe₄ (CZTSe) with oleylamine capping as the porous hydrophobic light-harvesting material/membrane, and a piece of commercial filter membrane as the hydrophilic substrate. A stacked hierarchical CTSe (or CZTSe) nanosphere superstructure was well formed during filtration, which benefits the vapor release from the membrane.

Fig. 1a shows the concept of HHNDL membranes for effective vapor generation and evaporation with salt resistance, where well-defined hydrophobic CTSe and CZTSe hierarchical nanosphere arrays were filtrated on the hydrophilic filter membrane as the light absorber, and a piece of polyurethane foam was used as a thermal insulator and a self-floater with a bundle of infiltrative nonwoven fabrics as a water pipe (Fig. S1, ESI†). The excellent solar water generation, evaporation and ultra-stability in water purification under high salinity of this design will be presented and demonstrated subsequently.

Fig. 1 (a) Schematic illustration of the hydrophilic/hydrophobic nanoporous double layer solar water purification device. (b) Optical photograph the of CTSe membrane on a filter membrane. (c) Top-view SEM image of the CTSe membrane. (d) Cross-section SEM image of the hydrophilic/hydrophobic nanoporous double layer.

Solar-to-heat conversion materials, CTSe and CZTSe, were synthesized by a one-pot colloidal synthesis using oleylamine as a capping agent. The as-synthesized hierarchical crystalline nanospheres had diameters of 300–500 nm (Fig. S2a–d, ESI†), which was the perfect size to be a little larger than the pore size of the filter membrane (Fig. S3, ESI†). X-ray diffraction (XRD) patterns and Raman spectra (Fig. S2e and f, ESI†) were used to identify the monoclinic structure of CTSe (JCPDS No. 72-8034) and the tetragonal kesterite structure of CZTSe (JCPDS No. 70-8930). The Cu⁺, Zn²⁺, Sn⁴⁺ and Se²⁻ chemical states with certain peak separations were in good agreement with the expected X-ray photoelectron spectroscopy (XPS) results (Fig. S4, ESI†). The hierarchical CTSe and CZTSe nanospheres can be filtrated finely on a cellulose filter membrane and formed a HHNDL membrane (Fig. 1b). It can be observed from the top-view SEM image that the CTSe nanospheres were uniformly distributed on the cellulose membrane (Fig. 1c). The unique wing structures can prevent the spheres from packing too closely, and thus naturally form the vapor pathways for evaporation. The cross-section view, colored back-scattering SEM image clearly shows the hydrophilic/hydrophobic nanoporous structure. The blue area indicates the hydrophobic CTSe layer, while the yellow area is the hydrophilic cellulose filter membrane (Fig. 1d). This unique HHNDL design plays a key role in salt-resistance, water purification, and bacteria removal, which will be discussed in detail.

The high evaporation performance of both materials is associated with both their physically porous HHNDL structures, and their unique optical characteristics, i.e. their broad efficient solar absorption (CTSe > 97% and CZTSe > 95.5%) for water heating from CTSe and CZTSe membranes which covers the entire UV/vis/NIR region (Fig. 2a) due to their favorable interior electronic band structures, density of states (DOS) and high absorption coefficients (Fig. S5, ESI†). The evaporation performance of the HHNDL membranes with optimized loading masses of 0.8 mg cm⁻² CTSe and CZTSe deliver the highest average solar evaporation rates (Bohai Sea) of 1.657 and 1.643 kg m⁻² h⁻¹, respectively (Fig. 2b and Fig. S6a, ESI†), which are 3.93 and 3.89 times higher than that of pristine seawater (0.422 kg m⁻² h⁻¹). The evaporation rate of the device in a dark environment shows a relatively low value of 0.276 kg m⁻² h⁻¹, which reflects the solar energy dominated nature of thermal water evaporation. After calculation, the achieved solar thermal conversion efficiencies applicable to steam are up to 86.6% and 85.7% under only one sun, and are superior to those of almost all previous reports (Fig. S6b and Table S1, ESI†). As shown in Fig. 2c, the cycle performance of the device is exhibited for each cycle, and is sustained for over 1 h. It is clear that steady evaporation performance is demonstrated for at least 20 cycles. Upon light irradiation, the temperature of the membrane increases significantly and rapidly, and reaches the highest temperature within 3–5 min (Fig. S7 and S8, ESI†). A stable vapor temperature of ∼43 °C on the membrane surface is achieved in 4 min. In stark contrast, the upper bulk water has only a 2 °C increase even after 1 h irradiation (Fig. 2d). The temperature of the membrane also quickly rises to ∼37 °C in 3 min and reaches beyond 40 °C after 1 h of heating (Fig. 2e). These findings are attributed to the well-insulated localized heating of our delicately designed device structures. With the non-wetting surface, heat is generated and localized on the light absorption materials, and then efficiently heats the HHNDL interface where water is stored, and vapor is generated, leading to the high evaporation rate of heated steam through the porous membrane (Fig. 2f). The heat transfer processes involved in the solar evaporator, including radiative (∼4.0%) and convective (∼3.0%) heat loss to the surrounding and conductive (∼2.3%) heat loss to the underlying bulk water (Fig. 2g, detailed calculations shown in the ESI†) confirm that the heat is highly localized on the absorber. The efficient solar utilization, low thermal loss, abundant evaporation channels and rapid water transport endow the design with excellent photothermal transduction, a timely water supply, and solute resistance for efficient and steady solar vapor generation.

Fig. 2 (a) UV-vis-NIR absorption spectra of the HHNDL membranes. (b) The mass change of water through vapor evaporation over time. (c) Evaporation cycle performance of the devices. (d) Temperature variation of the vapor and bulk water over time. (e) IR images of the CTSe membrane over time during the water evaporation process. (f) Optical photograph of the device operating at an illumination intensity of 4 suns. (g) Energy balance and heat transfer diagram of the device during solar evaporation.

It is generally reported that hydrophilic structures are beneficial to evaporation efficiency. The continuous water supply to the hydrophilic nano/micropores of the materials by capillary effect can be easily heated. However, only a few research reports have mentioned long-term desalination using high salinity water (Table S1, ESI†). In actual desalination through solar evaporation, the solutes such as NaCl, KCl, Na₂SO₄ can continuously crystalize on the hydrophilic membrane surfaces.^8,19,27,43 In order to demonstrate the significant differences between hydrophilic and hydrophobic surfaces, we used the system with a CTSe membrane as a solar evaporator. Fig. 3a–c are the hydrophilic membrane (Fig. S9, ESI†) before and after 10 hours of desalination. A white layer of salt crystals can be clearly observed on the membrane surface (Fig. 3b). The magnified SEM image shows that the diameters of the salt crystals range from over nanometers to micrometers (Fig. 3c). In contrast, the hydrophobic CTSe membrane keeps a clean and dry surface after 10 hours of desalination (Fig. 3e). The corresponding SEM image further demonstrates a salt-free surface, which retained a good porous structure (Fig. 3f). A schematic illustration of the evaporation process for the two configurations is plotted in Fig. 3g. In the all-hydrophilic configuration, salt keeps crystalizing as the water evaporates. The obvious increase of the salt concentration and salt crystals on the membrane significantly decrease the vapor pressure (detailed analysis is shown in the ESI†), block the vapor pathways as well as light absorption, and then seriously hinder the evaporation process and destroy the membrane. In the HHNDL structure, the hydrophobic CTSe layer converts solar light to heat. Salt cannot crystalize on the membrane surface because water, along with salt, is kept underneath the CTSe layer by the hydrophobic effect, and the hydrophilic filter membrane offers an adequate and steady water supply. Therefore, the salt concentration on the membrane and the vapor pressure cannot increase obviously. Therefore, the heating efficiency is not affected and hence results in a high evaporation rate. Eventually, the water vapor is transported through the channels formed by the interconnected nanospheres. Fig. 3h shows the long-term performance of the hydrophilic and hydrophobic membrane for 10 hours. The evaporation rate of the hydrophilic membrane significantly decreases during operation, while the evaporation rate of the hydrophobic membrane keeps a steady performance throughout the entire process. We also evaluated the differences between the hydrophilic and hydrophobic membranes by fitting the curves using exponential decay:

where A, t₀, and t₁ are constants, m stands for the evaporation mass per square meter, t is the evaporation time, and m₀ represents the evaporation quantity when t = ∞. The m₀ values for the hydrophilic and hydrophobic membranes are calculated from eqn (1) to be −23.52 and −584.23 kg m⁻², respectively. This indicates that the hydrophilic membrane is quickly disabled under continuous working conditions. However, the water production of the hydrophobic membrane is ultrastable under the same solar illumination conditions. Theoretically, the mass change of the HHNDL membrane over time was not ideally linear in this test, and the evaporation rate shows a slow exponential decay over time (Fig. 3h and eqn (1)) due to the slight decrease of vapor pressure with a little increase in the salt concentration as the water volume reduces. Although the capillary effect may also be slightly reduced with increasing salt concentration, the solar evaporation rate still remains unchanged, which could be ascribed to the ultrastrong capillary effect, namely the required water pumping by the nanostructure of the hydrophilic layer and nonwoven fabrics is much faster than the solar evaporation. As a result, the solar evaporation rate seems to be nearly constant within every 12 h period (Fig. 3i). Moreover, the evaporation rate of the HHNDL structure was found to be stable over 15 days (solar/dark conditions for 12/12 h per day, with refreshed seawater every 12 h) of continuous desalination, and worked without decay (Fig. 3i). After continuous running for 15 days, salt-free hydrophobic and hydrophilic layers, a nonwoven fabric surface and a hydrophilic/hydrophobic interface can be seen (Fig. S10, ESI†), which is attributed to the combination of the salt-blocking of the hydrophobic layer and adequate water pumping of the hydrophilic layer. Besides, despite the expectations that the chemical potential and the vapor pressure would decrease with an increasing concentration of the solution (detailed analysis shown in ESI†) due to water loss during the evaporation process, the evaporation rate remained unaffected within 1 h when the water contained salt, organic dye and heavy metal solutes, respectively (Fig. S11, ESI†), indicating a temperature rather than a concentration dominated evaporation process.

Fig. 3 Hydrophilic and hydrophobic membranes (a and d) before and (b and e) after 10 h solar desalination. SEM images of the (c) hydrophilic and (f) hydrophobic membranes after 10 h solar desalination. (g) Schematic illustration of the two kinds of membrane during the solar desalination process. (h) Solar desalination rate of the hydrophilic and hydrophobic membranes and their simulated decay curves. (i) Long-term desalination stability of the HHNDL membrane.

Constructing abundant vapor pathways is also a necessary step for a high rate of water vapor generation. In order to demonstrate its importance, we synthesized CTSe nanoplates as a control. The surface of the nanoplate membrane showed a metallic luster due to its high light reflection (Fig. 4a), while the surface of nanosphere membrane was free of luster (Fig. 4b). The evaporation rate was obviously reduced when nanoplates were used due to the low vapor flux of the close-stacked structure of the nanoplate assembled membrane (Fig. 4c–e). We also roughly estimated the amount of transport channels by using cyclohexane as an analogue for water vapor. It can be clearly observed that during vacuum filtration, cyclohexane passed through the nanoplate membrane much slower than through the nanosphere membrane (Fig. 4f and g).

Fig. 4 Optical photographs and schematic illustrations of light reflection and vapor flux of CTSe (a) nanoplate and (b) nanosphere membranes. (c) Top-view and (d) cross-section SEM images of the CTSe nanoplate/filter membrane double layer membrane. (e) Mass change of water through nanosphere (red line) and nanoplate (blue line) membranes. Time course filtration rates of (f) nanoplate and (g) nanosphere membranes.

We evaluated the solar desalination of the device using natural water, and three water samples with different salinities (Bohai Sea, Qinghai Lake and Caka Salt Lake) were employed and carefully tracked, as shown in Fig. 5a–c. The concentrations of the four primary ions, Na⁺, K⁺, Mg²⁺ and Ca²⁺ were significantly decreased with ion rejection beyond 99.5%, and were below the salinity levels defined by the World Health Organization (WHO)⁴⁹ and the US Environmental Protection Agency (EPA) standards,⁵⁰ and below the values typically obtained through other membrane-based (10–500 ppm) and distillation-based (1–50 ppm) seawater desalination technologies.⁵¹ Besides, elemental analysis was conducted to assess the risk of metal ions leaching from the material. There was no presence of Cu or Se in either the vapor or the bulk seawater (Table S2, ESI†), thus excluding the possibility of the metals leaching from the membrane. The HHNDL membrane can not only generate clean water from seawater, but also works well in water remediation. The device achieves rejections close to 100% for Rhodamine B (RhB), methyl orange (MO) and methyl blue (MB) dyes, and Fe³⁺, Cu²⁺ and Cr⁶⁺ heavy metal ions in simulated wastewater (Fig. 5d–f and Fig. S12, ESI†). The most two common bacteria in water, Escherichia coli and Staphylococcus aureus, can also be completely rejected through the device, even with ultrahigh numbers of 10⁶ (Fig. 5g–i). Solar vapor generation through the HHNDL membranes not only showed the efficient and steady solar water evaporation performance without solute, but also demonstrated an effective solute blocking function without loss of evaporation rate.

Fig. 5 Measured salinities of three actual seawater samples from: (a) Bohai Sea, (b) Qinghai Lake, (c) Caka Salt Lake before (red column) and after (blue column) solar desalination. (d) Purification of organic dyes. (e) Purification of heavy metal ions. (f) Organic dye and heavy metal ion rejection performance. (g) Optical photographs of E. coli removal. (h) Optical photographs of S. aureus removal. (i) Anti-bacterial performance. (j) Schematic illustration of water purification under natural sun illumination. (k) Optical photograph of the all-in-one solar distillation setup for freshwater evaporation and collection during the daytime. (l) Freshwater collection performance of the all-in-one solar distillation setup under natural sunlight.

For the practical use of the HHNDL membrane, an all-in-one solar distillation set-up was designed and fabricated. Fig. 5j displays a schematic illustration of the set-up and freshwater collection during a continuous daytime. Under natural sunlight, vapor was generated from the membrane, freshwater then condensed on the chamber, and finally collected on the bottom (Fig. 5k). After 4 h of evaporation, ∼15 mL of freshwater can be collected, which is 45.2% of the theoretical value. Moreover, ∼35 mL of freshwater can be obtained after 8 h of work, which reaches 53.7% of the theoretical evaporation. The practical test conditions and collection performances are summarized in Fig. 5l. The self-floating solar evaporation device containing the HHNDL membrane was quite stable in structure and water purification performance under simulated solar light or natural sunlight, indicating the practical applications of the device for point-of-use freshwater production.

Conclusions

Novel HHNDL membranes composed of well-defined hydrophobic hierarchical nanosphere arrays as a light absorption and vapor evaporation layer and a piece of hydrophilic filter membrane as a water supply and vapor generation substrate have been fabricated. The HHNDL membrane based self-floating solar evaporation device has been developed with low-cost commercial materials and facile up-scalable fabrication. Unique characteristics, such as broad efficient solar utilization, abundant evaporation channels, rapid water transport in a hydrophilic layer and salt resistance on hydrophobic surfaces, enable this solar evaporation device to exhibit remarkable solar evaporation performance and extraordinary stability under rigorous conditions. This ensures highly efficient and enduring solar water purification of seawater and wastewater containing organic dyes, heavy metals and bacteria under natural conditions. This design has achieved superb high solar thermal conversion efficiency of up to 86.6% under one sun. Besides CTSe and CZTSe, the HHNDL design is believed to be feasible for other solar vapor generation materials, and is expected to provide a facile, green and robust approach for the development of advanced designs of solar water evaporation devices for practical sustainable water purification.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (Grant No. 61774122), the Science and Technology Developing Project of Shaanxi Province under Grant No. 2015KW-001, and the 111 Project from China (B18013, B14040). Y. P. Du gratefully acknowledge the NSFC (Grant No. 21371140), the National Key R&D Program of China (2017YFA0208000), 111 Project (B18030) from China, and the China National Funds for Excellent Young Scientists (Grant No. 21522106). The SEM and TEM work were conducted at the International Center for Dielectric Research, Xi’an Jiaotong University, Xi’an, China. We thank Dr Liu at Instrument Analysis Center of Xi'an Jiaotong University for her assistance with XPS analysis.

Notes and references

1. H. Cheng, Y. Hu and J. Zhao, Environ. Sci. Technol., 2009, 43, 240–244.
2. E. P. Glenn, J. J. Brown and J. W. O'Leary, Sci. Am., 1998, 279, 76–81.
3. M. Elimelech and W. A. Phillip, Science, 2011, 333, 712–717.
4. C. Fritzmann, J. Löwenberg, T. Wintgens and T. Melin, Desalination, 2007, 216, 1–76.
5. M. Kumar, S. S. Adham and W. R. Pearce, Environ. Sci. Technol., 2006, 40, 2037–2044.
6. H. M. Qiblawey and F. Banat, Desalination, 2008, 220, 633–644.
7. H. Kunze, Desalination, 2001, 139, 35–41.
8. X. Li, W. Xu, M. Tang, L. Zhou, B. Zhu, S. Zhu and J. Zhu, Proc. Natl. Acad. Sci. U. S. A., 2016, 113, 13953–13958.
9. K. Bae, G. Kang, S. K. Cho, W. Park, K. Kim and W. J. Padilla, Nat. Commun., 2015, 6, 10103.
10. R. Li, L. Zhang, L. Shi and P. Wang, ACS Nano, 2017, 11, 3752–3759.
11. O. Neumann, A. S. Urban, J. Day, S. Lal, P. Nordlander and N. J. Halas, ACS Nano, 2013, 7, 42–49.
12. Y. Zeng, J. Yao, B. A. Horri, K. Wang, Y. Wu, D. Li and H. Wang, Energy Environ. Sci., 2011, 4, 4074–4078.
13. L. Zhang, B. Tang, J. Wu, R. Li and P. Wang, Adv. Mater., 2015, 27, 4889–4894.
14. L. Zhou, Y. Tan, D. Ji, B. Zhu, P. Zhang, J. Xu, Q. Gan, Z. Yu and J. Zhu, Sci. Adv., 2016, 2, e1501227.
15. L. Zhou, Y. Tan, J. Wang, W. Xu, Y. Yuan, W. Cai, S. Zhu and J. Zhu, Nat. Photonics, 2016, 10, 393–398.
16. C. Chen, Y. Li, J. Song, Z. Yang, Y. Kuang, E. Hitz, C. Jia, A. Gong, F. Jiang, J. Y. Zhu, B. Yang, J. Xie and L. Hu, Adv. Mater., 2017, 29, 1701756.
17. F. Chen, A. S. Gong, M. Zhu, G. Chen, S. D. Lacey, F. Jiang, Y. Li, Y. Wang, J. Dai, Y. Yao, J. Song, B. Liu, K. Fu, S. Das and L. Hu, ACS Nano, 2017, 11, 4275–4282.
18. X. Gao, H. Ren, J. Zhou, R. Du, C. Yin, R. Liu, H. Peng, L. Tong, Z. Liu and J. Zhang, Chem. Mater., 2017, 29, 5777–5781.
19. M. Zhu, Y. Li, G. Chen, F. Jiang, Z. Yang, X. Luo, Y. Wang, S. D. Lacey, J. Dai, C. Wang, C. Jia, J. Wan, Y. Yao, A. Gong, B. Yang, Z. Yu, S. Das and L. Hu, Adv. Mater., 2017, 29, 1704107.
20. X. Hu, W. Xu, L. Zhou, Y. Tan, Y. Wang, S. Zhu and J. Zhu, Adv. Mater., 2017, 29, 1604031.
21. Y. Ito, Y. Tanabe, J. Han, T. Fujita, K. Tanigaki and M. Chen, Adv. Mater., 2015, 27, 4302–4307.
22. H. Ghasemi, G. Ni, A. M. Marconnet, J. Loomis, S. Yerci, N. Miljkovic and G. Chen, Nat. Commun., 2014, 5, 4449.
23. Y. Li, T. Gao, Z. Yang, C. Chen, W. Luo, J. Song, E. Hitz, C. Jia, Y. Zhou, B. Liu, B. Yang and L. Hu, Adv. Mater., 2017, 29, 1700981.
24. Y. Liu, S. Yu, R. Feng, A. Bernard, Y. Liu, Y. Zhang, H. Duan, W. Shang, P. Tao, C. Song and T. Deng, Adv. Mater., 2015, 27, 2768–2774.
25. O. Neumann, C. Feronti, A. D. Neumann, A. Dong, K. Schell, B. Lu, E. Kim, M. Quinn, S. Thompson, N. Grady, P. Nordlander, M. Oden and N. J. Halas, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, 11677–11681.
26. G. Ni, G. Li, S. V. Boriskina, H. Li, W. Yang, T. Zhang and G. Chen, Nat. Energy, 2016, 1, 16126.
27. H. Ren, M. Tang, B. Guan, K. Wang, J. Yang, F. Wang, M. Wang, J. Shan, Z. Chen, D. Wei, H. Peng and Z. Liu, Adv. Mater., 2017, 29, 1702590.
28. G. Wang, Y. Fu, A. Guo, T. Mei, J. Wang, J. Li and X. Wang, Chem. Mater., 2017, 29, 5629–5635.
29. N. Xu, X. Hu, W. Xu, X. Li, L. Zhou, S. Zhu and J. Zhu, Adv. Mater., 2017, 29, 1606762.
30. P. Zhang, J. Li, L. Lv, Y. Zhao and L. Qu, ACS Nano, 2017, 11, 5087–5093.
31. M. S. Zielinski, J.-W. Choi, T. La Grange, M. Modestino, S. M. H. Hashemi, Y. Pu, S. Birkhold, J. A. Hubbell and D. Psaltis, Nano Lett., 2016, 16, 2159–2167.
32. G. Ni, N. Miljkovic, H. Ghasemi, X. Huang, S. V. Boriskina, C. T. Lin, J. Wang, Y. Xu, M. M. Rahman, T. Zhang and G. Chen, Nano Energy, 2015, 17, 290–301.
33. P. Yang, K. Liu, Q. Chen, J. Li, J. Duan, G. Xue, Z. Xu, W. Xie and J. Zhou, Energy Environ. Sci., 2017, 10, 1923–1927.
34. Y. Yang, R. Zhao, T. Zhang, K. Zhao, P. Xiao, Y. Ma, P. M. Ajayan, G. Shi and Y. Chen, ACS Nano, 2018, 12, 829–835.
35. Q. Jiang, L. Tian, K.-K. Liu, S. Tadepalli, R. Raliya, P. Biswas, R. R. Naik and S. Singamaneni, Adv. Mater., 2016, 28, 9400–9407.
36. J. Wang, Y. Li, L. Deng, N. Wei, Y. Weng, S. Dong, D. Qi, J. Qiu, X. Chen and T. Wu, Adv. Mater., 2017, 29, 1603730.
37. H. C. Jin, G. P. Lin, L. Z. Bai, A. Zeiny and D. S. Wen, Nano Energy, 2016, 28, 397–406.
38. X. Li, R. Lin, G. Ni, N. Xu, X. Hu, B. Zhu, G. Lv, J. Li, S. Zhu and J. Zhu, Natl. Sci. Rev., 2017, 5, 70–77.
39. L. Zhu, M. Gao, C. K. N. Peh and G. W. Ho, Mater. Horiz., 2018, 5, 323–343.
40. H. Liu, C. Chen, G. Chen, Y. Kuang, X. Zhao, J. Song, C. Jia, X. Xu, E. Hitz, H. Xie, S. Wang, F. Jiang, T. Li, Y. Li, A. Gong, R. Yang, S. Das and L. Hu, Adv. Energy Mater., 2018, 8, 1701616.
41. S. Zhuang, L. Zhou, W. Xu, N. Xu, X. Hu, X. Li, G. Lv, Q. Zheng, S. Zhu, Z. Wang and J. Zhu, Adv. Sci., 2018, 5, 1700497.
42. X. Yang, Y. Yang, L. Fu, M. Zou, Z. Li, A. Cao and Q. Yuan, Adv. Funct. Mater., 2018, 28, 1704505.
43. M. Zhu, Y. Li, F. Chen, X. Zhu, J. Dai, Y. Li, Z. Yang, X. Yan, J. Song, Y. Wang, E. Hitz, W. Luo, M. Lu, B. Yang and L. Hu, Adv. Energy Mater., 2018, 8, 1701028.
44. Y. Li, T. Gao, Z. Yang, C. Chen, Y. Kuang, J. Song, C. Jia, E. Hitz, B. Yang and L. Hu, Nano Energy, 2017, 41, 201–209.
45. H. Liu, X. Zhang, Z. Hong, Z. Pu, Q. Yao, J. Shi, G. Yang, B. Mi, B. Yang, X. Liu, H. Jiang and X. Hu, Nano Energy, 2017, 42, 115–121.
46. M. Gao, C. K. N. Peh and G. W. Ho, Energy Environ. Sci., 2016, 9, 3151–3160.
47. L. Zhu, M. Gao, C. K. N. Peh, X. Wang and G. W. Ho, Adv. Energy Mater., 2018, 8, 1702149.
48. W. Xu, X. Hu, S. Zhuang, Y. Wang, X. Li, L. Zhou, S. Zhu and J. Zhu, Adv. Energy Mater., 2018, 8, 1702884.
49. Safe Drinking-Water from Desalination, WHO, 2011, http://who.int/water_sanitation_health/publications/2011/desalination_guidance/en/.
50. A. H. Smith, P. A. Lopipero, M. N. Bates and C. M. Steinmaus, Science, 2002, 296, 2145–2146.
51. S. P. Surwade, S. N. Smirnov, I. V. Vlassiouk, R. R. Unocic, G. M. Veith, S. Dai and S. M. Mahurin, Nat. Nanotechnol., 2015, 10, 459–464.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c8mh00386f

‡ These authors contributed equally to this work.

---