Extinction Coefficients: Reading Concentration From Absorbance

Absorbance is not a concentration. It is a ratio of light in to light out across a defined path, reported in dimensionless absorbance units. Turning it into a concentration requires a constant that says how strongly one mole per litre of that particular molecule absorbs at that particular wavelength across a 1 cm path. That constant is the molar extinction coefficient, written as epsilon and carrying units of M⁻¹ cm⁻¹.

For peptides the useful wavelengths fall into two groups. Near 280 nm, absorbance comes almost entirely from aromatic side chains, so the coefficient can be estimated by counting residues. Near 205 to 214 nm, absorbance comes from the peptide bond itself, which every sequence has, so the method works on peptides with no aromatic residues at all. The two regimes have different failure behaviour and are not interchangeable.

Path length matters as much as the coefficient and is the easier of the two to get wrong. A standard cuvette is 10.00 mm. Microvolume instruments that hold a droplet between two pedestals use path lengths from 0.05 to 1.0 mm, set by the instrument rather than by the operator, and a reading taken on the assumption of 10 mm when the instrument used 1 mm is wrong by a factor of ten.

The coefficients that do the work at 280 nm

Pace and colleagues measured molar extinction coefficients for the absorbing residues in 6.0 M guanidine hydrochloride at 280 nm and reported values that are still the standard reference set. Only three contributors matter in practice, and one of them, cystine, only counts when two cysteines are paired into a disulfide bridge.

The estimate is a sum: coefficient equals 5,500 times the count of tryptophan, plus 1,490 times the count of tyrosine, plus 125 times the count of disulfide bridges. A sequence with one tryptophan and two tyrosines is estimated at 8,480 M⁻¹ cm⁻¹. Concentration in mol L⁻¹ is then absorbance divided by the product of that coefficient and the path length in cm, and mass concentration follows by multiplying through by the molecular weight.

Because the count is of whole residues, the arithmetic depends on the sequence being what the label says it is. A deletion at a tryptophan position removes 5,500 M⁻¹ cm⁻¹ from the true coefficient of that molecule while the assumed coefficient stays where it was.

The failure mode: a sequence with nothing to see

This is the one failure worth naming, because it is silent. A peptide containing no tryptophan, no tyrosine and no disulfide bridge has an estimated coefficient at 280 nm of zero. The instrument still returns a number, typically 0.00 to 0.02 absorbance units, which is noise plus whatever scatter the solution carries. Divided by a coefficient near zero, that noise becomes an arbitrary concentration, and divided by a coefficient of exactly zero the calculation is undefined.

How it shows up: a reading that barely moves when the sample is diluted twofold, or a calculated concentration that swings by an order of magnitude between replicate loadings of the same solution. Both say the signal being divided is not the peptide. Many sequences of interest fall into this group, including peptides built entirely from aliphatic, basic and acidic residues.

Moving down to 205 and 214 nm

Below about 230 nm the amide bond absorbs strongly, and since every residue after the first contributes one amide bond, every peptide has signal there. Anthis and Clore derived a coefficient at 205 nm from sequence composition and reported it as roughly 31,000 M⁻¹ cm⁻¹ per mole of protein per 100 residues as a first approximation, refined by per-residue terms. The trade is sensitivity for interference.

Trifluoroacetate left over from purification is the interference that catches people out at the low wavelengths, and it is also why the same window is chosen for chromatographic detection with care. The choice between detection wavelengths in reverse-phase work runs on the same physics.

What a UV concentration is and is not

A UV reading answers one question: how much material absorbing at this wavelength is in this solution. It does not distinguish the intended sequence from a closely related impurity that carries the same aromatic residues. A truncated fragment that retained its single tryptophan absorbs exactly as much per mole as the full-length molecule does, so it counts as though it were product.

This is the same distinction that separates chromatographic area percent from the fraction of a vial's contents that is actually peptide, and it is worth keeping the two apart when reading any specification. Purity by area and content by mass are different quantities.

• Absorbance above roughly 1.0 AU on a 10 mm path is outside the reliable linear range of most instruments and should be diluted rather than trusted.
• Scatter from undissolved or aggregated material raises absorbance across the whole spectrum and inflates the result; a rising baseline toward 320 to 340 nm is the tell.
• The blank must be the same solvent, from the same bottle, in the same cuvette orientation, or the subtraction carries a systematic offset.
• Coefficients measured in 6.0 M guanidine hydrochloride are used routinely for peptides in water or buffer; the transfer is an approximation, not an identity.

Where our own numbers stop

We publish purity as a chromatographic figure, and uV-derived concentration or an extinction coefficient for any catalogue item is not among the specifications Aurum publishes. Two reasons, both worth stating plainly. First, the extinction coefficient at 280 nm is an estimate from residue counts rather than a measured constant for that sequence, and quoting it would lend it a precision it does not have. Second, amino acid analysis is not among the specifications Aurum publishes, so we have no measured net peptide content against which a UV figure could be checked.

The label mass on a vial is a fill target verified by balance, not a UV assay result. Anyone deriving a concentration by spectrophotometry is measuring their own solution with their own coefficient and their own path length, and the uncertainty in that number belongs to that measurement rather than to the specification sheet.

Common questions

Why is absorbance reported without units?

It is a base-10 logarithm of a ratio of light intensities, so the units cancel. The units live in the extinction coefficient and the path length instead.

Does 1 absorbance unit mean the same thing on every instrument?

Only if the path length matches. Microvolume pedestal instruments commonly use 0.05 to 1.0 mm and normalise their display to a 10 mm equivalent, which is convenient until a raw reading is taken from one and used as though it were a cuvette value.

Can phenylalanine be counted?

Not at 280 nm, where its absorbance has effectively fallen away. It contributes in the 250 to 260 nm region, which is not used for concentration because tyrosine and tryptophan dominate there and the coefficients are less stable.

Is mass spectrometry an alternative?

For identity, yes, and the two are complementary rather than competing. Mass spectrometry confirms what the molecule is; UV estimates how much of something absorbing is present.

Why do replicate readings drift upward?

Usually adsorption or slow aggregation in the cuvette, or evaporation from a small droplet on an open pedestal. A spectrum rather than a single wavelength reading will show whether scatter is rising.