Research guide

HPLC Purity vs Mass Spectrometry: What Each Test Proves

Aminent Labs Research·Updated July 2026·13 min read·For Research Use Only

Two numbers sit at the top of nearly every peptide certificate of analysis (COA): a purity figure from high-performance liquid chromatography (HPLC), and a mass value from mass spectrometry (MS). They are frequently read as if they say the same thing. They do not. RP-HPLC answers the question how much of what is in this vial is a single dominant species? Mass spectrometry answers a completely different question: is that species the molecule it is supposed to be? One is a measure of homogeneity; the other is a measure of correct mass. A material can score well on one and fail the other, which is exactly why a defensible COA reports both. This article walks through what each technique physically measures, how to read the output at a working level, and where each one quietly misleads. Everything here is framed for in-vitro analytical and laboratory characterization of research-use-only (RUO) reference materials.

The core distinction: purity is not identity

Purity and identity are orthogonal properties. Purity is a statement about the proportion of your target relative to everything else that the detector sees. Identity is a statement about the chemical nature of the main component. A synthesis can deliver a beautifully homogeneous single peak that happens to be the wrong sequence, a deletion analog, or a diastereomer; conversely, the correct molecule can be present but contaminated with related impurities. RP-HPLC is built to resolve and quantify the first situation. Mass spectrometry is built to confirm or refute the second.

The practical consequence is that neither result is interpretable alone. A reported "99% by HPLC" with no mass confirmation tells you the sample is homogeneous but says nothing about what the homogeneous thing is. A clean mass match with no chromatographic purity tells you the target is present but says nothing about how much co-purified junk rides along with it. The U.S. Pharmacopeia's work on synthetic-peptide reference standards makes this layered approach explicit, pairing chromatographic purity measurement with orthogonal mass-based confirmation rather than treating either as sufficient on its own.

What RP-HPLC actually measures

Reversed-phase HPLC separates molecules by hydrophobicity. The sample is pushed through a column packed with a nonpolar stationary phase (typically C18-derivatized silica) while the mobile phase is ramped from mostly aqueous to increasingly organic. A common setup uses solvent A of 0.1% trifluoroacetic acid (TFA) in water and solvent B of 0.1% TFA in acetonitrile, run as a gradient. More hydrophobic species partition into the stationary phase and elute later; more hydrophilic species elute earlier. Because retention depends on subtle surface chemistry, RP columns can resolve peptides that differ by as little as a single amino acid substitution.

The detector is almost always a UV absorbance detector set low in the ultraviolet, around 214-220 nm. That wavelength is chosen deliberately: it is where the peptide (amide) bond backbone absorbs, so every peptide species responds regardless of whether it contains aromatic residues. A validated RP-HPLC assay for the decapeptide cetrorelix acetate, for example, uses a C18 column with a TFA/acetonitrile gradient, which is representative of the general approach (that particular method reads absorbance higher in the UV, at 275 nm, exploiting the compound's aromatic residues; backbone-wavelength detection near 214-220 nm is the more generic choice when uniform peptide response is wanted).

Reading a chromatogram

The output is a chromatogram: absorbance (y-axis) versus retention time (x-axis). Each resolved species appears as a peak. Purity is reported by area normalization: the software integrates the area under each peak, and purity is the main peak's area expressed as a percentage of the total integrated area of all peaks.

So "98.5% purity by HPLC" means the main peak accounts for 98.5% of the total UV-absorbing peak area detected under those conditions. When you read a chromatogram, look for a few things: a single, sharp, symmetric main peak; good baseline separation (peaks returning to baseline between species) rather than shoulders or unresolved humps; and a flat, low baseline elsewhere. Small early- or late-eluting peaks are the impurities being counted against the main peak. Peak shape matters too: fronting or tailing, or a peak that is broad relative to its neighbors, can hide co-eluting species that inflate the apparent purity.

What HPLC purity does not tell you

The area-normalization number rests on an assumption that is convenient but not strictly true: that every impurity absorbs UV light the same way the main compound does. In reality each species has its own molar absorptivity (its own response factor) at the detection wavelength. If an impurity absorbs more weakly than the target, its true mass fraction is underestimated; if more strongly, overestimated. Rigorous impurity quantitation requires relative response factors, but early-stage area-percent reporting typically assumes they are equal. A single-wavelength area-percent chromatogram is a fast, useful comparison of homogeneity, but it lacks specificity for the absolute amount of any given substance.

Two further blind spots matter for peptides. First, UV detection at 214-220 nm only sees things that absorb there. Counterions (the TFA or acetate salt), residual water, inorganic salts, and other non-chromophore contaminants are invisible to the detector and are therefore not counted in the "purity" figure. A peptide can be "99% pure" by peak area while a substantial fraction of the vial's dry mass is salt and water — which is why net peptide content is a separate assay from chromatographic purity. Second, and most importantly, HPLC cannot tell you the identity of any peak. A deletion analog missing one residue, an epimerized diastereomer, or an entirely different sequence with similar hydrophobicity can co-elute with, or masquerade as, the target. Deletion peptides and short fragments are known to co-elute under strong TFA ion-pairing conditions. The chromatogram will happily report a single clean peak and a high purity number for the wrong molecule.

What mass spectrometry actually measures

Mass spectrometry measures mass-to-charge ratio (m/z). It does not separate by hydrophobicity or count relative amounts by UV area; it ionizes the molecules, sorts the resulting ions by m/z, and reports where they land. From that you reconstruct the molecular mass and compare it to the theoretical mass calculated from the intended sequence. If observed and theoretical masses agree within the instrument's accuracy, you have strong evidence the molecule is the intended one. This is the identity check HPLC cannot provide.

Two soft-ionization methods dominate peptide work: electrospray ionization (ESI) and matrix-assisted laser desorption/ionization (MALDI). Both are gentle enough to lift intact peptides into the gas phase without shredding them.

Monoisotopic vs average mass

Before reading any spectrum you need to know which "mass" you are comparing against, because a peptide does not have one mass — it has an isotopic distribution. Carbon is ~98.9% ¹²C and ~1.1% ¹³C, so any molecule with many carbons shows a cluster of peaks one dalton apart. The monoisotopic mass is computed using the lightest, most abundant isotope of each element with its exact mass (¹H = 1.007825, ¹²C = 12.000000, ¹⁴N = 14.003074, ¹⁶O = 15.994915). The average mass uses the abundance-weighted atomic weights (C = 12.011, and so on). For a small peptide below ~1500 Da the monoisotopic peak is the tallest in the cluster, and the two masses differ by only about 1 Da; the gap widens with size. Angiotensin I, for instance, has a monoisotopic mass of 1295.6775 Da (formula C₆₂H₈₉N₁₇O₁₄). A frequent misreading is comparing an observed monoisotopic value against a theoretical average value (or vice versa) and declaring a mismatch that is really just a definition error.

Reading an ESI spectrum

ESI produces multiply charged ions. A peptide typically shows up as a family of signals, each carrying a different number of protons — commonly (M+2H)²⁺ through (M+5H)⁵⁺, landing between roughly m/z 300 and 1500. The number of charges tracks the number of basic residues plus the N-terminus. Because each peak of the series carries different charge, the raw ESI spectrum looks like a "ladder," and the true molecular mass is not read directly off any single peak — it is reconstructed (deconvoluted) from the series.

The reconstruction is straightforward arithmetic. Any two adjacent charge states are enough. If two neighboring peaks appear at m/z values x₁ and x₂ (with x₁ > x₂), the charge z of the x₁ peak is z = (x₂ − m_H) / (x₁ − x₂), and the mass follows as M = z·x₁ − z·m_H, where m_H is the mass of a proton. As a worked example from the reference literature, apomyoglobin gives adjacent peaks at x₁ = 998.2 and x₂ = 942.8; the arithmetic yields z = 17, and the deconvoluted mass of 16,951 Da can be recovered to better than 50 ppm. In practice the instrument software does this for you, but knowing the logic is what lets you sanity-check a deconvoluted result. A related trick: when the isotopes of a single charge state are resolved, the spacing between adjacent isotope peaks equals 1/z — so a spacing of 0.3 m/z units means a triply charged ion. That spacing alone assigns the charge state and, from there, the monoisotopic mass.

Reading a MALDI spectrum

MALDI is the simpler spectrum to read because it predominantly generates singly charged ions. For peptides in positive-ion mode, the dominant signal is almost always (M+H)⁺ — the protonated molecule — so the observed m/z is essentially the mass plus one proton, and you don't have to deconvolute a charge ladder. That single-charge behavior makes MALDI especially convenient for mixtures, since each component gives basically one peak. The trade-offs: you will often see less intense sodium and potassium adducts ([M+Na]⁺, [M+K]⁺) sitting to the high-mass side of the main peak, and sample-spot heterogeneity ("hot spots") can affect reproducibility. With delayed-extraction reflectron instruments, mass resolution above 10,000 and accuracy in the 10-50 ppm range are achievable for peptides up to several thousand daltons.

How the identity call is made

The confirmation is a comparison: observed mass versus theoretical mass, expressed in parts per million (ppm) of error. Modern peptide-range analyzers routinely reach the low-ppm range, and that accuracy is what makes the identity assignment meaningful. High accuracy is not academic — it can distinguish amino acid substitutions that are nearly isobaric. Lysine and glutamine differ by only 0.0364 Da, which at mass 1000 demands about 36 ppm accuracy to tell apart; accurate-mass MALDI has resolved a Gln-for-Lys swap in a peptide where the two forms differed by just 0.0036 Da, with experimental values within about 1.5 ppm of theory. That is identity confirmation at a level HPLC simply cannot reach.

Why you genuinely need both

The two techniques fail in opposite directions, which is precisely why they are complementary rather than redundant.

HPLC catches what MS can miss on quantity. A standard MS experiment is not a reliable measure of relative abundance — ionization efficiency differs between species, so a large impurity may ionize poorly and look small, or vice versa. HPLC-UV, integrating a backbone chromophore, gives a far more even-handed count of how much impurity is present.
MS catches what HPLC misses on identity and on co-elution. When impurities co-elute with the target under a given gradient, the chromatogram shows one peak and reports a falsely high purity. Coupling the separation to high-resolution MS resolves those hidden species by mass even when chromatography cannot separate them in time. In one peptide-drug study, bivalirudin and two related peptides co-eluted as a single chromatographic band, yet LC-HRMS extracted-ion analysis distinguished all of them; the same approach detected impurities present at less than 0.1% of the main component and uncovered ten additional related impurities beyond what conventional methods found. Sequence changes from deletion, insertion, substitution, truncation, oxidation, or deamidation "typically result in detectable mass shifts" — the exact events that leave HPLC purity looking fine.

Put concretely: HPLC tells you the vial contains one dominant species and roughly how dominant it is. MS tells you that species is the intended molecule and flags mass-shifted relatives hiding under the main peak. Take away either and you have a real blind spot. This is why a serious COA presents an RP-HPLC chromatogram and a mass spectrum, not one standing in for the other — and why mass confirmation should never be reported as a "purity."

Common misreadings to avoid

Treating the mass match as a purity result. A correct observed mass confirms the main component's identity. It says little about how much impurity co-exists, because minor species may ionize poorly or fall outside the acquisition window.
Treating HPLC area-percent as absolute quantity. Area normalization assumes equal UV response factors across species. Without relative response factors, the number is a homogeneity comparison, not an exact weight fraction.
Forgetting the invisible fraction. Salt counterions, water, and non-UV-absorbing contaminants are not counted in HPLC purity. Net peptide content is a distinct measurement.
Mixing up monoisotopic and average mass. Compare like with like; a "mismatch" is often just the wrong reference value.
Reading a raw ESI ladder as multiple compounds. A series of evenly progressing charge states is usually one molecule at several charges, not several molecules. Deconvolute before concluding.
Ignoring adducts in MALDI. Sodium and potassium adducts to the high-mass side are expected and should not be mistaken for a modified species or an impurity.
Trusting a single peak. A lone symmetric HPLC peak can still be a co-eluting mixture. Orthogonal confirmation by MS is what tests that assumption.

FAQ

Does a high HPLC purity number prove the peptide is correct? No. HPLC purity is a measure of homogeneity — how much of the UV-absorbing material is one dominant species. It cannot confirm that the species is the intended sequence. Only mass spectrometry (or orthogonal structural methods) establishes identity.

Does a matching mass prove the peptide is pure? No. A mass match confirms the main component's identity but is a poor measure of relative abundance, since different species ionize with different efficiencies. Purity is best assessed chromatographically.

Why is UV detection usually set to ~214-220 nm? Because the peptide (amide) bond backbone absorbs strongly in that region, so all peptides respond regardless of aromatic content. The cost is that non-peptide, non-absorbing contaminants such as salts and water are invisible to the detector.

What is the difference between monoisotopic and average mass? Monoisotopic mass uses the lightest, most abundant isotope of each element (exact masses); average mass uses abundance-weighted atomic weights across the isotopic distribution. For small peptides they differ by roughly 1 Da, widening with molecular size. Always compare observed against the correct theoretical value.

ESI or MALDI — which is better for identity confirmation? Both confirm identity by comparing observed to theoretical mass. ESI produces multiply charged ions requiring deconvolution and couples naturally to LC; MALDI gives mostly singly charged ions that are simpler to read and tolerant of some impurities. They are complementary, and each reaches low-ppm accuracy for peptides.

Can two different peptides co-elute as one HPLC peak? Yes. Species with similar hydrophobicity — including deletion and truncation analogs — can co-elute, especially under strong ion-pairing conditions, producing a single peak and an inflated purity figure. High-resolution MS resolves these by mass even when chromatography cannot.

References

Related research

How to Read a Peptide COA Where the HPLC and mass-spec results sit on a certificate, section by section. Peptide Impurities Explained The species behind the extra peaks and the mass shifts these two methods resolve. Net Peptide Content vs. Purity The invisible fraction — salt and water — that chromatographic purity never counts. GLP-3 RT (Retatrutide) — Research Overview A triple-agonist peptide whose identity confirmation depends on both tests. Peptide Nomenclature and Sequence Notation Why a stated sequence and a measured mass only match once the notation is parsed correctly.

For Research Use Only. All Aminent Labs materials are intended solely for in-vitro research and laboratory use — not for human or veterinary use. See our GLP-3 RT page, browse the catalog, or view certificates of analysis.