UV quantitation of peptides: what A280 measures, and what it quietly assumes

Ask how much peptide is in a solution and the fastest answer comes from a UV spectrophotometer. Load a few microliters, read the absorbance at 280 nm, divide by an extinction coefficient, and a concentration appears in seconds. The method is non-destructive, requires no calibration curve in the usual sense, and is embedded in virtually every protein and peptide workflow — which is exactly why its assumptions are so rarely examined. A280 quantitation is not a primary measurement of peptide mass. It is a measurement of aromatic chromophore content, converted to concentration through a coefficient that is usually calculated rather than determined. For some sequences that conversion is accurate to a few percent. For others it is unusable. This post works through where the number comes from, which peptides it serves well, and how UV-derived concentrations should be read against orthogonal methods like amino acid analysis.

The physics: Beer–Lambert and its terms

UV quantitation rests on the Beer–Lambert law: absorbance A equals the molar extinction coefficient ε times the path length l times the molar concentration c. Path length is fixed by the instrument — 1 cm in a standard cuvette, or a known short path in microvolume instruments. Absorbance is what the detector reports. Everything therefore turns on ε, the proportionality constant that describes how strongly one mole of the analyte absorbs light at the chosen wavelength.

At 280 nm, peptide absorbance is dominated by three side chains. Tryptophan contributes the most, with a molar extinction coefficient around 5,500 M⁻¹cm⁻¹ in the commonly used Pace parameters. Tyrosine contributes roughly 1,490 M⁻¹cm⁻¹. Cystine — the disulfide-bonded form of cysteine, not the free thiol — adds a small term near 125 M⁻¹cm⁻¹. Phenylalanine absorbs so weakly at 280 nm that it is conventionally ignored. The standard practice is to sum these contributions across the sequence: count the tryptophans, tyrosines, and disulfides, multiply by the residue coefficients, and use the total as the peptide’s ε₂₈₀.

The immediately visible weakness is that this is a calculated value, derived from model-compound studies on folded proteins in defined buffers. It carries the assumption that residue contributions are additive and environment-independent. In practice, the local environment of a tryptophan — buried versus solvent-exposed, near a quenching group or not — shifts its absorptivity by several percent. For folded proteins the Pace coefficients were empirically tuned to minimize this error, with a reported average deviation under 4 percent. Short peptides, which lack persistent tertiary structure, generally sit closer to the fully solvent-exposed limit, and calculated coefficients tend to perform reasonably — but “reasonably” here means low single-digit percent error under favorable conditions, not exactness.

The sequence problem: peptides without chromophores

The more fundamental limitation is categorical. A peptide with no tryptophan and no tyrosine has essentially no A280 signal. This is not an edge case in research peptide chemistry — it describes a meaningful fraction of commonly studied sequences. GHK-Cu contains glycine, histidine, and lysine: no 280 nm chromophore at all. Many short fragment peptides and collagen-derived sequences are similarly blind at 280 nm. For these, an A280 reading reports buffer background and scattering, and any concentration derived from it is noise.

The usual fallback is absorbance at 205–214 nm, where the peptide bond itself absorbs. Peptide-bond absorbance has the advantage of being present in every sequence and roughly proportional to chain length, with per-bond coefficients near 2,780 M⁻¹cm⁻¹ at 205 nm in common estimation schemes. The disadvantages are severe in practice: almost everything else absorbs there too. Common buffer components, residual TFA from purification, dissolved oxygen, and many excipients contribute background at low wavelengths, and instrument stray-light performance degrades. Far-UV quantitation can work in clean, defined systems, but it demands matched blanks and careful technique in a way that A280 does not.

Between these extremes sit sequences with a single tyrosine and no tryptophan — a common situation for mid-size research peptides. With ε₂₈₀ around 1,490 M⁻¹cm⁻¹, absorbance is measurable but small, so the relative contribution of scattering, background drift, and pipetting error grows. A peptide read at 0.05 absorbance units is being quantified in a regime where a 0.005 AU baseline offset is a 10 percent concentration error.

What the vial actually contains: the gravimetric gap

A second category of error has nothing to do with spectroscopy. UV quantitation, when it works, reports the molar concentration of dissolved peptide. It is frequently used in reverse — to check how much peptide a vial contained after reconstituting a nominal quantity — and here it collides with the distinction between gross mass and peptide content.

Lyophilized peptide is never 100 percent peptide by weight. Counterions (TFA or acetate), residual moisture, and any excipients all contribute to the gravimetric mass. Peptide content — the fraction of the powder that is actually peptide — commonly runs 70 to 90 percent for research-grade material, a topic covered in depth in the earlier post on peptide content versus chromatographic purity. A UV measurement that returns 85 percent of the “expected” concentration is therefore not necessarily evidence of degradation or a short-filled vial; it may simply be measuring the difference between net peptide and gross powder. Conversely, agreement with the nominal value can be coincidental if the extinction coefficient is biased in the compensating direction. Interpreting a UV concentration against a label claim requires knowing which basis the label uses — and certificates of analysis differ on exactly this point.

Matrix effects, scattering, and the shape of the spectrum

Reading a single wavelength discards most of what the spectrophotometer measures, and the discarded information is often diagnostic. A full scan from roughly 240 to 340 nm costs seconds and reveals the failure modes that a single A280 number hides.

Light scattering from aggregates produces a sloping baseline that rises toward shorter wavelengths and, critically, extends beyond 320 nm where neither aromatic residues nor peptide bonds absorb. Absorbance at 320–340 nm that is meaningfully above zero indicates particulates or soluble aggregates inflating the A280 reading. Simple corrections extrapolate the scattering contribution from the 320–340 nm region back to 280 nm, but a strongly scattering sample is better treated as a solubility or aggregation problem than a quantitation problem.

The 260/280 ratio, familiar from nucleic acid work, has a peptide-side use as well: oxidation products of tryptophan such as kynurenine and N-formylkynurenine absorb at wavelengths where intact tryptophan does not, distorting the spectrum’s shape. A tryptophan-containing peptide whose spectrum has grown a shoulder above 300 nm has a photochemistry or oxidation history worth investigating — a thread that connects to the earlier posts on oxidative degradation and photodegradation.

Buffer contributions matter more than commonly assumed. Imidazole, some reducing agents, and nucleotide contaminants absorb strongly in the 260–280 nm window. The blank must be the actual reconstitution matrix, not water; benzyl alcohol in bacteriostatic diluent, for instance, has its own aromatic absorbance in the 250–270 nm region that a water blank does not subtract.

Where UV sits in the hierarchy of content methods

It is worth being explicit about what each concentration method actually anchors to. Amino acid analysis ties concentration to certified amino acid standards through complete hydrolysis — it is the absolute reference, slow and destructive. Gravimetry with content correction ties it to a balance plus a set of ancillary assays. UV ties it to a calculated coefficient and a clean spectrum — fastest, non-destructive, and entirely dependent on the quality of those two inputs.

The methods are complementary rather than competitive. A defensible workflow for a laboratory that relies on UV routinely is to qualify the extinction coefficient once against AAA for each sequence of interest: measure the same stock by both methods, compute the empirical ε that reconciles them, and use that empirical value thereafter. This converts UV from a calculated estimate into a transferred calibration, retaining its speed while inheriting AAA’s accuracy. Published studies comparing calculated and experimentally determined coefficients suggest the correction is usually small for tryptophan-containing sequences and occasionally substantial for tyrosine-only ones — which is precisely the population where the check matters most.

Read this way, an A280 value on a certificate of analysis or a bench worksheet is neither authoritative nor worthless. It is a rapid, chromophore-dependent estimate whose error budget is knowable: coefficient uncertainty of a few percent in good cases, categorical failure in chromophore-free sequences, and matrix and scattering artifacts that a full spectrum exposes. The measurement rewards the same habit that serves everywhere else in peptide analytics — asking not just what number the instrument produced, but what physical quantity it was actually measuring when it produced it.