Building a Calibration Curve
A calibration curve is the foundation of quantitative spectrophotometric analysis. By measuring a series of standards with known concentrations and plotting their instrument response — typically absorbance — you create a mathematical model that converts any future absorbance reading into a concentration value. Without a well-constructed calibration curve, even the most precise instrument cannot deliver accurate results.
The Beer-Lambert Foundation
Over a limited concentration range, the relationship between absorbance and concentration follows the Beer-Lambert law: A = ε × l × c, where ε is the molar absorptivity, l is the path length, and c is concentration. This linear relationship is why a simple straight-line calibration works well for many applications. However, at high concentrations stray light, detector non-linearity, and molecular interactions cause deviation from linearity, making curve-fitting models essential.
Choosing and Preparing Standards
A reliable calibration curve requires at least five to seven standards spanning the expected concentration range of your unknowns. Best practice includes a blank (zero-concentration) standard to anchor the curve at the origin. Standards should be prepared from a certified reference material, using the same matrix (buffer, solvent) as the samples to minimise matrix effects. Bracketing your samples — ensuring no unknown falls outside the range of the standards — is critical; extrapolation beyond the calibration range introduces unquantified error.
Running the Curve on K LAB Instruments
K LAB UV-Vis instruments (Alpha and POP series) include a dedicated Quantitation mode. After measuring each standard the software plots the curve in real time, fits the selected model, and reports the R² correlation coefficient — values up to 0.9999 are routinely achieved with high-quality standards. For immunoassay and ELISA workflows, the MRX A2000 microplate reader extends this with sigmoidal curve-fitting options (4PL and 5PL) that handle the characteristically non-linear dose-response of antibody-based assays.
Validating and Troubleshooting Your Curve
Once fitted, validate the curve by checking: R² (should exceed 0.99 for most quantitative work), visual inspection of residuals (no systematic trend), and back-calculated concentrations of each standard (within ±15% of the nominal value, or ±20% at the LLOQ). Common problems include:
- Hook effect at high concentrations — absorbance flattens or drops; dilute high samples and add more high-concentration standards.
- Poor low-end linearity — detector noise dominates; increase path length or analyte concentration.
- Outlier standards — a single poorly prepared standard can distort the entire fit; rerun and, if still an outlier, exclude with documented justification.
A well-validated calibration curve is not just a computational step — it is the primary evidence that your measurement system is fit for purpose and that reported concentrations are traceable to known reference values.