Achieving a flat baseline which does not exhibit spikes, ghost peaks, drift or wander in an unpredictable manner should be a primary goal when performing HPLC analysis or developing methods. Methods which result in flat baselines and have well defined, sharp peaks allow for accurate sample area integration. Integration algorithms perform poorly in quantifying peaks on sloped, drifting or noisy baselines. Excessive baseline noise contributes too many problems, including poor quantitation, high %RSD errors, peak identification errors, retention time variation and many other critical problems. Properly developed HPLC methods are reproducible methods which apply and utilize good chromatography fundamentals.
In this article, we will discuss how temperature fluctuations, inadequate mixing, inadequate degassing and flow cell contamination can result in excessive baseline noise. We will provide suggestions on how to reduce or eliminate these problems.