The Art of Noise, by Maria Ofitserova, Senior Research Chemist

Baseline noise is a common and often frustrating problem in HPLC analysis. It makes integration difficult and adversely affects reproducibility and sensitivity of analysis. The most common sources of baseline noise are from pumps and bubbles in the lines or detectors. Not all noisy baselines can be easily explained but understanding common noise patterns will help you to determine what part of your system has a problem.

The Sine Wave
To observe the noise pattern you need to find a portion of the chromatogram that does not have any peaks and zoom in to look at about 5-10 min of the baseline. Baseline noise caused by reciprocating HPLC or post-column pump looks like a fairly regular sine wave. This kind of noise is usually due to old/poorly installed seals or bad piston. Ups and downs in the baseline follow flow/pressure variations as the piston moves. The period of the sine wave is different for different pumps. Measure the interval (in seconds) between the maximums of two waves to determine which pump is causing the noise. Most HPLC pumps have an interval of 6 -13 seconds. Pickering PCX5200 and Vector PCX reagent pumps have 2 sec and 4 sec intervals respectively. Pinnacle PCX contains a syringe pump which moves the piston in a single stroke hence it does not produce sine wave noise.

Most HPLC software programs record the column pressure during the analysis. It is very helpful to look at the trace to check if pressure variations have a similar pattern to your baseline noise. Pinnacle PCX users can also take advantage of log files collected by the Pickering software. Reagent pump pressure recorded in the log file helps Pickering technical support to evaluate the performance of the post-column system and determine if the syringe pump needs maintenance.

Baseline noise caused by bubbles consists of random spikes of varying amplitudes. Bubbles can occur in solvent lines or in the detector flow cell and are often caused by solutions outgassing. To prevent this from happening use a properly working degasser and install a backpressure regulator on the detector waste line to prevent boiling and outgassing in the heated reactor.

Detector Noise
Detector noise is always present and can be visible even on a “good” baseline if you zoom in deep enough. It is random and looks about the same throughout the chromatogram. An old detector lamp, dirt in a flow cell or problems with electronics can greatly increase noise level. If detector noise is suspected make sure the flow cell is clean and check the lamp hours. Built-in detector tests are also useful in assessing detector performance.

Shooting in the Dark
A common mistake people make when troubleshooting baseline noise in post-column applications is turning off the post-column reagent pump. Noise in the baseline is essentially variations in signal so it is proportional to background signal. Common eluants don’t fluoresce or absorb light in the visible range so when eluants alone go through the detector there is no signal and hence no noise. Post-column reagents, on the other hand, are often either colored or have background fluorescence so elevated noise caused by any part of HPLC system becomes visible. Turning off the reagent pump is akin to turning off the detector lamp and taking a shot in the dark – the noise is still there but we just can’t see it.

Let Us Help
When contacting Pickering support about elevated baseline noise please be ready to fax or e-mail your chromatogram and zoomed in portion of the baseline. For Pinnacle PCX users sending the log files will also help us to find the problem. You can email or send a fax to 650-968-0749.