#### Introduction to calibration

Glow Discharge Optical Emission Spectroscopy (GDOES) is a relative method and therefore requires calibration. The calibration procedure consists of measuring the emitted light for a series of samples with known elemental composition, usually CRM's. The calibration function is determined using a regression algorithm. The quality of the final analysis will strongly depend on the quality of the calibration function. The choice of an appropriate calibration function and a statistically sound regression procedure, including a sufficiently large set of CRM are therefore crucial.

The accepted calibration function in GD-OES (ignoring spectral interferences) is

where *c*_{i} is the concentration in the sample, *q*_{M} is the sputtering rate, *k*_{i}, *k'*_{i}, *k''*_{i}, and *b*_{i} are constants determined by regression, and *R*_{i} is the emission yield.

The constant bi is commonly called **B**ackground **E**quivalent **C**oncentration (BEC). It is directly linked to the background intensity, the intensity that measured for a blank-sample not containing the element under study.

The linear term in the equation is linked to the sensitivity of the spectral line, where a steep slope indicates a poor sensitivity.

The higher order terms are mainly caused by self-absorption of the emitted light in the plasma. Both terms should be positive, the slope increases in this case with increasing element concentration. Selfabsorption will reduce the sensitivity of the spectral line.