Why two different instruments
Nearly every research peptide certificate rests on two measurements. One, from high-performance liquid chromatography (HPLC), says how much of the material is a single substance. The other, from mass spectrometry (MS), says what that substance weighs, and therefore whether it is the molecule it is supposed to be. Neither alone is sufficient: HPLC can tell you a sample is 99% one thing without telling you what the thing is, and MS can confirm a mass without telling you how much else is in the vial.
This guide explains what each technique does, assuming no background beyond school chemistry. For how the results are laid out on the document, see how to read a certificate of analysis; for the methods used on our batches, see the methods page.
Reversed-phase HPLC
What the instrument does
Chromatography separates a mixture by passing it through a material to which its components stick with differing strength. In HPLC the material is packed into a steel column a few millimetres wide and 5 to 25 cm long, a pump pushes liquid through it at high pressure, a small volume of dissolved sample is injected at the top, and a detector at the bottom records what comes out and when.
"Reversed-phase" describes the chemistry. The packing is silica particles, typically 2 to 5 µm across, coated with hydrocarbon chains eighteen carbons long, which is what "C18" means. This surface is oily. The mobile phase flowing past it starts out mostly water and becomes progressively richer in acetonitrile over the run; this change in composition is the gradient. A peptide injected at the start sticks to the C18 surface through its hydrophobic side chains. As the acetonitrile fraction rises, at some characteristic composition it lets go and travels to the detector. More hydrophobic peptides need more acetonitrile and come off later, and two molecules that differ even slightly in hydrophobicity, say a peptide and the same peptide with one oxidised methionine, reach the detector at different times.
Both mobile phases almost always contain 0.1% trifluoroacetic acid (TFA). It holds the pH low so the peptide's acidic groups behave consistently, and the trifluoroacetate anion pairs with the peptide's positive charges, which sharpens the peaks. It is also why lyophilised peptides arrive as TFA salts: the acid follows the peptide through the purification column and is not fully removed by freeze-drying.
Detection at 214 or 220 nm
The detector is a UV absorbance cell. The peptide bond, the amide link between every pair of residues, absorbs ultraviolet light strongly at around 190 to 220 nm, so detection at 214 nm (or 220 nm, slightly less sensitive but with a steadier baseline under acetonitrile gradients) sees essentially every peptide regardless of sequence, with a response roughly proportional to the number of peptide bonds. A second wavelength, 280 nm, is sometimes recorded as well; only tryptophan, tyrosine and cystine absorb there.
Retention time
The chromatogram is a plot of absorbance against time. Each substance appears as a peak, and the time at which the top of the peak passes the detector is its retention time. The same peptide on the same column with the same gradient comes out at the same time, run after run, to within a few hundredths of a minute. It is a fingerprint of the method rather than a universal constant; change the column, gradient or temperature and the number moves. That is why a certificate should state the method, and why retention times from different laboratories cannot be compared unless the methods match.
Area percentage, and what it leaves out
Purity by HPLC comes from integrating the chromatogram. Software draws a baseline, finds each peak and measures the area under it; the purity reported is the main-peak area divided by the sum of all peak areas, times 100. A result of 99.41% means that 99.41% of the UV-absorbing area at 214 nm that was integrated belongs to the main peak. That is a precise statement with three qualifications.
Co-elution. An impurity that comes off the column at the same time as the main peptide is counted as main peptide. Deletion sequences, which lack one residue out of thirty or forty, are often hydrophobically almost identical to the parent and are the usual culprits. Good laboratories run a second, orthogonal method (a different column chemistry, pH or gradient) to catch what the first one merges. A shoulder on the main peak, visible when the chromatogram is plotted with an expanded y-axis, is the tell-tale sign.
Non-absorbing impurities. Anything that does not absorb at 214 nm is invisible to the detector: water, inorganic salts, the TFA counter-ion and residual solvents. HPLC purity describes the peptide fraction of the sample, not the whole powder, which is why net peptide content is reported separately.
Response factors. Area percentage assumes every peak absorbs equally per unit mass. A truncated peptide has fewer peptide bonds and absorbs less; a peptide with an aromatic protecting group still attached absorbs more. For closely related impurities the error is small, which is why the convention is accepted in pharmacopoeial methods, but it is an approximation. Area percentage is nonetheless the right tool for the question it answers, "how much of the peptide material is the right peptide".
LC-MS and electrospray ionisation
From chromatography to mass
In LC-MS the liquid leaving an HPLC column is sprayed into a mass spectrometer. Each peak in the chromatogram can then be assigned a mass, so the main peak is confirmed as the expected molecule and minor peaks can often be identified as, for example, "parent plus 16 Da", an oxidised species.
A mass spectrometer cannot weigh neutral molecules. It manipulates ions with electric fields and measures their mass-to-charge ratio, written m/z. Getting a fragile 5,000 Da peptide out of a liquid and into the gas phase as an intact ion was for decades the obstacle to using MS on biomolecules at all. Electrospray ionisation (ESI), developed by John Fenn and colleagues in the late 1980s, solved it and earned Fenn a share of the 2002 Nobel Prize in Chemistry.
How electrospray works
The liquid is pushed through a fine needle held at several kilovolts, which pulls it into a mist of tiny charged droplets. As the solvent evaporates, each droplet shrinks until its surface charge can no longer be contained and it breaks up; repeated cycles leave the peptide molecules in the gas phase carrying protons on their basic sites (the N-terminus and the lysine, arginine and histidine side chains). The process is gentle enough that the peptide stays intact.
Multiply charged ions
Because a peptide has several basic sites, electrospray produces a family of ions from the same molecule, each with a different number of protons. A peptide of mass M carrying z protons has a mass of M plus z times the proton mass (about 1.007 Da) and a charge of z, so it appears at
m/z = (M + z × 1.007) / z
A 4,813 Da peptide therefore does not show up at m/z 4,813. It appears at around 1,605 as the 3+ ion, around 1,204 as the 4+ ion, around 964 as the 5+ ion, and so on. This is useful rather than inconvenient: most routine instruments measure accurately only up to a few thousand m/z, and multiple charging brings a large molecule into that range.
A worked example of charge-state arithmetic
Take tirzepatide, average mass 4813.45 Da.
For the 4+ ion: (4813.45 + 4 × 1.007) / 4 = 4817.48 / 4 = 1204.37.
For the 5+ ion: (4813.45 + 5 × 1.007) / 5 = 4818.49 / 5 = 963.70.
Going the other way, from a spectrum to a mass, you first need the charge of each peak. If two adjacent peaks are at m/z 1204.37 and 963.70, the charge of the higher-m/z peak is
z = (963.70 − 1.007) / (1204.37 − 963.70) = 962.69 / 240.67 = 4.00
so it is the 4+ ion, and the neutral mass is 4 × 1204.37 − 4 × 1.007 = 4813.45 Da. The next peak is then 5+ and gives the same answer. When several charge states agree on one neutral mass, the assignment is secure.
Deconvolution
Doing that arithmetic for every peak is what deconvolution software does: it examines the whole charge-state envelope, finds the single neutral mass that explains all the peaks, and reports it as the "deconvoluted" or "zero-charge" mass. A certificate may give this number, the individual m/z values, or both. The deconvoluted mass is the one to compare directly with the theoretical value on the compound page; the m/z values are the raw evidence it was derived from.
Monoisotopic versus average mass
Carbon is mostly carbon-12, but about 1.1% of it is carbon-13, and hydrogen, nitrogen and oxygen have heavy isotopes too. A molecule with 225 carbons will on average contain two or three carbon-13 atoms, so the population of molecules in a sample is a spread of masses one dalton apart.
The monoisotopic mass is the mass of the molecule made entirely from the lightest isotopes; for tirzepatide (C225H348N48O68) it is about 4810.5 Da. The average mass weights each element by its natural isotope abundance; for tirzepatide it is 4813.45 Da. At this size the two differ by about 3 Da.
Which one a certificate reports depends on the instrument. A high-resolution instrument resolves the individual isotope peaks and usually quotes the monoisotopic mass; a lower-resolution instrument sees the cluster as one hump and quotes something close to the average. A reader who does not know this can see "4810.5 observed, 4813.45 expected" and conclude the sample is 3 Da light, when it is exactly right. A 3 Da gap on a 5,000 Da peptide is almost always this, not an error.
MALDI-TOF as an alternative
Matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry is the other common way to get a peptide mass. The sample is mixed with a UV-absorbing matrix, dried on a metal plate and hit with a laser pulse; the matrix carries the peptide into the gas phase, mostly as singly charged ions, which are accelerated down a flight tube and timed to the detector. MALDI is fast, tolerant of salts, and gives simple spectra: a 4,813 Da peptide appears at m/z around 4,814 as the 1+ ion, with no deconvolution needed. It is not coupled to a chromatographic separation, so it cannot easily tie masses to HPLC peaks, and routine mass accuracy is somewhat lower than a well-calibrated electrospray system. On a certificate it serves the same purpose as LC-MS, identity confirmation, and the same tolerances apply.
Amino acid analysis for content
Neither HPLC nor MS tells you how much peptide is in the vial by mass. For that the standard method is amino acid analysis (AAA). A weighed portion of powder is hydrolysed in hot hydrochloric acid, which releases the free amino acids; these are separated and quantified against calibrated standards. Because the sequence is known, the amount of each amino acid recovered gives the amount of peptide present, independent of water, counter-ions and other non-peptide mass. The result is net peptide content, typically 70% to 90% of the gross mass for a TFA salt, and it is the figure to use when making a solution of defined concentration.
LAL for endotoxin
The limulus amoebocyte lysate test uses an extract of horseshoe-crab blood cells that clots in the presence of bacterial endotoxin. In the kinetic chromogenic format the sample is incubated with the lysate and a colour-forming substrate, and the time taken for colour to develop is read against a standard curve. The method is defined in USP general chapter 85 and European Pharmacopoeia chapter 2.6.14, both of which now admit recombinant Factor C as an animal-free equivalent. Results are given in endotoxin units per milligram (EU/mg). The test measures a specific bacterial contaminant that HPLC does not remove and that matters for cell-based assays; it is run only when commissioned, so its absence from a research certificate is normal.
What "greater than or equal to 99%" actually means
Product pages often state a specification, such as purity of 99% or above. That is a threshold the batch must meet, not a measurement. The measurement is the number on the certificate, and it should be read for what it is.
Take a reported purity of 99.41%. It means that, on the stated column with the stated gradient at the stated wavelength, 99.41% of the integrated peak area belonged to the main peak. It does not mean that 99.41% of the powder is peptide (that is net peptide content), that the remaining 0.59% has been identified, or that an orthogonal method would give the same figure. It does mean that the laboratory found no single impurity larger than 0.59% by area on that method, which for a synthetic peptide of forty residues is a good result. Two decimal places is appropriate precision for an integrated area and signals a real result; "99%" flat is a rounded claim. A result of 99.41% is not meaningfully different from 99.35% or 99.50%, since run-to-run variation in integration alone accounts for a tenth of a percent. And the number describes the batch at the time of testing; what a vial contains after months of storage depends on how it was kept.
The certificate register lists each batch's independent report so that these checks can be made against the data rather than a summary. All material is supplied for in-vitro laboratory research only and is not for human or veterinary use.
References
- Snyder LR, Kirkland JJ, Dolan JW. Introduction to Modern Liquid Chromatography. 3rd ed. Hoboken, NJ: John Wiley & Sons; 2010. ISBN 978-0-470-16754-0.
- Fenn JB, Mann M, Meng CK, Wong SF, Whitehouse CM. Electrospray ionization for mass spectrometry of large biomolecules. Science. 1989;246(4926):64–71. doi:10.1126/science.2675315.
- International Council for Harmonisation. ICH Q2(R2): Validation of Analytical Procedures. Step 4, November 2023. EMA/CHMP/ICH/82072/2006.
- D'Hondt M, Bracke N, Taevernier L, et al. Related impurities in peptide medicines. Journal of Pharmaceutical and Biomedical Analysis. 2014;101:2–30. doi:10.1016/j.jpba.2014.06.012.
- United States Pharmacopeia. General chapter <1503> Quality Attributes of Synthetic Peptide Drug Substances. USP–NF. Rockville, MD: United States Pharmacopeial Convention.
- United States Pharmacopeia. General chapter <85> Bacterial Endotoxins Test. USP–NF. Rockville, MD: United States Pharmacopeial Convention.
This monograph summarises the published literature for laboratory researchers. It is not medical advice, and nothing on this site is for human or veterinary use.