Heating water to its boiling point is an important first step not only for preparing a cup of tea or a bowl of pasta, but for a range of applications fundamental to an industrial society, including distillation, sterilization, and power generation. In a solar economy, one could boil water with an electric heater powered by a photovoltaic cell. But it would be far more efficient to use solar energy to heat the water directly.
That’s manifestly possible. For decades solar steam turbines in wide-open sunny spaces have used arrays of mirrors to concentrate sunlight from a large area onto a small volume of water. But those mirrors are expensive: They must be precisely machined to focus light over several hundred meters, and they must be mounted on motors to track the Sun’s position in the sky. Because the motors require that a powerful source of electricity already be available, optical concentrating arrays aren’t suitable to smaller-scale or off-the-grid applications, such as sterilizing medical instruments in a clinic in the developing world.
Now MIT’s Gang Chen, George Ni, and their colleagues have demonstrated a different approach: concentrating not the Sun’s light but its heat.1 Because their steam generator consists entirely of commonly available materials—a conscious choice on their part—they estimate that per unit area, it could be built for just 1–3% of the cost of an array of motorized mirrors.
Heat trap
The device is sketched in figure 1. It works by absorbing solar energy over a large area but giving it nowhere to escape except through a small slot where the absorber is in contact with a reservoir of ambient-temperature water. If the absorption area is large enough and the contact area is small enough, the water is locally brought to a boil to release steam before the heat can diffuse out into the bulk liquid. The challenge, then, is to keep the absorber from losing too much heat to conduction, convection, and radiation. Normally—and not unfortunately—those losses prevent any object heated by unconcentrated sunlight from getting anywhere near 100 °C.
Figure 1. A three-layer steam generator consists of a selective absorber insulated above with bubble wrap and below with polystyrene foam. Because conductive, convective, and radiative losses are suppressed, most of the solar heat captured by the absorber is channeled to a small slot where the absorber is in contact with water. (Adapted from ref. 1.)
Figure 1. A three-layer steam generator consists of a selective absorber insulated above with bubble wrap and below with polystyrene foam. Because conductive, convective, and radiative losses are suppressed, most of the solar heat captured by the absorber is channeled to a small slot where the absorber is in contact with water. (Adapted from ref. 1.)
To limit conductive and convective losses, the researchers insulated the top and bottom of the absorbing layer. For the bottom layer, they used ordinary polystyrene foam, which also kept the device afloat. The choice of top layer was a bit more constrained, because they needed something optically transparent. So they tried bubble wrap. “I was surprised by how well the bubble wrap worked,” said Ni. “Most researchers are using high-performance materials, and here we were, testing out bubble wrap, which wasn’t designed for maximum optical clarity.” Indeed, the bubble wrap transmits only 80% of the light that hits it. But its insulation benefits far outweighed that modest optical inefficiency.
The radiative losses depend on the absorbing material itself. A natural first choice of absorber would be a blackbody, which absorbs strongly and indiscriminately across the electromagnetic spectrum. But absorption strength at any given wavelength goes hand in hand with emissivity at the same wavelength. Blackbody absorbers are good at capturing optical energy, but they lose a lot of that energy through thermal emission in the IR.
The solution is to use a selective absorber, designed to absorb strongly at visible wavelengths but neither absorb nor emit in the IR. There are several ways such an absorption spectrum can be engineered, including with layered materials, photonic crystals, and textured metal surfaces. But the most commercial success so far has come from ceramic–metal composites, or cermets, composed of plasmonic metal nanoparticles in a dielectric matrix; nanoparticles are known to be strong absorbers at wavelength ranges that depend on their size and shape. Cermet selective absorbers have been on the market for decades for use in solar home water-heating systems. The one Chen and colleagues used for their steam generator emits 93% less thermal radiation than a blackbody does.
Getting steamed
For the past few years, researchers have been pursuing various ways of using materials to concentrate solar heat and generate 100 °C steam from water that remains cool in bulk. Although Chen and colleagues’ new work is the first to use unconcentrated sunlight, others have reduced the necessary concentration from a factor of 1000 or more, which typically requires an expensive array of mirrors, to a factor of 10 or less, which can be achieved with a single inexpensive lens.
Naomi Halas, Peter Nordlander, and their colleagues at Rice University, for example, developed an approach based on nanoparticles dispersed in the water.2 The nanoparticles efficiently capture solar energy, convert it locally to heat, and produce steam, all while the bulk liquid is heated only a little. The nanoparticles are neither damaged by the process nor carried away with the steam, so the method can be used for distillation. The Rice researchers developed their concept into a working autoclave for off-the-grid sterilization.3
The first foray by Chen and his group into solar steam generation used a double-layer foam structure floating in a beaker of water.4 They designed the top layer to be optically absorbing and the bottom to be thermally insulating. Water was carried up through the pores of the foam and was heated by the top layer. A 10-fold concentration of sunlight was enough to boil off about one gram of water per hour for each square centimeter of the absorber surface.
In the new work, Chen and colleagues’ proof-of-principle steam generator was a disk 10 cm in diameter. To tune the degree of thermal concentration, the researchers varied the size and number of evaporation slots: By cutting a single slot 1 mm wide and 7.5 mm long, for example, they concentrated the heat by a factor of 1000. In that case, the absorber reached 100 °C after about 5 minutes of exposure to direct sunlight. It boiled off about 5 g of steam per hour.
Because the device heats up so quickly, it works even when sunlight is intermittent. Figure 2 shows data taken on the roof of Chen’s MIT lab on one such day in August 2015. The solar flux, shown in green, varies widely as clouds pass in front of the Sun. The absorber temperature, shown in red for the same period, gets close to 100 °C several times, even though most periods of full sun last less than five minutes.
Figure 2. Sunshine on a cloudy day can be enough to bring water nearly to its boiling point, as shown here by the time-varying solar flux and absorber temperature for the same period. (Adapted from ref. 1.)
Figure 2. Sunshine on a cloudy day can be enough to bring water nearly to its boiling point, as shown here by the time-varying solar flux and absorber temperature for the same period. (Adapted from ref. 1.)
Bought in bulk, the materials that make up the device cost a total of $6/m2—which Chen and colleagues estimate could be reduced to $2/m2—compared with $200/m2 for an array of sunlight-concentrating mirrors. As the researchers work to implement their process on a larger scale, the challenges they face will depend on the application they target. So far they’ve worked only with clean, fresh water, for example, so they don’t yet know if the device will work for desalination or water purification without getting clogged with salt or contaminants. And some applications—such as electrical power generation, which requires large amounts of pressurized steam—might not be feasible at all. But, says Ni, “we can potentially apply the concepts we’ve demonstrated in another way.”
