A high-ux, self-cleaning solar water evaporation system

Solar interfacial evaporation is a promising technology to produce clean water from seawater or polluted water. However, the evaporation rate upon existing photothermal materials is normally limited to 2 kg m -2 h -1 or below, which is hard to fulfil the daily water demand of a family, let alone enormous water requirements in agricultural and industrial sectors. Herein, under the guidance of computational fluid dynamics (CFD) simulations, multichannel photothermal rod (MCPR) is proposed as solar interfacial material for evaporative disposal of real municipal sewage and concurrent production of freshwater. The production rate of the freshwater over 10-centimetre-high MCPR is 18.8 kg m -2 h -1 under 1 sun, with water quality equivalent to that of commercial pure water. When the light is incident at oblique angles, faster evaporation rates are obtained (e.g., 31.3 kg m -2 h -1 at 30° angle). Furthermore, the freshwater generation rate can even be promoted to 126.5 kg m -2 h -1 in outdoor environment via a magnified evaporation system constructed of 35-centimetre-high MCPRs array. These values were achievable due to a series of properties of the MCPR, including antigravity water transport, omnidirectional collection of solar energy, minimization of heat dissipation, and maximization of the evaporation interface. d e Comparisons among the evaporation rates of the MCPR arrays and advanced solar-driven evaporation systems in the literatures ( R stands for reference; 1 sun radiation). f Effect of the height of the MCPR array on air turbulence. g Effects of the height of the MCPR array on the moisture transport. between top irradiation intensity and evaporation rate at different time. e-h Irradiation intensities at the directions of east, south, west, and north, respectively. The inset represents the deviation of the sampling point from the azimuth angle. i Long-term stability of the evaporation system. of further the a striking evaporation rate of 126.5 kg m -2 h -1 is achieved in an outdoor due the reinforcement air turbulence and the steam escape from the evaporation interface. This work identified that the key for high-performance solar evaporation systems is the presence for both VCA and HCA structures, and the total heigh of 1D evaporators. These findings will guide the development of solar evaporation systems suitable for high-throughput purification of water from household, industrial wastewater, or seawater.


Introduction
Clean water is essential for the survival and well-being of human society 1,2 . However, industrialization, urbanization, and population growth have exposed modern society to face the intertwined challenges of water scarcity and energy crisis 3 . Freshwater production via solar-assisted interfacial evaporation is a promising way to address the issues with minimized carbon footprint 4 , as the sole power input of this technology is renewable solar energy, whereas the raw water can be seawater, brine, and even polluted water 5 . Seawater cannot be accessed easily by more than 3 billion residents who live away from the coast, while wastewater (e.g., municipal sewage, surface water, industrial wastewater, etc.) are ubiquitous 6 . Apart from freshwater production, solar evaporation of wastewater integrates the process of wastewater disposal, thus gaining extra environmental profit (namely kill two birds with one stone).
In the past few years, a variety of solar interfacial evaporators based on carbonaceous materials 7 , hierarchical gels 8 , plasmonic nanoparticles 9,10 , and semiconductors have been proposed 11 , and a record high evaporation rate of 10.9 kg m -2 h -1 is achieved over three-dimensional (3D) interconnected porous carbon foam 12 . If these evaporators were employed to supply daily water for a typical household with four people (assuming 150 litre freshwater per person per day and 8 hours of sunshine) 13,14 , the area of the setup will exceed a few dozen square meters (Supplementary Table 1).
Such a bulky size not only increases the manufacturing cost but also takes up a lot of space. Therefore, it is worth further exploring more efficient solar-driven evaporation system. Generally, solar evaporators are designed in the form of floating two-dimensional (2D) membranes or three-dimensional (3D) blocks with thermal-insulating bottom-layer to immerse in the water and photothermal top-layer out of water (Fig. 1a) 15 . Floating structures not only sacrifice the evaporation area (ca. 50% utilization of the total surface) but also accelerate heat dissipation to the bulk water via the large water/evaporator interface and short heat transfer distance. Computational fluid dynamics (CFD) simulations ( Fig. 1) suggest that one-dimensional photothermal rod combing both vertical channel array (VCA) and horizontal channel array (HCA) would be a better photothermal material, as will be explained in detail later. In comparison to random and tortuous pores, regular VCA can significantly enhance hydraulic conductivity and diffusion flux according to the Hagen-Poiseuille law and Fick's law of diffusion (Fig. 1b, c) 16 . While HCA behaves as water reservoir to counteract the gravity effect and enables the water to be uplifted to a higher altitude (Fig.   1d, Supplementary Fig. 1-3, and Supplementary Movie 1). Under the regulation of VCA & HCA structure, the multichannel photothermal rod (MCPR) will exhibit faster, longer, and antigravity water transport properties. Meanwhile, the irradiation and evaporation area will be maximized and the heat dissipation to the bulk water will be minimized.
Periodic channels naturally exist in rattan, a kind of plant growing up to 170 metres tall 17 , with incredible power to pump groundwater and nutrients to the top through the slender trunk 18 . Rattan sticks have been widely used as room fragrance diffusers based on their outstanding mass transfer performance. The distinct structure of rattan, in combination with the preceding theoretical prediction, encouraged us to synthesize MCPR using rattan as naturally available material. Herein, we report MCPR prepared through carbonization of rattan. Freshwater was generated through evaporative disposal of real municipal sewage over MCPR. A domelike MCPR protype evaporator (height 15 cm, diameter 10 cm) yielded freshwater at an astonishing rate of 18.8 kg m -2 h -1 under vertical radiation of 1 sun (1 kW m -2 ), with water quality equivalent to commercial pure water. When the light was incident at oblique angles, the water evaporation rate increased to 31.3 kg m -2 h -1 at 30° angle, 66.5% faster than that under vertical irradiation. Freshwater generation rate can even be accelerated to 126.5 kg m -2 h -1 via a magnified evaporation system constructed of 35-centimetre-long MCPR. The MCPR was prepared from the rattan (diameter 6 mm) through a two-step process: delignification in the white liquor (2.5 mol L -1 NaOH and 0.4 mol L -1 Na2SO3) followed by pyrolysis at 800 °C (Fig. 2a). As shown in Supplementary Fig. 4, the resulting MCPR perfectly inherited the natural channels of rattan. Scanning electron microscopy (SEM) images show that the MCPR has a range of directional channels (VCA, parallel to the rod axis) (Fig. 2b-d and Supplementary Fig. 5).

Fabrication and characterization of the MCPR
The statistical frequencies for large channels (radius > 50 μm), medium channels (radius in 5~50 μm), and small channels (radius <5 μm) are 0.69%, 15.22%, and 84.09%, respectively ( Supplementary Fig.   6). The total projection area of large channels constitutes approximately 21.93% of the total channel area. Such high area of large channels favours the hydraulic conductivity of the MCPR according to the Hagen-Poiseuille law (Equation 1) 19 , where Q is the hydraulic conductivity, G the ratio of the pressure gradient across the length of the pipe, r the pipe radius, and μ the fluid viscosity. For example, the hydraulic conductivity of the 185 μm channel is 14 460 000 times larger than that of 3 μm channel. The high hydraulic conductivity enables the water supply to keep up with rapid water evaporation, thus guaranteeing steady solar vapor generation and high efficiency. Further SEM observations of the surface and section of the MCPR reveal that the VCA is interconnected through periodically aligned holes (HCA) with a planar density ≈2 200 000 holes cm -2 ( Fig. 2d and Supplementary Fig. 5d). HCA provides lateral pathways for water diffusion and evaporation. In control experiments, the MCPR was prepared without the delignification procedure while kept other conditions unchanged. With this simplified scheme, only VCA structure is present but the periodically aligned holes (HCA) are scarce ( Supplementary Fig. 7). Given this work is dedicated to using municipal sewage as the raw water to produce freshwater, the antifouling and self-cleaning properties would be indispensable to the MCPR. Consequently, photocatalytic TiO2 needles capable of non-selectively decomposing organic pollutants were grown on the surface of the MCPR (TiO2 loading: 1.5wt%) 20,21 , followed by annealing treatment in hydrogen atmosphere 22  near infrared (NIR) spectra ( Supplementary Fig. 11). The long wavelength bands interact with the MCPR to transfer the energy to the carbon lattice by electron-phonon coupling, which generate heat for water evaporation. The short wavelength bands excite H-TiO2 to generate electrons and holes for photocatalytic effect and pollutant decomposition (Fig. 2h).

Water transport performance of the MCPR
To understand the structure-dependent water transport, CFD simulations of the diffusion of fluid in the MCPR were carried out. The models for CFD study are three virtual rods featuring high curved channels, low curved channels, and vertical channels (VCA), respectively (Supplementary Fig. 1).  Fig. 3), facilitating the fluid to overcome gravity and to diffuse longer distance. for the carbonized balsa wood. The excellent water transport capacity of the MCPR confirms that vertical channels reduce the water transport resistance, in good agreement with CFD simulations (Fig.   1b, c).
To highlight the importance of the HCA in the MCPR, we performed a control experiment by comparing water transport performance with a sample with only VCA structures (prepared via pyrolyzing rattan without delignification, see above and Supplementary Fig. 7). As shown in Fig. 3b, the testing liquid was uplifted to a height of 11 cm by the MCPR with both VCA & HCA structure, almost threefold of the sample with only VCA structure. This agrees with the CFD simulations suggesting a 10-fold difference (Fig. 1d). The discrepancy is attributed to the small amount of HCA and mesopores in the VCA structured rod. These holes generate capillary actions to pull the liquid to move upward. Taken together, the significant height difference underpins the importance of VCA & HCA structure for long-distance water transport.  (Fig. 3d). The temperature of the side of the MCPR can also increase by more than 10 ℃ after 30 minutes irradiation. This is because the fully exposed sides of the MCPRs can capture the diffuse reflection light and scattered light. When the MCPR was immersed in the water and became moistened, the top temperature was stabilized at 66 ℃ after illumination for 20 minutes (Fig. 3d). The stabilization of the temperature is owing to the energy balance between the adsorption of irradiation energy by water and the energy consumption of volatized steam generation (cooling effect). These results reveal that the MCPR is a promising solar-driven interfacial evaporation material. To support this perspective, the MCPRs array immersed in the water was placed in a closed container and the environmental humidity in the container was monitored by a hygrometer. As depicted in Fig. 3e, MCPR array can increase the environmental humidity from 45.2% to 72.4% within 300 seconds driven by indoor light, and 28.7% to 70.3% by outdoor light. In contrast, the environmental humidity only increased by 3% when water was naturally volatilized through the surface with the same area. The rapid humidity change in a short period of time indicates that the MCPR efficiently utilizes the irradiation to heat the water and generate the steam.

Photocatalytic property of the MCPR
Photocatalytic activity of the MCPR was evaluated by using methylene blue (MB) and methyl orange (MO) solution as model wastewater (Supplementary Fig. 13). By soaking the MCPR in the dye solution overnight under dark conditions, the intensity of the characteristic absorption band was only decreased by ~10%, which is ascribed to the adsorption of the dye by porous MCPR (Fig. 3f, g).
Under the illumination of simulated sunlight, both MO and MB solution were decoloured within 1 hour, indicative of the photocatalytic decomposition of the dyes by H-TiO2 grown on the surface of the MCPR. After 5 cycles of testing (Fig. 3h, i), the degradation efficiencies of MB and MO were still maintained over 99%, signifying the stable photocatalytic performance of the MCPR.
Consequently, through H-TiO2 modification, MCPR acquires solar-driven self-cleaning and antifouling properties. In the preceding section, CFD simulations ( Fig. 1b-d) and water-transport experiments (Fig. 3a, b) revealed that the synergy of VCA and HCA enables long-distance, vertical, and rapid water transport, hence permitting to use long MCPR to construct the evaporator. Long MCPR will expand the steam/evaporator/water interface 12 , promote the convection with the environment, and accelerate the escape of steam from the evaporation interface. To confirm this conjecture, the microenvironments (air turbulence field and moisture distributions) surrounding the MCPR array were simulated using the CFD model. The results indicate that, with the increase of the evaporator height and the MCPR/air interfaces, air turbulence is reinforced and the flow velocity above the evaporator speeds up (Fig. 4f and Supplementary Fig. 15). According to Bernoulli's principle 35  accelerate the outward diffusion of the steam ( Fig. 4g and Supplementary Fig. 16). On the other hand, the forest-like MCPR array contributes the sides to capture obliquely incident light and scattered light in the ambient 36 , which supplies additional radiation energy to the evaporator for photothermal process ( Supplementary Fig. 17) 34 . This accounts for the fact that oblique solar radiation gives a faster evaporation rate relative to vertical illumination.   (Fig. 5e-h), the evaporation rate of the system did not fluctuate considerably (Fig. 5d), consistent with the fact that MCPRs array can capture omnidirectional solar radiation. Throughout the day, water is produced by the system at an average rate of 34 kg m -2 h -1 . Moreover, when we used higher MCPR array (height 35 cm) to fabricate a larger evaporator (Supplementary Table 2 and Supplementary Fig. 18), the elongation of the MCPR further increase the area of MCPR-air interface, reinforces the air turbulence, and accelerates the steam escape. Consequently, the freshwater 20 production rate was significantly increased to 126.5 kg m -2 h -1 (Supplementary Fig. 18). With such high flux of freshwater production, a photothermal evaporator with only 0.6 m 2 working area will be enough to produce enough water for a typical household with four people (126.5 kg m -2 h -1 × 0.6 m 2 × 8 h = 607.2 kg).

Actual sewage purification with MCPR evaporator
Stability and durability are also critical parameters for the solar evaporation system. According to five consecutive experiments (Fig. 5i), the evaporation rate only fluctuated in a narrow range (31.23 ~ 34.67 kg m -2 h -1 ), indicating the stability of the system. Self-cleaning and antifouling properties are supported by thermogravimetric analysis (TGA) (Supplementary Fig. 19). Prior to sewage treatment, the weight loss is about 11.3% at 800 °C for the MCPR. After treating 1 litre of sewage, the weight loss is only increased to 14.7%. Such small difference (~3.4%) in the weight loss suggests that the MCPR can decompose organic pollutants due to the surface modification with photocatalytic H-TiO2 needles. Conversely, when an evaporator is built from TiO2-free MCPR, the loss of the MCPR after treating 1 litre of municipal sewage increased to 39% (Supplementary Fig.   19). The extra 27.7% weight loss (39% -11.3%) is attributed to the micropollutants accumulated in the MCPR. This control experiment underscored the importance for self-cleaning and antifouling properties from H-TiO2 needles for a highly reusable system.
The quality of purified water collected from the solar evaporation system is assessed. As shown in Supplementary Fig. 20, the conductivity of the purified water is only 0.12 mS m -1 compared to the initial municipal sewage of 277 mS m -1 , due to the significant reduction of ionizable salts and organics. From a conductivity standpoint, this water quality is far superior the standards for drinking water of < 200 mS m -1 and < 250 mS m -1 in China (GB 5749-2006) and Europe (98/83/EEC), respectively ( Supplementary Fig. 20a) 37 , and is close to the international standard of the water for analytical laboratories (IOS 3696-1987: Grade 1 <0.01 mS m -1 , Grade 2 <0.1 mS m -1 , and Grade 3 <0.5 mS m -1 ). The COD value of the purified water is almost 0 mg L -1 , far below the municipal sewage (8600 mg L -1 ) and the WHO drinking water guidelines for drinking water quality ( Supplementary Fig. 20b). Moreover, municipal sewage is a reservoir for pathogenic microorganisms.
The Luria-Bertani medium inoculated with the sample of the municipal sewage was full of bacterial colonies according to microbial antibacterial (bacillus coli) experiment ( Supplementary Fig. 20c). In contrast, bacterial colonies were not found in the purified water from this study or commercial pure water. Taken together, these experiments indicate that the freshwater produced from the sewage by the MCPR based solar evaporator has excellent purity and quality.

Discussion and outlook
In summary, we use CFD simulations to guide the design for solar water evaporator materials.
Simulations identified that 1D rod materials with both VCA and HCA internal structures are superior.
Accordingly, rattan-a plant with natural vertical and horizontal channels is chosen to derive MCPR through delignification followed by hydrolysis. The surface of the MCPR is modified by hydrogenated TiO2 needles to engender photocatalytic self-cleaning and anti-fouling functions that are critical for sewage treatment. These surface-modified MCPRs were used to build in a protype kg m -2 h -1 under oblique light with incident angles of 60° and 30°, respectively due to the omnidirectional collection of solar energy by fully exposed surface of the MCPR. By further increase the height of MCPRs, a striking evaporation rate of 126.5 kg m -2 h -1 is achieved in an outdoor environment, due to the reinforcement of air turbulence and the steam escape from the evaporation interface. This work identified that the key for high-performance solar evaporation systems is the presence for both VCA and HCA structures, and the total heigh of 1D evaporators. These findings will guide the development of solar evaporation systems suitable for high-throughput purification of water from household, industrial wastewater, or seawater.

Materials.
Rattan wood was purchased from Qingcun Rattan Factory (Shanghai, China). All chemical reagents were purchased from China Sinopharm Chemical Reagent Co., Ltd. Deionized water was made by the laboratory.

Preparation of MCPR.
First, the natural rattan was soaked in a mixed solution of 2.5M NaOH and 0.4M Na2SO3 at 95 °C for 8 h. The volume ratio of the rattan to the mixed solution is about 1:10.
Then the sample was washed with deionized water several times and then freeze-drying. The sample was carbonized at 800 °C for 2 h in a nitrogen atmosphere to obtain MCPR. Next, 7 mL tetrabutyl titanate was mixed with 2 mL n-butanol and stirred for 1 h to obtain solution A. Simultaneously, 1 mL of deionized water, 1 mL n-butanol and 9 mL glacial acetic acid were mixed and stirred for 1 h to obtain solution B, then solution B and solution A were mixed and stirred for 0.5 h to obtain TiO2 sol.
The prepared MCPR was pulled three times in the TiO2 sol, and the surface was covered with a layer of TiO2 sol, and then transferred to a tube furnace at 450 °C for heat preservation in nitrogen for 1 hour. Then the sample was immersed in a pressure test tube containing tetrabutyl titanate, concentrated hydrochloric acid and deionized water (volume ratio 1:40:40) and heated at 160 °C for 6 h. After that, the sample was cleaned and dried and then purged with H2 at 550 °C for 24 h to obtain a H-TiO2 modified MCPR.
Characterization of the MCPR evaporator. As shown in Fig. 4a, the MCPR device was placed on an analytical balance and irradiated from the top by a solar simulator (CEL-HXF300-T3, Beijing Ceaulight), and record the quality of steam loss in real-time. The data of natural volatilization under dark conditions should be subtracted when calculating the steam efficiency. The photothermal temperature data of the MCPR was recorded through an IR (FLIR One Pro). In addition, the entire device was placed in a light reaction chamber to prevent environmental interference. The detailed characterization strategies are shown in Supplementary Note 1.