HPLC vs Mass Spectrometry: What Each Test Actually Proves

Analytical characterization of a synthetic peptide rests on a pair of methods that are frequently discussed as if they were interchangeable. They are not. They answer different questions, and neither answer substitutes for the other.

High performance liquid chromatography separates a mixture into its components and reports how much of each is present relative to the others. It is a counting method. It tells you that the sample contains one dominant species and some number of minor ones, and it quantifies that ratio. What it cannot do is name any of them.

Mass spectrometry ionizes molecules and measures their mass to charge ratio. It is an identification method. It tells you the molecular mass of what is present, which for a peptide of known sequence is a strong identity check. What it does not do well, at least not without careful method development, is tell you the relative quantity of each species, because ionization efficiency varies from molecule to molecule.

The workhorse configuration for peptide analysis is reverse-phase chromatography on a C18 column. Porous silica particles, commonly three to five micrometers in diameter with 100 or 300 angstrom pores, are surface-bonded with octadecyl chains to create a nonpolar stationary phase. A larger pore size is generally preferred for peptides because it allows the molecule access to the internal surface area of the particle.

The mobile phase is a gradient. Solvent A is typically water with about 0.1 percent trifluoroacetic acid, and solvent B is acetonitrile with a similar acid concentration. The run begins at low organic content and increases the proportion of acetonitrile over time. Peptides adsorb to the C18 surface at low organic strength and release when the mobile phase becomes sufficiently nonpolar to compete. More hydrophobic species require more acetonitrile and therefore elute later.

Trifluoroacetic acid does two jobs. It keeps the mobile phase acidic, which protonates basic residues and suppresses silanol interactions that would cause peak tailing. It also acts as an ion-pairing agent, forming ion pairs with positively charged residues and increasing their effective hydrophobicity, which sharpens peaks considerably. The tradeoff is that TFA suppresses ionization in mass spectrometry, which is why LC-MS methods often substitute formic acid and accept slightly worse peak shape.

Detection is by ultraviolet absorbance at 214 nm, where the amide bond absorbs. Some methods add 280 nm for tryptophan and tyrosine as a secondary channel, but 214 nm is the quantitative one because every residue contributes rather than only the aromatic ones.

A clean chromatogram proves homogeneity under one specific set of separation conditions. It demonstrates that when this sample was pushed through this column with this gradient, the great majority of the UV-absorbing material moved as a single band. That is genuinely informative, and it is the basis of every purity percentage you will see on a COA.

The first limitation is co-elution. Two species of similar hydrophobicity travel together and appear as one peak, and this is not an edge case for peptides, because the impurities that matter most are structurally similar to the target. A deamidated variant differs by one dalton. A diastereomer differs only in the stereochemistry of one center. A methionine-oxidized form is slightly more polar and often elutes just before the main peak as an unresolved shoulder. Under a shallow gradient these separate. Under a steep production-style gradient they may not.

The second limitation is invisibility. Chromatographic purity only counts what the detector sees and what the column releases. Trifluoroacetate counterions, chloride, acetate, residual water, and inorganic salts contribute essentially nothing at 214 nm. Highly hydrophobic aggregates may not elute at all within the gradient window and simply never appear. Both cases inflate the apparent purity.

The third limitation is definitional. HPLC cannot identify anything. If a supplier synthesized an entirely different peptide, purified it well, and ran it on a C18 column, the chromatogram would be excellent. Purity and identity are independent properties, and the chromatogram addresses only the former.

Mass spectrometry requires the analyte to be an ion in the gas phase. For peptides, two soft ionization techniques dominate because they transfer the molecule intact rather than fragmenting it.

Electrospray ionization sprays the sample from a fine charged capillary, producing droplets that evaporate until the peptide carries multiple protons. The resulting spectrum shows a charge state envelope, a family of peaks corresponding to the same molecule at 2+, 3+, 4+ and higher charge. Deconvolution software collapses that envelope into a single neutral mass. Because the charge state series is internally consistent, the presence of a coherent envelope is itself a quality signal, and its absence is a reason to look harder.

Matrix-assisted laser desorption ionization with time-of-flight detection, or MALDI-TOF, co-crystallizes the sample with a UV-absorbing matrix and fires a laser at the spot. It produces mainly singly charged ions, so the observed peak sits roughly one dalton above the neutral mass. MALDI tolerates salts better than electrospray and handles larger peptides comfortably, at the cost of lower mass accuracy on standard instruments.

The comparison that matters is observed mass against theoretical mass calculated from the stated sequence. Get the basis right: average mass against average mass, monoisotopic against monoisotopic. Then check the tolerance. Single quadrupole ESI on a routine instrument is typically good to about one dalton on a two kilodalton peptide. Time-of-flight and Orbitrap instruments reporting monoisotopic mass should agree to a few parts per million.

Intact mass measurement is a strong falsification tool. If the observed mass does not match the theoretical mass, the material is not what the label says, and no further discussion is required. That alone justifies the test.

What it cannot do is confirm sequence order. Mass is a sum over composition. Any rearrangement of the same residues produces the same mass, so a scrambled sequence is invisible to intact MS. Leucine and isoleucine have identical elemental composition and identical mass, so substitutions between them are invisible. Racemization from L to D at any center changes no atoms at all and is completely invisible. Glutamine and lysine differ by only 0.036 daltons, which a low resolution instrument cannot resolve.

Resolving sequence requires tandem mass spectrometry. In MS/MS a precursor ion is isolated and fragmented, usually by collision-induced dissociation, generating a ladder of b and y ions whose mass differences read out the residue sequence. That is standard practice in proteomics and is not standard on routine peptide COAs, which is worth knowing when a document claims to have confirmed sequence. Amino acid analysis confirms composition, but it also does not establish order.

Intact MS is likewise a poor quantitative tool for impurity levels. Ionization efficiency depends on the molecule, so a species present at one percent may produce a signal comparable to the main peak or nearly none at all. Reading relative peak heights in a mass spectrum as if they were relative abundances is a common and significant error.

Consider the failure modes each method leaves open. HPLC alone permits a scenario in which a well purified but entirely incorrect peptide passes as excellent material. Mass spectrometry alone permits a scenario in which the correct peptide is present but constitutes a modest fraction of a crude mixture, because a strong signal at the right mass says nothing about how much else is in the vial.

Running both closes each gap with the other. Chromatography establishes that one species dominates. Mass spectrometry establishes that the dominant species is the intended molecule. Best practice is to acquire them on the same sample preparation, and ideally on the same injection using LC-MS, so that the mass measurement is unambiguously assigned to the main chromatographic peak rather than to the sample as a whole.

This is why the standard characterization package for a research-grade peptide lot is RP-HPLC with UV detection plus an ESI or MALDI mass spectrum, and why a document offering only one of the two should be treated as partial rather than complete.

Several additional techniques appear on more thorough documentation. None are universally necessary, since a cell culture experiment has different requirements than a stability program. The point is to know which questions your documentation answered and which it left open, rather than assuming a clean chromatogram and a matching mass constitute complete characterization.