I will elaborate my protocol. In order to maintain completely anaerobic conditions in a reactor run, I sparge absolute nitrogen into the reactor so as to displace air in the headspace. In earlier reports, researchers have used absolute nitrogen for this and hence I have been following the same. However, some of the recent reports mention a combination of 95% N2 and 5% CO2 for this purpose, instead of absolute N2. Hence, I wanted to know why one is preferred over the other.

I agree N2 is not used by fermenters. N2 is only used to displace O2 from the reactor so as to maintain anaerobic conditions. Some people use pure Nitrogen for this, while others use a combination of N2 and CO2

The only reason I can think of is to (partially) simulate the effect of depletion of O2 and 1:1 substitution of CO2, may pH be consideration?

...or has it something to with the differences in density of gases.  CO2 is much heavier than N2.  Perhaps a clue to this question may be found here:

http://www.wineqc.com/papers/inertgas/inertgas.html

In the papers, no explanation is given.

Thank you so much for the clues. I read up the link sent by Susantha and I feel this could be a very plausible explanation. CO2 being heavier forms a blanket over the liquid medium thus ensuring complete anaerobic conditions. But of course, one cannot use only CO2, as it will add to pH changes. Also, since CO2 is much more soluble in media (containing water), I feel even the medium becomes scrubbed off dissolved oxygen. Thanks Susantha for that link.

My guess is that the use of N2/CO2 mixtures started either because someone was working with autotrophs or because the use of CO2 was a convenient way to adjust the pH. But, I am only speculating. Perhaps you could send an e-mail to one of the corresponding authors of the PLOS Genetics paper and ask them why they chose a N2/CO2 mixture instead of pure N2.

Also, I am skeptical that the density of the gases has anything to do with their use in this situation. Gases in the atmosphere don't separate by density. Gases in high-pressure gas cylinders don't separate by density. I can't imagine why gases in the headspace of an incubation vessel would separate by density. Molecular diffusion (and probably convection as well) should keep the headspace well mixed.

I had written to the authors, but I have not got a consensus answer. Some say, CO2 is added so that the enzyme PEP carboxylase is not limited by the dissolved CO2 content available in the medium. Another answer I got is that if CO2 is less, then it can be limiting for Fatty acid metabolism when cells grow anaerobically.

We used  N2/CO2 mixtures  because CO2 was a convenient way to adjust the pH in Ringer's solution.