detection limit (LOD) should not be confused with the instrument's minimum response value. The signal-to-noise ratio can be used as an important indicator for examining the performance of the instrument, but it is not applicable to the evaluation of detection limit (LOD). detection limit (LOD) can be divided into two parts: method detection limit (MDL) and instrument detection limit (IDL).
method detection limit MDL
reliably uses a specific method to reliably identify or distinguish the analyte determination signal from a specific matrix background, that is, MDL is to use this method to determine the lowest value greater than the relevant uncertainty . All matrix interferences should be taken into account when determining MDL. Currently, most analyses use instruments, each of which has a certain detection capability for the amount of detected analytes. This detection capability is expressed by instrument detection limit (IDL) . As the sensitivity increases, the IDL decreases, the instrument noise decreases, and the IDL decreases. There are many methods to determine the detection limit, and other methods can also be accepted in addition to the methods listed below.
visual evaluation method evaluation LOD
visual evaluation can be applied to non-instrumental methods and instrument analysis methods. At the same time, it can also be applied to LOD evaluation of qualitative testing methods. LOD is then evaluated by adding a known concentration analyte to the sample blank, which can reliably detect the minimum concentration value of the analyte. Add a series of analytes of different concentrations to the sample blank. At each concentration point, about 7 independent tests are usually required. The repetitive tests of each concentration point should be performed in a random order. For qualitative analysis, the response curve corresponding to the concentration of the positive (or negative) result percentage is drawn to check whether the threshold concentration is reliable, that is, whether the detection is reliable at that concentration.
blank standard deviation method evaluation LOD
can be determined by analyzing large amounts of sample blanks or adding sample blanks of the lowest acceptable concentration. The number of independent tests should be no less than 10 times (n≥10). The standard deviation (s) of the detection result is calculated. See the table for details of the calculation method of LOD.
Method of representation of LOD in quantitative detection
Table: Calculation method of LOD
"Minimum acceptable concentration" is the minimum concentration added when the obtained uncertainty is acceptable; assuming that the sample and the blank are measured separately in actual detection, and the concentration corresponding to the blank signal is deducted for blank correction; the average value and standard deviation of the sample blank value are affected by the sample matrix, so the minimum detection limit is also affected by the type of sample matrix; therefore, if the compliance judgment is used for this condition, the precision value needs to be updated regularly with the actual detection data.
Scope of application of calibration equation evaluation LOD
If sample data in LOD or near LOD cannot be obtained, the parameters of the calibration equation can be used to evaluate the instrument's LOD. If the blank mean value is added to 3 times the standard deviation of the blank, the instrument's response to the blank is the intercept a of calibration, and the standard deviation of the instrument's response is the standard deviation of calibration (sy/x). Therefore, the equation yLOD=a+3sy/x=a+bxLOD can be used, then xLOD=3sy/x/b. This equation can be widely used in analytical chemistry . However, since this is the extrapolation method , when the concentration is close to the expected LOD, the results are not as reliable as the results obtained from the experiment. Therefore, it is recommended to analyze samples with concentrations close to LOD, and it is necessary to confirm that they can be detected in the analyte at an appropriate probability.
signal-to-noise ratio method evaluation LOD
Since the instrument analyzes process will have background noise, the commonly used method is to compare the measured signal of analyte samples with known low concentrations with the blank samples to determine the minimum concentration that can be reliably detected. Typically acceptable signal-to-noise ratios are 2:1 or 3:1.
is unreliable for qualitative methods when selectivity is below the critical concentration. This critical value will vary depending on the reagent, scaling amount, matrix, etc. in the test conditions. When determining the LOD of the qualitative method, you can add several standard liquids of different concentration levels to the blank sample, and randomly detect 10 times at each level, and record the detection results (positive or negative).Draw the curve of the positive rate (%) or negative rate (%) detected by the sample against the addition concentration, and the critical concentration is the turning point when the detection result is unreliable.
Example: When the concentration of the substance to be tested in the sample is less than 100μg/g, the positive test results are no longer 100% reliable.
Table 2: Qualitative Analysis - Determining the critical value
Summary: Detection limit There are four commonly used representations
(1) The lower limit of instrument detection
can detect the minimum signal of the instrument, usually expressed by the signal-to-noise ratio. When the ratio of signal to noise is greater than or equal to 3, the sample concentration equivalent to the signal intensity is defined as the lower limit of instrument detection.
(2) Method detection lower limit
, i.e., the lowest concentration that a certain method can detect. The lower limit is usually detected by the method that can be obtained by using the method of extrapolation of the low concentration curve.
The detection limit for method specified in the US EPASW-846 is: MDL=3.143δ (δ repeated measurements 7 times)
(3) Lower sample detection limit
i.e., the minimum content of the sample that can be detected relative to the blank. The lower limit for sample detection is defined as the concentration when its signal is equal to three times the standard deviation of the signal measured in the blank solution. The lower limit of detection is an important factor in choosing an analysis method. The lower limit of sample detection is not only related to the lower limit of method detection, but also to the blank content and fluctuations in the blank sample. Only when the blank content is zero, the lower limit of sample detection is equal to the lower limit of method detection. However, the blank content is often not equal to zero, and the blank size is affected by factors such as environmental pollution to the sample, reagent purity, water quality purity, container texture and operation. Therefore, it can be determined by the extrapolation method that the lower limit of the method detection may be very low, but due to the existence of the blank content and the fluctuation of the blank content, the lower limit of the sample detection may be much larger than the lower limit of the method detection. From the perspective of practicality, the lower limit of sample detection is more useful and practical.
(4) Quantitative Lower Limit
US EPASW-846 stipulates that 4MDL is the quantitative lower limit RQL, that is, the concentration of 4 times the detection limit is used as the measurement lower limit, and the relative standard deviation of the measurement value is about 10%. Japan JIS stipulates that the quantitative lower limit is 10 times MDL.
