Plastics Technology - Dedicated to improving Plastics Processing.
Issue link: http://pty.epubxp.com/i/105397
Calculating MDL & LOQ Calculations: Sample Replicate Result, mg StanDarD DEviatiOn: LiMit OF QuantitatiOn (LOQ): 10 1 22.9 2 23.4 3 20.6 4 21.2 MEthOD DEtECtiOn LiMit (MDL): 5 21.8 6 22.5 7 21.2 t = "Student's t value" for the number of replicates at the 99% confdence level. (Can be found in a reference table in a statistics textbook, or on the internet. The value changes with the confdence level and the number of replicates used. For this example the student's t value equals 3.143). = The sample standard deviation N = The number of values in the data set x i = An individual value of the data set x-bar = The mean of the values in the data set Using these calculations with the example data set: Mean value = 21.94 mg Standard deviation = 1.02 mg MDL = 3.2 mg LOQ = 10.2 mg Since the instrument detects mass, the detection limits are in units of mass. This means that the percent moisture that can be detected is a function of the sample size. In other words the MDL for a 10gram sample size would be 3.2 mg/10 g = 0.032%. For a 50-g sample size, the MDL would be 3.2 mg/50 g = 0.0064%. The larger the sample size, the lower the percent moisture you will be able to detect. MDL Explanation Figure MDl region 1 loQ region 2 region 3 increasing Mass lost region 1: Starting from the left at 0 mg, up to the MDL, the instrument can't reliably differentiate the amount of water lost from the measurement noise. Any result in this region is less than the instrument can detect. region 2: In this region, between the MDL and the LOQ, the amount of mass loss is greater than 0 with a 99% Some additives may evaporate out of the plastic at higher temperatures and that weight loss will be interpreted by the instrument as losing moisture. By sending out a split sample you'll be able to measure and compare the results from your moisture balance to the amount of actual water to make sure the material isn't undergoing some other process that is producing a biased result. The moisture balance is only capable of detecting weight loss, regardless of the kind of mass loss that is occurring. A true moisture analyzer is capable of detecting only water loss by utilizing the RH detector, which can discern the difference between water and other volatile vapors being released from the sample. Table 2 shows that the sample is losing roughly 0.1% of its mass that is not water. confdence level, but the actual numerical value has a large degree of uncertainty. region 3: This region is where the numerical values assigned to the mass loss have a reasonable degree of confdence. You should be able to determine whether your materials are dry or not by producing values from your instrument in Region 3. As shown in Table 2, the % RSD is much larger for the moisture analyzer due to the smaller sample size, making it more susceptible to slight changes in the moisture content of the sample. By performing 10 replicates, variations in the samples can be averaged out and a true value for the overall sample can be measured. The moisture balance required a much larger sample size for testing. Step 4: Evaluation. Once all the data is collected it needs to be evaluated. Ideally your blank results are at or near 0 mg so that there are no biases or interferences within the instrument or the method. If there are biases, they need to be identifed and eliminated to the best of your ability. This may mean moving the Continued on p. 47... Plastics technology february 2013 37