Ask a purchasing manager what a peptide certificate of analysis should show and the first answer is usually a number: 98 percent, 99 percent, higher. The number is real and it matters. It also answers only one of the two questions a laboratory buyer has to settle before a vial goes on the bench. The first is how clean the material is. The second is whether it is the right molecule at all.

Two instruments, two different questions

High-performance liquid chromatography, or HPLC, pushes a dissolved sample through a packed column and records what comes out, and when. Different molecules travel at different speeds, so a mixture separates into peaks. A detector, commonly set to ultraviolet wavelengths that peptide bonds absorb, draws each peak as it passes. Purity is then a ratio: the area of the main peak divided by the area of everything detected. A reading of 98 percent means 98 percent of the detected signal belonged to the main peak.

That ratio says nothing about what the main peak is. Retention time, the moment the peak appears, reflects how a molecule interacts with the column, and several different peptides can behave almost identically on one. A vial holding the wrong sequence, or a close relative of the right one, can produce a single sharp peak and a flattering percentage. The chromatogram has told the truth about cleanliness and stayed silent on identity.

What mass spectrometry adds

Mass spectrometry measures the mass of molecules by weighing ions. For large biological molecules this became practical with soft ionization methods, and the 2002 Nobel Prize in Chemistry went in part to John Fenn and Koichi Tanaka for developing them. Electrospray ionization, Fenn’s method, gives a peptide a series of charges and lets the instrument report it as a set of peaks, which software converts into a single molecular mass.

The check is arithmetic a buyer can follow. A peptide’s sequence fixes its theoretical mass, because every amino acid contributes a known amount. If the measured mass matches the calculated one within the instrument’s stated tolerance, the vial contains something with the expected composition. If it does not, the gap is informative. A missing amino acid shows up as a lower mass by that residue’s weight. A methionine that picked up an oxygen adds about 16 daltons. Both are visible in a mass reading and easy to miss in a purity percentage alone.

Reading a spectrum without panic

Real spectra are messier than the arithmetic suggests. Peptides pick up sodium and potassium ions, so extra peaks about 22 and 38 daltons above the expected mass are common and usually harmless. Electrospray also spreads one molecule across several charge states, which is why the raw output looks like a comb of peaks before software collapses it into one mass. A reviewer should look for the deconvoluted mass, the theoretical mass and the difference between them, and be wary of a report that shows an impressive spectrum with no numbers beside it.

A mismatch is not always a bad vial. Sample handling, a mislabeled file or a transcription error can produce one, which is why the report has to tie back to the lot.

What it still cannot settle

Mass is not sequence. Two peptides built from the same amino acids in a different order have the same mass, so a matching reading cannot rule out a scrambled chain. Tandem mass spectrometry breaks the molecule into fragments and reads them to confirm the order, and amino acid analysis measures composition and how much of the powder is peptide at all. A certificate that names its method lets a buyer see which of these questions it answered.

That is the honest way to read any certificate. Each analytical method answers one question well and others poorly. The weakness lies in files that present a single number as if it settled everything.

The business case for asking

Procurement teams already apply this discipline to other inputs, and the cost of unreliable inputs has been estimated at scale. In a 2015 paper in PLOS Biology, Leonard Freedman, Iain Cockburn and Timothy Simcoe estimated that more than half of US preclinical research spending, roughly 28 billion dollars a year, went to work that could not be reproduced, and they named biological reagents and reference materials as one of the four main sources of the problem.

A wrong or degraded peptide is a small version of the same loss. Weeks of bench time go to a result that cannot be trusted, controls fail for no visible reason, and the repeat run costs the reagents again. For a contract lab, the failed batch becomes a client problem. Set against that, asking a supplier for one more page of data in advance is inexpensive insurance.

When paper and product drift apart

Peptides can change hands before they reach a bench: manufacturer, distributor, repackager, end user. At each step a certificate can stay attached to the wrong lot or describe an earlier batch. That is an administrative failure rather than a chemical one, and it is why the first check on any certificate is the boring one. Does the lot number on the page match the label on the vial? A spectrum for the right molecule in the wrong lot protects no one.

What a usable certificate looks like

A few suppliers now publish their acceptance criteria where buyers can read them before they ask. The quality standards page at usapeptidedepot.com is one example, and reading a supplier’s stated criteria first makes the questions to ask sharper.

A complete file pairs the two measurements for the same lot. A public database of lot reports shows the format in practice, with each entry tied to a lot number, a stated method (HPLC with electrospray mass spectrometry) and a test date. Whatever the source, a buyer reviewing a certificate can run through four checks.

  • The lot number on the report matches the label on the vial, and the test date falls after the lot was made.
  • The chromatogram is included with its method and detection wavelength, so the percentage can be traced to integrated peaks rather than asserted.
  • The measured mass sits beside the theoretical mass, with the instrument’s tolerance stated.
  • The testing laboratory is named, and it is clear whether it is independent of the seller.

Buyers can also ask how a supplier handles failures. A company that describes its testing process in writing, as the USA Peptide Depot about page does, gives a purchasing team something to hold it to. For work where a wrong compound is expensive, some teams also retest incoming lots independently and compare reports across lots to spot drift. That adds cost on arrival and removes an assumption.

Putting it into the purchase specification

The tidiest fix sits in the supplier agreement rather than the lab. A purchase specification can name the tests required for each lot, such as HPLC purity at a stated wavelength and a mass reading within a stated tolerance of the theoretical value, and say that shipments without them are rejected. Quality teams in other industries do this as routine. It turns an optional extra into a condition of supply and gives both sides a written standard to point to when a result is disputed.

Where this leaves a purchasing desk

A purity figure and a mass reading belong in the same file. The first says how much of the signal belongs to the main peak, and the second says what that peak weighs. The next time a certificate reaches a purchasing desk with a single percentage on it, the right reply is one sentence long: send the mass spectrum.