Graphene Array-Based Anti-fouling Solar Vapour Gap Membrane Distillation with High Energy Efficiency

Highlights New concept of solar vapour gap membrane distillation (SVGMD) is based on synergizing of nanochannel-guided water transport, localized heating, and membrane separation from feed solution. First-time introduction of the gap enables long-term stability and non-fouling membrane. SVGMD exhibits a solar-water energy efficiency higher than state-of-the-art solar vapour systems. Electronic supplementary material The online version of this article (10.1007/s40820-019-0281-1) contains supplementary material, which is available to authorized users.


S1 Microorganisms-removal Experiments
The microorganisms-removal experiments of SVGMD and conventional photothermal steam generation system are conducted using the standard commercial bio-indicator. The experimental setups are shown in Fig. S1. The bacteria carrier (thermophilic bacillus) is placed into the feed water. Then the purified water is transferred into the bio-indicator and incubated at 56 °C for 48 h. The color of bio-indicator after incubation is applied to indicate the microorganisms-removal results, in which the purple color indicates that the microorganisms can be completely rejected through the membrane and the spore survival does not occur, while the yellow color indicates the spore survival. As shown in Fig. S1, the bio-indicator containing the purified water from conventional photothermal steam generation system changes its color from purple to yellow. Conversely, the bio-indicator containing the purified water after SVGMD process has no change in color (yellow), revealing the microorganisms-removal performance of the membrane.

Fig. S1
a Schematic of the conventional photothermal steam generation system. The standard commercial bio-indicator changes its color from purple to yellow. b Schematic of the SVGMD system. The standard commercial bio-indicator has no change in color, indicating the microorganisms-removal performance. All tests were performed at an illumination of 1 kW m -2 . Table S1 A comparison between SVGMD and conventional photothermal MD in terms of system design, water flow, and heat transfer Top-left: In conventional photothermal MD, the dispersion of light-absorbing materials (e.g., nanoflakes and nanoparticles) is deposited onto the surface of membrane or even incorporated into the membrane. It may block the membrane pores, leading to higher vapour transport resistance.

Middle-left:
In conventional photothermal MD, the feed saline/contaminated water directly contacts with the membrane, leading to the contaminant deposition and thus the fouling problem.

Bottom-left:
In conventional photothermal MD, the feed water flows through the photothermal membrane. The heat concentrated in the light absorber is transferred to the bulk feed water. Substantial amount of heat is lost to the bulk water.
Top-right: Our new conceptual design of free-standing absorber, which enables a gap between the membrane and the absorber. Graphene array is seamlessly anchored on the surface of robust three-dimensional nickel foam.

Middle-right:
Water is transported by the free-standing light absorber, achieving the uniquely engineered water pathways. A thin water layer with self-guided transport from the bulk feed solution surrounds the absorber (i.e., graphene array). The feed saline/contaminated water no longer directly contacts with the membrane, thus resolving the fouling problem.

Bottom-right:
With the engineered water pathways, the as-generated thermal energy is effectively confined in the thin water layer surrounding graphene nanosheets, leading to the surpressed heat loss to bulk water beneath and the directional heat transfer from the light absorber to the thin water layer.

S3 The Fabrication Procedure and Material Characterization of P-G-Nifoam
Fig. S2 a Schematic illustration of preparation of P-G-Nifoam via plasma-enhanced chemical vapor deposition and coating. P-G-Nifoam becomes superhydrophilic (water contact angle θ = 1.8°) after PEDOT-PSS coating. SEM images of P-G-Nifoam b before and c after PEDOT-PSS coating

S4 Underwater Oil Contact Angle Measurements of P-G-Nifoam
The underwater oil contact angle was recorded using a computerized contact angle analyzer (DropMeter A-200, MAIST) and a high-speed camera (MotionXtra HG-100K, REDLAKE). Motor oil droplets (5 L) were injected onto the surface using a motor-driven syringe device. Contact angles were calculated by the Young-Laplace equation.

S5 Excellent Light Harvesting Ability of P-G-Nifoam
As shown in Fig. S4, the graphene array can trap incident light within the nanochannels instead of bouncing off or penetrating the absorber, thereby effectively suppressing the light transmission and reflection and maximizing the light absorption capability. The effect of the vapour gap width on flux and solar-water energy efficiency. The increase of the permeate flux is negligible as the gap width is reduced from 2 to 0.5 mm. So we choose a gap width of 1 mm. All tests were performed at room temperature and atmospheric pressure.

Fig. S6
The permeate flux and solar-water energy efficiencies at 5 sun and 10 sun of the SVGMD system. a Permeate water obtained at an illumination of 1 kW m -2 and under the dark environment. The permeate flux under the dark environment is subtracted from all the measured data under the solar illumination. b The vapour temperature as a function of time at an illumination of 1, 5, and 10 kW m -2 . c The permeate flux and solar-water energy efficiencies of the SVGMD system at an illumination of 1, 5 and 10 kW m -2 . All the tests were performed at a room temperature and atmospheric pressure. Figure S6 shows the performance of SVGMD at 5 sun and 10 sun. Vapour temperature increases obviously with the illumination density, from 36.1 °C at 1 sun to 79.9 °C at 10 sun. Driven by the larger temperature gradient across the membrane, the permeate flux is prominently exaggerated by 10 folds, which are calculated to be 1.1, 5.9, and 11.4 L m -2 h -1 at 1, 5, and 10 sun, respectively. The corresponding energy efficiency is 73.4%, 79.3%, and 82.3%, respectively.

S8 Thermal Analysis
Thermal analysis of the solar-vapour conversion process has been conducted, including radiative (qrad), convective (qconv) heat loss to the surroundings and conductive heat loss to bulk water (qcond). The energy dynamic equilibriumcan be expressed as Eq. S1 (S1) Where α and qsolar denote the light absorption of the absorber and solar flux, respectively. The heat flux terms in Eq. S1 are as follows.
The radiative heat loss (qrad) to the adjacent environment can be calculated as Eq. S2: Where is the emission of the absorber, is the Stefan-Boltzmann constant. is the wet surface temperature of the absorber. is the ambient temperature [S1, S2] The conductive heat loss (qcond) to bulk water can be calculated through the temperature gradient in the underlying water: Where k is the thermal conductivity of feed water. The temperature gradient in the underlying water below the sample is measured by two embedded thermocouples (i.e., Tl1 and Tl2) placed in the water tank. The distance between the thermocouples Δl is 10 mm.
The convective heat loss ( qconv) to the adjacent environment can be calculated as Eq. S4: Where h is the convection heat transfer coefficient and is assumed to be 5 W m -2 K -1 .

Fig. S7
The wet surface temperature of P-G-Nifoam as a function of time in the SVGMD system. The temperature is stabilized at about 39.1 °C at an illumination of 1 kW m -2 , which could be approximated as the wet surface temperature of the absorber (Tα)

S9 Salt Rejection in the Current SVGMD System
Salt rejection is defined as a measure of the quality of the flux and the percent reduction in salinity after distillation process, which can be calculated as Eq. S5: where is the conductivity of the feed, is the conductivity of the distillate.

S12 Cyclic Performance Tests of P-G-Nifoam
The salt deposited on the sample would affect the desalination performance [S3]. As shown in the Fig. S10, only a few salt particles deposit on the surface of P-G-Nifoam after processing the natural seawater for 72 h. The long-term stability can be mainly attributed to two factors. Firstly, the surface of G-Nifoam coated with PEDOT-PSS has the characteristics of preventing adhesion of the salt particles to the surface. Secondly, under illumination, the salt ions remained in the P-G-Nifoam result in increased salt concentrations over the natural seawater. Meanwhile, the accumulated ions at the surface of P-G-Nifoam continuously diffuse through the waterways back into the bulk water. The spontaneous ion diffusion driven by salt gradient avoids salt accumulation at the surface of P-G-Nifoam.