I remember reading about a membrane years ago that would allow oxygen to pass through. But it required quite a large surface area to get enough oxygen for a person to breath, it might be suitable for an underwater dwelling. I don't know if anyone has improved on that.

In a different vein, yet another nanoscale device no one has come close to making is the respirocyte, a little machine you would inject into your blood stream in the zillions, when oxygen levels were normal it would store oxygen and when they dropped a bit, they would release it.

They are called Rebreathers and cost several thousand $$.
http://www.scuba.com/scuba-gear-157/131708/VR-Technology-Sentinel-Rebreather.html

While they don't act like a fish's gills to pull O2 from the water. They do recycle gases giving divers a much lighter pack and in some cases more diving time.

If we could create a device that pulls O2 from the water it probably wouldn't be able to do it fast enough to give use breathable air. It would probably have to be connected to like a Heart and Lunge bypass machine so it could transfer the oxygen directly into our blood stream. Not a very fun way to go diving.

Come to think of it we have devices like that already. Problem is they are the size of an 18 wheeler. They are used on nuclear subs. They use high voltage to break the water down to its base elements. The system stores the oxygen and dumps the hydrogen overboard. Wouldn't be able to get it down to the size of a backpack cause you need a good power source, a way to filter the water, storage tank for the oxygen and many pumps and control valves to make it all feasible. You would have to have a full tank of oxygen before you started your dive since the system wouldn't be able to keep up with how much you are using. Which is why i mention connecting the system directly to your circulatory system. Its a more efficient way of giving you the oxygen you need. But you don't want your lungs to stop working.

I'm thinking electrolysis might work. It involves using a lot of electricity to pull hydrogen and oxygen out of water. The problem is that it would require really high current electricity at a pretty high voltage (at least 12V or 20V) go get enough oxygen to breathe. Not to mention, the hydrogen and the fact that the leads in the water would corrode, AND salt buildup would probably remove conductivity.

Liquid breathing with perchlorethylene fluid has been done experimentally in human volunteers undergoing lung lavage procedures apparently, but breathing liquids in humans is very difficult, and it was extremely hard to get the flow through needed for oxygentation at anything but rest.

Steve