The schematic diagram for the construction of SVGs in our strategy is shown in Fig. 1. Owing to the unique interaction between the polymer chains and the water in the swelled cryogel, the cryogel was first considered as an ideal SVG. However, owing to the high density of the swelled cryogels, it is difficult to float them on water for a long period of time (Fig. S1). Therefore, a crosslinked CNF/PLA composite aerogel based on Pickering emulsion technology was proposed as the cryogel substrate owing to its advantages: the CNF/PLA composite aerogels can provide higher water resistance and anti-compression properties, while still maintaining similar porosity, thermal conductivity, and hydrophilicity as the pure CNF aerogel (Fig. S2, S3) (S. Li, Y. He, et al., 2020; S. Li, C. Zhou, et al., 2020). Additionally, their specific shapes can be obtained through an ingeniously designed mold (Fig. S4).
As an important derivative of cellulose, CMC is one of the most widely used eco-friendly materials with a rich content of carboxyl and hydroxyl groups on the cellulose backbone, and is also available in large amounts (Shu et al., 2020). Thus, in this study, CMC was considered as an ideal candidate for preparing cryogels. As shown in Fig. 1a, the CMC cryogel containing polyaniline (PANI) nanoparticles were prepared by a simple one-pot in situ crosslinking reaction through glutaraldehyde (GA). The detailed preparation process and formulations for the cryogels (CG) are presented in Sect. 2.4 and Table S1, respectively, with detailed abbreviations. The chemical composition of the CGs was verified by the Fourier transform infrared (FTIR) spectrum shown in Fig. 1b. In the CMC spectrum, the peaks at 3414 cm− 1, 1617 cm− 1, 1054 cm− 1, and 1420 cm− 1 represent the stretching vibrations of O–H, –COOH, and C–O–C, and the scissoring vibration of –CH2, respectively. In the CG-0 spectrum, a new absorption peak corresponding to the aldehyde peak (CHO) appeared at 1716 cm− 1, which confirms the formation of a crosslinked network (Fig. 1c). For the PANI spectrum, the characteristic absorption peaks at 1561 cm− 1, 1480 cm− 1, 1301 cm− 1, 1246 cm− 1, and 1137 cm− 1 correspond to the C = C stretching deformation of the quinoid rings, C = C stretching deformation of the benzenoid rings, C–N stretching vibration of the aromatic amine, doped C-N.+ aromatic amine peak, and in-plane bending vibration of the plane deformation of the C–H in the benzene ring, respectively. The characteristic peaks of CMC, PANI, and GA are clearly observed in the spectrum of CG-5, which confirms the formation of CMC cryogel crosslinked by GA and the successful introduction of PANI in CG-5.
When the CNF/PLA composite aerogel with a specific shape was introduced into the CMC viscous aqueous solution containing crosslinkers, the viscous solution was infiltrated into the bottom of the aerogel by the capillary force. Accompanied by the subsequent cross-linking and gelation reaction, the aerogel and hydrogels were firmly combined through chemical bonds and hydrogen bonds. The cross-linking interaction was further enhanced by repeated freezing and thawing treatment, and an integrated A-CG was finally prepared through the freeze-drying and rehydrating process (Fig. 1a). The resultant A-CG exhibited excellent interface bonding, as can be initially inferred from the phenomenon shown in Fig. S5. When the cylindrical aerogel (diameter of 1.2 cm) was held and lifted, the cryogel after being completely swollen with water and Petri dish (total weight of approximately 35 g) were also lifted, and the A-CG still maintained a complete structure. To further evaluate the robustness and stability of the A-CG and interface bonding, the A-CG with a particular shape shown in Fig. S6 was continuously subjected to a series of severe conditions, such as ultrasonic agitation (400 W, 8 h), boiling water (98°C, 4 h), strong oxidizing acid (1M H2SO4, 24 h) environments, and even soaking in water for 30 days. After these rigorous tests, the appearance of A-CG did not significantly change, and the A-CG still exhibited an excellent self-floating property, which again confirms the structural stability and strong interface bonding force of the A-CG.
Additionally, the morphologies of the vertical section of the A-CG-5 shown in Fig. 2a, b were also investigated using scanning electron microscopy (SEM). In the A-CG, the cryogel part (CG-5) as the solar absorber of SVG had a typical macroporous structure with a pore diameter of approximately several hundreds microns (Fig. 2c). The PANI photothermal material was randomly distributed onto the pore wall (Fig. 2d). The aerogel part also had a large number of interconnected micron pore structures. Compared with the cryogel, the aerogel had smaller pore diameter, and smoother and flatter pore wall structure (Fig. 2e, f). These relatively small pore structures will have stronger capillary forces to ensure sufficient water supply during the evaporation process (Z. Wang, Wu, He, Peng, & Li, 2021). When the bottom of the aerogel was placed in an aqueous solution saturated with methylene blue, the excellent hydrophilicity and strong capillary force could overcome gravity to draw water to the top of the aerogel, which indicates its excellent water transport ability (Fig. S7). In the cross-sectional SEM image, a hierarchical pore structure between the aerogel and the cryogel can be clearly observed (Fig. 2g). The high magnification SEM micrograph (Fig. 2h) also shows that the aerogels and cryogels were effectively combined, which ensures the structural stability and facilitates continuous water transport.
As has been previously reported, the water in the cryogel after being completely swollen with water can be divided into bound water, intermediate water, and free water (X. Zhou et al., 2020; X. Zhou, Zhao, et al., 2019). Additionally, the evaporation enthalpy of the water in the cryogels is greatly influenced by the intermediate water owing to the weak hydrogen bond interaction between the water and the polymer molecular chains. Hence, differential scanning calorimetry (DSC) was used to reveal the phase change of water in the CGs. In Fig. 3a, the melting point peaks corresponding to the intermediate water (blue dotted curves) and free water (blue curves) are clearly observed, which is consistent with the result reported by Yu et al (X. Zhou et al., 2020; X. Zhou, Zhao, et al., 2019). This confirms that the CMC molecular chains can facilitate water activation transformation from free water to intermediate water owing to the formation of hydrogen bond interaction between the carboxyl group in the CMC and water, which leads to the lower evaporation enthalpy of CGs compared with pure water. Hence, to accurately calculate the water steam generation efficiency of the final obtained SVG, the precise enthalpy of the evaporation of A-CGs with different PANI content was investigated (see Supplementary Materials Sect. 1.1 for details). The evaporation rate (\(\dot{m})\)can be calculated as follows (Y. Wang et al., 2018):
where m is the mass of the evaporated water, \(S\) is the evaporation area, and t is the evaporation time. The evaporation rate of pure water, A-CG-0, A-CG-2, A-CG-5, and A-CG-8 without solar irradiation was 0.043, 0.063, 0.0586, 0.0582, and 0.0567 kg m− 2 h− 1, respectively (Fig. 3b). The evaporation enthalpies of water in all A-CGs were significantly lower than that of water with the same evaporation area. Moreover, with the continuous increase of PANI content, the evaporation enthalpy of water in A-CG-0, A-CG-2, A-CG-5, and A-CG-8 also increased, and was 1.50, 1.65, 1.66, and 1.70 kJ g− 1, respectively. The increase of the evaporation enthalpy is attributed to the fact that the N-H in PANI can also form hydrogen bonds with COOH in CMC, which may weaken the hydrogen bond interaction between the CMC and water and diminish the formation of intermediate water. Compared with A-CG-8, the A-CG-2 and A-CG-5 had the lower water evaporation enthalpy for accelerating the water evaporation of the final obtained SVG.
Additionally, high solar absorption is an essential prerequisite for an SVG with high performance. Therefore, the effect of the PANI content in the CGs on the light absorption of the CGs was also investigated using an ultraviolet-visible near-infrared (UV-Vis-NIR) spectrophotometer equipped with an integrating sphere. The transmission and reflection spectra of CGs with different PANI content between 200 nm and 2500 nm were measured to further calculate their absorption spectrum. (Fig. 3c, d) The A-CG-0 without PANI exhibited a high diffuse reflection (approximately 40%) in the visible light region, which indicates lower absorption. After adding the PANI light-absorbing material, the diffuse reflection of A-CG-2 to light was significantly reduced; therefore, high light absorption (approximately 98%) was achieved. Notably, the absorption of A-CG-5 and A-CG-8 wasslightly higher than that of A-CG-2. Moreover, the water contact angle test confirms that all CGs with different PANI content have excellent hydrophilicity and wettability (Fig. S8). In conclusion, a higher PANI content can enhance the light absorption of the CG. However, a higher PANI content can also weaken the interaction between the CMC and the water molecules. Therefore, CG-5 was selected to construct the SVG used in the subsequent experiment.
Apart from the high solar absorption, the thermal management capacity of the SVGs is also as an important factor influencing the acceleration of evaporation. For comparison, three SVG types with different shape were designed and their diagrams are shown in Fig. 4. As stated above, once the CG-5 were put in a breaker full of water, they sank to the bottom of the breaker (Fig. S1). Although the A-CGs can self-float on water, all A-CGs were still located below the water surface (Fig. 4, S6). Hence, the CG and A-CG as SVGs maybe result in more heat loss owing to the high thermal conductivity of water. Thus, we propose a mushroom-shaped evaporator supported by insulated polystyrene (PS) foam (A-CG/F) (Fig. 4). Compared with A-CG, A-CG/F can protrude above the water surface through the buoyancy of the PS foam to achieve a larger evaporation area. Additionally, PS can be used as a heat insulation layer to further limit the heat loss to the lower bulk water.
To further verify this conclusion, the thermal conductivity of the CG-5 and obtained CNF/PLA aerogel was calculated in accordance with the experimental method described in Supplementary Materials Sect. 1.2 and the corresponding results shown in Fig. S9. Obviously, the thermal conductivity of the CNF/PLA aerogel filled with water is 0.305 W m− 1 K− 1. This is much lower than that of the fully swollen CG-5 (0.570 W m− 1 K− 1), which is close to that of pure water (0.609 W m− 1 K− 1) (N. Shalkevich, A. Shalkevich & Burgi, 2010). Undoubtedly, the higher thermal conductivity accelerated the heat loss to the bulk water and weakened the evaporation rate. An infrared (IR) thermal imaging camera was used to dynamically monitor the temperature changes of pure water, CG, A-CG, and A-CG/F under the irradiation of 1 Sun. Owing to the high thermal conductivity and high light transmittance of water, the temperature of pure water under irradiation increased almost uniformly (Fig. 4). The CG sank to the bottom of the water after being completely swollen, and the sunlight was absorbed by the CG through the water. Hence, the temperature of the water surface and CG was higher than that of other areas, which is consistent with the bottom heating phenomenon proposed by Deng et al (Tao et al., 2018). With the introduction of the CNF/PLA aerogel, on one hand, the A-CG easily floated on the water surface, absorbed solar energy, and converted this energy into heat. On the other hand, a small part of the generated heat was conducted through the aerogel to the bulk water owing to the lower thermal conductivity of the aerogels, which led to the rapid increase of the temperature of the top water surface and the slow increase of the temperature of the bulk water. Which reached up to 35.1°C and 21.6°C, respectively, when the evaporation process was stable. Interestingly, A-CG/F can further reduce the heat loss to the bulk water. After a long period of irradiation, much less heat was lost to the bulk water, which caused the temperature of the bulk water to increase only slightly by 0.1°C. Hence, A-CG/F is considered to be an ideal SVG with excellent thermal management capability.
The vapor evaporation performances of the SVGs under the irradiation of 1 Sun were further evaluated. As shown in Fig. 5a and b, the average evaporation rate was 0.46, 0.48, 1.84, and 2.16 kg m-2 h-1, corresponding to pure water, CG, A-CG, and A-CG/F, respectively. This high evaporation rate for the A-CG/F is attributed to the following reasons. First, the abundant hydrophilic groups on the CMC surface can reduce the evaporation enthalpy of water. As stated above, the evaporation enthalpy of water in A-CG-5 was 1.66 kJ g-1, which is much lower than that of pure water (approximately 2.26 kJ g-1). However, macroporous cryogels have high light absorption within a broad band. Secondly, the aerogel with excellent hydrophilicity and relatively small pore structure endows the A-CG/F with stronger capillary action for pumping water. Additionally, a larger evaporation area and a minor heat loss to the bulk water were achieved owing to the protrusive structure of the A-CG/Fs through the buoyancy of the PS foam, which can extremely accelerate the evaporation of water. he steam generation efficiency (ηth) is defined as follows:
All steam generation efficiencies were calculated by subtracting the evaporation rate without solar irradiation (Fig. 5b). Notably, Copt is the optical concentration and qi is the standard solar irradiation (1 kw/m2), which can be calibrated using a full spectrum glare optical power meter. Here, hLV was determined by the actual evaporation enthalpy in the experiment described in Supplementary Materials Section 1.1. Thus, the steam generation efficiency of the SVGs was calculated as 26.1%, 31.9%, 82.1%, and 93.6%, corresponding to pure water, CG, A-CG, and A-CG/F, respectively (Fig. 5c).
Energy balance analysis was conducted to further demonstrate the excellent performance of A-CG and A-CG/F in terms of the effective use of heat (see Supplementary Materials Sect. 1.3 for details). As shown in Fig. 5d, the A-CG device had a radiation loss of 5.5%, convection loss of 4.5%, and conduction loss of 11.4%. Compared with A-CG, the radiation and convection losses of A-CG/F slightly increased to 5.7% and 5.8%, respectively (Fig. 5e). This is attributed to the increase in the area of heat exchange with the environment owing to the raised A-CG/F structure, which resulted in an increased evaporation area. In contrast, the conduction loss sharply decreased to 0.5%, and this was the primary reason for the PS ring with an extremely low thermal conductivity of 0.03 W m− 1 K− 1(T. Li et al., 2018) exhibiting a better thermal insulation effect compared with the water-filled aerogel, whose thermal conductivity was 0.305 W m− 1 K− 1.
A cyclic performance test was also performed on the A-CG/F irradiated under 1 Sun for 2 h each day over two weeks. After 15 cycles, the evaporation rate and steam generation efficiency of the A-CG/F were still maintained at approximately 2.15 kg m− 2 h− 1 and 93.3% (Fig. 5f), respectively, which indicates that the sample has excellent reusability. Additionally, the salt rejection performance of A-CG/F was investigated. Various NaCl particles were placed on the surface of A-CG/F, owing to the good hydrophilicity of A-CG/F and the macroporous structure of the cryogel, almost no salt was observed on the surface of A-CG/F after being subjected to an irradiation of 1 Sun for 4 hours (Fig. S10).
As is known, the air flow in the evaporation environment is another important stimulus (N. Li et al., 2020). The effects of air flow on solar evaporation, such as the vapor temperature on the light-absorbing surface, windward surface, and leeward surface of the evaporator, were systematic investigated. Therefore, using the device shown in Fig. S11, the evaporation rate of the evaporator at different wind speeds and the vapor temperature at different evaporation positions were detected in real time. With the continuous enhancement of wind energy, the evaporation rate of A-CG/F with the wind speed of 1, 2, 3 m/s reached 2.66, 3.90, and 5.67 kg m− 2 h− 1, respectively (Fig. 6a). The reason for this is the fact that the air flow in the evaporative environment pushes away air with higher humidity and brings in air with lower humidity such that the water in the A-CG/F can evaporate into the environment in a free manner. Therefore, even in the absence of light, the increase of wind speed significantly increased the evaporation rate of A-CG/F (Fig. 6b). On the other hand, under windy energy the surface temperature of the A-CG/F was lower than the ambient temperature, meaning that the water steam generation can further be facilitated by the additional heat energy form the environment. This can be concluded from the facts shown in Fig. 6d, e, f, S12. When there was no wind, the A-CG/F benefited from its excellent light-heat effect. The temperature of the light-absorbing surface significantly increased and heat was conducted to the side surface. The temperature of the side surface also gradually increased even higher than the ambient temperature (Fig. 4, 6c). When the wind energy was input, the water evaporation process was accelerated accompanied with more heat removed, leading to the surface temperature of A-CG/F was lower than the ambient temperature (Fig. 6d, e, f, S12). This means that, during the evaporation process assisted by wind energy, A-CG/F cannot only receive solar energy from the sun, but can also absorb heat from the environment, which can contribute to the water steam generation. Additionally, the temperature on the windward side was significantly higher than that on the leeward side (Fig. 6d-f, S12), which is attributed to the changes in the wind speed and evaporation rate at different locations. Notably, the evaporation rate of the A-CG/F is very competitive compared with previously reported interfacial steam generators (Table S2) (Chen et al., 2020; Jiang et al., 2018; N. Li et al., 2020; X. Li et al., 2018; Mu et al., 2019; Tan, Wang, Song, Fang, & Zhang, 2019; Y. Wang et al., 2018; Xiao et al., 2019; Yin et al., 2018; X. Zhou, Zhao, et al., 2019; Xingyi Zhou, Zhao, Guo, Zhang, & Yu, 2018). This is an inevitable consequence of our comprehensive consideration of the lower evaporation enthalpy of water in the cryogel swelled by water, greater evaporation areas, and additional energy input.
To demonstrate the solar water purification performance of A-CG/F, solar evaporation tests were conducted on various types of raw water under the irradiation of 1 Sun, and the quality of the collected water was evaluated. Figure 7a shows the solar desalination ability of A-CG/F on 3.5wt% NaCl, 3wt% mixed salt solution and AgNO3 solution. The initial concentration of Na+, K+, Ca2+, and Mg2+ was significantly reduced by approximately four orders, which is far less than that specified in the drinking water standard of the World Health Organization. Even for Ag+, which is required in drinking water, the requirements for pure water after solar desalination are fully satisfied. Notably, A-CG/F cannot only remove the ions in the brine, but can also purify various types of raw water and maintain a high steam generation rate (Fig. 7b). After the purification of wastewater rich in methylene blue and rhodamine B, the absorption peak disappeared, which again confirms the powerful purification ability of A-CG/F (Fig. 7c, d). These results demonstrate the potential of A-CG/F for use in seawater desalination, and domestic and industrial wastewater purification.