Question: The baseline noise in my GC-MS suddenly increased, and I was told the ion source was dirty. After cleaning the source and replacing the filaments, the noise was slightly reduced but still excessively high. Then I checked for leaks, cleaned the injector, replaced the column and traps, and installed a new gas cylinder, but the noise was still high. Finally I disassembled and rinsed the transfer line, which solved the problem. The column passes through the heated transfer line and

more »

... tes at the source. Since no sample comes in contact with the transfer line, how did it become dirty? Answer: Probably one or more small column ferrule particles became trapped in the transfer line. The ferrule particle was probably generated during column installation. When the transfer line column nut was tightened, a small piece of the ferrule was broken off. This typically occurrs when the column nut is overtightened. Some transfer line designs are prone to producing ferrule particles upon overtightening of the column nut. The vacuum of the mass spectrometer pulls the ferrule particle into the body of the transfer line. Depending on the design of the transfer line, the ferrule particle then becomes lodged in the line, trapped at the end of the line, or passes completely through and enters the ion source. Column ferrules contain graphite, which is a very good absorbent. Compounds absorbed by the graphite slowly desorb when the ferrule is heated, especially under vacuum. These released compounds are swept into the mass spectrometer, thus producing a response. Because the ferrule impurities are released on a somewhat constant basis, the overall effect is a rise in the baseline and/or an increase in baseline noise. Over time, the ferrule impurities often become depleted, and the background signal may improve. Sometimes the background signal remains high until the responsible particle is removed. If the ferrule particle enters the ion source, the noise level usually remains high until source cleaning removes the particle. It is possible, but more difficult and less likely, to get a ferrule particle trapped in a standard GC detector.