Calibration is essential for establishing the degree of reliance that can be placed on a measurement technique. Similarly it is absolutely necessary to prepare a calibration plot using a certified reference material before you decide to carry out quantitative estimations no matter how sophisticated is the analysis technique that you decide upon.
A calibration plot in simple language is a graphical representation of the change in response of the measurement device with respect to change in concentration of the analyte under determination. A clear understanding of the concepts of calibration and a visual examination of the calibration plot will help you decide whether you should proceed with the adopted analytical technique for estimations over the desired concentration range of the analyte or look for alternate options. For this there is no necessity to go into the mathematical analysis of the plot but a simple visual look should help you take your decision.
The steeper the calibration line the higher is the sensitivity of the adopted technique.
The calibration plot or graph can be manually plotted on a graph paper after taking signal readings at different known concentrations of it can be displayed on the system screen automatically using application software available on most modern-day analytical instruments. At this stage it would be a good idea to make a quick reference to the earlier articles on the topic ‘Guidelines on generation and interpretation of calibration plots‘ and ‘Understanding the linearity of a calibration plot’.
The ideal graph shown above is one in which all the measurements recorded are in direct proportion to respective changes in concentration. All the points lie in a straight line and it passes through the origin. Such a plot can be ideally used for arriving at a known concentration over a wide range of measurements. However, in real situations, the calibration plot deviates from ideal behaviour due to several reasons and may appear like the representation below:
- The straight line may intersect the response or Y axis at a point different from origin. In other words there can be a residual response even at zero concentration due to matrix interferences.
- The observed points do not necessarily lie in a straight line. This in essence makes it clear that the straight line is an approximation which is drawn manually to pass through maximum number of points or generated automatically through system software which generates the line of best- fit. It is important to bear in mind that a calibration line cannot be generated by joining the observed dots as in any quiz exercise.
- Beyond a certain concentration the plot is no longer linear and tends to flatten out. This means that inference can be drawn with a certain degree of confidence only in the linear range of the plot.
- The longer the linear portion of the plot the better is the chosen technique for a given analysis. This is also referred to as Linear Dynamic Range and it can extend to several orders of magnitude, say \(10^3\) to \(10^6\), for sensitive analysis techniques.
Calibration plots are prepared using standards and the approaches used to overcome matrix interference will be discussed in a subsequent article.