I think you should describe your conditions in more detail if this is really of interest. or if not, try another process to go from the niitro group to the amino group;.
1. Was it a simple hydrogenation or did you use a metal (ie tin chloride) and what metal?
2. Was the experiment run under nitrogen and at what temperature (THF reflux?)?
3. What are the properties of the byproduct? Is it very dark or white crystals? Can you run an NMR to establish if all of the carbons or hydrogens are aromatic? Can you measure N content, mp?
The easiest explanation (although likely incorrect if no other analysis is done) would be to assume scenario 2 (C16H12O) and that the THF somehow coupled to the biphenyl along with loss of N and two H's. This would most consistent if a metal reductant was used. There would probably need to be an alcohol functional group (ie THF ring opening) to explain a higher polarity (presumably that is defined by TLC migration rates).