Analytical-Chemistry Formulae

Introduction

Analytical formulae turn a signal into a defensible sample result only when calibration and sample preparation are valid. A line equation gives concentration in the measured solution; dilution may be needed to reconstruct the original sample. Recovery and percent error describe different checks. Detection limit is a statistical and operational concept that cannot be reduced to one universal number without specifying blanks, noise, decision rule and matrix. This sheet keeps each denominator and assumption explicit.

Core explanation

A common linear calibration equation is S = mc + b , with measured signal S , slope m , analyte concentration c and intercept b . Rearranged, c = (S − b)/m . The slope's units are signal units divided by concentration units, while intercept has signal units. If the sample was diluted, multiply the measured concentration by the dilution factor V final/V aliquot to estimate the original concentration, assuming no analyte loss or reaction and appropriate volume handling. A calibration line should be validated over the signal and concentration range used. An excellent numerical fit does not correct a matrix effect if standards and sample respond differently.

For a spike recovery , add a known amount of analyte and compare the measured increase: recovery fraction = (measured spiked result − measured unspiked result)/(known added amount) on a consistent amount basis. Multiply by 100% for percent recovery. The spike should be processed at a stage relevant to the losses under investigation. A spike added immediately before instrument reading may not test extraction losses in earlier steps. Recovery need not be exactly 100% to be useful, but acceptable limits are method- and decision-dependent. It also does not prove absence of every interference.

Percent error compares a measured value with a valid reference. A signed version is 100(x meas − x ref)/x ref ; an absolute percent error uses the absolute value of the numerator. State which is reported. If the reference is uncertain, the comparison needs more than this simple formula. Relative standard deviation , often 100s/x̄ , reports repeatability and should not be mistaken for accuracy. Uncertainty can combine random and systematic contributions; extra display digits do not remove them.

A detection limit addresses the low signal at which the analyte can be distinguished from background under a specified decision rule and false-positive/false-negative balance. Simple instructional formulae sometimes use a multiple of blank standard deviation divided by calibration slope, but the multiplier and procedure depend on the method and standard. A quantitation limit is usually higher because reliable numerical estimation requires acceptable precision and bias, not mere detection. A reported limit must include analyte, matrix, unit and method. If a sample result falls below quantitation limit, reporting many decimal places is not justified even if the instrument produces a number.

Step-by-step reasoning

1. Name analyte, matrix, measured solution and original sample. 2. Check blank, calibration range and units of slope and intercept. 3. Solve for concentration and apply dilution or extraction factor exactly once. 4. Compute recovery or error using a clearly labeled denominator. 5. Compare results with detection/quantitation criteria and include uncertainty.

Visual explanation

Draw a calibration graph with standards along a line, a blank near the intercept and a sample point inside the validated range. An arrow from a small original aliquot to a larger diluted flask labels the volume ratio. A separate bracket compares unspiked and spiked results, subtracting to isolate the spike response. The image shows why the calibration result and original-sample result can differ.

Real-world analogy

A ruler calibrated in centimeters cannot be read as inches without conversion, and a stretched copy may need its scale factor reversed to infer original length. That resembles calibration and dilution correction. The analogy does not cover matrix interference or detection statistics, which require chemical validation.

Real-world example

A water laboratory measures a trace metal in a salty sample. Standards in pure water produce a straight line, but salt suppresses the instrument response. Applying the pure-water line would underestimate metal concentration. Matrix matching or standard additions can diagnose this, while an independent reference material tests broader accuracy. The calibration equation is mathematically correct but chemically unsuitable until its sample-response assumption is checked.

Why?

Why subtract the intercept before dividing by slope? The intercept represents predicted signal at zero analyte concentration under the fitted model. Dividing the total signal by slope without subtracting it attributes background to analyte. The effect is especially large near the detection range, where blank signal may be a substantial fraction of the sample signal.

Common misconception

“A high calibration correlation proves accurate samples.” It does not address matrix or preparation bias. “Recovery and yield are the same.” Recovery tests analytical procedure; chemical yield compares actual product with a theoretical reaction amount. “A detected peak is a reliable numerical concentration.” Detection may occur below quantitation. “Percent error always reports a positive value.” Signed and absolute conventions differ.

Worked example

A validated calibration gives S = 0.050 + 2.00c , where c is in mmol L⁻¹. A diluted sample has S = 0.450 , so c diluted = (0.450 − 0.050)/2.00 = 0.200 mmol L⁻¹ . If 10.0 mL original sample was diluted to 50.0 mL, the dilution factor is five and original concentration is 1.00 mmol L⁻¹, assuming full analyte recovery. A separate spike adds 0.100 mmol L⁻¹ equivalent to the measured solution and raises measured concentration from 0.200 to 0.290 mmol L⁻¹. Recovery is (0.290 − 0.200)/0.100 × 100% = 90% . This suggests some bias under that spike procedure, but does not identify its cause without further controls.

Quick check

1. For S = mc+b , what is c in terms of measured signal? Answer: (S−b)/m within the calibration's valid range. 2. Does a detection limit automatically establish a quantitation limit of the same value? Answer: No. Reliable quantitation typically requires stronger precision and bias performance.

Exam focus

Label every sample stage and apply dilution factors in the correct direction. Write calibration units and subtract the intercept. State signed or absolute error convention. Use a known spike amount with the correct denominator for recovery. Distinguish detection from reliable quantitation and report uncertainty at the level supported by the full method.

Advanced insight

Calibration errors can be correlated across every sample if all standards derive from one misprepared stock solution. Blank variability may change with concentration, invalidating a constant-noise assumption. Standard additions can reduce matrix-response bias but depend on added analyte behaving like native analyte. An analytical result is traceable through sampling, preparation, calibration and interpretation; the detector readout is only one link.

Summary

Calibration maps signal to concentration under validated conditions, dilution returns to the original sample, recovery tests procedure response and percent error compares with a reference. Detection and quantitation describe different low-level claims. Units, blanks and matrix checks decide whether the formula result is chemically meaningful.

Practice questions

1. If S = 0.10 + 4.0c and S = 0.90 , find c . Answer: (0.90−0.10)/4.0 = 0.20 in the calibration's concentration unit. 2. A 5.0 mL sample diluted to 25.0 mL measures 0.40 mg L⁻¹. Estimate original concentration. Answer: 0.40 × (25.0/5.0) = 2.0 mg L⁻¹, assuming no loss. 3. A 10.0 μg spike increases measured analyte by 8.0 μg. Find percent recovery. Answer: 80%. 4. Why must a percent error calculation name the reference value? Answer: The denominator is the reference; a different or invalid reference changes the meaning of the percentage.