Peak integration in practice

Understanding how much of a reported purity figure is the sample and how much is the integration.

Purpose

Understanding how much of a reported purity figure is the sample and how much is the integration.

This page exists so that a discussion about a number can start from a shared understanding of what produced it. It is written for a reader who is technically literate but not a practising analyst.

How it works

Integration is the step where a continuous detector trace becomes a set of numbers, and it involves judgement. Where the baseline is drawn under a peak, whether a shoulder is skimmed tangentially or split valley-to-valley, and whether a small feature is integrated at all, all change the reported area percent.

On a clean chromatogram with well-resolved peaks the judgement is inconsequential. On a chromatogram with a trailing shoulder or a rising baseline it is not, and differences of one to two percentage points between defensible integrations are ordinary.

What determines the result

This is the main reason a purity figure without the underlying chromatogram is a weaker piece of evidence than it appears, and the main reason two competent laboratories can report different numbers on the same vial without either being wrong.

The practical response is not cynicism. It is to ask for the chromatogram, to compare relative retention times rather than absolute ones when comparing laboratories, and to treat differences under about two percentage points as uninformative unless you have the raw data.

Typical criteria

These are typical values, not universal standards. The criteria that apply to a specific result are the ones in the validated method that produced it, which is why a report without a method identifier is harder to interpret than one with.

What it cannot tell you

Every analytical method answers one question. The most common error in this community's discussions is treating an answer to one question as an answer to another: a mass match read as a purity result, a purity figure read as a content determination, a sterility test read as a guarantee of safety.

When you read a report, write down which question was asked. The rest follows.

Related pages

Other method notes in this commons cover the neighbouring techniques, and the certificate interpretation page covers how these results get reported. The hplc subcategory is where the discussion happens.