Solar steam generation (SSG) has been proposed as one of the most advanced techniques to trigger solar energy desalination of sea water. Although many efforts have been dedicated to develop SSG devices, the efficiency remains relatively low. Previous work was mainly focused on thermal insulation film and light absorption. Attention has seldom been concentrated on device structure. Inspired by the manner of water transportation within flowers, we designed an artificial SSG unit which can effectively speed up the water transpiration from the bulk to the surface. Another advantage of such a device is that steam generation is separated from the bulk salty solution and thereby the solar thermal evaporation can be improved greatly. As demonstrated via the desalination experiment, the mass change and evaporation rate under 1 solar irradiation can reach as high as 2.51 kg/m2 and 1.26 kg/m2·h−1, respectively. Meanwhile, the evaporation efficiency is 74%. These values are much higher than those of traditional SSG devices and bulk water.
Solar-energy-related techniques have been developed and applied in many fields owing to the abundance of solar energy (Bae et al. 2015; El-Bialy et al. 2016; Jiang et al. 2016; Tian et al. 2016; Wang et al. 2016; Wang et al. 2017; Al-harahsheh et al. 2018; Pouyfaucon & García-Rodríguez 2018). Potable water scarcity has become a worldwide issue hindering sustainable development of society with increasing requirements from growing populations and water pollution (Yang et al. 2016; Yue et al. 2019). Due to the massive sea water resource on our planet, desalination is envisioned to be the preferable, environmentally benign manner to obtain fresh water. Among all desalination techniques, solar steam generation (SSG) is regarded as one of the most promising desalination technologies, profiting from the direct use of solar irradiation as the energy source, which is inexhaustible and widely distributed, and thus it can be utilized anywhere with ultralow capital-cost. Nevertheless, traditional SSG suffers from poor efficiency. Actually, in a typical SSG process, the direction of energy is roughly divided in three ways: (1) total enthalpy of the liquid–vapor phase change (sensible heat and phase-change enthalpy); (2) optical losses including reﬂection and transmission of incident light; (3) thermal losses occurring in the transfer and exchange with the environment. In this case, most SSG performances have been obtained by directly tiling the film on the bulk water. As a result, photo-thermal energy might be utilized to heat the bulk water, leading to low desalination efficiency. For instance, Ghasemi et al. (2014) fabricated a double-layer structured film consisting of carbon foam and exfoliated graphite as SSG, revealing an efficiency of 85% under 4 suns. Zhou et al. (2016a) prepared SSG absorbers by depositing metal nanoparticles onto a nanoporous template, demonstrating a high efficiency of 90% under 4 suns. Further, Zhou et al. (2016b) proposed a three-dimensional porous membrane as an SSG film, reaching an efficiency of 80% under 3 suns, but when the desalination testing was performed under only 1 sun, the efficiency was as low as ∼57%.
Inspired by the water transpiration behavior in ‘auto-flowerpot’-water automatically pumped from the flowerpot by artificial flower roots, manufactured by low density polyethylene (LDPE) filled with cotton, transported from the bottom to the top through flower roots and finally released to the atmosphere through the leaves in this work we proposed a flowerpot-inspired design concept of an SSG device based on vacuum-filtrated graphene film. The ultra-thin graphene film is capable of maximizing light absorption to promote photo-thermal conversion. The evaporation zone is separated from the bulk water by utilizing expanded polystyrene (EPS) as a water transportation tunnel. In this case, the device is able to restrict heat into a small amount of water in the evaporating surface, and may substantially improve the efficiency of solar energy utilization. As a result, the evaporation rate of the modified SSG reaches 1.26 kg/m2·h−1 and the photo-thermal conversion efficiency reaches 74% under 1 sun illumination.
Preparation of graphene film
Graphene oxide (GO) was prepared according to the modified Hummers' method (Li et al. 2014). GO powder was obtained by freeze-drying for 48 hours (−55 °C), and 84.25 mg GO was dissolved in 50 mL deionized water (DI) by ultrasound for 2 hours. Then, GO was reduced by adding 15 mL NaOH (4 mol/L) in a water bath at 90 °C and holding for 15 minutes. Finally, reduced GO (reduced graphene oxide (rGO)) film was obtained by employing vacuum filtration (0.5 mg/mL, 4 mL).
The morphology of rGO film was identified on a scanning electron microscope (SEM, S4800) and transmission electron microscope (TEM, HT7700). The surface temperature of the membrane and infrared (IR) images were captured by E8 FLIR. The absorbance was measured on an ultraviolet-visible-near-infrared diffuse reflectance tester (Lambda750). The speed of water evaporation was measured by a system consisting of an analytical balance connected to a computer and a source simulation. Source simulation was provided by a xenon lamp (CEL-PE300 L-3A, China). The sample applied in the SSG experiment floated flat on the aquifer. The simulated sun illuminated the film perpendicularly from the top. For simplicity, the traditional SSG device and modified SSG device were denoted as T-SSG and M-SSG devices, respectively.
T is time; m is the cumulative change of water mass in t time; s is the effective area of evaporation.
Hv (2,260 kJ kg−1) is the heat of evaporation of water, Hx is the latent heat of evaporation of water, ve is the rate of water evaporation (directly obtained from Equation (1)), Q (1 kW m−2) is the energy simulating the vertical radiation of the solar instrument on the surface of the SSG film.
RESULTS AND DISCUSSION
Figure 1(a) and 1(b) are photographs of the T-SSG device and M-SSG device, respectively. As illustrated, when the sunlight illuminates the rGO film, the black-natural surface will absorb the sunlight and commence heating the water–air interface. Then the temperature of the illuminated zone starts to increase and evaporation is generated, running to the atmosphere. Basically, there are three dominant heat consumptions accompanying the SSG process, which can be classified as water heating, parasitic thermal loss, and water vaporization. Only water vaporization is effective. Regarding the T-SSG device, it suffers from inefficient utilization of converted heat since evaporation and water transportation simultaneously occur in the same local zone, leading to large energy loss during the SSG process. However, in terms of the M-SSG device, as the evaporation is continuously carried out, water from the bottom will be pumped up simultaneously through the LDPE vessels, promoting the continuous supply of water for evaporation. The process is very similar to water transpiration in an ‘auto-flowerpot’. The mechanism may be proposed as follows: the negative pressure at the top of the rGO film caused by water evaporation ensures a large capillary force inside the rGO film. On the other hand, rGO film is a superior sunlight collector as well as solar–thermal convertor and thereby the converted heat will be localized on the surface of rGO film, and effective heat for steam generation thus can be achieved. In this case, high efficiency is expected for the rGO-film-based M-SSG device due to the effective thermal management, water transportation and evaporation.
The TEM image of as-prepared GO is shown in Figure 2(a). Large-area and transparent thin layers can be observed, indicating the successful synthesis of GO. Figure 2(b) is the digital image of vacuum-filtrated rGO film, which is black and smooth. The inset of Figure 2(b) exhibits the SEM image of rGO which illustrates the layered morphology with wrinkled surface. Such structure is beneficial for enhancing the absorption of sunlight. Figure 2(c) draws the Raman spectra of GO and rGO for comparison. It is found that a broadened 2D peak emerged in rGO, implying the existence of a large amount of graphitic carbon. Additionally, the intensity of the G peak over the D peak is obtained as 2.0, demonstrating the presence of defects in the rGO film. Figure 2(d) shows that the average absorbance of the rGO film is 0.7, indicating the high utilization of incident light.
The temperature distribution of the bulk water, T-SSG device and M-SSG device is recorded by an IR camera and shown in Figure 3. Before commencement of solar illumination, the surface temperature of pure water (Figure 3(a), 17.2 °C) is about the same as that of the T-SSG device (Figure 3(b), 17.5 °C), whereas the temperature of the M-SSG device (14.7 °C) shows an obvious decrease due to the bulk water being separated from the rGO film. After the 1 sun illumination had lasted for 120 minutes, the surface temperature of the bulk water, T-SSG device and M-SSG device increased from 17.2 °C to 38.4 °C, 17.5 °C to 40.6 °C, and 14.7 °C to 42.1 °C, respectively. These data suggest that heat is localized only on the top surface of the M-SSG device.
Figure 4(a) plots the surface temperature versus time curves for the bulk water, T-SSG device and M-SSG device under the irradiation of 1 sun. As compared with the bulk water and T-SSG device, the M-SSG device increased quickly in the initial 900 s and became mostly steady afterwards. Although the same sunlight absorption layer (rGO) was employed, the M-SSG device demonstrates a higher maximum steady temperature than that of the T-SSG device due to the optimized device structure. Although the difference is not huge, it is adequate to improve the photo-thermal conversion effectively.
The SSG performance was investigated in a home-made system. The mass changes with respect to time for the bulk water, T-SSG device and M-SSG device under 1 sun irradiation of 1 kW m−2 are shown in Figure 4(b). It should be mentioned that the mass was carefully tracked by a balance as a function of irradiation time. Clearly, as irradiation time goes up, the mass change for all devices increases quickly. However, the increment step for the T-SSG device is the highest of all. Moreover, the evaporation rate of the M-SSG device increases quickly in the initial 30 minutes and then stabilizes afterwards under 1 sun, whereas the evaporation rate for both the bulk water and T-SSG device slowly increases within 120 minutes (Figure 4(c)). The steady evaporation rate of the M-SSG device after 120 minutes is 1.26 kg/m2·h−1, which is much higher than those of the T-SSG device (0.84 kg/m2·h−1) and the bulk water (0.55 kg/m2·h−1).
The solar vapor conversion efficiency η can be obtained through Equation (2) and is exhibited in Figure 4(d) (note that the environmental evaporation rate in darkness was 0.0768 kg/m2·h−1). Accordingly, the evaporation efficiency of the M-SSG device is 74%, which is 1.54 times higher than that of the T-SSG device (48%) and 2.5 times higher than that of the bulk water (30%).
To improve the water evaporation efficiency of rGO film, a modified solar thermal generation (SSG) device has been designed. In the SSG device, the water transpiration from the bulk to the surface can be effectively enhanced. On the other hand, the steam generation zone is separated from the bulk salty solution and thereby the solar thermal evaporation can be improved significantly. As a result, the evaporation rate and mass change of rGO film as well as evaporation efficiency have been demonstrated to be much higher than that of traditional SSG devices and bulk water.
This work was supported by Ningxia Key R&D Program (2017BY064).