Wait, What? A colorimeter can give you a precise absorbance and still leave you with the wrong concentration.

The instrument measures an optical signal. Concentration is inferred only after that signal is connected to standards of known concentration under matched conditions. The practical job is therefore not “read the machine.” It is to build and validate a calibration relationship strong enough to support an unknown.

The laboratory job is calibration

For a coloured species in a suitable concentration range, Beer-Lambert behaviour links absorbance A to concentration c and path length l:

A = εcl

The deeper spectroscopy mechanism already has its own canonical owner. Here the practical question is narrower: how do standards, blanks, wavelength choice and cuvette handling turn absorbance into defensible concentration evidence?

Standards create the ruler

Prepare a set of solutions with known concentrations that span the expected unknown. Measure their absorbances under identical conditions and plot absorbance against concentration.

The Royal Society of Chemistry emphasises exactly this sequence: make standards, blank the instrument, choose a suitable wavelength and build a satisfactory calibration plot before using it to determine an unknown. See the RSC Beer-Lambert practical resource.

A blank removes background that is not the analyte

The blank contains the solvent and relevant reagents without the coloured analyte. Zeroing with a blank means the later absorbance is interpreted relative to that background.

A wrong blank can shift every reading. This is systematic error: repeating the same wrong blank gives beautifully precise but biased results.

Why wavelength choice matters

Measurements are commonly made near a wavelength where the analyte absorbs strongly. Stronger absorbance can improve sensitivity, but the wavelength must also avoid interference from other coloured species.

If the instrument uses coloured filters rather than a scanning spectrophotometer, choose the filter that gives a useful response to the solution colour, often the complementary region of the spectrum.

Cuvettes create hidden geometry

Beer-Lambert behaviour assumes a fixed optical path length. Fingerprints, droplets, scratches, bubbles or rotating a non-equivalent cuvette face into the beam can alter transmitted light without changing concentration.

Handle cuvettes by the textured or upper surfaces, wipe optical faces clean, remove bubbles and keep orientation consistent.

The calibration range has boundaries

A straight calibration line over one range does not guarantee linearity at all concentrations. Highly concentrated solutions can absorb so strongly that little light reaches the detector, while chemical equilibria or instrumental stray light can also cause deviation.

If an unknown lies above the highest standard, do not simply extend the line indefinitely. Dilute the unknown into the validated range and multiply back by the dilution factor.

Quantitative window

Suppose standards give an approximately linear relation:

A = 0.080c + 0.005

where c is in mmol dm⁻³. An unknown gives A = 0.405.

c = (0.405 − 0.005)/0.080 = 5.00 mmol dm⁻³

If the sample had first been diluted tenfold, the original concentration would be 50.0 mmol dm⁻³.

Why forcing the line through the origin can be dangerous

Ideal Beer-Lambert behaviour predicts zero analyte concentration should give zero analyte absorbance after proper blanking. But a real calibration may show a small intercept because of blank mismatch, instrument offset or preparation error.

Do not automatically force a line through zero unless the method and data justify it. The intercept itself can be diagnostic evidence.

Replicates and residuals

Repeated measurements reveal short-term scatter. More importantly, inspect how far each standard lies from the best-fit line. A curved residual pattern may indicate non-linearity even when the graph looks approximately straight by eye.

Observation versus inference

Observation: “The unknown absorbance was 0.405 at the selected wavelength.”

Inference: “Its concentration is about 5.00 mmol dm⁻³ in the measured dilution, because that absorbance lies inside the validated calibration range.”

Overclaim: “The sample contains exactly 5.00 mmol dm⁻³ of one pure substance.” A colour signal may not be uniquely specific unless the chemistry establishes specificity.

Failure modes

• Using an incorrect blank.
• Fingerprints or bubbles in cuvettes.
• Standards prepared inaccurately.
• Unknown outside the calibration range.
• Changing wavelength between standards and unknown.
• Using naturally coloured or turbid samples without accounting for background.
• Treating a high R² value as proof that every assumption is valid.

Unfamiliar transfer: measuring a reaction rate

If absorbance changes as a reaction proceeds, colorimetry can become a kinetic method. The new challenge is timing: the calibration must connect absorbance to concentration, and measurements must be fast enough to resolve the reaction. Instrument response and mixing time now become part of the error model.

Secondary → JC → deeper Chemistry

Secondary: compare colour intensity consistently, use standards and recognise that darker colour can indicate more absorbing species under matched conditions.

JC: prepare calibration standards, use blanks, apply Beer-Lambert relationships, dilute unknowns into range and quantify uncertainty.

Deeper Chemistry: quantitative spectroscopy extends to wavelength scans, multicomponent analysis, limits of detection, matrix effects, standard additions and full validation.

Checkpoint

Your standards range from absorbance 0.05 to 0.80. An unknown reads 1.42. What is the strongest next step?

Answer key and WHY reasoning

Dilute the unknown so its absorbance falls inside the validated calibration range, remeasure it and apply the dilution factor. Extrapolating far beyond the standards assumes linear behaviour where you have not tested it.

How to study this practical

Practise the chain known standards → calibrated optical response → unknown signal → interpolated concentration. Then diagnose what happens if the blank is wrong, the unknown is too concentrated or the cuvette is dirty.

Evidence boundaries

A calibration curve supports concentration inference only for the measured species, wavelength, matrix and validated range. It does not automatically prove chemical identity or linearity beyond the standards.

Authoritative next steps

• Royal Society of Chemistry: Beer-Lambert practical
• Royal Society of Chemistry: spectrophotometry and calibration
• SEAB A-Level syllabus directory

Teaching Guide

For teachers and parents: give students a calibration graph with one unknown above range and another with a non-zero intercept. Ask them to decide what can be interpolated, what must be diluted, and what the intercept might diagnose. That is much stronger than simply reading a concentration from a line.