We experience a massive baseline drift, it more or less looks like a hill, when injecting human serum albumin purified from Pichia pastoris in a Q-TOF SYNAPT G2 HDMS (Waters). I have attached a spectrum. The buffer was exchanged from PBS to 50% acetonitrile, 0.1% formic acid for the measurement and the protein was diluted 1:50 before injection. So, why do we see this hill? Normally if we would have protein impurities we expect more peaks. With salt, we expect broader peaks. But with what can we expect this "baseline hill"?