Here's a nicer writeup of this development in the University of Texas at Austin's own "Texas Science News":
http://web5.cns.utexas.edu/news/2013/06/desalting-the-ocean/
Thursday, June 27th, 2013
It's a nice start, but they have a few hurdles left:
Thus far Crooks and his colleagues have achieved 25 percent desalination. Although drinking water requires 99 percent desalination, they are confident that goal can be achieved.
OK, maybe they can put these chips in series, desalinate the desalinated water...
The other major challenge is to scale up the process. Right now the microchannels, about the size of a human hair, produce about 40 nanoliters of desalted water per minute. To make this technique practical for individual or communal use, a device would have to produce liters of water per day. The authors are confident that this can be achieved as well.
At least the authors are confident...
40 nanoliters of 25% desalinated water per minute.
Let's assume they get it working to 99% desalination. “This was a proof of principle,” said Knust. “We’ve made comparable performance improvements while developing other applications based on the formation of an ion depletion zone. That suggests that 99 percent desalination is not beyond our reach.” OK, let's say that's doable - and doesn't slow down the process too much.
Now, the speed - there are 1440 minutes per day.
"nanoliter" is 1 billionth of a liter.
A billion of these 'water chips' could do 40 liters per minute, or 57,600 liters per day.
Divide by 1000 - a million of these chips could give 57.6 liters per day.
Let's say 100,000 chips - that would give 5.76 liters/day. About 1.5 gallons of water.
That's still a lot of store-bought batteries.
If they could make each chip desalinate 99%, and 10 times faster, that's still 10,000 chips needed working 24 hours to produce 5.76 liters of potable water. How much power ? How long before the chips clog or degrade ? Did going from 25% to 99% desalination make the chips slower, not faster ? How reliable is the tubing taking seawater (already de-sedimented) into thousands of tiny human-hair-width microchannels ?
Most current methods for desalinating water rely on expensive and easily contaminated membranes. The membrane-free method we’ve developed still needs to be refined and scaled up, but if we can succeed at that, then one day it might be possible to provide fresh water on a massive scale using a simple, even portable, system.”
True - I like that the approach involves no vaporization/condensation or membranes.
still needs to be refined and scaled up - yes, and let's hope it is still affordable when it is.
Good luck. 
Edit history
= new reply since forum marked as read
