Every analytical method has a point below which it cannot see the analyte. The limit of detection calculator takes the standard deviation of the blank, the calibration slope and the ICH factors, and gives the limit of detection (LOD), the limit of quantification (LOQ) and a comparison with your measured concentration.
How the calculation works
- Enter the standard deviation σ of the blank (or of the regression intercept) and the slope m of your calibration curve, in the same signal units.
- The limit of detection is LOD = k × σ ÷ m, with the factor k set to 3.3 by the ICH Q2 guideline. The limit of quantification is LOQ = kq × σ ÷ m, with kq equal to 10.
- Enter the concentration measured in your sample to see how it compares with both limits.
- A result below the LOD is reported as not detected, one between the LOD and the LOQ as present but not quantified, and one above the LOQ as a reliable number.
Example
A calibration curve has a slope of 0.085 absorbance units per mg/L, and the blank has a standard deviation of 0.012. The LOD is 3.3 × 0.012 ÷ 0.085 = 0.4659 mg/L and the LOQ is 10 × 0.012 ÷ 0.085 = 1.4118 mg/L. A sample measured at 2.1 mg/L is 4.5 times the LOD and 1.5 times the LOQ, so it can be reported.
Why it matters
Method validation in analytical chemistry, pharmaceutical quality control, environmental testing and clinical labs all require the LOD and LOQ. Reporting results below these limits without care can mislead regulators and customers.
Good to know
This signal-to-slope approach is one of several accepted ways to estimate the limits, and others use signal-to-noise ratios or visual evaluation. The values should be confirmed by measuring samples near the computed limits. The units of σ and the slope must be consistent.
Related tools
The standard error calculator gives the spread of an average, and the coefficient of variation calculator measures precision in relative terms.