I have always wanted to understand how desalination works. I actually owned an RO filter when I first started keeping fish, but quickly learned that it wasn’t good for the species I kept, so gave it to a friend who kept discus.
I know that there are semi-permeable membranes involved, and that they are mostly polymeric these days, but used to be acetate.
In short, I know WHAT happens, but not why it works. How does flowing a solution through a membrane filter out the salt/minerals? Is it just a mechanical process? Like these membranes are so perfectly punctured that the water molecules fit through but the others don’t?
And if the water is (in most cases) pushed through under pressure, why wouldn’t they just clog up immediately and either stop working or tear bigger holes through which the salt would flow?
I am also given to (not) understand that some of these membranes are penetrated not due to pressure, but to a “drawing” solution which draws the water through but not the salt/minerals. . . Huh?!? How on earth does THAT work? And how does one then get the water out of a solution which draws it so powerfully?
Help me out with this, Dopers! Any explanations are appreciated, as are links to good ones that may already exist. My (usually excellent) google-fu has turned up a whole lotta “what happens” and almost no “why it works.”
It’s essentially pores that are big enough for water but not sodium, potassium, and their associated hydration spheres. Molecules in solution move a lot. They’re not like particles. There isn’t really a mechanism for clogging by solute. Any concentration gradient will quickly diffuse away.
Draw solutions are used for forward osmosis. For example you could have seawater on one side and concentrated ammonium carbonate on the other. Water will pass through the membrane to equalize osmotic pressure. You can then heat the ammonium carbonate to remove ammonia and carbon dioxide.
Note that if you put saltwater one side and fresh water on the other, the fresh water will naturally flow into the saltwater. That’s plain old osmosis.
You need to apply a lot of pressure from the salt water side to overcome this (and get a good flow rate) and to force the water to flow the opposite direction. Some of the pressure on the other side can be recovered to save energy in applying the pressure.
Not all the pressured salt water goes thru the membrane. Some is drawn off as brackish water to assist in keeping the membrane clear.
The process of osmosis is analogous to partial pressure - that is, the solvent flows from the weak solution to the strong one because of osmotic pressure
I think when you look at this on a particle level, it’s the same thing that makes pressure and evaporation work - the particles are jumping around, and pressure (or heat) makes them more likely to jump in one direction than the other - so solvent particles jumping through from the weak to the strong solution have an easier time than particles jumping in the opposite direction.
Increasing the physical pressure on the strong-solution side changes the balance - so it’s like filtration, but it’s not actually filtration.
We can take these statistical behaviours of particles to weird extremes - when you open the neck of a balloon, there’s a nonzero chance that the air particles outside will all happen to move in, further inflating it - but the statistical behaviour of the gas particles means it’s overwhelmingly more likely that the ones on the inside will find their way out.
How are polymers chosen to give the desired selective permeability?
I’m familiar with how evolution does it: proteins like aquaporin form water-molecule sized pores in lipid membranes. Those pores get their selectivity by having just the right arrangement of hydrophillic, hydrophobic, and charged residues, which do not favorably interact with common charged solutes.
Certainly polymers can be made with almost any desired chemical properties, but AFAIK they don’t form precise and complicated structures like proteins.
So what have the polymer chemists done to make good selective membranes?
Any Cape Coral residents here? My daughter just moved down there and said the drinking water is horrible and attributes it to the RO system. She also claims the RO water is behind the dearth of good pizza in SW Florida. Bad water = bad crusts.
It could be done this way…just boil the saltwater, collect the steam, condense it to get fresh water, and dump the concentrated brine back in the ocean.
But that obviously takes a lot of energy to boil the water, and reverse osmosis is just energetically more efficient.
Then run the steam through a turbine on the way to the condenser and you’ve got a power plant that returns some of the energy necessary to boil the water.
There are extensions of this idea like multi-stage flash. But reverse osmosis is generally the most energy efficient. But if you have a free source of heat (waste heat from a power plant, geothermal, etc.) then the numbers might add up right.
A lot of the economics gets dominated by where you are in the world. Some of the people most in need of fresh water live in areas where the costs to bring in and maintain this sort of tech are quite high.
You might want a closed loop for a steam turbine; that water is often RO’d. They also work best at much higher temperature and pressure. For water purification, you have to leave much of the water behind so that the salts are carried away. Heating that extra water has a cost.
So your choices are run a shitty turbine or unnecessarily heat a bunch of water.
But if you have ok-quality heat available, multi-stage flash, membrane distillation, or something else could still work.
I’m not up to speed with the membranes in question, but you can add pore-formers (e.g. bicarb + acid), make block copolymers that phase-separate, coat a fibrous mat. Probably some other options. The polymer work I do have some exposure to has a trial and error component to it, despite the PIs all insisting they can engineer everything perfectly.
The problem with municipal RO water is that they will only desalinate it to a base level quality. So it may remain with a salt and other dissolved solids content that is not so much unpalatable, but still enough to affect the taste.
RO systems used domestically and commercially can deliver very pure water. Indeed too pure for some applications. An example I am familiar with is coffee. Using RO water in an espresso machine results in a brew that seems to extract a bit too much of the wrong stuff. The trick is the harden the water a tiny bit. The usual way of doing this is to add another unit to the RO systems output that is little more than a cylinder full of crushed calcite. This will add a small amount of calcium back into the water, enough to make the water produce really good coffee.
Talk to brewers and they are very concerned about water. Huge brewery near me uses a massive RO plant to make the water they use. They produce 1.8 MOhm water - which is pretty seriously pure. (18 MOhm is as pure as it gets - that is evil stuff.)
I sort of doubt it is the RO water that affects pizza dough. There is salt in the dough, and by the time you are done, the water is not going to affect things, not unless other residuals are a problem. But like with beer - there are nuances that are difficult to define. (Guy I knew was a full professor of chemistry, and an enthusiastic home brewer. He would cheerfully admit that the water made a huge difference, and he had no clue as to why.)