Dear  Arusha, the cause of air cavities is  local overheating that is following by light beam scattering. It means that You need to decrease the local heating of aqueous protein solution. For that You have two  choices: a) to apply defocusing of IR laser beam b)  to make  the closed circulation of solution  at optically homogeneous laminar stream (without light  scattering). But if You need use  the high intensive laser light  in your experiment without above mentioned  methods  You should  apply to  picosecond  laser..

Wish You success

Thanks Eugene for your suggestions. Actually my aim is to locally heat up the solution by about 10 degree centigrade. I did not have a major problem when I was using D2O instead of water with the same laser alignment, but I checked the solubility of air is not very different in those two. I will try your suggestions.

Sparge with a helium flow. Helium will cause all of the air to be removed from the liquid. I use this technique to prepare my HPLC solvets and it works great.

Thanks Ross for your suggestion. I can definitely try that.

You can detect oxygen electrochemically and those who want to remove it simply bubble  N2 through it until O2 is no longer detected.  So three short things:  1) unless N2 is passed over the sample,  O2 gradually will replace the N2 (or He for that matter due to O2 partial pressure.  2) For those who wanted to increase O2 in water, was thought that the bubbles increases O2 solubility; turned out to be a false assumption: bubbles worked by perturbing the surface tension of the water which then allowed O2 via partial pressure to maximize.   So shaking your aqueous sample even mildly will increase, not decrease O2 levels. 3) Raising temperature will decrease O2 solubility, i.e. as longs as temperature is not perfectly uniform [localized],  O2 levels most likely will be localized as well.  Aqueous solutions have very large air pockets actually.  Can do a very simple experiment.  Mix  70 ml of  95% ethanol with 30 ml of water.  Huge amounts of bubbles arise from the two liquids.  30% water in ethanol has lost the 3D structure that hold gas. Raising the temperature of aqueous solutions just makes the 3D packing gradually more tight.   The size and shape of the packing most probably is not altered when He or N2 replaces the air that is present .    .      .