Is This Thermodynamically Possible?

Much as I hate to interrupt the gloom with unexpected good news, an optimistic report on Gizmodo is worth some consideration. My trouble is that it puts me in mind of Maxwell’s demon.

The Reverse Osmosis process works on a simple principle: molecules within a liquid will flow across a semipermeable membrane from areas of higher concentration to lower until both sides reach an equilibrium.But that same membrane can act as a filter for large molecules and ions if outside pressure is applied to one side of the system. For desalinization, the process typically employs a sheet of thin-film composite (TFC) membrane which is made from an active thin-film layer of polyimide stacked on a porous layer of polysulfone. The problem with these membranes is that their thickness requires the presence of large amounts of pressure (and energy) to press water through them.

Lockheed Martin’s Perforene, on the other hand, is made from single atom-thick sheets of graphene. Because the sheets are so thin, water flows through them far more easily than through a conventional TFC. Filters made through the Perforene process would incorporate filtering holes just 100 nm in diameter—large enough to let water molecules through but small enough to capture dissolved salts. It looks a bit like chicken wire when viewed under a microscope, John Stetson, the Lockheed engineer credited with its invention, told Reuters. But ounce for ounce, its 1000 times stronger than steel.

“It’s 500 times thinner than the best filter on the market today and a thousand times stronger,” Stetson explained to Reuters. “The energy that’s required and the pressure that’s required to filter salt is approximately 100 times less.”

This potentially limitless supply of fresh water could do more than Heartland Institute ever could to put Peter Gleick out of business…


  1. give it a regular flush with fresher water to get rid of salt accumulation on the surface and could be. takes energy of course. possibly by two inlets for ocean water. 3 valves working in sync per membrane to control where the saltiest water goes. i know this is possible with larger molecules such as proteins but with salts? no idea.

  2. Something is not right here, possibly just in the reporting. The pore diameter looked way too large to me so I dug a bit. According to this report ( reverse osmosis requires nominal pore diameters in the range of 0.1 to 1.1 nm, not 100 nm as claimed in the Gizmodo article.

  3. Reverse osmosis requires energy to (a) overcome the osmotic pressure, and (b) overcome dissipative losses in the system. The first part is entirely thermodynamic - for seawater the osmotic pressure is about 27atm, which translates to 0.75kWh of energy per cubic metre of seawater. The second part will depend on the setup.

    A graphene membrane could significantly reduce the dissipative losses, but there's no chance of it reducing energy input by 99% unless the previous systems were wasting more than 75kWh per cubic metre on overcoming friction. (R.O. plants can desalinate sea water at under $1 per cubic metre, so that is clearly not the case.)

    Your concerns over the thermodynamics appear to be valid - these are misleading claims, unfortunately

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.