Reading an HPLC chromatogram: peaks and baselines

Reading an HPLC chromatogram: peaks and baselines starts with time, signal, and the baseline between peaks. The trace can support a finding about identity or amount, but its appearance alone cannot prove either.

Key takeaways

• Read the time and signal axes before judging a peak.
• Use retention time to support an identity, not to prove it.
• Check separation and peak shape before treating a signal as a separate compound.
• Apply the same baseline and integration rules when comparing peak areas.
• Use standards to link detector response to an amount.
• Review blank, standard, sample, and system checks before reporting results.

Read the axes before judging the peaks

High performance liquid chromatography (HPLC) separates parts of a liquid sample as they pass through a column. A chromatogram is a graph of detector response, the signal recorded by a detector, over time.

The x axis usually shows time after injection. The y axis shows the detector response, such as ultraviolet absorbance or another signal. That response is not a direct measure of amount unless the method links it to a known amount.

The baseline is the trace when no clear peak signal rises above it. Find the injection point and the baseline before calling a rise a sample peak. A peak is a feature that rises above or falls below that trace.

Review the whole run, not only its largest peak. Early or late signals may matter, and the tallest peak is not automatically the compound being measured.

Compare traces only when the method, detector settings, and integration rules match. Integration is the process of marking a peak's start and end, then calculating its size.

Use retention time as a clue, not proof

Retention time is the time from injection to the top of a peak. Compare it with a reference standard, a known sample of the compound, run under the same conditions.

A close match supports an identity but does not prove it. Different compounds can leave a column at similar times and form peaks in nearly the same place.

A blank run contains no sample compound. It helps show whether a signal comes from the solvent, injection, carryover from an earlier run, or the instrument system.

Compare the blank, sample, and reference standard. A peak that appears in the blank needs a different explanation from one that appears only in the sample.

Dead time is the time an unretained compound takes to pass through the column. The capacity factor, a measure of how much longer a compound stays than an unretained one, is calculated as (retention time minus dead time) divided by dead time.

Judge peak shape and separation with care

Two peaks are better separated when the valley between them comes close to the baseline. A visual check helps, but resolution, a measure of separation between peaks, gives a firmer basis for judgment.

A common calculation is twice the gap between the peak times, divided by the sum of their baseline widths. Resolution can help judge baseline separation.

Co elution means that two or more compounds leave the column at the same time or too close together to distinguish. A shoulder or double top may point to overlap, but injection or column effects can also change a peak's shape.

Check method suitability, meaning the tests that show whether the method meets its set limits. If overlap remains likely, test the separation or use a different analytical method before calling the feature a second compound.

Peak tailing has a long back edge. It can point to slow movement through the column or unwanted binding. Peak fronting has a steep back edge and can occur when too much sample reaches the column.

A bell shaped peak follows a smooth, near symmetric curve. For this shape, baseline width is measured in a consistent way so peak areas can be compared fairly.

By convention, the peak width at baseline is approximately 4σ, which includes 95.4% of the total peak area.

Source: LCGC International

Check the baseline and how peaks are drawn

A steady baseline makes small peaks easier to judge. Noise, or small unwanted changes in the signal, can come from air bubbles, changes in pump flow, a dirty detector cell, or electrical interference.

A rising or falling baseline can follow changes in the liquid mix or temperature. It is a reason to investigate the run, not proof that the sample is at fault.

Peak integration estimates where a peak starts and ends, then calculates its area. A different baseline choice can change that area, so use one clear rule across the traces being compared.

When a small feature could be a peak, shoulder, or noise, compare blank and repeat runs. If an analyst changes the integration by hand, the record should state what changed and why, especially when peaks are split or joined.

Choose peak area or height for measurement

Peak area adds the detector signal across the whole peak. It is usually preferred for measuring amount because small changes in peak width affect peak height more.

Peak height can work for narrow, well shaped peaks. Use it for comparisons only when peak shape and width stay consistent.

A calibration curve or known standards link detector response to an amount. Peak area alone does not give the mass of a compound.

Area percentage is one integrated peak's area divided by the total area counted. It does not prove mass purity, and it leaves out compounds the detector does not see.

Different compounds can produce different detector responses. Compare peak areas as amounts only when a suitable calibration accounts for differences in detector response.

Review the trace before reporting a result

Start with the blank, then review the reference standard, the sample, and the system suitability results. These checks test different parts of the run.

Compare repeat injections for baseline, peak shape, retention time, and integration. A change in peak size or an extra signal is a reason to investigate, not proof of an impurity or instrument fault.

If overlap could affect the result, test the method's separation or use a second analytical technique. Do not rely on the trace's appearance alone.

A certificate of analysis is a batch report of test results. For related help with that document, see how to read a certificate of analysis.

Conclusion

Reading an HPLC chromatogram: peaks and baselines works best as a set of checks, not a quick scan for the tallest signal. Read the axes, test peak identity and separation, inspect the baseline, and apply consistent integration rules.

Report what the trace shows, then separate that observation from what it cannot prove. Use standards, blanks, and suitable checks before drawing conclusions about identity, amount, or purity.