Most discussions of research-peptide documentation stop at the summary table: a purity percentage, an observed mass, a lot number. That table is a transcription. Someone read an instrument output and typed numbers into a template. The instrument output itself — the chromatogram and the mass spectrum, usually appended as images on page two — is the primary data, and it is the only part of the document that can contradict the table.
Almost nobody reads it. The traces are treated as decoration, proof that testing happened rather than a record of what testing found. But a chromatogram encodes the conditions that produced the purity figure, and a mass spectrum encodes how confidently the identity was assigned. Both are readable without any analytical chemistry background, as long as you know what the axes mean and which choices were made by an operator rather than by the sample.
The summary table is downstream of everything
A purity percentage is not measured directly. It is calculated from a chromatogram by dividing the area of the main peak by the total integrated peak area. That calculation depends on how much sample was injected, what wavelength the detector was set to, how long the run lasted, how the solvent gradient was programmed, and where the integration software was told to draw its baselines.
Change any of those and the same physical vial produces a different number. This is not misconduct — it is inherent to a relative-area method, which is why HPLC purity and mass spec answer different questions and neither answers the mass-accounting question that net peptide content covers. The trace is where you find out which version of the number you were handed.
Anatomy of a chromatogram: two axes and a lot of context
The x-axis is retention time in minutes: how long a species stayed on the column before the detector saw it. The y-axis is detector response, usually milli-absorbance units at a stated UV wavelength. Peaks are species; area under a peak is proportional to how much of that species passed the detector.
Three features are worth locating immediately:
The void volume / solvent front. Everything unretained by the column washes through together at the start of the run, producing an early disturbance. Salts, some scavengers, and injection artifacts live here. Whether that early region is integrated or excluded is an operator decision that moves the final percentage.
The baseline. In reversed-phase peptide work, mobile phase B (acetonitrile with trifluoroacetic acid) absorbs in the same low-UV region used for detection, so a gradient run typically shows a rising or drifting baseline. That drift is normal. It also creates genuine ambiguity about where a small peak ends and the baseline begins.
Peak shape. Symmetrical peaks suggest the method is working. Heavy tailing, fronting, or a flat-topped peak (detector saturation from an overloaded injection) means the areas being integrated are distorted. USP General Chapter <621> defines the system-suitability parameters analysts use to judge this — resolution between adjacent peaks, tailing factor, theoretical plates, and repeatability across replicate injections. A COA that reports system-suitability results is showing you that the method was in control on the day the sample ran.
Where the purity number is actually decided
Four operator-controlled variables sit between the sample and the percentage.
Injection load. Impurities below the detection or integration threshold are reported as absent. A deliberately dilute injection produces a cleaner-looking trace and a higher percentage. A properly loaded one reveals small peaks that were always there. Nothing about the sample changed.
Wavelength. Peptide detection at 210–220 nm works because the backbone amide bond absorbs there, so response scales roughly with peptide bonds and most peptidic impurities show up. Detection at 280 nm depends on tryptophan and tyrosine. A peptide with neither aromatic residue — BPC-157, Epithalon, KPV, Selank — is nearly invisible at 280 nm, and so are its impurities. The wavelength must be stated for the number to mean anything.
Gradient and run length. Retention time is meaningless without the gradient program that produced it. More importantly, a species more hydrophobic than the end of the gradient may never elute during the run at all. A short generic gradient produces fast, clean-looking chromatograms and is exactly the condition under which late-eluting material goes uncounted.
Integration parameters. Where baselines are dropped, whether a shoulder is split from the main peak or absorbed into it, what area threshold counts as a peak — these are software settings. Two competent analysts working from the same raw file can defensibly report figures that differ in the first decimal place, and sometimes more.
The practical consequence: a purity percentage is only comparable to another percentage generated under the same method. Ranking two vendors on 99.1% versus 98.6% without reading both methods is not a comparison.
Anatomy of a mass spectrum
Electrospray ionization does not report molecular weight directly. It reports mass-to-charge ratio (m/z) for ions carrying one or more added protons. A small peptide typically appears as [M+H]⁺ and often [M+2H]²⁺; a larger species such as IGF-1 LR3 at roughly 9 kDa appears as a whole envelope of charge states, which software deconvolutes into a single neutral mass.
Read the spectrum for three things:
Which mass was reported. Monoisotopic mass (the lightest isotopic peak) and average mass (isotope-weighted) differ by roughly 0.06% — under a decimal Dalton for a tetrapeptide, a couple of Daltons for a mid-sized peptide. A COA comparing an observed average mass against a theoretical monoisotopic mass will look like a mismatch when nothing is wrong, and can equally absorb a real discrepancy.
Adducts. Sodium (+22 relative to the protonated ion) and potassium (+38) adducts are common and mostly tell you about sample handling and matrix. Recognising them prevents misreading an adduct as an impurity — or an impurity as an adduct.
Whether the spectrum belongs to the main peak. In an LC-MS run, the spectrum is taken from a specified chromatographic peak, which links identity to the peak carrying the purity figure. A direct-infusion spectrum of the bulk sample confirms the dominant species is right but says nothing about which peak on the chromatogram it corresponds to, or what the minor peaks are.
The near-mass problem
Several of the most common peptide defects change mass by about one Dalton or not at all:
- Deamidation of asparagine or glutamine: +0.98 Da
- C-terminal free acid where an amide was intended: +0.98 Da — relevant to Oxytocin and other amidated sequences
- Loss of an N-terminal acetyl group: −42 Da, resolvable, but the theoretical mass on the COA must be the modified peptide, not the bare sequence — a live issue for Thymosin Alpha-1
- Epimerisation (an L-residue racemised to D): 0 Da. Isobaric. Invisible to any mass spectrometer.
- Disulfide scrambling and misfolding: 0 Da. Isomeric. Also invisible.
A unit-resolution instrument reporting a rounded average mass cannot distinguish a one-Dalton modification from rounding. High-resolution data can. Chirality and folding require entirely different methods — chiral amino acid analysis, or a folding-sensitive assay — which is why manufacturing route determines which tests are meaningful, and why intact mass alone is weaker evidence for short, proline-rich, or D-amino-acid-containing sequences such as SS-31. MS/MS sequence confirmation, not intact mass, is what closes that gap.
Does the trace belong to your lot?
Documentation integrity is the cheapest check available and takes about fifteen seconds.
- Does the trace carry its own lot number, acquisition date, and method identifier, or has it been cropped to just the peaks?
- Are the axes labelled with units, or stripped?
- Is the method section specific — column chemistry and dimensions, mobile phase composition, gradient table, flow rate, column temperature, detection wavelength, injection volume, run time — or is it a generic paragraph that would fit any peptide?
- Does the percentage printed on the trace match the percentage in the summary table?
- Does the same image, including its noise pattern, appear across multiple lots? Identical noise means an identical file, which means one lot was tested.
A method section that reads identically for a 4-residue peptide and a 9 kDa recombinant protein is a template, not a characterisation — the same failure pattern as a COA that lists the same three assays regardless of what the compound is and how it was made.
What no trace can tell you
Chromatograms and mass spectra are silent on counterion identity and load, residual water, endotoxin, sterility, folding, chirality, and biological potency. They are also a snapshot of the material at the time of testing, not at the time of receipt — degradation pathways run on their own schedule after the certificate is printed. Reading the raw data well tells you how much confidence the purity and identity claims deserve. It does not extend those claims to cover questions they never addressed.
FAQ
Is a chromatogram with one sharp peak proof of a single compound? No. It is evidence that under those specific conditions one species dominated the detected material. Co-eluting near-isomers, species that never eluted within the run, and impurities below the integration threshold all produce the same picture. Resolution is a property of the method, not the sample.
Why do two labs report different purity for the same vial? Because area-normalised purity is method-dependent. Different gradients, wavelengths, loads, and integration settings produce different denominators. Reproducible disagreement in the first decimal place is expected; a multi-point gap usually means the methods differ in a way one of the documents is not disclosing.
If the mass matches, is identity settled? For a straightforward sequence with no stereochemical or folding questions, a high-resolution mass match is strong evidence of composition. It is not evidence of sequence order, chirality, or disulfide connectivity. Deletion and insertion impurities have different masses and are caught; epimers and scrambled disulfides are not.
Our documentation standards are described on the Quality & Testing page, and compound-level references are collected in the Peptide Library.
This article is educational and for the laboratory research community. Trulogic Labs products are sold for laboratory and research use only and are not for human consumption.