Determining the LOD and LOQ based on the comparison of standard deviation (SD) of the peak area and the slope of calibration curve.
LOD and LOQ definitions are as follows:
LOD:
The lowest analyte concentration of a sample which can still be detected by the analysis method but do not have to quantitate as an appropriate value.
LOQ:
The lowes sample concentration which can still be quantitatively detected with accuracy and an acceptable precision.
The formula for the LOD and LOQ are as follows:
Description: σ is the standard deviation and S is the slope of the calibration curve.
Value of 3 and 10 in above equation in accordance with the analysis method provision of regulation maker in your country.
Case example:
Injection of caffeine standard solution runs on the HPLC, obtained caffeine calibration data, Which LOD and LOQ concentration value depending on the value of the standard solution series that we make as follows:
Table 1. Correlation between caffeine concentration and its area.
Above calibration data will produce caffeine calibration curve as follow:
In the picture above, we have line equation: Y=66132X-3940, where the value of slope obtained from the equation is S=66132.
Then calculate the standard deviation of Peak Area as follows:
Table 2. Deviation of measurement of caffeine peak area with linier regression.
Yi is obtained from the regression equation Y=66132X-3940, for example: for X=25.075 ppm, then Yi=(66132 x 25.075) – 3940 = 1654319.90.
Standard deviation can be calculated by:
Σ(Y-Yi)^2 value is obtained from table 2.
σ= (49987352.84/3)^(1/2)
σ= 4081.97 –> (Standard Deviation)
Thus:
Gràcies!!!!
ReplyDeleteIN CALCULATING THE STANDARD DEVIATION HOW WAS IT DETERMINED TO USE N-2 INSTEAD OF N-1.
ReplyDeletePlease, write the source of this way of calculation!
ReplyDeleteThanks!
Another way of understanding the degrees of freedom is to note that we are estimating two parameters from the regression – the slope and the intercept. Therefore, ν = n − 2 and we need at least three points to perform the regression analysis.
ReplyDeletehttp://www.chem.utoronto.ca/coursenotes/analsci/stats/ErrRegr.html