The main limitation that can be pointed out in the mixed methods studies is that data from different perspectives (qualitative and quantitative) are not always fully integrated but just combined so that potential sources of bias are not directly addressed, and the potential of mixed methods is underutilized.
Can you say more about what you mean by bias and where it could occur in a mixed methods study? It would also help if you could be more specific about the research design you have in mind.