How to Read an HPLC Chromatogram on a Peptide COA
Most peptide COAs report purity as a single number, such as "99.8%", with a chromatogram printed underneath. Part 1 of this guide covered the COA as a whole: identity, purity, batch number and testing lab. It also recommended looking for the chromatogram and not just the percentage. This part explains what that chromatogram actually shows, how the purity figure is derived from it, and what it cannot tell you.
What the instrument is doing
High-performance liquid chromatography (HPLC) pushes a dissolved sample through a packed column. Components of the sample interact with the column packing to different degrees, so they leave it at different times. Reversed-phase HPLC is the most widely used mode for peptide separations. In it, more hydrophobic molecules are held longer, and most laboratories run it with acidic water/acetonitrile mobile phases containing trifluoroacetic acid (TFA) (Mant et al., 2007).
As each component leaves the column it passes a detector. For peptides this is usually a UV detector. Peptide bonds absorb strongly in the far ultraviolet, below 220 nm, so detection is generally set at 210–220 nm. Aromatic residues (tyrosine, phenylalanine, tryptophan) additionally absorb at 250–290 nm (Mant et al., 2007). A COA may name a single wavelength, or a photodiode-array (PDA) detector, which records many wavelengths at once.
Reading the axes
The chromatogram is the detector's signal plotted against time:
- Horizontal axis: retention time, in minutes from injection. Each component appears at its own time. The same compound run on the same method should appear at a consistent time.
- Vertical axis: detector response, usually in milli-absorbance units (mAU). A taller, wider peak means more UV-absorbing material at that time.
- Baseline: the flat line between peaks. A clean trace returns to a flat baseline on both sides of each peak.
The features that matter
- Solvent front. A disturbance in the first minute or so, from the injection itself and anything that isn't retained by the column. It is normally excluded from the purity calculation. Very small, very polar peptides can elute close to it, where separation from other unretained material is weakest. That is one reason identity confirmation by mass spectrometry matters alongside HPLC.
- Related impurities. Small peaks near the main peak. In synthetic peptides these are typically structural relatives of the target. Synthesis-related impurities include deletion sequences (a missing amino acid), insertions (an extra one), diastereomers from racemization, and incompletely removed protecting groups. Degradation products include oxidized residues, dimers and cyclization products (D'Hondt et al., 2014). Because they resemble the target, they often elute close to it.
- Main peak. The compound the COA is about. It should be labeled, or its retention time given, so you can tell which peak the percentage refers to.
- Shoulders. A bump on the side of the main peak, rather than a separate peak, can be a closely related impurity that the method did not fully resolve. How the lab splits or includes that area changes the reported purity.
- Late eluters. More hydrophobic components come off later. A run that ends shortly after the main peak cannot show anything that would have eluted after it. So check that the time axis extends well past the main peak.
How the purity percentage is calculated
The usual figure is area percent: the area under the main peak divided by the total area of all integrated peaks, times 100. A COA reading "99.8% (HPLC)" is saying that 99.8% of the UV-absorbing material the method detected, and integrated, came out as the main peak.
The same logic sets other limits on the number. Material that barely absorbs at the detection wavelength contributes little area. An impurity that co-elutes exactly under the main peak is counted as main peak. And different laboratories use different columns, gradients and integration settings, so a 99.1% from one lab and a 99.4% from another do not mean the second sample is purer. A second decimal place is a statement of arithmetic, not of accuracy.
A checklist for any chromatogram
- Labeled axes: time in minutes and a detector response, with the detection wavelength or detector type stated.
- An identified main peak: labeled, or with its retention time given.
- A percentage consistent with the trace. Clearly visible secondary peaks don't square with a near-100% figure. A lone peak on a flat baseline doesn't square with a low one.
- A run long enough to show late-eluting material after the main peak.
- A matching batch. The chromatogram belongs to the same lot number as the vial (see Part 1).
- Identity confirmed separately. One clean peak shows one dominant UV-absorbing component, not which component. Mass spectrometry answers that.
Red flags
- The same trace on different documents. An identical chromatogram, down to the baseline noise, across different batches or different compounds points to a reused image, not a new test.
- A trace with no axes, units or scale, which cannot be checked against anything.
- A purity figure with no chromatogram, and none available on request.
- A run cropped just after the main peak, which hides whatever would have eluted later.