Beer–Lambert Law Calculator: A=ε·c·l with Calibration

You just pulled a cuvette out of the spectrophotometer and the display reads A = 0.74 at 280 nm. Now what? A Beer–Lambert law calculator takes that absorbance reading and, given two of the three remaining variables — molar absorptivity (ε), path length (l), and concentration (c) — solves for whichever one you are missing. In most bench situations that means solving for concentration, but the relationship works in any direction.

The mistake that burns new lab members more than anything else is a unit mismatch. You look up ε for your protein in the literature, punch in the absorbance, and get a concentration that is off by a factor of a thousand. What happened? The ε was reported in M⁻¹cm⁻¹ but you entered your concentration target in mM, or the paper used a mass-based extinction coefficient (mL·mg⁻¹cm⁻¹) and you treated it as molar. That single slip cascades through every downstream dilution.

What the calculator gives you: the unknown variable, fully unit-resolved, so you can move straight to the next step — whether that is diluting a protein stock, adjusting a dye concentration for an assay, or reporting a sample concentration in a lab notebook.

Molar absorptivity tells you how strongly a molecule absorbs light at a particular wavelength. The canonical unit is L·mol⁻¹·cm⁻¹ (equivalently M⁻¹cm⁻¹). A chromophore with ε = 50,000 at its λ_max is a very strong absorber — even micromolar solutions will give measurable absorbance. Something like NADH at 340 nm (ε ≈ 6,220) is moderate. A weak absorber might sit below 100.

Where do you find ε? For proteins, tools like ProtParam on the ExPASy server predict ε at 280 nm from the amino acid sequence (based on Trp, Tyr, and Cys content). For small molecules, the supplier datasheet or the original paper reporting the compound usually lists ε at the relevant wavelength. For nucleic acids, a common shortcut is A260 = 1 for ~50 µg/mL dsDNA or ~40 µg/mL ssRNA in a 1 cm cuvette — these are mass-based coefficients, not molar, so do not plug them into A = εcl without converting.

One thing to watch: ε values in older papers may use log base e (the Napierian molar absorption coefficient) instead of log base 10, which is the standard in spectrophotometry. If the number looks about 2.3× bigger than expected, that is probably why. Divide by 2.303 to convert to the decadic ε that the Beer–Lambert formula expects.

Standard cuvettes are 1 cm across, which is why most published ε values assume l = 1 cm. But if you are reading absorbance on a plate reader, the path length is whatever the liquid depth happens to be in the well. In a standard 96-well plate with 200 µL per well, the path is roughly 0.5–0.6 cm. In a half-area plate with 100 µL, it can drop below 0.3 cm.

If you take an A280 reading on a plate reader and divide by ε × 1 cm, you will underestimate concentration by however much the real path length is shorter than 1 cm. Many modern plate readers have a path-length correction feature — they measure absorbance at 900 nm and 977 nm (a water absorption peak) and compute the actual liquid depth. Turn that feature on, or manually correct: A_1cm = A_measured × (1 / path_cm).

Instruments like the NanoDrop use a very short path (0.1 cm for the 1 mm pedestal, 0.05 cm for the 0.5 mm pedestal). The software corrects internally, but if you export raw absorbance values, be aware that the numbers look much lower than cuvette readings for the same sample. Always know your path length before running the Beer–Lambert calculation.

When you do not know ε — or when the chromophore is part of a complex mixture where ε is not well defined — you build a standard curve. Prepare a series of known concentrations of the analyte, measure absorbance for each, and plot A versus c. If Beer–Lambert holds, the points fall on a straight line through the origin with slope = ε·l.

In practice, the line rarely passes exactly through the origin because of stray light, baseline drift, or cuvette imperfections. A least-squares fit gives you the slope and intercept, and you read unknown concentrations off that line. The key quality check is R² — it should be above 0.99 for a good curve. If it is below 0.98, look for bubbles in the cuvette, inconsistent pipetting, or standards that were not mixed properly.

The calculator does not build calibration curves for you (that is a separate workflow), but it applies the Beer–Lambert relationship for single-point readings. If you already know ε from your own standard curve, plug it in and solve for unknown concentrations directly.

Beer–Lambert is a straight line — but only over a range. Below about A = 0.1, the detector is measuring a tiny difference between reference and sample, so noise dominates and your concentration calculation bounces around. Above about A = 1.5–2.0, so little light reaches the detector that stray light (photons that bypassed the sample) becomes a significant fraction of the signal, and the reading plateaus. Some benchtop spectrophotometers cap out around A = 3; NanoDrops go higher because of the very short path.

The practical sweet spot for most instruments is A between 0.1 and 1.0. If your reading falls outside that range, dilute the sample (or concentrate it) and re-measure. Always record the dilution factor so you can back-calculate the original concentration.

There are also chemical reasons for non-linearity at high concentrations: molecular interactions (dimerization, aggregation) can shift ε, and solute–solute effects change the effective refractive index. For most routine lab work at moderate concentrations, though, the instrument limitation kicks in well before the chemistry does.

I blanked with water but my buffer absorbs at this wavelength.
Always blank with the exact solvent/buffer the sample is in. If your protein is in 50 mM Tris + 150 mM NaCl, blank with that buffer, not water. Tris absorbs in the low UV (< 230 nm), so at 280 nm it is usually fine, but additives like DTT, imidazole, or DMSO can absorb significantly and throw off your reading.

My absorbance keeps drifting upward.
Check for air bubbles on the cuvette wall, particulates settling into the light path, or temperature-dependent solubility changes. If the sample is a protein that tends to aggregate, the scattering component increases over time and the apparent absorbance creeps up.

I get a different reading every time I insert the same cuvette.
Orientation matters. Optical-quality cuvettes have two polished faces and two frosted faces. The light must pass through the polished faces. Also, fingerprints scatter light — wipe the cuvette with a lint-free tissue before every reading.

My A280/A260 ratio looks wrong for pure protein.
Pure protein typically gives A280/A260 around 1.5–2.0. If the ratio is closer to 1.0 or below, nucleic acid contamination is pulling A260 up. If it is above 2.0 with a very low A260, your concentration may just be very low and noise is dominating the A260 measurement.

The core equation and its three rearrangements:

Transmittance ties in through:

Unit check: [L·mol⁻¹·cm⁻¹] × [cm] × [mol/L] = unitless. If the units do not cancel to give a dimensionless absorbance, something is mismatched in your inputs.

Scenario: You purified a His-tagged recombinant protein (MW 35 kDa) and need to know its concentration before setting up a binding assay. You measured A280 = 0.92 in a 1 cm quartz cuvette, blanked against elution buffer. ProtParam gives ε = 43,890 M⁻¹cm⁻¹ for this protein based on its sequence.

Step 1 — Solve for molar concentration.
c = A / (ε × l) = 0.92 / (43,890 × 1) = 2.096 × 10⁻⁵ M ≈ 21.0 µM.

Step 2 — Convert to mg/mL.
mg/mL = c (mol/L) × MW (g/mol) = 2.096 × 10⁻⁵ × 35,000 = 0.73 mg/mL.

Step 3 — Sanity check.
An A280 just under 1.0 for a protein with a moderately high ε should give a concentration in the low-µM range and under 1 mg/mL. The numbers line up. If you had gotten 7.3 mg/mL or 0.073 mg/mL, you would want to double-check whether the ε was entered correctly or whether the path length was actually 1 cm.

What to do next: Dilute the stock to working concentration for your assay and record the aliquot volumes. Label the tube with the concentration, date, and protein name. Freeze at −80 °C if you are not using it immediately.