Chromatogram Interpretation

Retention time (where it elutes), peak height, peak width at half-height, and peak area together describe a peak. Purity = main peak area ÷ total integrated area, in the elution window of interest.

Front-shoulder: deletion sequence (a residue short). Rear-shoulder: oxidation product or diastereomer. Late-eluting hump: hydrophobic aggregate. Multiple small peaks before main: synthesis precursors.

≥99% main peak is our release threshold. Borderline batches (98.5–99%) are re-tested; below 98.5% is rejected. We never blend batches to hit a number.

Start at the axes: time runs left to right, signal intensity runs bottom to top. The large peak somewhere in the middle is the target sequence; its position on the time axis is the retention time, and it should sit close to where the method's reference places it. Everything to its left is more hydrophilic material eluting early — often solvent, salts and very short fragments; everything to its right is more hydrophobic material, which is where truncated and modified sequences tend to appear.

Next, look at peak shape. A healthy main peak is roughly symmetrical; fronting or tailing suggests column overload or strong interaction with the stationary phase. Then scan the baseline between peaks — it should be genuinely flat. Rising or wavy baselines mean the integration percentages carry more uncertainty than the headline number suggests.

Finish with the impurity peaks themselves: their heights relative to the main peak, and whether the report identifies any of them. A named impurity at 0.3% is more reassuring than an unnamed cluster totalling 0.3%, because identification means the lab looked, not just measured.

Three patterns justify pausing before use. A main peak with a visible shoulder — a smaller peak fused to the main one — suggests incomplete separation and means the stated purity may understate the related-impurity burden. A forest of small peaks across the whole time axis suggests a badly degraded or badly synthesized batch, whatever the main-peak percentage claims. And a retention time far from the method's reference position raises the question of whether the main peak is the named compound at all.

None of these is proof of a problem — but each is a reason to read the rest of the report closely, compare against the certificate archive, or ask the supplier's technical team for the raw data. Pausing costs minutes; an interpretive error in a protocol costs the protocol.