How to optimize the baseline with a classic algorithm

The first choice when you want to optimize the peak integration is to change the baseline parameters. This section describes how to optimize the baseline with a classical algorithm.

The Classic algorithm searches for all parts of the source curve that are longer than a defined minimum baseline segment and fall within limiting parameters. Together, the parameter values define the limits for a rectangular box. A part of the source curve must fit entirely inside this rectangular box to be identified as a baseline segment.

The Classic algorithm is particularly useful when you need to integrate curves with negative peaks and when quantitative data from negative peaks are important.

The parameters for the Classic algorithm are:

• Shortest baseline segment

• Noise window

• Max baseline level

• Slope limit

See more information about the parameters below.

The table below describes how to set a Classic algorithm and define a baseline.

Note: The same settings can be edited in the Calculate Baseline dialog box when a new baseline is created. Choose Integrate:Calculate Baseline to open the dialog box.

The best way to optimize the baseline is to change the baseline parameters step by step and then check the resulting baseline after each change. When the desired effect is accomplished it is best to go back and try a parameter value in between the two last settings to avoid an unnecessarily low or high value.

How much the values should be changed depends on the cause of the peak integration problem. The table below is a general guideline.

Note: If necessary, click the Default button to restore the default values.

If a too high Shortest baseline segment value is set, short curve segments between peaks in the middle of the chromatogram are not identified as baseline segments. The calculated baseline does not follow the source curve, see below:

The Shortest baseline segment value is decreased by 50% in this example:

A changed Slope limit will often improve the baseline calculation. The Slope limit sets the maximum slope of the curve to define when a peak is recognized. A too high Slope limit will cause the up-slopes of the peaks to be recognized as baseline segments.

The example above was improved by the shorter baseline segments but the high slope of the short segments in the region between the second and the fourth peak still makes the baseline unacceptable. In the example below the Slope limit is increased by a factor of 2.5, which produces a correct baseline:

A too high Slope limit value can cause peak limits too high up on the peaks. This can be the case when the chromatogram includes a very large flow-through or solvent peak. The large peak affects the calculation of the default parameters and leads to too high values for the Slope limit.

Note: A too high value for the Noise window can have the same effect and be caused by the same situation, often also in combination with a high Slope limit.

Peak limits are defined on peaks in the example below due to the high Slope limit:

The example below has a much lower Slope limit, and a lower Noise window:

Sometimes you get too many peaks after the peak integration, usually because noise on the baseline is erroneously detected as peaks.

The solution to this is to increase the Noise window parameter. However, this can result in peak limits too high up on the peak slopes.

The illustration below is an example of noise detected as peaks (A) and the result of a second peak integration with an increased Noise window (B).

Note: You can also use the Reject peaks function in the Integrate dialog box to reduce the number of peaks based on the total number of accepted peaks or the minimum peak height.

Sometimes obvious peaks are not detected in the peak integration. The probable cause is that the Noise window is set too high. See the illustration below:

All peaks are detected if the Noise window is decreased, see example below:

Note: Missing peaks can also be caused by improper settings for Reject peaks in the Integrate dialog box, or Filter peaks in the Chromatogram layout dialog box.

In rare cases the top of a broad, flat peak can be incorporated as a baseline segment. This is one of the very few situations where it is useful to change the Max baseline level. The illustration below is an example:

The table below describes how to set the Max baseline level.

The illustration below is an example of a correct baseline after the Max baseline level has been changed: