That is awkward, but could you provide more info? Type of column? Are you injecting the pure solvent? Is it a treated sample? What type of ionization mode are you using? Are there other ions in the same peaks? How is the peak shape? What is the carrier gas and how do you obtain it?
I have used Agilent DB 5MS UI 30m column. Ester-quart sample were extracted to gc grade hexane and pass through sodium sulfate bed. EI mode was used to ionization. Two fronting peak were appeared and library match as ammonia and water.
Sorry, but at this point I don't have a better guess than a possible leak in your MS system and lot's of ammonia in the environment air. But this should cause a systematic noise.....
The best advice I can give is do systematic trouble shooting. Check for leaks, trick the instrument with a blank manual sample, inject directly the hexane, do a treatment blank with hexane, and extract a blank sample. Compare everything and you may be able to get more clues.
Also, maybe someone else might give better insight. Good luck!