New approach to water desalination

A new approach to desalination being developed by researchers at MIT and in Korea could lead to small, portable desalination units that could be powered by solar cells or batteries and could deliver enough fresh water to supply the needs of a family or small village. As an added bonus, the system would also remove many contaminants, viruses and bacteria at the same time.

The new approach, called ion concentration polarisation, is described in a paper by Postdoctoral Associate Sung Jae Kim and Associate Professor Jongyoon Han, both in MIT's Department of Electrical Engineering and Computer Science, and colleagues in Korea.

The system works at a microscopic scale, using fabrication methods developed for microfluidics devices - similar to the manufacture of microchips, but using materials such as silicone (synthetic rubber). Each individual device would only process minute amounts of water, but a large number of them - the researchers envision an array with 1,600 units fabricated on an 8-inch-diameter wafer - could produce about 15 litres of water per hour, enough to provide drinking water for several people. The whole unit could be self-contained and driven by gravity - salt water would be poured in at the top, and fresh water and concentrated brine collected from two outlets at the bottom.

That small size could actually be an advantage for some applications, Kim explains. For example, in an emergency situation like Haiti's earthquake aftermath, the delivery infrastructure to get fresh water to the people who need it was largely lacking, so small, portable units that individuals could carry would have been especially useful.

So far, the researchers have successfully tested a single unit, using seawater they collected from a Massachusetts beach. The water was then deliberately contaminated with small plastic particles, protein and human blood. The unit removed more than 99 percent of the salt and other contaminants. 'We clearly demonstrated that we can do it at the unit chip level,' says Kim.

While the amount of electricity required by this method is actually slightly more than for present large-scale methods such as reverse osmosis, there is no other method that can produce small-scale desalination with anywhere near this level of efficiency, the researchers say. If properly engineered, the proposed system would only use about as much power as a conventional lightbulb.

The basic principle that makes the system possible, called ion concentration polarisation, is a ubiquitous phenomenon that occurs near ion-selective materials (such as Nafion, often used in fuel cells) or electrodes, and this team and other researchers have been applying the phenomenon for other applications such as biomolecule preconcentration. This application to water purification has not been attempted before, however.

Potable water is often in high demand and short supply following a natural disaster like the Haiti earthquake or Hurricane Katrina. In both of those instances, the disaster zones were near the sea, but converting salty seawater to potable fresh water usually requires a large amount of dependable electrical power and large-scale desalination plants - neither of which were available in the disaster areas.

One of the leading desalination methods, called reverse osmosis, uses membranes that filter out the salt, but these require strong pumps to maintain the high pressure needed to push the water through the membrane, and are subject to fouling and blockage of the pores in the membrane by salt and contaminants. The new system separates salts and microbes from the water by electrostatically repelling them away from the ion-selective membrane in the system - so the flowing water never needs to pass through a membrane. That should eliminate the need for high pressure and the problems of fouling, the researchers say.

Having proved the principle in a single-unit device, Kim and Han plan to produce a 100-unit device to demonstrate the scaling-up of the process, followed by a 10,000-unit system. They expect it will take about two years before the system will be ready to develop as a product. 'After that,' says Kim, 'we'll know if it's possible' for this to work as a robust, portable system, 'and what problems might need to be worked on.'