A lot of folks get skeptical when they read about a new technological innovation that's promising to save the world or whatever. I can sympathize with that - it's the goal of every start-up out there to get media attention. They succeed perhaps more often than would be merited. (And if their pitch isn't sensationalist enough, journalists will often put a sensationalist spin on it themselves).
The stuff that gets media attention runs the full gamut from "production-ready" to "works in the lab" to "a nice theory" to "pure crackpottery". (I sometimes wonder whether I shouldn't try to make a buck as a consultant for investors in these things - as scientific merit and investment don't seem to be highly related.)
But I was quite impressed by a story in The Economist on a new idea for desalination, coming from a company named Saltworks Technologies (site). It's not often this kind of 'press-release journalism' has that effect on a jaded chemist (and at least on paper, chemical engineer) such as myself. But this idea seems to be just plain clever. I'll explain why below the fold.
The fundamental problem of desalination - the solution's the problem.
I'm going to take the risk of giving a small lesson in chemical thermodynamics here. Before anyone runs away screaming in horror: I'll try to keep the difficult stuff and equations out of it.
The problem with desalination in general is that it's simply difficult. 70% of the Earth's surface is covered with water and 97% of that water is salty. Yet, as many people know, salt was a valuable commodity until modern times. (cf. the origin of the word 'salary') That's not a coincidence.
Salt is sodium chloride (NaCl), which when dissolved in water separates into a positively-charged sodium ion (Na+) and a negatively charged chloride ion (Cl-). Thermodynamically, it's energetically favorable for it to do so (it goes to a lower energy state). Basic chemistry says that the fact that salt is very water-soluble (up to about 35% by weight in water) means it loses a lot of energy.
So conversely, basic chemistry/thermodynamics says that removing salt from water requires a lot of energy. On top of that, it's not really a technically simple thing to do either. Sodium and chloride ions don't react with many things. They don't form insoluble compounds with many other substances (or you could add that substance and filter out the solid precipitate).
An alternate method mentioned in the article are evaporation - removing the water from the solution. Which is costly - because you're not only 'paying' (in energy) for the cost of removing the salt. You're also paying the larger additional 'cost' of turning water into steam. (The boiling point of water is 100 C. The boiling point of seawater is about 104 C. So as you can tell, it costs more energy to remove water from salty water - but relatively little compared to the cost of vaporizing water in general)
Another alternate method mentioned is reverse osmosis; which is a sort-of fancy word for 'filtering out the salt'. Now, besides the energy you 'get' from dissolving salt in water, you have an energy corresponding to the amount of salt dissolved in water. If you have a 'filter' with salty water on one side, and less salty water on the other, that the salt is not going to voluntarily (spontaneously is the adverb chemists use) flow from the less salty water to the more salty water, any more than heat is going to flow from a cold object to a hot one - and for the same reason.
(The Nernst equation describes the basic relationship between energy-concentration, although the calculations you'd need to do to properly describe the Saltworks process are a bit trickier)
So if you want to filter out the salt, you're going to need to make up for the energy difference, in other words, force it through the filter. That's supplied in the form of pressure. You simply need to push it through the filter.
It turns out this requires pretty high pressures, and correspondingly, a lot of energy. On top of that it's long been a major technical obstacle to create 'filters' that can tolerate the high water pressures needed.
This tech
As explained, salt will spontaneously flow from a solution with high concentration to a solution with low concentration - because there's an energetic benefit in doing so. What the Saltworks guys have hit upon, is that you can use that energy - directly - to lower the concentration of salt in water. How, you ask? Well, that's the clever thing.
Here's a picture of their cycle, slightly adapted from the one in the Economist article:
The key here are the 'ion bridges'. What's that? Concretely, an ion bridge is a tube filled with some material that allows ions of a certain charge (positive or negative) to flow freely through it (usually in either direction). But I'll get to the details of that in a second.
Starting at the bottom, the idea here is that you use the sun to evaporate water from your seawater and produce brine - water with an even higher concentration of salt. Now, you connect your brine to two reservoirs of sea water with ion bridges, one for the sodium ions (Na+) and one for the chloride ions (Cl-).
Now given that (the bottom half only), they won't flow. Even though the concentrations of sodium and chloride are lower in the seawater than in the brine. The reason for that, besides the fact that nature 'wants' to even out the concentrations, it also 'wants' to keep the charges neutral. If the ions of each respective type started flowing to one and/or the other container, it'd end up becoming electrically charged! Which is very energetically unfavorable, even more so than the concentration difference.
What they do then, is connect our desalination stream (the top water reservoir) to these sea-water reservoirs with the opposite ion bridges. So, if a sodium or potassium ion flows from the brine to a seawater reservoir, the corresponding ion of opposite charge has to enter the seawater from our desalination stream - where the concentration is actually lower.
(Sidenote: Our bodies are powered by a similar difference in hydrogen ion concentrations, inside the mitochondria of our cells, it's used to power the production of ATP - the molecule that in turn powers almost everything else in our cells. Figuring out how this works garnered the 1997 Nobel prize in chemistry. A future Nobel is likely in waiting for the one to figure out exactly how the 'ion pump', cytochrome c oxidase, that creates this concentration difference works.)
This will happen as long as the brine's concentration is sufficiently high relative the seawater, compared to the seawater's concentration relative the desalination stream. In other words, the energy gained by going from the high-concentration brine to the mid-concentration seawater has to be bigger than the energy lost by going from the desalination stream to the seawater.
This is pretty basic chemistry, but if you ask me, it's very clever. You are now in effect using the brine as the source of energy for your desalination. Given that you have enough brine in high enough concentration, no additional energy is required (apart from the relatively small amounts required to move water around).
Now, about those ion bridges. Which would really seem to be the biggest technical hurdle here. I don't think they're actually that big a problem: It's more or less existing technology. What you do (and what it seems they've done), is fill a tube with a gel (or gel-like substance) which only lets a certain ion through. Often a polymer (i.e. a 'plastic'), a big chain of a molecule that has a lot of positively or negatively charged bits attached to it. If the polymer had lots of negative groups (for instance), then the negatively charged chloride ions wouldn't want to get near it (since equal charges repel). The positively charged sodium ions would enter and bind to the negative groups of the polymer. By moving from group to group, they'd be able to make their way through the substance. (the polymer itself stays put) These kinds of polymer gels are used in cell phone batteries nowadays, but have for a long time been used for other chemical-engineering applications (e.g. ion-exchange columns).
So, while I'm sure the folks behind this are hard at work finding the best kind of ion bridges for their technology, rest assured that it's mostly an engineering problem.
Efficiency
The important question, you'd think is: What about the efficiency?
Well to quote Homer Simpson: "In this house, we obey the laws of thermodynamics!". It's easily predicted that this process does not, in fact, use less energy in itself. In fact, given that the brine would be produced by evaporation (which I'll reiterate, requires substantially more energy than that involved in raising the concentration), it's virtually guaranteed to be rather inefficient in terms of total energy.
But therein lies much the cleverness of the whole thing! Evaporating water and condensing it to get desalinated water costs energy - 'real' energy, both from the evaporation itself, but also in terms of electricity to run condensers. If you'd use solar energy for the evaporating, you'd need a big structure to collect all the vapor, etc. Using reverse osmosis uses lots of electricity energy to achieve the high pressures used.
This technique, on the other hand, uses very little 'real' energy. Just what's needed to move the water around. In addition, since you're not interested in collecting the vapor, you can produce your brine in the simplest of circumstances - releasing seawater into an open reservoir. Which requires little cost or energy. Efficiency isn't a big deal if your energy is 'free'.
Some final thoughts
The things that really inspires my imagination are the possibilities not directly mentioned in the Economist article. Obviously, the big thing here is the many possibilities to produce brine in sunny, arid places - those which typically are in greatest need of water from desalination. For instance, you could flood a low-lying desert area with seawater or create an artificial lagoon by walling off a shallow area of the sea. Or make use of an existing natural lagoon, such as the Garabogazköl (Turkmenistan). Or make use of existing sources of brine that are near the sea, such as the Dead Sea, or Lake Assal in Djibouti.
It'll be interesting to see if this technology takes off. I can't see why it wouldn't at the very least be able to find specific applications. And as always with clever ideas, I'm left wondering: Why didn't I think of that?