Empower Tip: Baseline Noise

In this tip for Empower Chromatography Data System (CDS), we will conclude our series on Baseline Noise.

Using one chromatogram, you can see how the technique used to measure noise will result in different values of noise reported. Over a relatively flat baseline region you would expect Baseline Noise and Peak to Peak Noise to be similar, and that Detector Noise, the RMS value, is considerably lower. Also, you can expect that the Average Peak to Peak Noise, and Average Detector Noise calculations will be somewhat smaller than their respective non-Averaged calculations. If we use Baseline Noise as our noise calculation of choice and set our LOD to be 3 times noise, then a peak having height of 3.75X10-5 AU would have a measured signal-to-noise value of 3. You can see how there would be some confusion if the same calculation were performed using one of the other noise techniques. For example, using Peak to Peak Noise for a 3.75X10-5 AU peak would calculate a signal-to-noise value of 2.18 (below the stated LOD of 3X signal-to-noise). Likewise, if you assume LOQ at 10 times noise using Detector Noise or Average Detector Noise as the noise calculation would be reporting quantified results for our peak at 3.75X10-5AU.

Measuring Noise on the flat portion of the chromatographic baseline, you can see the resulting signal-to-noise calculations for the three peaks.

Measuring Noise on a drifting portion of the chromatographic baseline, you can see the resulting signal-to-noise calculations for the three peaks produce different values.

Even though all the displayed signal-to-noise calculations show some improvement when noise is measured in a flat portion of the baseline as compared to that of a sloping portion of the baseline, none is as significantly different as Baseline Noise. If you wish to use Baseline Noise for the determination of signal-to-noise, it is recommended noise be measured on a non-sloping portion of the baseline so that baseline drift does not negatively impact the calculation. You can easily see that drift comes into play with all the calculations used. When comparing the Detector Noise calculation to the average Detector Noise calculation, they are not all that different when noise is measured on a flat portion of the baseline, but they are largely different when measured on the sloping portion of the baseline. You can make a similar comparison to the Peak to Peak and Average Peak to Peak calculations. Although it is impossible to eliminate drift from the calculation of chromatographic noise, your choice of algorithms for measuring noise can have an impact. Understanding the calculations and how different chromatographic anomalies impact the noise calculation is important to the calculation of signal-to-noise. All the calculations described in this tip are acceptable and are used for the measurement of noise. Remember that when selecting USP Noise or any of the other pharmacopeias, Empower defaults to Peak to Peak Noise for the signal-to-noise calculation.

It’s that easy!