I don't think we can have an accurate "theoretical minimum mass" for information because it is highly dependent on our (incomplete) understanding of small particles.