Air accumulation in CO2 reactor

Hi all,

I have a CO2 reactor setup in my 55gal freshwater tank. I have a
powerhead pulling water in through 4 foam filters (providing some
biological filtration), exiting the powerhead into a 1/2" tube going
across to the other side of the tank. At the end of that tube is a 1
litre soda bottle reactor. CO2 is piped in the bottom through a wooden
airstone and the stream of water coming from the top churns it up. A
piece of coarse filter media is at the bottom of the soda bottle.

Here's the problem... air (or some gas other than injected CO2) seems
to accumulate in the reactor. It's not excess injected CO2, because it
accumulates even when the CO2 is not hooked up. The air seems to
accumulate throughout the day, then dissipate at night. Early in the
AM, the reactor is empty save for the stream of water and a few CO2
bubbles (if the CO2 is actually connected). By late afternoon the
reactor is about 1/3 full of air. By the next morning, the air is gone
again.

I don't think it's a stretch to suggest the light (4 T5's currently)
may have something to do with it.

If someone can offer some insight into what may be happening and, even
better, what to do about it, I would be most grateful.

- Joe Barta